Disclosure of Invention
The embodiment of the application mainly aims to provide a vehicle abnormal sound identification method and device, a vehicle and a storage medium, and aims to improve the accuracy of vehicle abnormal sound identification.
In order to achieve the above object, a first aspect of an embodiment of the present application provides a vehicle abnormal sound identification method, where the method is applied to a cloud server, and the method includes:
Acquiring reference sound data of a vehicle;
performing abnormal sound prediction on the reference sound data through a preset abnormal sound prediction model to obtain abnormal sound prediction data;
Verifying the abnormal sound prediction data to obtain an abnormal sound verification result, wherein the abnormal sound verification result is used for representing whether the abnormal sound prediction data is error data or not;
If the abnormal sound verification result represents that the abnormal sound prediction data is error data, correcting the reference sound data to obtain target sound data, and uploading the target sound data to a preset abnormal sound database;
performing model parameter adjustment on the abnormal sound prediction model based on the abnormal sound database to obtain an abnormal sound identification model;
And carrying out abnormal sound recognition on the current sound data acquired in advance according to the abnormal sound recognition model.
In some embodiments, the abnormal sound prediction data includes sound feature data, abnormal sound occurrence time and abnormal sound type, and the verifying the abnormal sound prediction data to obtain an abnormal sound verification result includes:
If the abnormal sound type is a parameter type abnormal sound, acquiring vehicle operation parameters of the abnormal sound at the occurrence time of the abnormal sound, wherein the parameter type abnormal sound is used for representing abnormal sound generated by the vehicle when the vehicle operation parameters change;
and verifying the parameter type abnormal sound according to the sound characteristic data and the vehicle running parameters to obtain the abnormal sound verification result.
In some embodiments, the vehicle operation parameters include a first engine speed, a second engine speed, a first balance shaft frequency and a second balance shaft frequency, the abnormal sound generation time includes a first time and a second time, the first engine speed is a speed of an engine at the first time, the second engine speed is a speed of the engine at the second time, the first balance shaft frequency is a frequency of a balance shaft at the first time, the second balance shaft frequency is a frequency of the balance shaft at the second time, the verifying the parameter abnormal sound type according to the sound feature data and the vehicle operation parameters, and the obtaining the abnormal sound verification result includes:
If the parameter abnormal sound is abnormal sound of the balance shaft of the engine, acquiring a spectrogram of the sound characteristic data;
acquiring first abnormal sound energy corresponding to the first moment and the first balance axis frequency on the spectrogram, and second abnormal sound energy corresponding to the second moment and the second balance axis frequency on the spectrogram;
when the first engine speed is greater than the second engine speed and the first abnormal sound energy is smaller than the second abnormal sound energy, or the first engine speed is smaller than the second engine speed and the first abnormal sound energy is greater than the second abnormal sound energy, or the first engine speed is equal to the second engine speed and the first abnormal sound energy is not equal to the second abnormal sound energy, the abnormal sound verification result represents that the abnormal sound prediction data are error data.
In some embodiments, the vehicle operating parameters include a first gearbox rotational speed, a second gearbox rotational speed, a first gearbox frequency and a second gearbox frequency, the abnormal sound generating time includes a third time and a fourth time, the first gearbox rotational speed is a rotational speed of the gearbox at the third time, the second gearbox rotational speed is a rotational speed of the gearbox at the fourth time, the first gearbox frequency is a frequency of the gearbox at the third time, the second gearbox frequency is a frequency of the gearbox at the fourth time, the verifying the parameter type abnormal sound based on the sound characteristic data and the vehicle operating parameters to obtain the abnormal sound verification result includes:
if the parameter abnormal sound is abnormal rotation sound of the gearbox, acquiring a spectrogram of the sound characteristic data;
acquiring third abnormal sound energy corresponding to the third moment and the first gearbox frequency on the spectrogram, and fourth abnormal sound energy corresponding to the fourth moment and the second gearbox frequency on the spectrogram;
When the first gearbox rotating speed is greater than the second gearbox rotating speed and the third abnormal sound energy is smaller than the fourth abnormal sound energy, or the first gearbox rotating speed is smaller than the second gearbox rotating speed and the third abnormal sound energy is greater than the fourth abnormal sound energy, or the first gearbox rotating speed is equal to the second gearbox rotating speed and the third abnormal sound energy is not equal to the fourth abnormal sound energy, the abnormal sound verification result represents that the abnormal sound prediction data are error data.
In some embodiments, the vehicle operation parameters include a third engine speed, a fourth engine speed, a first engine frequency and a second engine frequency, the abnormal sound generating time includes a fifth time and a sixth time, the third engine speed is a speed of the engine at the fifth time, the fourth engine speed is a speed of the engine at the sixth time, the first engine frequency is a frequency of the engine at the fifth time, the second engine frequency is a frequency of the engine at the sixth time, the verifying the parameter abnormal sound type according to the sound characteristic data and the vehicle operation parameters, and the obtaining the abnormal sound verification result includes:
if the parameter abnormal sound is engine knock abnormal sound, acquiring a spectrogram of the sound characteristic data;
obtaining fifth abnormal sound energy corresponding to the fifth moment and the first engine frequency on the spectrogram, and sixth abnormal sound energy corresponding to the sixth moment and the second engine frequency on the spectrogram;
And when the third engine speed is greater than the fourth engine speed and the fifth abnormal sound energy is smaller than the sixth abnormal sound energy, or the third engine speed is smaller than the fourth engine speed and the fifth abnormal sound energy is greater than the sixth abnormal sound energy, or the third engine speed is equal to the fourth engine speed and the fifth abnormal sound energy is not equal to the sixth abnormal sound energy, the abnormal sound verification result represents that the abnormal sound prediction data are error data.
In some embodiments, the vehicle operation parameters include a shift signal, the abnormal sound generating time includes a seventh time, shift signal information indicates whether the shift signal exists at the seventh time, the verifying the parameter abnormal sound type according to the sound feature data and the vehicle operation parameters, and obtaining the abnormal sound verification result includes:
the parameter type abnormal sound is a gear shifting impact abnormal sound;
and obtaining the abnormal sound verification result according to the gear shifting signal information, wherein if the gear shifting signal does not exist at the seventh moment, the abnormal sound verification result represents that the abnormal sound prediction data are error data.
In some embodiments, the vehicle operation parameters include an accelerator pedal signal, the abnormal sound generating time includes an eighth time, accelerator pedal signal information indicates whether the accelerator pedal signal exists at the eighth time, the verifying the parameter type abnormal sound according to the sound feature data and the vehicle operation parameters, and obtaining the abnormal sound verification result includes:
the parameter abnormal sound is the whistle abnormal sound of the supercharger;
and obtaining the abnormal sound verification result according to the accelerator pedal signal information, wherein if the accelerator pedal signal does not exist at the eighth moment, the abnormal sound verification result represents that the abnormal sound prediction data is error data.
To achieve the above object, a second aspect of the embodiments of the present application provides a vehicle abnormal sound recognition device, including:
the acquisition module is used for acquiring the reference sound data of the vehicle;
The first identification module is used for carrying out abnormal sound prediction on the reference sound data through a preset abnormal sound prediction model to obtain abnormal sound prediction data;
the abnormal sound verification module is used for verifying the abnormal sound prediction data to obtain an abnormal sound verification result, wherein the abnormal sound verification result is used for representing whether the abnormal sound prediction data is error data or not;
The correction module is used for correcting the reference sound data to obtain target sound data if the abnormal sound verification result represents that the abnormal sound prediction data are error data, and uploading the target sound data to a preset abnormal sound database;
The updating module is used for carrying out model parameter adjustment on the abnormal sound prediction model based on the abnormal sound database to obtain an abnormal sound identification model;
And the second recognition module is used for carrying out abnormal sound recognition on the current sound data acquired in advance according to the abnormal sound recognition model.
To achieve the above object, a third aspect of the embodiments of the present application proposes a vehicle including a memory storing a computer program and a processor implementing the method according to the first aspect when the processor executes the computer program.
To achieve the above object, a fourth aspect of the embodiments of the present application proposes a computer-readable storage medium storing a computer program which, when executed by a processor, implements the method of the first aspect.
The abnormal sound identification method and device for the vehicle, the vehicle and the storage medium are characterized in that abnormal sound prediction data are obtained by acquiring reference sound data of the vehicle and carrying out abnormal sound prediction on the reference sound data through an abnormal sound prediction model. And verifying the abnormal sound prediction data to obtain an abnormal sound verification result. And if the abnormal sound verification result represents that the abnormal sound prediction data is error data, correcting the reference sound data to obtain target sound data, and uploading the target sound data to a preset abnormal sound database. And carrying out model parameter adjustment on the abnormal sound prediction model based on the abnormal sound database to obtain an abnormal sound identification model, and carrying out abnormal sound identification on the current sound data acquired in advance according to the abnormal sound identification model. Therefore, the abnormal sound prediction model can be continuously collected to predict incorrect sound data, and the abnormal sound recognition model with higher accuracy is obtained by training the abnormal sound prediction model in a targeted manner through the incorrect sound data. When the current sound data is identified according to the abnormal sound identification model, the accuracy of abnormal sound identification can be improved.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
It should be noted that although functional block division is performed in a device diagram and a logic sequence is shown in a flowchart, in some cases, the steps shown or described may be performed in a different order than the block division in the device, or in the flowchart. The terms first, second and the like in the description and in the claims and in the above-described figures, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the application only and is not intended to be limiting of the application.
By monitoring and identifying the vehicle sound, abnormal sound in the vehicle sound can be identified, and the abnormal sound type is judged, so that a fault source is rapidly positioned, and the fault is conveniently detected. In the related art, a pre-trained recognition model is used for recognizing the sound of the vehicle, so that the abnormal sound type is recognized. But the recognition accuracy of the recognition model is lower due to less training data.
Based on the above, the embodiment of the application provides a vehicle abnormal sound identification method and device, a vehicle and a storage medium, aiming at enabling an abnormal sound prediction model to conduct abnormal sound prediction on reference sound data in the vehicle, and verifying abnormal sound prediction data after outputting the abnormal sound prediction data so as to obtain an abnormal sound verification result. And when the abnormal sound verification result represents that the abnormal sound prediction data is error data, the abnormal sound prediction model is indicated to output an error prediction result. At this time. And marking the correct abnormal sound type for the reference sound data to obtain target sound data, and uploading the target sound data to an abnormal sound database. Training the abnormal sound prediction model based on the abnormal sound database again to obtain an abnormal sound identification model. And through continuously collecting training data, the model is iteratively updated, and the recognition accuracy of the abnormal sound recognition model is improved. And then abnormal sound recognition is carried out on the current sound data according to the abnormal sound recognition model, so that the accuracy of abnormal sound recognition is improved.
The method and device for identifying abnormal sound of a vehicle, the vehicle and the storage medium provided by the embodiment of the application are specifically described through the following embodiments, and the method for identifying abnormal sound of a vehicle in the embodiment of the application is described first.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Wherein artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) is the theory, method, technique, and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend, and expand human intelligence, sense the environment, acquire knowledge, and use knowledge to obtain optimal results.
Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.
The vehicle abnormal sound identification method provided by the embodiment of the application can be applied to a terminal, a server side and software running in the terminal or the server side. In some embodiments, the terminal may be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc., the server may be configured as an independent physical server, may be configured as a server cluster or a distributed system formed by a plurality of physical servers, and may be configured as a cloud server for providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs, and basic cloud computing services such as big data and artificial intelligent platforms, and the software may be an application for implementing a vehicle abnormal sound recognition method, but is not limited to the above form.
The application is operational with numerous general purpose or special purpose computer system environments or configurations. Such as a personal computer, a server computer, a hand-held or portable device, a tablet device, a multiprocessor system, a microprocessor-based system, a network PC, a minicomputer, a mainframe computer, a distributed computing environment that includes any of the above systems or devices, and the like. The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The application may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
Referring to fig. 1, fig. 1 is an optional flowchart of a vehicle abnormal sound identification method according to an embodiment of the present application, where the vehicle abnormal sound identification method is applied to a cloud server, and the method in fig. 1 may include, but is not limited to, steps S101 to S106.
Step S101, acquiring reference sound data of a vehicle;
step S102, abnormal sound prediction is carried out on reference sound data through a preset abnormal sound prediction model, and abnormal sound prediction data are obtained;
Step S103, verifying abnormal sound prediction data to obtain an abnormal sound verification result, wherein the abnormal sound verification result is used for representing whether the abnormal sound prediction data is error data or not;
Step S104, if the abnormal sound verification result represents that the abnormal sound prediction data is error data, correcting the reference sound data to obtain target sound data, and uploading the target sound data to a preset abnormal sound database;
Step S105, carrying out model parameter adjustment on the abnormal sound prediction model based on the abnormal sound database to obtain an abnormal sound identification model;
and S106, carrying out abnormal sound identification on the current sound data acquired in advance according to the abnormal sound identification model.
In step S101 of some embodiments, the reference sound data refers to sound data collected by a sound collection system on a vehicle. Different sound collection systems can be selected to collect reference sound data according to different vehicle type configurations. For example, for a vehicle with intelligent cabin functions, reference sound data is acquired through the cabin microphone without additional equipment. For vehicles without intelligent cabin functions, microphones and audio data collectors are additionally installed to collect reference sound data.
In practical application, the abnormal sound identification model is deployed on a cloud server, and after a sound acquisition system on a vehicle acquires reference sound data, the intelligent cabin uploads the reference sound data to the cloud server through the Internet of vehicles for the vehicle with the intelligent cabin function. For vehicles without intelligent cabin functions, the vehicle-mounted terminal uploads reference sound data to the cloud server.
In step S102 of some embodiments, since the collected sound data contains other sounds that do not need to be identified, such as multimedia sounds, speaking sounds. Therefore, before abnormal sound recognition is performed on the reference sound data, voice processing is further required on the reference sound data, and the processing process comprises audio filtering, noise reduction, audio slicing, pre-emphasis, framing, windowing, fourier transformation, power spectrum calculation, mel filter bank filtering, logarithm taking, discrete cosine transformation and MFCC feature extraction, so that sound feature data is finally obtained. And carrying out abnormal sound prediction on the sound characteristic data by the abnormal sound prediction model to obtain abnormal sound prediction data.
Specifically, an abnormal sound database is pre-built on the cloud server, and the abnormal sound data and the corresponding abnormal sound types are stored in the abnormal sound database. The abnormal sound data are data after voice processing, and the abnormal sound type comprises abnormal sound of engine knocking sound, abnormal sound of metal knocking sound and abnormal sound of whistle of the supercharger, and the transmission case rotates. And dividing the abnormal sound data into a training set and a verification set, training the neural network model through the training set, and verifying through the verification set until the model progress reaches the target, so as to obtain a trained abnormal sound prediction model.
In step S103 of some embodiments, the sound feature data is input to the abnormal sound prediction model to perform abnormal sound prediction, and the abnormal sound prediction result is output. The abnormal sound prediction result comprises an abnormal sound type and abnormal sound generation time, and the abnormal sound prediction data comprises sound characteristic data and a corresponding abnormal sound prediction result. After-sales personnel can judge whether the abnormal sound type of the sound corresponding to the sound characteristic data at the abnormal sound generation moment is consistent with the abnormal sound type output by the model, so that an abnormal sound verification result is obtained. And uploading the abnormal sound verification result to the cloud server. Or the cloud server verifies the abnormal sound prediction data according to a preset parameter judgment rule to obtain an abnormal sound verification result. When the abnormal sound type judged by the after-sales personnel is consistent with the abnormal sound type output by the model, the abnormal sound type output by the abnormal sound prediction model is correct. Otherwise, the abnormal sound type output by the abnormal sound prediction model is indicated to be wrong, namely the abnormal sound prediction data is wrong data. The after-sales personnel can conduct the test by repeatedly replacing the parts, so that the correct abnormal sound type is obtained, or the after-sales personnel experienced the after-sales personnel can judge the abnormal sound type through the subjective sense of the human ear.
In step S104 of some embodiments, the abnormal sound verification result includes the reference sound data and the correct abnormal sound type, when the abnormal sound verification result characterizes that the abnormal sound prediction data is error data, the correct abnormal sound type is marked for the reference sound data, so as to obtain target sound data, and the target sound data is uploaded to the abnormal sound database.
In step S105 of some embodiments, after uploading the target sound data to the abnormal sound database, the abnormal sound prediction model is subjected to model parameter adjustment again based on the sound data in the abnormal sound database, so as to obtain an abnormal sound identification model.
In practical application, after uploading the preset number of target sound data to the abnormal sound database, the abnormal sound prediction model is subjected to model parameter adjustment based on the sound data in the abnormal sound database. Preventing too frequent training results in wasted computing resources.
In step S106 of some embodiments, the current sound data is also sound data collected by a sound collection system on the vehicle. Abnormal sound recognition is carried out on the current sound data through the abnormal sound recognition model, abnormal sound recognition data are obtained, and the output abnormal sound recognition result can be more accurate.
In one example, abnormal sound prediction data is continuously verified, so that sound data with incorrect prediction of the abnormal sound prediction model is collected. And training the abnormal sound prediction model in a targeted manner through the sound data with the misprediction to obtain an abnormal sound recognition model, so that the accuracy of model recognition can be greatly improved. When the verification times reach the preset times or the proportion of the abnormal sound prediction data to the correct data reaches the preset proportion, the abnormal sound type can be predicted correctly by the abnormal sound recognition model, and verification is not needed. In another example, when the number of times of verification reaches a preset number of times, or the proportion of the abnormal sound prediction data to the correct data reaches a preset proportion, the abnormal sound prediction data may be verified every preset period. If the abnormal sound verification result represents that the abnormal sound prediction data is error data, the accumulated verification times are restarted, or the statistics of the proportion of the abnormal sound prediction data to correct data is restarted. And verifying the abnormal sound prediction data each time until the verification times reach the preset times or the proportion of the abnormal sound prediction data which is correct data reaches the preset proportion, and then starting to verify the abnormal sound prediction data every preset period.
In the steps S101 to S106 shown in the embodiment of the application, the abnormal sound prediction data is obtained by acquiring the reference sound data of the vehicle and carrying out abnormal sound prediction on the reference sound data through an abnormal sound prediction model. And verifying the abnormal sound prediction data to obtain an abnormal sound verification result. And if the abnormal sound verification result represents that the abnormal sound prediction data is error data, correcting the reference sound data to obtain target sound data, and uploading the target sound data to a preset abnormal sound database. And carrying out model parameter adjustment on the abnormal sound prediction model based on the abnormal sound database to obtain an abnormal sound identification model, and carrying out abnormal sound identification on the current sound data acquired in advance according to the abnormal sound identification model. Therefore, the abnormal sound prediction model can be continuously collected to predict incorrect sound data, and the abnormal sound recognition model with higher accuracy is obtained by training the abnormal sound prediction model in a targeted manner through the incorrect sound data. When the current sound data is identified according to the abnormal sound identification model, the accuracy of abnormal sound identification can be improved.
Referring to fig. 2, in some embodiments, the abnormal sound prediction data includes sound feature data, abnormal sound occurrence time and abnormal sound type, and step S103 may include, but is not limited to, steps S201 to S202:
Step S201, if the abnormal sound type is a parameter type abnormal sound, acquiring vehicle operation parameters at the occurrence time of abnormal sound, wherein the parameter type abnormal sound is used for representing abnormal sound generated when the vehicle operation parameters change;
Step S202, verifying the parameter abnormal sound type according to the sound characteristic data and the vehicle operation parameters to obtain an abnormal sound verification result.
In step S201 of some embodiments, the parameter abnormal sound type is abnormal sound generated when the vehicle operation parameters change, including abnormal sound of an engine, abnormal sound of a gearbox, abnormal sound of a steering gear, etc., and the parameter abnormal sound type is generally characterized in that the magnitude of abnormal sound changes along with the change of the vehicle operation parameters, and when an operation is performed, abnormal sound is generated or amplified, such as stepping on a throttle, steering wheel, etc. Therefore, if the abnormal sound type predicted by the abnormal sound prediction model is the parameter abnormal sound type, whether the abnormal sound type predicted by the model is correct or not can be verified through the sound characteristic data and the vehicle operation parameters at the abnormal sound generation moment.
In an example, the non-parametric alien type is the rest of alien types other than the parametric alien type, such as assembly problems resulting in a knocked alien, etc. If the abnormal sound type predicted by the abnormal sound prediction model is the abnormal sound type without parameters, the abnormal sound type predicted by the model can be verified by after-sales personnel continuously.
In step S202 of some embodiments, a corresponding parameter determination rule may be obtained according to a parameter class abnormal sound type. In an example, the parameter determination rule is set in advance according to the feature corresponding to the parameter abnormal sound type, for example, if the magnitude of the abnormal sound corresponding to the parameter abnormal sound type changes with the change of the vehicle operation parameter, the corresponding parameter determination rule is that at different moments when the abnormal sound occurs, the change trend of the value of the vehicle operation parameter is the same as the change trend of the magnitude of the abnormal sound, that is, when the value of the vehicle operation parameter becomes larger, the abnormal sound also becomes larger, which indicates that the parameter abnormal sound type predicted by the abnormal sound prediction model is correct. If the parameter abnormal sound is an abnormal sound generated when a certain operation is performed, for example, the gear shift impact abnormal sound is an abnormal sound generated when a gear shift operation is performed. And if the corresponding parameter judgment rule is that the vehicle running parameter representation at the abnormal sound generation moment is executing the corresponding operation, the abnormal sound prediction model is correctly predicted by the abnormal sound prediction model.
In steps S201 to S202 illustrated in the present embodiment, when the abnormal sound type is a parameter type abnormal sound, the vehicle operation parameters at the occurrence time of the abnormal sound are obtained. Therefore, the parameter abnormal sound type can be verified according to the sound characteristic data and the vehicle running parameters, and whether the parameter abnormal sound type predicted by the abnormal sound prediction model is correct or not is judged.
Referring to fig. 3, in some embodiments, the vehicle operating parameters include a first engine speed, a second engine speed, a first balance shaft frequency, and a second balance shaft frequency, the abnormal sound generation time includes a first time and a second time, the first engine speed is a speed of the engine at the first time, the second engine speed is a speed of the engine at the second time, the first balance shaft frequency is a frequency of the balance shaft at the first time, the second balance shaft frequency is a frequency of the balance shaft at the second time, and step S202 may include, but is not limited to including steps S301 to S303:
step S301, if the parameter abnormal sound is abnormal sound of an engine balance shaft, acquiring a spectrogram of sound characteristic data;
step S302, obtaining first abnormal sound energy corresponding to a first moment and a first balance axis frequency on a spectrogram, and obtaining second abnormal sound energy corresponding to a second moment and a second balance axis frequency on the spectrogram;
In step S303, when the first engine speed is greater than the second engine speed and the first abnormal sound energy is less than the second abnormal sound energy, or the first engine speed is less than the second engine speed and the first abnormal sound energy is greater than the second abnormal sound energy, or the first engine speed is equal to the second engine speed and the first abnormal sound energy is not equal to the second abnormal sound energy, the abnormal sound verification result represents that the abnormal sound prediction data is error data.
In step S301 of some embodiments, engine balance shaft abnormal sound is a common abnormal sound of a vehicle, and the magnitude of the abnormal sound generally varies with the variation of the engine speed, so that the abnormal sound is periodically generated. Therefore, a plurality of abnormal sound generation timings exist in the sound characteristic data. The first time and the second time are different abnormal sound generation times, respectively. In order to obtain the rotation speed of the engine and the frequency of the balance shaft when abnormal sound occurs, the frequency of the balance shaft and the rotation speed of the engine at the first moment are obtained, and the first balance shaft frequency and the first engine rotation speed are obtained. And obtaining the frequency of the balance shaft at the second moment and the rotating speed of the engine to obtain the frequency of the second balance shaft and the rotating speed of the second engine.
In one example, the frequency of the balance axis can be calculated by the following formulas (1) - (2):
Nb=Na×t, (1)
fb=Nb/60, (2)
where Nb represents the rotation speed of the balance shaft, na represents the rotation speed of the engine, t represents the gear ratio of the balance shaft, and fb represents the frequency of the balance shaft.
Specifically, when the parameter type abnormal sound is abnormal sound of the balance shaft of the engine, the corresponding abnormal sound is changed along with the change of the rotating speed of the engine. Therefore, by acquiring the spectrogram of the sound feature data, the change of the abnormal sound size is judged. Wherein the fourier analysis display pattern of the speech signal is called a spectrogram. A spectrogram is a three-dimensional spectrum that represents a graph of the speech spectrum over time. The abscissa of the spectrogram is time, the ordinate is frequency, and the coordinate point value represents the energy (in db) of the voice data. In this embodiment, the coordinate point value represents abnormal sound energy, and the larger the abnormal sound energy is, the larger the abnormal sound representing sound is.
In step S302 of some embodiments, in order to obtain the magnitude of the sound generated by the abnormal sound of the balance shaft of the engine at the abnormal sound occurrence time, the abnormal sound energy corresponding to the moment on the spectrogram with the abscissa being the first time and the ordinate being the first balance shaft frequency is obtained, so as to obtain the first abnormal sound energy. And acquiring abnormal sound energy corresponding to the second moment on the spectrogram with the abscissa being the second balance shaft frequency and the ordinate being the second balance shaft frequency, and obtaining the second abnormal sound energy.
In step S303 of some embodiments, the first engine speed is denoted as Na1, the second engine speed is denoted as Na2, the first abnormal sound energy is denoted as P1, the first abnormal sound energy is denoted as P2, and when Na1> Na2 and P1< P2, or Na1< Na2 and P1> P2, or Na 1=na2 and p1+.p2, it indicates that the variation trend of the abnormal sound is different from the variation trend of the engine speed, and it indicates that the abnormal sound type is not the abnormal sound of the engine balance shaft, that is, the abnormal sound verification result indicates that the abnormal sound prediction data is error data.
In an example, if Na1 = Na2 and P1 = P2 exist, the rotation speed of the engine at the time of generating the abnormal sound in the next period may be obtained until the rotation speeds of the engine at different times of generating the abnormal sound are different, and then the abnormal sound verification result is obtained in the above manner. Since there may be an error in the calculated abnormal sound energy, when the difference between the first abnormal sound energy P1 and the second abnormal sound energy P2 is within a certain range, p1=p2 may be considered.
In the steps S301 to S303 illustrated in this embodiment, it can be accurately determined whether the abnormal noise predicted by the abnormal noise prediction model is a prediction error of the abnormal noise of the balance shaft of the engine.
Referring to fig. 4, in some embodiments, the vehicle operating parameters include a first gearbox rotational speed, a second gearbox rotational speed, a first gearbox frequency, and a second gearbox frequency, the abnormal sound generation time includes a third time and a fourth time, the first gearbox rotational speed is a rotational speed of the gearbox at the third time, the second gearbox rotational speed is a rotational speed of the gearbox at the fourth time, the first gearbox frequency is a frequency of the gearbox at the third time, the second gearbox frequency is a frequency of the gearbox at the fourth time, and step S202 may include, but is not limited to including steps S401 through S403:
step S401, if the parameter abnormal sound is abnormal sound of the rotation of the gearbox, acquiring a spectrogram of sound characteristic data;
step S402, obtaining third abnormal sound energy corresponding to a third moment and a first gearbox frequency on a spectrogram, and fourth abnormal sound energy corresponding to a fourth moment and a second gearbox frequency on the spectrogram;
Step S403, when the first gearbox rotational speed is greater than the second gearbox rotational speed and the third abnormal sound energy is less than the fourth abnormal sound energy, or the first gearbox rotational speed is less than the second gearbox rotational speed and the third abnormal sound energy is greater than the fourth abnormal sound energy, or the first gearbox rotational speed is equal to the second gearbox rotational speed and the third abnormal sound energy is not equal to the fourth abnormal sound energy, the abnormal sound verification result represents that the abnormal sound prediction data is error data.
In step S401 of some embodiments, the abnormal noise generated by the rotation of the gearbox is a common abnormal noise of the vehicle, and the magnitude of the abnormal noise generally varies with the variation of the rotation speed of the gearbox, so that the abnormal noise is generated periodically. Therefore, a plurality of abnormal sound generation timings exist in the sound characteristic data. The third moment and the fourth moment are different abnormal sound generating moments respectively, and in order to obtain the frequency and the rotating speed of the gearbox when the abnormal sound is generated, the frequency and the rotating speed of the gearbox at the third moment are obtained, and the frequency and the rotating speed of the first gearbox are obtained. And acquiring the frequency and the rotating speed of the gearbox at the fourth moment to obtain the frequency and the rotating speed of the second gearbox.
In one example, the frequency of the gearbox may be calculated by the following equation (3):
fc=Nc/60, (3)
Where Nc denotes the rotational speed of the gearbox and fc denotes the frequency of the gearbox.
Specifically, when the parameter type abnormal sound is abnormal sound of rotation of the gearbox, the corresponding abnormal sound is changed along with the change of the rotating speed of the gearbox. Therefore, by acquiring the spectrogram of the sound feature data, the change of the abnormal sound size is judged.
In step S402 of some embodiments, in order to obtain the magnitude of the sound generated by the abnormal sound generated by the rotation of the gearbox at the occurrence time of the abnormal sound, the abnormal sound energy corresponding to the moment on the spectrogram with the abscissa being the third moment and the ordinate being the first gearbox frequency is obtained, so as to obtain the third abnormal sound energy. And acquiring abnormal sound energy corresponding to the fourth moment on the spectrogram with the abscissa being the frequency of the second gearbox and the ordinate being the frequency of the second gearbox, and obtaining fourth abnormal sound energy.
In step S403 of some embodiments, the first gearbox rotational speed is denoted as Nc1, the second gearbox rotational speed is denoted as Nc2, the third abnormal sound energy is denoted as P3, the fourth abnormal sound energy is denoted as P4, and when Nc1> Nc2 and P3< P4, or Nc1< Nc2 and P3> P4, or Nc 1=nc 2 and p3 not equal to P4, it is indicated that the trend of change of the abnormal sound level is different from the trend of change of the gearbox rotational speed, which is indicative of that the abnormal sound type is not the gearbox rotational abnormal sound, that is, the abnormal sound verification result indicates that the abnormal sound prediction data is erroneous data.
In an example, if nc1=n2 and p3=p4 exist, the rotation speed of the gearbox at the occurrence time of the abnormal sound of the next period can be obtained until the rotation speeds of the gearboxes at different moments of the occurrence of the abnormal sound are different, and then the abnormal sound verification result is obtained through the above method. Since there may be an error in the calculated abnormal sound energy, when the difference between the third abnormal sound energy P3 and the fourth abnormal sound energy P4 is within a certain range, p3=p4 may be considered.
In the steps S401 to S403 illustrated in this embodiment, it can be accurately determined whether the abnormal noise of the parameter predicted by the abnormal noise prediction model is a rotation abnormal noise of the gearbox.
Referring to fig. 5, in some embodiments, the vehicle operating parameters include a third engine speed, a fourth engine speed, a first engine frequency, and a second engine frequency, the abnormal sound generating time includes a fifth time and a sixth time, the third engine speed is a speed of the engine at the fifth time, the fourth engine speed is a speed of the engine at the sixth time, the first engine frequency is a frequency of the engine at the fifth time, the second engine frequency is a frequency of the engine at the sixth time, and step S202 includes, but is not limited to including steps S501 to S503:
step S501, if the parameter abnormal sound is engine knock abnormal sound, a spectrogram of sound characteristic data is obtained;
step S502, obtaining fifth abnormal sound energy corresponding to a fifth moment and a first engine frequency on a spectrogram, and sixth abnormal sound energy corresponding to a sixth moment and a second engine frequency on the spectrogram;
In step S503, when the third engine speed is greater than the fourth engine speed and the fifth abnormal sound energy is less than the sixth abnormal sound energy, or the third engine speed is less than the fourth engine speed and the fifth abnormal sound energy is greater than the sixth abnormal sound energy, or the third engine speed is equal to the fourth engine speed and the fifth abnormal sound energy is not equal to the sixth abnormal sound energy, the abnormal sound verification result characterizes the abnormal sound prediction data as error data.
In step S501 of some embodiments, engine knock noise is a noise that is common to vehicles, and the magnitude of the noise generally varies with the variation of the engine speed, resulting in periodic generation of noise. Therefore, a plurality of abnormal sound generation timings exist in the sound characteristic data. The fifth time and the sixth time are respectively different abnormal sound generation time, and in order to obtain the frequency and the rotating speed of the engine when the abnormal sound is generated, the frequency and the rotating speed of the engine at the fifth time are obtained, and the first engine frequency and the third engine rotating speed are obtained. And obtaining the frequency and the rotating speed of the engine at the sixth moment to obtain the second engine frequency and the fourth engine rotating speed.
In one example, the frequency of the engine may be calculated by the following equation (4):
fa=Na/60, (4)
Where Na denotes the rotational speed of the engine and fa denotes the frequency of the engine.
Specifically, when the parameter type abnormal sound is abnormal sound of engine knock, the corresponding abnormal sound is changed along with the change of the engine rotating speed. Therefore, by acquiring the spectrogram of the sound feature data, the change of the abnormal sound size is judged.
In step S502 of some embodiments, in order to obtain the magnitude of the sound generated by the engine knock abnormal sound at the abnormal sound occurrence time, the abnormal sound energy corresponding to the moment on the spectrogram with the abscissa being the fifth moment and the ordinate being the first engine frequency is obtained, so as to obtain the fifth abnormal sound energy. And obtaining abnormal sound energy corresponding to the sixth moment on the spectrogram with the abscissa being the second engine frequency and the ordinate being the second engine frequency, and obtaining the sixth abnormal sound energy.
In step S503 of some embodiments, the third engine speed is denoted as Na3, the fourth engine speed is denoted as Na4, the fifth abnormal sound energy is denoted as P5, and the sixth abnormal sound energy is denoted as P6, when Na3> Na4 and P5< P6, or Na3< Na4 and P5> P6, or Na 3=na 4 and p5+.p6, it indicates that the variation trend of the abnormal sound size is different from the variation trend of the engine speed, and it indicates that the abnormal sound type is not the engine knock abnormal sound, that is, the abnormal sound verification result indicates that the abnormal sound prediction data is erroneous data.
In an example, if Na3 = Na4 and p5 = P6 exist, the rotation speed of the engine at the time of generating the abnormal sound in the next period may be obtained until the rotation speeds of the engine at different times of generating the abnormal sound are different, and then the abnormal sound verification result is obtained in the above manner. Since there may be an error in the calculated abnormal sound energy, when the difference between the fifth abnormal sound energy P5 and the sixth abnormal sound energy P6 is within a certain range, p5=p6 may be considered.
In the steps S501 to S503 illustrated in the present embodiment, it can be accurately determined whether the abnormal noise predicted by the abnormal noise prediction model is a misprediction of engine knock abnormal noise.
Referring to fig. 6, in some embodiments, where the vehicle operating parameters include a shift signal and the abnormal sound generation time includes a seventh time, the shift signal information characterizes whether the shift signal is present at the seventh time, step S202 may include, but is not limited to, steps S601 to S602:
step S601, the parameter abnormal sound is gear shifting impact abnormal sound;
And step S602, obtaining an abnormal sound verification result according to the gear shift signal information, wherein if the gear shift signal does not exist at the seventh moment, the abnormal sound verification result represents that the abnormal sound prediction data is error data.
In steps S601 to S602 illustrated in the present embodiment, the shift shock abnormal sound is a kind of abnormal sound common to vehicles, and is generally generated at the time of shifting. If the abnormal gear shifting impact sound occurs at the seventh moment and a gear shifting signal exists at the seventh moment, the abnormal gear shifting impact sound is correctly predicted by the abnormal gear shifting prediction model. If the abnormal gear shifting impact sound occurs at the seventh moment and no gear shifting signal exists at the seventh moment, the abnormal gear shifting impact sound is wrong when the parameter type abnormal sound predicted by the abnormal sound prediction model is the abnormal gear shifting impact sound, namely, the abnormal sound verification result represents that abnormal sound prediction data are wrong data. According to the embodiment, whether the abnormal noise predicted by the abnormal noise prediction model is the abnormal noise of the gear shifting impact or not can be accurately judged.
Referring to fig. 7, in some embodiments, the vehicle operating parameters include an accelerator pedal signal, the abnormal sound generation time includes an eighth time, and the accelerator pedal signal information characterizes whether the accelerator pedal signal is present at the eighth time, and step S202 may include, but is not limited to, steps S701 to S702:
step S701, the parameter abnormal sound is the whistle abnormal sound of the supercharger;
And step S702, obtaining an abnormal sound verification result according to the accelerator pedal signal information, wherein if the accelerator pedal signal does not exist at the eighth moment, the abnormal sound verification result represents that abnormal sound prediction data are error data.
In steps S701 to S702 illustrated in the present embodiment, the supercharger whistle abnormal sound is a kind of abnormal sound common to vehicles, and the abnormal sound generally generates a sharp whistle when the throttle is released quickly. If the abnormal sound of the whistle of the supercharger occurs at the eighth moment, and the accelerator pedal signal exists at the eighth moment, the abnormal sound prediction model predicts the parameter type abnormal sound to be the abnormal sound of the whistle of the supercharger and is correct. If the abnormal sound of the whistle of the supercharger occurs at the eighth moment, and the accelerator pedal signal does not exist at the eighth moment, the abnormal sound prediction model predicts that the parameter type abnormal sound is wrong, namely the abnormal sound verification result represents that abnormal sound prediction data are wrong data. According to the embodiment, whether the abnormal noise predicted by the abnormal noise prediction model is wrong or not can be accurately judged.
Referring to fig. 8, an embodiment of the present application further provides a device for identifying abnormal sound of a vehicle, which can implement the method for identifying abnormal sound of a vehicle, where the device includes:
an acquisition module 801 for acquiring reference sound data of a vehicle;
The first identifying module 802 is configured to perform abnormal sound prediction on the reference sound data through a preset abnormal sound prediction model, so as to obtain abnormal sound prediction data;
the verification module 803 is configured to verify the abnormal sound prediction data to obtain an abnormal sound verification result, where the abnormal sound verification result is used to characterize whether the abnormal sound prediction data is error data;
the correction module 804 is configured to correct the reference sound data to obtain target sound data if the abnormal sound verification result indicates that the abnormal sound prediction data is error data, and upload the target sound data to a preset abnormal sound database;
The updating module 805 is configured to perform model parameter adjustment on the abnormal sound prediction model based on the abnormal sound database to obtain an abnormal sound identification model;
The second recognition module 806 is configured to perform abnormal sound recognition on the current sound data acquired in advance according to the abnormal sound recognition model.
The specific implementation manner of the vehicle abnormal sound identification device is basically the same as the specific embodiment of the vehicle abnormal sound identification method, and is not repeated here.
The embodiment of the application also provides a vehicle, which comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the abnormal sound identification method of the vehicle when executing the computer program.
Referring to fig. 9, fig. 9 illustrates a hardware structure of a vehicle of another embodiment, the vehicle including:
The processor 901 may be implemented by a general purpose CPU (Central Processing Unit ), a microprocessor, an Application SPECIFIC INTEGRATED Circuit (ASIC), or one or more integrated circuits, etc. for executing related programs, so as to implement the technical solution provided by the embodiments of the present application;
The Memory 902 may be implemented in the form of a Read Only Memory (ROM), a static storage device, a dynamic storage device, or a random access Memory (Random Access Memory, RAM). The memory 902 may store an operating system and other application programs, and when the technical solution provided in the embodiments of the present disclosure is implemented by software or firmware, relevant program codes are stored in the memory 902, and the processor 901 invokes the method for identifying abnormal sound of a vehicle to execute the embodiments of the present disclosure;
an input/output interface 903 for inputting and outputting information;
The communication interface 904 is configured to implement communication interaction between the device and other devices, and may implement communication in a wired manner (e.g. USB, network cable, etc.), or may implement communication in a wireless manner (e.g. mobile network, WIFI, bluetooth, etc.);
a bus 905 that transfers information between the various components of the device (e.g., the processor 901, the memory 902, the input/output interface 903, and the communication interface 904);
wherein the processor 901, the memory 902, the input/output interface 903 and the communication interface 904 are communicatively coupled to each other within the device via a bus 905.
The embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program realizes the vehicle abnormal sound identification method when being executed by a processor.
The memory, as a non-transitory computer readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory remotely located relative to the processor, the remote memory being connectable to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The embodiments described in the embodiments of the present application are for more clearly describing the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application, and those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.
It will be appreciated by persons skilled in the art that the embodiments of the application are not limited by the illustrations, and that more or fewer steps than those shown may be included, or certain steps may be combined, or different steps may be included.
The above described apparatus embodiments are merely illustrative, wherein the units illustrated as separate components may or may not be physically separate, i.e. may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
Those of ordinary skill in the art will appreciate that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof.
The terms "first," "second," "third," "fourth," and the like in the description of the application and in the above figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be understood that in the present application, "at least one (item)" means one or more, and "a plurality" means two or more. "and/or" is used to describe an association relationship of an associated object, and indicates that three relationships may exist, for example, "a and/or B" may indicate that only a exists, only B exists, and three cases of a and B exist simultaneously, where a and B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one of a, b or c may represent a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.
In the several embodiments provided by the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the above-described division of units is merely a logical function division, and there may be another division manner in actual implementation, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including multiple instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method of the various embodiments of the present application. The storage medium includes various media capable of storing programs, such as a USB flash disk, a removable hard disk, a Read-only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk.
The preferred embodiments of the present application have been described above with reference to the accompanying drawings, and are not thereby limiting the scope of the claims of the embodiments of the present application. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and spirit of the embodiments of the present application shall fall within the scope of the claims of the embodiments of the present application.