Disclosure of Invention
Solves the technical problem
Aiming at the defects in the prior art, the invention provides the industrial equipment voice control method and the system based on the robust voice enhancement algorithm, and solves the technical problems in the background technology.
Technical scheme
In order to achieve the purpose, the invention is realized by the following technical scheme:
in a first aspect, a robust speech enhancement algorithm based industrial equipment speech control method comprises the following steps:
step 1: analyzing an industrial equipment deployment area, setting an area boundary, and constructing a local area network in the industrial equipment deployment area after the area boundary is set;
step 2: acquiring industrial equipment position information in an industrial equipment deployment area, and generating an industrial equipment topological structure image according to the position information;
and 3, step 3: capturing point positions with equal distance and short distance between every two adjacent groups of industrial equipment in the topological image;
and 4, step 4: acquiring point locations captured in the topological structure image of the industrial equipment in thestep 3, acquiring point locations corresponding to the industrial equipment consisting of industrial equipment position information in the topological structure image of the industrial equipment, and installing voice recognition equipment in the acquired point locations;
and 5: analyzing the signal-to-noise ratio of the industrial equipment in the industrial equipment deployment area;
the signal-to-noise ratio calculation formula is as follows:
wherein, in the formula in the step 5: m is the frame number of the voice; n is the length of each frame of speech;
is the starting point of each frame of speech; s (i) is a clean speech signal;
the application signal after noise reduction;
and 6: acquiring the signal-to-noise ratio obtained by analyzing each analysis target, and confirming the industrial equipment voice recognition equipment corresponding to the analysis target with the signal-to-noise ratio within the range of the set signal-to-noise ratio safety threshold;
and 7: acquiring the signal-to-noise ratio obtained by analyzing each analysis target, and recording the voice recognition equipment of the industrial equipment corresponding to the analysis target of which the signal-to-noise ratio is not in the range of the safety threshold value of the set signal-to-noise ratio;
and step 8: analyzing the signal-to-noise ratio of the operating state of the industrial equipment in the industrial equipment deployment area;
and step 9: analyzing whether the signal-to-noise ratio of the industrial equipment in each current running state is in a signal-to-noise ratio safety threshold, analyzing the ratio of the industrial equipment which is not in the signal-to-noise ratio safety threshold to all industrial equipment, and skipping to the step 4 to execute when the ratio is more than 5%;
said step 8 is executed,step 5 is executed again to provide the execution of step 8 with the newly set safety threshold.
Furthermore, thestep 3 is executed for a plurality of times according to the user's own setting, and in the repeated execution process after thestep 3 is executed for the first time, two groups adjacent to each other in the point location obtained by the last operation of thestep 3 are applied when the point location is captured.
Furthermore, in the step 4, when the voice recognition device is installed, more than one group of voice recognition devices are deployed to the corresponding point locations of the industrial device in the topological result image of the industrial device, and more than two groups of voice recognition devices are deployed to the point locations captured in thestep 3.
Still further, saidstep 5 subordinate is provided with a sub-step comprising the steps of:
step 51: acquiring adjacent industrial equipment according to corresponding point positions in the topological structure images of the industrial equipment and the industrial equipment, enabling each two groups of adjacent industrial equipment to serve as signal-to-noise ratio analysis targets, and setting signal-to-noise ratio safety thresholds;
and the signal-to-noise safety threshold is manually edited and set by the user side.
Furthermore, the step 6 captures the execution status of thestep 5 and the step 7 in real time before the execution, and the step 6 is executed in the state that the execution of thestep 5 is finished, the execution of the step 7 is finished, or the execution of thestep 5 is finished and the step 7 is not executed.
Still further, said step 7 is provided with a sub-step, comprising the steps of:
step 71: monitoring the running state of thestep 5 in real time, and acquiring recorded voice recognition equipment from the step 7 after the running of thestep 5 is finished;
step 72: analyzing the difference value between the corresponding signal-to-noise ratio of each recorded voice recognition device and a set signal-to-noise ratio safety threshold value, and carrying out adaptive optimization on the voice recognition devices according to the difference value;
and 9, when the ratio of the industrial equipment which is not at the signal-to-noise ratio safety threshold value to all the industrial equipment is less than or equal to 5%, skipping to the next substep 72 of the step 7 for execution.
Further, the adaptive optimization of the voice recognition device in step 72 is manually coordinated and set by the user end;
the method for adaptively optimizing the voice recognition equipment by manually coordinating and setting the user side comprises the following steps: the method comprises the steps of coordinating operation parameters of voice recognition equipment components, coordinating the number of the voice recognition equipment components, coordinating circuit design of the voice recognition equipment components, coordinating new deployment of anti-interference shielding elements and coordinating deployment positions and intervals of sensitive components in the voice recognition equipment components.
Furthermore, the voice recognition device installed on the point position corresponding device in the topological structure image of the industrial device in the step 4 and the point position captured in thestep 3 monitors the circuit state of the voice recognition device in real time, and triggers another group of voice recognition devices to operate under the circuit short circuit and open circuit state.
In a second aspect, a robust speech enhancement algorithm based industrial equipment speech control system includes:
the control terminal is a main control end of the system and is used for sending out a control command;
the deployment module is used for acquiring a system service area and configuring the voice recognition equipment for the industrial equipment in the system service area;
the editing module is used for editing the recognizable voice of the voice recognition equipment;
the matching unit is used for identifying executable operation of the industrial equipment; the voice recognition device is used for receiving the recognizable voice edited by the editing module and matching the recognizable voice of the voice recognition device with the executable operation of the industrial equipment;
the sharing module is used for acquiring the voice received by any voice recognition device in real time and sending the voice to other voice recognition devices in a local area network;
a driving module: the voice matching device is used for driving the industrial equipment to run according to the executable operation matched with the voice.
Furthermore, the control terminal is electrically connected with a deployment module and an editing module through a medium, the editing module is internally and electrically connected with a matching unit through the medium, and the editing module is electrically connected with a sharing module and a driving module through the medium.
Advantageous effects
Compared with the known public technology, the technical scheme provided by the invention has the following beneficial effects:
the invention provides a robust speech enhancement algorithm-based industrial equipment speech control method, which can effectively identify the deployment area of industrial equipment through step execution in the method, thereby further constructing a local area network in the identified deployment area of the industrial equipment, and enabling the speech recognition equipment deployed on the industrial equipment to have interconnection and data sharing conditions, so that a user can issue a control command of any industrial equipment at any speech recognition equipment deployment position in the deployment area of the industrial equipment.
In the execution process of the steps of the method, a plurality of groups of voice recognition equipment can be further configured in the specified range of the industrial equipment deployment area by recognizing and acquiring the position information of each industrial equipment, so that the aim of issuing the industrial equipment control command by a convenient user through the voice recognition equipment is fulfilled.
In the execution process of the steps of the method, the voice recognition equipment deployed on the industrial equipment can be continuously coordinated twice, the voice recognition equipment is ensured to be suitable for the industrial equipment, and the application quantity of the voice recognition equipment is more suitable, so that the aim of saving the cost for configuring the voice recognition equipment required by system implementation is fulfilled.
4. The invention provides an industrial equipment voice control system based on a robust voice enhancement algorithm, which provides necessary step execution conditions for the step execution of the method in the invention through the operation of the system, so that voice recognition equipment deployed on industrial equipment can have a specified voice recognition function, and the industrial equipment is driven to operate by the specified voice recognition function.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive step based on the embodiments of the present invention, are within the scope of protection of the present invention.
The present invention will be further described with reference to the following examples.
Example 1
The robust speech enhancement algorithm-based industrial equipment speech control method of the embodiment, as shown in fig. 1, includes the following steps:
step 1: analyzing an industrial equipment deployment area, setting an area boundary, and constructing a local area network in the industrial equipment deployment area after the area boundary is set;
step 2: acquiring industrial equipment position information in an industrial equipment deployment area, and generating an industrial equipment topological structure image according to the position information;
and step 3: capturing point positions with equal distance and short distance between every two adjacent groups of industrial equipment in the topological image;
and 4, step 4: acquiring point locations captured in the topological structure image of the industrial equipment in thestep 3, acquiring point locations corresponding to the industrial equipment consisting of industrial equipment position information in the topological structure image of the industrial equipment, and installing voice recognition equipment in the acquired point locations;
and 5: analyzing the signal-to-noise ratio of the industrial equipment in the industrial equipment deployment area in the closed state;
the signal-to-noise ratio calculation formula is as follows:
wherein, in the formula in step 5: m is the frame number of the voice; n is the length of each frame of speech;
is the starting point of each frame of speech; s (i) is a clean speech signal;
the application signal after noise reduction; step 6: acquiring the signal-to-noise ratio obtained by analyzing each analysis target, and confirming the industrial equipment voice recognition equipment corresponding to the analysis target with the signal-to-noise ratio within the set signal-to-noise ratio safety threshold range;
and 7: acquiring the signal-to-noise ratio obtained by analyzing each analysis target, and recording the voice recognition equipment of the industrial equipment corresponding to the analysis target of which the signal-to-noise ratio is not in the range of the safety threshold value of the set signal-to-noise ratio;
and step 8: analyzing the signal-to-noise ratio of the operating state of the industrial equipment in the industrial equipment deployment area;
and step 9: analyzing whether the signal-to-noise ratio of the industrial equipment in each current running state is in a signal-to-noise ratio safety threshold, analyzing the ratio of the industrial equipment which is not in the signal-to-noise ratio safety threshold to all industrial equipment, and skipping to the step 4 to execute when the ratio is more than 5%;
step 8 when executedstep 5 is executed again providing the execution of step 8 with the newly set safety threshold.
Example 2
In a specific implementation level, on the basis of embodiment 1, this embodiment further specifically describes, with reference to fig. 1, the robust speech enhancement algorithm-based speech control method for industrial equipment in embodiment 1:
and 3, executing for a plurality of times according to the autonomous setting of the user, and applying two adjacent groups of point positions obtained in the last operation of thestep 3 when the point positions are captured in the repeated execution process after thestep 3 is executed for the first time.
Through the arrangement, more quantity can be configured for the voice recognition equipment when the voice recognition equipment is deployed in an industrial equipment deployment area, on one hand, the purpose that other voice recognition equipment can be used for replacement when the voice recognition equipment breaks down can be ensured, and on the other hand, the convenience and the location of a user when the industrial equipment control command is issued through the voice recognition module can be improved.
As shown in fig. 1, in step 4, when the speech recognition device is installed, more than one group of speech recognition devices is deployed to the corresponding point locations of the industrial device in the topological result image of the industrial device, and more than two groups of speech recognition devices are deployed to the point locations captured instep 3.
By limiting the number of the voice recognition devices during deployment, a user can control the industrial device to a certain extent when a fault occurs in a specific voice recognition device, so that the robustness of the voice recognition device controlled and managed by the method in the operation process is better.
As shown in fig. 1, thestep 5 is provided with sub-steps at the lower level, including the following steps:
step 51: acquiring adjacent industrial equipment according to corresponding point positions in the topological structure images of the industrial equipment and the industrial equipment, enabling each two groups of adjacent industrial equipment to serve as signal-to-noise ratio analysis targets, and setting signal-to-noise ratio safety thresholds;
wherein, the signal noise safety threshold is manually edited and set by the user terminal.
As shown in fig. 1, step 6 captures the execution status ofstep 5 and step 7 in real time before execution, and step 6 is executed when execution ofstep 5 ends, execution of step 7 ends, or execution ofstep 5 ends and step 7 does not end.
As shown in fig. 1, the step 7 is provided with sub-steps at the lower level, including the following steps:
step 71: monitoring the running state of thestep 5 in real time, and acquiring recorded voice recognition equipment from the step 7 after the running of thestep 5 is finished;
step 72: analyzing the difference between the signal-to-noise ratio corresponding to each recorded voice recognition device and a set signal-to-noise ratio safety threshold, and performing adaptive optimization on the voice recognition device according to the difference;
and 9, when the ratio of the industrial equipment which is not at the signal-to-noise ratio safety threshold value to all the industrial equipment is less than or equal to 5%, skipping to the next substep 72 of the step 7 for execution.
As shown in fig. 1, the adaptive optimization of the speech recognition device in step 72 is manually coordinated and set by the user end;
the method for adaptively optimizing the voice recognition equipment by manually coordinating and setting the user side comprises the following steps: the method comprises the steps of coordinating operation parameters of voice recognition equipment components, coordinating the number of the voice recognition equipment components, coordinating circuit design of the voice recognition equipment components, coordinating new deployment of anti-interference shielding elements and coordinating deployment positions and intervals of sensitive components in the voice recognition equipment components.
As shown in fig. 1, the voice recognition devices installed on the devices corresponding to each point in the topological structure image of the industrial device in step 4 and the points captured instep 3 monitor the circuit state of the device in real time, and trigger another group of voice recognition devices to operate in the short-circuit and open-circuit states of the circuit.
Example 3
In a specific implementation level, on the basis of embodiment 1, this embodiment further specifically describes, with reference to fig. 2, the robust speech enhancement algorithm-based speech control method for industrial equipment in embodiment 1:
an industrial equipment voice control system based on a robust voice enhancement algorithm, comprising:
the control terminal 1 is a main control end of the system and is used for sending out a control command;
thedeployment module 2 is used for acquiring a system service area and configuring the voice recognition equipment for the industrial equipment in the system service area;
theediting module 3 is used for editing the recognizable voice of the voice recognition equipment;
amatching unit 31 for identifying an operation executable by the industrial equipment; the voice recognition device is used for receiving the recognizable voice of the voice recognition device edited by theediting module 3 and matching the recognizable voice of the voice recognition device with the executable operation of the industrial equipment;
the sharing module 4 is used for acquiring the voice received by any voice recognition device in real time and sending the voice to other voice recognition devices in a local area network;
the driving module 5: the voice matching device is used for driving the industrial equipment to run according to the executable operation matched with the voice.
In this embodiment, the control terminal 1 controls thedeployment module 2 to operate and acquire a system service area, and configures the voice recognition device for the industrial device in the system service area, theediting module 3 operates the recognizable voice of the editing voice recognition device in real time in a rear manner, the matchingunit 31 recognizes the executable operation of the industrial device, receives the recognizable voice of the voice recognition device edited by theediting module 3, and matches the recognizable voice of the voice recognition device with the executable operation of the industrial device, the synchronous sharing module 4 acquires the voice received by any voice recognition device in real time, and sends the voice to other voice recognition devices in the local area network, and finally thedriving module 5 drives the industrial device to operate according to the executable operation matched with the voice.
As shown in fig. 1, the control terminal 1 is electrically connected to adeployment module 2 and anediting module 3 through a medium, theediting module 3 is electrically connected to amatching unit 31 through a medium, and theediting module 3 is electrically connected to a sharing module 4 and adriving module 5 through a medium.
Example 4
In a specific implementation level, on the basis of embodiment 1, this embodiment further specifically describes, with reference to fig. 2, the robust speech enhancement algorithm-based speech control method for industrial equipment in embodiment 1:
an industrial equipment voice control system based on a robust voice enhancement algorithm, comprising:
the control terminal 1 is a main control end of the system and is used for sending out a control command;
thedeployment module 2 is used for acquiring a system service area and configuring voice recognition equipment for industrial equipment in the system service area;
theediting module 3 is used for editing the recognizable voice of the voice recognition device;
amatching unit 31 for identifying an operation executable by the industrial equipment; the voice recognition device is used for receiving the recognizable voice of the voice recognition device edited by theediting module 3 and matching the recognizable voice of the voice recognition device with the executable operation of the industrial equipment;
the sharing module 4 is used for acquiring the voice received by any voice recognition device in real time and sending the voice to other voice recognition devices in a local area network;
the driving module 5: the voice matching device is used for driving the industrial equipment to run according to the executable operation matched with the voice.
When the method in the above embodiment is actually applied, any existing speech enhancement algorithm can be used as a speech enhancement algorithm used by the speech recognition device during operation as data support, and the existing speech enhancement algorithm is selected as an example as follows:
auditory masking method
The auditory masking effect means psychologically that when a strong signal is present in the vicinity of a weaker signal, the weak signal will not be perceived, i.e. the perception threshold of sound X is increased by the presence of sound B, a phenomenon known as the auditory masking effect of the human ear. For example, in a factory building with a noisy machine sound, the workers need to increase the voice to make the other party hear the voice. In the auditory masking effect, the maximum sound pressure level of the masked imperceptible signal is called the masking threshold, below which all signals will be masked. Due to the presence of the masking sound, a signal in the vicinity of which the energy is below the masking threshold cannot be perceived even if the energy is not zero, because of the masking effect. Currently, there are three main types of speech enhancement methods based on auditory masking: the method based on the probability of the noise being masked: when the noise component is masked by the voice, the voice signal with the noise is not processed, and when the noise component is not masked by the voice, the voice is processed by using a voice enhancement algorithm. Auditory masking effects are combined with spectral subtraction: the noise weight coefficient a and the power smoothing coefficient b in the spectral subtraction are dynamically adjusted through the auditory masking threshold, and the balance is achieved between noise reduction and distortion reduction. The perceptual filter method comprises: the perception filter does not completely eliminate the residual noise, and utilizes the masking effect of human ears to control the residual noise to be below a masking threshold, so that the residual noise is not perceived by the human ears, and the distortion of voice can be reduced while the noise is suppressed.
There are three main categories of current speech enhancement methods based on auditory masking.
1. The method based on the probability of the noise being masked: when the noise component is masked by the voice, the voice signal with the noise is not processed, and when the noise component is not masked by the voice, the voice is processed by using a voice enhancement algorithm.
2. Auditory masking effects are combined with spectral subtraction: the noise weight coefficient a and the power smoothing coefficient beta in the spectral subtraction are dynamically adjusted through the auditory masking threshold, and balance is achieved between noise reduction and distortion reduction.
3. The perceptual filter method comprises the following steps: the perception filter does not completely eliminate the residual noise, and utilizes the masking effect of human ears to control the residual noise to be below a masking threshold, so that the residual noise is not perceived by the human ears, and the distortion of voice can be reduced while the noise is suppressed.
In summary, the deployment areas of the industrial devices can be effectively identified through the above embodiments, so that a local area network is further constructed in the identified deployment areas of the industrial devices, and the subsequent voice recognition devices deployed on the industrial devices have interconnection and data sharing conditions, so that a user can issue a control command of any industrial device at any deployment position of the voice recognition device in the deployment areas of the industrial devices; in the execution process of the steps, by identifying and acquiring the position information of each industrial device, a plurality of groups of voice recognition devices can be further configured in the specified range of the industrial device deployment area, so that the aim of issuing the industrial device control command by a convenient user through the voice recognition devices is fulfilled; meanwhile, the voice recognition equipment deployed on the industrial equipment can be continuously coordinated twice, so that the voice recognition equipment is suitable for the industrial equipment, and the application quantity of the voice recognition equipment is more suitable, so that the aim of saving the cost for configuring the voice recognition equipment for system implementation is fulfilled; in addition, the robust speech enhancement algorithm-based industrial equipment speech control system described in the embodiment can provide necessary step execution conditions for the step execution of the method, so that the speech recognition equipment deployed on the industrial equipment can have a specified speech recognition function, and the industrial equipment is driven to operate by the specified speech recognition function.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.