Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method and a system for diagnosing the online faults of a calorimeter, and solves the problems in the background art.
The heat meter online fault diagnosis system comprises a data acquisition and preprocessing module, a feature extraction module, an abnormality detection module, a fault type judgment module and an execution iteration module;
the data acquisition and preprocessing module acquires the data of the heat meter operation area in real time through integrating a sensor group at the position of the heat meter operation area, and performs preprocessing to form an acquisition data group DAT;
The feature extraction module performs feature extraction on the preprocessed collected data set DAT to obtain a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF and a noise intensity feature NRF, and a feature vector FVEC is formed;
The abnormality detection module processes the obtained feature vector FVEC by using an online abnormality detection algorithm to obtain an abnormality detection index AS, and matches the abnormality detection index AS with a preset operation abnormal state threshold YZ to obtain a calorimeter operation state result;
The fault type judging module is used for fitting according to the abnormal detection index AS and the feature vector FVEC to obtain a fault feature probability index TYP;
The execution iteration module is used for matching the obtained fault characteristic probability index TYP with a preset heat meter operation characteristic abnormality evaluation threshold TZ, obtaining a heat meter operation abnormality characteristic response strategy scheme, executing specifically according to the content of the heat meter operation abnormality characteristic response strategy scheme, recording an execution result, and adjusting the fault characteristic probability index TYP in an iteration mode.
Preferably, the data acquisition and preprocessing module comprises a data acquisition unit and a data preprocessing unit;
the data acquisition unit acquires heat meter operation area data in real time through integrating a sensor group at the heat meter operation area position, wherein the heat meter operation area data comprises fluid temperature TMP, flow FLW, pipeline pressure PRS, environment temperature ENV, vibration intensity VIB and operation noise NRM;
the sensor group comprises a temperature sensor, an ultrasonic flow sensor, a pressure sensor, an environment temperature sensor, a vibration sensor and a noise sensor;
The data preprocessing unit preprocesses the collected operation area data, including noise removal processing, missing value processing and data standardization processing, to form a collected data group DAT;
The method comprises the steps of performing noise removal processing, namely performing noise removal processing on time sequence data by using moving average filtering, and meanwhile, keeping the trend of running region data, performing missing value processing on the time sequence data by using linear interpolation to fill in missing values by using the linear relation of adjacent data points so as to restore a complete data sequence, and performing data normalization preprocessing on the time sequence data by using Z-score normalization so as to adjust different parameters to be in the same dimension.
Preferably, the feature extraction module comprises a feature calculation unit and a feature integration unit;
The feature calculation unit performs feature extraction on the preprocessed acquired data set DAT to obtain a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF and a noise intensity feature NRF;
The feature integration unit marks the time t for the temperature difference feature TDF, the flow change feature FCF, the pressure fluctuation feature PWF, the vibration frequency feature VDF, and the noise intensity feature NRF, and integrates the same to form a feature vector FVEC.
Preferably, the temperature difference feature TDF is used for reflecting the difference state between the fluid temperature of the calorimeter and the ambient temperature, the flow change feature FCF is used for reflecting the change rate of the flow and the stability of the flow, the pressure fluctuation feature PWF is used for reflecting the stability of the pressure in the pipeline, the vibration frequency feature VDF is used for analyzing the vibration state of the equipment or the pipeline, particularly extracting the frequency component of a vibration signal through Fourier transformation for feedback, and the noise intensity feature NRF is used for reflecting the noise level of the equipment and the pipeline in the operation process.
Preferably, the abnormality detection module includes an abnormality detection unit and a state matching unit;
the abnormality detection unit performs weight assignment and weight calculation processing on the obtained temperature difference feature TDF, flow change feature FCF, pressure fluctuation feature PWF, vibration frequency feature VDF and noise intensity feature NRF in the feature vector FVEC by using an online abnormality detection algorithm, and obtains an abnormality detection index AS by integrating the weight deviation of the temperature difference feature TDF, the flow change feature FCF, the pressure fluctuation feature PWF, the vibration frequency feature VDF and the noise intensity feature NRF, and is used for triggering the heat meter operation state result matching of the state matching unit;
the state matching unit is used for matching the preset operation abnormal state threshold YZ with the abnormality detection index AS to obtain a heat meter operation state result, and triggering the execution of the fault type judging module according to the heat meter operation state result.
Preferably, the operation state result of the heat meter is marked as a return signal, and the execution of the fault type judging module is triggered according to the return signal;
when the return signal is 1, acquiring abnormal operation state of the heat meter, and triggering the execution of the fault type judging module;
When the return signal is 0, acquiring that the running state of the heat meter is not abnormal, and not triggering the execution of the fault type judging module.
Preferably, the fault type judging module performs verification on the abnormality detection index AS and the feature vector FVEC through a verification mechanism, verifies the integrity and consistency of data, temporarily stores the abnormality detection index AS and the feature vector FVEC in a buffer area synchronously, sequentially processes the abnormality detection index AS and the feature vector FVEC according to a time sequence, and comprehensively processes the abnormality detection index AS and the feature vector FVEC by using a line fitting algorithm to obtain a fault feature probability index TYP, thereby realizing quantitative evaluation on different fault types.
Preferably, the execution iteration module comprises an abnormality evaluation unit and an iteration regulation unit;
The abnormality evaluation unit is used for matching the obtained fault characteristic probability index TYP with a preset heat meter operation characteristic abnormality evaluation threshold TZ to obtain a heat meter operation abnormality characteristic response strategy scheme;
The iteration regulation and control unit performs specific execution according to the abnormal operation characteristic response strategy scheme content of the heat meter, and iteratively adjusts the fault characteristic probability index TYP by recording an execution result.
Preferably, the heat meter operation abnormal characteristic response strategy scheme is obtained by the following matching mode:
When the fault characteristic probability index TYP is more than or equal to the heat meter operation characteristic abnormality evaluation threshold TZ, acquiring a heat meter operation state abnormality response strategy scheme, wherein the strategy scheme comprises the steps of switching an operation mode, starting a self-diagnosis function of the heat meter, giving an alarm, recording a log and notifying related patrol personnel to overhaul;
And when the fault characteristic probability index TYP is smaller than the heat meter operation characteristic abnormality evaluation threshold TZ, acquiring a heat meter operation state abnormality non-response strategy scheme.
An online fault diagnosis method for a calorimeter comprises the following steps:
The method comprises the steps that firstly, a data acquisition and preprocessing module acquires heat meter operation area data in real time through integrating a sensor group at the heat meter operation area position, and performs preprocessing to form an acquisition data group DAT;
the feature extraction module performs feature extraction on the preprocessed acquired data set DAT to obtain a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF and a noise intensity feature NRF, and a feature vector FVEC is formed;
The abnormality detection module processes the obtained feature vector FVEC by using an online abnormality detection algorithm to obtain an abnormality detection index AS, and matches the abnormality detection index AS with a preset operation abnormal state threshold YZ to obtain a calorimeter operation state result;
Fitting by the fault type judging module according to the abnormality detection index AS and the feature vector FVEC to obtain a fault feature probability index TYP;
And fifthly, matching the obtained fault characteristic probability index TYP with a preset heat meter operation characteristic abnormality evaluation threshold TZ by an execution iteration module, obtaining a heat meter operation abnormality characteristic response strategy scheme, performing specific execution according to the content of the heat meter operation abnormality characteristic response strategy scheme, and iteratively adjusting the fault characteristic probability index TYP by recording an execution result.
The invention provides a heat meter online fault diagnosis method and system, which have the following beneficial effects:
(1) When the system operates, data are collected and preprocessed in real time through the integrated sensor group, key features are extracted to form feature vectors FVEC, and an on-line abnormality detection algorithm and a fault type judgment model are utilized to accurately calculate an abnormality detection index AS and a fault feature probability index TYP. The data not only can reflect the health state of the system in real time, but also further optimizes the response strategy and fault judgment by executing the iteration module, so that the system has the self-diagnosis and self-adaptive adjustment capabilities. The optimization process can effectively solve the defect that the traditional calorimeter system cannot process and detect faults on line in real time, improves the accuracy and response speed of fault identification, avoids equipment damage or energy waste caused by delay detection, further ensures continuous and stable operation of the system, and provides a more accurate and timely fault early warning and response mechanism.
(2) In the data acquisition and preprocessing stage, key parameters are acquired in real time by integrating various high-precision sensors, and preprocessing is performed, including moving average filtering, linear interpolation and Z-score standardization, so that high quality and consistency of data are ensured. In the feature extraction process, the system extracts multidimensional features reflecting the operation state of the heat meter through a complex feature calculation formula, and integrates the features into a feature vector FVEC with a time mark. The process not only ensures the integrity and accuracy of the data, but also enables the system to comprehensively capture the dynamic change of the running state. Compared with the traditional method, the system has higher sensitivity and robustness in the aspects of detecting tiny anomalies and various complex fault modes, thereby effectively avoiding misjudgment caused by inaccurate data or improper processing and ensuring stable operation and accurate metering of the heat meter.
(3) The feature vector FVEC is processed through an online abnormality detection algorithm, an abnormality detection index AS is accurately calculated, and the operation state of the heat meter is judged in real time through matching with a preset operation abnormal state threshold YZ. The intelligent monitoring system for the abnormal state of the heat meter is capable of ensuring the acute capture of various characteristic deviation degrees through strict standardization and weight calculation, accurately identifying the abnormal state, realizing the intelligent triggering of the fault type judging module through a return mechanism of a return signal, carrying out further fault analysis only when the abnormality is detected, avoiding the waste of system resources, being also suitable for continuous monitoring under complex working conditions, ensuring that the system can quickly judge and respond when the system faces an emergency, reducing unnecessary downtime and improving the overall operation efficiency and stability of the heat meter.
(4) The abnormal detection index AS and the feature vector FVEC are subjected to fitting calculation, so that the fault feature probability index TYP can be accurately obtained, and then an abnormal response strategy with strong pertinence is intelligently generated by matching with a preset threshold TZ. The system can be used for rapidly switching the operation mode and starting the self-diagnosis function when the actual fault occurs, notifying maintenance personnel in time, reducing unnecessary intervention under the condition of no serious abnormality, and optimizing the utilization rate of system resources.
Detailed Description
The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, but not all embodiments, and all other embodiments obtained by those skilled in the art without making any inventive effort based on the embodiments of the present invention are within the scope of protection of the present invention.
Example 1
The invention provides an online fault diagnosis system of a calorimeter, referring to FIG. 1, which comprises a data acquisition and preprocessing module, a feature extraction module, an abnormality detection module, a fault type judgment module and an execution iteration module;
the data acquisition and preprocessing module acquires the data of the heat meter operation area in real time through integrating a sensor group at the position of the heat meter operation area, and performs preprocessing to form an acquisition data group DAT;
The feature extraction module performs feature extraction on the preprocessed collected data set DAT to obtain a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF and a noise intensity feature NRF, and a feature vector FVEC is formed;
The abnormality detection module processes the obtained feature vector FVEC by using an online abnormality detection algorithm to obtain an abnormality detection index AS, and matches the abnormality detection index AS with a preset operation abnormal state threshold YZ to obtain a calorimeter operation state result;
The fault type judging module is used for fitting according to the abnormal detection index AS and the feature vector FVEC to obtain a fault feature probability index TYP;
The execution iteration module is used for matching the obtained fault characteristic probability index TYP with a preset heat meter operation characteristic abnormality evaluation threshold TZ, obtaining a heat meter operation abnormality characteristic response strategy scheme, executing specifically according to the content of the heat meter operation abnormality characteristic response strategy scheme, recording an execution result, and adjusting the fault characteristic probability index TYP in an iteration mode.
In this embodiment, data is collected and preprocessed in real time by an integrated sensor group, key features are extracted to form feature vectors FVEC, and an on-line abnormality detection algorithm and a fault type judgment model are utilized to accurately calculate an abnormality detection index AS and a fault feature probability index TYP. The data not only can reflect the health state of the system in real time, but also further optimizes the response strategy and fault judgment by executing the iteration module, so that the system has the self-diagnosis and self-adaptive adjustment capabilities. The optimization process can effectively solve the defect that the traditional calorimeter system cannot process and detect faults on line in real time, improves the accuracy and response speed of fault identification, avoids equipment damage or energy waste caused by delay detection, further ensures continuous and stable operation of the system, and provides a more accurate and timely fault early warning and response mechanism.
Example 2
The embodiment is explained in embodiment 1, please refer to fig. 1, specifically, the data acquisition and preprocessing module includes a data acquisition unit and a data preprocessing unit;
the data acquisition unit acquires heat meter operation area data in real time through integrating a sensor group at the heat meter operation area position, wherein the heat meter operation area data comprises fluid temperature TMP, flow FLW, pipeline pressure PRS, environment temperature ENV, vibration intensity VIB and operation noise NRM;
the sensor group comprises a temperature sensor, an ultrasonic flow sensor, a pressure sensor, an environment temperature sensor, a vibration sensor and a noise sensor;
the data preprocessing unit is used for preprocessing the acquired operation area data, and comprises noise removal processing, missing value processing and data standardization processing, so as to form an acquired data set DAT, specifically DAT= { TMP, FLW, PRS, ENV, VIB, NRM };
The method comprises the steps of performing noise removal processing, namely performing noise removal processing on time sequence data by using moving average filtering, and meanwhile, keeping the trend of running region data, performing missing value processing on the time sequence data by using linear interpolation to fill in missing values by using the linear relation of adjacent data points so as to restore a complete data sequence, and performing data normalization preprocessing on the time sequence data by using Z-score normalization so as to adjust different parameters to be in the same dimension.
The feature extraction module comprises a feature calculation unit and a feature integration unit;
The feature calculation unit performs feature extraction on the preprocessed acquired data set DAT to obtain a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF and a noise intensity feature NRF;
The feature integration unit marks the time t for the temperature difference feature TDF, the flow change feature FCF, the pressure fluctuation feature PWF, the vibration frequency feature VDF and the noise intensity feature NRF, and integrates the two features to form a feature vector FVEC, specifically FVEC = { TDF, FCF, PWF, VDF, NRF, t }.
The system comprises a heat meter, a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF, a noise intensity feature NRF, a temperature difference feature TDF, a pressure fluctuation feature PWF and a noise intensity feature NRF, wherein the temperature difference feature TDF is used for reflecting the difference state between the fluid temperature of the heat meter and the ambient temperature;
The temperature difference characteristic TDF is obtained by the following characteristic calculation formula:
Wherein TMPi represents the fluid temperature at the i-th time point, ENVi represents the ambient temperature at the i-th time point, Δti represents the time interval at the i-th time point, log represents a logarithmic function, TMPmax represents the fluid temperature peak value, and TMPmin represents the fluid temperature valley value;
the flow change characteristic FCF is obtained by the following characteristic calculation formula:
Where T denotes the total observation time length, in particular a number of time intervals Deltat, which comprise a number of time points i, FLWi denotes the flow rate at time point i, d denotes the derivative,Representing the rate of change of the flow FLW with time point i, FLWmax representing the flow peak, FLWmin representing the flow trough;
The pressure fluctuation feature PWF is obtained by the following feature calculation formula:
where n represents the total number of time points, PRSi represents the line pressure at the i-th time point,Representing the mean value of the pipeline pressure, PRSprev representing the pipeline pressure at the i-1 th time point;
the vibration frequency characteristic VDF is obtained by the following characteristic calculation formula:
Wherein w represents the total number of frequency components after Fourier transformation, VIBk represents the kth frequency component of the vibration intensity signal, F (VIBk) represents the Fourier transformation result of the vibration intensity VIB signal on the kth frequency component, and particularly represents the amplitude of the vibration intensity VIB;
the noise intensity characteristic NRF is obtained by the following characteristic calculation formula:
wherein NRMi denotes the operation noise at the i-th time point,Representing the running noise mean.
In this embodiment, in the data acquisition and preprocessing stage, key parameters are obtained in real time by integrating multiple high-precision sensors, and preprocessing is performed, including moving average filtering, linear interpolation and Z-score normalization, so as to ensure high quality and consistency of data. In the feature extraction process, the system extracts multidimensional features reflecting the operation state of the heat meter through a complex feature calculation formula, and integrates the features into a feature vector FVEC with a time mark. The process not only ensures the integrity and accuracy of the data, but also enables the system to comprehensively capture the dynamic change of the running state. Compared with the traditional method, the system has higher sensitivity and robustness in the aspects of detecting tiny anomalies and various complex fault modes, thereby effectively avoiding misjudgment caused by inaccurate data or improper processing and ensuring stable operation and accurate metering of the heat meter.
Example 3
In the explanation of embodiment 2, please refer to fig. 1, specifically, the abnormality detection module includes an abnormality detection unit and a state matching unit;
the abnormality detection unit performs weight assignment and weight calculation processing on the obtained temperature difference feature TDF, flow change feature FCF, pressure fluctuation feature PWF, vibration frequency feature VDF and noise intensity feature NRF in the feature vector FVEC by using an online abnormality detection algorithm, and obtains an abnormality detection index AS by integrating the weight deviation of the temperature difference feature TDF, the flow change feature FCF, the pressure fluctuation feature PWF, the vibration frequency feature VDF and the noise intensity feature NRF, and is used for triggering the heat meter operation state result matching of the state matching unit;
The abnormality detection index AS is obtained by the following calculation formula:
Where m represents the total number of features in feature vector FVEC, FVECj represents the j-th feature in feature vector FVEC,Representing the mean of the jth feature in feature vector FVEC, σ FVECj representing the standard deviation of the jth feature in feature vector FVEC, specifically representing the range of fluctuation of the feature values,The standard deviation of the jth feature in the feature vector FVEC is represented, the degree of deviation of the feature is specifically represented, and uj represents the preset weight value of the jth feature;
the state matching unit is used for matching the preset operation abnormal state threshold YZ with the abnormality detection index AS to obtain a heat meter operation state result, and triggering the execution of the fault type judging module according to the heat meter operation state result.
Marking the operation state result of the heat meter as a return signal, and triggering the execution of the fault type judging module according to the return signal;
the return signal passes throughObtaining a marking mode;
when the return signal is 1, acquiring abnormal operation state of the heat meter, and triggering the execution of the fault type judging module;
When the return signal is 0, acquiring that the running state of the heat meter is not abnormal, and not triggering the execution of the fault type judging module.
In this embodiment, the feature vector FVEC is processed by an online anomaly detection algorithm, the anomaly detection index AS is accurately calculated, and the running state of the heat meter is judged in real time by matching with a preset running anomaly state threshold YZ. The intelligent monitoring system for the abnormal state of the heat meter is capable of ensuring the acute capture of various characteristic deviation degrees through strict standardization and weight calculation, accurately identifying the abnormal state, realizing the intelligent triggering of the fault type judging module through a return mechanism of a return signal, carrying out further fault analysis only when the abnormality is detected, avoiding the waste of system resources, being also suitable for continuous monitoring under complex working conditions, ensuring that the system can quickly judge and respond when the system faces an emergency, reducing unnecessary downtime and improving the overall operation efficiency and stability of the heat meter.
Example 4
Referring to fig. 1, the fault type judging module checks the abnormality detection index AS and the feature vector FVEC through a checking mechanism, verifies the integrity and consistency of data, synchronously stores the abnormality detection index AS and the feature vector FVEC in a buffer area, sequentially processes the abnormality detection index AS and the feature vector FVEC according to a time sequence, and comprehensively processes the abnormality detection index AS and the feature vector FVEC by using a line fitting algorithm to obtain a fault feature probability index TYP so AS to realize quantitative evaluation of different fault types;
The failure feature probability index TYP is obtained by the following calculation formula:
wherein TYPk denotes a fault feature probability index TYP of the kth fault type, m denotes a total number of features in the feature vector FVEC, pk, j denotes a preset weight value between the kth fault type and the jth feature, FVECj denotes the jth feature in the feature vector FVEC, σ FVECj denotes a standard deviation of the jth feature in the feature vector FVEC, and αk denotes a sensitivity coefficient of the kth fault type;
The fault types include a temperature abnormal fault type, a flow abnormal fault type, a pressure abnormal fault type, a vibration abnormal fault type and a noise abnormal fault type.
The execution iteration module comprises an abnormality evaluation unit and an iteration regulation unit;
The abnormality evaluation unit is used for matching the obtained fault characteristic probability index TYP with a preset heat meter operation characteristic abnormality evaluation threshold TZ to obtain a heat meter operation abnormality characteristic response strategy scheme;
The iteration regulation and control unit performs specific execution according to the abnormal operation characteristic response strategy scheme content of the heat meter, and iteratively adjusts the fault characteristic probability index TYP by recording an execution result.
The abnormal characteristic response strategy scheme of the heat meter operation is obtained by the following matching mode:
When the fault characteristic probability index TYP is more than or equal to the heat meter operation characteristic abnormality evaluation threshold TZ, acquiring a heat meter operation state abnormality response strategy scheme, wherein the strategy scheme comprises the steps of switching an operation mode, starting a self-diagnosis function of the heat meter, giving an alarm, recording a log and notifying related patrol personnel to overhaul;
And when the fault characteristic probability index TYP is smaller than the heat meter operation characteristic abnormality evaluation threshold TZ, acquiring a heat meter operation state abnormality non-response strategy scheme.
In this embodiment, the anomaly detection index AS and the feature vector FVEC are subjected to fitting calculation, so that the fault feature probability index TYP can be accurately obtained, and then an anomaly response strategy with strong pertinence is intelligently generated by matching with a preset threshold TZ. The system can be used for rapidly switching the operation mode and starting the self-diagnosis function when the actual fault occurs, notifying maintenance personnel in time, reducing unnecessary intervention under the condition of no serious abnormality, and optimizing the utilization rate of system resources.
Example 5
Referring to fig. 2, the on-line fault diagnosis method for the heat meter specifically includes the following steps:
The method comprises the steps that firstly, a data acquisition and preprocessing module acquires heat meter operation area data in real time through integrating a sensor group at the heat meter operation area position, and performs preprocessing to form an acquisition data group DAT;
the feature extraction module performs feature extraction on the preprocessed acquired data set DAT to obtain a temperature difference feature TDF, a flow change feature FCF, a pressure fluctuation feature PWF, a vibration frequency feature VDF and a noise intensity feature NRF, and a feature vector FVEC is formed;
The abnormality detection module processes the obtained feature vector FVEC by using an online abnormality detection algorithm to obtain an abnormality detection index AS, and matches the abnormality detection index AS with a preset operation abnormal state threshold YZ to obtain a calorimeter operation state result;
Fitting by the fault type judging module according to the abnormality detection index AS and the feature vector FVEC to obtain a fault feature probability index TYP;
And fifthly, matching the obtained fault characteristic probability index TYP with a preset heat meter operation characteristic abnormality evaluation threshold TZ by an execution iteration module, obtaining a heat meter operation abnormality characteristic response strategy scheme, performing specific execution according to the content of the heat meter operation abnormality characteristic response strategy scheme, and iteratively adjusting the fault characteristic probability index TYP by recording an execution result.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.