An online fault diagnosis method for solenoid valves based on time-frequency domain characteristic analysis

Abstract

A method for online diagnosis of solenoid valve faults based on time-frequency domain characteristic analysis is presented. [Solution] Obtain the time-series current data of the solenoid valve driving side, perform time domain and frequency domain analysis respectively to obtain the feature vectors of the time domain and frequency domain, and combine both to form a composite time-frequency domain feature vector, and obtain multiple groups of labeled composite feature vectors by changing the operating state of the solenoid valve, and generate an isolated forest model evaluated using the majority voting method based on this. By integrating the current data of multiple states with the integrated feature vector from the time-frequency domain analysis, different fault types of the solenoid valve can be more accurately identified and distinguished, not only improving the feature characterization ability, but also improving the accuracy and reliability of fault diagnosis by comprehensively utilizing the information of the time domain and the frequency domain.

Description

[Technical field]

[0001] The present invention relates to the technical field of fault diagnosis of solenoid valves, and more particularly to an online fault diagnosis method for solenoid valves based on characteristic analysis in the time-frequency domain. [Background technology]

[0002] Due to the advantages of simple structure, low cost, low pollution, fast response speed, etc., solenoid valves are widely used in electrohydraulic control devices in important fields such as shipbuilding, petrochemical industry, national defense, and industrial equipment. However, due to factors such as long-term wear, corrosion, and electrolysis in the usage environment, solenoid valves are very prone to failure. This will seriously affect the performance of the control system, and even cause the system to shut down, endangering the normal operation of the system. Therefore, in order to improve the safety and reliability of the system, it is necessary to diagnose the failure of the solenoid valve online so that the failure can be found and corrected in a timely and accurate manner.

[0003] At present, the fault diagnosis method of the solenoid valve is mainly divided into a model-driven method and a data-driven method. The model-driven method mainly establishes the mapping relationship between the fault type of the solenoid valve and the solenoid valve model parameters through simulation, testing or actual engineering experience, and then performs fault diagnosis based on the mapping relationship. However, this method does not form a specific fault diagnosis standard and is greatly influenced by human subjective factors. The data-driven method extracts feature vectors based on multiple measurable data processing, and finally uses algorithms to realize fault diagnosis. For example, Kaku Seihashi et al. use sensors to measure the load pressure of the solenoid valve, the flow rate of the inlet and outlet oil ports, and the opening value, extract feature vectors based on a physical model, and finally perform online fault diagnosis based on a support vector machine (SVM). However, the method is complicated to operate, difficult to obtain the original data, and unlikely to be engineered. In fact, current signal measurement has the advantages of being low cost, providing a large amount of information, and being easy to obtain, and can also be used for fault diagnosis of electromagnetic directional control valves.

[0004] The existing invention patent application in Patent Document 1 discloses a solenoid valve fault diagnosis method based on current detection, which measures the front end driving current of the solenoid valve based on a sensor, performs wavelet energy decomposition to extract feature vectors, and finally performs fault diagnosis based on a BP neural network.

[0005] From the method described in the above Patent Document 1, it can be seen that the limitations of the prior art include at least the following problems. First, only wavelet packet decomposition is used to process the signal, which makes it less reliable. In addition, the reliability of fault diagnosis based on BP neural network depends heavily on the selection of initial parameters, which makes it difficult to apply in real engineering. Therefore, it is urgent to propose a method for quickly diagnosing the fault type of the solenoid valve online to ensure the safety and reliability of the system. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] China Patent Application Publication No. 103336189 Summary of the Invention [Problem to be solved by the invention]

[0007] The present invention has been made in consideration of the shortcomings of the prior art described above, and provides an online fault diagnosis method for solenoid valves based on characteristic analysis in the time-frequency domain, which solves the problems that only wavelet packet decomposition is used to process signals, resulting in low reliability, and further, the reliability of fault diagnosis based on BP neural network depends heavily on the selection of initial parameters, making it difficult to apply to actual engineering. [Means for solving the problem]

[0008] In order to achieve the above objectives, the present invention is achieved by the following technical solutions: A method for online diagnosis of a fault in a solenoid valve based on time-frequency domain characteristic analysis, comprising: A step of obtaining current time-series current data on the solenoid valve drive side, valve core normal time-series history current data, spring break time-series history current data, valve core non-operation time-series history current data, and valve core slightly stuck time-series history current data, and performing pre-processing on the data; A step of performing a time domain analysis and a frequency domain analysis on the preprocessed current time series current data of the solenoid valve drive side, the valve core normal time series history current data, the spring broken time series history current data, the valve core non-operation time series history current data, and the valve core slightly stuck time series history current data, thereby obtaining a current time domain feature vector and a frequency domain feature vector, a valve core normal time domain feature vector and a frequency domain feature vector, a spring broken time domain feature vector and a frequency domain feature vector, a valve core non-operation time domain feature vector and a frequency domain feature vector, and a valve core slightly stuck time domain feature vector and a frequency domain feature vector; performing fusion processes on the current time domain feature vector and frequency domain feature vector, the time domain feature vector and frequency domain feature vector of the normal valve core, the time domain feature vector and frequency domain feature vector of the broken spring, the time domain feature vector and frequency domain feature vector of the non-operating valve core, and the time domain feature vector and frequency domain feature vector of the slightly stuck valve core, to obtain a current composite feature vector, a composite feature vector of the normal valve core, a composite feature vector of the broken spring, a composite feature vector of the non-operating valve core, and a composite feature vector of the slightly stuck valve core; The method includes creating a vector data set based on the composite feature vector of a valve core normal, the composite feature vector of a spring broken, the composite feature vector of a valve core non-operating, and the composite feature vector of a valve core slightly stuck; using the vector data set to perform model training on a pre-created random forest model, and performing diagnostic identification of the current composite feature vector after training of the random forest model is completed.

[0009] Furthermore, the current time-series current data on the solenoid valve driving side is specifically current values ​​at multiple time points when the solenoid valve driving side is in a current state, the valve core normal time-series history current data on the solenoid valve driving side is specifically history current values ​​at multiple historical time points when the solenoid valve driving side is in a valve core normal state, the spring breakage time-series history current data on the solenoid valve driving side is specifically history current values ​​at multiple historical time points when the solenoid valve driving side is in a spring breakage state, the valve core non-operation time-series history current data on the solenoid valve driving side is specifically history current values ​​at multiple historical time points when the solenoid valve driving side is in a valve core non-operational state, and the valve core light sticking time-series history current data on the solenoid valve driving side is specifically history current values ​​at multiple historical time points when the solenoid valve driving side is in a valve core light sticking state.

[0010] In addition, the specific steps of obtaining the current time domain feature vector, the time domain feature vector of the valve core being normal, the time domain feature vector of the spring breaking, the time domain feature vector of the valve core not working, and the time domain feature vector of the valve core being slightly stuck are as follows: The current values ​​at multiple points in time when the preprocessed solenoid valve driving side is in the current state, the historical current values ​​at multiple historical points in time when the valve core is in a normal state, the historical current values ​​at multiple historical points in time when the spring is in a broken state, the historical current values ​​at multiple historical points in time when the valve core is in a non-operating state, and the historical current values ​​at multiple historical points in time when the valve core is in a slightly stuck state are read, and average value analysis and descending sorting are performed on them, to obtain the average current value in the current state, the effective current value in the current state, the maximum current value in the current state, the minimum current value in the current state, the average current value in the normal state of the valve core, the historical current value in the normal state of the valve core, obtaining an effective current value in a state where the valve core is in a normal state, a maximum current value in a normal state of the valve core, a minimum current value in a normal state of the valve core, an average current value in a spring broken state of the valve core, an effective current value in a normal state of the valve core, a maximum current value in a normal state of the valve core, a minimum current value in a normal state of the valve core, an average current value in a non-operating state of the valve core, an effective current value in a non-operating state of the valve core, a maximum current value in a non-operating state of the valve core, a minimum current value in a non-operating state of the valve core, and an average current value in a slightly stuck state of the valve core, an effective current value in a slightly stuck state of the valve core, a maximum current value in a slightly stuck state of the valve core, and a minimum current value in a slightly stuck state of the valve core; For the maximum current value and minimum current value in the current state, the maximum current value and minimum current value in the normal state of the valve core, the maximum current value and minimum current value in the spring break state, the maximum current value and minimum current value in the non-operating state of the valve core, and the maximum current value and minimum current value in the light sticking state of the valve core, by performing differential analysis respectively, steps of obtaining the current range value in the current state, the current range value in the normal state of the valve core, the current range value in the spring break state, the current range value in the non-operating state of the valve core, and the current range value in the light sticking state of the valve core; and based on the current average value in the current state, the current effective value in the current state, the current range value in the current state, the current average value in the normal state of the valve core, the current effective value in the normal state of the valve core, the current range value in the normal state of the valve core, the current average value in the spring break state, the current effective value in the spring break state, the current range value in the spring break state, the current average value in the non-operating state of the valve core, the current effective value in the non-operating state of the valve core, the current range value in the non-operating state of the valve core, and the current average value in the light sticking state of the valve core, the current effective value in the light sticking state of the valve core, the current range value in the light sticking state of the valve core, steps of respectively creating the feature vector in the current time domain, the feature vector in the time domain of the normal valve core, the feature vector in the time domain of the spring break, the feature vector in the time domain of the non-operating valve core, and the feature vector in the time domain of the light sticking valve core; including.

[0011] Furthermore, the specific formulas for calculating the current average value, the current effective value, and the current range value in the current state are as follows:

Number

[0012] In addition, the specific steps of obtaining the current frequency domain feature vector, the frequency domain feature vector of the normal valve core, the frequency domain feature vector of the spring break, the frequency domain feature vector of the non-operating valve core, and the frequency domain feature vector of the slightly stuck valve core are as follows: reading current values ​​at multiple time points when the pre-processed solenoid valve driving side is in a current state, historical current values ​​at multiple historical time points when the valve core is in a normal state, historical current values ​​at multiple historical time points when the spring is broken, historical current values ​​at multiple historical time points when the valve core is in a non-operating state, and historical current values ​​at multiple historical time points when the valve core is in a slightly stuck state, and performing second derivative analysis on them, thereby obtaining a second derivative of the current state, a second derivative of the valve core normal state, a second derivative of the spring broken state, a second derivative of the valve core in a non-operating state, and a second derivative of the valve core in a slightly stuck state; Performing binary discrete wavelet packet decomposition on the second derivative of the current state, the second derivative of the valve core normal state, the second derivative of the spring broken state, the second derivative of the valve core non-operating state, and the second derivative of the valve core slightly stuck state, respectively, to obtain a wavelet function of the current state, a wavelet function of the valve core normal state, a wavelet function of the spring broken state, a wavelet function of the valve core non-operating state, and a wavelet function of the valve core slightly stuck state; performing energy analysis on the wavelet function of the current state, the wavelet function of the valve core normal state, the wavelet function of the spring broken state, the wavelet function of the valve core non-operating state, and the wavelet function of the valve core slightly stuck state, to obtain an energy value of the current state, an energy value of the valve core normal state, an energy value of the spring broken state, an energy value of the valve core non-operating state, and an energy value of the valve core slightly stuck state; and a step of performing normalization processing on the energy value of the current state, the energy value of the valve core normal state, the energy value of the spring broken state, the energy value of the valve core not working state, and the energy value of the valve core slightly stuck state to obtain normalized energy features of the current state, the normalized energy features of the valve core normal state, the normalized energy features of the spring broken state, the normalized energy features of the valve core not working state, and the normalized energy features of the valve core slightly stuck state, and recording them as a current frequency domain feature vector, a frequency domain feature vector of the valve core normal, a frequency domain feature vector of the spring broken, a frequency domain feature vector of the valve core not working, and a frequency domain feature vector of the valve core slightly stuck, respectively.

[0013] In addition, the specific formulas for calculating the wavelet function of the current state, the energy value of the current state, and the normalized energy feature of the current state are as follows:

number

[0014] Further, the vector dataset is used to perform model training on a pre-created random forest model, specifically, the training is performed using a C4.5 algorithm, specifically, the vector dataset is sampled to generate multiple decision trees to construct the random forest model.

[0015] In addition, the specific steps of generating multiple decision trees include: reading the vector data set, randomly extracting multiple groups of data, and separately creating random vector data sets; respectively dividing the random vector data set of each group of data into a vector training data set and a vector validation data set; starting from the root node of the vector training data set for each group of data, dividing it using the feature with the highest information gain rate, and repeating this division process for each subset generated by the division to recursively build a tree; stopping the tree construction operation when a preset stopping condition is reached; adopting a post-pruning strategy for the constructed tree, which is performed after the tree is completely generated, and using the vector validation data set to evaluate and adjust the tree before and after pruning until the tree meets a predetermined criterion.

[0016] Furthermore, in the diagnosis identification of the current composite feature vector after the training of the random forest model is completed, the means of fault diagnosis by the random forest model uses the majority voting method to obtain the final result. Effect of the Invention

[0017] The present invention has the following beneficial effects: (1) The online fault diagnosis method for solenoid valves based on the characteristic analysis of the time-frequency domain can more accurately identify and distinguish different fault types of solenoid valves by integrating the current data of multiple states and the integrated feature vector obtained by time-frequency domain analysis, which not only improves the feature characterization ability, but also improves the accuracy and reliability of fault diagnosis by comprehensively utilizing the information in the time domain and the frequency domain. (2) The online fault diagnosis method for solenoid valves based on the characteristic analysis of the time-frequency domain monitors and analyzes the current characteristics of the solenoid valve in real time, and uses an efficient random forest model to perform online fault diagnosis, so that the method can provide immediate fault warnings and maintenance advice, not only to resolve the fault in a timely manner and reduce the equipment downtime, but also to prevent more serious equipment damage, thereby reducing maintenance costs and extending the life of the equipment. (3) The online diagnosis method for solenoid valve faults based on the time-frequency domain characteristic analysis is trained using a random forest model based on the C4.5 algorithm, and combined with the strategy of post-hoc pruning, overfitting is effectively avoided and the generalization ability of the model is improved. By integrating multiple decision trees, the random forest reduces the variance of a single decision tree, and the overall judgment accuracy is improved through the majority voting method. This allows the model to maintain efficient diagnosis performance even in the face of changes in the working environment and various types of faults. Of course, a product embodying the present invention need not necessarily achieve all of the above advantages simultaneously. [Brief description of the drawings]

[0018] [Figure 1] 2 is a flowchart of an online fault diagnosis method for a solenoid valve based on time-frequency domain characteristic analysis according to the present invention; [Diagram 2] 4 is a flowchart illustrating an example of a method for online diagnosis of a fault in a solenoid valve based on a characteristic analysis in a time-frequency domain according to the present invention; [Diagram 3] FIG. 13 is a diagram illustrating a random forest training and evaluation method for the online diagnosis method for faults of a solenoid valve based on the time-frequency domain characteristics of multi-dimensional driving side current according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] The embodiments of the present application use an online fault diagnosis method for solenoid valves based on characteristic analysis in the time-frequency domain to solve the problem that only wavelet packet decomposition is used to process signals, resulting in low reliability, and the reliability of fault diagnosis based on BP neural network depends heavily on the selection of initial parameters, making it difficult to apply to practical engineering.

[0020] The overall approach to the problem in the embodiment of this application is as follows.

[0021] First, collect time-series current data of the solenoid valve in various operating states, including normal, spring broken, valve core non-operating, and slightly stuck, in the current state and historical state, preprocess, clean and format these data to ensure that the data quality meets the analysis needs, and perform time domain and frequency domain analysis on the preprocessed data to extract key features such as the average value, effective value, maximum value, minimum value, and range value of the current. Then, the extracted time domain features and frequency domain features are fused to form an integrated feature vector, and these composite feature vectors are used to construct a vector data set, and a random forest model is applied to perform model training based on the set of decision trees constructed by the C4.5 algorithm. After training is completed, the trained model is used to perform fault diagnosis on the new feature vector or the current feature vector. The processes in the embodiments may be realized, for example, by a processor included in the processing device executing a program stored in a storage device. Alternatively, the processes in the embodiments may be executed by a hardware circuit included in the processing device.

[0022] Referring to Fig. 1, an embodiment of the present invention provides a technical solution. A method for online diagnosis of faults of a solenoid valve based on characteristic analysis in the time-frequency domain, which includes the following steps:Obtain the current time series current data of the solenoid valve driving side, the valve core normal time series history current data, the spring broken time series history current data, the valve core non-operation time series history current data, and the valve core lightly stuck time series history current data, and perform pre-processing on them, the valve core normal time series history current data is the current measurement data in the normal operating state of the solenoid valve valve core set by the operator, the spring broken time series history current data is the current measurement data in the solenoid valve spring broken operating state set by the operator, and the history valve core non-operation time series current data is the current measurement data in the solenoid valve valve core lightly stuck operating state set by the operator. The historical valve core lightly stuck time series current data is current measurement data in a solenoid valve core lightly stuck operating state set by an operator, and the preprocessed current time series current data of the solenoid valve driving side, the valve core normal time series historical current data, the spring break time series historical current data, the valve core non-operating time series historical current data, and the valve core lightly stuck time series historical current data are respectively analyzed in the time domain and the frequency domain to obtain feature vectors of the current time domain and the frequency domain, feature vectors of the valve core normal time domain and the frequency domain The time domain and frequency domain feature vector of the spring broken, the time domain and frequency domain feature vector of the valve core not working, and the time domain and frequency domain feature vector of the valve core slightly stuck are obtained, and the current time domain and frequency domain feature vector, the time domain and frequency domain feature vector of the valve core normal, the time domain and frequency domain feature vector of the spring broken, the time domain and frequency domain feature vector of the valve core not working, and the time domain and frequency domain feature vector of the valve core slightly stuck are respectively fusion processed to obtain a current composite feature vector, a composite feature vector of the valve core normal, the composite feature vector of the spring broken, the valve core not working, and the composite feature vector of the valve core slightly stuck. A vector data set is created according to the composite feature vector of the valve core normal, the composite feature vector of the spring broken, the valve core not working, and the composite feature vector of the valve core slightly stuck, and the vector data set is used to perform model training on the pre-created random forest model, and after the training of the random forest model is completed, a diagnosis identification of the current composite feature vector is performed.

[0023] The current time-series current data on the solenoid valve driving side is specifically current values ​​at multiple points in time when the solenoid valve driving side is in the current state, obtained by a current sensor installed in the solenoid valve driving circuit. The valve core normal time-series historical current data on the solenoid valve driving side is specifically historical current values ​​at multiple points in time when the solenoid valve driving side is in a valve core normal state, obtained by a current sensor installed in the solenoid valve driving circuit. The spring breakage time-series historical current data on the solenoid valve driving side is specifically historical current values ​​at multiple historical points when the solenoid valve driving side is in a spring breakage state, obtained by a current sensor installed in the solenoid valve driving circuit. The time-series historical current data of the valve core inoperative on the solenoid valve driving side is specifically historical current values ​​at multiple historical points in time when the solenoid valve driving side is in a valve core inoperative state, which is acquired by a current sensor installed in the solenoid valve driving circuit. The historical valve core light sticking time-series current data on the solenoid valve driving side is, specifically, historical current values ​​at multiple historical time points when the solenoid valve driving side is in a valve core light sticking state, which is acquired by a current sensor installed in the solenoid valve driving circuit.

[0024] Specifically, the specific steps of obtaining the feature vector of the current time domain, the feature vector of the time domain of the valve core being normal, the feature vector of the time domain of the spring broken, the feature vector of the valve core not working, and the feature vector of the time domain of the valve core being slightly stuck are as follows: current values ​​at multiple time points when the preprocessed solenoid valve driving side is in the current state, and historical current values ​​at multiple historical time points when the valve core is in the normal state, the spring broken state, the valve core not working, and the valve core is in the slightly stuck state are read, and average value analysis and descending sorting are performed on them, respectively, to obtain the average current value, effective current value, maximum current value, and minimum current value of the current state, and the average current value, effective current value, maximum current value, and minimum current value of the valve core being normal, the spring broken, the valve core not working, and the valve core being slightly stuck states. the step of obtaining a current range value for the current state, and a current range value for the normal valve core state, the broken spring state, the broken valve core state, and the slightly stuck valve core state, by performing a difference analysis on the maximum and minimum current values ​​in the current state and the maximum and minimum current values ​​in the normal valve core state, the broken spring state, the broken valve core state, and the slightly stuck valve core state; and the step of creating a feature vector of the current time domain, a feature vector of the normal valve core time domain, a feature vector of the broken spring time domain, a feature vector of the stuck valve core time domain, and a feature vector of the slightly stuck valve core time domain, based on the average current value, effective current value, and current range value of the current state, and the average current value, effective current value, and current range value of the normal valve core state, the broken spring state, the broken valve core state, and the slightly stuck valve core state, respectively.

[0025] The specific formulas for calculating the current average value, current effective value, and current range value in the current state are as follows:

number

[0026] Furthermore, the specific formulas for calculating the average current, effective current, and current range values ​​in the valve core normal state, spring broken state, valve core non-operating state, and valve core slightly stuck state are similar to the specific formulas for calculating the average current, effective current, and current range values ​​in the current state.

[0027] In this embodiment, the current data of the solenoid valve in various states (normal, spring broken, valve core not working, etc.) can be analyzed in detail to accurately identify abnormal current patterns. For example, the average, effective, maximum, minimum, and range values ​​of the current can be compared to effectively identify possible problems during the operation of the equipment. These indicators can reflect whether the operation status of the solenoid valve is abnormal, so that early detection of failures and possible equipment failures can be prevented. A detailed fault feature database can be built by analyzing the current feature vectors of various fault states (such as normal operation and various fault states). Such a database can not only help to accurately diagnose the type of current failure, but also identify the specific cause of the failure, such as whether the valve core is not working or whether the spring is broken, through comparative analysis. The accuracy of this diagnosis is very important for subsequent maintenance work, which can reduce purposeless inspections and unnecessary maintenance costs, and the data provided by the time domain analysis can be used to build and optimize predictive maintenance models. By tracking the current data of the solenoid valve in various states over the long term, the future maintenance needs of the equipment can be predicted and the maintenance plan and resource allocation can be optimized. For example, if data shows that a particular pattern of electrical current changes consistently causes equipment failure, preventive maintenance can be performed before a problem occurs, avoiding costly emergency repairs and production interruptions.

[0028] Specifically, the specific steps of obtaining the current frequency domain feature vector, the valve core normal frequency domain feature vector, the spring broken frequency domain feature vector, the valve core non-operating frequency domain feature vector, and the valve core slightly stuck frequency domain feature vector include the steps of reading current values ​​at multiple time points when the pre-processed solenoid valve driving side is in the current state, and historical current values ​​at multiple historical time points when the valve core is in the normal state, the spring broken state, the valve core non-operating state, and the valve core slightly stuck state, respectively, and performing second derivative analysis on them to obtain second derivatives of the current state, the valve core normal state, the spring broken state, the valve core non-operating state, and the valve core slightly stuck state; and performing binary discrete wavelet packet decomposition on the second derivatives of the current state, the valve core normal state, the spring broken state, the valve core non-operating state, and the valve core slightly stuck state, respectively, to obtain second derivatives of the current state, the valve core normal state, the spring broken state, the valve core non-operating state, and the valve core slightly stuck state. the step of obtaining wavelet functions of the valve core not working state and the valve core slightly stuck state; the step of obtaining energy values ​​of the current state, the valve core normal state, the spring broken state, the valve core not working state and the valve core slightly stuck state by performing energy analysis on the wavelet functions of the current state, the valve core normal state, the spring broken state, the valve core not working state and the valve core slightly stuck state respectively; and the step of obtaining normalized energy features of the current state, the valve core normal state, the spring broken state, the valve core not working state and the valve core slightly stuck state by performing normalization processing on the energy values ​​of the current state, the valve core normal state, the spring broken state, the valve core not working state and the valve core slightly stuck state respectively, and recording them as a current frequency domain feature vector, a normal valve core frequency domain feature vector, a broken spring frequency domain feature vector, a valve core not working frequency domain feature vector and a slightly stuck valve core frequency domain feature vector respectively.

[0029] The specific formulas for calculating the wavelet function, energy value, and normalized energy feature of the current state are as follows:

number

[0030] The specific formulas for calculating the wavelet function, energy value, and normalized energy feature of the valve core normal state, the wavelet function energy value and normalized energy feature of the spring broken state, the wavelet function energy value and normalized energy feature of the valve core non-operating state, and the wavelet function energy value and normalized energy feature of the valve core slightly stuck state are similar to the specific formulas for calculating the wavelet function, energy value, and normalized energy feature of the current state.

[0031] In this embodiment, by performing wavelet transform and second derivative analysis on the solenoid valve current data, it is possible to effectively capture the instantaneous changes and nonlinear characteristics of the current signal, which are difficult to identify by ordinary time series analysis. Wavelet transform can partially analyze the current signal at different time scales, so that even small changes can be detected, and the sensitivity and accuracy of fault detection can be greatly improved. By analyzing the feature vectors in the frequency domain under different fault conditions, this method can accurately distinguish various types of faults, such as valve core normal, spring broken, valve core non-operation and light sticking. Each state has its own frequency domain characteristics (e.g., changes in energy distribution and wavelet coefficients), and such detailed frequency domain information can help the maintenance team to more accurately diagnose certain types of faults, so that more accurate maintenance measures can be taken. By continuously monitoring the current characteristics of the solenoid valve and analyzing its time-frequency domain data, it is possible to grasp the operating state of the equipment in real time and timely discover changes in trends that lead to failure. This method is not only useful for real-time fault diagnosis, but also can predict the maintenance needs of the solenoid valve through long-term data analysis and realize preventive maintenance. Preventive maintenance helps avoid catastrophic failures, reduces unscheduled downtime and improves overall equipment reliability and efficiency.

[0032] Specifically, the vector dataset is used to perform model training on a pre-built random forest model, specifically, the training is performed using a C4.5 algorithm, specifically, the vector dataset is sampled to generate multiple decision trees, and the random forest model is constructed.

[0033] In this embodiment, the random forest improves the accuracy and reliability of prediction by integrating multiple decision trees. Each tree is trained on a random subset of the data set, and this integration method can significantly reduce the variance of the model, thereby avoiding overfitting. The information gain rate method used in the C4.5 algorithm when building each tree allows the splitting attributes to be selected more effectively, which further optimizes the performance of each tree and the stability of the entire model. The C4.5 algorithm natively supports the processing of continuous attributes and can automatically find the optimal binary split point for continuous attributes. Since the current data is usually continuous, the C4.5 algorithm is particularly useful for fault diagnosis of solenoid valves. In addition, C4.5 does not require prior data conversion when processing classification data, and can directly learn from the original data, reducing the complexity of data preprocessing. The flexibility of the random forest model allows the random forest model to adapt to variable input characteristics and provide reliable diagnosis results for various types of faults. The C4.5 algorithm optimizes the feature selection process so that the model can more accurately identify the key factors that lead to failure and maintain a high level of performance even in complex industrial environments. By using information gain rate instead of simple information gain to select split nodes, the C4.5 algorithm avoids the tendency of decision trees to select more categorical features, making the model more fair and balanced. This is particularly important in fault diagnosis, as it ensures that the model does not miss signs of potential failures due to feature diversity.

[0034] Specifically, the specific steps of generating multiple decision trees include the steps of reading a vector data set, randomly extracting multiple groups of data, creating random vector data sets separately, and splitting the random vector data set of each group of data into a vector training data set and a vector validation data set respectively; and for each group of data, starting from the root node of the vector training data set, splitting using the feature with the highest information gain rate, and repeating this splitting process for each subset generated by the splitting, and recursively building a tree. In C4.5, the criterion for selecting split features does not use pure information gain (used by the ID3 algorithm), but uses the information gain rate. The information gain rate takes into account the impurity of the features and helps to avoid the tendency of more categorical features to be selected. Here, the information gain calculates the change in entropy before and after the difference, and the inherent value takes into account the weighted sum of the size of each subset after splitting, and stops the construction operation when a pre-set stopping condition is reached. The stopping condition is to stop when all instances belong to the same category, or to stop when there is no more information gain, or to stop when a pre-set maximum depth of the tree or minimum number of samples of a node is reached. A post-pruning strategy is adopted for the constructed tree, which is performed after the tree is fully generated, and the vector validation data set is used to evaluate and adjust the tree before and after pruning until the tree meets a predetermined criterion. Specifically, statistical indicators such as accuracy, recall, precision, etc. are calculated, and parts that may improve the performance of the validation set after pruning are pruned until the tree meets the predetermined criterion, which is helpful in reducing overfitting and improving the generalization ability of the model.

[0035] In this embodiment, by randomly extracting subsets of data to construct multiple decision trees, the random forest model can greatly reduce the risk of overfitting and improve the performance of the model on unknown data. In the C4.5 algorithm, the information gain rate is taken into account when selecting features for splitting, which not only optimizes the quality of splitting, but also reduces the deviation caused by multiple feature values, making the construction of each tree more reasonable and improving the accuracy and robustness of the overall model. The use of the C4.5 algorithm, especially the introduction of the information gain rate in feature selection, helps the model to more accurately identify the features that are most useful for classification, and avoids the selection variation that may be caused by simple information gain. In addition, the information gain rate takes into account the size of the subsets after splitting, which helps to generate a more balanced tree structure, which is especially important when dealing with complex data sets, and by setting various stopping conditions (such as maximum depth, lower limit of the number of samples, reaching the information gain threshold, etc.), the tree growth can be effectively controlled to prevent the tree structure from becoming too complex, and overfitting can be avoided. Through the strategy of post-pruning, the tree model can be further optimized. By pruning the fully grown tree, nodes that contribute little or no contribution to the improvement of the model's performance can be removed, ensuring the simplicity and execution efficiency of the final model. In the strategy of post-pruning, the model first forms a decision basis based on a larger data frame, and then pruning can be used to remove unnecessary parts, which not only improves the efficiency of the application of the model to new data, but also optimizes the execution time and resource usage of the model, making the model more efficient in practical applications. In the application scenario of solenoid valve fault diagnosis, this method can accurately identify the specific fault type of the equipment based on the comprehensive analysis of historical data and real-time data, which is very important for formulating timely and effective maintenance measures and reducing equipment downtime and maintenance costs.

[0036] Specifically, in the diagnosis identification of the current composite feature vector after the training of the random forest model is completed, the means of fault diagnosis by the random forest model uses the majority voting method to obtain the final result.

[0037] In this embodiment, the majority voting method integrates the prediction results of each tree to determine the final classification output. Each tree is trained independently on a different random subset of the data set, so different prediction results may be obtained for the same input. The model leverages collective intelligence through majority voting, and even if some trees have errors in their predictions, the correct predictions of other trees compensate for these errors, improving the overall prediction accuracy. A single decision tree is easily affected by noise in the data and prone to overfitting, especially when the tree depth is deep. On the other hand, random forest effectively distributes the risk of overfitting by building multiple trees and using a voting mechanism. Multiple trees cover various data perspectives and reduce the sensitivity of the model to certain data samples, making the model's performance on unknown data more stable, and when faced with outliers or changes in feature distribution in the data set, random forests exhibit high robustness due to their integrated characteristics. The majority voting method ensures that even if some trees are affected by abnormal data, the output of the entire model is still determined by the majority of trees, thus maintaining the consistency and reliability of the prediction results, and random forests can handle data sets with high-dimensional features without the need for dimensionality reduction. Each tree in the model can be trained from a random subset of features in the data, which not only improves the model's ability to handle complex data, but also allows the model to capture complex interactions between features, which is very difficult to achieve with a single tree or simple model.

[0038] A specific example of an online diagnosis method for solenoid valve faults based on time-frequency domain characteristic analysis is as follows:

[0039] Specifically, see Figures 2 and 3.

[0040] In step 1, four typical solenoid valve operating states are set: valve core normal, spring broken, valve core not working, and valve core slightly stuck. 0 , c 1 , c2 , c 3 In order to ensure that the solenoid valve operates normally, a solenoid valve operation circuit for normal operation is set, a non-contact current sensor is connected in series to the driving side, and a sampling period T s , total acquisition time nT s Then, the time series current data i(t) = [i 1 ,i 2 ,···,i k ,···i n ] Collect.

[0041] In step 2, the average value of the time series current data i(t) is calculated. TIFF0007662921000008.tif85, rms i k , range R and first inflection time t are calculated to generate time series feature I, where

number

[0042] Taking a certain data set as an example, the final time-series feature vector I is obtained as follows:

number

[0043] In step 3, we choose N=3 and perform a binary discrete wavelet packet decomposition on the second derivative of the current Δi″(t) and calculate the energy as a frequency domain feature E according to the following equation:

number

[0044] In step 4, we combine the frequency domain features E and the time domain features I into a composite feature vector T 0 and record multiple groups.

number

[0045] In step 5, three typical fault conditions, namely, spring break, valve core stuck, and valve core slightly stuck, are simulated to obtain the composite feature vector T 1 , T 2 , T 3 is formed again and recorded.

[0046] In step 5.1, we simulate the spring breakage fault condition and obtain multiple sets of composite feature vectors T 1 Record the following. In step 5.2, the valve core non-operating fault condition is simulated and multiple sets of composite feature vectors T 2 Record the following. In step 5.3, the valve core sticking fault condition is simulated and multiple sets of composite feature vectors T 3 Record the following.

[0047] In step 6, we create a composite feature vector with labels for multiple groups. <T i ,C i >(i=0,1,2,3) is used as the training set DATA.

number

[0048] For ease of explanation, the following assumptions are made.

number

[0049] In step 7, we generate a random forest model (RF). In step 7.1, the training data set is sampled. For example, the training set DATA contains R groups of data. Each time, one group of data is sampled with replacement, and the process is repeated k times to generate k decision trees. The training set Train sampled in a certain time is i is shown as follows:

number

[0050] In step 7.2, a decision tree model is created. The present invention uses the C4.5 algorithm to build a decision tree unit in the random forest model. The C4.5 decision tree selects split attributes based on the amount of information gain, and i The decision trees corresponding to,generate a total of k decision trees.

[0051] In step 7.4, we build a random forest classification model. The application scenario is a 4-classification problem, and we can use majority voting to get the final prediction result. The core idea is to vote for the prediction results of all k decision trees in the random forest model, and the category that gets the most votes becomes the final output category.

[0052] In step 8, the trained random forest model is applied to perform online diagnosis of the failure mode of the solenoid valve, that is, the composite feature vector of the solenoid valve is taken as input and introduced into the random forest model, and all decision trees in the model are obtained, and the most frequent number of the operating state is determined and output as its operating state, and it is assumed that the input composite feature vector is as follows:

number

number

[0053] From the above, the present application has at least the following effects. By integrating the current data of multiple states and the integrated feature vector analyzed using the time-frequency domain, different fault types of the solenoid valve can be more accurately identified and distinguished, not only improving the feature characterization ability, but also improving the accuracy and reliability of fault diagnosis through the comprehensive utilization of the time domain and frequency domain information. By monitoring and analyzing the current characteristics of the solenoid valve in real time and using an efficient random forest model for online fault diagnosis, the method can provide immediate fault warning and maintenance advice, not only to resolve the fault in a timely manner and reduce the equipment downtime, but also to prevent more serious equipment damage, thereby reducing maintenance costs and extending the equipment life. Training is performed using a random forest model based on the C4.5 algorithm, and combined with the strategy of post-hoc pruning, overfitting is effectively avoided and the generalization ability of the model is improved. Random forests can reduce the variance of a single decision tree by integrating multiple decision trees, and the overall judgment accuracy is improved through the majority voting method. This allows the model to maintain efficient diagnosis performance even in the face of changes in the working environment and various types of faults.

[0054] Although the preferred embodiments of the present invention have been described, further variations and modifications may be made to these embodiments once the basic inventive concept is apparent to those skilled in the art. It is therefore intended that the appended claims be interpreted as including the preferred embodiments and all such variations and modifications that are within the scope of the present invention.

[0055] It is obvious that those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these changes and modifications of the present invention fall within the scope of the claims of the present invention and its equivalent technology, the present invention also includes these changes and modifications.

Claims

1. A method for online diagnosis of a fault in a solenoid valve based on time-frequency domain characteristic analysis, comprising: A step of acquiring current time-series current data of the solenoid valve drive side, valve core normal time-series history current data, spring break time-series history current data, valve core non-operation time-series history current data, and valve core slightly stuck time-series history current data, and performing pre-processing on the data; A step of performing a time domain analysis and a frequency domain analysis on the preprocessed current time series current data of the solenoid valve drive side, the valve core normal time series history current data, the spring broken time series history current data, the valve core non-operation time series history current data, and the valve core slightly stuck time series history current data, thereby obtaining a current time domain feature vector and a frequency domain feature vector, a valve core normal time domain feature vector and a frequency domain feature vector, a spring broken time domain feature vector and a frequency domain feature vector, a valve core non-operation time domain feature vector and a frequency domain feature vector, and a valve core slightly stuck time domain feature vector and a frequency domain feature vector; performing fusion processes on the current time domain feature vector and frequency domain feature vector, the time domain feature vector and frequency domain feature vector of the normal valve core, the time domain feature vector and frequency domain feature vector of the broken spring, the time domain feature vector and frequency domain feature vector of the non-operating valve core, and the time domain feature vector and frequency domain feature vector of the slightly stuck valve core, to obtain a current composite feature vector, a composite feature vector of the normal valve core, a composite feature vector of the broken spring, a composite feature vector of the non-operating valve core, and a composite feature vector of the slightly stuck valve core; Creating a vector data set based on the composite feature vector of the valve core being normal, the composite feature vector of the spring being broken, the composite feature vector of the valve core not working, and the composite feature vector of the valve core being slightly stuck; performing model training on a pre-created random forest model using the vector dataset, and performing diagnostic identification of the current composite feature vector after the random forest model training is completed; Including, The present invention relates to an online diagnosis method for faults in a solenoid valve based on characteristic analysis in the time-frequency domain.

2. The present time-series current data of the solenoid valve driving side is specifically current values ​​at multiple time points when the solenoid valve driving side is in a current state, the valve core normal time-series historical current data of the solenoid valve driving side is specifically historical current values ​​at multiple time points when the solenoid valve driving side is in a valve core normal state, the spring breakage time-series historical current data of the solenoid valve driving side is specifically historical current values ​​at multiple time points when the solenoid valve driving side is in a spring breakage state, the valve core non-operation time-series historical current data of the solenoid valve driving side is specifically historical current values ​​at multiple time points when the solenoid valve driving side is in a valve core non-operation state, and the valve core mild sticking time-series historical current data of the solenoid valve driving side is specifically historical current values ​​at multiple time points when the solenoid valve driving side is in a valve core mild sticking state.

2. The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 1.

3. The specific steps of obtaining the current time domain feature vector, the time domain feature vector of the valve core being normal, the time domain feature vector of the spring breaking, the time domain feature vector of the valve core not working, and the time domain feature vector of the valve core being slightly stuck are as follows: The current values ​​at multiple points in time when the preprocessed solenoid valve drive side is in the current state, the historical current values ​​at multiple historical points in time when the valve core is in a normal state, the historical current values ​​at multiple historical points in time when the spring is in a broken state, the historical current values ​​at multiple historical points in time when the valve core is in a non-operating state, and the historical current values ​​at multiple historical points in time when the valve core is in a slightly stuck state are read, and average value analysis and descending sorting are performed on these, to obtain the average current value in the current state, the effective current value in the current state, the maximum current value in the current state, the minimum current value in the current state, and the average current value in the valve core normal state. obtaining a current value when the valve core is in a normal state, an effective current value when the valve core is in a normal state, a maximum current value when the valve core is in a normal state, a minimum current value when the valve core is in a normal state, an average current value when the spring is broken, an effective current value when the spring is broken, a maximum current value when the spring is broken, a minimum current value when the spring is broken, an average current value when the valve core is in an inoperable state, an effective current value when the valve core is in an inoperable state, a maximum current value when the valve core is in an inoperable state, a minimum current value when the valve core is in an inoperable state, and an average current value when the valve core is in a slightly stuck state, an effective current value when the valve core is in a slightly stuck state, a maximum current value when the valve core is in a slightly stuck state, and a minimum current value when the valve core is in a slightly stuck state; a step of performing a difference analysis on the maximum and minimum current values ​​in the current state, the maximum and minimum current values ​​in the valve core normal state, the maximum and minimum current values ​​in the spring broken state, the maximum and minimum current values ​​in the valve core non-operating state, and the maximum and minimum current values ​​in the valve core slightly stuck state, thereby obtaining a current range value in the current state, a current range value in the valve core normal state, a current range value in the spring broken state, a current range value in the valve core non-operating state, and a current range value in the valve core slightly stuck state; creating a feature vector of the current time domain, a feature vector of the normal time domain of the valve core, a feature vector of the broken time domain of the valve core, a feature vector of the non-operating time domain of the valve core and a feature vector of the slightly stuck valve core based on the average current value in the current state, the effective current value in the current state, the current range value in the current state, the average current value in the normal state of the valve core, the effective current value in the normal state of the valve core, the current range value in the normal state of the valve core, the average current value in the broken spring state, the current range value in the broken spring state, the average current value in the non-operating state of the valve core, the effective current value in the non-operating state of the valve core, and the average current value in the slightly stuck valve core state, the effective current value in the slightly stuck valve core and the current range value in the slightly stuck valve core; Including, 3. The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 2.

4. The specific formulas for calculating the current average value, the current effective value, and the current range value of the current state are as follows: [0010] Where: 【number】 is the current average value in the current state, and i ｋ is the current value at the kth time point when the solenoid valve driving side is in the current state, and i ｎ is the current value at the nth time point when the solenoid valve driving side is in the current state, 2<k<n, n is the total number of times, I' is the current effective value in the current state, R is the current range value in the current state, and i Ｍａｘ is the maximum current value in the current state, and i Ｍｉｎ is the minimum current value in the current state, 4. The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 3.

5. The specific steps of obtaining the current frequency domain feature vector, the frequency domain feature vector of the normal valve core, the frequency domain feature vector of the broken spring, the frequency domain feature vector of the non-operating valve core, and the frequency domain feature vector of the slightly stuck valve core are as follows: a step of reading current values ​​at multiple time points when the pre-processed solenoid valve driving side is in a current state, historical current values ​​at multiple historical time points when the valve core is in a normal state, historical current values ​​at multiple historical time points when the spring is broken, historical current values ​​at multiple historical time points when the valve core is in a non-operating state, and historical current values ​​at multiple historical time points when the valve core is in a slightly stuck state, and performing second derivative analysis on these, thereby obtaining a second derivative of the current state, a second derivative of the valve core normal state, a second derivative of the spring broken state, a second derivative of the valve core in a non-operating state, and a second derivative of the valve core in a slightly stuck state; performing binary discrete wavelet packet decomposition on the second derivative of the current state, the second derivative of the valve core normal state, the second derivative of the spring broken state, the second derivative of the valve core non-operating state, and the second derivative of the valve core slightly stuck state, respectively, to obtain a wavelet function of the current state, a wavelet function of the valve core normal state, a wavelet function of the spring broken state, a wavelet function of the valve core non-operating state, and a wavelet function of the valve core slightly stuck state; performing energy analysis on the wavelet function of the current state, the wavelet function of the valve core normal state, the wavelet function of the spring broken state, the wavelet function of the valve core non-operating state, and the wavelet function of the valve core slightly stuck state, to obtain an energy value of the current state, an energy value of the valve core normal state, an energy value of the spring broken state, an energy value of the valve core non-operating state, and an energy value of the valve core slightly stuck state; performing normalization processing on the energy value of the current state, the energy value of the valve core normal state, the energy value of the spring broken state, the energy value of the valve core not working state, and the energy value of the valve core slightly stuck state to obtain normalized energy features of the current state, the normalized energy features of the valve core normal state, the normalized energy features of the spring broken state, the normalized energy features of the valve core not working state, and the normalized energy features of the valve core slightly stuck state, and recording them as a current frequency domain feature vector, a frequency domain feature vector of the valve core normal, a frequency domain feature vector of the spring broken, a frequency domain feature vector of the valve core not working, and a frequency domain feature vector of the valve core slightly stuck, respectively; 3. The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 2.

6. The specific formulas for calculating the wavelet function of the current state, the energy value of the current state, and the normalized energy feature of the current state are as follows: [0025] Here, Ψ ｊｒ (t) is the wavelet function of the r-th component of the j-th layer wavelet decomposition of the current state, g is the scale parameter of the wavelet transform, g=1, 2, 3..., s, s is the maximum scale, and 1≦s≦j, and r is the location parameter of the wavelet transform; 【number】 is the second derivative of the current state, and X ｊｒ is the coefficient obtained by the wavelet transform in the current state, and e ｊ is the energy value of the j-th layer wavelet decomposition of the current state, and e ｊ denotes the sum of the energy of the wavelet coefficients at all positions in the jth layer, r = 1, 2, 3, ..., Q, where Q is the total number of positions, and E ｊ is the normalized energy feature of the j-th layer wavelet decomposition of the current state, The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 5.

7. Using the vector dataset, perform model training on a pre-created random forest model, specifically, perform training using a C4.5 algorithm, specifically, sample the vector dataset to generate multiple decision trees and build a random forest model; 2. The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 1.

8. The specific steps to generate multiple decision trees are: reading the vector data set and randomly extracting a plurality of groups of data to generate a respective random vector data set; For each group of data, a random vector data set is divided into a vector training data set and a vector validation data set respectively; for each group of data, starting from the root node of the vector training data set, split using the feature with the highest information gain rate, repeating this splitting process for each subset generated by the split, and recursively building a tree; stopping the tree construction operation when a predefined stopping condition is reached; employing a post-pruning strategy on the constructed tree, which is performed after the tree has been fully generated, and using a vector validation data set to evaluate and adjust the tree before and after pruning until the tree meets a predetermined criterion; The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 7.

9. In the current diagnosis identification of the composite feature vector after the training of the random forest model is completed, the means of fault diagnosis by the random forest model uses the majority voting method to obtain the final result; 2. The method for online diagnosis of faults in a solenoid valve based on characteristic analysis in the time-frequency domain as claimed in claim 1.

Images