Electromechanical equipment detection method and device

Abstract

The invention provides an electromechanical equipment detection method and device, and the method comprises the steps: carrying out the fundamental frequency separation of a current signal of a monitoring node, extracting a high-frequency transient component of a current at the monitoring node, carrying out the inverse transformation of a propagation path of a pulse current of the monitoring node according to a cross correlation coefficient of the high-frequency transient component between adjacent monitoring nodes, and obtaining a detection result. Acquiring propagation attenuation characteristics of pulse current of the monitoring node; determining an abnormal label of magnetic flux leakage detection at each monitoring node position according to the gradient change rate of the magnetic flux leakage between the adjacent monitoring nodes, and further determining an abnormal detection tree of the magnetic flux leakage detection of the electromechanical equipment according to a topological correlation structure between all the abnormal labels and the monitoring nodes; and determining the confidence probability of the insulation defect at the monitoring node position according to all the propagation attenuation characteristics and the anomaly detection tree, and then positioning the insulation defect position of the electromechanical equipment based on the confidence probability of the insulation defect at the monitoring node position. By adopting the scheme of the invention, the multi-level association confidence detection of the insulation defect of the electromechanical equipment can be realized.

Description

Technical Field

[0001] This application relates to the technical field of electromechanical equipment detection. More specifically, this application relates to a method and device for detecting electromechanical equipment. Background Art

[0002] The detection of electromechanical equipment refers to the regular or irregular inspection and evaluation of the performance, structure, and operating status of electromechanical equipment, aiming to ensure the normal operation of the equipment, detect potential faults in advance, and perform preventive maintenance. With the improvement of industrial automation levels and the increase in equipment operation complexity, equipment failures may not only cause production interruptions and equipment damage but also lead to safety accidents. Therefore, accurate detection and analysis are particularly important. It can ensure the stability and safety of equipment operation, provide reliable guarantees for industrial production, and reduce economic losses and safety risks caused by sudden failures.

[0003] Insulation defects in the power supply circuit of electromechanical equipment can lead to electrical leakage, short circuits, or equipment failures, and in severe cases, even cause safety accidents or production interruptions. However, in the prior art, insulation defects usually rely on the monitoring of basic signals such as current, voltage, or temperature. However, these signals are often affected by external electrical noise and equipment complexity, resulting in poor stability and accuracy of detection results. Especially in large-scale and long-distance electromechanical equipment, the initial signals of defects are weak, making it difficult to accurately capture and locate them. In addition, existing insulation defect location methods usually focus on the monitoring of single signals and do not fully consider the comprehensive effects of multi-source data, which limits the location accuracy and is prone to false positives or false negatives, thereby affecting the normal operation and maintenance decisions of equipment. Therefore, how to achieve multi-level associated confidence detection of insulation defects in electromechanical equipment has become a difficult problem faced by the industry. Summary of the Invention

[0004] This application provides a method and device for detecting electromechanical equipment, which can achieve multi-level associated confidence detection of insulation defects in electromechanical equipment.

[0005] In a first aspect, this application provides a method for detecting electromechanical equipment, including the following steps: Set multiple monitoring nodes in the power supply circuit of the electromechanical equipment, and collect the current signals of each monitoring node through intelligent sensors; Perform fundamental frequency separation on the current signals of each monitoring node, and then extract the high-frequency transient components of the current at each monitoring node. Invert the propagation path of the pulse current at each monitoring node according to the cross-correlation coefficients of the high-frequency transient components between adjacent monitoring nodes to obtain the propagation attenuation characteristics of the pulse current at each monitoring node; Collect the magnetic flux leakage of the monitoring nodes through intelligent sensors, determine the abnormal labels of the magnetic flux leakage detection at the positions of each monitoring node according to the gradient change rate of the magnetic flux leakage between adjacent monitoring nodes, and then determine the abnormal detection tree of the magnetic flux leakage detection of the electromechanical equipment from all the abnormal labels and the topological association structure between all the monitoring nodes; Determine the confidence probability of the insulation defect at the position of each monitoring node according to the propagation attenuation characteristics of the pulse current of all the monitoring nodes and the abnormal detection tree, and then locate the insulation defect position of the electromechanical equipment based on the confidence probability of the insulation defect at the positions of all the monitoring nodes.

[0006] Preferably, perform fundamental frequency separation on the current signal of each monitoring node, and then extract the high-frequency transient component of the current at each monitoring node, which specifically includes: For each monitoring node, perform spectrum analysis on the current signal of the monitoring node to identify the fundamental frequency component and each harmonic component; Filter out the fundamental frequency component and each harmonic component from the current signal through a filter to obtain a residual component; Extract the high-frequency transient component of the current at the monitoring node from the residual component through wavelet transform, and then obtain the high-frequency transient component of the current at each monitoring node.

[0007] Preferably, the harmonic components specifically include the 2nd, 3rd, 4th, and 5th harmonic components.

[0008] Preferably, perform inverse transformation on the propagation path of the pulse current of each monitoring node according to the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes to obtain the propagation attenuation characteristics of the pulse current of each monitoring node, which specifically includes: Determine the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes; Determine the attenuation factor of the pulse current of each monitoring node according to all the cross-correlation coefficients; Invert and reconstruct the propagation path of the pulse current through the least squares method with the high-frequency transient component and the attenuation factor corresponding to each monitoring node; Determine the propagation attenuation characteristics of the pulse current of each monitoring node according to the intensity of the pulse current of each monitoring node in the propagation path.

[0009] Preferably, determining the abnormal label of the magnetic flux leakage detection at the position of each monitoring node according to the gradient change rate of the magnetic flux leakage between adjacent monitoring nodes specifically includes: Determine the attribute label of each monitoring node through threshold comparison of the gradient change rate; Determine the abnormal coefficient of the magnetic flux leakage detection at the position of each monitoring node according to the gradient change rate of the magnetic flux leakage between all adjacent monitoring nodes; Determine the abnormal tags for magnetic flux leakage detection at the positions of each monitoring node through the attribute tags of each monitoring node and all abnormal coefficients.

[0010] Preferably, determining the abnormal detection tree for magnetic flux leakage detection of the electromechanical equipment from all the abnormal tags and the topological association structure between all monitoring nodes specifically includes: Take each monitoring node as a tree node; Determine the structure tree for magnetic flux leakage detection of the electromechanical equipment according to the topological association structure between all monitoring nodes and all the tree nodes; Determine the connection relationship between each tree node through all the abnormal tags; Determine the abnormal detection tree for magnetic flux leakage detection of the electromechanical equipment according to the structure tree and the connection relationship between each tree node.

[0011] Preferably, the intelligent sensor specifically includes a current sensor and a magnetic flux leakage sensor.

[0012] In a second aspect, the present application provides an electromechanical equipment detection device, including: An acquisition module, configured to set a plurality of monitoring nodes in the power supply circuit of the electromechanical equipment and collect the current signals of each monitoring node through an intelligent sensor; A processing module, configured to perform fundamental frequency separation on the current signals of each monitoring node, and then extract the high-frequency transient component of the current at each monitoring node, perform inverse transformation on the propagation path of the pulse current of each monitoring node according to the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes, and obtain the propagation attenuation characteristics of the pulse current of each monitoring node; The processing module is further configured to collect the magnetic flux leakage of the monitoring node through an intelligent sensor, determine the abnormal tags for magnetic flux leakage detection at the positions of each monitoring node according to the gradient change rate of the magnetic flux leakage between adjacent monitoring nodes, and then determine the abnormal detection tree for magnetic flux leakage detection of the electromechanical equipment from all the abnormal tags and the topological association structure between all monitoring nodes; An execution module, configured to determine the confidence probability of the insulation defect at the position of each monitoring node according to the propagation attenuation characteristics of the pulse current of all monitoring nodes and the abnormal detection tree, and then locate the insulation defect position of the electromechanical equipment based on the confidence probability of the insulation defect at the positions of all monitoring nodes.

[0013] In a third aspect, the present application provides a computer device, which includes a memory and a processor, the memory stores code, and the processor is configured to obtain the code and execute the above-mentioned electromechanical equipment detection method.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the above-mentioned electromechanical equipment detection method is implemented.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In the embodiments of this application, a plurality of monitoring nodes are arranged in the power supply circuit of the electromechanical equipment, and the current signals of each monitoring node are collected by intelligent sensors; the fundamental frequency separation is performed on the current signals of each monitoring node, and then the high-frequency transient components of the current at each monitoring node are extracted. The propagation path of the pulse current at each monitoring node is inversely transformed according to the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes, and the propagation attenuation characteristics of the pulse current at each monitoring node are obtained; the leakage magnetic flux of the monitoring node is collected by an intelligent sensor, and the abnormal label of the leakage magnetic flux detection at the position of each monitoring node is determined according to the gradient change rate of the leakage magnetic flux between adjacent monitoring nodes. Furthermore, the abnormal detection tree of the leakage magnetic flux detection of the electromechanical equipment is determined by all the abnormal labels and the topological association structure between all monitoring nodes; the confidence probability of the insulation defect at the position of each monitoring node is determined according to the propagation attenuation characteristics of the pulse current of all monitoring nodes and the abnormal detection tree, and then the insulation defect position of the electromechanical equipment is located based on the confidence probability of the insulation defect at the position of all monitoring nodes.

[0016] It can be seen that the present application determines the confidence probability of the insulation defect at the monitoring node position through the propagation attenuation characteristics of the node pulse current and the abnormal detection tree of the magnetic flux leakage detection of the electromechanical equipment, and then locates the insulation defect position of the electromechanical equipment. First, the fundamental frequency of the current signal of the monitoring node is separated, and then the high-frequency transient component of the current at the monitoring node is extracted. By separating the high-frequency transient information in the current signal, the sensitivity to the weak fault signal of the electromechanical equipment can be improved. Secondly, the propagation path of the pulse current of each monitoring node is inversely transformed according to the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes, and the propagation attenuation characteristics of the pulse current of each monitoring node are obtained. By determining the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes, the subtle changes in the current signal propagation can be revealed, which helps to accurately identify and locate potential faults, reduce the interference of external noise, thereby improving the accuracy of fault detection. Through the inverse transformation, the propagation path of the pulse current in the power supply circuit can be accurately restored, thereby revealing the attenuation characteristics of the current, which helps to identify the possible abnormalities in the current propagation process, and further reveals potential equipment faults or insulation defects. Then, the abnormal detection tree of the magnetic flux leakage detection of the electromechanical equipment is determined through the abnormal label of the magnetic flux leakage detection at the monitoring node position and the topological association structure between the monitoring nodes. Through the topological association structure, the magnetic flux leakage abnormal information of each monitoring node is associated and analyzed, which can reveal the fault propagation relationship between each monitoring node, avoid misjudgment of single-point abnormalities, and at the same time enhance the ability to identify the magnetic flux leakage defects inside complex equipment, thereby improving the accuracy of abnormal detection and avoiding misjudgment or missed judgment caused by single-point detection. Finally, combining the propagation attenuation characteristics of the pulse current and the abnormal detection tree of the magnetic flux leakage detection of the electromechanical equipment, the confidence probability of the insulation defect at each monitoring node position is generated, and the insulation defect is located based on these confidence probabilities. Through the correlation analysis of multi-source signals and fully considering the mutual relationship between multi-source signals, the reliability of the equipment insulation defect detection can be improved. By quantifying the possibility of the defect occurrence through the confidence probability, the risk of misjudgment and missed judgment of the equipment insulation defect can be reduced, which helps to achieve more efficient and accurate fault diagnosis in the monitoring of large-scale and complex equipment. In summary, the solution of the present application can realize the multi-level associated confidence detection of the insulation defect of the electromechanical equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is an exemplary flowchart of an electromechanical equipment detection method according to some embodiments of the present application; Figure 2 is a detection logic diagram of an electromechanical equipment according to some embodiments of the present application; Figure 3 is a schematic flowchart of determining an abnormal detection tree according to some embodiments of the present application; Figure 4 is a schematic structural diagram of an electromechanical equipment detection device according to some embodiments of the present application; Figure 5 It is a schematic structural diagram of a computer device for implementing an electromechanical device detection method as shown in some embodiments of the present application. Detailed implementation manners

[0018] To better understand the technical solutions of the present application, the technical solutions of the present application will be described in detail below in conjunction with the specification drawings and specific implementation manners.

[0019] Refer to Figure 1 , which is an exemplary flowchart of an electromechanical device detection method as shown in some embodiments of the present application. The electromechanical device detection method 100 mainly includes the following steps: In step 101, multiple monitoring nodes are set in the power supply circuit of the electromechanical device, and the current signals of each monitoring node are collected through intelligent sensors.

[0020] It should be noted that the intelligent sensors in the present application specifically include current sensors and magnetic flux leakage sensors. Among them, the current sensor is a Hall effect sensor; specifically in implementation, multiple key positions in the power supply circuit of the electromechanical device can be selected as monitoring nodes, and the key positions are, for example, the transformer end, the switch end, and the cable joint. And intelligent sensors for monitoring current and magnetic flux leakage are set at each monitoring node, and then the current signals of each monitoring node are collected through the Hall effect sensor.

[0021] In some embodiments, refer to Figure 2 shown, which is a detection logic diagram of an electromechanical device as shown in some embodiments of the present application. This diagram shows the logical framework of the electromechanical device detection. At the monitoring end, the monitoring device collects the current signals and magnetic flux leakage of multiple monitoring nodes in the power supply circuit of the electromechanical device through intelligent sensors, stores the data in the storage device, and then transmits the data to the control end through the transmission device. At the control end, the processing system performs fundamental frequency separation on the current signals, extracts the high-frequency transient components, and inversely transforms the propagation path of the pulse current of each monitoring node according to the cross-correlation coefficient between adjacent monitoring nodes to obtain the propagation attenuation characteristics of the pulse current of each monitoring node; the calculation system determines the abnormal label according to the gradient change rate of the magnetic flux leakage, and constructs an abnormal detection tree; the analysis system comprehensively analyzes the propagation attenuation characteristics and the abnormal detection tree to determine the confidence probability of the insulation defect of each monitoring node; finally, the warning system locates the insulation defect position according to the confidence probability and issues a warning when necessary.

[0022] In step 102, fundamental frequency separation is performed on the current signals of each monitoring node, and then the high-frequency transient components of the current at each monitoring node are extracted. The propagation path of the pulse current of each monitoring node is inversely transformed according to the cross-correlation coefficient between the high-frequency transient components of each adjacent monitoring node to obtain the propagation attenuation characteristics of the pulse current of each monitoring node.

[0023] It should be noted that the high-frequency transient component in this application refers to the rapidly changing component in the current signal with a frequency higher than the fundamental frequency and a short duration, usually caused by a fault impact, and can be used to reflect the dynamic characteristics and abnormal states of electromechanical equipment.

[0024] In some embodiments, the fundamental frequency separation of the current signal of each monitoring node and then the extraction of the high-frequency transient component of the current at each monitoring node can be achieved by the following steps: For each monitoring node, perform a spectrum analysis on the current signal of the monitoring node to identify the fundamental frequency component and each harmonic component; Filter out the fundamental frequency component and each harmonic component from the current signal through a filter to obtain a residual component; Extract the high-frequency transient component of the current at the monitoring node from the residual component through wavelet transform, and then obtain the high-frequency transient component of the current at each monitoring node.

[0025] It should be noted that the fundamental frequency component in this application refers to the sine wave component with the lowest frequency and the highest amplitude in the current signal, usually the same as the power grid power frequency (such as 50 Hz), which represents the main energy component of the current and is the basic part of the current waveform; each harmonic component in this application refers to the component with a frequency that is an integer multiple of the fundamental frequency in the current signal. For example, when the fundamental frequency is 50 Hz, 100 Hz (2nd harmonic), 150 Hz (3rd harmonic), etc.; the high-frequency transient component in this application refers to the high-frequency component that appears in the current signal in a short time, usually exceeding the power frequency and its low-order harmonic range, with the characteristics of suddenness, non-periodicity, and rapid decay, mainly caused by the transient process of equipment faults, and can be used to detect the abnormal states of electromechanical equipment.

[0026] In specific implementation, for each monitoring node, the spectrum analysis of the current signal of the monitoring node and the identification of the fundamental frequency component and each harmonic component can be achieved in the following manner, that is: the fast Fourier transform can be used to perform time-frequency transformation on the current signal, converting the time-domain signal (i.e., the current signal) into a frequency-domain signal. By analyzing the peak distribution of the spectrum amplitude at the fundamental frequency and its integer multiple frequencies (i.e., each harmonic component), the amplitude and frequency components of the fundamental frequency component and each harmonic component are identified. Among them, the fundamental frequency of the current signal is 50 Hz, and here each harmonic component specifically includes the 2nd, 3rd, 4th, and 5th harmonic components; filtering the fundamental frequency component and each harmonic component from the current signal to obtain the residual component can be achieved in the following manner, that is: a band-stop filter (such as an IIR or FIR filter) can be used to filter the identified fundamental frequency component and each harmonic component, filtering out these known frequency components, and taking the remaining part of the current signal as the residual component. The design of the filter needs to ensure sufficient suppression ability for the target frequency while minimizing the impact on other frequency components as much as possible; extracting the high-frequency transient component of the current at the monitoring node from the residual component through wavelet transform, and then obtaining the high-frequency transient component of the current at each monitoring node can be achieved in the following manner, that is: the discrete wavelet transform can be performed on the filtered residual component, a suitable wavelet basis is selected (the Haar wavelet is selected in the solution of this application), and the decomposition level is set (set to 6 levels in the solution of this application) so that the high-frequency components can be effectively separated at a higher scale, filtering out the low-frequency part in the residual component, and taking the remaining high-frequency part as the high-frequency transient component of the current at the monitoring node, and its high-frequency transient component can reflect the mutation characteristics in the current signal.

[0027] It should be noted that the solution of this application can reduce the interference of the fundamental frequency signal through fundamental frequency separation and improve the accuracy of fault feature extraction.

[0028] In some embodiments, the inversion transformation of the propagation path of the pulse current at each monitoring node can be performed according to the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes, and the propagation attenuation characteristics of the pulse current at each monitoring node can be obtained through the following steps: Determine the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes; Determine the attenuation factor of the pulse current at each monitoring node according to all the cross-correlation coefficients; Invert and reconstruct the propagation path of the pulse current by the least squares method using the high-frequency transient component and the attenuation factor corresponding to each monitoring node; Determine the propagation attenuation characteristics of the pulse current at each monitoring node according to the intensity of the pulse current at each monitoring node in the propagation path.

[0029] It should be noted that the cross - correlation coefficient in the solution of this application is a statistic for measuring the correlation between adjacent high - frequency transient components; the attenuation factor in this application is a parameter used to quantify the degree of energy loss during the propagation of pulse current, and the larger the attenuation factor, the greater the degree of energy loss; the propagation attenuation characteristic in this application is an eigenvalue for measuring the energy loss of the path segment between monitoring nodes when pulse current propagates in electromechanical equipment.

[0030] Preferably, in the above - mentioned embodiment, the cross - correlation coefficient of high - frequency transient components between adjacent monitoring nodes can be determined by the following steps: For each pair of adjacent monitoring nodes, determine the Pearson correlation coefficient of high - frequency transient components between the adjacent monitoring nodes; According to the Pearson correlation coefficient and the physical distance between adjacent monitoring nodes, determine the cross - correlation coefficient of high - frequency transient components between adjacent monitoring nodes, and thus obtain the cross - correlation coefficient of high - frequency transient components between each pair of adjacent monitoring nodes.

[0031] In specific implementation, for each pair of adjacent monitoring nodes, the Pearson correlation coefficient of high - frequency transient components between the adjacent monitoring nodes can be determined in the following way, that is: for each pair of adjacent monitoring nodes, substitute the high - frequency transient components between the adjacent monitoring nodes into the Pearson correlation coefficient calculation formula, and take the calculation result as the Pearson correlation coefficient of high - frequency transient components between the adjacent monitoring nodes; according to the Pearson correlation coefficient and the physical distance between adjacent monitoring nodes, the cross - correlation coefficient of high - frequency transient components between adjacent monitoring nodes can be determined in the following way, that is: the adjustment coefficient can be determined according to the physical distance between adjacent monitoring nodes. Specifically, the natural exponential function value of the opposite number of the physical distance between adjacent monitoring nodes can be used as the adjustment coefficient, and then the product of the adjustment coefficient and the Pearson correlation coefficient is used as the cross - correlation coefficient of high - frequency transient components between adjacent monitoring nodes. Through the above method, the cross - correlation coefficient of high - frequency transient components between each pair of adjacent monitoring nodes can be obtained.

[0032] In specific implementation, the decay factor of the pulse current of each monitoring node can be determined according to all cross-correlation coefficients in the following way: a monitoring node can be selected as the target monitoring node, and the reciprocal of the average value of the cross-correlation coefficients between the target monitoring node and its adjacent monitoring nodes can be used as the decay factor of the pulse current of the target monitoring node. By repeating the selection of other monitoring nodes, the decay factor of the pulse current of each monitoring node can be obtained. Based on the high-frequency transient component and the decay factor corresponding to each monitoring node, the propagation path of the pulse current can be inversely reconstructed by the least squares method in the following way: the high-frequency transient component and the decay factor corresponding to each monitoring node can be used as independent variables in the inversion equation. The pulse current intensity at each monitoring node in the inversion equation = high-frequency transient component * decay factor + error term, where the error term can be obtained by performing regression fitting on the current propagation speed between monitoring points using the least squares method, and its fitting error can be used as the error term. Finally, the pulse current intensities obtained by fitting are connected according to the positions of the monitoring points to obtain the current propagation curve, and this curve is used as the propagation path of the pulse current. The propagation attenuation characteristics of the pulse current of each monitoring node can be determined according to the intensity of the pulse current of each monitoring node in the propagation path in the following way: the absolute difference between the pulse current intensities between every two adjacent monitoring nodes in the propagation path can be used as the energy loss of the path segment between every two adjacent monitoring nodes. For each monitoring node, the average value of the energy losses between the monitoring node and its adjacent monitoring nodes can be used as the propagation attenuation characteristic of the pulse current of the monitoring node, and thus the propagation attenuation characteristics of the pulse current of each monitoring node can be obtained.

[0033] It should be noted that the solution of this application uses cross-correlation coefficients to analyze the propagation path of the pulse current between monitoring nodes, and can inversely obtain the propagation attenuation characteristics of the current, thereby reflecting the conductivity and energy loss conditions at different positions inside the equipment, and providing a more powerful basis for the diagnosis of insulation defects.

[0034] In step 103, the magnetic flux leakage of the monitoring nodes is collected by intelligent sensors, and the abnormal labels of the magnetic flux leakage detection at the positions of each monitoring node are determined according to the gradient change rate of the magnetic flux leakage between adjacent monitoring nodes. Then, the abnormal detection tree of the magnetic flux leakage detection of the electromechanical equipment is determined by all the abnormal labels and the topological association structure between all monitoring nodes.

[0035] It should be noted that in this application, a Hall effect sensor (such as a HALL sensor) can be used to collect the magnetic flux leakage of each monitoring node. Among them, the magnetic flux leakage refers to the magnetic flux that leaks to the outside of the equipment due to incomplete magnetic field closure or abnormal conditions (such as insulation damage) in the electromechanical equipment.

[0036] In some embodiments, the abnormal labels of the magnetic flux leakage detection at the positions of each monitoring node can be determined according to the gradient change rate of the magnetic flux leakage between adjacent monitoring nodes by the following steps: Determine the attribute labels of each monitoring node by comparing the threshold of the gradient change rate; Determine the anomaly coefficient of the magnetic flux leakage detection at the position of each monitoring node according to the gradient change rate of the magnetic flux leakage between all adjacent monitoring nodes; Determine the anomaly label of the magnetic flux leakage detection at the position of each monitoring node through the attribute labels of each monitoring node and all anomaly coefficients.

[0037] It should be noted that the attribute label in this application is a status identifier used to reflect the magnetic flux leakage detection at the position of the monitoring node. The status identifier specifically includes normal and abnormal. The anomaly coefficient in this application is an index to measure the degree of abnormality of the magnetic flux leakage detection result at the position of the monitoring node. The larger the anomaly coefficient, the greater the degree of abnormality of the magnetic flux leakage detection result at the position of the monitoring node. The anomaly label in this application is a package quantity used to reflect the magnetic flux leakage detection result (i.e., the attribute identifier and the anomaly coefficient) at the position of the monitoring node.

[0038] When specifically implemented, determining the attribute labels of each monitoring node by comparing the threshold of the gradient change rate can be achieved by the following method, that is: set a safety threshold based on historical magnetic flux leakage data. This safety threshold can reflect the normal operating range and the health status of the device. This safety threshold represents the magnetic flux leakage fluctuation range of the device under normal operating conditions. Once the gradient change rate of the actual magnetic flux leakage exceeds this threshold, it indicates that the device may have abnormal conditions, such as insulation faults and magnetic circuit damage problems. For each monitoring node, compare the average value of all gradient change rates corresponding to the monitoring node with the safety threshold. When the average value is greater than the safety threshold, determine that the magnetic flux leakage detection result at the position of the monitoring node is abnormal, and mark the attribute label of the monitoring node with an abnormal magnetic flux leakage detection result as "1". When the average value is less than or equal to the safety threshold, determine that the magnetic flux leakage detection result at the position of the monitoring node is normal, and mark the attribute label of the monitoring node with a normal magnetic flux leakage detection result as "0". Determining the anomaly coefficient of the magnetic flux leakage detection at the position of each monitoring node according to the gradient change rate of the magnetic flux leakage between all adjacent monitoring nodes can be achieved by the following method, that is: for each monitoring node, normalize the average value of all gradient change rates corresponding to the monitoring node, that is, map the average value to the range of 0-1, and then use the normalized average value as the anomaly coefficient of the magnetic flux leakage detection at the position of the monitoring node. Through the above method, the anomaly coefficient of the magnetic flux leakage detection at the position of each monitoring node can be obtained. Determining the anomaly label of the magnetic flux leakage detection at the position of each monitoring node through the attribute labels of each monitoring node and all anomaly coefficients can be achieved by the following method, that is: for each monitoring node, package the attribute label of the monitoring node and the corresponding anomaly coefficient together, and use the packaged label as the anomaly label of the magnetic flux leakage detection at the position of the monitoring node, so as to obtain the anomaly label of the magnetic flux leakage detection at the position of each monitoring node.

[0039] It should be noted that the solution of this application can use the magnetic flux leakage to identify local abnormal areas inside the device. The gradient change rate of the magnetic flux leakage can quantify the magnetic field change situation, so as to accurately mark the monitoring nodes that may have defects.

[0040] In some embodiments, referring to Figure 3 As shown, this figure is a schematic flowchart of determining an anomaly detection tree in some embodiments of this application. In this embodiment, the anomaly detection tree for magnetic flux leakage detection of electromechanical equipment determined by all anomaly labels and the topological association structure between all monitoring nodes can be implemented by the following steps: In step 1031, each monitoring node is used as a tree node; In step 1032, a structure tree for magnetic flux leakage detection of electromechanical equipment is determined according to the topological association structure between all monitoring nodes and all tree nodes; In step 1033, the connection relationship between each tree node is determined through all anomaly labels; In step 1034, the anomaly detection tree for magnetic flux leakage detection of electromechanical equipment is determined according to the structure tree and the connection relationship between each tree node.

[0041] It should be noted that the structure tree in this application refers to a tree-shaped graphic structure constructed according to the physical and logical connection relationships of each monitoring node in the electromechanical equipment; the tree node in this application represents the fulcrum in the structure tree; the connection relationship in this application is an index to measure the connection strength between tree nodes; the anomaly detection tree in this application is a tree-shaped structure representing the abnormal propagation path in magnetic flux leakage detection.

[0042] In specific implementation, each monitoring node can be used as a tree node and can be implemented in the following ways: that is, each monitoring node can be marked as a tree node; the structure tree for magnetic flux leakage detection of electromechanical equipment can be determined according to the topological association structure among all monitoring nodes and all tree nodes, and can be implemented in the following ways: that is, the physical distance and logical connection relationship among monitoring nodes can be mapped to the topological association relationship among monitoring nodes. Specifically, the product of the reciprocal of the physical distance and the logical value that can reflect the logical connection relationship can be used as the topological association relationship among monitoring nodes. Among them, the logical connection relationship specifically includes the existence of a logical connection and the non - existence of a logical connection. The logical value for the existence of a logical connection is 1, and the logical value for the non - existence of a logical connection is 0. Then, the topological association relationship among monitoring nodes is used as the connection weight between tree nodes, and all tree nodes are connected by all the connection weights to obtain a tree structure, and this tree structure is used as the structure tree for magnetic flux leakage detection of electromechanical equipment; the connection relationship between each two tree nodes can be determined by all the abnormal tags and can be implemented in the following ways: that is, for every two tree nodes, obtain the abnormal tags of magnetic flux leakage detection at the positions of the monitoring nodes corresponding to the two tree nodes, and obtain the corresponding monitoring node attribute tags and abnormal coefficients from the abnormal tags. Then, compare the attribute tags. When the two attribute tags are the same, mark the attribute value between the two monitoring nodes as 1. When the two attribute tags are different, mark the attribute value between the two monitoring nodes as 0. Calculate the average value of the abnormal coefficients corresponding to the two monitoring nodes, and use the product of the attribute value between the two monitoring nodes and the average value as the connection relationship between the two tree nodes, so as to obtain the connection relationship between each two tree nodes; the abnormal detection tree for magnetic flux leakage detection of electromechanical equipment can be determined according to the structure tree and the connection relationship between each tree node, and can be implemented in the following ways: that is, compensate the connection weights between each tree node in the structure tree through the connection relationship between each tree node, that is, multiply the connection relationship between each tree node and the connection weights between each tree node, and use the product obtained by multiplication to replace and update the connection weights between the corresponding tree nodes in the structure tree. Then, use the updated structure tree as the abnormal detection tree for magnetic flux leakage detection of electromechanical equipment.

[0043] It should be noted that through the construction of the abnormal detection tree, the proposed solution of this application can associate multiple abnormal tags based on the topological structure of monitoring nodes, realize the systematic analysis of the overall magnetic flux leakage abnormality of electromechanical equipment, thereby improving the comprehensiveness and accuracy of detection and avoiding misjudgment of single - point data.

[0044] In step 104, according to the propagation attenuation characteristics of the pulse current of all monitoring nodes and the abnormal detection tree, determine the confidence probability of insulation defects at the positions of each monitoring node, and then locate the insulation defect positions of the electromechanical equipment based on the confidence probability of insulation defects at the positions of all monitoring nodes.

[0045] In some embodiments, the confidence probability of insulation defects at the positions of each monitoring node can be determined according to the propagation attenuation characteristics of the pulsed current of all monitoring nodes and the abnormal detection tree, which can be realized by the following steps: Determine the prior probability of insulation defects occurring at each monitoring node; According to the propagation attenuation characteristics of the pulsed current of all monitoring nodes and the abnormal detection tree, determine the posterior probability of insulation defects occurring at each monitoring node; Determine the confidence probability of insulation defects at the positions of each monitoring node based on the prior probability and the posterior probability of insulation defects occurring at each monitoring node.

[0046] It should be noted that the confidence probability in this application is an index for measuring the credible probability of insulation defects occurring at the position of the monitoring node. The greater the confidence probability, the greater the possibility of insulation defects occurring at the position of the monitoring node.

[0047] Specifically, the prior probability of insulation defects occurring at each monitoring node can be determined in the following way: collect the historical fault data of the electromechanical equipment, statistically count the historical insulation fault occurrence frequencies of each monitoring node through statistical methods, and input the historical insulation fault occurrence frequencies of each monitoring node and the status information of the monitoring node into the Bayesian network. Then, the insulation aging probability of each monitoring node can be inferred through the Bayesian network, and the inferred insulation aging probability can be used as the prior probability of insulation defects occurring at the corresponding monitoring node. The posterior probability of insulation defects occurring at each monitoring node can be determined according to the propagation attenuation characteristics of the pulsed current of all monitoring nodes and the abnormal detection tree in the following way: for each monitoring node, the product of the propagation attenuation characteristics of the pulsed current of the monitoring node and the abnormal coefficient of the corresponding tree node of the monitoring node in the abnormal detection tree can be used as the fault value of the monitoring node, and then the fault values of all monitoring nodes can be obtained. The ratio of the absolute difference between the fault value and the preset fault threshold to the preset fault threshold can be used as the posterior probability of insulation defects occurring at the corresponding monitoring node, where the value range of the ratio is between 0 and 1, and the preset fault threshold can be obtained through experiments. In this application, the average value of the historical experimental fault values can be used as the preset fault threshold. The confidence probability of insulation defects at the positions of each monitoring node can be determined based on the prior probability and the posterior probability of insulation defects occurring at each monitoring node in the following way: for each monitoring node, the prior probability and the posterior probability of insulation defects occurring at the monitoring node can be substituted into the Bayesian probability formula, and the output result can be used as the confidence probability of insulation defects at the position of the monitoring node, and then the confidence probability of insulation defects at the position of each monitoring node can be obtained.

[0048] It should be noted that in the solution of this application, by fusing multiple monitoring information, based on the propagation attenuation characteristics of the pulse current and the abnormal information of the magnetic flux leakage detection, the confidence probability of the existence of insulation defects at each monitoring node is comprehensively calculated. Furthermore, through the method of probability analysis, the risk of misjudgment can be reduced, thereby improving the accuracy of insulation defect identification.

[0049] In some embodiments, the positioning of the insulation defect location of the electromechanical equipment based on the confidence probability of the insulation defects at all monitoring node positions can be achieved by the following steps: Generate a location distribution map of the insulation defects of the electromechanical equipment according to the confidence probability of the insulation defects at all monitoring node positions; Locate the location of the insulation defect of the electromechanical equipment through the location distribution map.

[0050] In specific implementation, generating a location distribution map of the insulation defects of the electromechanical equipment according to the confidence probability of the insulation defects at all monitoring node positions can be achieved in the following way, that is: all monitoring nodes can be connected according to the physical connection relationship of the power supply circuits between the monitoring nodes to obtain a connection diagram, and then the confidence probability of the insulation defects at each monitoring node position can be marked on the connection diagram, and the marked connection diagram can be used as the location distribution map of the insulation defects of the electromechanical equipment; locating the location of the insulation defect of the electromechanical equipment through the location distribution map can be achieved by the following steps, that is: two monitoring nodes with a connection relationship can be selected from the location distribution map, and then the average value of the confidence probabilities of the insulation defects at the two monitoring node positions can be used as the probability of the occurrence of insulation defects in the path segment between the two monitoring nodes. Through this probability, it can be judged whether the path segment has insulation defects, and thus the location of the insulation defect of the electromechanical equipment can be completed.

[0051] It should be noted that the solution of this application realizes the precise positioning of equipment insulation defects by comprehensively analyzing current signals and magnetic flux leakage. First, by arranging multiple monitoring nodes in the power supply circuit and using intelligent sensors to collect current signals, the comprehensiveness and high precision of data collection are ensured; second, by separating the fundamental frequency to extract the high-frequency transient component and inversely calculating the propagation path of the pulse current based on the cross-correlation coefficient of the high-frequency signals, the propagation attenuation characteristics are obtained, thereby quantitatively analyzing the abnormal characteristics of the current signal and improving the sensitivity of insulation defect detection; then, using the gradient change rate of the magnetic flux leakage to mark the abnormal detection nodes and constructing an abnormal detection tree to identify systematic defects in a topological association manner, improving the systematicness and reliability of detection; finally, by fusing the propagation attenuation characteristics of the pulse current and the abnormal detection tree, calculating the insulation defect confidence probability of each monitoring node, realizing the precise positioning of insulation defects, improving the detection accuracy, reducing the risk of misjudgment, and optimizing the equipment operation and maintenance efficiency; in summary, the solution of this application combines the analysis of electrical and magnetic field signals, gives full play to the advantages of multi-source information fusion, and can improve the accuracy and intelligent level of electromechanical equipment insulation detection.

[0052] On the other hand, in some embodiments, the present application provides an electromechanical equipment detection device. Refer to Figure 4 , which is a schematic structural diagram of the electromechanical equipment detection device shown in some embodiments of the present application. The electromechanical equipment detection device 400 includes: an acquisition module 401, a processing module 402, and an execution module 403, which are described as follows: Acquisition module 401. In the present application, the acquisition module 401 is mainly used to set a plurality of monitoring nodes in the power supply circuit of the electromechanical equipment, and collect the current signals of each monitoring node through intelligent sensors; Processing module 402. In the present application, the processing module 402 is used to perform fundamental frequency separation on the current signals of each monitoring node, and then extract the high-frequency transient components of the current at each monitoring node. According to the cross-correlation coefficients of the high-frequency transient components between adjacent monitoring nodes, an inverse transformation is performed on the propagation path of the pulse current at each monitoring node to obtain the propagation attenuation characteristics of the pulse current at each monitoring node; In the present application, the processing module 402 is further used to collect the leakage magnetic flux of the monitoring nodes through intelligent sensors, determine the abnormal labels of the leakage magnetic detection at the positions of each monitoring node according to the gradient change rate of the leakage magnetic flux between adjacent monitoring nodes, and then determine the abnormal detection tree of the leakage magnetic detection of the electromechanical equipment from all the abnormal labels and the topological association structure between all monitoring nodes; Execution module 403. In the present application, the execution module 403 is mainly used to determine the confidence probability of the insulation defects at the positions of each monitoring node according to the propagation attenuation characteristics of the pulse current at all monitoring nodes and the abnormal detection tree, and then locate the insulation defect positions of the electromechanical equipment based on the confidence probability of the insulation defects at the positions of all monitoring nodes.

[0053] In addition, the present application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to obtain the code and execute the above-mentioned electromechanical equipment detection method.

[0054] In some embodiments, refer to Figure 5 , which is a schematic structural diagram of the computer device for implementing the electromechanical equipment detection method shown in some embodiments of the present application. The electromechanical equipment detection method in the above embodiments can be implemented by Figure 5 the computer device shown, and the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0055] The processor 501 can be a general-purpose central processing unit (CPU), or an application-specific integrated circuit (ASIC).

[0056] The communication bus 502 can be used to transfer information between the above components.

[0057] The memory 503 can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CDROM), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 503 can exist independently and be connected to the processor 501 through the communication bus 502. The memory 503 can also be integrated with the processor 501.

[0058] Among them, the memory 503 is used to store the program code for executing the solution of this application, and is controlled by the processor 501 to execute. The processor 501 is used to execute the program code stored in the memory 503. The program code can include one or more software modules. The electromechanical device detection method in the above embodiments can be implemented by one or more software modules in the program code in the processor 501 and the memory 503.

[0059] The communication interface 504, using any device such as a transceiver, is used to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0060] In a specific implementation, as an example, a computer device may include multiple processors, and each of these processors may be a single-core (single CPU) processor or a multi-core (multi CPU) processor. The processor here may refer to one or more devices, circuits, and / or processing cores for processing data (such as computer program instructions).

[0061] The above-mentioned computer device may be a general-purpose computer device or a special-purpose computer device. In a specific implementation, the computer device may be a desktop computer, a laptop computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of the present application do not limit the type of the computer device.

[0062] In addition, the present application also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the above-mentioned electromechanical device detection method is implemented.

[0063] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0064] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. A method for detecting an electromechanical device, characterized in that, Including the following steps: Set multiple monitoring nodes in the power supply circuit of the electromechanical equipment, and collect the current signals of each monitoring node through intelligent sensors; Perform fundamental frequency separation on the current signals of each monitoring node, and then extract the high-frequency transient components of the current at each monitoring node. Invert the propagation path of the pulsed current at each monitoring node according to the cross-correlation coefficients of the high-frequency transient components between adjacent monitoring nodes to obtain the propagation attenuation characteristics of the pulsed current at each monitoring node; Collect the leakage magnetic flux of the monitoring nodes through intelligent sensors, determine the abnormal labels for the leakage magnetic flux detection at the positions of each monitoring node according to the gradient change rate of the leakage magnetic flux between adjacent monitoring nodes, and then determine the abnormal detection tree for the leakage magnetic flux detection of the electromechanical equipment from all the abnormal labels and the topological association structure between all monitoring nodes; Determine the confidence probability of the insulation defects at the positions of each monitoring node according to the propagation attenuation characteristics of the pulsed current at all monitoring nodes and the abnormal detection tree, and then locate the insulation defect positions of the electromechanical equipment based on the confidence probabilities of the insulation defects at the positions of all monitoring nodes.

2. The method according to claim 1, wherein Performing fundamental frequency separation on the current signals of each monitoring node, and then extracting the high-frequency transient components of the current at each monitoring node specifically includes: For each monitoring node, perform spectrum analysis on the current signal of the monitoring node to identify the fundamental frequency component and each harmonic component; Filter out the fundamental frequency component and each harmonic component from the current signal through a filter to obtain a residual component; Extract the high-frequency transient components of the current at the monitoring node from the residual component through wavelet transform, and then obtain the high-frequency transient components of the current at each monitoring node.

3. The method according to claim 2, wherein The specific harmonic components include the 2nd, 3rd, 4th, and 5th harmonic components.

4. The method according to claim 1, wherein Inverting the propagation path of the pulsed current at each monitoring node according to the cross-correlation coefficients of the high-frequency transient components between adjacent monitoring nodes to obtain the propagation attenuation characteristics of the pulsed current at each monitoring node specifically includes: Determine the cross-correlation coefficients of the high-frequency transient components between adjacent monitoring nodes; Determine the attenuation factor of the pulsed current at each monitoring node according to all the cross-correlation coefficients; Invert and reconstruct the propagation path of the pulsed current through the least squares method using the high-frequency transient component and the attenuation factor corresponding to each monitoring node; Determine the propagation attenuation characteristics of the pulsed current at each monitoring node according to the intensity of the pulsed current at each monitoring node in the propagation path.

5. The method according to claim 1, characterized in that Determining the abnormal labels for the leakage magnetic flux detection at the positions of each monitoring node according to the gradient change rate of the leakage magnetic flux between adjacent monitoring nodes specifically includes: Determine the attribute labels of each monitoring node through threshold comparison of the gradient change rate; Determine the abnormal coefficients for the leakage magnetic flux detection at the positions of each monitoring node according to the gradient change rates of the leakage magnetic flux between all adjacent monitoring nodes; Determine the abnormal labels for the leakage magnetic flux detection at the positions of each monitoring node through the attribute labels of each monitoring node and all the abnormal coefficients.

6. The method according to claim 1, wherein Determining the abnormal detection tree for the leakage magnetic flux detection of the electromechanical equipment from all the abnormal labels and the topological association structure between all monitoring nodes specifically includes: Regard each monitoring node as a tree node; Determine the structure tree for the magnetic flux leakage detection of the electromechanical equipment according to the topological association structure among all monitoring nodes and all tree nodes; Determine the connection relationship between each tree node through all the abnormal tags; Determine the abnormal detection tree for the magnetic flux leakage detection of the electromechanical equipment according to the structure tree and the connection relationship between each tree node.

7. The method according to claim 1, wherein The intelligent sensor specifically includes a current sensor and a magnetic flux leakage sensor.

8. An electromechanical equipment detection device, characterized in that, It includes: An acquisition module, configured to set a plurality of monitoring nodes in the power supply circuit of the electromechanical equipment, and acquire the current signals of each monitoring node through the intelligent sensor; A processing module, configured to perform fundamental frequency separation on the current signals of each monitoring node, and then extract the high-frequency transient component of the current at each monitoring node, and perform inverse transformation on the propagation path of the pulsed current of each monitoring node according to the cross-correlation coefficient of the high-frequency transient components between adjacent monitoring nodes, so as to obtain the propagation attenuation characteristics of the pulsed current of each monitoring node; The processing module is further configured to acquire the magnetic flux leakage of the monitoring node through the intelligent sensor, determine the abnormal tag for the magnetic flux leakage detection at the position of each monitoring node according to the gradient change rate of the magnetic flux leakage between adjacent monitoring nodes, and then determine the abnormal detection tree for the magnetic flux leakage detection of the electromechanical equipment from all the abnormal tags and the topological association structure among all monitoring nodes; An execution module, configured to determine the confidence probability of the insulation defect at the position of each monitoring node according to the propagation attenuation characteristics of the pulsed current of all monitoring nodes and the abnormal detection tree, and then locate the insulation defect position of the electromechanical equipment based on the confidence probability of the insulation defect at the position of all monitoring nodes.

9. A computer device, the computer device includes a memory and a processor, the memory stores code, characterized in that, The processor is configured to obtain the code and execute the electromechanical equipment detection method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the electromechanical equipment detection method according to any one of claims 1 to 7.

Images