Partial discharge intelligent diagnosis system and method based on multi-mode signal fusion

Abstract

The invention discloses a partial discharge intelligent diagnosis system and method based on multi-modal signal fusion, and relates to the technical field of partial discharge intelligent diagnosis, and the system comprises an equipment end multi-modal partial discharge detection unit, a multi-modal data collection module, and an equipment end discharge autonomous processing module based on a chip. The partial discharge intelligent diagnosis system can detect partial discharge from multiple dimensions by fusing ultrasonic waves, electromagnetic waves and visual signals, and the accuracy and reliability of detection are effectively improved. The comprehensive analysis of the multi-modal data reduces misdiagnosis and missed diagnosis possibly caused by a single signal. And the equipment end discharge autonomous disposal module can detect in real time and carry out edge judgment based on a preset threshold value, so that a partial discharge event can be quickly identified. Independent arrangement of the side ends can be realized, real-time monitoring is beneficial to timely finding of abnormity, a power-off control signal is interacted with a controller of target equipment, and rapid response can prevent expansion of a partial discharge phenomenon from causing greater damage to electric parts in the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of partial discharge intelligent diagnosis, and in particular to a partial discharge intelligent diagnosis system and method based on multi-modal signal fusion. Background Art

[0002] Partial discharge (PD) is a common phenomenon in the insulation system of high-voltage power equipment, which may cause equipment performance degradation or even failure. Currently, partial discharge diagnosis technologies mainly include ultrasonic detection, high-frequency current detection, ultra-high frequency detection, optical detection and chemical detection.

[0003] At present, partial discharge generated inside the equipment is usually detected by ultrasonic detection, high-frequency current detection, ultra-high-frequency current detection, optical detection and other means;

[0004] Ultrasonic signals are used to identify partial discharge. This detection method is easily interfered by environmental noise, the signal is severely attenuated during propagation, and the positioning accuracy is not high.

[0005] High-frequency current detection detects high-frequency current signals caused by partial discharge. The detection range is limited and it is insensitive to long-distance discharge.

[0006] UHF detection uses UHF signals to detect partial discharge. The signal attenuates quickly in complex environments and the detection equipment is expensive.

[0007] Optical detection uses the optical signal generated by partial discharge for detection. It is greatly affected by ambient light and the detection equipment is complex to install.

[0008] Diagnosis is done by detecting chemical gases produced by partial discharge. The response time is long and real-time monitoring is not possible.

[0009] Therefore, developing a new partial discharge detection method to solve the problems of missed detection and misdiagnosis in traditional single partial discharge detection is a technical problem that needs to be solved urgently. Summary of the invention

[0010] In order to solve the technical problem of partial discharge detection and diagnosis of the above equipment, the present invention provides a partial discharge intelligent diagnosis system and method based on multi-modal signal fusion. The following technical solutions are adopted:

[0011] The intelligent diagnosis system for partial discharge based on multimodal signal fusion comprises a multimodal partial discharge detection unit on the equipment side, a multimodal data acquisition module and a chip-based equipment side discharge autonomous disposal module, wherein the multimodal partial discharge detection unit on the equipment side comprises an ultrasonic detection module, an electromagnetic wave signal detection module and a visual detection module, wherein the ultrasonic detection module is used to detect the ultrasonic signal change inside the target equipment, the electromagnetic wave signal detection module is used to detect the ultra-high frequency electromagnetic wave signal generated inside the target equipment, the visual detection module captures the visual picture inside the target equipment, the ultrasonic detection module, the electromagnetic wave signal detection module and the visual detection module are respectively connected to the signal input end of the multimodal data acquisition module in communication, and the multimodal The signal output end of the data acquisition module is communicatively connected with the device-side discharge autonomous handling module. The device-side discharge autonomous handling module performs edge judgment based on the threshold whether local discharge occurs, sets the ultrasonic change threshold and the ultra-high frequency electromagnetic wave signal threshold. If it is judged that the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current ultra-high frequency electromagnetic wave signal detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and at the same time, a luminous feature appears in the feature analysis of the visual image inside the device, then the output edge judgment shows a diagnosis result of local discharge. The device-side discharge autonomous handling module is communicatively connected with the controller of the target device. When the edge judgment of the device-side discharge autonomous handling module shows a diagnosis result of local discharge, it exchanges a power-off control signal with the controller of the target device.

[0012] By adopting the above technical solutions, the intelligent partial discharge diagnosis system can detect partial discharge from multiple dimensions by integrating ultrasonic, electromagnetic and visual signals, effectively improving the accuracy and reliability of detection. The comprehensive analysis of multimodal data reduces the misdiagnosis and missed diagnosis that may be caused by a single signal.

[0013] The device-side discharge autonomous disposal module performs edge judgment based on the threshold to determine whether partial discharge occurs, sets the ultrasonic change threshold and the ultra-high frequency electromagnetic wave signal threshold, and if it is determined that the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current ultra-high frequency electromagnetic wave signal detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and at the same time, the feature analysis of the visual image inside the device shows a luminous feature, then the diagnosis result of partial discharge at the edge judgment is output;

[0014] When the luminous features are detected, as long as the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current UHF electromagnetic wave signal detection value exceeds the UHF electromagnetic wave signal threshold, the probability of partial discharge in the target device is very high, and the diagnosis result of partial discharge can be output. At the same time, the power-off signal can be output. The device-side discharge autonomous disposal module can detect in real time and make edge judgments based on preset thresholds to quickly identify partial discharge events. Because only thresholds are used for judgment, the amount of calculation is small, and the independent arrangement of the edge can be realized. Real-time monitoring helps to detect anomalies in a timely manner, and the power-off control signal is exchanged to the controller of the target device. Rapid response can prevent the expansion of partial discharge phenomena and cause greater damage to the electrical components in the device.

[0015] Timely detection and treatment of partial discharge can prevent equipment damage and safety accidents caused by discharge, and improve the safety and reliability of the system.

[0016] Optionally, the ultrasonic detection module includes an ultrasonic sensor mounting base, an ultrasonic sensor, a filter and an ultrasonic data buffer, one end of the ultrasonic sensor mounting base is mounted at the top middle position of the inner wall of the device housing of the target device, the other end of the ultrasonic sensor mounting base is provided with a sensor mounting internal thread, the ultrasonic sensor is threaded on the sensor mounting internal thread, and the ultrasonic signal of the device housing of the target device is received through the ultrasonic sensor mounting base, the ultrasonic sensor is communicatively connected to the filter, the filter is communicatively connected to the ultrasonic data buffer, and the ultrasonic data buffer is communicatively connected to the data input end of the multimodal data acquisition module.

[0017] By adopting the above technical solution, partial discharge will generate ultrasonic signals. When partial discharge occurs inside the target device, the electrons in the discharge area move at high speed and collide with the surrounding medium, causing the medium to expand instantly due to heat, thereby generating mechanical stress waves, that is, ultrasonic waves. These ultrasonic signals will propagate inside the device in the form of waves. The ultrasonic sensor mount is used in combination with the ultrasonic sensor to receive these signals, and finally they are filtered by the filter and transmitted to the modal data acquisition module.

[0018] Optionally, the electromagnetic wave signal detection module includes an ultra-high frequency partial discharge sensor and an ultra-high frequency data buffer, the ultra-high frequency partial discharge sensor is arranged on the periphery of the main cable inside the target device, the ultra-high frequency partial discharge sensor is communicatively connected to the ultra-high frequency data buffer, and the ultra-high frequency data buffer is communicatively connected to the data input end of the multimodal data acquisition module.

[0019] By adopting the above technical solution, the high-frequency partial discharge sensor detects whether partial discharge occurs by coupling the ultra-high frequency electromagnetic wave signal generated by partial discharge in the electrical system. The basic principle of the ultra-high frequency detection method is to use the ultra-high frequency partial discharge sensor to detect the ultra-high frequency electromagnetic wave signal (the frequency of the electromagnetic wave is about 300MHz to 3GHz) generated by partial discharge in the power equipment.

[0020] Optionally, the visual detection module includes a visual camera and a visual data buffer, wherein the visual camera is installed on the inner wall of the shell of the target device through a bracket to capture the internal visual image of the target device, the visual camera is communicatively connected to the visual data buffer, and the visual data buffer is communicatively connected to the data input end of the multimodal data acquisition module.

[0021] By adopting the above technical solution, the visual camera can recognize the luminous features.

[0022] Optionally, the device-side multimodal partial discharge detection unit also includes a wireless communication attenuation detection module, which includes a first wireless communication module, a second wireless communication module and a wireless communication data buffer. The first wireless communication module and the second wireless communication module are respectively installed on the inner wall of the shell of the target device through a bracket. The first wireless communication module and the second wireless communication module are wirelessly connected to each other, and wirelessly communicate with each other to exchange standard data packets at set intervals. The first wireless communication module and the second wireless communication module store the received standard data packets in the wireless communication data buffer, and the wireless communication data buffer is communicatively connected to the data input end of the multimodal data acquisition module.

[0023] Optionally, the first wireless communication module and the second wireless communication module are both Zigbee modules.

[0024] By adopting the above technical solution, the Zigbee module has the advantages of low cost, low power consumption and small size. It can even work stably and continuously for several years under battery power. Due to its small size, it will not cause interference to the target device after deployment. The Zigbee module is greatly affected by the electromagnetic interference caused by partial discharge. Therefore, the detection of partial discharge can be achieved by receiving a standard data packet from another Zigbee sending module through the Zigbee sending module. The change analysis of signal strength is mainly based on the parsing result of the standard data packet. The standard data packet can be a small digital segment data, such as an Arabic data set from 1 to 10. The standard data packet is parsed to obtain the RSSI value of the wireless network transmitter. The structure where the RSSI is located is aflncomingMSGPacket_t. In this structure, there are two variables related to the communication quality. They are: aflncomingMSGPacket_t->rssi and rssirssiafincominaMSGPacket_t>inkQuality, where rssi: receivedsignalst rengthindicator is the received signal strength indication, which represents the signal strength value. The signal strength value is obtained based on the RSSI value. If a standard data packet is sent and received every 0.2 seconds, the signal strength change value can be obtained by comparing the difference in RSSI values ​​of the two standard data packets. The wireless communication signal attenuation value is set to 30%. For example, if the signal strength change value is greater than 30%, it can be determined that the signal strength change exceeds the standard.

[0025] Optionally, the device-side autonomous discharge handling module includes a data storage device, a data analysis chip and a visual analysis chip, the data storage device is communicatively connected to the data output end of the multimodal data acquisition module, and the data analysis chip and the visual analysis chip are respectively communicatively connected to the data storage device.

[0026] By adopting the above technical solution, the main visual feature of partial discharge is the luminescence phenomenon, and the visual analysis chip can realize the recognition of the luminescence visual features.

[0027] Optionally, the device-side discharge autonomous handling module also includes an audible and visual alarm. When the data analysis chip outputs an edge judgment that a diagnosis result of local discharge occurs, an alarm signal is exchanged with the audible and visual alarm.

[0028] By adopting the above technical solution, it is possible to implement an audible and visual alarm when a diagnosis result of partial discharge is determined at the output edge, thereby reminding the staff to take measures.

[0029] The local discharge intelligent diagnosis method based on multimodal signal fusion uses a local discharge intelligent diagnosis system based on multimodal signal fusion to detect the local discharge phenomenon of the target equipment, including the following steps:

[0030] Step 1: The device-side discharge autonomous disposal module analyzes the detection data packets of the ultrasonic detection module, the electromagnetic wave signal detection module, the visual detection module, and the wireless communication attenuation detection module respectively;

[0031] Step 2: The device-side discharge autonomous disposal module sets an ultrasonic change threshold, a UHF electromagnetic wave signal threshold, and a wireless communication signal attenuation threshold;

[0032] Step 3, setting three exceeding standard items: the ultrasonic current detection value exceeds the ultrasonic change threshold, the ultra-high frequency electromagnetic wave signal current detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and the wireless communication signal attenuation value exceeds the wireless communication signal attenuation threshold. If the device-side discharge autonomous disposal module determines that at least one of the three exceeding standard items occurs, and at the same time the device-side discharge autonomous disposal module analyzes the characteristics of the visual image inside the device and finds that a luminous feature occurs, then the output edge judgment shows that a partial discharge diagnosis result occurs, and the device-side discharge autonomous disposal module sends an interactive power-off control signal to the controller of the target device;

[0033] If it is determined that more than two of the three exceeded items appear, but the feature analysis of the visual image inside the device does not show any luminous features, the device-side discharge autonomous disposal module outputs the diagnosis result of partial discharge due to edge judgment, and sends a power-off control signal to the controller of the target device;

[0034] Step 4: The controller of the target device performs power-off control.

[0035] By adopting the above technical solution, assuming that the background noise level recorded by the ultrasonic detection module we use is 20dB under normal circumstances, and any signal exceeding this level by 10dB is considered a potential partial discharge signal. Then, the ultrasonic change threshold can be set to 30dB;

[0036] This means that if the detected ultrasonic signal strength exceeds 30dB, it is judged that the ultrasonic signal strength exceeds the standard and the system believes that partial discharge may have occurred.

[0037] For UHF electromagnetic wave signals, assuming that under normal circumstances, the detected signal strength is 1μV / m, and any signal exceeding 50 times this strength is considered a sign of partial discharge. Then, the UHF electromagnetic wave signal threshold can be set to 50μV / m;

[0038] If the detected UHF electromagnetic wave signal strength exceeds 50μV / m, the UHF electromagnetic wave signal is judged to be out of standard and the system will believe that partial discharge may have occurred.

[0039] The attenuation value of the wireless communication signal is 30%. If the signal strength change value is greater than 30%, it can be determined that the signal strength change exceeds the standard, and the system will believe that partial discharge may have occurred.

[0040] Combined with the recognition results of the luminous characteristics, more accurate partial discharge detection results of the target equipment can be achieved, and timely power-off measures can be taken to avoid serious damage to electrical components of the target equipment due to partial discharge.

[0041] Optionally, in step 3, when the device-side discharge autonomous disposal module outputs an edge judgment that a diagnosis result of partial discharge has occurred, an alarm signal is exchanged with the sound and light alarm device, and the sound and light alarm device performs a sound and light alarm action.

[0042] In summary, the present invention includes at least one of the following beneficial technical effects:

[0043] The present invention can provide a local discharge intelligent diagnosis system and method based on multimodal signal fusion. By fusing ultrasonic, electromagnetic and visual signals, the local discharge intelligent diagnosis system can detect local discharge from multiple dimensions, effectively improving the accuracy and reliability of detection. The comprehensive analysis of multimodal data reduces the misdiagnosis and missed diagnosis that may be caused by a single signal.

[0044] The device-side discharge autonomous disposal module can detect in real time and make edge judgments based on preset thresholds to quickly identify partial discharge events. Only thresholds are used for judgment, which requires less calculation and can achieve independent edge arrangement. Real-time monitoring helps to detect anomalies in a timely manner and send power-off control signals to the controller of the target device. Rapid response can prevent the expansion of partial discharge and cause greater damage to the electrical components in the equipment. Timely detection and disposal of partial discharge can prevent equipment damage and safety accidents caused by discharge, and improve the safety and reliability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the connection principle of electrical components of the partial discharge intelligent diagnosis system based on multi-modal signal fusion of the present invention;

[0046] Figure 2 It is a flow chart of the partial discharge intelligent diagnosis method based on multimodal signal fusion of the present invention.

[0047] Explanation of the accompanying drawings: 11. Ultrasonic detection module; 111. Ultrasonic sensor mounting base; 112. Ultrasonic sensor; 113. Filter; 114. Ultrasonic data buffer; 12. Electromagnetic wave signal detection module; 121. UHF partial discharge sensor; 122. UHF data buffer; 13. Visual detection module; 131. Visual camera; 132. Visual data buffer; 14. Wireless communication attenuation detection module; 141. First wireless communication module; 142. Second wireless communication module; 143. Wireless communication data buffer; 2. Multimodal data acquisition module; 3. Equipment-side discharge autonomous disposal module; 31. Data storage device; 32. Data analysis chip; 33. Visual analysis chip; 34. Sound and light alarm. DETAILED DESCRIPTION

[0048] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0049] The embodiment of the present invention discloses a partial discharge intelligent diagnosis system and method based on multi-modal signal fusion.

[0050] Reference Figure 1 and Figure 2 Embodiment 1, a partial discharge intelligent diagnosis system based on multimodal signal fusion, comprising a device-side multimodal partial discharge detection unit, a multimodal data acquisition module 2 and a chip-based device-side discharge autonomous disposal module 3, the device-side multimodal partial discharge detection unit comprising an ultrasonic detection module 11, an electromagnetic wave signal detection module 12 and a visual detection module 13, the ultrasonic detection module 11 is used to detect ultrasonic signal changes inside the target device, the electromagnetic wave signal detection module 12 is used to detect ultra-high frequency electromagnetic wave signals generated inside the target device, the visual detection module 13 captures the visual picture inside the target device, the ultrasonic detection module 11, the electromagnetic wave signal detection module 12 and the visual detection module 13 are respectively connected to the signal input end of the multimodal data acquisition module 2 The signal output end of the multimodal data acquisition module 2 is communicatively connected with the device-side discharge autonomous disposal module 3. The device-side discharge autonomous disposal module 3 performs edge judgment based on the threshold whether local discharge occurs, sets the ultrasonic change threshold and the ultra-high frequency electromagnetic wave signal threshold, and if it is judged that the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current ultra-high frequency electromagnetic wave signal detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and at the same time, a luminous feature appears in the feature analysis of the visual image inside the device, then the output edge judgment shows that a diagnosis result of local discharge occurs, and the device-side discharge autonomous disposal module 3 is communicatively connected with the controller of the target device. When the edge judgment of the device-side discharge autonomous disposal module 3 shows a diagnosis result of local discharge, a power-off control signal is exchanged with the controller of the target device.

[0051] By integrating ultrasonic, electromagnetic and visual signals, the PD intelligent diagnosis system can detect PD from multiple dimensions, effectively improving the accuracy and reliability of detection. The comprehensive analysis of multimodal data reduces the misdiagnosis and missed diagnosis that may be caused by a single signal.

[0052] The device-side discharge autonomous handling module 3 performs edge judgment based on the threshold to determine whether partial discharge occurs, sets the ultrasonic change threshold and the ultra-high frequency electromagnetic wave signal threshold, and if it is determined that the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current ultra-high frequency electromagnetic wave signal detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and at the same time, the feature analysis of the visual image inside the device shows a luminous feature, then the diagnosis result of partial discharge occurrence in the edge judgment is output;

[0053] When the luminous feature is detected, as long as the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current UHF electromagnetic wave signal detection value exceeds the UHF electromagnetic wave signal threshold, the probability of partial discharge in the target device is very high, and the diagnosis result of partial discharge can be outputted, and the power-off signal can be outputted at the same time. The device-side discharge autonomous disposal module 3 can detect in real time and make edge judgments based on preset thresholds to quickly identify partial discharge events. Because only thresholds are used for judgment, the amount of calculation is small, and the independent arrangement of the edge can be realized. Real-time monitoring helps to detect abnormalities in time, and the power-off control signal is exchanged to the controller of the target device. The rapid response can prevent the expansion of partial discharge phenomena and cause greater damage to the electrical components in the device.

[0054] Timely detection and treatment of partial discharge can prevent equipment damage and safety accidents caused by discharge, and improve the safety and reliability of the system.

[0055] Embodiment 2, the ultrasonic detection module 11 includes an ultrasonic sensor mounting base 111, an ultrasonic sensor 112, a filter 113 and an ultrasonic data buffer 114, one end of the ultrasonic sensor mounting base 111 is mounted at the top middle position of the inner wall of the device shell of the target device, the other end of the ultrasonic sensor mounting base 111 is provided with a sensor mounting internal thread, the ultrasonic sensor 112 is screwed on the sensor mounting internal thread, and the ultrasonic signal of the device shell of the target device is received through the ultrasonic sensor mounting base 111, the ultrasonic sensor 112 is communicatively connected to the filter 113, the filter 113 is communicatively connected to the ultrasonic data buffer 114, and the ultrasonic data buffer 114 is communicatively connected to the data input end of the multimodal data acquisition module 2.

[0056] Partial discharge will generate ultrasonic signals. When partial discharge occurs inside the target device, the electrons in the discharge area move at high speed and collide with the surrounding medium, causing the medium to expand instantly due to heat, thereby generating mechanical stress waves, that is, ultrasonic waves. These ultrasonic signals will propagate inside the device in the form of waves. The ultrasonic sensor mounting seat 111 is combined with the ultrasonic sensor 112 to receive these signals, and finally filtered by the filter 113 and transmitted to the modal data acquisition module 2.

[0057] Embodiment 3, the electromagnetic wave signal detection module 12 includes a UHF partial discharge sensor 121 and a UHF data buffer 122, the UHF partial discharge sensor 121 is arranged on the periphery of the main cable inside the target device, the UHF partial discharge sensor 121 is communicatively connected with the UHF data buffer 122, and the UHF data buffer 122 is communicatively connected with the data input end of the multimodal data acquisition module 2.

[0058] The high frequency partial discharge sensor 121 detects whether partial discharge occurs by coupling the ultra-high frequency electromagnetic wave signal generated by partial discharge in the electrical system. The basic principle of the ultra-high frequency detection method is to use the ultra-high frequency partial discharge sensor to detect the ultra-high frequency electromagnetic wave signal (the frequency of the electromagnetic wave is about 300MHz to 3GHz) generated by partial discharge in the power equipment.

[0059] Embodiment 4, the visual detection module 13 includes a visual camera 131 and a visual data buffer 132, the visual camera 131 is installed on the inner wall of the shell of the target device through a bracket to capture the internal visual image of the target device, the visual camera 131 is communicatively connected with the visual data buffer 132, and the visual data buffer 132 is communicatively connected with the data input end of the multimodal data acquisition module 2.

[0060] The visual camera 131 can realize the recognition of luminous features.

[0061] Embodiment 5, the multimodal partial discharge detection unit at the device end also includes a wireless communication attenuation detection module 14, the wireless communication attenuation detection module 14 includes a first wireless communication module 141, a second wireless communication module 142 and a wireless communication data buffer 143, the first wireless communication module 141 and the second wireless communication module 142 are respectively installed on the inner wall of the shell of the target device through a bracket, the first wireless communication module 141 and the second wireless communication module 142 are wirelessly connected, and wirelessly communicate with each other to exchange standard data packets at set intervals, the first wireless communication module 141 and the second wireless communication module 142 store the received standard data packets in the wireless communication data buffer 143, and the wireless communication data buffer 143 is communicatively connected to the data input end of the multimodal data acquisition module 2.

[0062] In Embodiment 6, the first wireless communication module 141 and the second wireless communication module 142 are both Zigbee modules.

[0063] The Zigbee module has the advantages of low cost, low power consumption and small size. It can even work stably for several years under battery power. Due to its small size, it will not interfere with the target device after deployment. The Zigbee module is greatly affected by the electromagnetic interference caused by partial discharge. Therefore, the detection of partial discharge can be achieved by receiving the standard data packet of another Zigbee sending module through the Zigbee sending module. The change analysis of signal strength is mainly based on the parsing results of the standard data packet. The standard data packet can be a small digital segment data, such as an Arabic data set from 1 to 10. The standard data packet is parsed to obtain the RSSI value of the wireless network transmitter. The structure where the RSSI is located is aflncomingMSGPacket_t. In this structure, there are two variables related to the communication quality. They are: aflncomingMSGPacket_t->rssi and rssirssiafincominaMSGPacket_t>inkQuality, where rssi: receivedsignalstrengthindicator is the received signal strength indicator, which represents the signal strength value. The signal strength value is obtained based on the RSSI value. If a standard data packet is sent and received every 0.2 seconds, the signal strength change value can be obtained by comparing the difference in RSSI values ​​of the two standard data packets. The wireless communication signal attenuation value is set to 30%. For example, if the signal strength change value is greater than 30%, it can be determined that the signal strength change exceeds the standard.

[0064] Embodiment 7, the device-side autonomous discharge handling module 3 includes a data storage device 31, a data analysis chip 32 and a visual analysis chip 33. The data storage device 31 is communicatively connected to the data output end of the multimodal data acquisition module 2, and the data analysis chip 32 and the visual analysis chip 33 are communicatively connected to the data storage device 31 respectively.

[0065] The main visual feature of partial discharge is the luminescence phenomenon, and the visual analysis chip 33 can realize the recognition of the luminescence visual feature.

[0066] In Embodiment 8, the device-side discharge autonomous handling module 3 further includes an audible and visual alarm 34 . When the data analysis chip 32 outputs a diagnosis result of edge judgment indicating the occurrence of partial discharge, an alarm signal is exchanged with the audible and visual alarm 34 .

[0067] When the diagnosis result of partial discharge is judged at the output edge, an audible and visual alarm can be issued to remind the staff to deal with it.

[0068] Embodiment 9, a partial discharge intelligent diagnosis method based on multimodal signal fusion, uses a partial discharge intelligent diagnosis system based on multimodal signal fusion to detect partial discharge phenomena of a target device, comprising the following steps:

[0069] Step 1, the device-side discharge autonomous disposal module 3 analyzes the detection data packets of the ultrasonic detection module 11, the electromagnetic wave signal detection module 12, the visual detection module 13, and the wireless communication attenuation detection module 14 respectively;

[0070] Step 2, the device-side discharge autonomous disposal module 3 sets an ultrasonic change threshold, a UHF electromagnetic wave signal threshold, and a wireless communication signal attenuation threshold;

[0071] Step 3, setting three exceeding standard items: the ultrasonic current detection value exceeds the ultrasonic change threshold, the ultra-high frequency electromagnetic wave signal current detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and the wireless communication signal attenuation value exceeds the wireless communication signal attenuation threshold. If the device-side discharge autonomous disposal module 3 determines that at least one of the three exceeding standard items occurs, and at the same time the device-side discharge autonomous disposal module 3 analyzes the characteristics of the visual picture inside the device and finds that a luminous feature occurs, then the output edge determines that a diagnosis result of partial discharge occurs, and the device-side discharge autonomous disposal module 3 sends an interactive power-off control signal to the controller of the target device;

[0072] If it is determined that more than two of the three exceeded items appear, but the characteristic analysis of the visual image inside the device does not show the luminous feature, the device-side discharge autonomous disposal module 3 outputs the diagnosis result of partial discharge in edge judgment, and sends a power-off control signal to the controller of the target device;

[0073] Step 4: The controller of the target device performs power-off control.

[0074] Assume that the background noise level recorded by the ultrasonic detection module we use is 20dB under normal circumstances, and any signal exceeding this level by 10dB is considered a potential partial discharge signal. Then, the ultrasonic change threshold can be set to 30dB;

[0075] This means that if the detected ultrasonic signal strength exceeds 30dB, it is judged that the ultrasonic signal strength exceeds the standard and the system believes that partial discharge may have occurred.

[0076] For UHF electromagnetic wave signals, assuming that under normal circumstances, the detected signal strength is 1μV / m, and any signal exceeding 50 times this strength is considered a sign of partial discharge. Then, the UHF electromagnetic wave signal threshold can be set to 50μV / m;

[0077] If the detected UHF electromagnetic wave signal strength exceeds 50μV / m, the UHF electromagnetic wave signal is judged to be out of standard and the system will believe that partial discharge may have occurred.

[0078] The attenuation value of the wireless communication signal is 30%. If the signal strength change value is greater than 30%, it can be determined that the signal strength change exceeds the standard, and the system will believe that partial discharge may have occurred.

[0079] Combined with the recognition results of the luminous characteristics, more accurate partial discharge detection results of the target equipment can be achieved, and timely power-off measures can be taken to avoid serious damage to electrical components of the target equipment due to partial discharge.

[0080] In Example 10, in step 3, when the device-side discharge autonomous handling module 3 outputs the diagnosis result of edge judgment that partial discharge occurs, an alarm signal is exchanged to the sound and light alarm 34, and the sound and light alarm 34 performs a sound and light alarm action.

[0081] The following is a Python environment code example based on the partial discharge intelligent diagnosis method.

[0082] #Assumed threshold setting

[0083] ultrasonic_threshold=30#Ultrasonic change threshold, unit: dB

[0084] uhf_threshold=50e-6#UHF electromagnetic wave signal threshold, unit V / m

[0085] wireless_attenuation_threshold=30#Wireless communication signal attenuation threshold, unit: %

[0086] #Assumed detection packet parsing function

[0087] def parse_detection_data(ultrasonic_data,uhf_data,visual_dat a,wireless_data):

[0088] #Parse the data packet and return the detection value

[0089] ultrasonic_value=ultrasonic_data['value']

[0090] uhf_value = uhf_data['value']

[0091] visual_feature=visual_data['feature']

[0092] wireless_attenuation=wireless_data['attenuation']

[0093] return ultrasonic_value,uhf_value,visual_feature,wireless ss_attenuation

[0094] #Hypothetical visual feature analysis function

[0095] def analyze_visual_feature(visual_data):

[0096] #Analyze visual data to determine whether luminous features appear

[0097] return 'glow' in visual_data

[0098] #Assumed control signal sending function

[0099] def send_control_signal(controller,command):

[0100] #Send control signal to the controller

[0101] print(f"Control signal sent to controller:{command}")

[0102] #Assumed sound and light alarm interaction function

[0103] def alert_alerter(alerter):

[0104] #Send an alarm signal to the sound and light alarm

[0105] print("Alert signal sent to sound and light alerter.")

[0106] #Logic of the device-side discharge autonomous disposal module

[0107] def discharge_diagnosis(ultrasonic_data,uhf_data,visual_data,wireless_data):

[0108] #Parse the detection data packet

[0109] ultrasonic_value, uhf_value, visual_feature, wireless_attenuation = parse_detection_data(ultrasonic_data, uhf_data, visual_data, wireless_data)

[0110] # Determine if it exceeds the standard

[0111] ultrasonic_exceeds = ultrasonic_value > ultrasonic_threshold

[0112] uhf_exceeds = uhf_value > uhf_threshold

[0113] wireless_exceeds = wireless_attenuation > wireless_attenuation_threshold

[0114] # Comprehensive judgment

[0115] if (ultrasonic_exceeds or uhf_exceeds or wireless_exceeds) and analyze_visual_feature(visual_feature):

[0116] # Output the diagnosis result and execute the power-off control

[0117] print("Local discharge detected. Initiating power-off sequence.")

[0118] send_control_signal(controller='device_controller', command='POWER_OFF')

[0119] alert_alerter(alerter='sound_light_alerter')

[0120] elif sum([ultrasonic_exceeds, uhf_exceeds, wireless_exceeds]) > 1:

[0121] #If more than two items exceed the limit, power off control will be executed even if there is no visual feature

[0122] print("Local discharge likely.Initiating power-off sequence.")

[0123] send_control_signal(controller='device_controller',command='POWER_OFF')

[0124] alert_alerter(alerter='sound_light_alerter')

[0125] else:

[0126] #No partial discharge

[0127] print("No local discharge detected.")

[0128] #Example detection data packet

[0129] ultrasonic_data={'value':35}#Ultrasonic detection value, unit: dB

[0130] uhf_data={'value':60e-6}#UHF electromagnetic wave signal detection value, unit V / mvisual_data={'feature':'glow detected'}#Visual feature data

[0131] wireless_data = {'attenuation':35} #Wireless communication signal attenuation value, unit: %;

[0132] #Execute discharge diagnosis

[0133] discharge_diagnosis(ultrasonic_data,uhf_data,visual_data,wireless_data).

[0134] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. Partial discharge intelligent diagnosis system based on multimodal signal fusion, characterized by: The invention comprises a device-side multimodal partial discharge detection unit, a multimodal data acquisition module (2) and a chip-based device-side discharge autonomous disposal module (3); the device-side multimodal partial discharge detection unit comprises an ultrasonic detection module (11), an electromagnetic wave signal detection module (12) and a visual detection module (13); the ultrasonic detection module (11) is used to detect ultrasonic signal changes inside the target device; the electromagnetic wave signal detection module (12) is used to detect ultra-high frequency electromagnetic wave signals generated inside the target device; the visual detection module (13) captures the visual image inside the target device; the ultrasonic detection module (11), the electromagnetic wave signal detection module (12) and the visual detection module (13) are respectively connected to the signal input end of the multimodal data acquisition module (2) for communication; The signal output end of the multimodal data acquisition module (2) is connected to the device-side discharge autonomous handling module (3) for communication. The device-side discharge autonomous handling module (3) performs edge judgment based on a threshold value to determine whether partial discharge occurs, sets an ultrasonic change threshold and an ultra-high frequency electromagnetic wave signal threshold, and if it is determined that the current ultrasonic detection value exceeds the ultrasonic change threshold and / or the current ultra-high frequency electromagnetic wave signal detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and at the same time, a luminous feature is found in the feature analysis of the visual image inside the device, then the edge judgment outputs a diagnosis result of partial discharge. The device-side discharge autonomous handling module (3) is connected to the controller of the target device for communication. When the edge judgment of the device-side discharge autonomous handling module (3) results in a diagnosis result of partial discharge, a power-off control signal is exchanged with the controller of the target device.

2. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 1 is characterized in that: The ultrasonic detection module (11) comprises an ultrasonic sensor mounting seat (111), an ultrasonic sensor (112), a filter (113) and an ultrasonic data buffer (114); one end of the ultrasonic sensor mounting seat (111) is mounted at the top middle position of the inner wall of a device housing of a target device; the other end of the ultrasonic sensor mounting seat (111) is provided with a sensor mounting internal thread; the ultrasonic sensor (112) is screwed onto the sensor mounting internal thread; an ultrasonic signal of the device housing of the target device is received via the ultrasonic sensor mounting seat (111); the ultrasonic sensor (112) is communicatively connected to the filter (113); the filter (113) is communicatively connected to the ultrasonic data buffer (114); and the ultrasonic data buffer (114) is communicatively connected to a data input end of a multimodal data acquisition module (2).

3. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 2 is characterized in that: The electromagnetic wave signal detection module (12) comprises an ultra-high frequency partial discharge sensor (121) and an ultra-high frequency data buffer (122); the ultra-high frequency partial discharge sensor (121) is arranged on the periphery of a main cable inside a device of a target device; the ultra-high frequency partial discharge sensor (121) is communicatively connected to the ultra-high frequency data buffer (122); and the ultra-high frequency data buffer (122) is communicatively connected to a data input terminal of a multi-modal data acquisition module (2).

4. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 3 is characterized in that: The visual detection module (13) comprises a visual camera (131) and a visual data buffer (132); the visual camera (131) is mounted on the inner wall of a housing of a target device via a bracket to capture a visual image of the interior of the target device; the visual camera (131) is communicatively connected to the visual data buffer (132); and the visual data buffer (132) is communicatively connected to a data input terminal of the multimodal data acquisition module (2).

5. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 4 is characterized in that: The device-side multimodal partial discharge detection unit also includes a wireless communication attenuation detection module (14), the wireless communication attenuation detection module (14) including a first wireless communication module (141), a second wireless communication module (142) and a wireless communication data buffer (143), the first wireless communication module (141) and the second wireless communication module (142) being respectively mounted on the inner wall of a shell of a target device through a bracket, the first wireless communication module (141) and the second wireless communication module (142) being wirelessly connected, exchanging standard data packets wirelessly with each other at set intervals, the first wireless communication module (141) and the second wireless communication module (142) storing received standard data packets in the wireless communication data buffer (143), and the wireless communication data buffer (143) being communicatively connected to a data input end of the multimodal data acquisition module (2).

6. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 5 is characterized in that: The first wireless communication module (141) and the second wireless communication module (142) are both Zigbee modules.

7. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 6 is characterized in that: The device-side autonomous discharge handling module (3) comprises a data storage device (31), a data analysis chip (32) and a visual analysis chip (33); the data storage device (31) is communicatively connected to a data output end of the multimodal data acquisition module (2); and the data analysis chip (32) and the visual analysis chip (33) are respectively communicatively connected to the data storage device (31).

8. The partial discharge intelligent diagnosis system based on multimodal signal fusion according to claim 7 is characterized in that: The device-side discharge autonomous handling module (3) also includes an audible and visual alarm (34), and when the data analysis chip (32) outputs a diagnosis result of edge judgment indicating the occurrence of local discharge, an alarm signal is exchanged with the audible and visual alarm (34).

9. The intelligent diagnosis method of partial discharge based on multimodal signal fusion is characterized by: The local discharge intelligent diagnosis system based on multimodal signal fusion as claimed in claim 8 is used to detect the local discharge phenomenon of the target device, comprising the following steps: Step 1, the device-side discharge autonomous handling module (3) respectively analyzes the detection data packets of the ultrasonic detection module (11), the electromagnetic wave signal detection module (12), the visual detection module (13), and the wireless communication attenuation detection module (14); Step 2, the device-side discharge autonomous handling module (3) sets an ultrasonic change threshold, a UHF electromagnetic wave signal threshold, and a wireless communication signal attenuation threshold; Step 3, setting three exceeding standard items: the ultrasonic current detection value exceeds the ultrasonic change threshold, the ultra-high frequency electromagnetic wave signal current detection value exceeds the ultra-high frequency electromagnetic wave signal threshold, and the wireless communication signal attenuation value exceeds the wireless communication signal attenuation threshold. If the device-side discharge autonomous handling module (3) determines that at least one of the three exceeding standard items occurs, and at the same time the device-side discharge autonomous handling module (3) analyzes the characteristics of the visual image inside the device and finds that a luminous feature occurs, then the output edge judgment shows that a diagnosis result of partial discharge occurs, and the device-side discharge autonomous handling module (3) sends an interactive power-off control signal to the controller of the target device; If it is determined that more than two of the three exceeded items are present, but the characteristic analysis of the visual image inside the device does not show any luminous features, the device-side discharge autonomous handling module (3) outputs a diagnosis result of partial discharge in edge judgment and sends a power-off control signal to the controller of the target device; Step 4: The controller of the target device performs power-off control.

10. The intelligent diagnosis method for partial discharge based on multimodal signal fusion according to claim 9, characterized in that: In step 3, when the device-side discharge autonomous handling module (3) outputs a diagnosis result of edge judgment that a partial discharge has occurred, an alarm signal is exchanged with the sound and light alarm (34), and the sound and light alarm (34) performs a sound and light alarm action.

Images