Identity recognition and state judgment method for high-voltage power distribution cabinet in complex environment

Abstract

The invention discloses an identity recognition and state judgment method for a high-voltage power distribution cabinet in a complex environment, and belongs to the technical field of electrical equipment internet of things. According to the method, a multi-mode sensing device is innovatively configured, a UWB identity recognition system and a composite sensing array are integrated, a door magnetic sensor and Hall sensor linkage framework is adopted for cabinet door state monitoring, and accurate sensing of the opening and closing state of a cabinet door is achieved through magnetic induction intensity threshold judgment and mechanical displacement dual-mode verification; in air switch position detection, non-contact mechanical position identification is realized by embedding an RFID positioning tag in a switch operation position and combining a multi-band reader array in the cabinet. Meanwhile, a vibration sensor and a voltage / current transformer are arranged to acquire state change information of the grounding knife and current and voltage change information after the grounding knife is switched on and off. The system transmits and fuses sensing data to the intelligent inspection robot in real time through LoRa wireless networking, and the current state of equipment is judged through processing of the embedded processing module in combination with historical data.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) technology for power equipment, specifically a method for identifying and determining the status of high-voltage distribution cabinets in complex environments. Background Technology

[0002] Existing high-voltage switchgear monitoring technologies have achieved certain results in terms of basic function implementation. Traditional door magnetic sensors have the advantages of simple structure and low cost, and can realize basic status monitoring of cabinet door opening and closing. Mechanical limit switches have the characteristics of intuitive reliability in detecting the open and closed positions of circuit breakers and air switches, and are especially suitable for low-complexity scenarios. In the early days, high-voltage switchgear relied on manual inspection, and the status was judged by indicator lights such as red light to indicate closed and green light to indicate open, and instrument readings. This was inefficient and prone to missed detections. The existing high-voltage switchgear identification and status judgment methods using traditional door magnetic sensors only rely on a single magnetic flux threshold for judgment, which is prone to misjudgment in the case of strong electromagnetic interference or cabinet vibration.

[0003] The existing methods for identifying and determining the status of high-voltage switchgear have the following drawbacks:

[0004] 1. Patent document CN120073990A discloses a remote monitoring system for high-voltage distribution cabinets based on the Internet of Things, which "includes modules such as sensor acquisition, distribution cabinet power optimization, temperature and smoke joint monitoring, and remote control; the sensor acquisition module adjusts sensor information and collects power and internal status data; the distribution cabinet power optimization module obtains power operation early warning values, divides load areas, marks risky equipment, and optimizes power according to equipment priority; the temperature and smoke joint monitoring module monitors faults through smoke and temperature parameters, the smoke detection unit calculates smoke hazard values, and the temperature detection unit locates dangerous parts; the remote control module receives maintenance signals, obtains equipment location, and assigns maintenance instructions; this invention achieves precise power optimization and early fault warning and precise location, improves the safety of high-voltage distribution cabinet operation, and has important significance in the field of power system operation management." However, the traditional door magnetic sensor of the existing high-voltage distribution cabinet identification and status judgment method only relies on a single magnetic flux threshold for judgment, which is prone to misjudgment in the case of strong electromagnetic interference or cabinet vibration.

[0005] 2. Patent document CN114268033A discloses a high-voltage distribution cabinet with remote control based on the Internet of Things, "including a cabinet body, a detection slot is opened at the top inside the cabinet body, an exhaust fan is fixedly installed inside the detection slot, a smoke sensor is fixedly installed inside the detection slot and at the top of the exhaust fan, a wireless sensor is connected to one side of the smoke sensor, the wireless sensor is connected to the main control device, and a main control board is connected to one side of the wireless sensor. By rotating the exhaust fan, the gas inside the cabinet body can be extracted in real time. Furthermore, the smoke sensor inside the detection slot will monitor the extracted gas in real time and simultaneously feed back the monitoring results to the main control device, which can further prevent the occurrence of fire." However, the existing high-voltage distribution cabinet identification and status judgment methods rely on a single vibration sensor, which cannot distinguish between vibration signals generated by normal operation of the grounding switch and accidental collisions.

[0006] 3. Patent document CN103837263A discloses a fiber optic temperature measurement system for high-voltage distribution cabinets based on IoT remote monitoring. It "includes a temperature sensor, which is connected to a switching circuit. The switching circuit is connected to a signal processing circuit, which is connected to a computer control center. The network server of the computer control center accesses an IoT computer via the Internet. The network server of the computer control center also accesses an IoT mobile phone via a 3G Wi-Fi network. The IoT mobile phone is equipped with a sensor information storage and display module, a networked mobile phone information transmission module, and a networked computer information transmission module. The network server is equipped with a network service module. The computer control center is equipped with an IoT mobile phone user management module. The IoT computer is equipped with an information receiving and processing module. Both the IoT mobile phone and the IoT computer are equipped with a network controller information transmission module. This invention realizes temperature measurement inside the high-voltage distribution cabinet and remote monitoring based on IoT." However, the existing threshold algorithms for high-voltage distribution cabinet identification and status judgment are difficult to integrate multi-source heterogeneous data to achieve switch cabinet status prediction. Summary of the Invention

[0007] The purpose of this invention is to provide a method for identifying and determining the status of high-voltage distribution cabinets in complex environments, in order to solve the technical problem in the existing high-voltage distribution cabinet identification and status determination methods mentioned in the background art that traditional door magnetic sensors rely solely on a single magnetic flux threshold for judgment, which is prone to misjudgment in scenarios with strong electromagnetic interference or cabinet vibration.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a method for identification and status determination of high-voltage distribution cabinets in complex environments, comprising:

[0009] S1. Information collection configuration of a multimodal sensing device, wherein the multimodal sensing device integrates a UWB identity recognition system and a composite sensor array;

[0010] S2. Monitor the cabinet door status through a linkage architecture of door magnetic sensor and Hall sensor, and achieve accurate perception of cabinet door opening and closing status based on magnetic induction intensity threshold judgment and mechanical displacement dual-mode verification.

[0011] S3. Embed RFID positioning tags in the air switch operating position, and combine them with the multi-band reader array in the cabinet to achieve non-contact mechanical position identification of the air switch position;

[0012] S4. Install vibration sensors to obtain information on changes in the status of the grounding blade;

[0013] S5. Real-time transmission of fused sensor data to computing devices via LoRa wireless networking, combined with historical data and processed by embedded processing modules for status analysis, spatiotemporal alignment of real-time monitoring data and historical operating data, and construction of device status prediction model based on LSTM network.

[0014] S6. By collecting data from the high-voltage distribution cabinet through a network, the system inputs the data into the status diagnosis model, performs matching analysis based on preset thresholds and historical data, and outputs the health status level of the high-voltage distribution cabinet.

[0015] Preferably, in S2, a three-axis Hall sensor array is set on the cabinet door frame, a door magnetic sensor is set on the cabinet door, and a permanent magnet calibration block is set on the movable end of the cabinet door. When the cabinet door displacement exceeds a set threshold, the door magnetic sensor is triggered, and at the same time, the magnetic flux change data of the Hall sensor is collected.

[0016] Preferably, in S3, the RFID positioning tags include RFID positioning tags of different frequency bands embedded in the closing and opening positions of the circuit breaker operating handle, and a dual-polarized multi-band reader array is symmetrically arranged inside the cabinet. The actual position of the operating handle is determined by a joint positioning algorithm of radio frequency signal strength RSSI and phase difference. S3 also includes an air switch positioning unit, which is a dual-frequency RFID tag and a dual-polarized multi-band reader array.

[0017] Preferably, a piezoelectric vibration sensor is installed at the drive shaft of the grounding switch in S4, and a deep learning algorithm is used to extract the operational impact characteristic spectrum.

[0018] Preferably, the UWB identification system in S1 includes a UWB anchor array deployed on the top of the high-voltage distribution cabinet. The UWB anchor array has a frequency band of 6-8GHz and achieves spatial positioning accuracy of ±10cm. The UWB signal receiver is set on the robot to determine whether the identity of the distribution cabinet reached by the robot is correct.

[0019] Preferably, the triaxial Hall sensor array, the permanent magnet calibration block, and the displacement calculation chip constitute a Hall linkage sensor group, and the displacement calculation chip integrates a Kalman filter algorithm.

[0020] Preferably, the piezoelectric vibration sensor has a built-in frequency response compensation circuit.

[0021] Preferably, the health status level of the high-voltage distribution cabinet in S6 includes three levels: normal level 1, warning level 2, and fault level 3. The high-voltage distribution cabinet lights indicate the following: Level 1: green indicator light is always on; Level 2: yellow indicator light flashes to indicate the maintenance cycle; Level 3: red indicator light rotates and flashes to trigger the circuit breaker trip protection.

[0022] Preferably, S5 further includes an adaptive transmission module and an intelligent analysis module.

[0023] Preferably, the adaptive transmission module includes a multi-hop relay node and an edge computing node. The multi-hop relay node supports the TDMA (Time Division Multiple Access) protocol, and the edge computing node performs data compression and feature extraction. The intelligent analysis module includes a spatiotemporal alignment database and an LSTM (Laser-Based Time-Series Prediction Network). The spatiotemporal alignment database stores historical operating data and device parameters, and the LSTM has an input dimension of 12 and 128 hidden nodes.

[0024] Compared with the prior art, the beneficial effects of the present invention are:

[0025] 1. This invention uses an array of UWB anchor points installed at the four corners of the top of a high-voltage distribution cabinet to form a diamond-shaped positioning network with a side length of 80cm. The anchor points are powered by PoE. The UWB signal receivers are arranged on the robot and between the robot and the anchor points. The TDOA algorithm is used to achieve three-dimensional positioning. The positioning data is filtered by Kalman to generate a real-time trajectory with an accuracy of ±10cm. Three sets of three-axis Hall sensors are installed horizontally on the cabinet door fixing frame, which realizes the function of reducing misjudgment in the case of strong electromagnetic interference or cabinet vibration.

[0026] 2. This invention utilizes a neodymium iron boron permanent magnet embedded in a corresponding movable end, maintaining a 5mm±0.2mm gap with the Hall sensor. The door magnetic sensor is a normally open reed switch, which is encapsulated in epoxy resin for shock protection. The displacement calculation chip is an STM32F4, which integrates a Kalman filter algorithm to output the cabinet door's three-dimensional displacement vectors ΔX, ΔY, ΔZ and the confidence level of the opening and closing status in real time. This enables the differentiation between vibration signals generated by normal operation of the floor door and those generated by accidental collisions.

[0027] 3. This invention embeds 125kHz low-frequency RFID tags and 13.56MHz high-frequency tags into the closing and opening positions of the circuit breaker operating handle, respectively. The low-frequency RFID tags are NXPNTAG213 and TITRPGR30TGC, respectively. The tag surface is covered with a 0.1mm polyimide insulating film. A dual-polarization multi-band reader array is symmetrically arranged at the four corners of the cabinet. The dual-polarization multi-band reader array is ImpinjR420, which achieves non-contact positioning through the LANDMARC algorithm. Three sets of piezoelectric vibration sensors are installed at 120° radially evenly spaced points on the drive shaft of the ground switch. The piezoelectric vibration sensors are PCB352C33, with a frequency response of 0.1-1kHz. The signal is conditioned by a programmable gain amplifier and then connected to a 16-bit ADC module, realizing the function of predicting the status of the switch cabinet by integrating multi-source heterogeneous data.

[0028] 4. This invention utilizes four sets of LoRa multi-hop relay nodes, each equipped with a Semtech SX1276 chip, SF=10, bandwidth 125kHz, and node spacing 50cm. Information is preprocessed using the TDMA (Time Division Multiple Access) protocol, and the preprocessed data packets are encrypted and transmitted to the host computer via the IPv6 over LoRaWAN protocol. The end-to-end transmission delay is <100ms. The spatiotemporal alignment database uses the NTP protocol to synchronize the clocks of each sensor, with a clock deviation of 0.3~1ms. Multidimensional datasets are stored in 5~8 minute time windows, including UWB information, Hall displacement variance with a resolution of 0.1mm, RFID positioning coordinates (X, Y, Z), and short-time energy (STE) of vibration signals. This invention effectively addresses the application bottlenecks in complex industrial scenarios. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the system flow of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front end," "rear end," "both ends," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0032] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand this according to the specific circumstances.

[0033] Example 1: Please refer to Figure 1 The present invention provides an embodiment of a method for identifying and determining the status of high-voltage distribution cabinets in complex environments, comprising:

[0034] S1. Information collection configuration of a multimodal sensing device, wherein the multimodal sensing device integrates a UWB identity recognition system and a composite sensor array;

[0035] S2. Monitor the cabinet door status through a linkage architecture of door magnetic sensor and Hall sensor, and achieve accurate perception of cabinet door opening and closing status based on magnetic induction intensity threshold judgment and mechanical displacement dual-mode verification.

[0036] S3. Embed RFID positioning tags in the air switch operating position, and combine them with the multi-band reader array in the cabinet to achieve non-contact mechanical position identification of the air switch position;

[0037] S4. Install vibration sensors to obtain information on changes in the status of the grounding blade;

[0038] S5. Real-time transmission of fused sensor data to computing devices via LoRa wireless networking, combined with historical data and processed by embedded processing modules for status analysis, spatiotemporal alignment of real-time monitoring data and historical operating data, and construction of device status prediction model based on LSTM network.

[0039] S6. By collecting data from the high-voltage distribution cabinet through a network, input the data into the status diagnosis model, perform matching analysis based on preset thresholds and historical data, and output the health status level of the high-voltage distribution cabinet.

[0040] In S2, a triaxial Hall sensor array is installed on the cabinet door frame, a door magnetic sensor is installed on the cabinet door, and a permanent magnet calibration block is installed at the movable end of the cabinet door. When the cabinet door displacement exceeds a set threshold, the door magnetic sensor is triggered, and the magnetic flux change data of the Hall sensor is collected simultaneously. In S3, the RFID positioning tags include RFID positioning tags of different frequency bands embedded in the closing and opening positions of the circuit breaker operating handle. A dual-polarization multi-frequency reader array is symmetrically arranged inside the cabinet. The actual position of the operating handle is determined by a joint positioning algorithm of radio frequency signal strength RSSI and phase difference. S3 also includes an air switch positioning unit, which is a dual-frequency RFID tag and a dual-polarization multi-frequency reader array. In S4, a piezoelectric vibration sensor is installed at the drive shaft of the grounding switch. The device employs a deep learning algorithm to extract the operational impact characteristic spectrum. The UWB identification system in S1 includes a UWB anchor array deployed on the top of the high-voltage distribution cabinet. The UWB anchor array operates in the 6-8GHz frequency band, achieving spatial positioning accuracy of ±10cm. A UWB signal receiver is mounted on the robot to determine whether the robot has correctly identified the distribution cabinet it has arrived at. The three-axis Hall sensor array, permanent magnet calibration block, and displacement calculation chip form a Hall linkage sensor group. The displacement calculation chip integrates a Kalman filter algorithm. The piezoelectric vibration sensor has a built-in frequency response compensation circuit. In S6, the high-voltage distribution cabinet's health status is classified into three levels: normal level 1, warning level 2, and fault level 3. The high-voltage distribution cabinet's indicator light represents level 1. Level 1: Green indicator light is constantly on; Level 2: Yellow indicator light flashes, indicating maintenance cycle; Level 3: Red indicator light rotates and flashes, triggering circuit breaker tripping protection. The S5 also includes an adaptive transmission module and an intelligent analysis module. The adaptive transmission module includes multi-hop relay nodes and edge computing nodes. The multi-hop relay nodes support the TDMA (Time Division Multiple Access) protocol, and the edge computing nodes perform data compression and feature extraction. The intelligent analysis module includes a spatiotemporal alignment database and an LSTM (Laser-Based Time-Series Prediction) network. The spatiotemporal alignment database stores historical operating data and equipment parameters. The LSTM network has an input dimension of 12 and 128 hidden nodes. UWB anchor point arrays are installed at the four corners of the top of the high-voltage distribution cabinet, forming a diamond-shaped positioning network with sides of 80cm. The anchor points are connected via... PoE power supply, UWB signal receiver is placed on the robot between the robot and the anchor point, and TDOA algorithm is used to realize three-dimensional positioning. The positioning data is filtered by Kalman to generate a real-time trajectory with an accuracy of ±10cm. Three sets of triaxial Hall sensors are installed horizontally on the fixed frame of the cabinet door, and neodymium iron boron permanent magnets are embedded in the corresponding moving ends, maintaining a distance of 5mm±0.2mm from the Hall sensors. The door magnetic sensor is a normally open reed switch. The door magnetic sensor is encapsulated with epoxy resin for shock protection. The displacement calculation chip is STM32F4, which integrates Kalman filtering algorithm and outputs the three-dimensional displacement vector ΔX, ΔY, ΔZ of the cabinet door and the confidence level of the opening and closing status in real time. The closing and opening positions of the air switch operating handle are embedded with 125kHz low frequency RFID tags and 13.The 56MHz high-frequency tag, low-frequency RFID tag, and high-frequency tag are NXPNTAG213 and TITRPGR30TGC respectively. The tag surface is covered with a 0.1mm polyimide insulating film. A dual-polarization multi-band reader array (ImpinjR420) is symmetrically arranged at the four corners of the cabinet, achieving contactless positioning via the LANDMARC algorithm. Three sets of piezoelectric vibration sensors (PCB352C33) are installed at 120° radial division points on the ground switch drive shaft, with a frequency response of [missing information]. The 0.1-1kHz signal is conditioned by a programmable gain amplifier and then fed into a 16-bit ADC module. Four LoRa multi-hop repeater nodes are deployed, using Semtech SX1276 chips with SF=10, bandwidth of 125kHz, and node spacing of 50cm. Information is preprocessed using TDMA (Time Division Multiple Access) protocol, and the preprocessed data packets are encrypted and transmitted to the host computer via IPv6 over LoRaWAN protocol, with an end-to-end transmission delay of <100ms. The spatiotemporal alignment database uses the NTP protocol. The clocks of each sensor are synchronized using the NTP protocol with a deviation of 0.3~1ms. A multidimensional dataset is stored in 5~8 minute time windows, containing UWB information, Hall displacement variance with 0.1mm resolution, RFID positioning coordinates (X, Y, Z), and short-time energy (STE) of the vibration signal. First, the aligned data is used to generate a standardized feature vector containing UWB information, Hall displacement variance, RFID positioning coordinates, and short-time energy of the vibration signal. The input layer of the LSTM time-series prediction model receives the standardized feature vector within a 60~70 second time window. After processing by a bidirectional LSTM network, the current operating status of the cabinet is output. The database for training the network is generated by collecting input data from a data acquisition system, followed by manual labeling of the cabinet status to form a complete dataset. This high-voltage distribution cabinet identification and status judgment method transmits fused sensor data to an intelligent inspection robot in real time via LoRa wireless networking. Combined with historical data, the data is processed by an embedded processing module to determine the current status of the equipment, achieving rapid and accurate understanding of the robot's status in high-voltage distribution rooms and solving its application bottleneck in complex industrial scenarios.

[0041] Example 2: Please refer to Figure 1One embodiment of this invention involves installing UWB anchor point arrays at the four corners of the top of a high-voltage distribution cabinet, forming a rhomboid positioning network with sides of 80cm. The anchor points are powered via PoE. A UWB signal receiver is positioned on a robot, and 3D positioning is achieved between the receiver and the anchor points using the TDOA algorithm. The positioning data is then filtered by Kalman filtering to generate a real-time trajectory with an accuracy of ±10cm. Three sets of triaxial Hall sensors are horizontally installed on the cabinet door's fixed frame, with neodymium iron boron permanent magnets embedded at the corresponding moving ends, maintaining a distance of 5mm ±0.2mm from the Hall sensors. The door magnetic sensor is a normally open type. The reed switch and door magnetic sensor are encapsulated in epoxy resin for shock protection. The displacement calculation chip is an STM32F4, which integrates a Kalman filter algorithm to output the cabinet door's three-dimensional displacement vectors ΔX, ΔY, and ΔZ, as well as the confidence level of the opening and closing status in real time. The closing and opening positions of the air switch operating handle are embedded with 125kHz low-frequency RFID tags and 13.56MHz high-frequency tags, respectively. The low-frequency RFID tag is NXPNTAG213, and the high-frequency tag is TITRPGR30TGC. The tag surface is covered with 0.1mm polyamide. The cabinet features an imide insulating film and a symmetrically arranged dual-polarized multi-band reader array (ImpinjR420) at its four corners, employing the LANDMARC algorithm for non-contact positioning. Three piezoelectric vibration sensors (PCB352C33) with a frequency response of 0.1-1kHz are installed at 120° radial intervals along the grounding switch drive shaft. The signals are conditioned by a programmable gain amplifier and then fed into a 16-bit ADC module. Four LoRa multi-hop repeater nodes (Sem) are also deployed. The TechSX1276 chip, with SF=10, bandwidth of 125kHz, and node spacing of 50cm, uses the TDMA (Time Division Multiple Access) protocol to preprocess information. The preprocessed data packets are then encrypted and transmitted to the host computer via the IPv6 over LoRaWAN protocol, with an end-to-end transmission delay of <100ms. The spatiotemporal alignment database uses the NTP protocol to synchronize the clocks of each sensor, with a synchronization deviation of 0.3~1ms. The multidimensional dataset is stored in 5-minute time windows, including UWB information and Hall displacement variance of 0.The system uses 1mm resolution, RFID positioning coordinates (X, Y, Z), and short-time energy (STE) of vibration signals. First, the aligned data is used to generate a standardized feature vector containing UWB information, Hall displacement variance, RFID positioning coordinates, and short-time energy of vibration signals. The input layer of the LSTM time-series prediction model receives the standardized feature vector within a 60-second time window. After processing by a bidirectional LSTM network, the current operating status of the cabinet is output. The database for training the network is generated by collecting input data from a data acquisition system, followed by manual labeling of the cabinet status to form a complete dataset. This high-voltage distribution cabinet identification and status judgment method transmits fused sensor data to an intelligent inspection robot in real time via LoRa wireless networking. Combined with historical data, the data is processed by an embedded processing module to determine the current status of the equipment, achieving rapid and accurate understanding of the robot's status in high-voltage distribution rooms and solving its application bottleneck in complex industrial scenarios.

[0042] Example 3: Please refer to Figure 1One embodiment of this invention involves installing UWB anchor point arrays at the four corners of the top of a high-voltage distribution cabinet, forming a rhomboid positioning network with sides of 80cm. The anchor points are powered via PoE. A UWB signal receiver is positioned on a robot, and 3D positioning is achieved between the receiver and the anchor points using the TDOA algorithm. The positioning data is then filtered by Kalman filtering to generate a real-time trajectory with an accuracy of ±10cm. Three sets of triaxial Hall sensors are horizontally installed on the cabinet door's fixed frame, with neodymium iron boron permanent magnets embedded at the corresponding moving ends, maintaining a distance of 5mm ±0.2mm from the Hall sensors. The door magnetic sensor is a normally open type. The reed switch and door magnetic sensor are encapsulated in epoxy resin for shock protection. The displacement calculation chip is an STM32F4, which integrates a Kalman filter algorithm to output the cabinet door's three-dimensional displacement vectors ΔX, ΔY, and ΔZ, as well as the confidence level of the opening and closing status in real time. The closing and opening positions of the air switch operating handle are embedded with 125kHz low-frequency RFID tags and 13.56MHz high-frequency tags, respectively. The low-frequency RFID tag is NXPNTAG213, and the high-frequency tag is TITRPGR30TGC. The tag surface is covered with 0.1mm polyamide. The cabinet features an imide insulating film and a symmetrically arranged dual-polarized multi-band reader array (ImpinjR420) at its four corners, employing the LANDMARC algorithm for non-contact positioning. Three piezoelectric vibration sensors (PCB352C33) with a frequency response of 0.1-1kHz are installed at 120° radial intervals along the grounding switch drive shaft. The signals are conditioned by a programmable gain amplifier and then fed into a 16-bit ADC module. Four LoRa multi-hop repeater nodes (Sem) are also deployed. The TechSX1276 chip, with SF=10, bandwidth of 125kHz, and node spacing of 50cm, uses the TDMA (Time Division Multiple Access) protocol to preprocess information. The preprocessed data packets are then encrypted and transmitted to the host computer via the IPv6 over LoRaWAN protocol, with an end-to-end transmission latency of <100ms. The spatiotemporal alignment database uses the NTP protocol to synchronize the clocks of each sensor, with a clock deviation of 0.3~1ms. The multidimensional dataset is stored in an 8-minute time window, including UWB information and Hall displacement variance of 0.The system uses 1mm resolution, RFID positioning coordinates (X, Y, Z), and short-time energy (STE) of vibration signals. First, the aligned data is used to generate a standardized feature vector containing UWB information, Hall displacement variance, RFID positioning coordinates, and short-time energy of vibration signals. The input layer of the LSTM time-series prediction model receives the standardized feature vector within a 70-second time window. After processing by a bidirectional LSTM network, the current operating status of the cabinet is output. The database for training the network is generated by collecting input data from a data acquisition system, followed by manual labeling of the cabinet status to form a complete dataset. This high-voltage distribution cabinet identification and status judgment method transmits fused sensor data to an intelligent inspection robot in real time via LoRa wireless networking. Combined with historical data, the data is processed by an embedded processing module to determine the current status of the equipment, achieving rapid and accurate understanding of the robot's status in high-voltage distribution rooms and solving its application bottleneck in complex industrial scenarios.

[0043] Example 4: Please refer to Figure 1One embodiment of this invention involves installing UWB anchor point arrays at the four corners of the top of a high-voltage distribution cabinet, forming a rhomboid positioning network with sides of 80cm. The anchor points are powered via PoE. A UWB signal receiver is positioned on a robot, and 3D positioning is achieved between the receiver and the anchor points using the TDOA algorithm. The positioning data is then filtered by Kalman filtering to generate a real-time trajectory with an accuracy of ±10cm. Three sets of triaxial Hall sensors are horizontally installed on the cabinet door's fixed frame, with neodymium iron boron permanent magnets embedded at the corresponding moving ends, maintaining a distance of 5mm ±0.2mm from the Hall sensors. The door magnetic sensor is a normally open type. The reed switch and door magnetic sensor are encapsulated in epoxy resin for shock protection. The displacement calculation chip is an STM32F4, which integrates a Kalman filter algorithm to output the cabinet door's three-dimensional displacement vectors ΔX, ΔY, and ΔZ, as well as the confidence level of the opening and closing status in real time. The closing and opening positions of the air switch operating handle are embedded with 125kHz low-frequency RFID tags and 13.56MHz high-frequency tags, respectively. The low-frequency RFID tag is NXPNTAG213, and the high-frequency tag is TITRPGR30TGC. The tag surface is covered with 0.1mm polyamide. The cabinet features an imide insulating film and a symmetrically arranged dual-polarized multi-band reader array (ImpinjR420) at its four corners, employing the LANDMARC algorithm for non-contact positioning. Three piezoelectric vibration sensors (PCB352C33) with a frequency response of 0.1-1kHz are installed at 120° radial intervals along the grounding switch drive shaft. The signals are conditioned by a programmable gain amplifier and then fed into a 16-bit ADC module. Four LoRa multi-hop repeater nodes (Sem) are also deployed. The TechSX1276 chip, with SF=10, bandwidth of 125kHz, and node spacing of 50cm, uses the TDMA (Time Division Multiple Access) protocol to preprocess information. The preprocessed data packets are then encrypted and transmitted to the host computer via the IPv6 over LoRaWAN protocol, with an end-to-end transmission delay of <100ms. The spatiotemporal alignment database uses the NTP protocol to synchronize the clocks of each sensor, with a clock deviation of 0.3~1ms. The multidimensional dataset is stored in 6-minute time windows, including UWB information and Hall displacement variance of 0.The system uses 1mm resolution, RFID positioning coordinates (X, Y, Z), and short-time energy (STE) of vibration signals. First, the aligned data is used to generate a standardized feature vector containing UWB information, Hall displacement variance, RFID positioning coordinates, and short-time energy of vibration signals. The input layer of the LSTM time-series prediction model receives the standardized feature vector within a 65-second time window. After processing by a bidirectional LSTM network, the current operating status of the cabinet is output. The database for training the network is generated by collecting input data from a data acquisition system, followed by manual labeling of the cabinet status to form a complete dataset. This high-voltage distribution cabinet identification and status judgment method transmits fused sensor data to an intelligent inspection robot in real time via LoRa wireless networking. Combined with historical data, the data is processed by an embedded processing module to determine the current status of the equipment, achieving rapid and accurate understanding of the robot's status in high-voltage distribution rooms and solving its application bottleneck in complex industrial scenarios.

[0044] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for identification and status determination of high-voltage distribution cabinets in complex environments, characterized in that, include: S1. Information collection configuration of a multimodal sensing device, wherein the multimodal sensing device integrates a UWB identity recognition system and a composite sensor array; S2. Monitor the cabinet door status through a linkage architecture of door magnetic sensor and Hall sensor, and achieve accurate perception of cabinet door opening and closing status based on magnetic induction intensity threshold judgment and mechanical displacement dual-mode verification. S3. Embed RFID positioning tags in the air switch operating position, and combine them with the multi-band reader array in the cabinet to achieve non-contact mechanical position identification of the air switch position; S4. Install vibration sensors to obtain information on changes in the status of the grounding blade; S5. Real-time transmission of fused sensor data to computing devices via LoRa wireless networking, combined with historical data and processed by embedded processing modules for status analysis, spatiotemporal alignment of real-time monitoring data and historical operating data, and construction of device status prediction model based on LSTM network. S6. By collecting data from the high-voltage distribution cabinet through a network, the system inputs the data into the status diagnosis model, performs matching analysis based on preset thresholds and historical data, and outputs the health status level of the high-voltage distribution cabinet.

2. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 1, characterized in that: In S2, a three-axis Hall sensor array is set on the cabinet door frame, a door magnetic sensor is set on the cabinet door, and a permanent magnet calibration block is set on the movable end of the cabinet door. When the cabinet door displacement exceeds a set threshold, the door magnetic sensor is triggered, and at the same time, the magnetic flux change data of the Hall sensor is collected.

3. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 1, characterized in that: The RFID positioning tags in S3 include RFID positioning tags of different frequency bands embedded in the closing and opening positions of the circuit breaker operating handle. A dual-polarized multi-band reader array is symmetrically arranged inside the cabinet. The actual position of the operating handle is determined by a joint positioning algorithm of radio frequency signal strength RSSI and phase difference. S3 also includes an air switch positioning unit, which is a dual-frequency RFID tag and a dual-polarized multi-band reader array.

4. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 1, characterized in that: A piezoelectric vibration sensor is installed at the drive shaft of the grounding switch in S4, and a deep learning algorithm is used to extract the operation impact feature spectrum.

5. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 1, characterized in that: The UWB identification system in S1 includes a UWB anchor array deployed on the top of the high-voltage distribution cabinet. The UWB anchor array has a frequency band of 6-8GHz and achieves spatial positioning accuracy of ±10cm. The UWB signal receiver is set on the robot to determine whether the robot has correctly identified the distribution cabinet it has arrived at.

6. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 2, characterized in that: The triaxial Hall sensor array, permanent magnet calibration block, and displacement calculation chip constitute a Hall linkage sensor group, and the displacement calculation chip integrates a Kalman filter algorithm.

7. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 4, characterized in that: The piezoelectric vibration sensor has a built-in frequency response compensation circuit.

8. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 1, characterized in that: The health status level of the S6 high-voltage distribution cabinet includes three levels: normal (Level 1), warning (Level 2), and fault (Level 3). The indicator lights on the high-voltage distribution cabinet represent Level 1: a constantly lit green indicator light. Level 2: Yellow indicator light flashes, indicating the maintenance cycle; Level 3: Red indicator light rotates and flashes, triggering the circuit breaker tripping protection.

9. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 1, characterized in that: The S5 also includes an adaptive transmission module and an intelligent analysis module.

10. The method for identification and status determination of high-voltage distribution cabinets in complex environments according to claim 9, characterized in that: The adaptive transmission module includes multi-hop relay nodes and edge computing nodes. The multi-hop relay nodes support the TDMA (Time Division Multiple Access) protocol, and the edge computing nodes perform data compression and feature extraction. The intelligent analysis module includes a spatiotemporal alignment database and an LSTM (Laser-Based Time-Series Prediction Network). The spatiotemporal alignment database stores historical operating data and device parameters, and the LSTM (Laser-Based Time-Series Prediction Network) has an input dimension of 12 and 128 hidden nodes.

Images