Multi-sensor-based electrical cabinet fault diagnosis and fire safety interlock protection system

Abstract

The application relates to the technical field of electrical cabinet security and protection, in particular to an electrical cabinet fault diagnosis and fire safety interlocking protection system based on multiple sensors, which comprises a data acquisition unit, which is used for determining the sensing correlation data set of each electrical device according to the cabinet layout data of the electrical cabinet and the device parameters of each electrical device in the electrical cabinet; determining an initial device correlation topology graph according to the sensing correlation data set; and performing path deduplication processing on the initial device correlation topology graph to obtain a full-device correlation topology graph. Through the correlation analysis of the temperature, smoke, arc light and other features in the abnormal feature set and the preset fault mode library, the core device determined according to the power consumption level and the node position in the full-device correlation topology graph, the application can accurately judge whether the fault belongs to a thermal runaway fault or an arc fault, and accurately locate the specific node in the device correlation network, thereby avoiding the problems of false reporting or missing reporting of traditional single sensors, and providing a reliable basis for subsequent fire protection.

Description

Technical Field

[0001] This invention relates to the field of electrical cabinet security technology, specifically to a multi-sensor-based electrical cabinet fault diagnosis and fire safety interlock protection system. Background Technology

[0002] Currently, traditional systems typically rely on single sensors such as temperature probes or smoke detectors, which cannot integrate multi-sensor data for correlation analysis. This makes it difficult to accurately distinguish between thermal runaway faults and arc faults, and is prone to false alarms or missed alarms. At the same time, traditional systems lack a node location mechanism based on the equipment correlation topology map, which makes it impossible to pinpoint the fault to a specific device or circuit. This leads to blind fire protection measures, often requiring large-scale power outages or full cabinet fire suppression, which affects the normal operation of non-faulty equipment.

[0003] Furthermore, traditional fire protection systems often trigger independently or execute in a simple, fixed sequence. For example, they may directly activate fire suppression upon detecting smoke, neglecting the safety logic of first cutting off power. This can easily lead to conflicting actions or incorrect sequence, which could escalate the fire or damage equipment. Additionally, traditional systems typically do not perform retesting after a single protection action, making it impossible to confirm whether the fault has been truly eliminated. They also lack the ability to automatically call upon surrounding emergency resources for coordinated response. If the protection fails or the fault reignites, it can only be detected through manual inspection, delaying emergency response. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution: a multi-sensor-based electrical cabinet fault diagnosis and fire safety interlock protection system, comprising: The data acquisition unit is used to determine the sensor association dataset of each electrical device based on the cabinet layout data of the electrical cabinet and the equipment parameters of each electrical device in the electrical cabinet; determine the initial device association topology map based on the sensor association dataset; and perform path deduplication processing on the initial device association topology map to obtain the full device association topology map. The feature extraction unit is used to receive multi-sensor data from various electrical devices in the electrical cabinet within a preset acquisition period; the multi-sensor data includes temperature time-series data, smoke concentration time-series data, current and voltage time-series data, humidity time-series data, and arc light sensor data; feature extraction is performed on the multi-sensor data to obtain an abnormal feature set; The fault diagnosis unit is used to perform fusion analysis on the abnormal feature set and the overall equipment association topology map to obtain fault diagnosis information. The fault diagnosis information includes fault type, risk level, the equipment association network in which the fault type is located, and the node location of the fault type in the equipment association network. The fire protection unit is used to generate a fire protection strategy package based on the risk level and node location in the fault diagnosis information; send the fire protection strategy package to the electrical cabinet intelligent terminal and execute the fire safety protection operation to obtain the protection execution result; The fault re-examination unit is used to send fault re-examination instructions to the intelligent terminal of the electrical cabinet to perform fault re-examination and obtain fault re-examination results; based on the fault re-examination results and protection execution results, it is determined whether the fault type has been eliminated and whether the fire risk has been controlled. The risk confirmation unit is used to perform emergency response operations based on the node location and fault re-inspection results in the fault diagnosis information if the fault type has not been eliminated or the fire risk has not been controlled; if the fault type has been eliminated and the fire risk has been controlled, the fault diagnosis and fire safety protection of the electrical cabinet are confirmed to be completed.

[0005] Preferably, the elements in the sensor association dataset are two electrical devices that have a signal transmission relationship; Based on the sensor association dataset, an initial device association topology is determined, including: Based on the sensor association dataset, multiple sensor adjacency datasets are determined; Among them, the elements in the sensor adjacency dataset are electrical devices, and the order of all elements in the sensor adjacency dataset indicates the signal transmission order of multiple electrical devices in the device association network. Based on multiple sensor adjacency datasets, an initial device association topology is determined.

[0006] Preferably, the equipment parameters include the thermal sensitivity data and power consumption level of the electrical equipment; the full equipment topology diagram displays the network of equipment connections within the electrical cabinet; Before fusing and analyzing the abnormal feature set and the overall equipment topology to obtain fault diagnosis information, the following steps are also included: Based on the thermal sensitivity data and the equipment association network where the fault type is located in the full equipment association topology diagram, determine the temperature field distribution data of all electrical equipment in the equipment association network; Determine whether there are any devices at risk of thermal runaway in the temperature field distribution data; Based on the pre-defined mapping relationship between temperature field distribution and thermal runaway risk, the thermal runaway risk level of the equipment is determined; the risk level is determined comprehensively based on the thermal runaway risk level and the fault type.

[0007] Preferably, the abnormal feature set and the overall equipment topology diagram are fused and analyzed to obtain fault diagnosis information, including: Based on the equipment parameters, the electrical device with the highest power consumption level among multiple electrical devices in the sensor association dataset is identified as the core device; Based on the temperature anomaly features in the anomaly feature set, query the preset temperature fault mode library to obtain the temperature fault mode; Based on the smoke anomaly features in the anomaly feature set, query the preset smoke fault mode library to obtain the smoke fault mode; Correlation analysis was performed based on temperature fault mode and smoke fault mode to obtain the correlation analysis results; If the correlation analysis results indicate a causal relationship between the temperature fault and the smoke fault, then the fault type is determined to be thermal runaway fault by combining the node position of the core equipment in the overall equipment correlation topology diagram. If the correlation analysis results indicate that there is no causal relationship between the temperature fault and the smoke fault, then the arc fault mode is obtained by querying the preset arc fault mode library based on the arc sensor data in the abnormal feature set, and the fault type is determined to be an electric arc fault based on the arc fault mode.

[0008] Preferably, a fire protection strategy package is generated based on the risk level and node location in the fault diagnosis information, including: Fire protection levels are determined based on risk levels; Based on the node locations in the fault diagnosis information, determine the scope of the fault type's impact on the device's associated network; Match the corresponding fire action sequence from the preset fire strategy library according to the fire protection level and the scope of impact; The fire-fighting actions in the fire-fighting action sequence are time-aligned with the fault handling actions of electrical equipment with adjacent node positions in the overall equipment association topology diagram to obtain the aligned fusion protection strategy. A fire protection strategy package is generated based on the aligned fusion protection strategy.

[0009] Preferably, the multi-sensor data also includes real-time anomaly information and historical clearing information. Real-time anomaly information is used to indicate that there is equipment anomaly in the electrical cabinet, and historical clearing information is used to indicate historical fault diagnosis information that needs to be cleared. Before receiving multi-sensor data from various electrical devices in the electrical cabinet within a preset acquisition period, the process also includes: Determine whether the multi-sensor data carries real-time anomaly information or historical cleared information; If multiple sensor data carry real-time anomaly information, determine whether the abnormal device indicated by the real-time anomaly information is an electrical device in the full device association topology diagram; If the abnormal device indicated by the real-time anomaly information is an electrical device in the full device association topology diagram, determine whether there is historical fault diagnosis information about the abnormal device; If historical fault diagnosis information about abnormal equipment exists, the real-time abnormal information is associated with the historical fault diagnosis information, and the abnormal feature set is updated based on the associated historical fault diagnosis information. If no historical fault diagnosis information for abnormal equipment exists, perform feature extraction on multi-sensor data to obtain an abnormal feature set.

[0010] Preferably, a fault re-inspection command is sent to the intelligent terminal of the electrical cabinet to perform a fault re-inspection, and the fault re-inspection results are obtained, including: Determine the corresponding re-inspection sensor type based on the fault type in the fault diagnosis information; A re-inspection data acquisition command is generated based on the re-inspection sensor type; the re-inspection data acquisition command is used to control the electrical cabinet intelligent terminal to collect data from the sensor corresponding to the re-inspection sensor type of the electrical cabinet. Send the re-inspection and data collection instruction to the electrical cabinet's intelligent terminal; Receive re-inspection sensor data generated by the intelligent terminal of the electrical cabinet in response to the re-inspection data acquisition command; By comparing and analyzing the re-inspection sensor data and fault types, the fault re-inspection results are obtained.

[0011] Preferably, the determination of whether the fault type has been eliminated and whether the fire risk has been controlled is based on the fault re-inspection results and protection execution results, including: Determine whether all fire-fighting actions in the protection execution results have been completed; If all fire-fighting actions have been completed, compare and analyze the fault re-inspection results and abnormal feature set to obtain the elimination status of the fault type. Based on the fire risk control data in the elimination status and protection execution results, determine whether the fault type has been eliminated and whether the fire risk has been controlled.

[0012] Preferably, emergency response operations are performed based on the node location in the fault diagnosis information and the fault re-inspection results, including: Obtain the location information of the electrical cabinet to determine its location; Search the preset emergency resource database for emergency response resources within a preset radius of the electrical cabinet location to obtain n emergency response resources; where n is an integer greater than 1; the emergency resource database stores the geographical location information and resource type of each emergency response resource in advance; The distance and resource matching degree between the location of the electrical cabinet and n emergency response resources are calculated to obtain n comprehensive evaluation values; Based on the node locations in the fault diagnosis information, determine the set of electrical devices affected by the fault type in the device association network; The matching degree between the resource requirements of each electrical device in the set of electrical devices and n comprehensive evaluation values ​​is calculated to obtain the corrected n comprehensive evaluation values; Determine the optimal comprehensive evaluation value among the n corrected comprehensive evaluation values; Obtain the emergency response resources corresponding to the optimal comprehensive evaluation value; The node location, fault re-inspection results, and electrical cabinet location in the fault diagnosis information are sent to the emergency response center corresponding to the emergency response resources.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This invention analyzes the correlation between abnormal feature sets such as temperature, smoke, and arc light and a preset fault mode library. Combined with the core equipment determined by power consumption level and the node positions in the overall equipment association topology, it can accurately determine whether the fault is a thermal runaway fault or an arc fault, and precisely locate the specific node in the equipment association network. This avoids the problem of false alarms or missed alarms of traditional single sensors, and provides a reliable basis for subsequent fire protection. This invention determines the fire protection level by risk level, and then combines the impact range of the fault node location to match the fire action sequence from the strategy library. These actions are then aligned with the fault handling actions of adjacent equipment at the millisecond level to generate a fusion protection strategy, avoiding the spread of fire or secondary damage to equipment caused by action conflicts or incorrect sequence. At the same time, after the protection is executed, a re-inspection test is conducted. If the fault is not eliminated, it can automatically call on surrounding emergency resources for linkage response, forming a closed-loop control, which significantly improves the timeliness and effectiveness of fire safety protection for electrical cabinets. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall system architecture in one embodiment of the present invention; Figure 2 This is a flowchart illustrating the steps of a fault diagnosis unit in one embodiment of the present invention; Figure 3 This is a schematic diagram of the steps of the fire protection unit in one embodiment of the present invention.

[0015] In the diagram: 1. Data acquisition unit; 2. Feature extraction unit; 3. Fault diagnosis unit; 4. Fire protection unit; 5. Fault re-inspection unit; 6. Risk confirmation unit. Detailed Implementation

[0016] 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.

[0017] Example 1, please refer to Figure 1This invention provides a technical solution: a multi-sensor-based electrical cabinet fault diagnosis and fire safety interlock protection system, comprising: Data acquisition unit 1 is used to determine the sensor association dataset of each electrical device based on the cabinet layout data of the electrical cabinet and the equipment parameters of each electrical device in the electrical cabinet; determine the initial device association topology map based on the sensor association dataset; and perform path deduplication processing on the initial device association topology map to obtain the full device association topology map. Feature extraction unit 2 is used to receive multi-sensor data of various electrical devices in the electrical cabinet within a preset collection period; wherein, the multi-sensor data includes temperature time series data, smoke concentration time series data, current and voltage time series data, humidity time series data and arc light sensor data; feature extraction is performed on the multi-sensor data to obtain an abnormal feature set; Fault diagnosis unit 3 is used to perform fusion analysis on the abnormal feature set and the overall equipment association topology map to obtain fault diagnosis information; wherein, the fault diagnosis information includes fault type, risk level, the equipment association network in which the fault type is located, and the node position of the fault type in the equipment association network. Fire protection unit 4 is used to generate a fire protection strategy package based on the risk level and node location in the fault diagnosis information; send the fire protection strategy package to the electrical cabinet intelligent terminal and execute the fire safety protection operation to obtain the protection execution result; Fault re-inspection unit 5 is used to send fault re-inspection instructions to the intelligent terminal of the electrical cabinet to perform fault re-inspection and obtain fault re-inspection results; based on the fault re-inspection results and protection execution results, it is determined whether the fault type has been eliminated and whether the fire risk has been controlled. Risk confirmation unit 6 is used to perform emergency response operations based on the node location and fault re-inspection results in the fault diagnosis information if the fault type is not eliminated or the fire risk is not controlled; if the fault type has been eliminated and the fire risk has been controlled, the fault diagnosis and fire safety protection of the electrical cabinet are confirmed to be completed.

[0018] It should be noted that there is a low-voltage electrical cabinet in the factory's power distribution room, which is equipped with electrical equipment such as circuit breakers, contactors, thermal relays and frequency converters. Based on the cabinet layout data, such as the equipment installation location, spacing and cable routing, as well as the equipment parameters of each electrical device, such as rated current and operating temperature range, the sensor association dataset for each device is determined. For example, a circuit breaker is associated with the current transformer of the incoming cable, and a contactor is associated with the arc sensor on the outgoing side. Based on these associations, an initial equipment association topology is constructed, where nodes are devices and edges are sensor association paths. Then, the topology is deduplicated to eliminate redundant connections, resulting in a complete equipment association topology. The full equipment interconnection topology diagram clearly shows the electrical connections and sensing and monitoring relationships between the equipment. For example, the circuit breaker node is connected to both current and voltage sensors and temperature sensors, while the frequency converter node is mainly connected to humidity sensors and smoke concentration sensors. It receives multi-sensor data within a preset acquisition period of 5 seconds; temperature time-series data comes from the patch thermocouple at the circuit breaker contact, smoke concentration time-series data comes from the smoke detector on the top of the cabinet, current and voltage time-series data comes from the current transformer at the incoming line end, humidity time-series data comes from the humidity-sensitive element on the inner wall of the cabinet, and arc light sensing data comes from the ultraviolet light sensor at each equipment interval. The system extracts features from these time-series data. For example, it finds that the current data has three consecutive spikes exceeding 1.5 times the rated value, the temperature data rises by 12 degrees Celsius in the last two seconds, and the arc sensor detects an abnormal light intensity that lasts for 80 milliseconds. These features are summarized into an abnormal feature set, which includes current overshoot, temperature surge, and short-term arc. The abnormal feature set was fused with the topology diagram of the entire equipment for analysis. By comparing the equipment node where the abnormality occurred with the path in the topology diagram, it was determined that the current overshoot phenomenon occurred on the line between the circuit breaker and the contactor, the temperature rise node was located at the outgoing end of the circuit breaker, and the arc signal came from near the arc-extinguishing chamber of the contactor. The system therefore obtains the following fault diagnosis information: the fault type is local overheating and overcurrent caused by contactor contact arcing; the risk level is high; the equipment associated network of the fault type is the main circuit of circuit breaker-contactor-thermal relay; the node location of the fault type in the equipment associated network is the contactor node and its upstream connecting cable segment. Fire protection strategy packages are generated based on high-risk levels and node locations. The strategy includes three actions: first, disconnecting the circuit breaker upstream of the fault node; second, activating the suspended dry powder extinguishing device located directly above the contactor, with the spray direction aimed at the arc detection point; and finally, turning on the exhaust fan inside the cabinet to reduce humidity and prevent arc reignition. The system sends the strategy package to the intelligent terminal of the electrical cabinet, and the terminal performs the following operations: circuit breaker trips, fire extinguishing device is triggered, and fan starts; the protection execution results are fed back as circuit breaker tripped successfully, fire extinguishing agent is released normally, and fan is running normally. The system then sent a fault re-inspection command to the smart terminal, requesting the re-collection of the temperature, smoke concentration and arc signal near the contactor. The re-inspection results showed that the temperature was still 58 degrees Celsius, which was too high. The smoke concentration had dropped to a safe range and the arc signal had disappeared. The protection execution results showed that the extinguishing agent had been released but the high temperature had not been eliminated. Therefore, the fault re-inspection results determined that the fault type had not been completely eliminated and the risk of reignition in the fire risk still existed. Based on the fault diagnosis information, the contactor at the node location and the fault re-inspection results showing that the high temperature had not subsided, an emergency response was initiated: the carbon dioxide gas injection system inside the cabinet was restarted via remote command from the backend to force cooling, and an audible and visual alarm was issued to notify on-site personnel to wear protective equipment and intervene manually; at the same time, the system recorded the entire emergency response process. If, in another simulated scenario, the re-inspection results show that the temperature is normal, the smoke is zero, and the arc light is normal, then it is confirmed that the fault type has been eliminated and the fire risk has been controlled. The system outputs a report on the completion of fault diagnosis and fire safety protection of the electrical cabinet.

[0019] In an optional embodiment, the elements in the sensor association dataset are two electrical devices that have a signal transmission relationship; Based on the sensor association dataset, an initial device association topology is determined, including: Based on the sensor association dataset, multiple sensor adjacency datasets are determined; Among them, the elements in the sensor adjacency dataset are electrical devices, and the order of all elements in the sensor adjacency dataset indicates the signal transmission order of multiple electrical devices in the device association network. Based on multiple sensor adjacency datasets, an initial device association topology is determined.

[0020] It should be noted that the electrical cabinet contains a thermostat, a heater, a fan, and a controller; the signal transmission relationship is as follows: the thermostat sends a temperature signal to the controller, the controller sends a start signal to the heater, and the controller sends a speed control signal to the fan; thus, the sensor association dataset contains three pairs of devices: thermostat and controller, controller and heater, and controller and fan; Based on these device pairs, starting from the signal source, two sensor adjacency datasets are generated: the first dataset consists of a thermostat, a controller, and a heater, with the indication signal transmitted from the thermostat to the controller and then to the heater; the second dataset consists of a thermostat, a controller, and a fan, with the indication signal transmitted from the thermostat to the controller and then to the fan. Based on the order relationship in these two datasets, an initial device association topology graph is obtained: the graph contains four nodes: thermostat, controller, heater, and fan; the directed edges are thermostat pointing to controller, controller pointing to heater, and controller pointing to fan; this topology graph reflects the signal transmission network.

[0021] In an optional embodiment, the device parameters include thermal sensitivity data and power consumption levels of the electrical equipment; a full device topology diagram displays the device association network within the electrical cabinet; Before fusing and analyzing the abnormal feature set and the overall equipment topology to obtain fault diagnosis information, the following steps are also included: Based on the thermal sensitivity data and the equipment association network where the fault type is located in the full equipment association topology diagram, determine the temperature field distribution data of all electrical equipment in the equipment association network; Determine whether there are any devices at risk of thermal runaway in the temperature field distribution data; Based on the pre-defined mapping relationship between temperature field distribution and thermal runaway risk, the thermal runaway risk level of the equipment is determined; the risk level is determined comprehensively based on the thermal runaway risk level and the fault type.

[0022] It should be noted that the overall equipment topology diagram within the electrical cabinet contains a device network that includes circuit breakers, contactors, and thermal relays. The fault type diagnosis indicates that contactor contact arcing has occurred in this network. Among the equipment parameters, the contactor's thermal sensitivity data shows an upper limit of 85 degrees Celsius, a thermal time constant of 120 seconds, and a high power consumption rating; the circuit breaker's thermal sensitivity data shows an upper limit of 70 degrees Celsius and a medium power consumption rating; and the thermal relay's thermal sensitivity data shows an upper limit of 65 degrees Celsius and a low power consumption rating. The system calculates the temperature field distribution data of each device in the network based on the heat-sensitive data and the network associated with the device; for example, a contactor generates a local high temperature of 120 degrees Celsius due to an arcing fault, which is conducted along the copper busbar to the circuit breaker side at about 95 degrees Celsius and to the thermal relay side at about 80 degrees Celsius. Based on the thermal sensitivity data of each device, it was determined that the circuit breaker temperature of 95 degrees Celsius exceeded its upper limit of 70 degrees Celsius, and the thermal relay temperature of 80 degrees Celsius exceeded its upper limit of 65 degrees Celsius. Therefore, both devices are at risk of thermal runaway, with the circuit breaker accumulating heat faster. The system further determines the risk of thermal runaway based on a preset mapping relationship between temperature field distribution and thermal runaway risk. For example, a temperature exceeding the upper limit by less than 10% is considered low risk, 10% to 30% is considered medium risk, and above 30% is considered high risk. A circuit breaker exceeding 71% corresponds to high risk, and a thermal relay exceeding 23% corresponds to medium risk. Therefore, among the equipment identified as having a risk of thermal runaway, circuit breakers were classified as high-risk and thermal relays as medium-risk. Finally, considering the high risk level of the fault type itself, the overall risk level was determined to be extremely high, requiring immediate implementation of fire protection strategies.

[0023] In an optional embodiment, a fusion analysis is performed on the abnormal feature set and the overall device association topology to obtain fault diagnosis information, including: A1. Based on the equipment parameters, the electrical device with the highest power consumption level among multiple electrical devices in the sensor association dataset is identified as the core device; A2. Based on the temperature anomaly features in the anomaly feature set, query the preset temperature fault mode library to obtain the temperature fault mode; based on the smoke anomaly features in the anomaly feature set, query the preset smoke fault mode library to obtain the smoke fault mode. A3. Perform correlation analysis based on temperature fault mode and smoke fault mode to obtain the correlation analysis results; A4. If the correlation analysis results indicate a causal relationship between the temperature fault and the smoke fault, then the fault type is determined to be thermal runaway fault by combining the node position of the core equipment in the overall equipment correlation topology diagram. A5. If the correlation analysis results indicate that there is no causal relationship between the temperature fault and the smoke fault, then the arc fault mode is obtained by querying the preset arc fault mode library based on the arc sensor data in the abnormal feature set, and the fault type is determined to be an electric arc fault based on the arc fault mode.

[0024] It should be noted that the equipment association network of the electrical cabinet includes circuit breakers with high power consumption levels, contactors with medium power consumption levels, and frequency converters with low power consumption levels; the circuit breaker with the highest power consumption level is identified as the core equipment. Temperature anomaly features were extracted from the abnormal feature set: the circuit breaker contact temperature rose from 45 degrees Celsius to 110 degrees Celsius within 3 seconds; smoke anomaly features: the smoke concentration inside the cabinet rose from 0.1 mg / m³ to 0.8 mg / m³; the preset temperature fault mode library was queried, and the temperature fault mode was found to be contact overheating leading to high temperature; the preset smoke fault mode library was queried, and the smoke fault mode was found to be insulation material pyrolysis generating smoke. Correlation analysis was performed on the two fault modes: overheating of the contacts led to pyrolysis of the insulation material, and there was a causal relationship between the two; the correlation analysis results showed that there was a causal relationship between the temperature fault and the smoke fault; at this time, combined with the fact that the core equipment, namely the circuit breaker, is located on the main incoming line side in the overall equipment correlation topology diagram, the fault type was determined to be thermal runaway fault. In another scenario, the abnormal features showed that the temperature abnormality was a local temperature rise of 5 degrees Celsius in the contactor, and the smoke abnormality was no smoke or extremely low concentration; the temperature fault mode was a slight temperature rise, and the smoke fault mode was no abnormality; there was no causal relationship between the two; the system then queried the arc fault mode library, and the arc sensor data was a strong light pulse lasting 200 milliseconds. The matching arc fault mode was found to be interphase arc discharge, and the fault type was determined to be an arc fault.

[0025] In an optional embodiment, a fire protection strategy package is generated based on the risk level and node locations in the fault diagnosis information, including: B1. Determine the fire protection level based on the risk level; B2. Based on the node locations in the fault diagnosis information, determine the scope of the fault type's impact on the device's associated network; B3. Match the corresponding fire action sequence from the preset fire strategy library according to the fire protection level and the scope of impact; B4. Align the fire-fighting actions in the fire-fighting action sequence with the fault handling actions of electrical equipment adjacent to the node positions in the overall equipment association topology diagram to obtain the aligned fusion protection strategy, and generate a fire protection strategy package based on the aligned fusion protection strategy.

[0026] It should be noted that in the electrical cabinet, the fault diagnosis information shows that the fault type is thermal runaway fault, the risk level is high, and the fault node location is the circuit breaker; the system first maps the high risk level to fire protection level two, corresponding to rapid local fire extinguishing and power outage; Based on the location of the fault node being a circuit breaker, determine its scope of influence in the overall equipment topology diagram: including the incoming circuit where the circuit breaker is located, downstream connected contactors and thermal relays, as well as adjacent cable channels; Based on the fire protection level of Level II and the scope of impact, the corresponding fire action sequence is matched from the preset fire strategy library: the first action is to disconnect the main switch upstream of the circuit breaker; the second action is to activate the suspended dry powder fire extinguishing device and spray it towards the circuit breaker area; the third action is to shut down the exhaust fan in the cabinet to prevent oxygen from supporting combustion. The sequence of fire-fighting actions is aligned with the fault handling actions of electrical equipment adjacent to the circuit breaker node in the overall equipment topology diagram. The disconnection action of adjacent equipment, including the upstream main switch, must be executed within 0 milliseconds, the circuit breaker's own protection tripping action must be completed within 50 milliseconds, the dry powder fire extinguishing device must be triggered 100 milliseconds after the circuit breaker trips, and the exhaust fan must be shut down 500 milliseconds after the fire extinguishing spray begins to avoid interfering with the diffusion of the fire extinguishing agent. Align these time points to obtain the integrated protection strategy: disconnect the upstream main switch at 0 milliseconds, confirm the circuit breaker trip at 50 milliseconds, activate the fire extinguishing device at 100 milliseconds, and shut down the fan at 600 milliseconds; finally, generate a fire protection strategy package based on the aligned integrated protection strategy, which includes equipment identification, action instructions and timestamps, and send it to the electrical cabinet intelligent terminal for execution.

[0027] In an optional embodiment, the multi-sensor data further includes real-time anomaly information and historical clearing information. The real-time anomaly information is used to indicate that there is a device anomaly in the electrical cabinet, and the historical clearing information is used to indicate historical fault diagnosis information that needs to be cleared. Before receiving multi-sensor data from various electrical devices in the electrical cabinet within a preset acquisition period, the process also includes: Determine whether the multi-sensor data carries real-time anomaly information or historical cleared information; If multiple sensor data carry real-time anomaly information, determine whether the abnormal device indicated by the real-time anomaly information is an electrical device in the full device association topology diagram; If the abnormal device indicated by the real-time anomaly information is an electrical device in the full device association topology diagram, determine whether there is historical fault diagnosis information about the abnormal device; If historical fault diagnosis information about abnormal equipment exists, the real-time abnormal information is associated with the historical fault diagnosis information, and the abnormal feature set is updated based on the associated historical fault diagnosis information. If no historical fault diagnosis information for abnormal equipment exists, perform feature extraction on multi-sensor data to obtain an abnormal feature set.

[0028] It should be noted that the electrical cabinet's full equipment topology diagram includes contactors and circuit breakers; the system has stored one historical fault diagnosis record, which shows that the contactor experienced a contact overheating fault three days ago. In the next data collection cycle, in addition to time-series data such as temperature, smoke, and current, the multi-sensor data also carries a real-time anomaly message, which states that the contactor temperature has once again surged above the threshold, as well as a history clearing message, which requires the removal of old fault records from three days ago. The system analyzes multi-sensor data to determine that it contains both real-time anomaly information and historical clearing information. Based on the historical clearing information, the system clears the old historical fault diagnosis information of the contactor. Processing real-time anomaly information: This information indicates that the abnormal device is a contactor; the system confirms that the contactor belongs to a node in the overall equipment association topology diagram; then it queries whether there is any historical fault diagnosis information about the contactor; since the old records have been cleared, there is no historical fault diagnosis information about the contactor at this time; therefore, the system directly performs feature extraction on the conventional time-series data in the multi-sensor data to obtain a new set of abnormal features for subsequent fault diagnosis. In another scenario, assuming the historical clearing information has not been sent, the system still retains the contactor's historical fault record from three days ago. At this time, the received real-time anomaly information is still a sudden rise in contactor temperature. After confirming that the contactor has historical fault diagnosis information, the system associates the current real-time anomaly information with this historical record. For example, it increments the original fault count by one, updates the time of the most recent anomaly, and updates the anomaly feature set based on the associated historical information, so that the set includes the evolution trend of two consecutive overheating faults. Subsequently, the system uses this updated anomaly feature set for fusion analysis.

[0029] In an optional embodiment, a fault re-inspection command is sent to the intelligent terminal of the electrical cabinet to perform a fault re-inspection and obtain the fault re-inspection result, including: Determine the corresponding re-inspection sensor type based on the fault type in the fault diagnosis information; A re-inspection data acquisition command is generated based on the re-inspection sensor type; the re-inspection data acquisition command is used to control the electrical cabinet intelligent terminal to collect data from the sensor corresponding to the re-inspection sensor type of the electrical cabinet. Send the re-inspection and data collection instruction to the electrical cabinet's intelligent terminal; Receive re-inspection sensor data generated by the intelligent terminal of the electrical cabinet in response to the re-inspection data acquisition command; By comparing and analyzing the re-inspection sensor data and fault types, the fault re-inspection results are obtained.

[0030] It should be noted that the fault diagnosis information in the electrical cabinet shows the fault type as arc fault, which occurs at the contactor node; based on the arc fault, the corresponding re-inspection sensor type is determined to be arc light sensor and current and voltage sensor; a re-inspection acquisition command is generated, which requires the electrical cabinet intelligent terminal to perform secondary data acquisition on the ultraviolet light sensor near the contactor arc extinguishing chamber and the current transformer and voltage transformer at the contactor's inlet and outlet terminals, with an acquisition time of 2 seconds; The re-inspection data collection command is sent to the electrical cabinet's intelligent terminal. After the terminal responds to the command, it collects the re-inspection sensor data: the arc signal is zero, the current value is 0 amps, and the voltage value is 380 volts, which is normal. The re-inspection sensor data is compared and analyzed with the fault type arc fault: the absence of arc and the zero current indicate that the arc has been extinguished, and the normal voltage indicates that the upstream power supply has been restored. Therefore, the fault re-inspection result is that the fault type has been eliminated.

[0031] In an optional embodiment, determining whether the fault type has been eliminated and whether the fire risk has been controlled based on the fault re-inspection results and protection execution results includes: Determine whether all fire-fighting actions in the protection execution results have been completed; If all fire-fighting actions have been completed, compare and analyze the fault re-inspection results and abnormal feature set to obtain the elimination status of the fault type. Based on the fire risk control data in the elimination status and protection execution results, determine whether the fault type has been eliminated and whether the fire risk has been controlled.

[0032] It should be noted that after a thermal runaway failure in the electrical cabinet, the system generates a protection strategy package containing three fire-fighting actions: disconnecting the incoming circuit breaker, activating the dry powder fire extinguishing device on the top of the cabinet, and shutting down the fan; after the electrical cabinet's intelligent terminal executes these actions, it returns the protection execution result; the protection execution result confirms that all three actions are marked as completed, meaning that all fire-fighting actions have been executed. The fault re-inspection results were compared with the abnormal feature set. The fault re-inspection results showed that the circuit breaker contact temperature was 42 degrees Celsius and the smoke concentration was 0.02 mg / m³. The abnormal feature set originally recorded that the contact temperature at the time of the fault was 115 degrees Celsius and the smoke concentration was 0.9 mg / m³. The comparison showed that the temperature dropped by 73 degrees and the smoke concentration decreased to a safe range. The fault type was determined to be eliminated. Finally, combining the fire risk control data in the protection execution results, such as the dry powder fire extinguishing device spraying coverage area including the entire fault node, the oxygen content in the cabinet dropping to 15% and no signs of reignition, it was comprehensively determined that the fault type had been eliminated and the fire risk had been controlled.

[0033] In an optional embodiment, emergency response operations are performed based on the node location and fault re-inspection results in the fault diagnosis information, including: Obtain the location information of the electrical cabinet to determine its location; Search the preset emergency resource database for emergency response resources within a preset radius of the electrical cabinet location to obtain n emergency response resources; where n is an integer greater than 1; the emergency resource database stores the geographical location information and resource type of each emergency response resource in advance; The distance and resource matching degree between the location of the electrical cabinet and n emergency response resources are calculated to obtain n comprehensive evaluation values; Based on the node locations in the fault diagnosis information, determine the set of electrical devices affected by the fault type in the device association network; The matching degree between the resource requirements of each electrical device in the set of electrical devices and n comprehensive evaluation values ​​is calculated to obtain the corrected n comprehensive evaluation values; Determine the optimal comprehensive evaluation value among the n corrected comprehensive evaluation values; Obtain the emergency response resources corresponding to the optimal comprehensive evaluation value; The node location, fault re-inspection results, and electrical cabinet location in the fault diagnosis information are sent to the emergency response center corresponding to the emergency response resources.

[0034] It should be noted that when a thermal runaway fault occurs in the factory's electrical cabinet, the location information of the electrical cabinet is first obtained, which is found to be the power distribution room on the second floor of Building 3 in the factory area. The pre-set emergency resource database is then searched for emergency response resources within a 500-meter radius of the electrical cabinet's location. Three resources are found: Resource A is the factory's fire brigade, located 300 meters away, and its resource type is fire rescue team; Resource B is the factory's electrical repair team, located 150 meters away, and its resource type is electrical maintenance; Resource C is a mobile fire extinguisher station located 450 meters away from the adjacent warehouse, and its resource type is fire extinguishing equipment. Distance and resource matching degree calculations were performed on the location of the electrical cabinet and the three resources respectively; the closer the distance, the higher the score. The resource type matching degree was scored according to thermal runaway fault: fire brigade matching degree 0.9, electrical emergency repair team matching degree 0.7, fire extinguisher station matching degree 0.8; after comprehensive evaluation, the comprehensive evaluation values ​​of A, B, and C were 0.85, 0.88, and 0.78 respectively. Based on the fault diagnosis information indicating a contactor as the node location, the affected electrical equipment set is determined to include the contactor, upstream and downstream cables, and thermal relays. The resource requirements of this set of equipment are analyzed: power outage isolation, continuous cooling and fire suppression, and insulation testing are needed. This requirement is then matched with three comprehensive evaluation values ​​in a secondary calculation. Although the electrical emergency repair team is the closest, it lacks continuous fire suppression capabilities, resulting in a score of 0.70. The fire brigade possesses both fire suppression and cooling capabilities, raising its score to 0.92. The fire extinguisher station only provides equipment but lacks operators, resulting in a score of 0.60. After correction, the optimal comprehensive evaluation value is 0.92 for Resource A (fire brigade). The emergency response resources corresponding to the optimal comprehensive assessment value, namely the factory fire brigade, are obtained. Then, information such as the location of the fault node contactor, the fault re-inspection result that the high temperature has not been completely eliminated, and the location of the electrical cabinet being the power distribution room on the second floor of Plant No. 3, are sent to the fire brigade's emergency response center to trigger personnel to be dispatched for disposal.

[0035] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A multi-sensor-based electrical cabinet fault diagnosis and fire safety interlock protection system, characterized in that, include: The data acquisition unit is used to determine the sensor association dataset of each electrical device based on the cabinet layout data of the electrical cabinet and the equipment parameters of each electrical device in the electrical cabinet. Based on the sensor association dataset, determine the initial device association topology map; Path deduplication is performed on the initial device association topology map to obtain the full device association topology map; The feature extraction unit is used to receive multi-sensor data from various electrical devices in the electrical cabinet within a preset acquisition period; the multi-sensor data includes temperature time-series data, smoke concentration time-series data, current and voltage time-series data, humidity time-series data, and arc light sensor data; feature extraction is performed on the multi-sensor data to obtain an abnormal feature set; The fault diagnosis unit is used to perform fusion analysis on the abnormal feature set and the overall equipment association topology map to obtain fault diagnosis information. The fault diagnosis information includes fault type, risk level, the equipment association network in which the fault type is located, and the node location of the fault type in the equipment association network. The fire protection unit is used to generate a fire protection strategy package based on the risk level and node location in the fault diagnosis information; send the fire protection strategy package to the electrical cabinet intelligent terminal and execute the fire safety protection operation to obtain the protection execution result; The fault re-examination unit is used to send fault re-examination instructions to the intelligent terminal of the electrical cabinet to perform fault re-examination and obtain fault re-examination results; based on the fault re-examination results and protection execution results, it is determined whether the fault type has been eliminated and whether the fire risk has been controlled. The risk confirmation unit is used to perform emergency response operations based on the node location and fault re-inspection results in the fault diagnosis information if the fault type has not been eliminated or the fire risk has not been controlled; if the fault type has been eliminated and the fire risk has been controlled, the fault diagnosis and fire safety protection of the electrical cabinet are confirmed to be completed.

2. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 1, characterized in that, The elements in the sensor association dataset are two electrical devices that have a signal transmission relationship; Based on the sensor association dataset, an initial device association topology is determined, including: Based on the sensor association dataset, multiple sensor adjacency datasets are determined; Among them, the elements in the sensor adjacency dataset are electrical devices, and the order of all elements in the sensor adjacency dataset indicates the signal transmission order of multiple electrical devices in the device association network. Based on multiple sensor adjacency datasets, an initial device association topology is determined.

3. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 2, characterized in that, Equipment parameters include thermal sensitivity data and power consumption levels of electrical equipment; the overall equipment topology diagram displays the network of equipment connections within the electrical cabinet; Before fusing and analyzing the abnormal feature set and the overall equipment topology to obtain fault diagnosis information, the following steps are also included: Based on the thermal sensitivity data and the equipment association network where the fault type is located in the full equipment association topology diagram, determine the temperature field distribution data of all electrical equipment in the equipment association network; Determine whether there are any devices at risk of thermal runaway in the temperature field distribution data; Based on the pre-defined mapping relationship between temperature field distribution and thermal runaway risk, the thermal runaway risk level of the equipment is determined; the risk level is determined comprehensively based on the thermal runaway risk level and the fault type.

4. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 3, characterized in that, By fusing and analyzing the abnormal feature set and the overall equipment topology diagram, fault diagnosis information is obtained, including: Based on the equipment parameters, the electrical device with the highest power consumption level among multiple electrical devices in the sensor association dataset is identified as the core device; Based on the temperature anomaly features in the anomaly feature set, query the preset temperature fault mode library to obtain the temperature fault mode; Based on the smoke anomaly features in the anomaly feature set, query the preset smoke fault mode library to obtain the smoke fault mode; Correlation analysis was performed based on temperature fault mode and smoke fault mode to obtain the correlation analysis results; If the correlation analysis results indicate a causal relationship between the temperature fault and the smoke fault, then the fault type is determined to be thermal runaway fault by combining the node position of the core equipment in the overall equipment correlation topology diagram. If the correlation analysis results indicate that there is no causal relationship between the temperature fault and the smoke fault, then the arc fault mode is obtained by querying the preset arc fault mode library based on the arc sensor data in the abnormal feature set, and the fault type is determined to be an electric arc fault based on the arc fault mode.

5. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 4, characterized in that, Based on the risk level and node locations in the fault diagnosis information, a fire protection strategy package is generated, including: Fire protection levels are determined based on risk levels; Based on the node locations in the fault diagnosis information, determine the scope of the fault type's impact on the device's associated network; Match the corresponding fire action sequence from the preset fire strategy library according to the fire protection level and the scope of impact; The fire-fighting actions in the fire-fighting action sequence are time-aligned with the fault handling actions of electrical equipment with adjacent node positions in the overall equipment association topology diagram to obtain the aligned fusion protection strategy. A fire protection strategy package is generated based on the aligned fusion protection strategy.

6. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 5, characterized in that, Multi-sensor data also includes real-time anomaly information and historical clearing information. Real-time anomaly information is used to indicate equipment anomalies in the electrical cabinet, while historical clearing information is used to indicate historical fault diagnosis information that needs to be cleared. Before receiving multi-sensor data from various electrical devices in the electrical cabinet within a preset acquisition period, the process also includes: Determine whether the multi-sensor data carries real-time anomaly information or historical cleared information; If multiple sensor data carry real-time anomaly information, determine whether the abnormal device indicated by the real-time anomaly information is an electrical device in the full device association topology diagram; If the abnormal device indicated by the real-time anomaly information is an electrical device in the full device association topology diagram, determine whether there is historical fault diagnosis information about the abnormal device; If historical fault diagnosis information about abnormal equipment exists, the real-time abnormal information is associated with the historical fault diagnosis information, and the abnormal feature set is updated based on the associated historical fault diagnosis information. If no historical fault diagnosis information for abnormal equipment exists, perform feature extraction on multi-sensor data to obtain an abnormal feature set.

7. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 6, characterized in that, Send a fault re-inspection command to the intelligent terminal of the electrical cabinet to perform a fault re-inspection, and obtain the fault re-inspection results, including: Determine the corresponding re-inspection sensor type based on the fault type in the fault diagnosis information; A re-inspection data acquisition command is generated based on the re-inspection sensor type; the re-inspection data acquisition command is used to control the electrical cabinet intelligent terminal to collect data from the sensor corresponding to the re-inspection sensor type of the electrical cabinet. Send the re-inspection and data collection instruction to the electrical cabinet's intelligent terminal; Receive re-inspection sensor data generated by the intelligent terminal of the electrical cabinet in response to the re-inspection data acquisition command; By comparing and analyzing the re-inspection sensor data and fault types, the fault re-inspection results are obtained.

8. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 7, characterized in that, Based on the results of fault re-inspection and protection execution, determine whether the fault type has been eliminated and whether the fire risk has been controlled, including: Determine whether all fire-fighting actions in the protection execution results have been completed; If all fire-fighting actions have been completed, compare and analyze the fault re-inspection results and abnormal feature set to obtain the elimination status of the fault type. Based on the fire risk control data in the elimination status and protection execution results, determine whether the fault type has been eliminated and whether the fire risk has been controlled.

9. The electrical cabinet fault diagnosis and fire safety interlock protection system based on multi-sensor technology according to claim 8, characterized in that, Based on the node locations and fault re-inspection results in the fault diagnosis information, emergency response procedures will be performed, including: Obtain the location information of the electrical cabinet to determine its location; Search the preset emergency resource database for emergency response resources within a preset radius of the electrical cabinet location to obtain n emergency response resources; where n is an integer greater than 1; the emergency resource database stores the geographical location information and resource type of each emergency response resource in advance; The distance and resource matching degree between the location of the electrical cabinet and n emergency response resources are calculated to obtain n comprehensive evaluation values; Based on the node locations in the fault diagnosis information, determine the set of electrical devices affected by the fault type in the device association network; The matching degree between the resource requirements of each electrical device in the set of electrical devices and n comprehensive evaluation values ​​is calculated to obtain the corrected n comprehensive evaluation values; Determine the optimal comprehensive evaluation value among the n corrected comprehensive evaluation values; Obtain the emergency response resources corresponding to the optimal comprehensive evaluation value; The node location, fault re-inspection results, and electrical cabinet location in the fault diagnosis information are sent to the emergency response center corresponding to the emergency response resources.

Images