Fire hazard multistage early warning system of energy storage equipment

Abstract

The invention, which relates to the technical field of the energy storage equipment, discloses a fire multi-stage early warning system for energy storage equipment, and the system comprises the steps: collecting the operation temperature data of a lithium battery pack, generating an operation temperature value, and starting a fire early warning mechanism if the operation temperature value exceeds an operation temperature threshold value; if the abnormal degree of the fire monitoring data of the potential risk area exceeds the expectation, a first-level alarm instruction is sent out and corresponding feedback data is received, if the feedback data is not consistent with the expectation, a second-level alarm instruction is sent out, the existing fire risk is processed, and after a traversal area is identified according to the fire state data, the traversal area is identified. An emergency escape path is planned in the early warning area, and if the emergency escape path does not exist, a third-level alarm instruction is sent out; and if the secondary alarm instruction is not received, the battery pack temperature adjustment knowledge graph is matched with a corresponding processing strategy. The lithium battery pack is maintained in a targeted manner, the running state of the lithium battery pack is improved, and subsequent potential safety hazards and fire risks are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage equipment, and in particular to a multi-level fire warning system for energy storage equipment. Background Art

[0002] The multi-level fire warning mechanism of energy storage equipment is an advanced safety management system that aims to achieve comprehensive monitoring and prevention of fire risks in energy storage systems through multi-level and multi-level warning methods. The mechanism integrates a variety of sensor technologies, including high-precision temperature sensors, sensitive smoke detectors, and professional gas concentration detectors. These devices can monitor various key parameters of energy storage equipment and its surrounding environment in real time, such as abnormal temperature rise, changes in smoke concentration, and excessive combustible gas concentration.

[0003] Based on the monitored data, the multi-level fire warning system will automatically assess the fire danger level, which is usually divided into low danger level, medium danger level and high danger level. At different danger levels, the system will trigger corresponding warning measures, such as sounding an alarm, starting the emergency cooling system, starting the smoke exhaust device or triggering the automatic fire extinguishing device, so as to achieve early detection, timely warning and effective control of fire risks.

[0004] In the Chinese invention patent with application publication number CN116484187A, a thermal runaway early warning method and system for energy storage equipment is disclosed, wherein the method includes the following steps: obtaining the operation information before and after the thermal runaway occurs in the operation data of the energy storage equipment; establishing the basis for judging the abnormal operation of relevant parameters when the energy storage equipment is in thermal runaway; collecting the operation parameters related to thermal runaway during the operation of the energy storage equipment; judging whether the energy storage equipment has thermal runaway. Compared with the prior art, the present invention has the advantages of fast and accurate judgment of thermal runaway of energy storage equipment, can give early warning of possible thermal runaway events of energy storage equipment, and has broad application prospects for improving the safety and effectiveness of energy storage equipment operation and maintenance.

[0005] When energy storage equipment, especially lithium battery packs or lead-acid battery packs, are in operation and short circuits or similar faults occur, they are particularly prone to spontaneous combustion or even fire. In particular, the battery also has the characteristic of explosive combustion. Once there is a fire hazard, the risk is very high. Therefore, fire monitoring and early warning devices and fire extinguishing systems need to be fully arranged in the coverage area of ​​the battery pack to facilitate timely handling and response in the event of a fire.

[0006] In the existing multi-level early warning method for fire of energy storage equipment, a sensor network is usually arranged in the area where the battery pack is located, especially in the potential risk area, and the sensor network monitors various data, such as temperature, light and smoke data, etc. When one or more of them are abnormal, an early warning or alarm is immediately issued to the outside; However, there are certain anomalies in the charging scenarios in the area where the battery pack is located. In particular, when the ambient temperature is high, the lithium battery is overcharged to a certain extent, or even when dust accumulates on the external surface, the operating temperature of the battery pack is likely to rise, and the risk of operating failures or even spontaneous combustion will be further increased. In this relatively abnormal charging scenario, when conducting fire warnings and alarms in potential risk areas, if only a static first-level alarm or repeated alarm mode is still used, the staff in the potential risk area may neglect it and find it difficult to take timely and effective measures.

[0007] To this end, the present invention provides a multi-level fire warning system for energy storage equipment. Summary of the invention

[0008] 1. Technical issues to be resolved In view of the deficiencies of the prior art, the present invention provides a multi-level fire warning system for energy storage equipment. By collecting the operating temperature data of the lithium battery pack and generating an operating temperature value, if the operating temperature value exceeds the operating temperature threshold, the fire warning mechanism is activated; if the abnormality of the fire monitoring data in the potential risk area exceeds expectations, a first-level alarm instruction is issued and the corresponding feedback data is received. If the feedback data is inconsistent with expectations, a second-level alarm instruction is issued to process the existing fire risk. After identifying the traversable area based on the fire status data, an emergency escape path is planned in the warning area. If there is no emergency escape path, a third-level alarm instruction is issued; if the second-level alarm instruction is not received, the battery pack temperature adjustment knowledge map matches the corresponding processing strategy. Targeted maintenance is performed on the lithium battery pack to reduce the subsequent safety hazards and fire risks, thereby solving the technical problems recorded in the background technology.

[0009] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: The multi-level fire warning method of energy storage equipment includes collecting the current charging scene data in the warning area, constructing a deviation value based on the difference between the current charging scene data and the preset scene, and if the deviation value exceeds the expectation, screening out the potential risk area in the warning area and issuing a data collection instruction to the outside; Collect fire monitoring data in potential risk areas. If the proportion of abnormal data exceeds expectations, collect the operating temperature data of the lithium battery pack and generate an operating temperature value. If the operating temperature value exceeds the operating temperature threshold, automatically start the fire warning mechanism; If the abnormality of fire monitoring data in the potential risk area exceeds expectations, a first-level alarm command is issued and the corresponding feedback data is received. If the feedback data is inconsistent with expectations, a second-level alarm command is issued to the outside. If a second-level alarm command is received, the existing fire risk is processed, and after identifying the traversable area based on the fire status data, an emergency escape route is planned in the warning area. If there is no emergency escape route, a third-level alarm command is issued to the outside. If no secondary alarm command is received, when there is an abnormality in the lithium battery pack, the key influencing factors that cause the lithium battery pack to enter a high temperature state are analyzed and obtained. Based on the key influencing factors, the battery pack temperature regulation knowledge graph matches the corresponding processing strategy.

[0010] Furthermore, charging scene data of the lithium battery pack is collected, including at least ambient temperature, charging status and dust accumulation data, and summarized to generate a charging scene data set; after setting the corresponding qualified interval for each charging scene parameter, the corresponding preset scene is obtained, and a deviation value is constructed according to the gap between the charging scene and the preset scene. If the deviation value exceeds the preset deviation threshold, an area identification instruction is issued to the outside.

[0011] Furthermore, after receiving the area recognition instruction, the monitoring point is marked with the charging scene data. If the deviation value at the current monitoring point and its adjacent area exceeds the expectation, it is regarded as a risk point and marked on the electronic map; Use a pre-trained clustering algorithm to cluster several risk points and obtain the corresponding potential risk areas.

[0012] Further, after receiving the data collection instruction, a sensor network is deployed in the potential risk area, the sensor network collects data in the potential risk area, and the acquired fire monitoring data is aggregated to generate a fire monitoring data set; Taking fire monitoring data as input, the trained abnormal data recognition model is used to identify abnormal data. If the proportion of abnormal data exceeds the preset proportion, the operating temperature data of each lithium battery pack is collected and summarized to generate a battery operating temperature data set.

[0013] Furthermore, the operating temperature data in the battery operating temperature data set are arranged along the time axis, and the operating temperature value is constructed according to the change of the operating temperature data under dimensionless conditions. , as follows: ; Where: is the temperature of the i-th group of batteries at time t, is the initial temperature, is the ambient temperature at time t-1, is the load power at time t-1, and The influence coefficients of the ambient temperature meter load power are all between 0 and 1. is the random noise term at time t-1.

[0014] Furthermore, the sensor network collects fire monitoring data in the potential risk area. If there is an abnormality in the fire monitoring data, the abnormality degree is constructed based on the abnormal state of the data. If the abnormality degree exceeds the preset abnormality threshold, a first-level alarm command is issued to the outside and the fire monitoring data is sent to the cloud.

[0015] Furthermore, when the number of first-level alarm instructions currently received continuously exceeds expectations, or when the continuously acquired abnormality degrees are arranged along the time axis and the continuously acquired abnormality degrees are in an upward trend, or when the area where the abnormality degrees exceed the abnormality threshold is in a spreading state, a second-level alarm instruction is issued to the outside.

[0016] Furthermore, if a secondary alarm command is received, feature extraction is performed on the fire monitoring data, charging scene data, and operation data of the lithium battery pack in the potential risk area to obtain fire emergency features; The basis is the correspondence between fire emergency characteristics and emergency plans, and emergency measures are matched in the pre-built emergency plan library.

[0017] Furthermore, in the potential risk area that meets the expected scenario, after identifying and acquiring the fire data and the location data of the corresponding lithium battery pack, the automatic fire extinguishing system is turned on to extinguish the fire; The sensor network collects fire status data in the warning area, takes the fire status data as input, uses the trained area recognition model to identify the traversable area, and marks the obtained traversable area on the electronic map.

[0018] Further, if the secondary alarm command is not received, the status data of the lithium battery pack is checked and collected; With the status data as input, the trained battery abnormality recognition model is used to perform abnormality recognition. If the lithium battery pack currently has an abnormality, an analysis instruction is issued to the outside after the current abnormality type is identified.

[0019] Furthermore, after receiving the analysis instruction, the historical temperature data and charging scene data of the lithium battery pack are collected, and the key influencing factors affecting the high temperature state are obtained through principal component analysis; Using regulating the operating temperature of lithium battery packs as keywords, a knowledge graph for battery pack temperature regulation is constructed.

[0020] The multi-level fire warning system of energy storage equipment includes a scene difference analysis unit, which collects the current charging scene data in the warning area, builds a deviation value based on the difference between it and the preset scene, and if the deviation value exceeds expectations, screens out potential risk areas in the warning area and issues data collection instructions to the outside; The temperature warning unit collects fire monitoring data in the potential risk area. If the proportion of abnormal data exceeds expectations, the operating temperature data of the lithium battery pack is collected and an operating temperature value is generated. If the operating temperature value exceeds the operating temperature threshold, the fire warning mechanism is activated. The data anomaly analysis unit sends a first-level alarm command and receives corresponding feedback data if the abnormality of the fire monitoring data in the potential risk area exceeds expectations. If the feedback data is inconsistent with expectations, a second-level alarm command is sent to the outside. The path planning unit, if receiving a level 2 alarm command, processes the existing fire risk, identifies the traversable area based on the fire status data, and plans an emergency escape path within the warning area. If there is no emergency escape path, a level 3 alarm command is issued to the outside. The optimization processing unit, if no secondary alarm instruction is received, analyzes and obtains the key influencing factors that cause the lithium battery pack to enter a high temperature state when there is an abnormality in the lithium battery pack. Based on the key influencing factors, the battery pack temperature regulation knowledge graph matches the corresponding processing strategy.

[0021] (III) Beneficial effects The present invention provides a multi-level fire warning system for energy storage equipment, which has the following beneficial effects:

[0022] 1. Based on the deviation value, it can be judged whether there are any abnormalities in the current operating environment. When there are abnormalities, an early warning can be issued to the outside in advance; the risk points can be screened out and potential risk areas can be obtained, and emergency plans can be prepared in advance.

[0023] 2. Monitor the operating temperature data of the lithium battery pack and use the operating temperature as a representative parameter to determine the operating status of the lithium battery pack; generate an operating temperature value , make an overall judgment on the operating status of the lithium battery pack, enable the early warning mechanism when the operating temperature is gradually rising and the operating status is gradually deteriorating, and generate an alarm before the lithium battery pack creates an operating risk, which can reduce the occurrence of safety hazards.

[0024] 3. By identifying abnormalities in each data and determining the degree of abnormality, an abnormality degree is constructed to judge the abnormality of each fire monitoring data. When the degree of abnormality of the fire monitoring data exceeds expectations, an alarm is generated to facilitate early processing.

[0025] 4. By combining alarm feedback data to build a multi-level alarm mechanism and issue multi-level alarms, when the risk of fire in the warning area is high or a fire has already occurred, it ensures that staff or managers can pay full attention to and understand the current situation to prevent the alarm instructions from being ignored.

[0026] 5. When a fire occurs or there is a fire hazard in the warning area or the screened potential risk area, emergency measures are matched according to the emergency characteristics. By implementing emergency measures, the current fire risk can be alleviated.

[0027] 6. Based on the collected data, the traversable areas and corresponding evacuation routes are identified through data monitoring and identification to ensure the rapid evacuation of personnel in the warning area and protect their safety; and by issuing three-level alarm commands, the current alarm mechanism is enriched.

[0028] 7. By collecting various data such as the appearance of the lithium battery pack, verify whether there is any abnormality in the current appearance and physical state of the lithium battery pack. If there is any abnormality, it can be maintained or replaced in time to prevent fire caused by failure of the lithium battery pack; after screening out the key factors that cause the high operating temperature of the lithium battery pack through principal component analysis, match the corresponding processing strategy according to the battery pack temperature regulation knowledge graph, carry out targeted maintenance on the lithium battery pack, improve the operating state of the lithium battery pack, and reduce the subsequent safety hazards and fire risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a structural schematic diagram of a multi-level fire warning method for energy storage equipment of the present invention; Figure 2 It is a schematic diagram of the structure of the multi-level fire warning system of the energy storage device of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] See also Figure 1 The present invention provides a multi-level fire warning method for energy storage equipment, comprising: Step 1: Collect the current charging scenario data in the warning area, and construct a deviation value based on the difference between the current charging scenario and the preset scenario. If the deviation value exceeds the expectation, select the potential risk area in the warning area and issue a data collection instruction to the outside. The step 1 includes the following contents: Step 101: Taking the energy storage device distribution area as the early warning area, constructing an electronic map covering the early warning area; collecting and recording charging scene data of energy storage devices, such as lithium battery packs or lead-acid battery packs, etc., wherein the following takes the lithium battery pack as an example, and the charging scene data includes the ambient temperature, charging state and dust accumulation data in the adjacent area, and the charging scene data set in the early warning area is aggregated and generated; Step 102: According to the fire hazard management expectations in the warning area, set the corresponding qualified interval for each charging scenario parameter and obtain the corresponding preset scenario. Under dimensionless conditions, construct the deviation value according to the gap between the current charging scenario and the preset scenario. , as follows: ; ; Where: is the state vector, including the ambient temperature , Charging status and dust accumulation ; is the deviation vector; is the time decay factor, is the time weighting function, , is the weight coefficient, , and The values ​​of fall between 0 and 1; The duration of the charging scenario; , , Charging status , dust accumulation and ambient temperature The minimum value of , , Charging status , dust accumulation and ambient temperature The maximum value of If the deviation value If the deviation exceeds the preset threshold, it means that the gap between the current charging scenario and the preset scenario is large, which may have a certain impact on the subsequent charging safety. At this time, an area identification command is issued to the outside. When in use, when the lithium battery pack is in operation, considering that the lithium battery pack can be in a better operating scenario, the current scenario data of the lithium battery pack is compared with the preset scenario data to obtain the deviation value. , according to the deviation value It can judge whether there are any abnormalities in the current operating environment, and issue early warnings to the outside world when there are any abnormalities. Step 103: After receiving the area recognition instruction, the monitoring point is marked with the charging scene data. If the deviation value between the current monitoring point and its adjacent area If it exceeds expectations, it will be regarded as a risk point and marked on the electronic map; Use pre-trained clustering algorithms to cluster several risk points, obtain corresponding potential risk areas, and issue data collection instructions to the outside world; When using, combine the contents in steps 101 to 103: As a further content, after obtaining various data from various monitoring points, the risk points are screened out and the potential risk areas are obtained. Considering that there are certain fire hazards in the potential risk areas, in order to reduce the risk of fire, emergency plans can be prepared in advance.

[0032] Combined with the above application and the contents of the prior art: When energy storage equipment, especially lithium battery packs or lead-acid battery packs, are in operation, when the battery packs have a short circuit or similar faults, it is particularly prone to fire or spontaneous combustion. In particular, the battery also has the characteristics of high-temperature explosion. Once there is a fire hazard, the risk is very high. Therefore, fire monitoring and early warning devices and fire extinguishing systems need to be fully arranged in the coverage area of ​​the battery pack to facilitate timely handling and response when a fire may occur.

[0033] In the existing multi-level early warning method for fire of energy storage equipment, a sensor network is usually arranged in the area where the battery pack is located, especially in the potential risk area, and the sensor network monitors various data, such as temperature, light and smoke data, etc. When one or more of them are abnormal, an early warning or alarm is immediately issued to the outside; However, there are certain anomalies in the charging scenarios in the area where the battery pack is located. In particular, when the ambient temperature is high, the lithium battery is overcharged to a certain extent, or even when dust accumulates on the external surface, the operating temperature of the battery pack is likely to rise, and the risk of operating failures or even spontaneous combustion will be further increased. In this relatively abnormal charging scenario, when conducting fire warnings and alarms in potential risk areas, if only a static first-level alarm or repeated alarm mode is still used, the staff in the potential risk area may neglect it and find it difficult to take timely and effective measures.

[0034] Step 2: Collect fire monitoring data in the potential risk area. If the proportion of abnormal data exceeds expectations, collect the operating temperature data of the lithium battery pack and generate an operating temperature value. If the operating temperature value exceeds the operating temperature threshold, activate the fire warning mechanism; The step 2 includes the following contents: Step 201: after receiving the data collection instruction, a sensor network, such as a thermistor, a temperature sensor, an infrared sensor, a smoke sensor, and a gas detector, is deployed at least in the potential risk area to ensure that the sensor network can cover all potential risk areas; The sensor network collects data at least in the potential risk area, such as temperature data, smoke data, humidity data, and sound data, and aggregates the acquired fire monitoring data to generate a fire monitoring data set; fire monitoring can be achieved at least in the potential risk area; Step 202: training a convolutional neural network with the labeled sample data to obtain a trained abnormal data recognition model; Using fire monitoring data as input, the trained abnormal data recognition model is used to identify abnormal data. If the proportion of abnormal data exceeds the preset proportion, the operating temperature data of each lithium battery pack is monitored and collected, and the continuously acquired operating temperature data is aggregated to generate a battery operating temperature data set; When in use, when abnormal data is continuously generated in the potential risk area, it indicates that a certain degree of abnormality may have already existed in the potential risk area. As a feedback method, the operating temperature data of the lithium battery pack can be monitored, and the operating temperature can be used as a representative parameter to judge the operating status of the lithium battery pack; Step 203: Arrange the operating temperature data in the battery operating temperature data set along the time axis, and construct an operating temperature value according to the change of the operating temperature data under dimensionless conditions. , as follows: ; Where: is the temperature of the i-th group of batteries at time t, is the initial temperature, is the ambient temperature at time t-1, is the load power at time t-1, and The influence coefficients of the ambient temperature meter load power are all between 0 and 1. is the random noise term at time t-1; Based on historical data and expectations for the operating temperature management of lithium battery packs, the operating temperature threshold is pre-set; if the operating temperature value If the operating temperature exceeds the threshold, it means that the operating temperature of the lithium battery pack has not been reduced, and the potential safety hazards of the lithium battery pack in the warning area are continuously increasing. At this time, the fire warning mechanism for the lithium battery pack is activated; When using, combine the contents in steps 201 to 203: By generating operating temperature values , the operating status of the lithium battery pack can be judged as a whole. When the operating temperature is gradually rising and the operating status is gradually deteriorating, the early warning mechanism can be activated, and an alarm can be generated before the lithium battery pack creates an operating risk, which can reduce the occurrence of safety hazards.

[0035] Step 3: If the abnormality of the fire monitoring data in the potential risk area exceeds expectations, a first-level alarm command is issued and corresponding feedback data is received. If the feedback data is inconsistent with expectations, a second-level alarm command is issued to the outside; The step three includes the following contents: Step 301: The sensor network collects fire monitoring data in the potential risk area. If the fire monitoring data is abnormal, under dimensionless conditions, an abnormality degree is constructed according to the abnormal state of the data. , as follows: ; Where: To observe the number of nodes, is the observed value of the fire monitoring parameter at time node t, is the corresponding mean, is the standard deviation of the fire monitoring parameter observations, is the interval between two abnormal nodes, is the distance attenuation factor, is the weight coefficient of the i-th fire monitoring parameter, and its value falls between 0 and 1. The number of parameters monitored for fire; If the abnormality obtained If the abnormality exceeds the preset threshold, it means that there is a certain fire risk in the current warning area, which needs to be dealt with in time. A first-level alarm command is issued to the outside, and the fire monitoring data is sent to the cloud. When in use, based on the collection of fire monitoring data in the potential risk area, the abnormality degree is constructed by identifying the abnormality of each data and determining the abnormality degree. By judging the abnormality of various fire monitoring data, an alarm can be generated when the abnormality of the fire monitoring data exceeds expectations, so as to facilitate early processing; Step 302: When the number of first-level alarm instructions currently received continuously exceeds the expected number, or the abnormality degrees continuously obtained are arranged along the time axis, the abnormality degrees continuously obtained are arranged along the time axis, When the fire is on an upward trend, or after marking each monitoring point in the potential risk area with fire monitoring data, if the area in the vicinity of the potential risk area where the abnormality exceeds the abnormal threshold is in a spreading state, a secondary alarm instruction is issued to the outside; When using, combine the contents in steps 301 to 302: On the basis of the first-level alarm command that has been issued, feedback data continues to be collected, and based on the feedback data, it is confirmed whether a new alarm command needs to be issued, thereby completing the construction of a multi-level alarm mechanism; different from a single static alarm, by combining the alarm feedback data to build a multi-level alarm mechanism and issue multi-level alarms, when the risk of fire in the warning area is high or a fire has already occurred, it can ensure that staff or managers can pay full attention to and understand the current situation, prevent the alarm command from being ignored, and reduce safety hazards.

[0036] Step 4: If a level 2 alarm command is received, the existing fire risk is processed, and after identifying the traversable area based on the fire status data, an emergency escape route is planned within the warning area. If no emergency escape route exists, a level 3 alarm command is issued to the outside. The steps include the following: Step 401: If a level 2 alarm command is received, feature extraction is performed on the fire monitoring data, charging scene data, and operation data of the lithium battery pack in the potential risk area to obtain fire emergency features; Based on the correspondence between fire emergency characteristics and emergency plans, emergency measures are matched in the pre-built emergency plan library, such as emergency power outage or heat dissipation, to deal with the existing fire risks; When in use, when a fire occurs or there is a fire hazard in the warning area or the screened potential risk area, emergency measures are matched according to the emergency characteristics, and the current fire risk can be alleviated by executing the emergency measures.

[0037] Step 402: When the potential risk area meets the expected scenario, such as smoke, flames, and high temperature, after identifying and obtaining detailed fire data and the location data of the corresponding lithium battery pack, the automatic fire extinguishing system is turned on to extinguish the fire; as a further content, after the fire location is monitored, the fire is extinguished through the automatic fire extinguishing system to reduce the impact of the fire; After training the machine learning algorithm with the labeled sample data, a trained area recognition model is obtained; the fire status data such as temperature and humidity data, air quality data and light data in the warning area are collected by the sensor network, and the fire status data is used as input to identify the traversable area using the trained area recognition model; The traversable area is marked on the electronic map, and the pre-trained path planning algorithm plans an emergency escape route for the staff in the warning area, and gives guidance prompts in real time. If there is no emergency escape route, a third-level alarm command is issued to the outside. When using, combine the contents in steps 401 and 402: As a further step, after a fire has occurred and rapid emergency fire extinguishing has been carried out, based on the collected data, the traversable areas and corresponding evacuation routes can be confirmed through data monitoring and identification, which can ensure the rapid evacuation of personnel in the warning area and protect the safety of the personnel; and by issuing three-level alarm commands, the current alarm mechanism can be enriched.

[0038] Step 5: If no secondary alarm command is received, when the lithium battery pack is abnormal, analyze and obtain the key influencing factors that cause the lithium battery pack to enter a high temperature state, and match the corresponding processing strategy from the battery pack temperature regulation knowledge graph based on the key influencing factors; The step five includes the following contents: Step 501: If no secondary alarm instruction is received, check and collect status data of the lithium battery pack, including leakage, expansion, dust accumulation and rust, etc.; train a convolutional neural network with the labeled sample data to obtain a trained battery abnormality recognition model; With the status data as input, the trained battery abnormality recognition model is used to identify abnormalities. If the lithium battery pack currently has an abnormality, an analysis instruction is issued to the outside after the current abnormality type is identified; When in use, when no fire occurs or the fire is handled quickly, the appearance and other data of the lithium battery pack are collected to verify whether there is any abnormality in the current appearance and physical state of the lithium battery pack. If there is any abnormality, maintenance or replacement can be carried out in time to prevent fire caused by failure of the lithium battery pack; Step 502: after receiving the analysis instruction, collect the historical temperature data and charging scene data of the lithium battery pack, and obtain the key influencing factors affecting the high temperature state, such as overload and insufficient heat dissipation, through principal component analysis; Taking regulating the operating temperature of lithium battery packs as the keyword, after deep retrieval and entity relationship building, a knowledge graph of battery pack temperature regulation is constructed; according to the correspondence between key influencing factors and lithium battery temperature regulation solutions, the battery pack temperature regulation knowledge graph is matched with corresponding processing strategies, such as considering optimizing equipment layout, adding ventilation equipment, and improving heat dissipation materials; When using, combine the contents in steps 501 and 502: When a lithium battery pack has a certain degree of abnormality and its operating temperature exceeds expectations, after screening out the key factors that cause the lithium battery pack to operate at a higher temperature through principal component analysis, the corresponding processing strategy can be matched according to the battery pack temperature regulation knowledge graph. Targeted maintenance can be performed on the lithium battery pack to improve the operating status of the lithium battery pack and reduce subsequent safety hazards and fire risks.

[0039] The construction method of the battery pack temperature regulation knowledge graph can refer to the following: Clearly define your build goals First, we need to clearly define the goal of building a knowledge graph for battery pack temperature regulation, for example, is it to improve the performance of the battery pack, extend its service life, or meet the needs of a specific application scenario? Clarifying the goal helps determine the content and scope of the knowledge graph.

[0040] Gather relevant knowledge Basic knowledge of batteries: including battery types (such as lithium-ion batteries, lead-acid batteries, etc.), working principles, chemical composition, etc. Battery thermal management: Understand the basic principles, methods and technologies of battery thermal management, such as cooling (air cooling, liquid cooling, direct cooling, etc.), heating, temperature balance, etc. Material science: Understand the materials related to battery thermal management, such as coolant, heat pipe, phase change material, etc. Heat conduction and convection: Master the basic principles of heat conduction and convection, and their applications in battery thermal management. Intelligent temperature control system: Understand the composition, working principle and adjustment method of the intelligent temperature control system.

[0041] Determine the structure of the knowledge graph Hierarchical structure: Divide the knowledge graph into multiple levels, such as basic knowledge layer, technical method layer, application scenario layer, etc. Relationship network: Determine the relationship between different knowledge points, such as cause-effect relationship, parallel relationship, inclusion relationship, etc., to form a relationship network. Attributes and attribute values: Add attributes and attribute values ​​to each knowledge point, such as the type of battery (attribute value: lithium-ion battery), the type of coolant (attribute value: water and ethylene glycol mixed solution), etc.

[0042] Data acquisition and processing Data collection: Collect relevant data and information by consulting literature, patents, technical reports, etc. Data cleaning: Remove duplicate, erroneous or irrelevant data to ensure the accuracy and completeness of the data. Data integration: Integrate the collected data into the knowledge graph to form a complete knowledge system.

[0043] Construction and visualization of knowledge graph Select tools: Choose appropriate tools (such as Neo4j, Ontolingua, etc.) to build the knowledge graph. Build the graph: Based on the collected knowledge and data, build the nodes, relationships, and attributes of the knowledge graph. Visualize: Use visualization tools (such as D3.js, Graphviz, etc.) to present the knowledge graph for easy understanding and application.

[0044] See also Figure 2 The present invention provides a multi-level fire warning system for energy storage equipment, comprising: The scenario difference analysis unit collects the current charging scenario data in the warning area, and constructs a deviation value based on the difference between the current charging scenario and the preset scenario. If the deviation value exceeds the expectation, the potential risk area is screened out in the warning area, and a data collection instruction is issued to the outside. The temperature warning unit collects fire monitoring data in the potential risk area. If the proportion of abnormal data exceeds expectations, the operating temperature data of the lithium battery pack is collected and an operating temperature value is generated. If the operating temperature value exceeds the operating temperature threshold, the fire warning mechanism is activated. The data anomaly analysis unit sends a first-level alarm command and receives corresponding feedback data if the abnormality of the fire monitoring data in the potential risk area exceeds expectations. If the feedback data is inconsistent with expectations, a second-level alarm command is sent to the outside. The path planning unit, if receiving a level 2 alarm command, processes the existing fire risk, identifies the traversable area based on the fire status data, and plans an emergency escape path within the warning area. If there is no emergency escape path, a level 3 alarm command is issued to the outside. The optimization processing unit, if no secondary alarm instruction is received, analyzes and obtains the key influencing factors that cause the lithium battery pack to enter a high temperature state when there is an abnormality in the lithium battery pack. Based on the key influencing factors, the battery pack temperature regulation knowledge graph matches the corresponding processing strategy.

[0045] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0046] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0047] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0048] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0049] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A multi-level fire warning method for energy storage equipment, characterized in that: include, Collect the current charging scenario data in the warning area, and build a deviation value based on the difference between it and the preset scenario. If the deviation value exceeds expectations, screen out potential risk areas in the warning area and issue data collection instructions to the outside. Collect fire monitoring data in potential risk areas. If the proportion of abnormal data exceeds expectations, collect the operating temperature data of the lithium battery pack and generate an operating temperature value. If the acquired operating temperature value exceeds the operating temperature threshold, activate the fire warning mechanism. If the abnormality of fire monitoring data in the potential risk area exceeds expectations, a first-level alarm command will be issued and corresponding feedback data will be received. If the feedback data is inconsistent with expectations, a second-level alarm command will be issued to the outside. If a level 2 alarm command is received, the existing fire risk is processed, and after identifying the traversable area based on the fire status data, an emergency escape route is planned within the warning area. If there is no emergency escape route, a level 3 alarm command is issued to the outside. If no level 2 alarm command is received, when there is an abnormality in the lithium battery pack, the key influencing factors that cause the lithium battery pack to enter a high temperature state are analyzed and obtained, and based on the key influencing factors, the battery pack temperature regulation knowledge graph is matched with the corresponding processing strategy.

2. The multi-level fire warning method for energy storage equipment according to claim 1, characterized in that: Collect charging scenario data of lithium battery packs, including at least ambient temperature, charging status and dust accumulation data, and summarize them to generate a charging scenario data set; set the corresponding qualified interval for each charging scenario parameter and obtain the corresponding preset scenario, and construct a deviation value based on the gap between the charging scenario and the preset scenario. If the deviation value exceeds the preset deviation threshold, issue an area recognition instruction to the outside.

3. The multi-level fire warning method for energy storage equipment according to claim 2, characterized in that: After receiving the area identification command, the monitoring point is marked with the charging scene data. If the deviation value of the current monitoring point and its adjacent area exceeds the expectation, it is regarded as a risk point and marked on the electronic map; Use a pre-trained clustering algorithm to cluster several risk points and obtain the corresponding potential risk areas.

4. The multi-level fire warning method for energy storage equipment according to claim 3 is characterized in that: After receiving the data collection instruction, a sensor network is deployed in the potential risk area, and the sensor network collects data in the potential risk area, and the acquired fire monitoring data is aggregated to generate a fire monitoring data set; Taking fire monitoring data as input, the trained abnormal data recognition model is used to identify abnormal data. If the proportion of abnormal data exceeds the preset proportion, the operating temperature data of each lithium battery pack is collected and summarized to generate a battery operating temperature data set.

5. The multi-level fire warning method for energy storage equipment according to claim 4, characterized in that: Arrange the operating temperature data in the battery operating temperature data set along the time axis, and construct the operating temperature value according to the change of the operating temperature data under dimensionless conditions. , as follows: ； Where: is the temperature of the i-th group of batteries at time t, is the initial temperature, is the ambient temperature at time t-1, is the load power at time t-1, and The influence coefficients of the ambient temperature meter load power are all between 0 and 1. is the random noise term at time t-1.

6. The multi-level fire warning method for energy storage equipment according to claim 5, characterized in that: The sensor network collects fire monitoring data in potential risk areas. If there is an abnormality in the fire monitoring data, the abnormality degree is constructed based on the abnormal state of the data. If the abnormality degree exceeds the preset abnormal threshold, a first-level alarm command is issued to the outside and the fire monitoring data is sent to the cloud.

7. The multi-level fire warning method for energy storage equipment according to claim 6, characterized in that: When the number of first-level alarm instructions currently received continuously exceeds expectations, or when the continuously acquired abnormality degrees are arranged along the time axis and the abnormality degrees are on an upward trend, or when the area where the abnormality degrees exceed the abnormality threshold is in a spreading state, a second-level alarm instruction is issued to the outside.

8. The multi-level fire warning method for energy storage equipment according to claim 7, characterized in that: If a level 2 alarm command is received, feature extraction is performed on the fire monitoring data, charging scene data, and lithium battery pack operation data in the potential risk area to obtain fire emergency features; The basis is the correspondence between fire emergency characteristics and emergency plans, and emergency measures are matched in the pre-built emergency plan library.

9. The multi-level fire warning method for energy storage equipment according to claim 8, characterized in that: When the potential risk area meets the expected scenario, after identifying and obtaining the fire data and the corresponding location data of the lithium battery pack, the automatic fire extinguishing system is activated to extinguish the fire; The sensor network collects fire status data in the warning area, takes the fire status data as input, uses the trained area recognition model to identify the traversable area, and marks the obtained traversable area on the electronic map.

10. The multi-level fire warning method for energy storage equipment according to claim 9, characterized in that: If no secondary alarm command is received, check and collect status data of the lithium battery pack; With the status data as input, the trained battery abnormality recognition model is used to perform abnormality recognition. If the lithium battery pack currently has an abnormality, an analysis instruction is issued to the outside after the current abnormality type is identified.

11. The multi-level fire warning method for energy storage equipment according to claim 10, characterized in that: After receiving the analysis command, the historical temperature data and charging scene data of the lithium battery pack are collected, and the key influencing factors affecting the high temperature state are obtained through principal component analysis; Using regulating the operating temperature of lithium battery packs as keywords, a knowledge graph for battery pack temperature regulation is constructed.

12. A multi-level fire warning system for energy storage equipment, characterized by: include, The scenario difference analysis unit collects the current charging scenario data in the warning area, and constructs a deviation value based on the difference between the current charging scenario and the preset scenario. If the deviation value exceeds the expectation, the potential risk area is screened out in the warning area, and a data collection instruction is issued to the outside. The temperature warning unit collects fire monitoring data in the potential risk area. If the proportion of abnormal data exceeds expectations, the operating temperature data of the lithium battery pack is collected and an operating temperature value is generated. If the operating temperature value exceeds the operating temperature threshold, the fire warning mechanism is activated. The data anomaly analysis unit sends a first-level alarm command and receives corresponding feedback data if the abnormality of the fire monitoring data in the potential risk area exceeds expectations. If the feedback data is inconsistent with expectations, a second-level alarm command is sent to the outside. The path planning unit, if receiving a level 2 alarm command, processes the existing fire risk, identifies the traversable area based on the fire status data, and plans an emergency escape path within the warning area. If there is no emergency escape path, a level 3 alarm command is issued to the outside. The optimization processing unit, if no secondary alarm instruction is received, analyzes and obtains the key influencing factors that cause the lithium battery pack to enter a high temperature state when there is an abnormality in the lithium battery pack. Based on the key influencing factors, the battery pack temperature regulation knowledge graph matches the corresponding processing strategy.

Images