Energy Storage Equipment Fault Monitoring Platform under Remote Identification

Abstract

The present invention discloses a fault monitoring platform for energy storage devices under remote identification, which relates to the technical field of fault monitoring. The platform includes: a path analysis unit for obtaining a fault transmission analysis tree; a node configuration unit for obtaining a fault monitoring architecture; an activation permission configuration unit for configuring hierarchical activation permissions; a fault analysis unit for obtaining real-time energy storage faults; and a load distribution unit for performing load distribution of energy storage devices. It solves the technical problems of being unable to accurately identify the fault transmission path and respond in a timely manner during the fault monitoring process of energy storage devices, and realizes the technical effect of constructing an edge monitoring architecture through a fault transmission analysis tree and performing fault location and analysis based on hierarchical activation permissions, thereby achieving fast and accurate fault monitoring and intelligent load distribution of energy storage devices.

Description

Technical Field

[0001] This application relates to the technical field of fault monitoring, and particularly to a fault monitoring platform for energy storage devices under remote identification. Background Art

[0002] With the rapid development of new energy technologies, energy storage devices play an important role in the power system. Energy storage devices can stabilize the grid frequency, alleviate power fluctuations, and improve the overall efficiency and stability of the power system by storing and releasing energy under the condition of unbalanced peak and valley demands of the power grid. However, energy storage devices consist of a large number of battery modules, which are prone to failures during long-term operation, resulting in a decline in device performance and even potential safety hazards. Existing fault monitoring systems for energy storage devices usually rely on a centralized monitoring mode, that is, a remote center uniformly monitors and processes faults of the devices. However, this centralized fault monitoring method has significant deficiencies: when a system failure occurs, it is difficult for the remote center to respond and isolate the fault in a timely manner due to data transmission delays and resource allocation bottlenecks. This lag may lead to an expansion of the fault range, affecting the reliability and safety of energy storage devices. In addition, the fault transmission paths inside energy storage devices are complex, and there are various mutual influence transmission relationships such as electrical and thermal between different battery modules. Traditional monitoring means lack a comprehensive analysis of the fault transmission paths and cannot accurately judge the fault diffusion situation, resulting in inaccurate fault location, thus increasing the difficulty and cost of repair and maintenance. Summary of the Invention

[0003] This application provides a fault monitoring platform for energy storage devices under remote identification, solves the technical problems of being unable to accurately identify the fault transmission path and respond in a timely manner during the fault monitoring process of energy storage devices, realizes the construction of an edge monitoring architecture through a fault transmission analysis tree, and conducts fault location and analysis based on hierarchical activation permissions, thereby achieving the technical effects of fast and accurate fault monitoring and intelligent load allocation for energy storage devices.

[0004] The present application provides a fault monitoring platform for energy storage devices under remote identification. The platform includes: a path analysis unit for analyzing the fault transmission path of the energy storage device to obtain a fault transmission analysis tree; a node configuration unit for configuring edge detection nodes for the energy storage device according to the fault transmission analysis tree to obtain a fault monitoring architecture; an activation permission configuration unit for, when a primary edge detection node identifies a battery module fault, activating constraints for secondary nodes with the fault monitoring architecture as the basis, locating K local edge nodes, and configuring hierarchical activation permissions for the K local edge nodes according to the fault transmission analysis tree; a fault analysis unit for the K local edge nodes to perform in-depth fault analysis on K battery modules based on the hierarchical activation permissions to obtain real-time energy storage faults; and a load distribution unit for, after an edge management coordination node receives the real-time energy storage faults, sending the real-time energy storage faults to a remote operation and maintenance center and performing load distribution for the energy storage device according to the real-time energy storage faults, where the edge management coordination node is the root node of the fault monitoring architecture.

[0005] It is intended to analyze the fault transmission path of the energy storage device through the present application to obtain a fault transmission analysis tree; configure edge detection nodes for the energy storage device according to the fault transmission analysis tree to obtain a fault monitoring architecture; when a primary edge detection node identifies a battery module fault, activate constraints for secondary nodes with the fault monitoring architecture as the basis, locate K local edge nodes, and configure hierarchical activation permissions for the K local edge nodes according to the fault transmission analysis tree; perform in-depth fault analysis on K battery modules by the K local edge nodes based on the hierarchical activation permissions to obtain real-time energy storage faults; after an edge management coordination node receives the real-time energy storage faults, send the real-time energy storage faults to a remote operation and maintenance center and perform load distribution for the energy storage device according to the real-time energy storage faults, where the edge management coordination node is the root node of the fault monitoring architecture. This solves the technical problems of unable to accurately identify the fault transmission path and respond in a timely manner during the fault monitoring process of energy storage devices, and realizes the technical effect of constructing an edge monitoring architecture through a fault transmission analysis tree and performing fault location and analysis based on hierarchical activation permissions, so as to achieve fast and accurate fault monitoring and intelligent load distribution of energy storage devices. Brief Description of the Drawings

[0006] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments of the present disclosure will be briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the platform according to the embodiments of the present application. It should be understood that the operations described above or below do not necessarily need to be executed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or several steps can be removed from these processes.

[0007] Figure 1 Schematic structural diagram of the energy storage device fault monitoring platform under remote identification provided by the embodiment of the present application;

[0008] Figure 2 Schematic diagram of the execution process for obtaining the fault transfer analysis tree in the energy storage device fault monitoring platform under remote identification provided by the embodiment of the present application.

[0009] Explanation of reference numerals: Path analysis unit 1, node configuration unit 2, activation permission configuration unit 3, fault analysis unit 4, load distribution unit 5. Specific implementation manners

[0010] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific implementation manners of the present application are hereby exemplified below.

[0011] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0012] In the following description, "some embodiments" are involved, which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. The terms "first\second" involved are only used to distinguish similar objects and do not represent a specific order for the objects. The terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, platform, product or server including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules not clearly listed or inherent to these processes, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application.

[0013] The embodiment of the present application provides an energy storage device fault monitoring platform under remote identification, as Figure 1 shown, the platform includes:

[0014] A path analysis unit, configured to perform fault transfer path analysis on the energy storage device to obtain a fault transfer analysis tree.

[0015] In one embodiment, in the path analysis unit, in order to comprehensively understand the propagation mode and influence range of faults inside the device, the system terminal identifies the connection relationships between various components (such as battery modules, lines, interfaces, etc.) in the energy storage device, constructs one or more potential fault propagation paths, and forms an initial fault transfer tree. These paths can show how a fault spreads from a certain component to other connected components when a component fails, thus affecting the entire device. Subsequently, the initial fault transfer tree is updated through protection delay and thermal runaway simulation to form a fault transfer analysis tree. Through this tree structure, the possible diffusion direction of the fault, the affected area, and the priority order of fault propagation can be clearly understood. This analysis tree provides a clear path reference for subsequent fault detection and handling, which helps to achieve accurate fault monitoring and rapid response.

[0016] Further, as Figure 2 shown, this application provides a method for analyzing the fault transfer path of an energy storage device to obtain a fault transfer analysis tree, including:

[0017] Performing fault transfer path fitting according to the electrical connection relationship of the energy storage device to generate an initial fault transfer tree; performing fault diffusion analysis based on protection delay and updating the initial fault transfer tree according to the analysis result to obtain a fault diffusion transfer tree; performing thermal runaway simulation according to the distribution of battery modules in the energy storage device and merging nodes of the fault diffusion transfer tree according to the simulation result to obtain the fault transfer analysis tree.

[0018] Preferably, the system terminal analyzes the electrical connection relationships of each battery module in the energy storage device. The electrical connection relationships of the energy storage device usually include series or parallel connections between modules, and these connections determine the distribution paths of current and voltage. Based on the identified electrical connection relationships, the fault propagation paths between each module are analyzed. For each module, the modules directly connected to it are determined as the first-level propagation nodes, that is, if this module fails, the fault may directly transfer to these first-level propagation nodes, and then the second-level propagation nodes, third-level propagation nodes, and so on are analyzed in turn to form a complete fault transfer path. This step can help identify how a fault spreads from one module to other modules under the electrical connection relationship. Subsequently, according to the identified fault transfer paths, each battery module and their fault propagation relationships are drawn into a tree structure to obtain an initial fault transfer tree. In this tree structure, each node represents a battery module, and the edge represents the fault transfer relationship between modules. The root node of the tree can be the first module to fail, and the first-level, second-level, and deeper nodes represent other modules affected by this faulty module, arranged in the order of propagation.

[0019] Preferably, after determining the initial fault propagation tree, analyze the impact of the delay in the protection mechanism on fault spread. In an energy storage device, the response time of the protection mechanism directly affects the degree of fault spread. For example, a protection mechanism with a large delay may cause the fault to spread to adjacent modules before being detected. The system terminal collects the time from the occurrence of a fault to its detection and response for other identical modules based on the protection system delay of each battery module in the energy storage device. For example, the overcurrent protection delay of a certain module may be several milliseconds, while the tripping time of the circuit breaker for isolating the fault may be longer. Combine this delay data with the initial fault propagation tree as the initial condition for the next analysis. In the initial fault propagation tree, gradually analyze how the protection delay of each module may affect fault propagation. For each node (module), count the downstream nodes to which the fault can spread within the delay time in the case where the fault is not completely isolated within the protection system delay of that node, and use the propagation paths between these downstream nodes and the fault node to update the initial fault propagation tree. For example, if the protection system delay of node A is long, and during this delay period, the counted fault spreads to nodes B and C, then at this time, these two propagation paths will be used to supplement the initial fault propagation tree. By performing similar analyses on each node, identify the propagation paths that may be triggered by the delay, thereby gradually improving the initial fault propagation tree to obtain the fault spread propagation tree.

[0020] Preferably, after obtaining the fault diffusion transfer tree, the system terminal further analyzes the fault diffusion characteristics under thermal runaway conditions. When a battery module experiences thermal runaway, heat diffuses to modules with a relatively short physical distance through the thermal coupling path between the modules, causing the temperature of adjacent modules to rise and potentially leading to their failure. To accurately understand the nodes that may be involved during thermal runaway, the system terminal determines the physical positions of each battery module in the energy storage device and the distance relationship with adjacent modules. Since the battery modules in the energy storage device are closely arranged, the distance relationship between the modules determines how heat is transferred between them. This step is to clearly identify the adjacent modules that may be affected by heat for each module. Subsequently, the determined distance relationship is input into the simulation software to adjust the distances between the respective simulation batteries in the simulation software, and a predetermined number of thermal runaway simulations are performed. The nodes that each node can diffuse to through the thermal coupling path are counted, and the number of times each diffused node is reached is also counted. Then, the nodes that do not meet the minimum number requirement are excluded from these counted diffusion node times, and the remaining nodes are merged. For example, if the nodes that node A can diffuse to without electrical connection are node D, node E, node F, and node G, and the diffusion times are 3 times, 5 times, 2 times, and 6 times respectively, since the diffusion times of node D and node F do not meet the minimum number requirement of 3 times, node D and node F are excluded, and the remaining ones are merged. The purpose of this merging process is to reflect the unique impact of thermal runaway on fault transmission and further improve the fault transmission tree. Through the above steps, the construction of the final fault transmission analysis tree is completed. The fault transmission analysis tree combines all potential fault propagation paths, including electrical connection paths, protection delay impact paths, and thermal runaway transmission paths. The analysis tree presents a branched structure, where the secondary nodes of each node include other modules that the fault of this node may affect, showing a complete fault diffusion relationship network. This tree structure helps to accurately identify the fault propagation path and provides a clear decision-making basis for fault detection and monitoring.

[0021] A node configuration unit, configured to perform edge detection node configuration on the energy storage device according to the fault transmission analysis tree to obtain a fault monitoring architecture.

[0022] In one embodiment, in the node configuration unit, the system terminal configures multiple local edge nodes on multiple battery modules of the energy storage device for real-time monitoring on the critical fault transfer paths to ensure that each potential fault path is within the monitoring coverage. Subsequently, the speed of the fault transfer paths is analyzed, and the fault transfer analysis tree is disassembled into several local delay topologies. High-level edge nodes are assigned to the paths with faster transfer speeds, while lower-level edge nodes are assigned to the slower paths to achieve optimized resource allocation. Then, according to the local delay topologies, edge detection nodes at different levels are configured to form a hierarchical monitoring network, which is communicatively connected to the edge management coordination node. As the core of the fault monitoring architecture, it is used to receive and integrate the fault data of each edge node and report to the remote operation and maintenance center when necessary, providing guarantee for the fault location and rapid response of the energy storage device.

[0023] Further, the present application provides edge detection node configuration for the energy storage device according to the fault transfer analysis tree to obtain a fault monitoring architecture, including:

[0024] Configuring multiple local edge nodes on multiple battery modules in the energy storage device; disassembling the fault transfer analysis tree into M local delay topologies according to the fault transfer delay; configuring M first-level edge detection nodes according to the M local delay topologies; dividing the multiple local edge nodes into M groups of local edge nodes according to the M local delay topologies; pre-building an edge management coordination node, and completing the construction of the fault monitoring architecture by hierarchically communicatively connecting the edge management coordination node, the M first-level edge detection nodes, and the M groups of local edge nodes.

[0025] Optionally, multiple local edge nodes are distributedly configured on each battery module in the energy storage device. Each of these nodes includes a fault identification model inside, which is responsible for real-time status monitoring of a single battery module, facilitating the capture of fault signals in the initial stage of a fault and providing a basis for subsequent multi-level monitoring. Subsequently, based on the fault transfer delay and fault detection delay of the energy storage device, the fault transfer analysis tree is disassembled into multiple local delay topologies. Each topology region represents the fault propagation delay range between a group of battery modules, making the fault transfer path within the region clearer and facilitating the subsequent setting of edge detection nodes. After that, according to the disassembled M local delay topologies, a first-level edge detection node is configured in each topology region. These nodes contain a macroscopic fault identification model inside, which has a high response priority and can quickly detect faults within the region, ensuring that faults on important propagation paths can be quickly responded to. Then, according to each local delay topology, the local edge nodes are divided into M groups. These nodes are responsible for monitoring the module status within the region in each topology region and reporting the fault information level by level to the first-level edge detection nodes in their respective regions, constructing a hierarchical fault monitoring network. Finally, an edge management coordination node is set up. This node is connected to the hierarchical fault monitoring network constructed by M first-level edge detection nodes and M groups of local edge nodes through hierarchical communication, completing the construction of the fault monitoring architecture. As the core node, the edge management coordination node undertakes multiple key tasks, including sending fault information to the remote operation and maintenance center, coordinating the load distribution of the remaining battery modules, isolating and managing the fault modules, etc., to ensure the stable operation and fault response efficiency of the entire energy storage device. This fault monitoring architecture ensures that faults can be quickly and accurately located and responded to when a fault occurs through multi-level node configuration and hierarchical management, improving the reliability and safety of the energy storage device.

[0026] For the fault identification model, the system terminal collects historical data containing key operating parameters such as current, voltage, and temperature, and labels them as faulty or normal states. Subsequently, the collected data is normalized to ensure that different parameters are within the same numerical range, thereby improving the training efficiency and stability of the model. The model adopts a multi-layer perceptron structure. The input layer receives these parameters, and after being processed by several hidden layers, it outputs a prediction result (0 represents normal, 1 represents faulty). The model uses binary cross-entropy as the loss function and selects the Adam optimizer to achieve fast convergence. During the training process, the system terminal inputs the training data into the model, calculates the error between the predicted output and the true label according to the loss function, and adjusts the weights of the model layer by layer through the backpropagation algorithm. This training process is iterated multiple times to gradually optimize the fault identification ability of the model. After each round of training, the performance of the model is evaluated using the validation set, and the hyperparameters of the model are adjusted through metrics such as accuracy and F1 score to improve the generalization ability. After training is completed, the prediction effect of the model is evaluated on the test set. If the expected accuracy is achieved, the model weights are saved for real-time monitoring of the device status and fault identification in practical applications.

[0027] Further, the present application provides for disassembling the fault transfer analysis tree into M local delay topologies according to the fault transfer delay, including:

[0028] Calculating the fault transfer delay by traversing the fault transfer analysis tree according to the fault transfer type to obtain the fault delay topology; interacting with the remote operation and maintenance center to obtain the fault detection delay; and aggregating the fault delay topology using the fault detection delay to obtain the M local delay topologies.

[0029] Optionally, the system terminal traverses each node (battery module) in the tree according to the fault transmission analysis tree of the energy storage device, and analyzes the propagation time of the fault on different types of transmission paths. The fault transmission types include electrical connection, heat conduction, etc. In each traversal process, identify the fault transmission type of the current transmission path, and through the same simulation process as described above, perform transmission simulations for a preset number of times, record the simulated fault time delays of each transmission simulation, and then calculate the average value of all the recorded simulated fault time delays to obtain the fault time delay of the starting node corresponding to this transmission path. Then, according to the positional relationship of each node, combine the corresponding fault time delays for topological connection to form a preliminary topological structure including the fault propagation time delays between nodes, that is, the fault time delay topology. After completing the construction of the fault time delay topology, interact with the remote operation and maintenance center to obtain the fault detection time delay. The fault detection time delay refers to the total time from detecting the fault signal to feeding back this information to the remote center. This data helps to understand the speed and efficiency of the system response during the fault propagation process. Using the obtained fault detection time delay, the system terminal performs topological aggregation processing on the fault time delay topology. Specifically, the agglomerative hierarchical clustering algorithm is used, and the fault detection time delay of each node is used as the eigenvalue of clustering, and gradually aggregate the nodes with similar detection time delays together to form several local time delay topology regions. Each region represents a group of nodes within a similar time delay range, which is convenient for the subsequent setting and response optimization of the edge detection nodes. Through the shortest distance aggregation method of the hierarchical clustering algorithm, the system terminal first finds the nodes or node groups with the shortest distance and merges them. After each aggregation, recalculate the distance between the newly generated cluster and other clusters, and continue to merge until the set clustering level is reached. For example, a maximum time delay difference threshold (such as 10 ms) can be set, and when the detection time delay difference of the nodes within the cluster exceeds this threshold, stop merging, and finally form several clusters, that is, M local time delay topology regions. These local time delay topology regions all contain nodes within a specific time delay range, which is convenient for configuring corresponding edge detection nodes for each region to achieve hierarchical monitoring and more efficient fault response.

[0030] Furthermore, the present application provides for configuring M first-level edge detection nodes according to the M local time delay topologies, including:

[0031] Interactively obtain multiple sample fault fusion images; construct a fault dimensionality reduction recognition layer based on a convolutional neural network, and optimize the recognition performance of the fault dimensionality reduction recognition layer using the multiple sample fault fusion images; call a time series chart plugin, and obtain multiple image conversion plugins by copying the time series chart plugin; connect the multiple image conversion plugins in parallel to complete the construction of the image conversion layer; after introducing the image fusion layer, complete the construction of the macro fault recognition model by cascading the image conversion layer, the image fusion layer, and the fault dimensionality reduction recognition layer; copy the macro fault recognition model to the M first-level edge detection nodes.

[0032] Optionally, the system terminal obtains fused images of multiple sample faults through interaction. These images contain comprehensive feature information under different fault states, providing rich sample data for subsequent recognition. Subsequently, a fault dimensionality reduction and recognition layer is constructed through a Convolutional Neural Network (CNN) architecture. This layer aims to extract the most representative fault features from the fault fused images for use in subsequent analysis. Specifically, the CNN hierarchical structure includes multiple convolutional layers and pooling layers. The convolutional layers are used to extract low-level features (such as edges and textures) from the input images, while the pooling layers are used to reduce the data size, lower the computational complexity, and achieve dimensionality reduction. The system terminal inputs multiple sample fault fused images into the fault dimensionality reduction and recognition layer. Each fused image contains comprehensive features of different fault states, providing diverse fault data for the model. Through the convolutional kernels in the CNN, the feature patterns in the images are scanned and learned layer by layer, and the size of the feature maps is reduced through the pooling layers. In this process, the most important feature information is gradually compressed and retained, completing the dimensionality reduction of the image data. To improve the recognition performance of the fault dimensionality reduction and recognition layer, the backpropagation algorithm is used to optimize the recognition layer, that is, the recognition results are compared with the actual labels, the loss function is calculated, and the convolutional kernel parameters are adjusted through gradient descent, so that the feature extraction ability of the fault dimensionality reduction and recognition layer for the fault fused images is continuously improved. This optimization process will be carried out in a loop until the recognition accuracy reaches the expected requirement. The fault dimensionality reduction and recognition layer constructed through this process can effectively perform dimensionality reduction and feature extraction on the input fused images, achieving accurate recognition of different fault states. This dimensionality reduction and recognition layer provides a basis for the construction of the subsequent macroscopic fault recognition model, ensuring that the system terminal can efficiently and accurately perform fault detection. To process the time series characteristics of the fault images, the system terminal calls a time series graph plugin, which is used to capture and process the dynamic characteristics of the fault images changing over time. During the fault recognition process, the fault state may change gradually or abruptly over time, and a single image is difficult to reflect the complete fault process. The time series graph plugin can extract the image change patterns at different time points by analyzing the time series characteristics of the images, ensuring sensitivity to changes in the time dimension. To achieve parallel processing, the system terminal copies the time series graph plugin to generate multiple image conversion plugins (such as Pillow, OpenCV) to form a parallel image conversion layer. Each image conversion plugin is responsible for image processing in a specific time period, independently processing the changes in the fault images in a specific time series. For example, one plugin can focus on capturing the features in the initial stage of the fault, while another plugin focuses on the changes during the fault expansion process. This parallel architecture ensures that the model does not miss any key time series information, thereby enhancing the sensitivity to time changes. After constructing the image conversion layer, an image fusion layer (such as Pyramid Fusion, Wavelet Transform Fusion, etc.) is introduced to fuse the image features processed by each plugin.After that, a macroscopic fault recognition model is constructed layer by layer through a cascaded image conversion layer, an image fusion layer, and a fault dimensionality reduction recognition layer. This model has the ability to identify and judge faults from an overall perspective and can quickly respond to the complex fault characteristics of energy storage devices. Finally, the constructed macroscopic fault recognition model is copied to M first-level edge detection nodes to achieve real-time monitoring and rapid response to macroscopic faults in each region. Each first-level edge detection node, with this model, can independently complete the macroscopic recognition of fault characteristics, improving the monitoring accuracy and efficiency.

[0033] The activation permission configuration unit is used to activate constraints for secondary nodes with the fault monitoring architecture when a first-level edge detection node identifies a battery module fault, locate K local edge nodes, and configure hierarchical activation permissions for the K local edge nodes according to the fault propagation analysis tree.

[0034] In one embodiment, in the activation permission configuration unit, when a first-level edge detection node detects a fault in a certain battery module, the system terminal activates the secondary nodes related to the fault based on the fault monitoring architecture to form a gradually expanding activation chain. First, according to the fault monitoring architecture, the K local edge nodes associated with the fault node are quickly located through the configuration relationship between the local edge nodes and the battery module, ensuring that these nodes can centrally monitor the affected area. Subsequently, through the forward node traversal and backward node traversal of the fault propagation analysis tree, the fault propagation relationships and positions of these local nodes are evaluated, and hierarchical activation permissions are set for the K local edge nodes one by one, enabling nodes with a higher response priority to obtain activation permissions prior to the spread of the fault impact, thereby effectively controlling the spread of the fault at different levels and ensuring timely status feedback.

[0035] Furthermore, this application provides that when a first-level edge detection node identifies a battery module fault, it activates constraints for secondary nodes with the fault monitoring architecture, locates K local edge nodes, and configures hierarchical activation permissions for the K local edge nodes according to the fault propagation analysis tree, including:

[0036] According to the configuration relationship between the local edge nodes and the battery module, locate the K associated batteries of the K local edge nodes in the fault propagation analysis tree; construct the forward node traversal order and the backward node traversal order according to the K associated batteries; construct the hierarchical activation permissions according to the parallel relationship between the forward node traversal order and the backward node traversal order.

[0037] Optionally, the system terminal locates K local edge nodes and their K associated battery modules associated with the faulty node in the fault transfer analysis tree according to the configuration relationship between the local edge nodes and the battery modules. Through these associated battery modules, the fault detection is divided into traversal paths in two directions: positive and negative. On the positive path, the system terminal starts from the faulty node and traverses step by step to each associated battery module along the direction of fault transfer to construct the positive node traversal order. On the negative path, it traces back from the associated battery module to the fault source in the reverse direction to form the negative node traversal order. This two-way traversal can not only cover the fault diffusion path but also ensure cross-verification of each node during detection. According to the parallel relationship between the positive node traversal order and the negative node traversal order, the system terminal constructs a hierarchical activation permission, preferentially activates the core nodes on the fault transfer path, and gradually activates the peripheral nodes. Through parallel fault detection in both forward and reverse directions, the system terminal effectively reduces the detection range, realizes efficient cross-verification and parallel detection, thereby accelerating the positioning speed of the faulty battery and improving the fault response and handling efficiency.

[0038] A fault analysis unit for performing in-depth fault analysis on the K battery modules by the K local edge nodes based on the hierarchical activation permission to obtain real-time energy storage faults;

[0039] In one embodiment, in the fault analysis unit, after obtaining the hierarchical activation permission, the K local edge nodes perform in-depth fault analysis on their respective associated battery modules according to different priorities. Each node sequentially collects and analyzes the detailed operation data of the battery module, such as key parameters like current, voltage, and temperature, and identifies potential fault characteristics through the internal fault identification model. Through this hierarchical and step-by-step in-depth analysis, real-time fault information can be quickly collected, the status of the battery module can be comprehensively evaluated, and a real-time fault report of the battery module, that is, a real-time energy storage fault, can be generated to ensure timely and accurate identification of the fault location.

[0040] A load distribution unit for the edge management coordination node to send the real-time energy storage fault to the remote operation and maintenance center after receiving the real-time energy storage fault and perform load distribution of the energy storage device according to the real-time energy storage fault, where the edge management coordination node is the root node of the fault monitoring architecture.

[0041] In one embodiment, in the load distribution unit, when the edge management coordination node receives real-time energy storage fault information, it first transmits the fault data to the remote operation and maintenance center in a timely manner, so that the remote monitoring team can obtain the real-time status of the device and make preparations for fault handling. At the same time, the edge management coordination node will adjust the load distribution of the energy storage device according to the received fault information, and preferentially reduce or isolate the load of the battery modules affected by the fault, so as to ensure the overall stable operation of the energy storage device. As the root node of the fault monitoring architecture, the edge management coordination node is responsible for overall coordination of fault handling and load adjustment to ensure that the entire energy storage device can effectively mitigate the impact of faults and maintain core functions in the fault state.

[0042] Furthermore, the present application also provides that after receiving the real-time energy storage fault, the edge management coordination node sends the real-time energy storage fault to the remote operation and maintenance center and performs load distribution of the energy storage device according to the real-time energy storage fault, wherein the edge management coordination node is the root node of the fault monitoring architecture, and includes:

[0043] After receiving the real-time energy storage fault, the edge management coordination node locates and isolates the faulty battery module according to the real-time energy storage fault; obtains the fault delay parameters of the faulty battery module and the remaining battery modules according to the fault delay topology to perform load weight distribution, and obtains a load distribution strategy; and uses the load distribution strategy to update the load distribution of the energy storage device.

[0044] Optionally, when the edge management coordination node receives real-time energy storage fault information, it identifies the specific faulty battery module according to the location information in the real-time energy storage fault, and immediately takes isolation measures to remove the module from the main load link of the energy storage system to avoid fault spread and further damage. Subsequently, the edge management coordination node uses the fault delay topology to analyze the fault delay parameters between the faulty battery module and other normal battery modules. Through these parameters, the node can determine the degree of delay of each module affected by the fault, and accordingly assign corresponding load weights to each module. For example, modules with lower delay and less impact will obtain higher load weights, while modules with higher delay will be assigned lower load weights. Then, according to the calculated load weights, a load distribution strategy is generated, and the load is preferentially assigned to the battery modules with higher weights. In this process, the node grades the modules according to the fault delay parameters, sets the modules with lower delay as high priority, and assigns more power loads, so that these modules can bear the core load requirements after the faulty module is isolated. Then, the power upper limit of each module is set to prevent overload. For modules with lower weights, the load distribution strategy will assign less power to reduce their operating pressure. This weight-based distribution platform ensures that the energy storage device can continue to operate under unbalanced conditions while minimizing the impact of faults on the overall system. Finally, the system terminal will dynamically update the load of the energy storage device according to this strategy, transferring more load to healthy modules, thereby ensuring the stability and continuous operation ability of the energy storage device while minimizing the impact of faults on the entire energy storage device.

[0045] Furthermore, this application also includes:

[0046] Optimizing the load weights of the faulty battery module and the remaining battery modules according to the distribution of battery modules in the energy storage device to obtain an optimized distribution strategy; presetting a thermal runaway time domain; when the execution of the load distribution update of the energy storage device reaches the thermal runaway time domain, using the optimized distribution strategy to perform dynamic load update of the energy storage device.

[0047] Optionally, based on the physical distribution of battery modules in the energy storage device, the system terminal optimizes the load weights of the faulty battery module and the remaining battery modules around it. The key to this step is to utilize the spatial relationship between the faulty battery module and the remaining battery modules to identify the battery modules within the priority distance in terms of physical distance or heat conduction path, ensuring that the load is preferentially allocated to the adjacent modules around the faulty battery module so that the modules within the fault impact range can effectively share the load and mitigate the impact of the fault on the overall system. Subsequently, based on these spatial distribution relationships, the system terminal generates an optimized load distribution strategy, allocating a higher load weight to the remaining battery modules that are within the priority distance from the faulty battery module and not affected by the fault. This strategy not only takes into account the spatial distribution of the battery modules but also reduces the risk of secondary faults in the battery modules under high load by optimizing the load path. At the same time, a thermal runaway time domain is preset, that is, when the temperature and load state of the energy storage device reach this thermal runaway time domain, it indicates the possibility of thermal runaway. When the load distribution update of the system terminal reaches this thermal runaway time domain, it will automatically switch to the previously generated optimized distribution strategy to perform real-time load dynamic updates. Through this real-time adjustment of the optimization strategy, the system terminal can still reasonably distribute the load spatially under high load and fault conditions, prevent the spread of thermal runaway, and ensure the stable operation of the energy storage device.

[0048] The energy storage device fault monitoring platform under remote identification according to an embodiment of the present invention is used to solve the technical problems of inability to accurately identify the fault transmission path and timely response during the fault monitoring process of the energy storage device, and realizes the construction of an edge monitoring architecture through a fault transmission analysis tree, and performs fault location and analysis according to hierarchical activation permissions, so as to achieve the technical effects of fast and accurate fault monitoring and intelligent load distribution of the energy storage device. The energy storage device fault monitoring platform under remote identification includes: a path analysis unit 1, a node configuration unit 2, an activation permission configuration unit 3, a fault analysis unit 4, and a load distribution unit 5.

[0049] Although the present application makes various references to certain modules in the platform according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or the server. The included individual units and modules are only divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0050] The above specific embodiments do not constitute a limitation to the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. The energy storage equipment fault monitoring platform under remote identification is characterized by: The platform includes: A path analysis unit, used to perform fault transmission path analysis on the energy storage device to obtain a fault transmission analysis tree; A node configuration unit is used to configure edge detection nodes of the energy storage device according to the fault transfer analysis tree to obtain a fault monitoring architecture. The platform includes: Configuring multiple local edge nodes for multiple battery modules in the energy storage device; Decomposing the fault transfer analysis tree into M local delay topologies according to the fault transfer delay; Configure M first-level edge detection nodes according to the M local delay topologies; Dividing the multiple local edge nodes into M groups of local edge nodes according to the M local delay topologies; Pre-constructing an edge management coordination node, and completing the construction of the fault monitoring architecture by connecting the edge management coordination node, M first-level edge detection nodes and M groups of local edge nodes in a hierarchical communication manner; an activation authority configuration unit, for locating K local edge nodes using the fault monitoring architecture as a secondary node activation constraint when the primary edge detection node identifies a battery module fault, and configuring hierarchical activation authorities for the K local edge nodes according to the fault transfer analysis tree; A fault analysis unit, configured for the K local edge nodes to perform in-depth fault analysis of the K battery modules based on the hierarchical activation authority to obtain real-time energy storage faults; A load distribution unit is used for sending the real-time energy storage fault to a remote operation and maintenance center after the edge management coordination node receives the real-time energy storage fault, and distributing the load of the energy storage device according to the real-time energy storage fault, wherein the edge management coordination node is the root node of the fault monitoring architecture.

2. The energy storage device fault monitoring platform under remote identification as claimed in claim 1, characterized in that: Performing fault transmission path analysis on the energy storage device to obtain a fault transmission analysis tree, the platform includes: Perform fault transmission path fitting according to the electrical connection relationship of the energy storage device to generate an initial fault transmission tree; Performing fault diffusion analysis based on protection delay, and updating the initial fault propagation tree according to the analysis result to obtain a fault diffusion propagation tree; A thermal runaway simulation is performed according to the distribution of battery modules in the energy storage device, and nodes of the fault diffusion propagation tree are merged according to the simulation results to obtain the fault propagation analysis tree.

3. The energy storage device fault monitoring platform under remote identification as claimed in claim 1, characterized in that: Decomposing the fault transfer analysis tree into M local delay topologies according to the fault transfer delay, the platform includes: According to the fault transmission type, the fault transmission analysis tree is traversed to calculate the fault transmission delay to obtain the fault delay topology; Interact with the remote operation and maintenance center to obtain a fault detection delay; The fault detection delay is used to perform topology aggregation on the fault delay topology to obtain the M local delay topologies.

4. The energy storage device fault monitoring platform under remote identification as claimed in claim 1, characterized in that: M first-level edge detection nodes are configured according to the M local delay topologies, and the platform includes: Interactively obtain multiple sample fault fusion images; Building a fault dimension reduction recognition layer based on a convolutional neural network, and optimizing the recognition performance of the fault dimension reduction recognition layer by using the multiple sample fault fusion images; Calling a timing chart plug-in, and obtaining a plurality of image conversion plug-ins by copying the timing chart plug-in; Connecting the plurality of image conversion plug-ins in parallel to complete the construction of the image conversion layer; After the image fusion layer is introduced, the construction of the macro fault identification model is completed by cascading the image conversion layer, the image fusion layer and the fault dimension reduction identification layer; The macro fault identification model is copied to the M first-level edge detection nodes.

5. The energy storage device fault monitoring platform under remote identification as claimed in claim 1, characterized in that: When the primary edge detection node identifies a battery module fault, the fault monitoring architecture is used as a secondary node activation constraint, K local edge nodes are located, and hierarchical activation permissions are configured for the K local edge nodes according to the fault transfer analysis tree. The platform includes: According to the configuration relationship between the local edge nodes and the battery modules, locating the K associated batteries of the K local edge nodes in the fault transfer analysis tree; Constructing a positive node traversal order and a negative node traversal order according to the K associated batteries; The hierarchical activation authority is constructed according to the parallel relationship between the positive node traversal order and the negative node traversal order.

6. The energy storage device fault monitoring platform under remote identification as claimed in claim 3, characterized in that: After receiving the real-time energy storage fault, the edge management coordination node sends the real-time energy storage fault to the remote operation and maintenance center, and performs load distribution of the energy storage device according to the real-time energy storage fault, wherein the edge management coordination node is the root node of the fault monitoring architecture, and the platform includes: After receiving the real-time energy storage fault, the edge management coordination node locates and isolates the faulty battery module according to the real-time energy storage fault; Obtaining the fault delay parameters of the faulty battery module and the remaining battery modules according to the fault delay topology to perform load weight distribution and obtain a load distribution strategy; The load distribution strategy is adopted to update the load distribution of the energy storage device.

7. The energy storage device fault monitoring platform under remote identification as claimed in claim 6, characterized in that: The platform includes: Optimizing the load weights of the faulty battery module and the remaining battery modules according to the distribution of the battery modules in the energy storage device to obtain an optimized allocation strategy; Preset thermal runaway time domain; When the load distribution update of the energy storage device reaches the thermal runaway time domain, the optimized distribution strategy is used to dynamically update the load of the energy storage device.

Images