A Visualized Low-Carbon Energy Dispatch System and Method for Industrial Parks Based on the Internet of Things

Abstract

This invention relates to the field of energy management technology, and more particularly to a low-carbon energy dispatching system and method for industrial parks based on Internet of Things (IoT) visualization. The method includes the following steps: acquiring renewable energy equipment data; extracting communication protocol data from the renewable energy equipment data; parsing the equipment connection topology based on the communication protocol data; identifying abnormal energy supply nodes based on the equipment connection topology; detecting energy storage battery charge drift faults based on the abnormal energy supply nodes; determining the electrolyte degradation level based on the energy storage battery charge drift faults; optimizing the electrolyte flow channel structure based on the electrolyte degradation level; and updating the energy storage battery status based on the renewable energy equipment data according to the electrolyte flow channel structure to obtain energy storage battery status data. This invention, based on energy management technology, improves the accuracy and response speed of energy dispatching in industrial parks, reduces carbon emissions, and enhances the operating efficiency of energy storage systems.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, and in particular to a low-carbon energy dispatching system and method for industrial parks based on Internet of Things visualization. Background Technology

[0002] Traditional IoT-based visualized low-carbon energy dispatching in industrial parks typically deploys various types of sensing nodes to collect real-time data on electricity, heat, water, gas, new energy sources (such as photovoltaic and wind power), and energy storage devices. This data is preprocessed and standardized via edge computing gateways, employing low-power wide-area networks (such as LoRa and NB-IoT) or 5G communication technologies to ensure wide-area coverage and high-frequency transmission of energy data, while maintaining data link stability and low latency. However, the periodic data collection method lacks real-time capability and cannot meet the dynamic response requirements of rapidly changing energy scenarios. Furthermore, the lack of unified standards for communication protocols and data interfaces among different types of energy devices leads to integration difficulties when connecting various heterogeneous devices, affecting the comprehensiveness and accuracy of overall data coverage. Dispatching based on manually set rules or experience-based models lacks the ability to learn and adapt based on real-time data, and lacks energy supply and load forecasting models. This makes it impossible to achieve optimal dispatching strategies for low-carbon goals, especially in multi-energy systems such as wind-solar-storage-load coupling, which can easily lead to resource waste or carbon emission peaks. Although it has energy consumption visualization capabilities, it mostly stays at the level of data display and lacks the ability to calculate and analyze key indicators such as carbon emission intensity and marginal carbon benefits, making it difficult to provide quantitative support for low-carbon management. Summary of the Invention

[0003] Therefore, it is necessary for the present invention to provide a low-carbon energy dispatching system and method for industrial parks based on Internet of Things visualization, in order to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a low-carbon energy dispatching method for industrial parks based on Internet of Things (IoT) visualization includes the following steps:

[0005] Step S1: Acquire renewable energy equipment data; extract communication protocol data from the renewable energy equipment data; parse the device connection topology based on the communication protocol data;

[0006] Step S2: Identify abnormal energy supply nodes based on the device connection topology; detect energy storage battery charge drift faults based on abnormal energy supply nodes; determine the electrolyte degradation level based on the energy storage battery charge drift faults; optimize the electrolyte flow channel structure based on the electrolyte degradation level.

[0007] Step S3: Update the energy storage battery status based on the electrolyte flow channel structure of the renewable energy equipment data to obtain energy storage battery status data; construct a low-carbon energy dispatch model based on the energy storage battery status data; simulate charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data.

[0008] Step S4: Predict thermal runaway risk based on charge and discharge behavior data; identify carbon emission impact areas in the park based on thermal runaway risk; conduct low-carbon energy visualization scheduling based on carbon emission impact areas in the park to obtain low-carbon energy scheduling data for the park.

[0009] This invention constructs a connection topology for energy devices by accurately acquiring renewable energy equipment data and analyzing communication protocols, providing fundamental support for subsequent scheduling, fault detection, and optimization. By identifying abnormal energy supply nodes, it can promptly detect charge drift faults in energy storage batteries, and then optimize the battery flow structure based on the degree of electrolyte degradation, improving battery performance and lifespan. The optimized electrolyte flow channel structure can update the energy storage battery status in real time, and based on this, a low-carbon energy scheduling model can be constructed to simulate charging and discharging behavior, thereby accurately acquiring charging and discharging data. Predicting thermal runaway risks and identifying carbon emission impact areas can effectively reduce carbon emissions in the park and ensure the rational allocation of energy resources. Low-carbon energy scheduling in the park can achieve refined management. Through real-time visualized scheduling data, managers can flexibly adjust according to real-time changes, avoiding energy waste and carbon emission peaks. It overcomes the real-time insufficiency and integration difficulties in traditional energy scheduling, realizing intelligent scheduling based on real-time data, promoting the efficient utilization of low-carbon energy and minimizing carbon emissions in the park.

[0010] Preferably, this specification also provides an IoT-based visualization-based low-carbon energy dispatching system for industrial parks, used to execute the IoT-based visualization-based low-carbon energy dispatching method described above. This IoT-based visualization-based low-carbon energy dispatching system includes:

[0011] The connection topology parsing module acquires renewable energy equipment data; extracts communication protocol data from the renewable energy equipment data; and parses the equipment connection topology based on the communication protocol data.

[0012] The electrolyte flow channel structure optimization module identifies abnormal energy supply nodes based on the equipment connection topology; detects battery charge drift faults based on these abnormal energy supply nodes; determines the degree of electrolyte degradation based on the battery charge drift faults; and optimizes the electrolyte flow channel structure based on the degree of electrolyte degradation.

[0013] The low-carbon energy dispatch model construction module updates the energy storage battery status based on the electrolyte flow channel structure of renewable energy equipment data to obtain energy storage battery status data; constructs a low-carbon energy dispatch model based on the energy storage battery status data; and simulates charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data.

[0014] The park's low-carbon energy dispatch module predicts thermal runaway risk based on charging and discharging behavior data; identifies areas affected by carbon emissions in the park based on thermal runaway risk; and performs visualized low-carbon energy dispatch based on these areas to obtain low-carbon energy dispatch data for the park.

[0015] The present invention relates to an IoT-based visualization-based low-carbon energy dispatching system for industrial parks. This system can implement any of the IoT-based visualization-based low-carbon energy dispatching methods of the present invention. It serves as a medium for coordinating the operation and signal transmission between various modules to complete the IoT-based visualization-based low-carbon energy dispatching method for industrial parks. The internal modules of the system cooperate with each other, which improves the accuracy and response speed of energy dispatching in the industrial park, reduces carbon emissions, and improves the operating efficiency of the energy storage system. Attached Figure Description

[0016] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0017] Figure 1 This is a schematic diagram of the steps of a low-carbon energy dispatching method for industrial parks based on Internet of Things visualization according to the present invention.

[0018] Figure 2 This is a detailed flowchart of step S1 in the present invention;

[0019] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] The technical method of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0021] In addition, the attached drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0022] It should be understood that although terms such as "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0023] To achieve the above object, please refer to Figures 1 to 2 , the present invention provides a low-carbon energy scheduling method for a park based on Internet of Things visualization, and the method includes the following steps:

[0024] Step S1: Obtain renewable energy device data; extract communication protocol data of the renewable energy device data; analyze the device connection topology structure according to the communication protocol data;

[0025] In this embodiment, the Internet of Things data acquisition terminals supporting solar photovoltaic panels, wind turbines and other renewable energy devices installed inside the park are used to collect the operation status data of each device through the Modbus-TCP or IEC 61850 communication protocol. The data content includes fields such as device voltage, current, frequency, power generation power, operating temperature, device address, operation timestamp and communication node ID. Subsequently, the protocol data in the communication messages of each device is extracted by using a preset communication protocol parsing template. Based on the field matching and message frame format recognition functions, a connection topology structure diagram between devices is constructed according to information such as slave device address, bus number, transformer interface number, etc. The topology structure data is stored in the graph database in the form of a directed graph. Each device is a node, and the connection relationship between nodes is represented by an edge, and the attribute parameters of the edge such as current direction, link impedance, cable length, etc. are attached to form a complete device physical connection model and data communication link model.

[0026] Step S2: Identify abnormal energy supply nodes based on the device connection topology; detect energy storage battery charge drift faults based on abnormal energy supply nodes; determine the electrolyte degradation level based on the energy storage battery charge drift faults; optimize the electrolyte flow channel structure based on the electrolyte degradation level.

[0027] In this embodiment, based on the device connection topology data, a topology traversal algorithm (such as Depth-First Search (DFS)) combined with a time series comparison method is used to analyze the energy flow change trend between device nodes within a specific sampling period (e.g., every 10 seconds). Nodes exhibiting abnormal supply characteristics such as voltage surges (surge amplitude greater than 10% of rated voltage), reverse energy flow (reversal of power flow direction), and current interruptions (zero current but device not idle). Among the identified abnormal nodes, the state of charge (SOC) change rate collected by the deployed energy storage battery BMS system is further analyzed. If the SOC still experiences an irregular decrease (decrease amplitude greater than 5%) within a unit time (e.g., 10 minutes) without charging or discharging, it is determined to be a charge drift fault. For detected charge drift units in energy storage batteries, sensors are used to collect real-time data on electrolyte conductivity, voltage deviation (single-cell voltage variance exceeding 0.02V in the battery pack), and internal resistance (measured via AC impedance). A combined calculation of these three parameters is then performed, using the formula: Deterioration coefficient = (Percentage decrease in conductivity + Percentage increase in internal resistance + Rate of change in voltage deviation) / 3. If the degradation coefficient exceeds 30%, the electrolyte is considered to have significantly deteriorated. Based on this, and referencing data on internal cooling pipe diameter, circulating pump flow rate (L / min), coolant temperature difference, and local bubble detection, CFD (Computational Fluid Dynamics) simulation tools are used to evaluate different flow structures. The flow structure with the shortest total flow path length, lowest pressure loss, and least heat accumulation is selected as the optimal solution and updated in the electrolyte control unit's control strategy.

[0028] Step S3: Update the energy storage battery status based on the electrolyte flow channel structure of the renewable energy equipment data to obtain energy storage battery status data; construct a low-carbon energy dispatch model based on the energy storage battery status data; simulate charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data.

[0029] In this embodiment, the thermal conductivity, cooling efficiency parameters, and electrochemical stability coefficient of the energy storage battery are updated based on the electrolyte flow channel structure. The operating data of the renewable energy equipment is also updated synchronously every 10 seconds. This data is uniformly aggregated through a data receiving gateway and fused with the energy storage battery data. At the data fusion layer, using time window alignment and data synchronization rules, data fields such as power generation, peak current, SOC status, electrolyte temperature, and resistance change rate are input sequentially to generate an energy storage battery status data table. The status data is encapsulated in a JSON structure and pushed to the energy consumption platform data warehouse via the MQTT protocol. Subsequently, a low-carbon energy scheduling model is constructed based on this status data. The construction parameters include unit power generation carbon emission intensity (gCO2 / kWh), battery cycle life index (the ratio of actual cycle count to design life), and current load demand (data collected by a real-time energy consumption collector). During the simulation of charging and discharging behavior, the following charging and discharging scheduling control rules are used: priority charging periods are those with irradiance > 600 W / m² and electrical load below average; priority discharging periods are those with load above 80% of rated power and insufficient photovoltaic output. The energy and duration of each scheduling event are allocated based on the battery's maximum charging power, remaining capacity, and current load power. The start time, duration, energy (kWh), voltage fluctuation range, and load matching degree of each charging and discharging event during the simulation are stored as charging and discharging behavior data in a log table.

[0030] Step S4: Predict thermal runaway risk based on charge and discharge behavior data; identify carbon emission impact areas in the park based on thermal runaway risk; conduct low-carbon energy visualization scheduling based on carbon emission impact areas in the park to obtain low-carbon energy scheduling data for the park.

[0031] In this embodiment, based on charge and discharge behavior data, parameters such as battery temperature change curves, current surge amplitude, cycle frequency, and individual cell voltage fluctuation range are extracted. A risk warning method based on thermal threshold detection is adopted, where the key thresholds for predicting thermal runaway risk are set as follows: a risk prediction flag is triggered if any one of the following conditions is met: battery temperature rise rate exceeds 3℃ / min, temperature reaches 60℃, individual cell voltage exceeds 4.2V or falls below 2.5V, or current surge amplitude exceeds 1.5 times the rated current. For battery cells that have triggered risk flags, their geographical installation location and connection relationship with deployed equipment are combined to spatially mark their location on the park map using a 3D modeling system (such as Revit or OpenStreetMap GIS), thereby identifying the impact range of abnormal carbon emission sources. When identifying the impact area, the carbon emission data per unit time in the past 30 minutes of that node (calculated by power generation carbon factor × discharge capacity) is superimposed and evaluated with the total emission ratio of that area to determine the carbon emission impact area. Finally, based on the location of the affected area and the status of available power, the low-carbon energy dispatch strategy table is retrieved, and the photovoltaic load transfer and low-carbon load shifting strategies are prioritized. The dispatch instructions are then sent to the relevant subsystems (such as photovoltaic inverters, load switching controllers, and air conditioning energy-saving systems). At the same time, the updated dispatch paths and carbon emission statistics are displayed in real time on the visualization platform, generating low-carbon energy dispatch data record entries for the park that include fields such as dispatch instruction number, dispatch time, dispatch energy amount, and dispatch target equipment.

[0032] Preferably, step S1 specifically includes:

[0033] Step S11: Obtain renewable energy device data;

[0034] In this embodiment, multiple edge acquisition devices are deployed within the park to periodically sample the operating parameters of distributed renewable energy equipment (including solar photovoltaic modules, wind power generation equipment, and biomass power generation units). Each acquisition device is equipped with a communication module supporting RS485, CAN bus, and Ethernet interfaces, and uses a sampling frequency of 5 seconds to acquire the following parameters: device operating voltage (unit: V, range 0-1000V), device output current (unit: A, range 0-100A), output power (unit: kW), device status code, abnormal alarm code, device number (number format is hexadecimal ID, length is 8 digits), and channel occupancy rate (unit: %). After data acquisition, the data is sent to the access layer server of the park's energy management platform via MQTT protocol packets. The packet structure includes device ID, data content, timestamp, and transmission path tag fields. All acquired data enters the Kafka streaming data queue in the access layer server and is then forwarded to the real-time data processing module for parsing.

[0035] Step S12: Extract communication protocol data from renewable energy equipment;

[0036] In this embodiment, communication protocol data is parsed based on the field format of the acquired data packets. The protocols involved include Modbus RTU, Modbus TCP, IEC 61850, and DL / T 645-2007, each with differences in data structure. A protocol identifier module is used to identify the protocol based on the frame header field, function code (e.g., 03 in Modbus indicates reading the holding register), and CRC checksum. The identifier reads the start character (e.g., 0x68 or 0xAA), command code, address code, and data field length field of each data frame to extract the complete data payload. Under the Modbus protocol, the holding register values ​​are mapped to specific operating parameters; under the IEC 61850 protocol, communication data is extracted based on the three-layer structure of LN (Logical Node) + DO (Data Object) + DA (Data Attribute), and the output JSON structure contains communication protocol parameter data such as communication port, physical device address, protocol frame number, transmission length, and device type label, providing the original communication content for subsequent device identifier extraction and path reconstruction.

[0037] Step S13: Identify the communication identifier of the energy storage device according to the communication protocol data;

[0038] In this embodiment, data frames related to the energy storage device are filtered out based on communication protocol data. Each data frame contains a unique communication identifier, extracted from the protocol address field. For example, in the Modbus protocol, this is the address field (generally a decimal integer from 1 to 247); in the IEC 61850 protocol, it is identified by the device logical node name (e.g., "BAT1.QF1") and the device instance ID. To ensure uniqueness, the address field of each data frame is compared and verified with the corresponding device number recorded in the asset management table. The extracted communication identifier must meet the following constraints: length not less than 8 bits, hexadecimal encoding, with the first bit representing the type identifier (e.g., energy storage devices uniformly begin with "B"), and the remaining bits representing the manufacturer number and device serial number. For energy storage devices with multiple communication ports, the ports are sorted from highest to lowest priority based on the port ID field in the communication frame, retaining only the identifier of the primary communication port for path reconstruction. The final output is a list of five fields containing the device ID, communication identifier, port ID, data packet length, and data reception time.

[0039] Step S14: Trace the communication relationship based on the communication identifier of the energy storage device, and reconstruct the communication path of the energy storage device according to the communication relationship;

[0040] In this embodiment, within the park's energy dispatch network architecture, all energy storage devices are embedded with a unique Device Communication Identifier (DCI). This identifier is a uniform 16-digit hexadecimal string, with the first 8 digits representing the device type and manufacturer number, and the last 8 digits representing the device's unique serial number. Energy storage devices interact via LoRa or RS-485 communication protocols and periodically (every 5 minutes) upload communication messages to the central dispatch server through the local gateway. All reported energy storage device communication data packets from the past 24 hours are extracted from the communication log database. The filtering fields include the sending DCI, receiving DCI, timestamp, message type, RSSI (Receiving Signal Strength Indicator), and response delay (ms). A two-dimensional communication record matrix is ​​then established, sorted by time, and each "sending DCI - receiving DCI" pair is constructed as a communication edge, forming an initial set of communication edges. A directed graph structure is constructed using the NetworkX library in Python, with DCI as nodes and communication edges as edge relationships. By analyzing the in-degree and out-degree of the nodes in the graph, the master-slave communication structure can be identified. During the communication relationship tracing process, valid communication judgment conditions are set: if any group of DCIs has at least 3 bidirectional communications within 1 hour, and the average response latency is less than 150ms and the RSSI is not less than -80dBm, then a stable communication relationship is determined to exist between the DCIs. All node pairs that meet the stable communication relationship are extracted as valid communication pairs, and a stable communication map is constructed. To reconstruct the communication path, the shortest path algorithm (Dijkstra's algorithm) is executed on the stable communication map. Starting from the park's master control node DCI, the algorithm traces layer by layer to all energy storage device nodes, generating a complete set of energy storage device communication paths. The communication path is stored in the system in the form of a path linked list. Each path records the following fields: starting DCI, path level, number of intermediate nodes in the path, RSSI value of each communication segment, communication latency, and number of communication hops, which serve as important basis for subsequent communication diagnosis and energy dispatch path selection. Finally, the path structure is mapped to the park's 3D GIS platform, and visualized through the association of device coordinate fields, allowing maintenance personnel to manage it intuitively on the dispatch platform.

[0041] Step S15: Construct the device connection topology based on the energy storage device communication path and the energy storage device communication identifier.

[0042] In this embodiment, each node in the communication path of the energy storage device is spatially mapped according to its physical location number (based on the device geocoding in the park's GIS system). Then, all devices within the same physical area in the communication path are aggregated into a local area network subnet unit, generating a local topology subgraph. Subsequently, the overall device connection topology is constructed based on the communication paths between subnets. The topology consists of graph nodes and graph edges. Nodes represent individual energy storage devices or relay devices, and edges represent communication links. Edge attributes include link type (fiber optic, wireless, CAN bus), communication frequency (in Hz), link delay (in ms), and link load (in %). During the topology graph generation process, each device communication identifier is uniquely mapped using a hash mapping, and each hop in the path data is converted into edge information of the graph. The final device connection topology is output in GraphML format, including node ID, node type, node coordinates, edge source node, edge target node, and edge attributes, for subsequent loading and use by the scheduling and analysis module.

[0043] Preferably, identifying abnormal energy supply nodes in step S2 includes:

[0044] Collect device energy supply data based on the device connection topology;

[0045] In this embodiment, based on the established device connection topology, the energy supply paths between various energy storage devices, distributed generation units, and load devices are determined. According to the node IDs and edge attribute information in the topology, the start and end points of each energy transmission link are identified, clarifying the source-sink relationship. Subsequently, three-phase smart meters and Hall current sensors with current, voltage, and active power acquisition functions are deployed at each energy storage device and energy transmission node (including relay distribution cabinets, main feeders, and load interfaces). All sensing devices must have Level 1 accuracy under the IEC 62053-22 standard, with a fixed sampling period of 1 second. The collected energy supply data includes: three-phase voltage (unit: V), three-phase current (unit: A), total active power (unit: kW), power factor (unit: dimensionless), cumulative energy (unit: kWh), and frequency (unit: Hz). The collected raw data is encapsulated into a unified structured data format through an edge computing gateway, aggregated and organized according to path number using timestamps as indexes, and stored in the HBase distributed database to provide continuous input for energy flow tracing.

[0046] Track energy flow based on equipment energy supply data;

[0047] In this embodiment, the device connection topology constructed in the previous step is numbered and structured to ensure that the input and output paths of each device node are clear, forming a graph structure with directional edges. Then, energy supply data for all device nodes is continuously collected at a fixed frequency of once every 5 seconds. This data is statistically based on active power, in kilowatts, recording the input power (supplied by its upstream connected nodes) and output power (provided to its downstream connected nodes) of each node. All collected data is transmitted directly from the energy monitoring instrument to the Kafka message queue in the central server via the OPC UA interface and processed in real time by the Flink stream computing framework. During path tracing, the system calculates the difference between the input and output power for each node to obtain its current net energy supply. Subsequently, this difference is compared and analyzed historically using a sliding time window. The width of the time window is set to 60 seconds, meaning that the data for each node within 60 seconds forms a sliding time series to observe the fluctuation trend of its net supply value. If any node on a path experiences a sudden change in power difference within a given time window (defined as a change exceeding 15%), the entire path is marked as an "energy flow mutation path." During processing, the energy flow state of each node is evaluated in real-time using Flink's window functions for streaming computation (such as `SlidingEventTimeWindows`), and node pairs meeting the mutation criteria are recorded. All paths are ultimately converted into a Directed Acyclic Graph (DAG), where each edge represents an energy transfer path between two nodes, and the edge weight is the energy transfer rate per unit time, measured in kilowatts per second (kW / s). This DAG structure is stored and displayed in a Spark environment using the GraphX ​​framework. Subsequent steps will use this graph structure to identify abnormal paths, energy flow mutation nodes, and related abnormal energy behaviors.

[0048] Identify abnormal energy flows;

[0049] In this embodiment, the anomaly determination rule is based on two core features: one is the energy jump threshold, defined as a power difference of more than 3kW between the same node within two consecutive sampling periods; the other is path continuity anomaly, i.e., in a certain path, there is a situation where the upstream node is supplying power normally while the downstream node has zero energy input, and the path is determined to be an interrupted abnormal path. An abnormal energy flow direction must meet one of the above two features simultaneously to be recorded. The detection process is analyzed through the Spark data batch processing platform. The processing logic is based on window aggregation and conditional filtering operations to batch identify all energy path data and output a list of abnormal energy flow paths, including path number, start time, end time, anomaly type (jump / interruption), abnormal energy value, and corresponding topology path node number.

[0050] Detection of copper busbar metal oxidation characteristics based on abnormal energy flow direction;

[0051] In this embodiment, multiple types of electrical monitoring sensor modules are deployed for key physical equipment nodes involved in power flow abrupt changes to perform detailed monitoring of the status of the copper busbar connection area. The deployed monitoring modules mainly include three types of sensing devices: partial discharge sensors, infrared thermal imaging modules, and surface resistance sensors. The partial discharge sensors operate in the ultra-high frequency (UHF) band, supporting continuous monitoring of a frequency range of 300MHz to 1500MHz; the infrared thermal imaging module has a temperature measurement accuracy of ±0.5℃ and is used to record the temperature distribution on the surface of the copper busbar and its surrounding environment; the surface resistance sensor measures the resistance change of the metal conductor surface, with a measurement range covering 0 to 100 megohms. All three types of sensors are deployed on the inner wall of the copper busbar node's housing and near the cable joints, connected to the edge data acquisition terminal to achieve continuous monitoring of the copper busbar's operating status. Whenever a power flow abrupt change is detected in a certain equipment path in the previous step, a detection command is sent to the corresponding edge acquisition terminal targeting all physical equipment within that path. The detection time window is fixed at 30 minutes before and after the occurrence of the abnormal path. During this time period, the resistance value, thermal characteristics, and partial discharge signal spectrum data of the copper busbar surface are sampled at high frequency, with a sampling period uniformly set to 2 seconds. Metal oxidation characteristics are identified through a comprehensive judgment using three criteria. First, in terms of resistance detection, if the measured resistance value of the copper busbar surface is greater than or equal to 1 megohm, it is considered that the surface conductivity has decreased. Second, in terms of thermal imaging, if the detected area shows a local temperature anomaly point in the thermal image, the temperature value of this anomaly point must be more than 5 degrees Celsius higher than the surrounding area, and this state must persist for more than 15 seconds in the image frame sequence. Finally, in terms of the partial discharge spectrum, if the main peak frequency is continuously above 600MHz in the signal spectrum collected by the sensor, it is considered that there is a continuous discharge phenomenon. All three criteria must be met simultaneously to confirm that the copper busbar at that node has metal oxidation characteristics. All collected sensor data carries a unique device node identifier code and is uploaded to the central analysis platform in real time for data synchronization and identification processing through the binding mechanism between the sensor module and the device. After completing the determination, the platform will assign a unique number to the copper busbar nodes identified as having oxidation characteristics and output them to a list for use in subsequent steps.

[0052] Identify ablation marks based on the metal oxidation characteristics of copper busbars;

[0053] In this embodiment, infrared thermal imaging data is used for further image analysis. The copper busbar region is extracted from the infrared image, and a thermal differential mapping algorithm is used to calculate the temperature difference between images from adjacent time periods to identify regions with abrupt changes in temperature rise rate. If a region in the thermal differential mapping is found to have a temperature rise rate per unit area exceeding 2℃ / s and lasting for more than 5 seconds, then that region is identified as an ablation region. Simultaneously, an ultraviolet corona camera is used to acquire images of this region to analyze for the presence of arc discharge traces. The arc feature is judged with an image brightness exceeding 400 lux as the threshold. If a strong bright spot appears in the image, and the consistency of the overlapping area combined with the aforementioned thermal imaging identification results exceeds 80%, then it is finally identified as an ablation trace. All ablation identification results are output in the form of rectangular coordinate labels. Each data entry includes the device ID, thermal imaging coordinates, area (unit: cm²), and time stamp, serving as input for copper busbar anomaly localization.

[0054] Loose areas of copper busbars can be located based on ablation marks;

[0055] In this embodiment, based on the identified coordinates of the ablation marks, and combined with the installation site drawings and equipment structure diagrams, a matching analysis of the copper busbar connection parts is performed. If the overlap area between the ablation area and the connection node (such as bolt joint, lap area) is greater than 60%, it is initially determined to be a loose area. Further verification is performed using data collected by vibration sensors. If the three-dimensional vibration acceleration in this area is greater than 0.5g (g is the acceleration due to gravity) during equipment operation, and a significant frequency modulation phenomenon occurs (a frequency drift of 1~3Hz appears in the spectrum), then it is finally identified as a loose copper busbar area. The above sensor data is collected by a high-frequency vibration sensor (sampling frequency not less than 1kHz), and after joint processing with image processing data, a three-dimensional point cloud marker of the loose area is established in the spatial model.

[0056] By mapping the loose copper busbar area to the equipment connection topology, the abnormal energy supply node is obtained.

[0057] In this embodiment, the physical spatial coordinates of each node in the device connection topology are matched using the three-dimensional coordinates of the loose copper busbar area. Each device node in the topology contains a physical location field (represented by GIS coordinates or local device coordinates). If the Euclidean distance between the coordinates of the loose area and the corresponding physical coordinates of a device node is less than 0.5 meters, the loose area is considered to belong to that device node. After matching, the node is marked as an energy supply anomaly node. Subsequently, the anomaly node ID is added to the anomaly node list, and combined with the aforementioned energy supply anomaly path results, all energy supply paths are traced back to ultimately form a complete energy anomaly path map. The output includes the anomaly node number, corresponding coordinates, affected path number, anomaly start time, and duration fields. All data is encapsulated in a JSON structure for display and alarm calls by the park's energy dispatch platform.

[0058] Preferably, the detection of charge drift fault in the energy storage battery in step S2 includes:

[0059] Monitoring battery charging and discharging status based on abnormal energy supply nodes;

[0060] In this embodiment, the data acquisition module is invoked through the scheduling platform control logic to deploy a charge / discharge monitoring process for the energy storage battery system corresponding to the node. Each energy storage unit is equipped with a voltage acquisition module (accuracy ±0.1V) and a current acquisition module (range ±200A, accuracy ±0.5A), with a sampling period uniformly set to 1 second, continuously recording for 24 hours. The scheduling system binds to the device ID and synchronously uploads the collected voltage and current raw data to the analysis server. The current battery state is determined by combining the voltage change rate (ΔU / t) and the current direction: if the current direction is negative (i.e., entering the battery) and the voltage rises by more than 5% within 10 minutes, it is identified as a charging state; if the current direction is positive and the voltage drops by 5%, it is identified as a discharging state. The state identification does not rely on an estimation model, but directly determines it through the mathematical relationship of the actual collected data, and synchronously records it into the database with a timestamp.

[0061] Identify lithium-ion migration paths based on battery charge / discharge status;

[0062] In this embodiment, after identifying whether the battery is currently charging or discharging, the spatial distribution of its internal electrochemical reaction regions is analyzed to deduce the migration path of lithium ions between the positive and negative electrodes. This step relies on a thermal imager (accuracy ±0.3℃) installed outside the battery pack to collect thermal data from the cells every 5 seconds. By extracting continuously heating areas from the thermal images, lithium ion aggregation or active migration paths are identified. The temperature rise threshold is set to a continuous area that is more than 5℃ higher than the surrounding average temperature. Combined with the cell voltage change gradient (calculated at 5-second sampling intervals), the time difference between temperature and voltage changes is analyzed simultaneously. If the time delay is less than 15 seconds, the heating area can be considered an effective migration path. In addition, a two-dimensional annotation is performed using a temperature difference distribution map to mark the temperature rise path range (represented by pixel coordinates). Finally, the thermal image coordinate chain of this area is output for subsequent simulation input.

[0063] Obtain graphite interlayer structure data; perform graphite thermal expansion simulation based on lithium ion migration path to obtain graphite thermal expansion data; perform graphite contraction simulation based on lithium ion migration path to obtain graphite contraction data.

[0064] In this embodiment, a scheduling platform connects to the energy storage battery manufacturing database to read the anode material design parameters of abnormal batteries, particularly the graphite layer structure. The battery ID is automatically matched to its manufacturing batch, and material physical data, including the number of graphite layers (typically 60–120 layers), single-layer thickness (μm level), and interlayer spacing (average 0.34 nm), are retrieved from the database. This data originates from CT scan images and original material quality inspection reports during manufacturing. The reading operation is completed by the scheduling system, which automatically numbers each structural parameter and writes it into the simulation input module without on-site measurement. All structural parameters are not allowed to be modified to ensure consistency with the actual structure, providing accurate material basis data for subsequent thermal expansion and contraction simulations. The thermally induced deformation module in the finite element simulation software is called, and the input parameters are set as follows: graphite interlayer spacing (unit: nm), number of layers, and coefficient of thermal expansion (standard value 3.3 × 10⁻⁻⁻⁻⁴). 6 The simulation parameters include the temperature rise rate (set to 5℃ / min) and the total temperature rise amplitude (set to 15℃). During the simulation, the heat source location is set within the migration path coordinate range, the simulation time window is 180 seconds, and the unit output step size is 1 second. The output results include the thickness change of each graphite layer (unit: μm) and the boundary expansion deformation (unit: μstrain). The simulation data is output simultaneously as graphs and tables, and uniformly numbered for structural health diagnosis. Similar to the thermal expansion simulation, the simulation parameters for the cooling process are set to analyze the graphite contraction state after the discharge ends. The cooling rate is set to -5℃ / min, the simulation duration is consistent with the thermal expansion, and local cooling simulation is also performed based on the thermal migration path region. The graphite thermal contraction coefficient is 2.8 × 10⁻⁻⁻⁶. 6 / K, the simulation output includes the shrinkage amount for each layer thickness and the overall shrinkage gradient. All parameters in the simulation are completely independent of the thermal expansion stage, and the output is a separate set of shrinkage data files. The two sets of data are aggregated into a bidirectional deformation data package within the simulation platform, and numbered and recorded for subsequent crack detection.

[0065] Electrode plate crack detection was performed based on graphite thermal expansion and shrinkage data to obtain electrode plate crack data.

[0066] In this embodiment, an ultrasonic phased array detector (operating frequency 20MHz, detection depth 0–10mm) is used to perform acoustic scanning in the graphite deformation region to identify microcracks caused by stress deformation. The scanning angle range is set to 0–60 degrees, and each layer of the substrate is scanned once. If a signal interruption or abnormal echo occurs in the acoustic echo image and coincides with the location of the simulation area, it is recorded as a crack initiation point. The crack length is calculated using the difference in wave velocity and time, and the crack depth is calculated using the change in acoustic intensity. If the crack depth exceeds 0.5mm or the length exceeds 2mm, the plate is marked as a structurally damaged plate and included in the risk analysis module.

[0067] Predicting electrode plate fracture risk based on electrode plate crack data;

[0068] In this embodiment, after identifying an electrode plate with cracks, the crack growth trend is recorded, and fracture prediction is performed based on the historical crack propagation rate. The standard value is set to a daily growth rate of no more than 0.2 mm. If the actual measured propagation rate exceeds this value, the plate is determined to be at high risk of fracture. Then, the critical fracture length of the electrode plate at the factory (set at 3.5 mm) is compared. If the current crack length exceeds 2.8 mm and the growth trend is stable, it is directly marked as an electrode plate fracture warning plate, and its number is synchronized to the energy storage battery health database.

[0069] Electrode metal corrosion was detected based on the lithium-ion migration path to obtain electrode metal corrosion data.

[0070] In this embodiment, after removing the outer shell of the thermal migration region, a microelectrode array is mounted on the electrode surface to monitor the metal corrosion potential and current density in the lithium-ion active region. The monitoring period is 48 hours, and the sampling period is 1 minute. A corrosion process is considered to have started when the current density increase rate exceeds 10 μA / cm²·h, and the metal corrosion activation period is defined as a decrease in total potential exceeding 50 mV within 24 hours. All data is collected in real-time via an electrochemical workstation, automatically packaged and uploaded after connecting to the platform. The detection area and corrosion depth are calculated by integrating the corrosion current, with an error not exceeding 3%.

[0071] Predicting electrode detachment risk based on electrode metal corrosion data;

[0072] In this embodiment, after analyzing the corrosion area ratio and depth of the electrode plate, the remaining integrity is calculated based on the metal thickness (e.g., 5 mm). If the area covered by corrosion with a depth greater than 1 mm exceeds 20%, the remaining thickness is less than 80%. If, after further calculation, the remaining integrity is less than 70%, the electrode plate is recorded as a plate with a risk of detachment and added to the detachment risk list by number. All corrosion areas are marked and recorded by coordinates and linked to the risk level.

[0073] The actual capacity of the energy storage battery is determined based on the risk of plate breakage and plate detachment.

[0074] In this embodiment, all electrode plate numbers at risk of breakage or detachment are counted, and their corresponding capacities are eliminated. Assuming a total battery capacity of 60Ah and each electrode plate capacity of 2Ah, if three electrode plates are marked as unusable, the remaining capacity is 54Ah. The platform automatically calculates the current usable capacity by reading the actual cell structure configuration table and combining it with the eliminated values, recording this as the actual battery capacity.

[0075] Extract the displayed battery power based on the battery's charge and discharge status;

[0076] In this embodiment, the current battery percentage and total capacity parameters displayed on the Battery Management System (BMS) interface are read. The data acquisition period is set to once per minute, and each acquired value is recorded in the database. The displayed battery level is read by directly extracting the SOC field (with a fixed register address) from the Modbus protocol data packet, converting it into a battery value, and then synchronously comparing it with the aforementioned actual battery level.

[0077] The charge drift fault of the energy storage battery can be determined based on the actual charge level and the displayed charge level of the energy storage battery.

[0078] In this embodiment, the difference between the actual power consumption and the displayed power consumption is calculated. If the absolute value of the difference exceeds 8% of the total capacity (e.g., 4.8Ah for 60Ah), it is recorded as a charge drift fault. The difference calculation formula is uniformly deployed within the analysis platform and is not subject to manual judgment. If three consecutive sampling results exceed this error range, the fault is identified as "confirmed," and the energy storage unit is marked as requiring calibration. The data is recorded and output to the equipment health status list.

[0079] Preferably, determining the degree of electrolyte degradation in step S2 includes:

[0080] Based on the charge drift fault of the energy storage battery, faulty energy storage batteries are screened, and the electrolyte infrared spectroscopy of the faulty energy storage batteries is detected to obtain the electrolyte infrared spectral data.

[0081] In this embodiment, the IoT scheduling platform connects to the battery management system (BMS) and diagnostic data interface to read the difference between the displayed state of charge (SOC) value and the actual usable capacity value obtained in step ten from the database. If the difference exceeds a set fixed threshold of 8% and remains at this deviation for three consecutive independent detection cycles (10 minutes apart per week), the system automatically marks the corresponding energy storage battery number as a "charge drift fault unit". Subsequently, the detection process module is invoked to schedule the spectral detection equipment to extract electrolyte from these marked units. A 2ml sample of electrolyte is extracted through the drain valve on the top of the battery cell in an oxygen-free nitrogen environment. After sampling, the sample is immediately transferred to a liquid infrared spectral detection device (e.g., a PerkinElmer SpectrumTwo FT-IR) for infrared absorption spectral scanning. The scanning wavenumber range is set to 400cm⁻¹ to 4000cm⁻¹, the resolution is 4cm⁻¹, and the number of scans is set to 32 to improve the signal-to-noise ratio. The spectral data is transmitted in real time to the local processing platform via Ethernet and recorded in the electrolyte spectral data table using the energy storage battery number as an identifier. The entire process enables automatic equipment activation, parameter solidification, and automatic data archiving, ensuring that samples are not exposed to air and avoiding water vapor contamination.

[0082] Identifying moisture characteristic peaks based on electrolyte infrared spectral data;

[0083] In this embodiment, the acquired electrolyte infrared spectrum data is stored in CSV format, which is read by the data processing platform and analyzed using wavenumber-absorbance pairs. Based on the known characteristic absorption peaks of water in the infrared region, the OH stretching vibration peak at a wavenumber of 3420 cm⁻¹ and the HOH bending vibration peak at a wavenumber of 1640 cm⁻¹ are directly located. These two absorption peaks are characteristic bands of water, and their absorbance changes significantly with varying water content. During the identification process, a sliding window averaging algorithm is used to smooth the absorption curve (window size is 5 points) to remove high-frequency noise. Two fixed wavenumber windows are programmed, namely 3410–3430 cm⁻¹ and 1630–1650 cm⁻¹. The maximum absorbance value is searched within each window and marked as the intensity point of the water characteristic peak. All absorbance values ​​are not normalized to maintain the original spectral intensity and preserve absolute quantity information. After identification, the wavenumber positions of the two characteristic peaks and their corresponding maximum absorbance values ​​are written into the characteristic peak data table as the basic input for subsequent area calculation.

[0084] Calculate the area of ​​characteristic peaks based on moisture characteristics; estimate moisture content based on the area of ​​characteristic peaks.

[0085] In this embodiment, absorbance data points within the wavenumber ranges of 3410–3430 cm⁻¹ and 1630–1650 cm⁻¹ are extracted, and the area of ​​each characteristic peak is calculated using a numerical integration method. The trapezoidal rule is used for numerical integration, and the integration formula is as follows: ,in Absorbance The wavenumber is used, and the integration unit is absorption·cm⁻¹. The area of ​​each characteristic peak is calculated separately and then summed to obtain the total characteristic peak area (unit: absorption·cm⁻¹). In the electrolyte batch quality control data, it is known that the characteristic peak area corresponding to the dry anhydrous sample does not exceed 0.05 absorption·cm⁻¹, and the water content is 0%. In the calibration curve of laboratory samples, the total peak areas corresponding to water contents of 0.1%, 0.5%, 1%, 2%, and 5% are 0.12, 0.27, 0.51, 1.02, and 2.25 absorption·cm⁻¹, respectively. Using this standard curve, a linear fitting method is used to convert the characteristic peak area of ​​the current sample to water content. For example, if the measured total peak area is 0.8 absorption·cm⁻¹, the water content is calculated to be approximately 1.6% through curve fitting. All conversion formulas and calibration coefficients are stored in the platform model library. After being called, the water percentage value is directly returned and recorded in the corresponding item of the faulty battery.

[0086] Gas release data was obtained by simulating the gas release of a faulty energy storage battery based on its moisture content.

[0087] In this embodiment, after the moisture content is calculated, this value is used as an input parameter to the gas release simulation module. This module establishes a qualitative-quantitative relationship based on isothermal decomposition experimental data. In known typical lithium-ion electrolytes (e.g., EC:DMC = 1:1 system), when the moisture content exceeds 0.5%, a side reaction will occur at 35℃–55℃, mainly generating CO2, CH4, and trace amounts of H2 gas. In the simulation, the ambient temperature is set at 40℃, and the reaction duration is 60 minutes. Based on the input moisture content, the gas generation rate initiated by unit moisture is obtained from a table (e.g., 0.05 mol CO2 is released per 1% moisture in 60 minutes). Combining the sample volume of 2 ml and the mass density, the total mass is estimated, and finally converted into the amount of each gas released (in μmol), and the amount of CO2, CH4, H2, etc., generated is output in tabular form. All conversion data comes from a standard experimental parameter library and requires no model training. The output data is archived in the platform as a gas release data table and associated with the corresponding battery number for subsequent analysis.

[0088] Of particular importance, the gas release simulation includes the following steps:

[0089] Gas generation reaction simulation was performed on a faulty energy storage battery based on moisture content to obtain gas generation data.

[0090] In this embodiment, the internal moisture content of faulty energy storage battery cells within the target park's energy storage system is monitored in real time using a miniature moisture sensor integrated inside the battery or on the inner wall of the battery casing. The selected sensor is a thin-film polymer capacitive humidity sensor with a resolution of 0.01% RH and an accuracy of ±1% RH. Battery fault types include thermal runaway, internal short circuit, and polarization, which are identified by voltage drops (<2.0V), rapid temperature increases (>80°C), and abnormal increases in internal resistance (exceeding 5mΩ), respectively. For electrolyte environments with a water content >0.5%, standard reaction pathways from an electrochemical reaction database are used to simulate gas generation behavior within the battery, such as LiPF6 + H2O → POF3 + HF↑, and solvent reduction reactions to generate CO or CH4. During the simulation, the electrochemical reaction calculation formula ΔG=ΔH-TΔS is applied, and the generation rate and gas volume of different reaction paths are calculated by combining the standard potential value and the water content ratio. Finally, the types of gases generated (such as H2, O2, HF, CO2) and their generation amounts are recorded in a data table and stored as a structured gas generation data file.

[0091] Calculate the gas generation rate based on the gas generation data;

[0092] In this embodiment, assuming that a single energy storage battery generates 0.75 mmol of H2 within 10 seconds in a fault-activated state, the gas generation rate is 0.075 mmol / s. To ensure timing accuracy, high-speed data sampling with a 5 ms cycle is used to capture the reaction time. A high-speed data acquisition card performs continuous measurements of voltage, current, and temperature, which are then cross-validated using real-time concentration measurements from gas sensors. Fourier transform analysis is used to analyze the periodic fluctuations of the generation curve, eliminating short-term interference signals and retaining only the continuous generation signal segment for rate fitting. For battery cells containing multiple reaction paths, the rates for each type of gas are calculated and recorded according to the dominant reaction path, ultimately generating a gas generation rate list.

[0093] Determine the types of gases generated based on gas generation data;

[0094] In this embodiment, the mass concentration of components in the collected gas generation data, combined with the chemical material composition of the energy storage battery (e.g., NMC cathode, graphite anode, and electrolyte composition of LiPF6+EC / DEC), can determine the typical reaction product types. Gas component detection uses NDIR (non-dispersive infrared spectroscopy) for CO2 detection, mass spectrometry for light organic gases (e.g., CH4, C2H6), and an electrochemical sensor array for H2, O2, and HF ion gases. Taking H2 as an example, its generation concentration reaches 500 ppm within 10 seconds, and the mass spectrum peak appears at 2 amu, confirming it as hydrogen. All collected data is matched with a gas spectral database, and principal component analysis (PCA) is used to remove interference from mixed components, ultimately identifying the gas types and forming a structured "gas generation type" data item list, which is then mapped one-to-one with the corresponding reaction pathways and fault types.

[0095] Gas release simulations were performed based on the gas generation rate and the types of gases generated to obtain gas release data.

[0096] In this embodiment, a three-dimensional simulation model is established using CFD technology. The model includes: 1) the internal cavity structure of the battery, 2) the safety valve opening parameters (e.g., opening diameter φ=1.2mm), and 3) the gas diffusion coefficient (H2 is 6.24×10⁻⁻⁻⁶). 5 m² / s, CO2 is 1.39×10⁻ 54) Initial pressure (e.g., 101325 Pa), 5) Initial temperature (e.g., 323.15 K). The generation rates and physical properties (molecular weight, diffusion coefficient) of various gases are input into the simulation system. The Navier-Stokes equations and diffusion equations are solved in ANSYS Fluent to simulate the release of multiple gas components. In the boundary condition settings, the heat conduction and gas outflow paths are set according to the battery casing material (e.g., aluminum), casing thickness (1 mm), and release path (top vent). Finally, the gas release time curve, internal pressure change curve, and gas component distribution map on the time axis are output, forming "gas release data," which is used for subsequent risk assessment and data retrieval by the low-carbon energy dispatch module.

[0097] Obtain lithium metal data from the energy storage battery; simulate lithium dendrite formation based on the lithium metal data to obtain lithium dendrite data; detect hydrogen evolution components based on the lithium dendrite data to obtain hydrogen evolution data;

[0098] In this embodiment, the scheduling system reads the anode material formula, lithium metal content, anode thickness, and specific capacity parameters of the energy storage battery model through the battery production data interface. The lithium metal content is directly calculated from the anode material capacity (unit: mAh / g) and electrode active material mass (g) specified in the production formula. For example, if the anode active material of a certain battery model is pure lithium, the mass of a single active material is 0.85g, and the specific capacity is 3860mAh / g, then the total lithium metal capacity is 3281mAh. Additionally, the anode layer thickness (unit: μm) is read, with a default thickness of 80μm and an electrode area of ​​100cm². All parameters are obtained from the quality control table of the manufacturing process and are automatically synchronized from the ERP system to the analysis system via the data interface. All lithium metal-related data is stored in the lithium parameter database for subsequent simulation, ensuring that the analysis is based on the actual physical structure. During the simulation of lithium dendrite formation, analysis is performed according to the current density threshold control principle. A standard is set: when the charging current density exceeds 1.5mA / cm², the probability of local lithium dendrite formation increases significantly. The simulation input included a lithium metal layer thickness of 80 μm, an anode area of ​​100 cm², and an actual charging current of 8 A. The calculated actual current density was 0.08 A / cm², far below the critical value. However, the presence of charge deviation and increased electrolyte water content can lead to increased local resistance and induce uneven current density. The simulation used a uniformly distributed local resistance increment of 10 Ω / cm², calculated the corresponding voltage gradient change, and then mapped it to the current density distribution map. If the local current density exceeded 1.5 mA / cm² in the simulation, the area was marked as a dendrite growth risk zone. The maximum dendrite length within 24 hours was further estimated based on the anode metal migration capacity and dendrite migration rate (standard value: 10 μm / h). The output results included the dendrite growth location, predicted length, and coordinates of the covered area, and were uniformly numbered and included in the lithium dendrite database. If the water concentration in the dendrite growth area is greater than 0.5%, H₂ is easily generated at the dendrite tip due to reaction with water. The simulation reaction region was set at the dendrite tip with a diameter of 1 mm, and the standard reaction rate was set to 0.02 mol H2 / mol H2O·h. The input molar amount of water (derived from step three) was substituted into the calculation of the total amount of hydrogen gas evolved per unit time. Taking a water content of 1.6% as an example, the total water content is approximately 0.032 g, which is equivalent to 1.78 mmol H2O. The estimated hydrogen evolution amount over 24 hours is 0.71 mmol. All data were not normalized; the output unit is mmol, and the gas type (H2) was recorded to form a hydrogen evolution data table, which was then mapped one-to-one with the lithium dendrite numbers.

[0099] The degree of electrolyte degradation is determined based on gas release data and hydrogen evolution data.

[0100] In this embodiment, the standard gas molar volume is set to 24 L / mol. Under normal temperature and pressure, the molar number of each gas is converted into volume (unit: ml). If the total gas release volume exceeds 2 ml / 2 ml sample, the electrolyte is considered severely degraded. Further calculations of the proportions of various gases are performed. If the CO2 content exceeds 70%, it indicates that the side reaction is concentrated in carbonate decomposition; if the H2 content is higher than 30%, it indicates that the dendrite hydrogen evolution reaction is dominant. Based on these proportions, the degradation source type is labeled, and a descriptive field such as "Severe Degradation" + "Dendrite Initiated" or "Hydrolysis Dominant" is entered into the electrolyte health status table. All thresholds and calculations are derived from standard data, requiring no manual judgment or abstract model invocation. Finally, the output of this step is a degradation level rating of the energy storage battery electrolyte, which is used by the energy dispatch system in decision-making.

[0101] Preferably, optimizing the electrolyte flow channel structure in step S2 includes:

[0102] High-risk areas of the battery are identified based on the degree of electrolyte degradation.

[0103] In this embodiment, Fourier transform infrared spectroscopy (FTIR) is used to acquire the infrared spectral image of the electrolyte of the faulty energy storage battery. Based on the water characteristic peak area obtained in the previous step, combined with the intensity change of the carbonyl peak in the 1730cm⁻¹ to 1750cm⁻¹ band in the spectrum, the degradation threshold is set to a peak area shift of more than 10%, i.e., ΔS>0.1S0, where S0 is the carbonyl characteristic peak area of ​​the reference electrolyte. For cell units that meet this threshold condition, the corresponding spatial acquisition location of the infrared spectrum is marked on the three-dimensional structure diagram of the battery. A three-dimensional thermal imager is used in conjunction with abnormal temperature areas on the battery surface (temperature exceeding the ambient temperature by more than 20°C, measured at 45°C±2°C) for further cross-verification to complete the calibration of high-risk areas of the battery. The entire process requires spatial positioning through a laser target scanning system in conjunction with a thermal imaging zoning mapping device to ensure a one-to-one correspondence between data and structural regions.

[0104] Identifying electrolyte channels based on high-risk areas of the battery;

[0105] In this embodiment, for the identified high-risk areas, based on the internal design drawings of the energy storage battery, a scanning X-ray micro-computed tomography (Micro-CT) system was used to acquire cell structure slice data with a resolution of 5 μm. The interlayer structure of the electrodes was extracted using an image segmentation algorithm, and the electrolyte seepage path was binarized and modeled. A channel connectivity index C was defined as the number of connected paths of liquid electrolyte from the inlet to the outlet in a single high-risk area. If C < 2 in a certain area, it is identified as a blocked or narrow channel area. Combined with liquid injection dye testing, electrolyte dye (iodine solution) was injected into the environmental chamber using a low-pressure pump, and the seepage trajectory was recorded in real time using a high-magnification visualization microscopic imaging system to further confirm the channel location and direction, outputting an electrolyte channel topology map.

[0106] Identify stagnant areas based on electrolyte channels;

[0107] In this embodiment, using the aforementioned channel topology map and fluid dynamics simulation tools, a finite volume grid is established on the three-dimensional structure map. The incompressible flow Navier-Stokes equations are used for solution, with boundary conditions set as inlet pressure 0.1 MPa and outlet atmospheric pressure. By calculating the rate of change of flow velocity v′ per unit time, the stagnation zone is defined as a region within the local grid where v′ < 0.01 m / s² and duration Δt > 2 s. The corresponding structural coordinates are extracted and mapped to the cell structure. Furthermore, the retention amount L is calculated by combining the electrolyte viscosity parameter (0.9 mPa·s) and temperature field changes. Regions where L = ∫vdt < 3 mm³ are identified as high-incidence stagnation areas. By fusing the three-dimensional battery anatomy map with the stagnation point locations, the stagnation zone identification is completed.

[0108] Adjust the curvature of the electrolyte flow channel in the stagnant region; adjust the width of the electrolyte flow channel in the stagnant region;

[0109] In this embodiment, high-resolution three-dimensional image data of the region is first acquired using a Micro-CT scanning system with a resolution of 5 μm, and exported as a DICOM format image sequence. Then, image processing software (such as ImageJ) is used to slice the image sequence and extract the spatial coordinates of the electrolyte flow channel. These coordinates are imported into mathematical modeling software (such as MATLAB), and the Bezier curve fitting method is used to smoothly model the centerline of each flow channel segment. The curve fitting accuracy is set to cubic curves to ensure that the fitting residual is less than 5 μm. After obtaining the fitted curves, the local curvature of each centerline segment is calculated using the formula K=|x'y''-y'x''| / (x'²+y'²)^(3 / 2), where x' and y' represent the first derivative of the centerline in that region, and x'' and y'' are the second derivatives. The sampling interval is set to 50 μm. After obtaining the curvature data, the calculation results are compared with the average curvature of the established healthy flow channel. If a local curvature value K of a certain flow channel segment is found to be greater than 0.12 mm⁻¹, and the corresponding flow velocity distribution value is less than 0.2 mm / s, it is determined that there is stagnation caused by excessive curvature in that area. For these abnormal areas, CAD software (such as SolidWorks) is used to modify the structural geometry, adjusting the flow channel centerline path to a straight line or a low-curvature polygonal line transition. This ensures that all local curvatures K are within a stable range of 0.05 mm⁻¹ to 0.08 mm⁻¹ after adjustment, and the structural continuity error is controlled within ±0.01 mm. After modification, a geometric comparison diagram before and after modification is output, and a verification model is printed using a 3D printing device (50 μm resolution) to evaluate its forming feasibility and channel unobstructedness. After confirming that there are no structural closures or liquid accumulation dead zones, the new channel path is solidified as the updated flow channel geometry and applied to subsequent structural simulation analysis and manufacturing design.

[0110] The flow channel structure is constructed based on the curvature and width of the electrolyte flow channel.

[0111] In this embodiment, a continuous electrolyte flow channel structure model was established using the SolidWorks 3D modeling system. Each Bezier line segment was used as the central axis, with a uniform cross-sectional width of 130 μm and a height consistent with the original structure (200 μm). A swept solid model was used to generate the overall 3D flow channel. Buffer structures were added to the structural boundaries to prevent current concentration caused by sharp corners. The structure was imported into ANSYS Fluent for fluid simulation verification, with the inlet mass flow rate set to 5e-8 kg / s to simulate electrolyte flow behavior. The output pressure distribution map and velocity vector field were used to check the structural rationality, completing the flow channel structure construction.

[0112] Calculate the current density based on the high-risk areas of the battery; rearrange the electrode stacking method according to the current density; adjust the number of electrode layers according to the current density; construct the electrode stacking structure according to the electrode stacking method and the number of electrode layers.

[0113] In this embodiment, the surface temperature distribution map of the electrode sheet in the high-risk area is obtained. Combining the conductivity of the electrode sheet material (1.5e5S / m for LiCoO2 material) with the thermo-electric coupling relationship, the local current density J is inverted through temperature rise. Using the formula J=√(q / σΔT), where q is the heat generation rate per unit volume (obtained from the infrared temperature gradient), σ is the material conductivity, and ΔT is the temperature rise (the part above the reference temperature), the area with the maximum temperature gradient is mapped to the current density concentration area, and the maximum local current density is measured to be 1.2A / cm². Using this current density data as a reference, the stacking structure optimization stage is entered. A simulation unit of the local electrode layer of the cell is established using a thermal-electric field co-simulation tool. Under the current stacking method (Z-type symmetrical stacking), the thermal field distribution and current density distribution are simulated, and the stacking layer number corresponding to the high current density area is identified. Based on the principle of equal current density distribution, the stacking layers corresponding to the current density peak are adjusted to an alternating asymmetrical arrangement. A high-porosity coated electrode sheet is added every 5 layers to weaken the local charge concentration, and finally the stacking sequence optimization is achieved. After simulating and outputting the new stacking sequence, the CAD model was imported to reconstruct the entire structure. Based on the optimized stacking method, and according to the actual current density of each layer (in A / cm²), the original 50-layer structure was adjusted to 65 layers, keeping the total current load per unit area unchanged, while reducing the average current density per layer to 0.92 A / cm². The logic for adjusting the number of layers is as follows: the original total current density was 50 * A, and the target is that each layer should not exceed 1 A / cm², where A is the total current divided by the total area; therefore, the number of layers was set to 65. The adjusted number of layers was updated in the SolidWorks structural model, and the electrode layer thickness was set to 85 μm and the separator thickness to 25 μm, completing the design of the new layer structure. Combining the adjusted stacking method (asymmetric arrangement) and the number of layers (65 layers), a complete electrode structure was built in 3D CAD using a layer-by-layer modeling method. The core of the structure consists of graphite anode and LiCoO2 cathode layers, sandwiched with a ceramic membrane material. The thicknesses of the ceramic membrane are 60 μm for the anode, 75 μm for the cathode, and 25 μm for the membrane. The electrode width is 65 mm and the height is 100 mm. A transverse current collection channel is incorporated every 10 layers to create a ventilation structure, facilitating subsequent thermal management system integration. After model completion, the model was imported into an electrochemical simulation platform for verification of interlayer voltage drop, voltage field distribution, and thermal field changes.

[0114] By integrating the electrode stack structure and the flow channel structure, an electrolyte flow channel structure is obtained.

[0115] In this embodiment, the constructed electrode stack structure model and the three-dimensional flow channel structure model are assembled and aligned in ANSYS Workbench, with an assembly gap of 0.5 mm set to accommodate the electrolyte channels. A channel-electrode interlock is formed using volume Boolean operations. The initial electrolyte filling volume is set to 5 mL, the flow viscosity to 0.9 mPa·s, and the inlet pressure to 0.15 MPa in the structural simulation. The flow path coverage and electrode wetting degree are calculated. All structural parameters are exported as a .STEP file for subsequent manufacturing process design, completing the construction of the electrolyte flow channel structure.

[0116] Preferably, step S3 specifically includes:

[0117] Step S31: Integrate the energy storage battery structure data into the renewable energy equipment data based on the electrolyte flow channel structure to obtain energy storage battery structure data; monitor the energy storage battery status based on the energy storage battery structure data to obtain energy storage battery status data;

[0118] In this embodiment, data integration of renewable energy devices is performed based on the electrolyte flow channel structure. By combining the electrolyte flow path and structural data of various parts of the battery, 3D modeling software (such as SolidWorks) is used to integrate the geometry of the energy storage battery with the flow characteristics of the electrolyte flow channels. Fluid simulation software (such as ANSYS Fluent) is used to model the electrolyte flow characteristics, ensuring that the electrolyte flow path within the energy storage battery structure is unobstructed and meets the requirements of battery charging and discharging. The obtained energy storage battery structural data includes information such as the battery's electrode layout, flow channel geometry, and electrolyte flow velocity in each part. Based on the energy storage battery structural data, a series of monitoring sensors, such as temperature sensors, voltage sensors, current sensors, and partial discharge sensors, are deployed to monitor the working status of the energy storage battery in real time. By combining the data collected by the sensors with the changes in current, voltage, and temperature distribution within the battery, time-series data analysis methods are used to evaluate the state of the energy storage battery. In particular, by monitoring the battery's temperature changes and current fluctuations, the state changes of the energy storage battery during charging and discharging can be effectively captured, obtaining energy storage battery state data, including battery health status, charging efficiency, and fault warnings.

[0119] Step S32: Perform data preprocessing based on the energy storage battery status data to obtain the energy storage battery status data to be processed; construct a low-carbon energy dispatch data layer based on the energy storage battery status data to be processed;

[0120] In this embodiment, the energy storage battery status data is first preprocessed. Data cleaning methods are used to remove noise and outliers, ensuring the accuracy and validity of the data. Data normalization is performed to ensure all input data are within the same range for subsequent analysis. The preprocessed energy storage battery status data includes time-series data such as battery charge / discharge status, charging voltage, current, and temperature. Next, a low-carbon energy dispatch data layer is constructed based on this data. The data layer construction process includes data standardization, feature extraction, and analysis to ensure the data can be easily accessed and processed on the platform. The energy storage battery status data is integrated with renewable energy production data (such as solar and wind power) within the park to form the low-carbon energy dispatch data input layer. Through analysis of historical data, key features are extracted, such as hourly battery charge / discharge efficiency and the current battery SOC (State of Charge), providing foundational data for subsequent dispatch decisions.

[0121] Step S33: Predict low-carbon energy demand based on the state data of the energy storage batteries to be processed; construct a low-carbon energy dispatch decision layer based on the low-carbon energy demand;

[0122] In this embodiment, the state data of the energy storage batteries to be processed is used to predict the low-carbon energy demand of the park in the future. A predictive model is constructed, employing regression analysis and time series forecasting methods to model the historical state data of the energy storage batteries, in order to predict future energy demand trends. Required parameters include historical battery load data, solar power generation data, wind power generation data, and historical battery charge / discharge state data. Using this data, the predictive model can estimate future electricity demand and energy storage battery charging demand. After obtaining the predicted low-carbon energy demand data, a low-carbon energy dispatch decision layer is constructed. The core objective of the dispatch decision layer is to dynamically adjust energy usage strategies based on the state of the energy storage batteries, energy demand, and renewable energy generation. Through optimization algorithms (such as linear programming and genetic algorithms), combined with the supply of various energy sources within the park, the optimal energy dispatch scheme is determined, including energy storage battery charging / discharging plans and renewable energy priority usage plans. The dispatch decision layer needs to automatically adjust the park's energy usage strategy based on the predicted low-carbon energy demand to ensure efficient energy utilization and the achievement of low-carbon emission targets.

[0123] Step S34: Perform carbon emission minimization design based on the state data of the energy storage battery to be processed to obtain carbon emission minimization data; construct a low-carbon energy dispatch execution layer based on the carbon emission minimization data;

[0124] In this embodiment, based on the state data of the energy storage batteries to be processed, the energy management system (EMS) is used to assess the energy consumption of the park and identify potential high-carbon emission sources. A scheduling optimization model is used, employing a cost function based on emission factors, to optimize the charging and discharging schedule of the energy storage batteries. The design requires setting certain carbon emission limits, such as a maximum carbon emission of no more than 0.3 kg CO2 / kWh, to ensure that the system operation does not exceed this carbon emission threshold. By adjusting the charging and discharging time window and power output of the energy storage batteries, the use of high-carbon energy is reduced, thereby achieving the goal of minimizing carbon emissions. After obtaining the carbon emission minimization data, a low-carbon energy scheduling execution layer is constructed based on this data. This layer translates specific energy scheduling strategies into actual operation instructions, controlling the charging and discharging behavior of the energy storage batteries and the scheduling of renewable energy equipment. The scheduling execution layer ensures that the park meets its low-carbon energy needs while achieving the predetermined carbon emission minimization target by monitoring and adjusting the operating status of energy equipment in real time.

[0125] Step S35: Integrate the low-carbon energy dispatch data layer, low-carbon energy dispatch decision layer, and low-carbon energy dispatch execution layer to obtain the low-carbon energy dispatch model;

[0126] In this embodiment, the core of the scheduling model is the organic integration of the low-carbon energy scheduling data layer, the low-carbon energy scheduling decision layer, and the low-carbon energy scheduling execution layer, achieving seamless integration of data acquisition, decision analysis, and execution control. The low-carbon energy scheduling model first collects real-time status data of energy storage batteries and renewable energy generation data. After preprocessing by the data layer, this data is transmitted to the decision layer for prediction and analysis. Finally, the execution layer implements specific scheduling operations for the energy storage batteries. During the model design, considering the system's response speed and stability, a distributed data processing architecture (such as a cloud computing platform) is adopted to ensure rapid transmission and processing of real-time data.

[0127] Step S36: Simulate charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data.

[0128] In this embodiment, a low-carbon energy dispatch model is used to simulate the charging and discharging behavior of energy storage batteries. Based on the dispatch strategy output by the model, a battery management system (BMS) is used for simulation to calculate key data such as charging efficiency, current fluctuations, and energy loss during discharge under different dispatch conditions. The charging and discharging behavior simulation is performed stepwise over the entire dispatch cycle with a time step set to 5 minutes to ensure the accuracy and detail of the simulation data. The simulation also needs to consider factors such as the impact of temperature on battery performance and charging / discharging rate limitations. Through this simulation, charging and discharging behavior data of the energy storage batteries are obtained, including key indicators such as the battery's state of charge, energy flow, and system efficiency in each time period. This data can be further used to adjust the energy dispatch scheme and provide a basis for subsequent real-time dispatch decisions.

[0129] Preferably, step S36 specifically includes:

[0130] Step S361: Import the low-carbon energy dispatch model into the simulation software;

[0131] In this embodiment, the core components of the low-carbon energy dispatch model, such as energy storage battery charging and discharging management, energy demand forecasting models, and energy dispatching decision rules, are organized into a standard input format that can be recognized by simulation software. This can be achieved by importing the model-related mathematical formulas, parameter settings, and operational logic into a simulation platform, such as MATLAB / Simulink, PLEXOS, or other software specifically designed for energy management simulation, in text or code format. The input content includes, but is not limited to, parameters such as the capacity, voltage, and charging / discharging efficiency of the energy storage battery, as well as dispatching conditions such as load demand fluctuations and low-carbon energy supply. The model input file in the simulation software needs to accurately describe the dynamic changes in the park's energy demand, the status of the energy storage battery, and the supply of low-carbon energy to ensure that the simulation can reflect the actual operating scenario.

[0132] Step S362: In the simulation software, set the energy storage battery type to lithium battery, the capacity range to 50-500kWh, and the battery voltage range to 100V-500V;

[0133] In this embodiment, the energy storage battery type is set to lithium batteries because lithium batteries are widely used in park energy storage systems and have high charge / discharge efficiency and long service life. The capacity range of the energy storage battery is set to 50-500kWh, which covers the needs of parks of different sizes. The capacity setting of 50kWh is to simulate small-scale applications, while 500kWh represents the energy storage needs of large parks. In addition, the battery voltage range is set to 100V-500V, which is based on the common voltage specifications of lithium batteries. Precisely configuring these parameters through simulation software ensures that the simulated battery performance is consistent with the battery characteristics in actual applications. When making this setting, it is essential to strictly follow the specifications provided by the battery manufacturer to ensure that the voltage and capacity range of each battery cell are consistent with the lithium batteries in the actual energy storage system.

[0134] Step S363: In the simulation software, set the battery's charge and discharge efficiency to 90%-98%, the maximum charging power to 20kW-100kW, and the maximum discharge power to 20kW-100kW.

[0135] In this embodiment, the simulation software sets the charging and discharging efficiency and power of the energy storage battery in detail. The charging and discharging efficiency is set to 90%-98%, which corresponds to the common efficiency of lithium batteries under different usage conditions. The efficiency setting is based on the simulation of the actual battery performance under different temperatures and charging and discharging conditions. A low value of 90% is set to simulate the battery's performance under high load or low temperature environments, while a high value of 98% simulates the battery's optimal efficiency under standard operating conditions. In addition, the maximum charging power is set to 20kW-100kW, a range that adapts to the needs of energy storage systems from small-scale to large-scale. The specific value of the charging power needs to be set according to the energy demand of the park and the actual specifications of the energy storage battery. For example, the charging power is set to 20kW for smaller parks and 100kW for larger parks. Similarly, the maximum discharging power is set in the range of 20kW-100kW to simulate the discharge capacity of the energy storage battery, which is crucial for the park's power supply, ensuring that the energy storage battery can provide sufficient power as needed to cope with load fluctuations.

[0136] Step S364: In the simulation software, set the load demand fluctuation range to 5%-30% and the low-carbon energy supply range to 10%-80%;

[0137] In this embodiment, the load demand fluctuation range and low-carbon energy supply range are set to reflect the dynamic changes in energy use within the park. The load demand fluctuation is set at 5%-30%, meaning that the park's electricity demand fluctuates by 5% to 30% within a day. This fluctuation range is derived from historical electricity consumption data and actual load forecasting models, reflecting the impact of different times, seasons, and weather conditions on load demand. Within this range, the simulation software will simulate load demand fluctuations at different time periods and use this as a key input parameter for charge / discharge scheduling. Simultaneously, the low-carbon energy supply range is set at 10%-80%, indicating that the park's renewable energy supply (such as solar and wind power) varies within a range of 10% to 80%. This setting is based on the park's renewable energy production capacity and its volatility. Solar power generation is affected by weather, and wind power generation also has uncertainties; therefore, the low-carbon energy supply capacity exhibits significant fluctuations. Through this setting, the simulation software can simulate how the low-carbon energy dispatch system responds to load demand and coordinates the charging and discharging behavior of energy storage batteries under different energy supply scenarios.

[0138] Step S365: Run the charge and discharge simulation program in the simulation software and output the charge and discharge behavior data.

[0139] In this embodiment, based on input energy storage battery parameters (such as charging efficiency, discharge power, voltage, etc.) and load demand fluctuations and low-carbon energy supply data, the simulation software performs charge and discharge scheduling calculations according to the low-carbon energy dispatching model. During the simulation, the software simulates the behavior of the energy storage battery under different charge and discharge scenarios according to the set dispatching strategy, including charging time, discharge duration, and power changes during charging and discharging. The simulation process needs to run continuously for a certain period of time to evaluate the operating status of the energy storage battery under different dispatching conditions. After the simulation is completed, the software outputs charge and discharge behavior data. This data includes information such as the real-time battery capacity, power output, charging current, and battery temperature changes. All output data will be used to analyze the performance of the energy storage battery and the effectiveness of the low-carbon energy dispatching strategy. The charge and discharge behavior data will provide detailed information on battery charge and discharge efficiency, power changes, and load matching, so as to optimize the battery dispatching strategy in actual operation and ensure the efficient and low-carbon operation of the park's energy.

[0140] Preferably, step S4 specifically includes:

[0141] Step S41: Calculate battery heat based on charge and discharge behavior data; predict thermal runaway risk based on battery heat.

[0142] In this embodiment, the battery's heat is calculated using the charging and discharging behavior data of the energy storage battery. During charging and discharging, the battery's current and voltage directly affect its heat generation. In the simulation software, the battery's current and voltage data during the charging and discharging process are first acquired. This data is typically provided by the Battery Management System (BMS) and is already included in the charging and discharging behavior data from the previous steps. Based on the battery's current and voltage data, the following heat calculation formula is used:

[0143] ;

[0144] In this formula, Q represents the heat generated by the battery, I is the charging / discharging current, R is the internal resistance of the battery, and t is the charging / discharging time. The battery's internal resistance (R), provided by the battery manufacturer, typically varies with battery usage time and ambient temperature. Battery temperature changes are directly related to heat, therefore, further calculations of the battery's temperature change are needed based on this formula. Based on the calculated battery heat, the risk of thermal runaway needs to be predicted. Thermal runaway refers to an uncontrollable reaction that occurs during battery charging and discharging due to excessively high temperatures, leading to battery damage or fire. The risk of thermal runaway is usually assessed by monitoring and analyzing battery temperature, combined with the battery's thermal runaway threshold. This threshold is typically provided by the battery manufacturer and is set as the battery's maximum safe temperature (e.g., 85°C). Once the battery temperature exceeds this threshold, the risk of thermal runaway increases significantly. Therefore, based on the calculated battery temperature data, if the battery temperature reaches or exceeds the threshold, the risk of thermal runaway is predicted.

[0145] Step S42: Identify the thermal runaway risk area in the park based on thermal runaway risk;

[0146] In this embodiment, the thermal runaway risk of each area within the park is closely related to the charging and discharging status of the energy storage batteries within that area, the ambient temperature, and the heat generation of the batteries. First, based on the location, charging and discharging data, and battery type of each energy storage system in the park, the heat generation of the energy storage batteries in each area is analyzed. For each energy storage system, combining its battery pack heat data with the frequency and intensity of charging and discharging, its temperature change under different loads is predicted and compared with the thermal runaway threshold. The thermal runaway risk of each energy storage system can be assessed using the following formula:

[0147] ;

[0148] in, It is the heat generated by the energy storage system. That is the maximum heat that the energy storage system can withstand. A value greater than 1 indicates a high risk of thermal runaway in the energy storage system. The same calculation is performed on each energy storage area to determine the thermal runaway risk zones for the entire park. These zones are then identified and displayed on the park's management interface using an IoT visualization platform.

[0149] Step S43: Obtain the carbon emission coefficient and energy consumption of the park; calculate the carbon emission of the park based on the carbon emission coefficient and energy consumption; map the carbon emission of the park to the thermal runaway risk area of ​​the park to obtain the carbon emission impact area of ​​the park.

[0150] In this embodiment, energy sensing terminals (including smart meters, calorimeters, water and gas flow sensors, and energy storage unit monitoring and control nodes) deployed within the park collect energy consumption data for various types of energy. The data collection frequency is set to once per minute, and the data is transmitted back to the energy edge server using the NB-IoT protocol. Energy types include electricity, natural gas, industrial heat, and photovoltaic energy storage, and all raw data units are uniformly converted to megajoules (MJ). Simultaneously, a standard carbon emission coefficient table is extracted from the National Carbon Emission Accounting Guidelines, and carbon emission coefficients are matched item by item according to energy type. For example, the carbon emission coefficient for electricity is 0.997 kgCO2 / kWh, and for natural gas it is 2.162 kgCO2 / Nm³. This coefficient data is stored in the park's carbon emission factor database. By writing a carbon emission calculation task script, the collected energy consumption data is sequentially multiplied by the corresponding carbon emission coefficients, and the carbon emission calculation for each region is performed according to the formula "carbon emission = energy consumption × carbon emission coefficient". Energy consumption data for each region is categorized by site number, and the corresponding carbon emission calculation results are written to the carbon emission result table using the region ID as an index. The system has a threshold alert mechanism that marks regions when their daily carbon emissions exceed 800 kg CO2. A thermal runaway risk area layer is generated based on the thermal runaway risk analysis system for the park. This layer is generated through spatial clustering analysis of real-time temperature data (e.g., cell temperature exceeding 65℃), abnormal fluctuations in state of charge (SOC>90% or <10%), and historical failure rate data from energy storage units. It is divided into high, medium, and low risk levels and stored in GIS coordinates. Using the ArcGIS spatial mapping module, the carbon emission results are spatially aligned with the thermal runaway risk layer according to the region number. Polygon intersection operations are used to overlay high-carbon emission areas with high-risk thermal runaway areas, and the intersection is extracted to generate a park carbon emission impact area layer. Finally, this layer is written to the low-carbon dispatch platform database and used for subsequent visualization and dispatch decision-making.

[0151] Step S44: Dispatch coal resources according to the areas affected by carbon emissions in the industrial park to obtain coal resource dispatch data; dispatch natural gas resources according to the areas affected by carbon emissions in the industrial park to obtain natural gas resource dispatch data;

[0152] In this embodiment, coal resource scheduling needs to be adjusted based on the energy demand and carbon emission levels of different areas within the industrial park. Coal use directly impacts carbon emissions; therefore, in areas affected by carbon emissions, priority should be given to replacing some coal consumption with low-carbon or renewable energy sources. A coal resource scheduling strategy is formulated based on the park's energy demand forecast and carbon emission targets. Natural gas resource scheduling needs to be rationally arranged based on the energy demand, carbon emission levels, and the charging and discharging status of energy storage batteries in each area. Natural gas has a lower carbon emission factor than coal; therefore, the usage of natural gas should be appropriately increased in areas affected by carbon emissions to reduce overall carbon emissions. Coal and natural gas resource scheduling data will be generated based on scheduling algorithms and monitored and adjusted in real time through an IoT system. The scheduling data will include information such as the scheduling volume, usage time, and target carbon emission levels of coal and natural gas in each area.

[0153] Of particular importance, step S44 includes the following steps:

[0154] Step S441: Determine the energy demand and carbon emission intensity based on the carbon emission impact area of ​​the industrial park;

[0155] This embodiment analyzes the energy demand of each area within the park, involving real-time consumption data of energy such as electricity and heat in each area. Data on electricity, heat, and water consumption in each area of ​​the park is collected through deployed sensing nodes and transmitted and stored in real time using IoT technology. Based on the collected real-time data and historical energy consumption records, the energy demand of each area is determined, particularly the precise calculation of electricity and heat demand. To calculate the carbon emission intensity of each area, a carbon emission factor is used. This factor is set according to the type and usage of energy, such as the carbon emission coefficients for electricity, coal, and natural gas. Using existing carbon emission data and energy consumption data, combined with the specific usage of the park, the carbon emission intensity of each area is calculated. This calculation typically relies on the product of energy consumption and the carbon emission factor to obtain the carbon emission intensity value per unit time. This data will provide a basis for subsequent coal and natural gas scheduling.

[0156] Step S442: Calculate the coal resource demand based on energy demand and carbon emission intensity;

[0157] In this embodiment, energy demand within the industrial park needs to be prioritized, and the demand is weighted based on carbon emission intensity. High-carbon-emission areas prioritize coal resource allocation, and the demand for coal resources is calculated based on energy consumption and the region's carbon emission intensity. Specifically, the formula for calculating coal resource demand is: Demand = Regional Energy Demand * Carbon Emission Factor. Here, energy demand is derived from cumulative data on regional electricity and heat consumption, while the carbon emission factor is a standard value set according to specific energy usage types. This method allows for the determination of the amount of coal required for each region, ensuring that coal resource allocation matches carbon emissions.

[0158] Step S443: Determine the coal delivery method based on the demand for coal resources;

[0159] In this embodiment, the park's transportation network needs to be assessed, including highways, railways, and warehousing facilities. The choice of coal delivery method should be determined based on the coal demand in each area, the location of storage facilities, and the park's transportation convenience. For example, areas with higher demand should prioritize rail transport, while areas with lower demand can use highways or pipelines. During implementation, a suitable delivery method is selected based on coal demand, transportation timeliness, and cost. Parameters involved include transportation distance, transportation capacity, transportation cost, and timeliness for each method. The delivery method is dynamically optimized by combining actual transportation needs with the park's conditions to ensure that coal arrives in the required areas on time.

[0160] Step S444: Identify areas with high coal resource demand based on coal resource demand; determine the priority allocation level of coal resources for areas with high coal resource demand to obtain the priority allocation level of coal resources.

[0161] In this embodiment, regions are prioritized based on their coal demand, with regions exhibiting higher demand designated as high-priority regions. A scheduling strategy is established to allocate different coal resource priorities to each region according to demand, typically categorized into high, medium, and low priority levels. High-priority regions are those within the industrial park with high carbon emission intensity and energy demand; these regions should be allocated sufficient coal resources to ensure their energy needs are met promptly. Parameters involved in this process include coal resource demand, peak regional energy consumption, and carbon emission intensity. These parameters are used for dynamic priority adjustment to ensure efficient allocation of coal resources.

[0162] Step S445: Integrate coal delivery methods and coal resource priority allocation levels to obtain coal resource scheduling data;

[0163] In this embodiment, an optimized scheduling model is constructed based on the priority allocation level of coal resources and the delivery method for each region. This model assigns specific coal delivery routes and times to each region based on factors such as demand and delivery timeliness. By comprehensively considering factors such as transportation methods, regional demand, and priorities, final scheduling data is generated. This scheduling data will provide a basis for subsequent coal transportation, allocation, and tracking, ensuring that coal resources can accurately and timely reach every demand area within the park. The technologies involved include data integration and analysis, using scheduling algorithms to optimize delivery routes and times, and interfacing with the traffic network management system to track coal transportation in real time.

[0164] Step S446: Dispatch natural gas resources according to the carbon emission impact area of ​​the park to obtain natural gas resource dispatch data.

[0165] In this embodiment, it is necessary to determine the natural gas demand for each region, based on the region's energy consumption, carbon emission intensity, and priority dispatch strategy. During this process, factors such as natural gas supply sources, reserves, and transportation methods also need to be comprehensively considered. Based on the natural gas demand of each region and combined with the region's carbon emission intensity, natural gas resources are rationally allocated to ensure sufficient natural gas supply to high-carbon emission areas. Similar to coal resources, the natural gas resource dispatch plan also needs to be updated in real time to respond to changes in energy demand within the park. Generating natural gas resource dispatch data provides necessary support for the park's energy management and carbon emission control.

[0166] Step S45: Integrate coal resource scheduling data and natural gas resource scheduling data, and perform visualization processing to obtain low-carbon energy scheduling data for the park.

[0167] In this embodiment, the integrated data will include coal and natural gas dispatch volumes for each region, as well as related energy consumption and carbon emission data. The integrated data will undergo preprocessing to ensure that energy dispatch in each region aligns with the park's low-carbon goals. During visualization, all dispatch data will be displayed through an IoT visualization platform, allowing park managers to view the low-carbon energy dispatch status of each region in real time. By combining the dispatch data with the park's thermal runaway risk areas and carbon emission impact areas, park managers can intuitively see which areas have problems with carbon emissions and energy dispatch, enabling timely adjustments. Ultimately, the integrated low-carbon energy dispatch data will serve as a decision-making basis for optimizing energy use and carbon emission management within the park, promoting the park's development towards a greener and lower-carbon goal.

[0168] Preferably, this specification also provides an IoT-based visualization-based low-carbon energy dispatching system for industrial parks, used to execute the IoT-based visualization-based low-carbon energy dispatching method described above. This IoT-based visualization-based low-carbon energy dispatching system includes:

[0169] The connection topology parsing module acquires renewable energy equipment data; extracts communication protocol data from the renewable energy equipment data; and parses the equipment connection topology based on the communication protocol data.

[0170] The electrolyte flow channel structure optimization module identifies abnormal energy supply nodes based on the equipment connection topology; detects battery charge drift faults based on these abnormal energy supply nodes; determines the degree of electrolyte degradation based on the battery charge drift faults; and optimizes the electrolyte flow channel structure based on the degree of electrolyte degradation.

[0171] The low-carbon energy dispatch model construction module updates the energy storage battery status based on the electrolyte flow channel structure of renewable energy equipment data to obtain energy storage battery status data; constructs a low-carbon energy dispatch model based on the energy storage battery status data; and simulates charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data.

[0172] The park's low-carbon energy dispatch module predicts thermal runaway risk based on charging and discharging behavior data; identifies areas affected by carbon emissions in the park based on thermal runaway risk; and performs visualized low-carbon energy dispatch based on these areas to obtain low-carbon energy dispatch data for the park.

[0173] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0174] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A park low-carbon energy scheduling method based on Internet of Things visualization, characterized in that, Includes the following steps: Step S1: Acquire renewable energy equipment data; extract communication protocol data from the renewable energy equipment data; parse the device connection topology based on the communication protocol data; Step S2: Identify abnormal energy supply nodes based on the device connection topology; detect energy storage battery charge drift faults based on abnormal energy supply nodes; determine the electrolyte degradation level based on the energy storage battery charge drift faults; optimize the electrolyte flow channel structure based on the electrolyte degradation level. Step S3: Update the energy storage battery status based on the electrolyte flow channel structure of the renewable energy equipment data to obtain energy storage battery status data; construct a low-carbon energy dispatch model based on the energy storage battery status data; simulate charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data. Step S4: Predict thermal runaway risk based on charge and discharge behavior data; identify carbon emission impact areas in the park based on thermal runaway risk; conduct low-carbon energy visualization scheduling based on carbon emission impact areas in the park to obtain low-carbon energy scheduling data for the park. 2.The method of claim 1, wherein, Step S1 is as follows: Step S11: Obtain renewable energy device data; Step S12: Extract communication protocol data from renewable energy equipment; Step S13: Identify the communication identifier of the energy storage device according to the communication protocol data; Step S14: Trace the communication relationship based on the communication identifier of the energy storage device, and reconstruct the communication path of the energy storage device according to the communication relationship; Step S15: Construct the device connection topology based on the energy storage device communication path and the energy storage device communication identifier. 3.The method of claim 1, wherein, Step S2 involves identifying nodes with abnormal energy supply, including: Collect device energy supply data based on the device connection topology; Track energy flow based on equipment energy supply data; Identify abnormal energy flows; Detection of copper busbar metal oxidation characteristics based on abnormal energy flow direction; Identify ablation marks based on the metal oxidation characteristics of copper busbars; Loose areas of copper busbars can be located based on ablation marks; By mapping the loose copper busbar area to the equipment connection topology, the abnormal energy supply node is obtained. 4.The method of claim 1, wherein, Step S2, which detects charge drift faults in the energy storage battery, includes: Monitoring battery charging and discharging status based on abnormal energy supply nodes; Identify lithium-ion migration paths based on battery charge / discharge status; Obtain graphite interlayer structure data; perform graphite thermal expansion simulation based on lithium-ion migration path to obtain graphite thermal expansion data; perform graphite contraction simulation based on lithium-ion migration path to obtain graphite contraction data. Electrode plate crack detection was performed based on graphite thermal expansion and shrinkage data to obtain electrode plate crack data. Predicting electrode plate fracture risk based on electrode plate crack data; Electrode metal corrosion was detected based on the lithium-ion migration path to obtain electrode metal corrosion data. Predicting electrode detachment risk based on electrode metal corrosion data; The actual capacity of the energy storage battery is determined based on the risk of plate breakage and plate detachment. Extract the displayed battery power based on the battery's charge and discharge status; The charge drift fault of the energy storage battery can be determined based on the actual charge level and the displayed charge level of the energy storage battery. 5.The method of claim 1, wherein, Determining the degree of electrolyte degradation in step S2 includes: Based on the charge drift fault of the energy storage battery, faulty energy storage batteries are screened, and the electrolyte infrared spectroscopy of the faulty energy storage batteries is detected to obtain the electrolyte infrared spectral data. Identifying moisture characteristic peaks based on electrolyte infrared spectral data; Calculate the area of ​​characteristic peaks based on moisture characteristics; estimate moisture content based on the area of ​​characteristic peaks. Gas release data was obtained by simulating the gas release of a faulty energy storage battery based on its moisture content. Obtain lithium metal data from the energy storage battery; simulate lithium dendrite formation based on the lithium metal data to obtain lithium dendrite data; detect hydrogen evolution components based on the lithium dendrite data to obtain hydrogen evolution data; The degree of electrolyte degradation is determined based on gas release data and hydrogen evolution data. 6.The method of claim 1, wherein, Step S2, optimizing the electrolyte flow channel structure, includes: High-risk areas of the battery are identified based on the degree of electrolyte degradation. Identifying electrolyte channels based on high-risk areas of the battery; Identify stagnant areas based on electrolyte channels; Adjust the curvature of the electrolyte flow channel in the stagnant region; adjust the width of the electrolyte flow channel in the stagnant region; The flow channel structure is constructed based on the curvature and width of the electrolyte flow channel. Calculate the current density based on the high-risk areas of the battery; rearrange the electrode stacking method according to the current density; adjust the number of electrode layers according to the current density; construct the electrode stacking structure according to the electrode stacking method and the number of electrode layers. By integrating the electrode stack structure and the flow channel structure, an electrolyte flow channel structure is obtained. 7.The method of claim 1, wherein, Step S3 is as follows: Step S31: Integrate the energy storage battery structure data into the renewable energy equipment data based on the electrolyte flow channel structure to obtain energy storage battery structure data; monitor the energy storage battery status based on the energy storage battery structure data to obtain energy storage battery status data; Step S32: Perform data preprocessing based on the energy storage battery status data to obtain the energy storage battery status data to be processed; A low-carbon energy dispatch data layer is constructed based on the status data of the energy storage batteries to be processed. Step S33: Predict low-carbon energy demand based on the state data of the energy storage batteries to be processed; construct a low-carbon energy dispatch decision layer based on the low-carbon energy demand; Step S34: Perform carbon emission minimization design based on the state data of the energy storage battery to be processed to obtain carbon emission minimization data; construct a low-carbon energy dispatch execution layer based on the carbon emission minimization data; Step S35: Integrate the low-carbon energy dispatch data layer, low-carbon energy dispatch decision layer, and low-carbon energy dispatch execution layer to obtain the low-carbon energy dispatch model; Step S36: Simulate charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data.

8. The method for low-carbon energy dispatching in industrial parks based on Internet of Things visualization according to claim 7, characterized in that, Step S36 is as follows: Step S361: Import the low-carbon energy dispatch model into the simulation software; Step S362: In the simulation software, set the energy storage battery type to lithium battery, the capacity range to 50-500kWh, and the battery voltage range to 100V-500V; Step S363: In the simulation software, set the battery's charge and discharge efficiency to 90%-98%, the maximum charging power to 20kW-100kW, and the maximum discharge power to 20kW-100kW. Step S364: In the simulation software, set the load demand fluctuation range to 5%-30% and the low-carbon energy supply range to 10%-80%; Step S365: Run the charge and discharge simulation program in the simulation software and output the charge and discharge behavior data.

9. The method for low-carbon energy dispatching in industrial parks based on Internet of Things visualization according to claim 1, characterized in that, Step S4 is as follows: Step S41: Calculate battery heat based on charge and discharge behavior data; predict thermal runaway risk based on battery heat. Step S42: Identify the thermal runaway risk area in the park based on thermal runaway risk; Step S43: Obtain the carbon emission coefficient and energy consumption of the park; The carbon emissions of the park are calculated based on the park's carbon emission coefficient and energy consumption; the carbon emissions of the park are then mapped to the park's thermal runaway risk area to obtain the area affected by the park's carbon emissions. Step S44: Dispatch coal resources according to the areas affected by carbon emissions in the industrial park to obtain coal resource dispatch data; dispatch natural gas resources according to the areas affected by carbon emissions in the industrial park to obtain natural gas resource dispatch data; Step S45: Integrate coal resource scheduling data and natural gas resource scheduling data, and perform visualization processing to obtain low-carbon energy scheduling data for the park.

10. A low-carbon energy dispatching system for industrial parks based on Internet of Things visualization, characterized in that: For executing the IoT-based visualization-based low-carbon energy dispatching method for industrial parks as described in claim 1, the IoT-based visualization-based low-carbon energy dispatching system for industrial parks includes: The connection topology parsing module acquires renewable energy equipment data; extracts communication protocol data from the renewable energy equipment data; and parses the equipment connection topology based on the communication protocol data. The electrolyte flow channel structure optimization module identifies abnormal energy supply nodes based on the equipment connection topology; detects battery charge drift faults based on these abnormal energy supply nodes; determines the degree of electrolyte degradation based on the battery charge drift faults; and optimizes the electrolyte flow channel structure based on the degree of electrolyte degradation. The low-carbon energy dispatch model construction module updates the energy storage battery status based on the electrolyte flow channel structure of renewable energy equipment data to obtain energy storage battery status data; constructs a low-carbon energy dispatch model based on the energy storage battery status data; and simulates charging and discharging behavior based on the low-carbon energy dispatch model to obtain charging and discharging behavior data. The park's low-carbon energy dispatch module predicts thermal runaway risk based on charging and discharging behavior data; identifies areas affected by carbon emissions in the park based on thermal runaway risk; and performs visualized low-carbon energy dispatch based on these areas to obtain low-carbon energy dispatch data for the park.

Images