Mobile energy storage system cluster energy balance management method and system

Abstract

The invention relates to the technical field of energy storage, in particular to a mobile energy storage system cluster energy balance management method and system. According to the method, multi-mode monitoring data are collected in real time, comprehensive sensing of cluster dynamic changes, topological connection relations and unit states is achieved, the problem of unit mutation sensing lag is solved, a topological connection matrix and a topological disturbance field are obtained based on unit mutation data and topological structure data, the influence of unit mutation on topology is quantified, and the overall performance of unit mutation sensing is improved. Then, a cluster balance feature set is obtained in combination with unit state data, a dynamic adaptive topological structure is constructed, the dynamic adaptive topological structure is adjusted in real time along with unit state changes, then energy balance regulation and control parameters are generated accordingly, and the problem that regulation and control parameters are disjointed with the topological balance state is solved; and then a regulation and control expected index obtained through the energy balance regulation and control parameters is compared with a target balance index based on analogue simulation, and a regulation and control strategy immobilization problem is solved through a closed-loop adjustment mechanism, so that cluster energy balance stability is ensured.

Description

Technical Field

[0001] This invention relates to the field of energy storage technology, and in particular to a method and system for energy balance management of mobile energy storage system clusters. Background Technology

[0002] A "mobile energy storage system cluster" is a collaborative system formed by multiple mobile energy storage units (such as vehicles equipped with energy storage batteries, mobile energy storage cabins, etc.) connected through a communication and energy interaction network. Each mobile energy storage unit includes an energy storage module, an energy conversion device, a mobile carrier, and an intelligent control unit, capable of independently completing energy storage, release, and short-distance relocation. The cluster achieves status awareness, data interaction, and energy scheduling between units through wireless or wired connections, forming a unified management and control system. This cluster mode is widely used in scenarios such as emergency power supply, distributed grid peak shaving, and power supply for large-scale events. Compared with fixed energy storage systems, it is more adaptable to dynamically changing load demands and complex site environments, making it an important technical form for improving the flexibility and reliability of energy supply. "Dynamic stratification," as a form of energy balancing, ensures that the cluster can meet external load demands at any time through differentiated SOC interval distribution, while simultaneously achieving a dynamic balance between internal energy consumption and replenishment. The "energy balance management" of mobile energy storage system clusters is the key mechanism to achieve the goal of "dynamic stratification": it achieves overall energy supply and demand balance, reasonable distribution and efficient utilization of the cluster by regulating the energy status and interaction behavior of each mobile energy storage unit in the cluster. Its core objective is to avoid local energy surplus or shortage and ensure the cluster's continuous and stable response to external loads.

[0003] Existing energy balance management models, centered on static partitioning and fixed strategies, attempt to achieve a "dynamically layered" energy balance goal. Typically, energy storage units are pre-divided into different groups based on function or initial state, and energy scheduling during operation primarily relies on preset hierarchical rules. However, in this model, the realization of "dynamically layered" energy balance depends heavily on the initially set static equilibrium, leading to static and fragmented problems. This makes it difficult to cope with topological and energy state changes caused by sudden events such as unit additions or removals. Specifically, if the addition of a new energy storage unit is not detected in time, it can lead to energy assessment deviations and local imbalances caused by unreconstructed links. If the removal of existing units is not promptly eliminated, it can result in ineffective energy consumption and link breaks, leading to regional supply and demand imbalances. These problems undermine the balance foundation of "dynamically layered" energy balance, causing a disconnect between cluster energy assessment and reality, and a mismatch between links and topology, ultimately resulting in disordered or interrupted energy supply and disrupting the stability of cluster energy balance. Summary of the Invention

[0004] The main objective of this invention is to provide a method and system for energy balance management of mobile energy storage system clusters, aiming to solve the technical problems in the prior art.

[0005] This invention proposes an energy balance management method for a mobile energy storage system cluster, wherein the mobile energy storage system cluster includes multiple energy storage units connected by communication, comprising:

[0006] Acquire multimodal monitoring data of a mobile energy storage system cluster, wherein the multimodal monitoring data includes unit mutation data, topology data, and unit status data;

[0007] The topological perturbation field and the topological connectivity matrix are obtained based on the unit mutation data and the topological structure data.

[0008] The cluster equilibrium feature set is obtained based on the topological disturbance field, topological structure data, and cell state data.

[0009] A dynamically adapted topology is obtained based on the topology connection matrix and the unit state data;

[0010] Energy balance control parameters are obtained based on the dynamic adaptation topology and the cluster balance feature set.

[0011] Obtain the target equilibrium index set, and obtain the equilibrium deviation value based on the target equilibrium index set and the energy equilibrium control parameters;

[0012] Determine whether the balance deviation value is greater than a preset threshold;

[0013] If the balance deviation value is not greater than the preset threshold, then the control operation is performed according to the current energy balance control parameters;

[0014] If the balance deviation value is greater than a preset threshold, the energy balance control parameter is adjusted according to the balance deviation value until the balance deviation value is not greater than the preset threshold.

[0015] Preferably, the step of obtaining the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data includes:

[0016] Based on the unit mutation data, obtain the topological mutation site and mutation unit energy supply parameters corresponding to each mutation unit;

[0017] Based on the topology data, obtain the static adjacency matrix, topology constraint rule set, energy storage unit power supply parameters and energy terminal data, and obtain the terminal priority coefficient based on the energy terminal data;

[0018] The topology connectivity matrix is ​​obtained based on the topology mutation sites, the static adjacency matrix, and the topology constraint rule set;

[0019] The importance coefficients and power supply link influence sets of the mutation units are obtained based on the topological mutation sites and the static adjacency matrix, and the topological distortion rate of each mutation unit is obtained based on the power supply link influence set.

[0020] The power supply disturbance index of each mutation unit is obtained based on the power supply parameters of the mutation unit, the power supply parameters of the energy storage unit, the importance coefficient, and the terminal priority coefficient. The disturbance intensity value of each mutation unit is obtained based on the topology distortion rate and the power supply disturbance index.

[0021] The topological perturbation field is obtained based on the topological structure data and the perturbation intensity value.

[0022] Preferably, the step of obtaining the cluster equilibrium feature set based on the topology perturbation field, topology structure data, and cell state data includes:

[0023] The spatiotemporal scalar field is obtained based on the unit mutation data and the topological perturbation field, and the field strength spatial gradient and perturbation intensity time-varying rate of each mutation unit are obtained based on the spatiotemporal scalar field.

[0024] A set of high-perturbation unit identifiers is obtained based on the spatial gradient of the field strength and the time-varying rate of the perturbation intensity, wherein the set of high-perturbation unit identifiers includes spatially dominant units and time-dominant units;

[0025] The topology reconstruction potential value is obtained based on the spatial dominant unit, the temporal dominant unit, the topology structure data, and the topology connection matrix;

[0026] The target SOC and real-time power demand are obtained based on the topology data, and the real-time SOC and real-time power supply are obtained based on the cell status data.

[0027] The SOC stratification deviation is obtained based on the real-time SOC and the target SOC, and the terminal balance index is obtained based on the real-time power supply and the real-time power demand.

[0028] The topology reconstruction potential value, SOC hierarchical deviation, and terminal balance index are used as the cluster balance feature set.

[0029] Preferably, the step of obtaining the dynamically adaptable topology structure based on the topology connection matrix and the cell state data includes:

[0030] The power supply link information is obtained based on the topology connection matrix, and the equivalent distance of each power supply link is obtained based on the power supply link information.

[0031] Based on the unit status data and the power supply parameters of the mutation unit, the schedulable energy, maximum power supply distance and energy transmission efficiency of each power supply unit are obtained, and based on the topology data, the minimum energy demand, lower limit of energy consumption efficiency and power supply time window of each power-consuming terminal are obtained.

[0032] The reliability score of each power supply link is obtained based on the schedulable energy, maximum power supply distance, energy transmission efficiency, link equivalent distance, minimum energy requirement, and the lower limit of energy efficiency. The set of available links is obtained based on the reliability score and the topology connection matrix.

[0033] The power supply capacity value of each available link is obtained based on the schedulable energy, power supply time window, minimum required energy, link equivalent distance, energy transmission efficiency, and the lower limit of energy consumption efficiency. A dynamic adaptation topology is constructed based on the set of available links and the corresponding power supply capacity value.

[0034] Preferably, the step of obtaining energy balance control parameters based on the dynamically adapted topology and the cluster balance feature set includes:

[0035] The link weight, maximum transmission power, and energy transmission efficiency of each available link are obtained based on the dynamic adaptation topology, and the average transmission efficiency is obtained based on the energy transmission efficiency.

[0036] The initial power allocation matrix is ​​obtained based on the link weight and the maximum transmission power;

[0037] Obtain safety constraint data, and obtain a dynamic power allocation matrix based on the safety constraint data and the initial power allocation matrix;

[0038] Obtain time window classification rules, and obtain a baseline time window based on the time window classification rules and the cluster balance feature set. Then, compensate the baseline time window based on the topology reconstruction potential value to obtain the link switching time window.

[0039] The SOC distribution quantiles are obtained based on the unit status data, and the SOC distribution quantiles are adjusted based on the energy transmission efficiency, average transmission efficiency, and cluster equilibrium feature set to obtain quantile adjustment values.

[0040] The quantile adjustment value is adjusted based on the security constraint data to obtain the dynamic SOC hierarchical interval.

[0041] Preferably, the step of adjusting the energy balance control parameter according to the balance deviation value includes:

[0042] Historical regulation data is acquired, and an impact coefficient set is obtained based on the historical regulation data and the energy balance regulation parameters, wherein the impact coefficient set includes link weight impact coefficient, time window impact coefficient, and SOC stratification impact coefficient;

[0043] The parameter interaction matrix is ​​obtained based on the link weight influence coefficient, the time window influence coefficient, and the SOC hierarchical influence coefficient, and the parameter adjustment constraint set is obtained based on the parameter interaction matrix.

[0044] The adjustment strategy control coefficient is obtained based on the historical control data, and the initial adjustment step size set is obtained based on the adjustment strategy control coefficient, the set of influence coefficients, and the equilibrium deviation value.

[0045] The final adjustment step set is obtained based on the initial adjustment step set and the parameter adjustment constraint set.

[0046] This application also provides a mobile energy storage system cluster energy balance management system, including:

[0047] The first acquisition module is used to acquire multimodal monitoring data of the mobile energy storage system cluster, wherein the multimodal monitoring data includes unit mutation data, topology data and unit status data;

[0048] The topology analysis module is used to obtain the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data.

[0049] The feature extraction module is used to obtain a cluster equilibrium feature set based on the topological perturbation field, topological structure data, and cell state data.

[0050] The topology reconstruction module is used to obtain a dynamically adapted topology structure based on the topology connection matrix and the cell state data.

[0051] The second acquisition module is used to acquire energy balance control parameters based on the dynamic adaptation topology and the cluster balance feature set.

[0052] The evaluation module is used to obtain the target equilibrium index set and obtain the equilibrium deviation value based on the target equilibrium index set and the energy equilibrium control parameters.

[0053] The judgment module is used to determine whether the balance deviation value is greater than a preset threshold.

[0054] If the balance deviation value is not greater than the preset threshold, then the control operation is performed according to the current energy balance control parameters;

[0055] If the balance deviation value is greater than a preset threshold, the energy balance control parameter is adjusted according to the balance deviation value until the balance deviation value is not greater than the preset threshold.

[0056] Preferably, the second acquisition module includes:

[0057] The first acquisition unit is used to acquire the link weight, maximum transmission power and energy transmission efficiency of each available link according to the dynamic adaptation topology, and to acquire the average transmission efficiency according to the energy transmission efficiency.

[0058] A power allocation unit is used to obtain an initial power allocation matrix based on the link weight and the maximum transmission power.

[0059] The second acquisition unit is used to acquire safety constraint data and acquire a dynamic power allocation matrix based on the safety constraint data and the initial power allocation matrix.

[0060] The third acquisition unit is used to acquire time window classification rules, acquire a benchmark time window based on the time window classification rules and the cluster balance feature set, and compensate the benchmark time window based on the topology reconstruction potential value to obtain the link switching time window.

[0061] An adjustment unit is used to obtain the SOC distribution quantile based on the unit status data, and adjust the SOC distribution quantile based on the energy transmission efficiency, average transmission efficiency and the cluster equilibrium feature set to obtain the quantile adjustment value.

[0062] The fourth acquisition unit is used to adjust the quantile adjustment value according to the security constraint data to obtain the dynamic SOC hierarchical interval.

[0063] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described mobile energy storage system cluster energy balance management method.

[0064] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described mobile energy storage system cluster energy balance management method.

[0065] This invention has the following characteristics and beneficial effects:

[0066] (1) Real-time comprehensive perception: By collecting multimodal data, the cluster dynamics, topological relationships and unit status are fully perceived, solving the problem of "lagging perception of unit mutations" and ensuring the real-time and completeness of status perception.

[0067] (2) Quantification of topology impact: Construct a topology connection matrix and a disturbance field to transform the impact of unit mutation on topology into a quantifiable spatial field distribution, solve the problem that “the impact of unit mutation on topology cannot be quantified”, and avoid link planning failure.

[0068] (3) Breaking through traditional limitations: extracting spatial field characteristic parameters to form a cluster equilibrium feature set, and integrating topological and energy state data to solve the problem of "disconnect between topological changes and energy equilibrium".

[0069] (4) Dynamic topology adaptation: The topology is adjusted in real time according to the unit state, which solves the problem of "static topology" and avoids power supply imbalance caused by mismatch between link and energy state.

[0070] (5) Precise control parameters: Energy control parameters are generated by combining dynamic topology and equilibrium feature set to solve the problem of "disconnect between control parameters and topological equilibrium state" and avoid control failure.

[0071] (6) Closed-loop optimization mechanism: The equilibrium target is transformed into a calculable deviation value through simulation, and the closed-loop adjustment is combined to solve the problems of "fuzziness of equilibrium target" and "fixation of control strategy" to ensure the stability of energy balance. Attached Figure Description

[0072] Figure 1 This is a schematic diagram of a method flow according to an embodiment of the present invention.

[0073] Figure 2 This is a schematic diagram of the system structure according to an embodiment of the present invention.

[0074] Figure 3 This is a schematic diagram of the internal structure of a computer device according to an embodiment of this application.

[0075] 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

[0076] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0077] like Figure 1 As shown, this application provides a method for energy balance management of a mobile energy storage system cluster, wherein the mobile energy storage system cluster includes multiple energy storage units connected in communication, including:

[0078] S1. Acquire multimodal monitoring data of the mobile energy storage system cluster, wherein the multimodal monitoring data includes unit mutation data, topology data and unit status data;

[0079] S2. Obtain the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data;

[0080] S3. Obtain the cluster equilibrium feature set based on the topological disturbance field, topological structure data, and cell state data;

[0081] S4. Obtain the dynamically adapted topology structure based on the topology connection matrix and the unit state data;

[0082] S5. Obtain energy balance control parameters based on the dynamic adaptation topology and the cluster balance feature set;

[0083] S6. Obtain the target equilibrium index set, and obtain the equilibrium deviation value based on the target equilibrium index set and the energy equilibrium control parameters;

[0084] S7. Determine whether the equilibrium deviation value is greater than a preset threshold;

[0085] If the balance deviation value is not greater than the preset threshold, then the control operation is performed according to the current energy balance control parameters;

[0086] If the balance deviation value is greater than a preset threshold, the energy balance control parameter is adjusted according to the balance deviation value until the balance deviation value is not greater than the preset threshold.

[0087] As described in steps S1-S7 above, this invention collects multimodal monitoring data of the mobile energy storage system cluster in real time through a sensor network, communication module, and intelligent control unit deployed within the cluster. The multimodal monitoring data includes unit mutation data, topology data, and unit status data. Unit mutation data refers to recorded data of sudden state changes (such as the addition of a new unit or the departure of an existing unit) in each energy storage unit (e.g., mobile energy storage vehicle, mobile container, etc.) within the mobile energy storage system cluster. Topology data describes the connection relationships, communication links, and energy interactions between units within the mobile energy storage system cluster. Unit status data refers to the operating status parameters of each mobile energy storage unit. Through these data, the invention enables real-time monitoring of the cluster's dynamic state. Comprehensive perception of state changes (such as sudden additions or subtractions of units), topological connectivity, and the state of the units themselves provides a complete data foundation for subsequent dynamic adjustments. This addresses the problem of "lagging perception of unit mutations" in existing technologies (such as the failure to detect newly added units in a timely manner), avoids energy assessment biases caused by incomplete data, and ensures the real-time and completeness of cluster state perception. Then, based on unit mutation data and topological structure data, a topological connectivity matrix is ​​obtained. This matrix quantifies the connectivity and interaction capabilities between units within the cluster. By structuring the topological relationships in matrix form, a mathematical tool is provided for subsequent topological reconstruction. Furthermore, the impact of unit mutations on the topological structure is quantified, resulting in a topological perturbation field. The concept of a field representing the degree to which cluster topology is affected by sudden changes is introduced. By introducing the concept of a "topology perturbation field," the abstract "impact of unit mutations on topology" is transformed into a quantifiable spatial field distribution. This solves the problem in existing technologies where the "impact of unit mutations on topology cannot be quantified" (e.g., the impact of a unit going offline on the overall topology stability cannot be assessed). This avoids link planning failures caused by ambiguity in topology change assessments. Subsequently, spatial field characteristic parameters are extracted from the topology perturbation field and combined with topology data and unit state data. After quantification, a cluster equilibrium characteristic set is formed. This cluster equilibrium characteristic set refers to a set of quantitative indicators used to measure the overall energy balance state of the cluster, including topology reconfiguration potential, SOC hierarchical deviation, and terminal equilibrium index. This method, which correlates topological perturbations with the degree of energy balance in a cluster, quantifies the "potential impact of topological changes on energy balance" through an index set. It overcomes the limitations of traditional methods that only focus on energy state and ignore topological factors, thus solving the problem of "disconnection between topological changes and energy balance" in existing technologies (e.g., power outages caused by neglecting topologically vulnerable regions). Then, based on the topological connectivity matrix, it filters effective connections and combines them with cell state data to form a dynamic topology that "adapts to the current cell state and topological connections in real time." The dynamic topology refers to a dynamically updated and optimized topology that can adapt to changes in cluster state. This topology reconstruction method considers both physical connectivity and cell energy state.This approach allows the topology to adjust in real-time according to changes in the unit state, rather than remaining fixed, thus solving the problem of "static topology" in existing technologies (e.g., a unit is still connected to the power supply link after its power is depleted). This avoids power imbalance caused by the mismatch between the link and the actual energy state. Then, based on the dynamically adapted topology and the cluster equilibrium feature set, the dynamic topology is combined with equilibrium indices to obtain energy equilibrium control parameters. These parameters are a set of key parameters used to regulate the energy interaction behavior of the cluster, including the dynamic power allocation matrix, link switching time window, and dynamic SOC hierarchical interval. This ensures that the control parameters adapt to the current topology state and target weak points in the equilibrium, addressing the problem of "disconnect between control parameters and topology equilibrium state" in existing technologies (e.g., lack of stricter control constraints for vulnerable topological regions). This avoids control failure caused by unreasonable parameter settings. The implementation effect of the energy equilibrium control parameters is simulated to obtain a set of expected control indices. These indices are obtained through a prediction model, representing the expected energy equilibrium state indices the cluster will achieve after control. Finally, the target equilibrium indices are calculated. The equilibrium deviation value is obtained by calculating the deviation values ​​of the target equilibrium indicator set and the expected control indicator set. The target equilibrium indicator set refers to the pre-set set of indicators for the ideal energy balance state that the cluster needs to achieve. The equilibrium deviation value is a numerical value that quantitatively reflects the difference between the "expected control effect" and the "ideal target." This mechanism of "deviation quantification" transforms the abstract "equilibrium target" into a calculable deviation value, providing a clear quantitative basis for subsequent adjustments. This addresses the problem of "fuzzy equilibrium target" in existing technologies, making the adjustment direction clearer. Finally, the equilibrium deviation value is compared with a preset threshold. If the equilibrium deviation value is not greater than the preset threshold, control operations are performed according to the current energy balance control parameters to avoid unnecessary parameter adjustments and improve control efficiency. If the equilibrium deviation value is greater than the preset threshold, the energy balance control parameters are adjusted according to the equilibrium deviation value until the equilibrium deviation value is no greater than the preset threshold. This closed-loop adjustment mechanism ensures that the control effect ultimately meets the target, rather than being a one-time decision, thus solving the problem of "fixed control strategies" in existing technologies and guaranteeing the stability of the energy balance of the mobile energy storage system cluster.

[0088] In one embodiment, step S2, which involves obtaining the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data, includes:

[0089] S21. Obtain the topological mutation site and mutation unit power supply parameters corresponding to each mutation unit based on the unit mutation data;

[0090] S22. Obtain the static adjacency matrix, topological constraint rule set, energy storage unit power supply parameters and energy terminal data based on the topology data, and obtain the terminal priority coefficient based on the energy terminal data.

[0091] S23. Obtain the topological connectivity matrix based on the topological mutation sites, static adjacency matrix, and topological constraint rule set;

[0092] S24. Obtain the importance coefficient and power supply link influence set of the mutation unit according to the topological mutation site and the static adjacency matrix, and obtain the topological distortion rate of each mutation unit according to the power supply link influence set.

[0093] S25. Obtain the power supply disturbance index of each mutation unit according to the power supply parameters of the mutation unit, the power supply parameters of the energy storage unit, the importance coefficient and the terminal priority coefficient, and obtain the disturbance intensity value of each mutation unit according to the topology distortion rate and the power supply disturbance index.

[0094] S26. Obtain the topological disturbance field based on the topological structure data and the disturbance intensity value.

[0095] As described in steps S21-S26 above, this invention achieves "real-time perception" of mutation events by acquiring the topological mutation site and power supply parameters of each mutation unit, providing accurate initial data for subsequent energy balance adjustments. This solves the problem of "energy assessment deviation" caused by the failure to promptly perceive unit additions or departures in existing technologies. The topological mutation site refers to the specific physical location and network node number in the cluster topology network when a mobile energy storage unit undergoes a mutation (such as addition, departure, or failure). The power supply parameters of the mutation unit refer to the set of energy supply-related parameters possessed by the energy storage unit experiencing the mutation, including state of health (SOH), capacity change, and power change. Furthermore, the invention acquires a static adjacency matrix, a topological constraint rule set, energy storage unit power supply parameters, and energy terminal data, laying the foundation for subsequently constructing a dynamic topological connection relationship and energy regulation model that meets actual operational requirements. The static adjacency matrix is ​​a matrix constructed based on the initial physical connection relationship of the cluster. The elements in the matrix represent whether there is a direct energy transmission link between each energy storage unit, typically represented by "1" for connectivity and "0" for disconnection. The topological constraint rule set... This refers to a set of rules that define the physical boundaries of energy interaction. Energy storage unit supply parameters refer to the energy supply-related parameters of all variable energy storage units in the cluster, including rated capacity, SOC, charging and discharging power, etc. Energy terminal data refers to relevant information about the terminal equipment that uses energy, including terminal type (such as life support system, industrial control equipment, residential electrical equipment, etc.), power outage tolerance time, real-time power consumption, etc. Based on the energy terminal data, the type of energy terminal corresponding to each variable unit is obtained, and the energy terminal is classified according to the type of energy terminal. Then, combined with the priority level correspondence table, the basic priority coefficient of the energy terminal is obtained. Then, combined with the real-time load rate and the system critical power threshold, the basic priority coefficient is dynamically adjusted through the graded adjustment formula to obtain the terminal priority coefficient. By incorporating the energy priority of the energy terminal into the topology analysis, a correlation model of "energy supply unit-terminal demand" is established. This makes the topology data not only include physical connection information, but also integrate the priority logic of energy allocation. This solves the problem of regional supply and demand imbalance caused by "energy supply link and topology mismatch" in the existing technology. The terminal priority coefficient ensures that key energy terminals are given priority in topology adjustment.

[0096] Next, based on the static adjacency matrix and combined with topological mutation sites, the connection relationships of corresponding units in the matrix are updated, and invalid links are filtered according to the topological constraint rule set: for example, if the distance between two units exceeds the maximum transmission distance, even if they are physically connected, they are marked as "unavailable" (set to 0) in the matrix; otherwise, they are marked as "available" (set to a weighted value equal to the maximum transmission power of the link). Finally, a topological connection matrix is ​​generated, which is the core basis for the cluster to perform dynamic energy scheduling. By combining static topology with dynamic mutation and physical constraints to generate a "real-time available" topological connection matrix, the problem of "local imbalance caused by link unreconstruction" in the existing technology can be solved, ensuring that the topological connection relationship is updated in real time with unit mutation, avoiding energy allocation errors caused by invalid links. Then, all energy transmission links affected by the mutated unit are identified to obtain the power supply link influence set, where the power supply link... The path impact set refers to the set of all energy transmission links affected by the mutation unit. Then, based on the topological mutation site and the static adjacency matrix, the set of transmission links between each mutation unit and each energy-consuming terminal is obtained. The mutation unit is abstracted as a vertex, and the transmission link is abstracted as an edge (the edge weight is the link transmission capacity) to construct a topological graph model. Next, the betweenness centrality (statistical proportion of the shortest link passing through the mutation unit) and proximity centrality (calculated based on the shortest link hop count from the mutation unit to other units) are calculated. Finally, the betweenness centrality and proximity centrality are normalized to obtain the mutation unit importance coefficient, which is used to quantify the strategic position of the mutation unit in the topological network. The topological distortion rate is calculated based on the ratio of the number of links affected by the mutation unit to the total number of links. The topological distortion rate refers to the value that quantifies the degree of damage of the mutation unit to the overall structure of the cluster topology.

[0097] And through the formula " "The energy supply disturbance index was calculated, where, Indicates the first The energy supply disturbance index of each mutation unit, Indicates the first The change in capacity of each mutant unit, Indicates the total available capacity of the system. Indicates the first The importance coefficient of each mutation unit. Indicates the first The power change of each mutant unit, Indicates the maximum total power of the system. express Terminal priority coefficient of each mutation unit, The formula, which indicates the sequence number of the mutation unit, is based on the core logic that the degree of disturbance of a mutation unit depends not only on the magnitude of its own capacity and power changes, but also on its critical position in the network structure and the priority of the energy-consuming terminals it serves. The capacity and power changes in the formula represent the net change in energy storage capacity and the net change in charging / discharging power, respectively, when the energy storage unit joins or leaves the system (e.g., unit withdrawal leads to...). It is a negative value. The larger the absolute value of the variable (negative if not positive, positive if included), the more significant the impact of the variable unit on the total system capacity and real-time power output. The ratio of the capacity change to the total available system capacity normalizes the variable unit capacity change into a dimensionless proportion, directly reflecting the relative impact of variable unit capacity changes on the total system capacity. This value is positively correlated with the power supply disturbance index: the larger the ratio (regardless of sign), the stronger the impact of variable unit capacity changes on the overall system energy storage capacity. The variable unit importance coefficient quantifies the strategic position of the variable unit in the topology network structure; the larger the value, the stronger the core role of the variable unit in the energy transmission link, and the more critical its capacity change on the system topology connectivity and energy distribution. Similarly, the ratio of the power change to the maximum total system power quantifies the relative impact of variable unit power changes on the overall system power supply capacity, and is positively correlated with the power supply disturbance index: the larger the ratio, the more significant the impact of variable unit power changes on the real-time power supply. The terminal priority coefficient reflects the variable unit's... The criticality of the energy-consuming terminals served by the unit is indicated by a higher value, signifying that the power fluctuation of the mutation unit directly affects the power supply stability of high-priority energy-consuming terminals. Therefore, it is positively correlated with the power supply disturbance index: under the same power change, the disturbance of the mutation unit corresponding to the high-priority terminal is amplified. Finally, through the superposition of the above two product terms, the power supply disturbance index comprehensively quantifies the overall disturbance degree of the mutation unit's change (joining / leaving) on ​​the system. Finally, through weighted fusion, the disturbance intensity value of each mutation unit is obtained based on the topology distortion rate and the power supply disturbance index. This value comprehensively reflects the overall impact intensity of the mutation unit on the cluster's "topology and energy supply". Furthermore, the topology disturbance field is obtained by combining the topology data. By introducing the concept of "field theory" to describe the spatial distribution of topology disturbance, discrete mutation events are transformed into a continuous field distribution model, which facilitates the system to identify "disturbance-stricken areas" and provides a spatial basis for global regulation, thereby solving the problem of "lack of spatial range assessment" of mutation impact in existing technologies.

[0098] In one embodiment, step S3, which involves obtaining a cluster equilibrium feature set based on the topology perturbation field, topology data, and cell state data, includes:

[0099] S31. Obtain a spatiotemporal scalar field based on the unit mutation data and the topological perturbation field, and obtain the field strength spatial gradient and perturbation intensity time-varying rate of each mutation unit based on the spatiotemporal scalar field.

[0100] S32. Obtain a set of high-perturbation unit identifiers based on the spatial gradient of the field strength and the time-varying rate of the perturbation intensity, wherein the set of high-perturbation unit identifiers includes spatially dominant units and time-dominant units;

[0101] S33. Obtain the topology reconstruction potential value based on the spatial dominant unit, the temporal dominant unit, the topology structure data, and the topology connection matrix;

[0102] S34. Obtain the target SOC and real-time power demand based on the topology data, and obtain the real-time SOC and real-time power supply based on the cell status data.

[0103] S35. Obtain the SOC stratification deviation based on the real-time SOC and the target SOC, and obtain the terminal balance index based on the real-time power supply and the real-time power demand.

[0104] S36. The topology reconstruction potential value, SOC hierarchical deviation, and terminal balance index are used as the cluster balance feature set.

[0105] As described in steps S31-S36 above, this invention obtains the precise time of each energy storage unit's mutation based on the unit mutation data, thus obtaining the unit mutation timestamp. Then, the topology disturbance field is projected onto the power supply link direction to generate a topology scalar field. Based on the unit mutation timestamp and the topology scalar field, the "spatial distribution" and "time series" are fused to construct a spatiotemporal scalar field. Here, the spatiotemporal scalar field refers to a field model that integrates time and spatial dimensions, with spatial and time coordinates as dimensions. The scalar value at each spatiotemporal point represents the disturbance intensity at that location at that time (e.g., ...). Time coordinates ( , The perturbation intensity is 0.8. By introducing a time dimension, the perturbation field is upgraded to a "spatiotemporal dynamic model," achieving a complete characterization of "when, where, and how large" a mutation event occurs. This addresses the problem of "lack of temporal correlation analysis" in existing technologies for mutation events and avoids distortion in perturbation assessment caused by ignoring the temporal differences in mutation occurrences (such as multiple mutations occurring consecutively within a short period). Then, the spatial rate of change of the perturbation intensity around a mutation unit (such as partial derivatives along the x and y directions) is calculated using a spatiotemporal scalar field, thus obtaining the spatial gradient of the field strength. This reflects the rate of change and direction of the perturbation intensity around the mutation unit in space. The time dimension data in the spatiotemporal scalar field is differentiated to obtain the time-varying rate of the perturbation intensity of the mutation unit, which reflects the dynamic trend of the perturbation, such as whether the perturbation intensity is continuously increasing, gradually decreasing, or remaining stable. The two are then fused using a weighted summation formula to obtain a comprehensive perturbation index. This method, which quantifies the dynamic characteristics of perturbation from two dimensions—"spatial diffusion capability" and "temporal persistence capability"—solves the problem of existing... To address the issue of insufficient dynamic trend judgment in the technology regarding the impact of disturbances, and to avoid misjudging long-term effects by focusing only on instantaneous disturbance intensity (e.g., a slowly increasing disturbance may be more dangerous than a strong instantaneous disturbance), the technology iterates through the comprehensive disturbance index of all mutation units, marking regions with values ​​exceeding the dynamic threshold as "high disturbance units," forming a high disturbance unit identifier set containing region coordinates and related unit numbers. Then, mutation units with a field strength spatial gradient contribution (as a percentage of the comprehensive disturbance index) > 60% are classified as spatially dominant units, whose disturbance impact is mainly spatial diffusion (e.g., a newly added unit located in the core of the topology causes large-scale reconstruction of surrounding links). Mutation units with a disturbance intensity time-varying rate contribution > 60% are classified as time-dominant units, whose disturbance impact is mainly time-duration (e.g., the disturbance intensity of a faulty unit gradually increases over time). This classification provides a basis for subsequent targeted regulation, such as: spatially dominant units require topology optimization, while time-dominant units require accelerated response, which helps to break through the traditional "one-size-fits-all" disturbance handling mode.

[0106] Then, based on the topological connectivity matrix and the static adjacency matrix obtained from the topological structure data, the ideal and actual topological connectivity vectors of the spatial dominant unit are further obtained. The ideal topological connectivity vector reflects the original connectivity state of the spatial dominant unit under undisturbed conditions, while the actual topological connectivity vector reflects the actual connectivity state of the spatial dominant unit after being affected by disturbances. These are then determined using the formula… "Calculate the topology reconfiguration potential value, where, This refers to the topology reconfiguration potential value. Refers to mutation unit In actual topology, the connection vector Refers to mutation unit Connection vector in an ideal topology Refers to mutation unit Importance coefficient, It refers to the first One mutation unit, It refers to the dominant spatial unit. This refers to the maximum possible distortion value of the system. This refers to the sequence number of the mutation unit. This refers to the number of time-dominant units. This refers to the total number of mutation units in a highly perturbation region. The topology reconfiguration potential value is used to quantify the potential to improve system topology distortion through topology reconfiguration. A larger value indicates that topology reconfiguration can significantly improve the system's connectivity and energy transfer efficiency, providing a basis for decision-making regarding whether to perform topology reconfiguration. The core logic of the above calculation formula is: taking the distortion of the spatially dominant unit as the core, normalizing it through the system's maximum possible distortion, and then dynamically correcting it with the proportion of the temporally dominant unit, ultimately obtaining a quantitative index that reflects both the necessity of reconfiguration and adapts to the type of system perturbation. The formula contains the term "..." The L1 norm reflects the difference between the actual connection vector and the ideal vector of the mutation unit. The greater the difference, the greater the total distortion component, the smaller the topology reconfiguration potential, and the greater the importance coefficient. Because a small distortion in the core mutation unit may trigger a chain reaction, the higher the total distortion component, the smaller the topology reconfiguration potential. Summing the total distortion component of the spatial dominant unit can reflect the total distortion contribution of all spatial dominant units to the system topology. The larger the value, the more serious the deviation of the topology structure of the spatial dominant unit from the ideal state, and the more significant the defects of the current topology. The maximum possible distortion value is the maximum distortion component of the system under extreme conditions (such as all energy storage units being completely disconnected). As a normalization benchmark, by setting a reference value for the "worst-case" situation, the total distortion component is converted into a relative proportion, making the topology... The topology reconfiguration potential value is comparable across systems of different scales. The larger the value, the smaller the ratio under the same total distortion, indicating a larger tolerance space for distortion and a higher potential benefit from topology reconfiguration. Since disturbances of time-dominant units are primarily time-dependent (e.g., slow energy decay), they are more suitable for resolution through energy scheduling rather than topology reconfiguration. Therefore, by quantifying the proportion of time-dominant units in the formula, excessive reconfiguration resources are avoided for disturbances that can be resolved without reconfiguration, making decisions more aligned with actual needs. Thus, the final obtained topology reconfiguration potential value provides a clear decision-making basis for cluster energy balance management: avoiding over-reconfiguration in time-dominant disturbances and not ignoring reconfiguration needs in space-dominant disturbances, ensuring the accuracy and efficiency of subsequent topology adjustments.

[0107] Next, for each available energy storage unit in the mobile energy storage system cluster, the corresponding real-time SOC and real-time power supply are obtained from the unit status data. Based on the energy terminal data obtained from the topology data, the target SOC and real-time power demand of each energy terminal are further obtained. Then, the absolute deviation between the real-time SOC and the target SOC is calculated, and weighted according to the importance coefficient of the abruptly changing unit to obtain the weighted SOC deviation value. The SOC stratification deviation is then calculated based on the ratio of the weighted SOC deviation value to the system average SOC. The SOC stratification deviation reflects the degree of deviation between the actual SOC and the preset stratified target SOC, thus providing a clear "correction direction" (charging or discharging) and "correction magnitude" basis for energy balance regulation. Next, the terminal priority coefficient corresponding to the energy terminal served by each abruptly changing unit is compared with a judgment threshold. Energy terminals with a terminal priority coefficient greater than the judgment threshold (such as life support systems) are classified as high-priority terminals. The system calculates the power demand ratio of high-priority energy-consuming terminals to the real-time power supply of corresponding mutation units to obtain the power supply satisfaction rate. Then, the reciprocal of the standard deviation of multiple power supply satisfaction rates is taken as the terminal balance index. The terminal balance index is used to measure the degree of power supply balance among different terminals. The closer the index is to 1, the more balanced the power supply satisfaction of each high-priority terminal is. By quantifying the power supply balance among high-priority terminals, the traditional method breaks through the limitation of only focusing on "whether a single terminal is satisfied". This ensures the stability of the overall power supply of key energy-consuming terminals. Finally, the obtained topology reconfiguration potential value, SOC hierarchical deviation, and terminal balance index together constitute the cluster balance feature set of the mobile energy storage system cluster. Traditional methods for evaluating the balance of energy storage clusters are often limited to a single dimension and ignore the dynamic correlation between disturbance events, topology structure, and energy supply and demand. However, the index set of this invention forms a multi-dimensional and complete evaluation chain, thereby realizing a dynamic, three-dimensional, and accurate characterization of the cluster balance state.

[0108] In one embodiment, step S4, which involves obtaining the dynamically adaptable topology based on the topology connection matrix and the cell state data, includes:

[0109] S41. Obtain power supply link information based on the topology connection matrix, and obtain the equivalent distance of each power supply link based on the power supply link information;

[0110] S42. Based on the unit status data and the power supply parameters of the mutation unit, obtain the schedulable energy, maximum power supply distance and energy transmission efficiency of each power supply unit, and based on the topology data, obtain the minimum energy demand, lower limit of energy consumption efficiency and power supply time window of each power terminal.

[0111] S43. Obtain a reliability score for each power supply link based on the schedulable energy, maximum power supply distance, energy transmission efficiency, link equivalent distance, minimum energy requirement, and the lower limit of energy efficiency, and obtain a set of available links based on the reliability score and the topology connection matrix.

[0112] S44. Obtain the power supply capacity value of each available link based on the schedulable energy, power supply time window, minimum required energy, link equivalent distance, energy transmission efficiency, and the lower limit of energy consumption efficiency, and construct a dynamic adaptation topology based on the set of available links and the corresponding power supply capacity value.

[0113] As described in steps S41-S44 above, this invention obtains power supply link information, whereby the power supply link information refers to data related to energy transmission links extracted from the topology connection matrix, including multiple power supply links and the number of hops and single-hop physical distance for each power supply link, and uses the formula " "Calculate the equivalent distance of the link, where, Indicates the equivalent distance of the link. Indicates the link hop count. Indicates the sequence number of the single-hop link. Indicates the first The physical distance of a single-hop link can be converted into an equivalent distance through this "hop-weighted" method, overcoming the limitation of the traditional method of "evaluating the link only by physical distance". This method can more accurately reflect the energy loss characteristics of multi-hop transmission. Then, based on the unit status data and the power supply parameters of the abrupt unit, the schedulable energy, maximum power supply distance and energy transmission efficiency of each power supply unit are obtained. Among them, schedulable energy refers to the total amount of energy that the power supply unit can actually output. The maximum power supply distance refers to the maximum physical distance that the power supply unit can effectively transmit energy under the current technical conditions (such as the transmission method is laser, electromagnetic induction, etc.). Energy transmission efficiency refers to the effective utilization rate of energy in the process of transmitting energy from the power supply unit to the power user terminal. Based on the power user terminal data, the minimum energy requirement, the lower limit of energy efficiency and the power supply time window of each power user terminal are obtained. Among them, the minimum energy requirement refers to the minimum energy value required by the power user terminal to maintain basic functions. The lower limit of energy efficiency refers to the minimum energy transmission efficiency allowed when the power user terminal is working normally. The power supply time window refers to the specific time period during which the power user terminal needs to receive power.

[0114] Then according to the formula " "Obtain a reliability score for each power supply link, among which, Indicates reliability score, Indicates the equivalent distance of the link. Indicates the maximum power supply distance. Indicates the minimum energy required. Indicates dispatchable energy. This indicates the lower limit of energy efficiency. Indicates energy transfer efficiency, if " If I=0, it indicates that the energy is beyond the physical transmission range and cannot be effectively transferred, making the link unavailable. Conversely, if I=1, the distance constraint is satisfied. The schedulable energy reflects the actual output energy of the energy storage unit and must be greater than the minimum energy requirement of the energy demanding unit. If I=0, it means the energy storage unit cannot meet the basic load demand and the link is unavailable; otherwise, the energy constraint is met (I=1). Energy transmission efficiency reflects the degree of loss in energy conversion and transmission and must be higher than the lower limit of energy efficiency required by safety regulations. If the value is 0, it indicates that the energy loss is too high, the actual effective output is insufficient, and the link is unusable (I=0). Conversely, if the value is 1, the efficiency constraint is met (I=1). According to the above calculation formula, if any condition is not met, it indicates that the power supply link is unusable. If all conditions are met, it indicates that the power supply link is usable. By matching the link reliability in multiple dimensions of "distance + energy + efficiency", the reliability of the link is quantified, thereby solving the problem of power supply interruption caused by "one-sided link reliability assessment" in the existing technology (such as the link distance meets the standard but the energy is insufficient). Then, based on the topology connection matrix, all "usable links" are retained to obtain the set of usable links.

[0115] And through the formula " "Calculate the energy supply capacity value, where, This refers to the energy supply capacity value. It refers to dispatchable energy. This refers to the energy supply time window. This refers to the minimum energy required. This refers to the spatial attenuation coefficient, which is determined by the transmission technology. This refers to the link's equivalent distance. This indicates the lower limit of energy efficiency. Indicates energy transfer efficiency, This item constitutes the "energy sufficiency coefficient," reflecting the ability of an energy storage unit to meet real-time power demand within the energy supply time window. The larger the coefficient, the more abundant the energy reserves of the energy storage unit, the higher the energy redundancy, and the greater the energy supply capacity value. Since the energy transmission loss is higher with the larger the equivalent link distance, resulting in a decrease in effective output, the energy supply capacity value is inversely proportional to the equivalent link distance. This item constitutes the "efficiency compliance coefficient," reflecting the gap between the current efficiency and the design benchmark. The higher the efficiency, the higher the energy utilization rate, the better the link quality, and the greater the power supply capacity value. The obtained power supply capacity value can quantify the "power supply quality" of each link in the available link set. Therefore, based on the available link set, the matrix elements are updated to "power supply capacity values" to form a dynamically adaptable topology. Each non-zero element in the matrix represents both the availability of the link and its power supply capacity, which can be directly used for subsequent energy scheduling to solve the scheduling blindness caused by the "separation of topology and power supply capacity" in the existing technology (such as knowing that the link exists but not knowing how much power it can supply), making subsequent scheduling decisions more accurate.

[0116] In one embodiment, step S5, which involves obtaining energy balance control parameters based on the dynamically adapted topology and the cluster balance feature set, includes:

[0117] S51. Obtain the link weight, maximum transmission power and energy transmission efficiency of each available link according to the dynamic adaptation topology, and obtain the average transmission efficiency according to the energy transmission efficiency.

[0118] S52. Obtain the initial power allocation matrix based on the link weight and the maximum transmission power;

[0119] S53. Obtain safety constraint data, and obtain the dynamic power allocation matrix based on the safety constraint data and the initial power allocation matrix;

[0120] S54. Obtain the time window classification rules, and obtain the benchmark time window according to the time window classification rules and the cluster balance feature set, and compensate the benchmark time window according to the topology reconstruction potential value to obtain the link switching time window.

[0121] S55. Obtain the SOC distribution quantile based on the unit status data, and adjust the SOC distribution quantile based on the energy transmission efficiency, average transmission efficiency and the cluster equilibrium feature set to obtain the quantile adjustment value.

[0122] S56. Adjust the quantile adjustment value according to the security constraint data to obtain the dynamic SOC hierarchical interval.

[0123] As described in steps S51-S56 above, this invention obtains the power balancing weight of each available link through S-curve mapping, thereby achieving "smooth quantification" of SOC deviation and avoiding weight abrupt changes caused by linear mapping (such as small fluctuations in SOC causing large changes in weight). This makes subsequent power balancing control more stable. The terminal key weight is obtained by multiplying the power supply capacity value of each available link by the corresponding terminal priority coefficient. Then, the topology stability coefficient is obtained by the difference between the topology distortion rate and "1". Finally, the topology adaptation weight is obtained by multiplying the topology stability coefficient by the topology reconfiguration potential value. The product of the power balancing weight, the terminal key weight, and the topology adaptation weight is used as the chain weight of each available link. Link weight, where link weight is a quantitative indicator used to measure the importance of each available link in the dynamically adapted topology, addresses the problem of "high-priority terminals being assigned to unstable links" caused by "one-sided link selection" in existing technologies (e.g., only considering terminal priority coefficients while ignoring link reliability). It obtains the maximum transmission power and energy transmission efficiency of each available link based on the dynamically adapted topology. Maximum transmission power refers to the maximum energy transmission limit that each available link can carry under safe and stable operating conditions, and energy transmission efficiency is an indicator measuring the effective utilization rate of link energy transmission. The process begins by obtaining the average transmission efficiency. Then, the initial power of each available link is calculated by multiplying the link weight and the maximum transmission power. For each energy-consuming terminal, the initial power of all its input links is adjusted proportionally to ensure the total power equals the unit's required power, thus obtaining the initial power allocation matrix. This initial power allocation matrix is ​​a pre-constructed matrix describing the power allocation ratio of each link. Normalization ensures accurate matching between power allocation and demand, avoiding oversupply or undersupply caused by "power allocation mismatch" in existing technologies. Finally, safety constraint data is obtained, which refers to a series of limiting conditions that ensure the stable operation of the mobile energy storage system cluster. The data includes SOC constraint boundary values, single-link power limits, power supply unit capacity limits, minimum SOC value, and maximum SOC value. The initial power allocation matrix is ​​iteratively adjusted through single-link power limits and power supply unit capacity limits. For example, the power of overloaded links is reduced, and the remaining quota is allocated according to the link weight ratio until all constraints are satisfied, forming the final dynamic power allocation matrix. The dynamic power allocation matrix refers to the final power allocation scheme matrix obtained by adjusting the initial power allocation matrix in combination with safety constraint data. Through this dual safety constraint mechanism of "link level + unit level", both single-link overload and power supply unit overcapacity output are avoided, thereby ensuring the physical safety of the system.

[0124] Then, the topology reconfiguration potential value, SOC hierarchical deviation, and terminal balance index are used as time window classification evaluation indicators, and time window classification rules are obtained. The time window classification evaluation indicators are matched with the time window classification rules to obtain the baseline time window. The baseline time window refers to the initially determined link switching time baseline value. The baseline time window is compensated according to the link hop count, terminal critical weight, and topology reconfiguration potential value. For example, for each additional hop on the baseline time window, it is shortened by 10%. The purpose of this is to enable long links with many hops to respond to switching more quickly and reduce the power interruption problem that may be caused by many transmission links. For terminals with high critical weight, The baseline time window is shortened by 20%, but not less than 5 seconds after the shortening. This compensation ensures that the links of important terminals can switch quickly. When the topology reconstruction potential value is <0.4, the baseline time window is extended by 50%; when the topology reconstruction potential value is ≥0.4, it is not extended. This adapts to the switching requirements under different topology reconstruction capabilities. After adjustment, the critical link switching time window is obtained. Through this triple dynamic compensation mechanism, the time window adapts to the physical characteristics of the link and reflects the terminal priority coefficient and topology status, thereby achieving differentiated response and solving the problems of "long link switching delay" or "insufficient response of high-priority terminals" caused by "undifferentiated time windows" in the existing technology.

[0125] Next, the SOC distribution quantiles are calculated based on the real-time SOC of each power supply unit. The SOC distribution quantiles refer to the SOC distribution node values ​​divided according to statistical quantiles (such as quartiles or medians) based on the real-time SOC of all power supply units. A transmission efficiency weight is obtained based on the ratio of energy transmission efficiency to average transmission efficiency. Then, the SOC distribution quantiles are adjusted using the transmission efficiency weight, terminal criticality weight, and topology reconfiguration potential value to obtain the final quantile adjustment value. For example, since transmission efficiency directly affects the effective utilization of energy in the link, links with high efficiency have low energy loss, and the SOC distribution quantile can be appropriately relaxed. The quartile range allows high-SOC units to discharge more actively and low-SOC units to charge more easily. Therefore, if the transmission efficiency weight is >1, the first quartile is lowered by 5%~10% (expanding the discharge layer range), and the third quartile is raised by 5%~10% (expanding the charging layer range). If the transmission efficiency weight is <1 (efficiency below average), the first quartile is raised by 5%~10% (shrinking the discharge layer), and the third quartile is lowered by 5%~10% (shrinking the charging layer). Since energy-consuming terminals with high critical weights have extremely high requirements for power supply stability, to ensure their continuous power supply, the quartiles corresponding to these high-critical-weight terminals need to be adjusted. The power supply unit is in a better SOC state. Therefore, for links with high critical weight at the terminal, the first quartile is increased by 10%~15% (raising the lower limit of the discharge layer to ensure that the power supply unit has a high energy reserve), and the third quartile is increased by 5%~10% (expanding the upper limit of the charging layer to allow more energy reserves). For links with low weight, the first and third quartiles can be appropriately lowered to reduce energy redundancy. When the topology reconfiguration potential value is low, topology adjustment is difficult, and it is necessary to reduce energy interaction and reduce topology pressure by shrinking the SOC quartile range. Therefore, when the topology reconfiguration potential value is <0.4, the first quartile is increased by 10%. The third quartile is lowered by 10%, narrowing the SOC quartile range; when the topology reconfiguration potential value is ≥0.4, the first quartile is lowered by 5%~10%, and the third quartile is raised by 5%~10%, widening the range. This method of linking transmission efficiency with SOC quartile and enabling SOC layering to respond to both "terminal balance state" and "topology reconfiguration capability" is conducive to achieving dynamic adaptation of energy allocation and overall system state, improving energy utilization efficiency, and thus solving the problems of "no adjustment of energy allocation when terminals are unbalanced" or "forced energy interaction when the topology is fragile" caused by "static SOC layering" in the existing technology.

[0126] The adjusted quantiles (first quartile, second quartile, and third quartile) reflect the statistical distribution characteristics of the State of Charge (SOC) of each mobile energy storage unit within the cluster. For example: the first quartile represents the threshold of the low SOC range, with approximately 25% of the units having an SOC below this value; the second quartile represents the midpoint of the medium SOC range, with approximately 50% of the units having an SOC below this value; and the third quartile represents the threshold of the high SOC range, with approximately 75% of the units having an SOC below this value. These quantiles are naturally suitable as the basis for dividing the stratified ranges and can directly correspond to the boundaries of the "discharge layer, balance layer, and charging layer." If the first quartile is greater than the minimum SOC value, then the first quartile is used as the lower limit of the discharge layer; if the first quartile is less than the minimum SOC value, then... The minimum SOC (State of Charge) serves as the lower limit of the discharge layer, while the midline of the balance layer represents the "balance point" of the cluster's energy, typically corresponding to the second quartile. However, to ensure that the SOC remains within the range of the minimum and maximum values, if the third quartile is less than the maximum SOC, it is used as the upper limit of the charging layer; conversely, if the third quartile is greater than the maximum SOC, the maximum SOC is used as the upper limit of the charging layer. The lower limit of the discharge layer, the balance layer, and the upper limit of the charging layer constitute the dynamic SOC stratification interval. This dynamic SOC stratification interval, determined through the logic of "quartile division + safety range constraint + dynamic smoothing adjustment," reflects the real-time energy distribution characteristics of the cluster (through quartile adaptation) and ensures the safe operation of the energy storage units, while achieving stable regulation through a smoothing mechanism. This process solves the problem that static stratification in existing technologies cannot adapt to dynamic changes in cluster energy, making the "dynamic stratification" objective more aligned with actual operational needs and improving the flexibility and reliability of energy balance management.

[0127] In one embodiment, step S7, which adjusts the energy balance control parameter based on the balance deviation value, includes:

[0128] S71. Obtain historical control data, and obtain an impact coefficient set based on the historical control data and the energy balance control parameters, wherein the impact coefficient set includes link weight impact coefficient, time window impact coefficient and SOC stratification impact coefficient.

[0129] S72. Obtain the parameter interaction matrix based on the link weight influence coefficient, the time window influence coefficient, and the SOC hierarchical influence coefficient, and obtain the parameter adjustment constraint set based on the parameter interaction matrix;

[0130] S73. Obtain the adjustment strategy control coefficient based on the historical control data, and obtain the initial adjustment step size set based on the adjustment strategy control coefficient, the set of influence coefficients, and the equilibrium deviation value.

[0131] S74. Obtain the final adjustment step set based on the initial adjustment step set and the parameter adjustment constraint set.

[0132] As described in steps S71-S74 above, this invention acquires historical control data, which refers to various data records related to energy balance control during the past operation of the mobile energy storage system, including: balance operation control data, control feedback data, etc. The control feedback data includes historical balance deviation values, historical link weights, historical power deviation rates, historical switching time windows, historical average delays, historical discharge layer failure rates, historical charging layer failure rates, historical SOC dispersion, etc. Then, based on the historical control data and energy balance control parameters, a multiple linear regression model is constructed: with the historical balance deviation value as the dependent variable, and the historical link weights, the product of historical link weights and terminal priority coefficients, and the historical power deviation rate as independent variables. After selecting key influencing features through the model, the link weight influence coefficient is obtained by combining the current link weight and the product of the terminal priority coefficients with reference to historical regression coefficients and normalizing the calculation. The time window influence coefficient is obtained based on the ratio of the historical switching time window to the historical average delay. Simultaneously, the historical discharge layer failure rate and the historical charging layer failure rate are also obtained. The ratio is calculated and weighted at 0.6:0.4 to obtain the overall failure ratio. Then, a nonlinear regression model is constructed, with the historical equilibrium deviation value as the dependent variable. The input is the product of the historical overall failure ratio, the historical SOC dispersion, and "1 minus the historical topology reconstruction potential value". The current overall failure ratio (estimated based on the current SOC layer boundary and the historical unit SOC distribution law) is multiplied by the corresponding historical coefficient, plus the product of the current SOC dispersion (estimated based on the current unit SOC) and "1 minus the current topology reconstruction potential value" multiplied by the corresponding historical coefficient, and then divided by the sum of these two historical coefficients. After normalization, the SOC layer influence coefficient is obtained. Through this multi-dimensional influence coefficient quantification method driven by historical data, the potential influence of link weight, time window, and SOC layer boundary on equilibrium deviation can be accurately quantified before the implementation of adjustment strategy. This provides a quantifiable priority basis for parameter adjustment, solves the problem of blind adjustment caused by the reliance on experience judgment and lack of historical data support in the existing technology, and the technical pain point that the influence of different parameters on equilibrium deviation cannot be quantitatively distinguished.

[0133] Next, based on the link weight influence coefficient, time window influence coefficient, and SOC stratification influence coefficient, a multiple linear regression model is constructed using the least squares method to obtain the coupling coefficients. These coupling coefficients together constitute the parameter interaction matrix. The parameter interaction matrix is ​​a matrix constructed based on the link weight influence coefficient, time window influence coefficient, and SOC stratification influence coefficient. The elements in the matrix correspond to the coupling coefficients between the parameters, thus fully presenting the coupling relationship between link weight, time window, and SOC stratification. This provides a basis for subsequent correction of the initial step size based on this matrix. This method breaks through the limitation of "isolated parameter adjustment" in traditional technology. By quantifying the intrinsic correlation between parameters, it provides a mathematical basis for coordinated adjustment and avoids the chain imbalance caused by the adjustment of a single parameter. Then, based on the coupling coefficients, the dynamic relationship between parameters is analyzed. At the same time, combined with safety constraint data, a complete set of parameter adjustment constraints is formed. This method solves the problem of "mathematical model being out of touch with physical reality" in existing technologies, ensuring that the parameter correlation analysis conforms to the data rules and meets the requirements for safe system operation, thus improving the feasibility of the adjustment strategy.

[0134] Next, adjustment strategy control coefficients are obtained based on historical control data. These coefficients, derived from the analysis and processing of historical control data, are used to control energy balance regulation. They include a comprehensive convergence coefficient (used to control the adjustment speed), an urgency coefficient (which increases with the increase of balance deviation), and a stability coefficient (to prevent system oscillation). The link weight influence coefficient, time window influence coefficient, and SOC hierarchical influence coefficient are multiplied by the aforementioned adjustment strategy control coefficients and the balance deviation value to obtain an initial adjustment step size set. This initial adjustment step size set refers to the adjustment range of the initially determined energy balance regulation parameters. Through the synergistic effect of multi-dimensional coefficients, the adjustment range is precisely matched with the severity of the deviation and the system stability requirements, thus solving the problem that traditional fixed step sizes "cannot adapt to dynamic deviations." Finally, the initial step size is corrected based on the parameter adjustment constraint set to obtain the final adjustment step size set. Through a parameter optimization method with a three-layer mechanism of historical data modeling, parameter coupling analysis, and dynamic step size adjustment, the problem of "mutual conflict in parameter adjustments" in existing technologies is solved, ultimately forming an adjustment strategy that is both scientific and reliable.

[0135] This application also provides a mobile energy storage system cluster energy balance management system, including:

[0136] The first acquisition module is used to acquire multimodal monitoring data of the mobile energy storage system cluster, wherein the multimodal monitoring data includes unit mutation data, topology data and unit status data;

[0137] The topology analysis module is used to obtain the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data.

[0138] The feature extraction module is used to obtain a cluster equilibrium feature set based on the topological perturbation field, topological structure data, and cell state data.

[0139] The topology reconstruction module is used to obtain a dynamically adapted topology structure based on the topology connection matrix and the cell state data.

[0140] The second acquisition module is used to acquire energy balance control parameters based on the dynamic adaptation topology and the cluster balance feature set.

[0141] The evaluation module is used to obtain the target equilibrium index set and obtain the equilibrium deviation value based on the target equilibrium index set and the energy equilibrium control parameters.

[0142] The judgment module is used to determine whether the balance deviation value is greater than a preset threshold.

[0143] If the balance deviation value is not greater than the preset threshold, then the control operation is performed according to the current energy balance control parameters;

[0144] If the balance deviation value is greater than a preset threshold, the energy balance control parameter is adjusted according to the balance deviation value until the balance deviation value is not greater than the preset threshold.

[0145] In one embodiment, the second acquisition module includes:

[0146] The first acquisition unit is used to acquire the link weight, maximum transmission power and energy transmission efficiency of each available link according to the dynamic adaptation topology, and to acquire the average transmission efficiency according to the energy transmission efficiency.

[0147] A power allocation unit is used to obtain an initial power allocation matrix based on the link weight and the maximum transmission power.

[0148] The second acquisition unit is used to acquire safety constraint data and acquire a dynamic power allocation matrix based on the safety constraint data and the initial power allocation matrix.

[0149] The third acquisition unit is used to acquire time window classification rules, acquire a benchmark time window based on the time window classification rules and the cluster balance feature set, and compensate the benchmark time window based on the topology reconstruction potential value to obtain the link switching time window.

[0150] An adjustment unit is used to obtain the SOC distribution quantile based on the unit status data, and adjust the SOC distribution quantile based on the energy transmission efficiency, average transmission efficiency and the cluster equilibrium feature set to obtain the quantile adjustment value.

[0151] The fourth acquisition unit is used to adjust the quantile adjustment value according to the security constraint data to obtain the dynamic SOC hierarchical interval.

[0152] This invention achieves comprehensive perception of cluster dynamic changes, topological connections, and unit states by real-time acquisition of multimodal monitoring data from mobile energy storage system clusters. This addresses the problem of "lagging perception of unit mutations" in existing technologies, avoids energy assessment bias due to incomplete data, and ensures the real-time and completeness of cluster state perception. Based on this data, a topological connection matrix and topological perturbation field are obtained from unit mutation data and topological structure data. The topological relationships are structured in matrix form, and the impact of unit mutations on the topology is transformed into a quantifiable spatial field distribution, solving the problem of "the impact of unit mutations on the topology cannot be quantified" and avoiding link planning failure due to ambiguous topological change assessment. Then, spatial field feature parameters are extracted from the topological perturbation field and combined with topological structure data and unit state data to form a cluster equilibrium feature set. This overcomes the limitation of traditional methods that only focus on energy state and ignore topological factors, solving the problem of "disconnection between topological changes and energy equilibrium." A dynamically adaptable topology is formed based on the topology connection matrix and unit state data, allowing it to adjust in real time as the unit state changes. This solves the problem of "static topology" and avoids energy imbalance caused by mismatch between links and actual energy states. Subsequently, energy balance control parameters are obtained by combining the dynamically adaptable topology and the cluster equilibrium feature set, solving the problem of "disconnect between control parameters and topology equilibrium state" and avoiding control failure due to unreasonable parameter settings. Furthermore, the expected control index set is obtained through simulation, and its deviation value from the target equilibrium index set is calculated, transforming the abstract "equilibrium target" into a calculable deviation value, solving the problem of "fuzzy equilibrium target" and making the adjustment direction clearer. Finally, the equilibrium deviation value is compared with a preset threshold. If it is not greater than the threshold, the current parameters are executed to improve efficiency; if it is greater than the threshold, the parameters are adjusted until the target is met. This closed-loop adjustment mechanism solves the problem of "fixed control strategy" and ensures the stability of energy balance in the mobile energy storage system cluster.

[0153] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described mobile energy storage system cluster energy balance management method.

[0154] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described mobile energy storage system cluster energy balance management method.

[0155] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0156] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0157] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for energy balance management of a mobile energy storage system cluster, wherein the mobile energy storage system cluster comprises multiple energy storage units connected in communication, characterized in that, include: Acquire multimodal monitoring data of a mobile energy storage system cluster, wherein the multimodal monitoring data includes unit mutation data, topology data, and unit status data; The topological perturbation field and the topological connectivity matrix are obtained based on the unit mutation data and the topological structure data. The cluster equilibrium feature set is obtained based on the topological disturbance field, topological structure data, and cell state data. A dynamically adapted topology is obtained based on the topology connection matrix and the unit state data; Energy balance control parameters are obtained based on the dynamic adaptation topology and the cluster balance feature set. Obtain the target equilibrium index set, and obtain the equilibrium deviation value based on the target equilibrium index set and the energy equilibrium control parameters; Determine whether the balance deviation value is greater than a preset threshold; If the balance deviation value is not greater than the preset threshold, then the control operation is performed according to the current energy balance control parameters; If the balance deviation value is greater than a preset threshold, the energy balance control parameter is adjusted according to the balance deviation value until the balance deviation value is not greater than the preset threshold.

2. The mobile energy storage system cluster energy balance management method according to claim 1, characterized in that, The step of obtaining the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data includes: Based on the unit mutation data, obtain the topological mutation site and mutation unit energy supply parameters corresponding to each mutation unit; Based on the topology data, obtain the static adjacency matrix, topology constraint rule set, energy storage unit power supply parameters and energy terminal data, and obtain the terminal priority coefficient based on the energy terminal data; The topology connectivity matrix is ​​obtained based on the topology mutation sites, the static adjacency matrix, and the topology constraint rule set; The importance coefficients and power supply link influence sets of the mutation units are obtained based on the topological mutation sites and the static adjacency matrix, and the topological distortion rate of each mutation unit is obtained based on the power supply link influence set. The power supply disturbance index of each mutation unit is obtained based on the power supply parameters of the mutation unit, the power supply parameters of the energy storage unit, the importance coefficient, and the terminal priority coefficient. The disturbance intensity value of each mutation unit is obtained based on the topology distortion rate and the power supply disturbance index. The topological perturbation field is obtained based on the topological structure data and the perturbation intensity value.

3. The mobile energy storage system cluster energy balance management method according to claim 1, characterized in that, The step of obtaining the cluster equilibrium feature set based on the topology perturbation field, topology structure data, and cell state data includes: The spatiotemporal scalar field is obtained based on the unit mutation data and the topological perturbation field, and the field strength spatial gradient and perturbation intensity time-varying rate of each mutation unit are obtained based on the spatiotemporal scalar field. A set of high-perturbation unit identifiers is obtained based on the spatial gradient of the field strength and the time-varying rate of the perturbation intensity, wherein the set of high-perturbation unit identifiers includes spatially dominant units and time-dominant units; The topology reconstruction potential value is obtained based on the spatial dominant unit, the temporal dominant unit, the topology structure data, and the topology connection matrix; The target SOC and real-time power demand are obtained based on the topology data, and the real-time SOC and real-time power supply are obtained based on the cell status data. The SOC stratification deviation is obtained based on the real-time SOC and the target SOC, and the terminal balance index is obtained based on the real-time power supply and the real-time power demand. The topology reconstruction potential value, SOC hierarchical deviation, and terminal balance index are used as the cluster balance feature set.

4. The mobile energy storage system cluster energy balance management method according to claim 1, characterized in that, The step of obtaining the dynamically adapted topology structure based on the topology connection matrix and the cell state data includes: The power supply link information is obtained based on the topology connection matrix, and the equivalent distance of each power supply link is obtained based on the power supply link information. Based on the unit status data and the power supply parameters of the mutation unit, the schedulable energy, maximum power supply distance and energy transmission efficiency of each power supply unit are obtained, and based on the topology data, the minimum energy demand, lower limit of energy consumption efficiency and power supply time window of each power-consuming terminal are obtained. The reliability score of each power supply link is obtained based on the schedulable energy, maximum power supply distance, energy transmission efficiency, link equivalent distance, minimum energy requirement, and the lower limit of energy efficiency. The set of available links is obtained based on the reliability score and the topology connection matrix. The power supply capacity value of each available link is obtained based on the schedulable energy, power supply time window, minimum required energy, link equivalent distance, energy transmission efficiency, and the lower limit of energy consumption efficiency. A dynamic adaptation topology is constructed based on the set of available links and the corresponding power supply capacity value.

5. The mobile energy storage system cluster energy balance management method according to claim 1, characterized in that, The step of obtaining energy balance control parameters based on the dynamically adapted topology and the cluster balance feature set includes: The link weight, maximum transmission power, and energy transmission efficiency of each available link are obtained based on the dynamic adaptation topology, and the average transmission efficiency is obtained based on the energy transmission efficiency. The initial power allocation matrix is ​​obtained based on the link weight and the maximum transmission power; Obtain safety constraint data, and obtain a dynamic power allocation matrix based on the safety constraint data and the initial power allocation matrix; Obtain time window classification rules, and obtain a baseline time window based on the time window classification rules and the cluster balance feature set. Then, compensate the baseline time window based on the topology reconstruction potential value to obtain the link switching time window. The SOC distribution quantiles are obtained based on the unit status data, and the SOC distribution quantiles are adjusted based on the energy transmission efficiency, average transmission efficiency, and cluster equilibrium feature set to obtain quantile adjustment values. The quantile adjustment value is adjusted based on the security constraint data to obtain the dynamic SOC hierarchical interval.

6. The mobile energy storage system cluster energy balance management method according to claim 1, characterized in that, The step of adjusting the energy balance control parameters according to the balance deviation value includes: Historical regulation data is acquired, and an impact coefficient set is obtained based on the historical regulation data and the energy balance regulation parameters, wherein the impact coefficient set includes link weight impact coefficient, time window impact coefficient, and SOC stratification impact coefficient; The parameter interaction matrix is ​​obtained based on the link weight influence coefficient, the time window influence coefficient, and the SOC hierarchical influence coefficient, and the parameter adjustment constraint set is obtained based on the parameter interaction matrix. The adjustment strategy control coefficient is obtained based on the historical control data, and the initial adjustment step size set is obtained based on the adjustment strategy control coefficient, the set of influence coefficients, and the equilibrium deviation value. The final adjustment step set is obtained based on the initial adjustment step set and the parameter adjustment constraint set.

7. A mobile energy storage system cluster energy balancing management system, characterized in that, include: The first acquisition module is used to acquire multimodal monitoring data of the mobile energy storage system cluster, wherein the multimodal monitoring data includes unit mutation data, topology data and unit status data; The topology analysis module is used to obtain the topological perturbation field and the topological connectivity matrix based on the unit mutation data and the topological structure data. The feature extraction module is used to obtain a cluster equilibrium feature set based on the topological perturbation field, topological structure data, and cell state data. The topology reconstruction module is used to obtain a dynamically adapted topology structure based on the topology connection matrix and the cell state data. The second acquisition module is used to acquire energy balance control parameters based on the dynamic adaptation topology and the cluster balance feature set. The evaluation module is used to obtain the target equilibrium index set and obtain the equilibrium deviation value based on the target equilibrium index set and the energy equilibrium control parameters. The judgment module is used to determine whether the balance deviation value is greater than a preset threshold. If the balance deviation value is not greater than the preset threshold, then the control operation is performed according to the current energy balance control parameters; If the balance deviation value is greater than a preset threshold, the energy balance control parameter is adjusted according to the balance deviation value until the balance deviation value is not greater than the preset threshold.

8. The mobile energy storage system cluster energy balance management system according to claim 7, characterized in that, The second acquisition module includes: The first acquisition unit is used to acquire the link weight, maximum transmission power and energy transmission efficiency of each available link according to the dynamic adaptation topology, and to acquire the average transmission efficiency according to the energy transmission efficiency. A power allocation unit is used to obtain an initial power allocation matrix based on the link weight and the maximum transmission power. The second acquisition unit is used to acquire safety constraint data and acquire a dynamic power allocation matrix based on the safety constraint data and the initial power allocation matrix. The third acquisition unit is used to acquire time window classification rules, acquire a benchmark time window based on the time window classification rules and the cluster balance feature set, and compensate the benchmark time window based on the topology reconstruction potential value to obtain the link switching time window. An adjustment unit is used to obtain the SOC distribution quantile based on the unit status data, and adjust the SOC distribution quantile based on the energy transmission efficiency, average transmission efficiency and the cluster equilibrium feature set to obtain the quantile adjustment value. The fourth acquisition unit is used to adjust the quantile adjustment value according to the security constraint data to obtain the dynamic SOC hierarchical interval.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images