Self-adaptive energy management system based on cloud edge collaborative energy storage cabinet

Abstract

The invention discloses a self-adaptive energy management system based on a cloud-side collaborative energy storage cabinet, which is based on novel digital technologies such as cloud computing and cloud-side collaboration to construct the self-adaptive energy management system of the cloud-side collaborative energy storage cabinet so as to monitor the running state of a battery cell in real time. The health degree, the charge state, the dynamic internal resistance, the maximum temperature difference, the expansion stress, the energy efficiency and the inconsistency degree of the battery cells are monitored, stable and efficient operation of the battery cells is ensured, meanwhile, the fault battery cells are recognized and closed in time through a sensor and a controller, and the stability of the system is improved. A battery cell management system is optimized through a cloud platform in combination with a snake-shaped egret optimization algorithm (SBOA), so that the maximum health degree, the maximum state of charge, the minimum dynamic internal resistance, the minimum maximum temperature difference, the minimum expansion stress, the maximum energy efficiency and the minimum degree of inconsistency are realized, and the safety coefficient and the energy storage efficiency of the system are improved.

Description

Technical Field

[0001] The present invention relates to an adaptive energy management system based on a cloud-edge collaborative energy storage cabinet, belonging to the fields of cloud-edge collaborative energy storage and cloud computing. Background Art

[0002] With the acceleration of the energy structure transformation, the importance of energy storage systems in the power system has become increasingly prominent. Energy storage systems can not only effectively solve the problems of intermittency and uncertainty of renewable energy power generation, but also improve the stability and reliability of the power grid. However, traditional energy storage systems face many challenges during operation, such as differences in cell consistency, insufficient monitoring of system operating status, and untimely fault diagnosis. These problems not only affect the performance and lifespan of the energy storage system, but also increase the operation and maintenance costs and safety risks.

[0003] To address these challenges, an adaptive energy management system based on a cloud-edge collaborative energy storage cabinet has emerged. By integrating advanced technologies such as cloud computing, big data analysis, artificial intelligence, and edge computing, the system realizes real-time monitoring, fault diagnosis, health assessment, and predictive maintenance of the energy storage system's operating status. Specifically, the system uses edge computing devices to collect and preprocess the real-time data of the cells, and transmits the data to the cloud through a cloud-edge collaborative architecture for in-depth analysis and model training. Based on big data and AI technologies, the system can promptly identify potential problem cells, construct fault identification, warning, and predictive maintenance models, thereby achieving early diagnosis of battery faults and accurate assessment of health status. In addition, the system can dynamically adjust the balancing management strategy according to the health status and operating trends of the cells, optimize the system operation plan, and transmit the optimized plan to the cloud platform module. In this way, the system not only improves the safety and reliability of the energy storage system, but also extends the service life of the cells and reduces the operation and maintenance costs.

[0004] The present invention analyzes and processes the real-time data of cell operation using cloud-edge collaborative energy storage technology, provides preventive diagnostic analysis services, uses big data and AI technologies to promptly identify and screen potential problem cells, constructs fault identification, warning, and predictive maintenance models, conducts battery fault diagnosis, health status assessment, evolution trend, and warning judgment, and obtains the optimal operation plan for the cells through the cloud platform combined with an improved optimization algorithm, effectively ensuring the safe and stable operation of the cells, while optimizing the overall performance of the energy storage system.

[0005] Therefore, at the present stage, there is an urgent need for an adaptive energy management system based on a cloud-edge collaborative energy storage cabinet to manage the optimal operation plan of the cell operation system, effectively ensure the safe and stable operation of the cells, while optimizing the overall performance of the energy storage system, reducing the operation and maintenance costs, and extending the service life of the cells. Summary of the Invention

[0006] Objective of the Invention: Aiming at the problems existing in the existing battery cell operation management system, the present invention provides an adaptive energy management system based on a cloud-edge collaborative energy storage cabinet, which manages the optimal operation plan of the battery cell operation system, effectively guarantees the safe and stable operation of the battery cells, optimizes the overall performance of the energy storage system, reduces the operation and maintenance costs, and prolongs the service life of the battery cells.

[0007] Technical Solution: The present invention discloses an adaptive energy management system based on a cloud-edge collaborative energy storage cabinet, which includes a battery cell module, a battery cell monitoring module, a cloud-edge collaborative energy storage module, and a cloud platform module;

[0008] The battery cell monitoring module is used to obtain the health status, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, inconsistency, and position of each battery cell in the battery cell module in real time, and transmit the operation data and position of each battery cell to the cloud platform module and the cloud-edge collaborative energy storage module;

[0009] The cloud-edge collaborative energy storage module determines whether the operation data of the battery cells is within the normal range according to the health status, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency of the battery cells monitored in real time, and determines whether there are potential safety hazards in the battery cells. If there are potential safety hazards, the faulty battery cells are located according to the position, and the faulty battery cells are shut down;

[0010] The cloud platform module constructs a comprehensive performance index for the operation of the adaptive energy management system as a multi-objective function based on the operation data of each battery cell obtained by the battery cell monitoring module, and aims at maximizing the comprehensive performance index. The snake heron optimization algorithm SBOA is used to optimize the battery cell management system to achieve the maximum health degree, the maximum state of charge, the minimum dynamic internal resistance, the minimum maximum temperature difference, the minimum expansion stress, the maximum energy efficiency, and the minimum inconsistency, and obtain the optimal comprehensive performance index.

[0011] Furthermore, a comprehensive performance index for the operation of the adaptive energy management system is constructed as a multi-objective function, and the objective function is:

[0012] The comprehensive performance index Ο is decomposed into 7 sub-objectives to clarify the optimization direction:

[0013] O = αH + βF + γM - δN - εU + ηX - σZ - θ(M·U)

[0014]

[0015] Among them, H is the health score, F is the state of charge, M is the dynamic internal resistance, N is the maximum temperature difference, U is the expansion stress, X is the energy efficiency, Z is the inconsistency, and α, β, γ, δ, ε, η, σ are the weight coefficients corresponding to the health score, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency respectively, and θ is the cross-term coefficient of the dynamic internal resistance M and the expansion stress U.

[0016] Furthermore, while constructing the multi-objective function, dynamic weight design is also carried out, and the weight coefficients are dynamically adjusted according to the real-time working conditions:

[0017]

[0018] In the formula, α(t) represents the dynamic weight of the health weight decaying with time, and δ(t) represents the dynamic weight of preferentially optimizing the temperature difference at high temperatures.

[0019] Furthermore, the specific process of optimizing the battery cell management system using the Snake-Egret Optimization Algorithm (SBOA) is as follows:

[0020] Step 1: Population initialization, randomly generate the individual positions of the snake-egret population in the solution space The dimension is 7, corresponding to 7 optimization parameters: health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency

[0021]

[0022] In the formula, is the position of the j-th dimension of the i-th snake-egret, i = 1, 2,..., n, n is the size of the snake-egret population, u j , l j are respectively the upper and lower limits of the j-th dimension search space;

[0023] Step 2: Multi-objective fitness evaluation and Pareto ranking, calculate the individual fitness, and construct the Pareto front, specifically as follows:

[0024] Step 2.1: For each individual O i , calculate 7 sub-objective values through the real-time data of the sensor:

[0025] O i = αH i + βF i + γM i - δN i - εU i + ηX i - σZ i - θ(M i ·U i )

[0026] Step 2.2: Use fast non-dominated sorting to divide the population into multiple front ranks, and preferentially retain the non-dominated solutions;

[0027] Step 2.3: For the solutions in the same front rank, calculate their crowding distances in the objective space, and retain the individuals with sparse distributions:

[0028]

[0029] where, CrewdingDistance(O i ) is the crowding degree of the i-th solution, G i,k is the value of the i-th solution on the k-th sub-objective function, G i+1,k , G i-1,k are the values of the adjacent solutions of the i-th solution on the k-th sub-objective function, G k,max , G k,min are the maximum and minimum values of the k-th sub-objective function among all solutions respectively;

[0030] Step 3: Achieve the balance between global search and local development by integrating differential evolution and Levy flight strategy;

[0031] Step 3.1: Design a hybrid update strategy to update the position by combining differential evolution and Levy flight:

[0032]

[0033] where, is the optimized comprehensive performance index, O i is the current comprehensive performance index, o best is the historical optimal comprehensive performance index, o random1 and o random2 are the random candidate solutions in the search stage, Levy(β)~u = t -1-β , β ∈ [1, 3] is the random step size of Levy flight, is the dynamic balance factor, which gradually biases towards global search as the iteration number t increases;

[0034] Step 3.2: Impose dynamic constraint conditions on conflicting objectives. When the energy efficiency is high, limit the inconsistency, specifically:

[0035] Z ≤ 5% if X ≥ 90%

[0036] Step 4: Select the comprehensive optimal solution from the Pareto front, that is, obtain the maximized comprehensive performance index and its corresponding health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency.

[0037] Further, when optimizing the battery cell management system using the Secretary Bird Optimization Algorithm (SBOA), feedback is also provided through the cloud-edge collaborative energy storage module. The cloud platform module collects battery cell data in real time, dynamically updates the constraint thresholds and weight coefficients, and performs dynamic weight adjustment and parameter coupling compensation.

[0038] Further, the dynamic weight adjustment and parameter coupling compensation are specifically as follows:

[0039] 1) Weight adaptation rule:

[0040] Health priority principle: When H < 80%, increase α to 1.5α0, decrease β, and reduce the SOC optimization weight;

[0041] Temperature difference emergency response principle: If N ≥ 45°C, immediately increase δ to 2δ0 and trigger the forced cooling strategy;

[0042] 2) Coupling parameter compensation term:

[0043] For the strongly correlated dynamic internal resistance M and expansion stress U, add a cross term to the objective function:

[0044] O comp = γM + εU + θ(M·U)

[0045] In the formula, O comp is the comprehensive performance index after adding the cross term, and θ is fitted according to historical data.

[0046] Further, the process of selecting the comprehensive optimal solution from the Pareto front in step 4 is specifically as follows:

[0047] 1) Entropy weight TOPSIS decision-making:

[0048] Calculate the objective weight ω k based on the entropy weight method, and select the individual closest to the ideal solution:

[0049]

[0050] In the formula, p i,k is the value after normalizing the objective function, E k is the information entropy of the k-th sub-objective function, ω k is the weight of the k-th sub-objective function,

[0051] 2) Edge-side security verification. Send the candidate solution O best to the controller of the energy storage cabinet, simulate the operation and verify: the maximum temperature difference N ≤ 5°C, and the expansion stress U ≤ U safe ;

[0052] 3) Dynamic iteration mechanism. If the above two validations fail, return to step 3 and narrow the solution space range; if successful, output O best as the optimal operation plan.

[0053] Furthermore, a faulty cell isolation module is also provided to locate the faulty cell according to the inconsistency Z. If Z > 10%, trigger the replacement protocol:

[0054]

[0055] Furthermore, upload the current optimization data of the Snake Egrets Optimization Algorithm (SBOA) to the cloud to update the Levy flight parameter β of the SBOA algorithm for further optimization and update at the next moment. The current optimization data includes the optimal solution O best and the constraint violation record.

[0056] Beneficial effects:

[0057] Compared with the traditional cell operation management system, the present invention utilizes cloud-edge collaborative energy storage technology to optimize and upgrade the cell operation system, effectively ensuring the safe and stable operation of the cells. Implement refined and intelligent management for each cell, and the system operation is not affected by faulty cells, optimizing the overall performance of the system. Also, through the optimized Snake Egrets Optimization Algorithm (SBOA) under the comprehensive processing of adaptive energy management, preventive service diagnosis, and cell operation status monitoring in the cloud-edge collaborative energy storage cabinet, the comprehensive performance index of the system is the highest. Description of the Drawings

[0058] Figure 1 is the structural diagram of the system equipment of the present invention;

[0059] Figure 2 is the comparison chart of cell failure rates before and after optimization;

[0060] Figure 3 is the comparison chart of system energy utilization rates before and after optimization. Detailed Embodiments

[0061] The present invention will be further described below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be used to limit the protection scope of the present invention.

[0062] The present invention discloses an adaptive energy management system based on a cloud-edge collaborative energy storage cabinet, including a cell module, a cell monitoring module, a cloud-edge collaborative energy storage module, and a cloud platform module.

[0063] The cloud-edge collaborative energy storage technology is used to monitor the battery cells in real time and achieve balanced management of the battery cells. The battery cell monitoring module monitors the health status, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, inconsistency, and the position of the battery cells, and at the same time transmits the data to the cloud platform module. The cloud-edge collaborative energy storage module judges whether the operating data of the battery cells is within the normal range based on the health status, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency of the battery cells monitored in real time, and judges whether there are potential safety hazards in the battery cells. If there are potential safety hazards, the faulty battery cells are located according to the position and the faulty battery cells are shut down. If there are no potential safety hazards, the original battery cell balanced management is continued, that is, battery cell balance optimization and parameter update.

[0064] Based on the operating data of each battery cell obtained by the battery cell monitoring module, the cloud platform module constructs a comprehensive performance index for the operation of the adaptive energy management system as a multi-objective function, and aims to maximize the comprehensive performance index. The snake heron optimization algorithm SBOA is used to optimize the battery cell management system to achieve the maximum health degree, the maximum state of charge, the minimum dynamic internal resistance, the minimum maximum temperature difference, the minimum expansion stress, the maximum energy efficiency, and the minimum inconsistency, and obtain the optimal comprehensive performance index, effectively improving the energy efficiency ratio, stability, and reliability of the adaptive energy management system of the cloud-edge collaborative energy storage cabinet.

[0065] The health status of the battery cell (SOH) reflects the ratio of the current capacity of the battery to the initial capacity and is used to evaluate the aging degree of the battery. The state of charge (SOC) represents the percentage of the current remaining power of the battery to the total capacity. The dynamic internal resistance is the internal resistance exhibited by the battery during charge and discharge and is affected by temperature, SOC, and aging degree. The maximum temperature difference refers to the maximum temperature difference between the battery cells at different positions in the battery pack. The expansion stress is the mechanical stress generated by the volume change of the battery during charge and discharge. The energy efficiency is the ratio of the output energy of the battery to the input energy and reflects the energy conversion efficiency. The inconsistency refers to the differences in capacity, internal resistance, voltage, etc. of the battery cells in the battery pack. The battery cell monitoring module ensures the safe and stable operation of the battery cells, and at the same time makes predictions about potential safety hazards, shuts down or replaces the faulty battery cells in advance, and ensures the safety and stability of the system operation.

[0066] Construct a comprehensive performance index for the operation of the adaptive energy management system as a multi-objective function, and the objective function is:

[0067] Decompose the comprehensive performance index Ο into 7 sub-objectives to clarify the optimization direction:

[0068] O = αH + βF + γM - δN - εU + ηX - σZ - θ(M·U)

[0069]

[0070] Among them, H is the health score, F is the state of charge, M is the dynamic internal resistance, N is the maximum temperature difference, U is the expansion stress, X is the energy efficiency, Z is the inconsistency, and α, β, γ, δ, ε, η, σ are the weight coefficients corresponding to the health score, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency respectively, and θ is the cross-term coefficient between the dynamic internal resistance M and the expansion stress U.

[0071] While constructing the multi-objective function, dynamic weight design is also carried out, and the weight coefficients are dynamically adjusted according to the real-time working conditions:

[0072]

[0073] In the formula, α(t) represents the dynamic weight of the health weight decaying with time, and δ(t) represents the dynamic weight of preferentially optimizing the temperature difference at high temperatures.

[0074] The specific process of optimizing the battery cell management system using the snake and heron optimization algorithm SBOA is as follows:

[0075] Step 1: Population initialization, randomly generate the individual positions of the snake and heron population within the solution space The dimension is 7, corresponding to 7 optimization parameters: health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency

[0076]

[0077] In the formula, is the position of the j-th dimension of the i-th snake and heron, i = 1, 2,..., n, n is the size of the snake and heron population, u j 、l j are the upper and lower limits of the j-th dimension search space respectively;

[0078] Step 2: Multi-objective fitness evaluation and Pareto ranking, calculate the individual fitness, and construct the Pareto front, specifically as follows:

[0079] Step 2.1: For each individual O i , calculate 7 sub-objective values through the real-time data of the sensor:

[0080] O i = αH i + βF i + γM i - δN i - εU i + ηX i - σZ i - θ(M i ·U i )

[0081] Step 2.2: Use fast non-dominated sorting to divide the population into multiple front ranks, and preferentially retain the non-dominated solutions;

[0082] Step 2.3: For the solutions in the same front rank, calculate their crowding distances in the objective space and retain the individuals with sparse distribution:

[0083]

[0084] where, CrowdingDistance(O i ) is the crowding degree of the i-th solution, G i,k is the value of the i-th solution on the k-th sub-objective function, G i+1,k , G i-1,k are the values of the adjacent solutions of the i-th solution on the k-th sub-objective function, G k,max , G k,min are the maximum and minimum values of the k-th sub-objective function among all solutions respectively;

[0085] Step 3: Achieve the balance between global search and local development by integrating differential evolution and Levy flight strategy;

[0086] Step 3.1: Design a hybrid update strategy to update the position by combining differential evolution and Levy flight:

[0087]

[0088] where, is the optimized comprehensive performance index, O i is the current comprehensive performance index, o best is the historical optimal comprehensive performance index, o random1 and o random2 are the random candidate solutions in the search stage, Levy(β) ∼ u = t -1-β , β ∈ [1, 3] is the random step size of Levy flight, is the dynamic balance factor, which gradually biases towards global search as the iteration number t increases;

[0089] Step 3.2: Impose dynamic constraint conditions on conflicting objectives to limit the inconsistency degree at high energy efficiency, specifically:

[0090] Z ≤ 5% if X ≥ 90%

[0091] Step 4: Select the comprehensive optimal solution from the Pareto front, that is, obtain the maximized comprehensive performance index and its corresponding health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, inconsistency degree.

[0092] When using the Secretary Bird Optimization Algorithm (SBOA) to optimize the battery cell management system, feedback is also provided through the cloud-edge collaborative energy storage module. The cloud platform module collects battery cell data in real time, dynamically updates the constraint thresholds and weight coefficients, and performs dynamic weight adjustment and parameter coupling compensation.

[0093] The dynamic weight adjustment and parameter coupling compensation are as follows:

[0094] 1) Weight adaptation rules:

[0095] Health priority principle: When H < 80%, increase α to 1.5α0, decrease β, and reduce the SOC optimization weight;

[0096] Temperature difference emergency response principle: If N ≥ 45°C, immediately increase δ to 2δ0 and trigger the forced cooling strategy;

[0097] 2) Coupling parameter compensation term:

[0098] For the strongly correlated dynamic internal resistance M and swelling stress U, add a cross term to the objective function:

[0099] O comp = γM + εU + θ(M·U)

[0100] In the formula, O comp is the comprehensive performance index after adding the cross term, and θ is fitted according to historical data.

[0101] The process of selecting the comprehensive optimal solution from the Pareto front is as follows:

[0102] 1) Entropy weight TOPSIS decision:

[0103] Calculate the objective weight ω k based on the entropy weight method, and select the individual closest to the ideal solution:

[0104]

[0105] In the formula, p i,k is the value after normalizing the objective function, E k is the information entropy of the k-th sub-objective function, ω k is the weight of the k-th sub-objective function,

[0106] 2) Edge-side security verification. Send the candidate solution O best to the controller of the energy storage cabinet, simulate the operation and verify: the maximum temperature difference N ≤ 5°C, and the swelling stress U ≤ U safe ;

[0107] 3) Dynamic iteration mechanism. If the above two verifications fail, return to step 3 and narrow the solution space range; if successful, output O bestIt is the optimal operation plan.

[0108] The adaptive energy management system based on the cloud-edge collaborative energy storage cabinet of the present invention is also provided with a faulty battery cell isolation module, which locates the faulty battery cell according to the inconsistency degree Z. If Z > 10%, a replacement protocol is triggered:

[0109]

[0110] Upload the current optimization data of the snake egret optimization algorithm SBOA to the cloud, update the Levy flight parameter β of the SBOA algorithm for further optimization and update at the next moment. The current optimization data includes the optimal solution O best and the constraint violation record.

[0111] Under the simulated high-density energy storage power station scenario, the system of the present invention was continuously tested for 30 days. The experimental group used a cloud-edge collaborative energy storage cabinet (including 20 lithium iron phosphate battery cells with a single capacity of 100 Ah), and the control group used a traditional non-collaborative management scheme. The experimental group dynamically optimized the comprehensive performance index through the snake egret optimization algorithm SBOA, adjusted the weight coefficient in real time and triggered the fault isolation protocol; the control group only performed static equalization management based on the SOC threshold.

[0112] Fault rate test method:

[0113] Artificially inject battery cell faults (such as overcharging, short circuit, and excessive expansion stress), and record the system response time and isolation success rate. The experimental group located the faulty battery cell within 8 seconds and initiated the replacement protocol, and the SOC fluctuation was controlled within ±1.5% (the SOC fluctuation of the control group reached ±10% when not isolated).

[0114] Energy efficiency verification:

[0115] The charging and discharging energy was monitored in real time through a high-precision power analyzer (Yokogawa WT3000). In the low-temperature mode (-10°C), the experimental group adjusted the dynamic weight (increasing ω), and the efficiency was increased by 6% compared with the control group.

[0116] As Figure 2 shown, the battery cell fault rate of the experimental group always remained between 20% and 25%; while the fault rate of the control group always remained between 25% and 35%. As Figure 3 shown, under the dynamic load cycle condition, the energy efficiency of the experimental group increased from the initial 89% to 94%, verifying the effectiveness of the cloud-edge collaborative optimization and adaptive weight strategy.

[0117] The above embodiments are only for illustrating the technical concept and characteristics of the present invention, and their purpose is to enable those who are familiar with this technology to understand the content of the present invention and implement it accordingly, and it should not be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit and essence of the present invention should be covered within the protection scope of the present invention.

Claims

1. An adaptive energy management system based on a cloud-edge collaborative energy storage cabinet, characterized in that, It includes a battery cell module, a battery cell monitoring module, a cloud-edge collaborative energy storage module, and a cloud platform module; The battery cell monitoring module is used to obtain the health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, inconsistency, and position of each battery cell in the battery cell module in real time, and transmit the operating data and position of each battery cell to the cloud platform module and the cloud-edge collaborative energy storage module; The cloud-edge collaborative energy storage module determines whether the operating data of the battery cells is within the normal range and whether there are potential safety hazards based on the health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency of the battery cells monitored in real time. If there are potential safety hazards, the faulty battery cells are located according to the position and the faulty battery cells are shut down; The cloud platform module constructs a comprehensive performance index for the operation of the adaptive energy management system as a multi-objective function based on the operating data of each battery cell obtained by the battery cell monitoring module, and aims to maximize the comprehensive performance index. The snake egret optimization algorithm SBOA is used to optimize the battery cell management system to achieve the maximum health, the maximum state of charge, the minimum dynamic internal resistance, the minimum maximum temperature difference, the minimum expansion stress, the maximum energy efficiency, and the minimum inconsistency, and obtain the optimal comprehensive performance index.

2. The adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 1, wherein Construct a comprehensive performance index for the operation of the adaptive energy management system as a multi-objective function, and the objective function is: Decompose the comprehensive performance index Ο into 7 sub-objectives to clarify the optimization direction: O = αH + βF + γM - δN - εU + ηX - σZ - θ(M·U) Where, H is the health score, F is the state of charge, M is the dynamic internal resistance, N is the maximum temperature difference, U is the expansion stress, X is the energy efficiency, Z is the inconsistency, α, β, γ, δ, ε, η, σ are the weight coefficients corresponding to the health score, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency, and θ is the cross-term coefficient of the dynamic internal resistance M and the expansion stress U.

3. The self-adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 2, wherein, While constructing the multi-objective function, dynamic weight design is also carried out, and the weight coefficients are dynamically adjusted according to the real-time working conditions: In the formula, α(t) represents the dynamic weight of the health weight decaying with time, and δ(t) represents the dynamic weight of preferentially optimizing the temperature difference at high temperatures.

4. The adaptive energy management system based on the cloud-edge collaborative energy storage cabinet according to claim 3, characterized in that The specific process of using the snake egret optimization algorithm SBOA to optimize the battery cell management system is as follows: Step 1: Population initialization, randomly generate the individual positions of the secretarybird population within the solution space The dimension is 7, corresponding to 7 optimization parameters: health status, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency In the formula, is the j-th dimensional position of the i-th secretarybird, where i = 1, 2, ..., n, n is the secretarybird population size, and u j , l j are the upper and lower limits of the j-th dimensional search space, respectively; Step 2: Multi-objective fitness evaluation and Pareto ranking, calculate the individual fitness, and construct the Pareto front, specifically as follows: Step 2.1: For each individual O i , calculate 7 sub-goal values through the real-time data of the sensor: O i = αH i + βF i + γM i - δN i - εU i + ηX i - σZ i - θ(M i ·U i ) Step 2.2: Use fast non-dominated sorting to divide the population into multiple front levels, and preferentially retain the undominated solutions; Step 2.3: For the solutions in the same front level, calculate their crowding distance in the objective space, and retain the individuals with sparse distribution: wherein, CrewdingDistance(O i ) is the crowding degree of the i-th solution, G i,k is the value of the i-th solution on the k-th sub-objective function, G i+1,k , G i-1,k are the values of the adjacent solutions of the i-th solution on the k-th sub-objective function, G k,max , G k,min are the maximum and minimum values of the k-th sub-objective function among all solutions, respectively; Step 3: Achieve the balance between global search and local development by fusing differential evolution and Levy flight strategies; Step 3.1: Design a hybrid update strategy to update the position by combining differential evolution and Levy flight: In the formula, is the optimized comprehensive performance index, O i is the current comprehensive performance index, o best is the historical optimal comprehensive performance index, o random1 and o random2 are the random candidate solutions in the search stage, Levy(β)~u=t -1-β , where β∈[1,3] is the random step size of Levy flight, is the dynamic balance factor, which gradually biases towards global search as the iteration number t increases; Step 3.2: Apply dynamic constraint conditions to conflicting objectives. When the energy efficiency is high, the inconsistency is restricted, specifically: Z ≤ 5% if X ≥ 90% Step 4: Select the comprehensive optimal solution from the Pareto front, that is, obtain the maximized comprehensive performance index and its corresponding health state, state of charge, dynamic internal resistance, maximum temperature difference, expansion stress, energy efficiency, and inconsistency degree.

5. The adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 4, characterized in that, When using the Secretary Bird Optimization Algorithm (SBOA) to optimize the battery cell management system, feedback is also provided through the cloud-edge collaborative energy storage module. The cloud platform module collects battery cell data in real time, dynamically updates the constraint thresholds and weight coefficients, and performs dynamic weight adjustment and parameter coupling compensation.

6. The adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 5, wherein The dynamic weight adjustment and parameter coupling compensation are specifically as follows: 1) Weight self-adaptation rule: Health priority principle: When H < 80%, increase α to 1.5α0, decrease β, and reduce the SOC optimization weight. Temperature difference emergency response principle: If N ≥ 45°C, immediately increase δ to 2δ0 and trigger the forced cooling strategy. 2) Coupling parameter compensation term: For the strongly correlated dynamic internal resistance M and expansion stress U, add a cross term to the objective function: O comp = γM + εU + θ(M · U) where O comp is the comprehensive performance index after adding the cross term, and θ is fitted according to historical data.

7. An adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 4, characterized in that, The process of selecting the comprehensive optimal solution from the Pareto front in Step 4 is specifically as follows: 1) Entropy weight TOPSIS decision-making: Calculate the target weights ω based on the commercial rights law k , and select the individual closest to the ideal solution: where p i,k is the value after normalizing the objective function, E k is the information quotient of the k-th sub-objective function, ω k is the weight of the k-th sub-objective function, 2) Edge-side security verification, sending the candidate solution O best to the controller of the energy storage cabinet, simulating operation and verifying that: the maximum temperature difference N ≤ 5°C, and the expansion stress U ≤ U safe ; 3) Dynamic iterative mechanism. If the verification of the above two fails, return to step 3 and narrow the solution space range; if successful, output O best as the optimal operation plan.

8. An adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 1, characterized in that, A faulty battery cell isolation module is also provided to locate the faulty battery cell according to the inconsistency degree Z. If Z > 10%, trigger the replacement protocol:

9. An adaptive energy management system based on a cloud-edge collaborative energy storage cabinet according to claim 4, wherein, Upload the current optimization data of the Secretary Bird Optimization Algorithm (SBOA) to the cloud, and update the Levy flight parameter β of the SBOA algorithm for further optimization and update at the next moment. The current optimization data includes the optimal solution O best and the constraint violation record.

Images