An automatic balancing management system for battery packs in a power exchange cabinet

Abstract

The present invention belongs to the technical field of automatic balancing of battery packs. The present invention discloses an automatic balancing management system for battery packs in a battery swap cabinet, including a data acquisition module for collecting battery data; a data processing module for preprocessing battery data to obtain effective battery data; an influence model module for obtaining a current temperature influence coefficient based on a constructed current loss temperature influence model; an automatic balancing module for performing energy calculations based on effective battery data to obtain electric energy demand, performing energy domain division based on the electric energy demand to obtain an energy domain, constructing a battery undirected graph based on the energy domain and the current temperature influence coefficient, and performing battery balancing management based on the battery undirected graph; the overall energy efficiency of the battery swap cabinet is greatly improved, and the battery life is extended.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic balancing of battery packs, and more specifically, to an automatic balancing management system for battery packs in a power exchange cabinet. Background Art

[0002] The patent with application publication number CN116714474A discloses an intelligent monitoring and management system for battery swap cabinet operation and a battery swap cabinet. On the one hand, the battery swap operation safety factor of the target battery swap cabinet is comprehensively analyzed from the two aspects of the charging safety and discharging safety of the target battery swap cabinet. On the other hand, the risk control matching index, risk control benefit index and risk control risk index of the fan assembly of the target battery swap cabinet within the set monitoring period are combined to analyze the risk control adjustment safety factor of the target battery swap cabinet. The battery swap operation safety factor of the target battery swap cabinet is then combined with the risk control adjustment safety factor to comprehensively evaluate the operation safety factor of the target battery swap cabinet, and safety warnings are issued based on this, effectively achieving the accuracy and timeliness of the target battery swap cabinet operation safety assessment, thereby avoiding the occurrence of safety accidents in the target battery swap cabinet, and further improving and popularizing the current intelligent monitoring and management system for battery swap cabinets, thereby bringing more convenience and safety guarantees to the electric vehicle industry.

[0003] The existing automatic balancing technology balances the energy distribution in the battery pack by adjusting the current and voltage between batteries. In large-scale battery pack systems, the balancing process is usually slow and cannot respond quickly to real-time changes. Especially in high-frequency charging and discharging applications, the balancing efficiency will be reduced. The battery packs in the battery swap cabinet are frequently charged and discharged, which can easily accelerate battery aging, causing their capacity and performance to gradually decrease, thereby affecting the availability and efficiency of the entire system.

[0004] In view of this, the present invention proposes an automatic balancing management system for battery packs in a power exchange cabinet to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: an automatic balancing management system for battery packs in a power exchange cabinet, comprising: a data acquisition module: collecting battery data;

[0006] Data processing module: pre-process the battery data to obtain valid battery data;

[0007] Impact model module: obtain the current temperature impact coefficient based on the constructed current loss temperature impact model;

[0008] Automatic balancing module: Perform energy calculation based on effective battery data to obtain power demand, divide energy domain based on power demand to obtain energy domain, build a battery undirected graph based on energy domain and current-temperature influence coefficient, and balance the battery based on the battery undirected graph.

[0009] Furthermore, the battery data includes: total energy demand, battery voltage, battery current, battery temperature, charging time, battery health status and battery cell internal resistance.

[0010] Furthermore, the method of preprocessing the battery data includes:

[0011] The battery data is normalized by the Min-Max normalization algorithm to obtain normalized battery data, the normalized battery data is vectorized to obtain battery vector data, and global density calculation is performed on each data in the battery vector data. The formula for global density calculation is:

[0012] Among them, Global a represents the global density of the ath battery vector data, a represents the index of the battery vector data, h represents the bandwidth parameter, u represents the total amount of battery vector data, Avg represents the average value of the module of the battery vector data, and x a represents the ath battery vector data, λ represents the adjustment parameter, x w represents the w-th battery vector data, where w represents the index of the battery vector data;

[0013] The global profile of each data in the battery vector data is calculated based on the global density. The calculation formula of the global profile is: Among them, P a Represents the global profile of the ath battery vector data;

[0014] Preset neighborhood radius R, for battery vector data x a , use the Euclidean distance formula to calculate the battery vector data except x a Other battery vector data and x a The distance value of all other battery vector data whose distance value is less than or equal to R constitutes x a The neighborhood set N a ;

[0015] Based on the neighborhood set, local density calculation is performed on each data in the battery vector data. The formula for local density calculation is:

[0016] Among them, Local a Represents the local density of the a-th battery vector data, N a represents the neighborhood set of the a-th battery vector data, A represents the battery vector data in the neighborhood set of the a-th battery vector data, and z represents the total amount of data in the neighborhood set of the a-th battery vector data;

[0017] The local profile of each data in the battery vector data is calculated based on the local density. The calculation formula of the local profile is: Among them, L a represents the local cross section of the ath battery vector data;

[0018] The anomaly score of each data in the battery vector data is calculated based on the global profile and the local profile. The calculation formula of the anomaly score is: Among them, Grade a Represents the anomaly score of the a-th battery vector data; preset anomaly score threshold Grade level , and set Grade a ≥Grade level The battery data corresponding to the battery vector data is marked as abnormal data, and all battery data marked as abnormal data in the battery data are eliminated to obtain valid battery data.

[0019] Furthermore, the current loss temperature impact model is constructed in the following manner:

[0020] Collect historical battery group energy data of group B. The historical battery group energy data includes: historical first group battery temperature, historical first group battery starting current, historical second group battery temperature and historical second group battery receiving current. The current loss is calculated by the historical first group battery starting current and the historical second group battery receiving current. The formula for calculating the current loss is: Among them, Loss represents the historical current temperature influence coefficient, Q1 represents the historical starting current of the first group of batteries, and Q2 represents the historical receiving current of the second group of batteries;

[0021] Based on the historical battery pack energy data and the historical current temperature influence coefficient, the historical first group of battery temperatures and the historical second group of battery temperatures are used as the input of the current loss temperature influence model, the predicted current temperature influence coefficient is used as the output of the current loss temperature influence model, and the historical current temperature influence coefficient is used as the prediction target of the current loss temperature influence model for model training to construct the current loss temperature influence model. The current loss temperature influence model is a linear regression model, and the target loss function of the current loss temperature influence model is: Where MSE is the mean square error, Su is the total amount of historical battery pack energy data, and F g The gth historical current temperature influence coefficient, F' g is the g-th predicted current temperature influence coefficient; taking the value of minimizing the target loss function as the training goal, the current loss temperature influence model with the minimum value of the target loss function is obtained;

[0022] Based on the constructed current loss temperature influence model, the temperature combinations of all battery temperatures in the effective battery data are calculated by recursion method. The temperature combinations are used as the input of the current loss temperature influence model for prediction to obtain the current temperature influence coefficient.

[0023] Furthermore, the method of obtaining the electric energy demand includes:

[0024] The energy demand is calculated for each battery in the valid battery data. The formula for calculating the energy demand is:

[0025]

[0026] ; Among them, E b Represents the power demand of the bth valid battery data, V b Represents the battery voltage of the bth valid battery data, I b The battery current representing the bth valid battery data, t b Represents the charging time of the bth valid battery data, SOH b Represents the battery health status of the bth valid battery data, R b The internal resistance of the battery cell representing the bth valid battery data, Tar b represents the total energy demand of the bth valid battery data, α represents the growth coefficient of the internal resistance of the battery cell, and k represents the attenuation coefficient of the battery health status.

[0027] Furthermore, the energy domain division method includes:

[0028] The energy grouping value D is preset, and D data are selected from the electric energy demand as the initial energy center based on the energy grouping value. The data interval of the initial energy center is d, and the conditions are met: Where Sum represents the total amount of effective battery data; based on the initial energy center, the energy distance from each data point in the power demand to each initial energy center is calculated, d(p,c) = |Val p -Org c |; where d(p,c) represents the energy distance between the pth power demand and the cth initial energy center, Val p represents the pth power demand, Org c represents the cth initial energy center; assign each data point in the electric energy demand to the initial energy center with the smallest energy distance to form D energy domains; calculate the mean of the electric energy demand in each energy domain and use the mean as the value of the new initial energy center; repeat until the value of the initial energy center no longer changes.

[0029] Furthermore, the method of constructing the battery undirected graph includes:

[0030] Based on the current-temperature influence coefficient and the energy domain, a graph is constructed for each energy domain. The graph construction method includes: constructing a battery undirected graph with electric energy demand as nodes, energy transmission paths as edges, and current-temperature influence coefficient as edge weights.

[0031] Furthermore, the method of performing equalization management on the battery includes:

[0032] The energy domains are sorted from large to small based on the values ​​of the initial energy centers, and the energy domains are connected in sequence using the sorting order of the energy domains as the first energy order, an energy minimization objective function is defined, and an optimal energy transmission path within the energy domain is found based on the energy minimization objective function. An electric energy demand threshold is preset, and batteries within the energy domain are charged and discharged using the optimal energy transmission path. The battery charging and discharging methods include: when the electric energy demand of the battery is less than or equal to the electric energy demand threshold, the battery is discharged, and when the electric energy demand of the battery is greater than the electric energy demand threshold, the battery is charged; when the electric energy demand of all batteries within the current energy domain is less than or equal to the electric energy demand threshold, the energy of the current energy domain is transmitted to the next energy domain according to the first energy order, and the batteries within the next energy domain are charged and discharged until the electric energy demand of the batteries within all energy domains is less than or equal to the electric energy demand threshold.

[0033] Furthermore, the formula of the energy minimization objective function is:

[0034] Where ET represents the total energy demand of the energy domain, ES i,j Represents the sum of the power demand of the ith node and the jth node in the battery undirected graph, η i,j Represents the weight of the edge connecting the i-th node and the j-th node in the battery undirected graph, Sid represents the set of edges in the battery undirected graph, i and j represent the indexes of the nodes in the battery undirected graph, and i≠j.

[0035] Furthermore, the method for obtaining the optimal energy transmission path includes:

[0036] Step S1: Initialize G ants, each ant represents an exploration of a transmission energy path, and assign each ant to a different node of the battery undirected graph; Initialize pheromone: initialize the pheromone concentration for each edge, and assign an initial pheromone concentration to each ant;

[0037] Step S2: Each ant starts from the starting node and moves in the battery undirected graph, and selects the direction of the position to be updated next based on the edge selection probability. The calculation formula of the edge selection probability is:

[0038] Among them, P m,nrepresents the probability that an ant selects an edge from the mth node in the battery undirected graph to the nth node, τ m,n Represents the pheromone concentration of the edge connecting the mth node and the nth node in the battery undirected graph. Pheromones are chemical marks left by ants on the path. The higher the concentration, the more attractive the path. m and n represent the indices of the nodes in the battery undirected graph, and m≠n. Represents the weight of the edge connecting the mth node and the nth node in the battery undirected graph, m Represents the set of other nodes that have direct edges connected to the mth node in the battery undirected graph;

[0039] Step S3: After each ant completes path selection, the total energy requirement of each ant corresponding to the selected path is calculated through the energy minimization objective function according to the nodes and edges passed in the path selected by the ant;

[0040] Step S4: Update the pheromone concentration of the edge of the selected path according to the total energy value of each ant's corresponding selected path. The formula for updating the pheromone concentration is:

[0041] Among them, τ'm,n represents the updated value of the pheromone concentration of the edge connecting the mth node and the nth node in the battery undirected graph, ρ represents the pheromone volatility coefficient, S represents the initial pheromone concentration allocated to the ants, which is used to adjust the pheromone increment, and ET represents the total energy demand;

[0042] Step S5: When the maximum number of iterations is reached or the energy minimization objective function converges, the path with the highest pheromone concentration is the optimal energy transfer path in the energy domain. When the maximum number of iterations is not reached and the energy minimization objective function does not converge, steps S2, S3 and S4 are repeated.

[0043] The technical effects and advantages of the automatic balancing management system for battery packs in a power exchange cabinet of the present invention are as follows:

[0044] The present invention eliminates abnormal battery data through the calculation of global density and local density through the system, thereby improving the safety and reliability of the system and reducing the occurrence of battery failures; optimizes the battery charging and discharging management and improves energy utilization efficiency through the current loss temperature influence model; monitors the battery health status and power demand in real time by collecting and analyzing the voltage, current, temperature and other data of the battery pack; automatically finds the optimal path for energy transmission within the battery pack through the ant colony algorithm, realizes efficient energy distribution and balanced management between batteries, extends battery life and improves the overall energy efficiency of the battery swap cabinet; realizes efficient and automatic balanced charging and discharging of the entire battery pack by gradually transmitting the energy of the batteries within the energy domain, reducing the need for manual intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of an automatic balancing management system for battery packs in a power exchange cabinet of the present invention;

[0046] Figure 2 It is a schematic diagram of an automatic balancing management method for battery packs in a power exchange cabinet of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Example 1

[0049] See also Figure 1 As shown, the present embodiment is an automatic balancing management system for battery packs in a power exchange cabinet, including: a data acquisition module: collecting battery data; a data processing module: preprocessing the battery data to obtain effective battery data; an influence model module: obtaining a current temperature influence coefficient based on a constructed current loss temperature influence model; an automatic balancing module: performing energy calculation based on effective battery data to obtain electric energy demand, dividing the energy domain based on the electric energy demand to obtain an energy domain, constructing a battery undirected graph based on the energy domain and the current temperature influence coefficient, and performing battery balancing management based on the battery undirected graph;

[0050] Each module is connected by wired and / or wireless means to achieve data transmission between modules;

[0051] Battery data includes: total energy demand, battery voltage, battery current, battery temperature, charging time, battery health status and battery cell internal resistance;

[0052] The battery health status is a percentage of the battery capacity. The internal resistance of the battery cell is the resistance encountered when the current flows through the battery, and its size affects the performance and efficiency of the battery.

[0053] The battery data is normalized by the Min-Max normalization algorithm to obtain normalized battery data, the normalized battery data is vectorized to obtain battery vector data, and global density calculation is performed on each data in the battery vector data. The formula for global density calculation is:

[0054] Among them, Global arepresents the global density of the ath battery vector data, a represents the index of the battery vector data, h represents the bandwidth parameter, which is used to control the smoothness of the function, u represents the total amount of battery vector data, Avg represents the average value of the module of the battery vector data, and x a represents the ath battery vector data, λ represents the adjustment parameter used to control the frequency of the cosine function, x w represents the w-th battery vector data, where w represents the index of the battery vector data;

[0055] The global profile of each data in the battery vector data is calculated based on the global density. The calculation formula of the global profile is: Among them, P a Represents the global profile of the a-th battery vector data. The global profile is used to measure the degree of abnormality of a data point under the global data distribution. Points with lower global density will have larger profile values ​​and may be abnormal points;

[0056] Preset neighborhood radius R, for battery vector data x a , use the Euclidean distance formula to calculate the battery vector data except x a Other battery vector data and x a The distance value of all other battery vector data whose distance value is less than or equal to R constitutes x a The neighborhood set N a ;

[0057] Based on the neighborhood set, local density calculation is performed on each data in the battery vector data. The formula for local density calculation is:

[0058] Among them, Local a Represents the local density of the a-th battery vector data, N a represents the neighborhood set of the a-th battery vector data, A represents the battery vector data in the neighborhood set of the a-th battery vector data, and z represents the total amount of data in the neighborhood set of the a-th battery vector data;

[0059] The local profile of each data in the battery vector data is calculated based on the local density. The calculation formula of the local profile is: Among them, L a Represents the local profile of the a-th battery vector data. The local profile is used to measure the abnormality of the data point under the local data. Points with low local density are considered possible abnormal points.

[0060] The anomaly score of each data in the battery vector data is calculated based on the global profile and the local profile. The calculation formula of the anomaly score is: Among them, Grade aRepresents the anomaly score of the a-th battery vector data; preset anomaly score threshold Grade level , and set Grade a ≥Grade level The battery data corresponding to the battery vector data is marked as abnormal data, and all battery data marked as abnormal data in the battery data are removed to obtain valid battery data;

[0061] By calculating the global density and local density, the abnormal data points in the battery can be accurately detected, the batteries with problems can be identified, and these abnormal battery data can be removed in time to prevent bad batteries from affecting the operation of the overall system, improve the safety and reliability of the system, and reduce the occurrence of battery failures;

[0062] Collect historical battery group energy data of group B. The historical battery group energy data includes: historical first group battery temperature, historical first group battery starting current, historical second group battery temperature and historical second group battery receiving current. The current loss is calculated by the historical first group battery starting current and the historical second group battery receiving current. The formula for calculating the current loss is: Among them, Loss represents the historical current temperature influence coefficient, Q1 represents the historical starting current of the first group of batteries, and Q2 represents the historical receiving current of the second group of batteries;

[0063] Based on the historical battery pack energy data and the historical current temperature influence coefficient, the historical first group of battery temperatures and the historical second group of battery temperatures are used as the input of the current loss temperature influence model, the predicted current temperature influence coefficient is used as the output of the current loss temperature influence model, and the historical current temperature influence coefficient is used as the prediction target of the current loss temperature influence model for model training to construct the current loss temperature influence model. The current loss temperature influence model is a linear regression model, and the target loss function of the current loss temperature influence model is: Where MSE is the mean square error, Su is the total amount of historical battery pack energy data, and F g The gth historical current temperature influence coefficient, F' g is the g-th predicted current temperature influence coefficient; taking the value of minimizing the target loss function as the training goal, the current loss temperature influence model with the minimum value of the target loss function is obtained;

[0064] Based on the constructed current loss temperature influence model, the temperature combination of all battery temperatures in the effective battery data is calculated by recursion method, and the temperature combination is used as the input of the current loss temperature influence model for prediction to obtain the current temperature influence coefficient;

[0065] The current loss temperature impact model based on historical data takes into account the impact of battery temperature on energy transfer, allowing the system to accurately predict the current loss of the battery under different temperature conditions, thereby optimizing the battery charge and discharge management and improving energy utilization efficiency;

[0066] The energy demand is calculated for each battery in the valid battery data. The formula for calculating the energy demand is:

[0067]

[0068] ; Among them, E b Represents the power demand of the bth valid battery data, V b Represents the battery voltage of the bth valid battery data, I b The battery current representing the bth valid battery data, t b Represents the charging time of the bth valid battery data, SOH b Represents the battery health status of the bth valid battery data, R b The internal resistance of the battery cell representing the bth valid battery data, Tar b represents the total energy demand of the bth valid battery data, α represents the growth coefficient of the internal resistance of the battery cell, which is used to indicate the rate at which the resistance increases over time, and k represents the attenuation coefficient of the battery health state, which is used to indicate the rate at which the battery health state decays over time;

[0069] The energy grouping value D is preset, and D data are selected from the electric energy demand as the initial energy center based on the energy grouping value. The data interval of the initial energy center is d, and the conditions are met: Where Sum represents the total amount of effective battery data; based on the initial energy center, each data in the electric energy demand is divided into energy domains, and the steps of energy domain division include:

[0070] Step 1: Calculate the energy distance from each data point in the power demand to each initial energy center, d(p,c) = |Val p -Org c |; where d(p,c) represents the energy distance between the pth power demand and the cth initial energy center, Val p represents the pth power demand, Org c represents the cth initial energy center; each data point in the electric energy demand is assigned to the initial energy center with the smallest energy distance, forming D energy domains;

[0071] Step 2: Calculate the average value of the electric energy demand in each energy domain and use the average value as the value of the new initial energy center;

[0072] Step 3: Repeat steps 1 and 2 until the value of the initial energy center no longer changes;

[0073] The energy domains are sorted from large to small based on the values ​​of the initial energy centers, and the sorting order of the energy domains is used as the first energy order to sequentially connect the energy domains;

[0074] Based on the current temperature influence coefficient and the energy domain, a graph is constructed for each energy domain. The graph construction method includes: constructing a battery undirected graph with the electric energy demand as a node, the energy transmission path as an edge, and the current temperature influence coefficient as the edge weight;

[0075] Define the energy minimization objective function Where ET represents the total energy demand of the energy domain, ES i,j Represents the sum of the power demand of the ith node and the jth node in the battery undirected graph, η i,j represents the weight of the edge connecting the i-th node and the j-th node in the battery undirected graph, Sid represents the set of edges in the battery undirected graph, i and j represent the index of the node in the battery undirected graph, and i≠j;

[0076] Finding the optimal energy transfer path in the energy domain based on the energy minimization objective function, the steps of finding the optimal energy transfer path in the energy domain include:

[0077] Step S1: Initialize G ants, each ant represents an exploration of a transmission energy path, and assign each ant to a different node of the battery undirected graph; Initialize pheromone: initialize the pheromone concentration for each edge, and assign an initial pheromone concentration to each ant;

[0078] Step S2: Each ant starts from the starting node and moves in the battery undirected graph, and selects the direction of the position to be updated next based on the edge selection probability. The calculation formula of the edge selection probability is:

[0079] Among them, P m,n represents the probability that an ant selects an edge from the mth node in the battery undirected graph to the nth node, τ m,n Represents the pheromone concentration of the edge connecting the mth node and the nth node in the battery undirected graph. Pheromones are chemical marks left by ants on the path. The higher the concentration, the more attractive the path. m and n represent the indices of the nodes in the battery undirected graph, and m≠n. Represents the weight of the edge connecting the mth node and the nth node in the battery undirected graph, m Represents the set of other nodes that have direct edges connected to the mth node in the battery undirected graph;

[0080] Step S3: After each ant completes path selection, the total energy requirement of each ant corresponding to the selected path is calculated through the energy minimization objective function according to the nodes and edges passed in the path selected by the ant;

[0081] Step S4: Update the pheromone concentration of the edge of the selected path according to the total energy value of each ant's corresponding selected path. The formula for updating the pheromone concentration is:

[0082] Among them, τ'm,n represents the updated value of the pheromone concentration of the edge connecting the mth node and the nth node in the battery undirected graph, ρ represents the pheromone volatility coefficient, which is used to simulate the natural volatilization of pheromones, S represents the initial pheromone concentration allocated to ants, which is used to adjust the pheromone increment, and ET represents the total energy demand;

[0083] Step S5: When the maximum number of iterations is reached or the energy minimization objective function converges, the path with the highest pheromone concentration is the optimal energy transmission path in the energy domain. When the maximum number of iterations is not reached and the energy minimization objective function does not converge, steps S2, S3 and S4 are repeated;

[0084] By collecting and analyzing the voltage, current, temperature and other data of the battery pack, the system can monitor the health status and power demand of the battery in real time; based on the ant colony algorithm with the energy minimization objective function, the system can automatically find the optimal path for energy transmission within the battery pack and achieve efficient energy distribution and balanced management between batteries, thereby extending the battery life and improving the overall energy efficiency of the battery swap cabinet;

[0085] The battery is balanced based on the optimal energy transfer path within the energy domain. The balanced management methods include:

[0086] A power demand threshold is preset, and the batteries in the energy domain are charged and discharged using the optimal energy transmission path. The battery charging and discharging methods include: when the power demand of the battery is less than or equal to the power demand threshold, the battery is discharged, that is, the energy of the battery is transferred to the next battery according to the optimal energy transmission path, and when the power demand of the battery is greater than the power demand threshold, the battery is charged; when the power demand of all batteries in the current energy domain is less than or equal to the power demand threshold, the energy of the current energy domain is transferred to the next energy domain according to the first energy sequence, and the battery is charged and discharged in the next energy domain; through the sorting in the energy domain and the optimal energy transmission path, the energy of the batteries in the energy domain is gradually transmitted, so as to realize the efficient and automatic balanced charging and discharging of the entire battery pack, and reduce the need for manual intervention;

[0087] This embodiment eliminates abnormal battery data through the calculation of global density and local density through the system, thereby improving the safety and reliability of the system and reducing the occurrence of battery failures; optimizes the battery charging and discharging management and improves energy utilization efficiency through the current loss temperature influence model; monitors the battery health status and power demand in real time by collecting and analyzing the voltage, current, temperature and other data of the battery pack; automatically finds the optimal path for energy transmission within the battery pack through the ant colony algorithm, realizes efficient energy distribution and balanced management between batteries, extends battery life and improves the overall energy efficiency of the battery swap cabinet; realizes efficient and automated balanced charging and discharging of the entire battery pack by gradually transmitting the energy of the batteries within the energy domain, reducing the need for manual intervention.

[0088] Example 2

[0089] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, and a method for automatic balancing management of battery packs in a power exchange cabinet is provided, including:

[0090] A1. Collect battery data;

[0091] A2. Preprocess the battery data to obtain valid battery data;

[0092] A3. Obtain the current temperature influence coefficient based on the constructed current loss temperature influence model;

[0093] A4. Perform energy calculation based on effective battery data to obtain electric energy demand, divide the energy domain based on the electric energy demand to obtain the energy domain, construct a battery undirected graph based on the energy domain and the current-temperature influence coefficient, and balance the battery based on the battery undirected graph.

[0094] Example 3

[0095] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the above-mentioned method for automatic balancing management of battery packs in a battery swap cabinet is implemented.

[0096] Since the electronic device introduced in this embodiment is an electronic device used to implement a method for automatic balancing management of battery packs in a battery swap cabinet in the embodiment of the present application, based on the method for automatic balancing management of battery packs in a battery swap cabinet introduced in the embodiment of the present application, the technical personnel of the field can understand the specific implementation of the electronic device of the present embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application will not be described in detail here. As long as the technical personnel of the field implement the electronic device used in the method for automatic balancing management of battery packs in a battery swap cabinet in the embodiment of the present application, it belongs to the scope of protection of this application.

[0097] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0098] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. An automatic balancing management system for battery packs in a power exchange cabinet, characterized in that: include: Data acquisition module: collect battery data; Data processing module: pre-process the battery data to obtain valid battery data; Impact model module: obtain the current temperature impact coefficient based on the constructed current loss temperature impact model; Automatic balancing module: performs energy calculation based on effective battery data to obtain power demand, divides energy domains based on power demand to obtain energy domains, constructs a battery undirected graph based on energy domains and current-temperature influence coefficients, and performs battery balancing management based on the battery undirected graph; The method of constructing the battery undirected graph includes: Based on the current temperature influence coefficient and the energy domain, a graph is constructed for each energy domain. The graph construction method includes: constructing a battery undirected graph with the electric energy demand as a node, the energy transmission path as an edge, and the current temperature influence coefficient as the edge weight; The method of performing battery balancing management includes: The energy domains are sorted from large to small based on the values ​​of the initial energy centers, and the energy domains are connected in sequence using the sorting order of the energy domains as the first energy order, an energy minimization objective function is defined, and an optimal energy transmission path within the energy domain is found based on the energy minimization objective function. An electric energy demand threshold is preset, and batteries within the energy domain are charged and discharged using the optimal energy transmission path. The battery charging and discharging methods include: when the electric energy demand of the battery is less than or equal to the electric energy demand threshold, the battery is discharged, and when the electric energy demand of the battery is greater than the electric energy demand threshold, the battery is charged; when the electric energy demand of all batteries within the current energy domain is less than or equal to the electric energy demand threshold, the energy of the current energy domain is transmitted to the next energy domain according to the first energy order, and the batteries within the next energy domain are charged and discharged until the electric energy demand of the batteries within all energy domains is less than or equal to the electric energy demand threshold.

2. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 1 is characterized in that: The battery data includes: total energy demand, battery voltage, battery current, battery temperature, charging time, battery health status and battery cell internal resistance.

3. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 2 is characterized in that: The method of preprocessing the battery data includes: The battery data is normalized by the Min-Max normalization algorithm to obtain normalized battery data, and the normalized battery data is vectorized to obtain battery vector data. The global density calculation is performed on each data in the battery vector data. The formula for the global density calculation is: Among them, Global a represents the global density of the ath battery vector data, a represents the index of the battery vector data, h represents the bandwidth parameter, u represents the total amount of battery vector data, Avg represents the average value of the module of the battery vector data, and x a represents the ath battery vector data, λ represents the adjustment parameter, x w represents the w-th battery vector data, where w represents the index of the battery vector data; The global profile of each data in the battery vector data is calculated based on the global density. The calculation formula of the global profile is: Among them, P a Represents the global profile of the ath battery vector data; Preset neighborhood radius R, for battery vector data x a , use the Euclidean distance formula to calculate the battery vector data except x a Other battery vector data and x a The distance value of all other battery vector data whose distance value is less than or equal to R constitutes x a The neighborhood set N a ; Based on the neighborhood set, local density calculation is performed on each data in the battery vector data. The formula for local density calculation is: Among them, Local a Represents the local density of the a-th battery vector data, N a represents the neighborhood set of the a-th battery vector data, A represents the battery vector data in the neighborhood set of the a-th battery vector data, and z represents the total amount of data in the neighborhood set of the a-th battery vector data; The local profile of each data in the battery vector data is calculated based on the local density. The calculation formula of the local profile is: Among them, L a represents the local cross section of the ath battery vector data; The anomaly score of each data in the battery vector data is calculated based on the global profile and the local profile. The calculation formula of the anomaly score is: Among them, Grade a Represents the anomaly score of the a-th battery vector data; preset anomaly score threshold Grade level , and set Grade a ≥Grade level The battery data corresponding to the battery vector data is marked as abnormal data, and all battery data marked as abnormal data in the battery data are eliminated to obtain valid battery data.

4. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 3 is characterized in that: The current loss temperature influence model is constructed in the following manner: Collect historical battery group energy data of group B. The historical battery group energy data includes: historical first group battery temperature, historical first group battery starting current, historical second group battery temperature and historical second group battery receiving current. The current loss is calculated by the historical first group battery starting current and the historical second group battery receiving current. The formula for calculating the current loss is: Among them, Loss represents the historical current temperature influence coefficient, Q1 represents the historical starting current of the first group of batteries, and Q2 represents the historical receiving current of the second group of batteries; Based on the historical battery pack energy data and the historical current temperature influence coefficient, the historical first group of battery temperatures and the historical second group of battery temperatures are used as the input of the current loss temperature influence model, the predicted current temperature influence coefficient is used as the output of the current loss temperature influence model, and the historical current temperature influence coefficient is used as the prediction target of the current loss temperature influence model for model training to construct the current loss temperature influence model. The current loss temperature influence model is a linear regression model, and the target loss function of the current loss temperature influence model is: Where MSE is the mean square error, Su is the total amount of historical battery pack energy data, and F g The gth historical current temperature influence coefficient, F' g is the g-th predicted current temperature influence coefficient; taking the value of minimizing the target loss function as the training goal, the current loss temperature influence model with the minimum value of the target loss function is obtained; Based on the constructed current loss temperature influence model, the temperature combinations of all battery temperatures in the effective battery data are calculated by recursion method. The temperature combinations are used as the input of the current loss temperature influence model for prediction to obtain the current temperature influence coefficient.

5. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 4 is characterized in that: The method of obtaining the electric energy demand includes: The energy demand is calculated for each battery in the valid battery data. The formula for calculating the energy demand is: ； Among them, E b Represents the power demand of the bth valid battery data, V b Represents the battery voltage of the bth valid battery data, I b The battery current representing the bth valid battery data, t b Represents the charging time of the bth valid battery data, SOH b Represents the battery health status of the bth valid battery data, R b The internal resistance of the battery cell representing the bth valid battery data, Tar b represents the total energy demand of the bth valid battery data, α represents the growth coefficient of the internal resistance of the battery cell, and k represents the attenuation coefficient of the battery health status.

6. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 5 is characterized in that: The energy domain division method includes: The energy grouping value D is preset, and D data are selected from the electric energy demand as the initial energy center based on the energy grouping value. The data interval of the initial energy center is d, and the conditions are met: Where Sum represents the total amount of effective battery data; based on the initial energy center, the energy distance from each data point in the power demand to each initial energy center is calculated, d(p,c) = |Val p -Org c |; where d(p,c) represents the energy distance between the pth power demand and the cth initial energy center, Val p represents the pth power demand, Org c represents the cth initial energy center; assign each data point in the electric energy demand to the initial energy center with the smallest energy distance to form D energy domains; calculate the mean of the electric energy demand in each energy domain and use the mean as the value of the new initial energy center; repeat until the value of the initial energy center no longer changes.

7. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 6 is characterized in that: The formula of the energy minimization objective function is: Where ET represents the total energy demand of the energy domain, ES i,j Represents the sum of the power demand of the ith node and the jth node in the battery undirected graph, η i,j Represents the weight of the edge connecting the i-th node and the j-th node in the battery undirected graph, Sid represents the set of edges in the battery undirected graph, i and j represent the indexes of the nodes in the battery undirected graph, and i≠j.

8. The automatic balancing management system for battery packs in a power exchange cabinet according to claim 7 is characterized in that: The method for obtaining the optimal energy transmission path includes: Step S1: Initialize G ants, each ant represents an exploration of a transmission energy path, and assign each ant to a different node of the battery undirected graph; Initialize pheromone: initialize the pheromone concentration for each edge, and assign an initial pheromone concentration to each ant; Step S2: Each ant starts from the starting node and moves in the battery undirected graph, and selects the direction of the position to be updated next based on the edge selection probability. The calculation formula of the edge selection probability is: Among them, P m,n represents the probability that an ant selects an edge from the mth node in the battery undirected graph to the nth node, τ m,n Represents the pheromone concentration of the edge connecting the mth node and the nth node in the battery undirected graph. Pheromones are chemical marks left by ants on the path. The higher the concentration, the more attractive the path. m and n represent the indices of the nodes in the battery undirected graph, and m≠n. Represents the weight of the edge connecting the mth node and the nth node in the battery undirected graph, m Represents the set of other nodes that have direct edges connected to the mth node in the battery undirected graph; Step S3: After each ant completes path selection, the total energy requirement of each ant corresponding to the selected path is calculated through the energy minimization objective function according to the nodes and edges passed in the path selected by the ant; Step S4: Update the pheromone concentration of the edge of the selected path according to the total energy value of each ant's corresponding selected path. The formula for updating the pheromone concentration is: Among them, τ' m,n represents the updated value of the pheromone concentration of the edge connecting the mth node and the nth node in the battery undirected graph, ρ represents the pheromone volatility coefficient, S represents the initial pheromone concentration allocated to the ants, which is used to adjust the pheromone increment, and ET represents the total energy demand; Step S5: When the maximum number of iterations is reached or the energy minimization objective function converges, the path with the highest pheromone concentration is the optimal energy transfer path in the energy domain. When the maximum number of iterations is not reached and the energy minimization objective function does not converge, steps S2, S3 and S4 are repeated.

Images