Reconfigurable battery energy storage energy optimization method and system for realizing SOC balance

Abstract

The invention discloses a reconfigurable battery energy storage energy optimization method and system for realizing SOC equalization, and the method comprises the following steps: collecting the voltage U = [u1, u2,..., un], the current I = [i1, i2,..., in] and the SOC value SOC = [soc 1, soc 2,..., soc n] of n batteries, and randomly initializing the binary state vectors of the batteries; constraining the binary state vector of the battery by using population constraint; solving an optimal solution of the binary state vector by using a particle swarm optimization algorithm in combination with an optimization objective function and a constraint condition; and controlling the working state of the battery based on the optimal solution. According to the method, in the energy optimization process of the reconfigurable battery system, the population constraint is introduced into the optimization algorithm according to the SOC of the battery, so that SOC balance of the battery pack is realized, the inconsistency of the battery is reduced, and the energy utilization rate of the battery pack is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery energy storage, and in particular relates to a reconfigurable battery energy storage energy optimization method and system for achieving SOC balance. Background Art

[0002] Battery energy storage systems can be used as both "sources" and "loads". They can discharge when wind and solar power generation is insufficient, store electricity when wind and solar power generation is excessive, and can also be used to smooth peaks and valleys and output, playing an important role in maintaining the stability of the power system. However, the inconsistency of batteries has brought huge difficulties to the energy utilization of battery energy storage systems, and the available battery pack capacity depends on the weakest battery. Once any battery reaches the defined threshold, the entire system will be cut off, and the battery utilization efficiency is low. The OCV-SOC curve also shows that the inconsistency of SOC will cause the inconsistency of battery terminal voltage, which will further accumulate the voltage difference of the reconfigurable battery path when it is connected in series, and generate circulation when the path is connected in parallel, resulting in varying degrees of battery aging, further exacerbating the inconsistency of the battery.

[0003] Reconfigurable batteries have high flexibility and can make the battery pack SOC tend to be consistent by bypassing any unbalanced single cells. By selecting appropriate batteries for series and parallel connection, the voltage of the parallel branches can be reduced, the circulation of the parallel branches can be suppressed, and the overall performance of the reconfigurable battery pack can be improved. Select appropriate batteries for series and parallel connection, reduce the difference in SOC between batteries, connect as many branches in parallel as possible, and satisfy the path voltage constraints and the single cell voltage and current constraints at the same time. Therefore, the optimal path combination of reconfigurable batteries can be regarded as a non-deterministic polynomial problem, and the selection of the optimal battery can be solved by a multi-objective optimization algorithm. Summary of the invention

[0004] The present invention aims to solve the deficiencies of the prior art and provides the following solutions:

[0005] A reconfigurable battery energy storage energy optimization method for achieving SOC balancing includes the following steps:

[0006] Collect the voltage of n batteries U=[u1,u2,...,u n ], current I=[i1,i2,...,i n ] and SOC value SOC=[soc1,soc2,...,soc n ] and randomly initialize the binary state vector of the battery;

[0007] constraining the binary state vector of the battery using a population constraint;

[0008] In combination with the optimization objective function and the constraint conditions, a particle swarm optimization algorithm is used to solve the optimal solution of the binary state vector;

[0009] The operating state of the battery is controlled based on the optimal solution.

[0010] Preferably, the binary state vector is X=[x1, x2, ..., x n ],have:

[0011]

[0012] Wherein, n represents the number of batteries.

[0013] Preferably, the population constraints include:

[0014] When the battery pack is discharged, x max =1,x min =0;

[0015] When the battery pack is charging, x max =0,x min =1;

[0016] Among them, x max Indicates the working state of the battery when the SOC in the battery pack is the highest, x min Indicates the working status of the battery when the SOC in the battery pack is lowest.

[0017] Preferably, the optimization objective function includes:

[0018] min F=ω1f1+ω2f2

[0019]

[0020]

[0021] Wherein, f1 represents the first sub-objective function, ω1 represents the weight coefficient of the first sub-objective function, f2 represents the second sub-objective function, ω2 represents the weight coefficient of the second sub-objective function, n represents the number of batteries, P represents the number of parallel branches, p represents the pth parallel branch, and U p Indicates the voltage of each parallel branch, U req Indicates the required voltage;

[0022] The constraints include:

[0023] U min ≤U p ≤U max

[0024] u off ≤u j ≤u max

[0025] i disc≤i j ≤i c

[0026] Among them, U max and U min Indicates the upper and lower limits of the battery parallel branch voltage, u max and u off Indicates the upper and lower limits of the working battery voltage, i c and i disc Indicates the upper and lower limits of the current during charging and discharging.

[0027] Preferably, the method for solving the optimal solution includes:

[0028] Step 1: Initialize the binary state vector X and the algorithm parameters of the particle swarm optimization algorithm;

[0029] Step 2: Determine the difference between the maximum SOC value and the minimum SOC value in the battery pack and the balance threshold value. When the difference is greater than the balance threshold value, use the population constraint to re-constrain the binary state vector X. When the difference is less than the balance threshold value, do not constrain the binary state vector X.

[0030] Step 3: Calculate the initial individual optimal value and the global optimal fitness value of the objective function according to the binary state vector X, and enter the iteration to update the inertia weight and speed of the particles in the population:

[0031] ω=ω max -(ω max -ω min )*k / G

[0032]

[0033] Among them, ω represents the updated inertia weight, ω max and ω min represents the upper and lower limits of the inertia weight, k represents the current number of iterations, G represents the number of iterations, represents the velocity of particle j in d-dimensional space, c1 and c2 represent acceleration factors, is the optimal position of individual particles, represents the position of particle j in d-dimensional space, represents the global optimal position of the particle, and rand() represents a random number between [0,1];

[0034] Step 4: Map the updated particle velocity and binarize the mapped particles using random numbers to obtain the updated particle position: have when Sometimes, there is

[0035] Step 5: Determine whether the updated particle position requires the population constraint, and if necessary, re-constrain it;

[0036] Step 6: If not, determine whether the updated particle position meets the constraint conditions of the objective function. If so, update the global optimal fitness value corresponding to the particle. If not, set the objective function F=F+100 corresponding to the particle that does not meet the condition, and then update the global optimal fitness value corresponding to the particle.

[0037] Step 7: Determine whether the iteration is finished. If not, return to step 3. If the iteration is finished, output the particle population corresponding to the global optimal fitness value to obtain the optimal solution of the binary state vector.

[0038] Preferably, the method for controlling the working state of the battery based on the optimal solution includes: connecting the batteries whose binary state vector is 1 in series in sequence to form a battery string that meets the required voltage, then connecting each of the battery strings in parallel, and connecting the batteries whose binary state vector is 0 to a bypass.

[0039] The present invention also provides a reconfigurable battery energy storage energy optimization system for achieving SOC balancing, wherein the optimization system applies any of the above-mentioned optimization methods, including: a state vector initialization module, a population constraint module, an optimal solution solving module and a battery control module;

[0040] The state vector initialization module is used to collect the voltages of n batteries U = [u1, u2, ..., u n ], current I = [i1, i2, ..., in] and SOC value SOC = [soc1, soc2, ..., soc n ] and randomly initialize the binary state vector of the battery;

[0041] The population constraint module constrains the binary state vector of the battery using population constraints;

[0042] The optimal solution solving module is used to combine the optimization objective function and the constraint conditions to solve the optimal solution of the binary state vector using a particle swarm optimization algorithm;

[0043] The battery control module controls the operating state of the battery based on the optimal solution.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] In the process of energy optimization of the reconfigurable battery system, the present invention introduces population constraints to the optimization algorithm according to the SOC of the battery to achieve SOC balance of the battery pack, reduce battery inconsistency, and improve the energy utilization rate of the battery pack. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0047] Figure 1 A schematic diagram of a method flow chart of an embodiment of the present invention;

[0048] Figure 2 This is a flow chart of using a particle swarm optimization algorithm to find the optimal solution in an embodiment of the present invention;

[0049] Figure 3 A reconfigurable battery topology used in the algorithm optimization of an embodiment of the present invention;

[0050] Figure 4 The operating conditions used when the reconfigurable battery system according to the embodiment of the present invention works;

[0051] Figure 5 The SOC variation curve of the battery when the population constraint is not used in the optimization process of the embodiment of the present invention;

[0052] Figure 6 This is a battery SOC variation curve when population constraints are used in the optimization process of an embodiment of the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0055] Embodiment 1

[0056] In this embodiment, if Figure 1 As shown, a reconfigurable battery energy storage energy optimization method for achieving SOC balancing includes the following steps:

[0057] S1. Collect the voltage of n batteries U = [u1, u2, ..., u n ], current I=[i1,i2,...,i n ] and SOC value SOC=[soc1,soc2,...,soc n ] and randomly initialize the binary state vector of the battery.

[0058] The binary state vector is X = [x1, x2, ..., x n ],have:

[0059]

[0060] Wherein, n represents the number of batteries.

[0061] S2. Use population constraints to constrain the binary state vector of the battery.

[0062] The population constraints include: when the battery pack is discharged, x max =1,x min =0; when the battery pack is charging, x max =0,x min =1; where x max Indicates the working state of the battery when the SOC in the battery pack is the highest, x min Indicates the working status of the battery when the SOC in the battery pack is lowest.

[0063] S3. Combining the optimization objective function and constraints, the particle swarm optimization algorithm is used to solve the optimal solution of the binary state vector.

[0064] The optimization objective function includes:

[0065] min F=ω1f1+ω2f2

[0066]

[0067]

[0068] Wherein, f1 represents the first sub-objective function, ω1 represents the weight coefficient of the first sub-objective function, f2 represents the second sub-objective function, ω2 represents the weight coefficient of the second sub-objective function, n represents the number of batteries, P represents the number of parallel branches, p represents the pth parallel branch, and U p Indicates the voltage of each parallel branch, U reqRepresents the required voltage; in this embodiment, the sub-objective function f1 is designed to be the inverse of the sum of the binary state vectors of the battery in order to allow more batteries to participate in the work and fully utilize the energy of the battery pack; the sub-objective function f2 is designed to be the absolute value sum of the difference between the voltage of each parallel branch of the battery pack and the required voltage in order to reduce the circulating current of the parallel branch of the battery pack.

[0069] The constraints include:

[0070] U min ≤U p ≤U max

[0071] u off ≤u j ≤u max

[0072] i disc ≤i j ≤i c

[0073] Among them, U max and U min Indicates the upper and lower limits of the battery parallel branch voltage, u max and u off Indicates the upper and lower limits of the working battery voltage, i c and i disc Indicates the upper and lower limits of the current during charging and discharging.

[0074] like Figure 2 As shown, the method for finding the optimal solution includes:

[0075] Step 1: Initialize the binary state vector X and the algorithm parameters of the particle swarm optimization algorithm. In this embodiment, the algorithm parameters include: number of iterations, particle velocity, acceleration factor and other parameters;

[0076] Step 2: Determine the difference between the maximum SOC and the minimum SOC in the battery pack max -SOC min When the difference is greater than the equilibrium threshold e, the binary state vector X is re-constrained using the population constraint. When the difference is less than the equilibrium threshold e, the binary state vector X is not constrained.

[0077] Step 3: Calculate the initial individual optimal value and global optimal fitness value of the objective function according to the binary state vector X, and enter the iteration to update the inertia weight and speed of the particles in the population:

[0078] ω=ω max -(ω max -ω min )*k / G

[0079]

[0080] Among them, ω represents the updated inertia weight, ω max and ω min represents the upper and lower limits of the inertia weight, k represents the current number of iterations, G represents the number of iterations, represents the velocity of particle j in d-dimensional space, c1 and c2 represent acceleration factors, is the optimal position of individual particles, represents the position of particle j in d-dimensional space, represents the global optimal position of the particle, and rand() represents a random number between [0,1];

[0081] Step 4: Map the updated particle velocities:

[0082]

[0083] And the mapped particles are binarized with random numbers to obtain the updated particle positions: Sometimes, there is when Sometimes, there is

[0084] Step 5: Determine whether the updated particle position needs population constraints, and if necessary, re-constrain it;

[0085] Step 6: If not, determine whether the updated particle position meets the constraint conditions of the objective function. If so, update the global optimal fitness value corresponding to the particle. If not, set the objective function F=F+100 corresponding to the particle that does not meet the conditions, and then update the global optimal fitness value corresponding to the particle.

[0086] Step 7: Determine whether the iteration is finished. If not, return to step 3. If the iteration is finished, output the particle population corresponding to the global optimal fitness value and obtain the optimal solution of the binary state vector.

[0087] S4. Control the working state of the battery based on the optimal solution.

[0088] The method for controlling the working state of a battery based on an optimal solution includes: connecting batteries whose binary state vector is 1 in series in sequence to form a battery string that meets the required voltage, then connecting the battery strings in parallel, and connecting batteries whose binary state vector is 0 to a bypass.

[0089] Embodiment 2

[0090] In this embodiment, based on MATLAB / Simulink, refer to Figure 3The reconfigurable battery topology shown is used to build a reconfigurable battery system consisting of 12 battery modules. Figure 4 The energy storage condition shown makes the battery pack work, and the optimization algorithm is run every 5 seconds. After each run, the batteries with a state vector of 1 are connected in series and then in parallel according to the required voltage, and the excess batteries are bypassed. Figure 5 , Figure 6 It can be seen that under the energy storage condition, the reconfigurable battery system energy optimization method for achieving SOC balance proposed in the present invention achieves SOC balance in the energy optimization process.

[0091] Embodiment 3

[0092] In this embodiment, a reconfigurable battery energy storage energy optimization system for achieving SOC balancing includes: a state vector initialization module, a population constraint module, an optimal solution solving module and a battery control module.

[0093] The state vector initialization module is used to collect the voltages of n batteries U = [u1, u2, ..., u n ], current I=[i1,i2,...,i n ] and SOC value SOC=[soc1,soc2,...,soc n ] and randomly initialize the binary state vector of the battery.

[0094] The binary state vector is X = [x1, x2, ..., x n ],have:

[0095]

[0096] Wherein, n represents the number of batteries.

[0097] The population constraint module uses population constraints to constrain the binary state vector of the battery.

[0098] The population constraints include: when the battery pack is discharged, x max =1,x min =0; when the battery pack is charging, x max =0,x min =1; where x max Indicates the working state of the battery when the SOC in the battery pack is the highest, x min Indicates the working status of the battery when the SOC in the battery pack is lowest.

[0099] The optimal solution solving module is used to combine the optimization objective function and constraints and use the particle swarm optimization algorithm to solve the optimal solution of the binary state vector.

[0100] The optimization objective function includes:

[0101] min F=ω1f1+ω2f2

[0102]

[0103]

[0104] Wherein, f1 represents the first sub-objective function, ω1 represents the weight coefficient of the first sub-objective function, f2 represents the second sub-objective function, ω2 represents the weight coefficient of the second sub-objective function, n represents the number of batteries, P represents the number of parallel branches, p represents the pth parallel branch, and U p Indicates the voltage of each parallel branch, U req Represents the required voltage; in this embodiment, the sub-objective function f1 is designed to be the inverse of the sum of the binary state vectors of the battery in order to allow more batteries to participate in the work and fully utilize the energy of the battery pack; the sub-objective function f2 is designed to be the absolute value sum of the difference between the voltage of each parallel branch of the battery pack and the required voltage in order to reduce the circulating current of the parallel branch of the battery pack.

[0105] The constraints include:

[0106] U min ≤U p ≤U max

[0107] u off ≤u j ≤u max

[0108] i disc ≤i j ≤i c

[0109] Among them, U max and U min Indicates the upper and lower limits of the battery parallel branch voltage, u max and u off Indicates the upper and lower limits of the working battery voltage, i c and i disc Indicates the upper and lower limits of the current during charging and discharging.

[0110] In the optimal solution solving module, the methods for solving the optimal solution include:

[0111] Step 1: Initialize the binary state vector X and the algorithm parameters of the particle swarm optimization algorithm. In this embodiment, the algorithm parameters include: number of iterations, particle velocity, acceleration factor and other parameters;

[0112] Step 2: Determine the difference between the maximum SOC and the minimum SOC in the battery pack max -SOC minWhen the difference is greater than the equilibrium threshold e, the binary state vector X is re-constrained using the population constraint. When the difference is less than the equilibrium threshold e, the binary state vector X is not constrained.

[0113] Step 3: Calculate the initial individual optimal value and global optimal fitness value of the objective function according to the binary state vector X, and enter the iteration to update the inertia weight and speed of the particles in the population:

[0114] ω=ω max -(ω max -ω min )*k / G

[0115]

[0116] Among them, ω represents the updated inertia weight, ω max and ω min represents the upper and lower limits of the inertia weight, k represents the current number of iterations, G represents the number of iterations, represents the velocity of particle j in d-dimensional space, c1 and c2 represent acceleration factors, is the optimal position of individual particles, represents the position of particle j in d-dimensional space, represents the global optimal position of the particle, and rand() represents a random number between [0,1];

[0117] Step 4: Map the updated particle velocities:

[0118]

[0119] And the mapped particles are binarized with random numbers to obtain the updated particle positions: Sometimes, there is when Sometimes, there is

[0120] Step 5: Determine whether the updated particle position needs population constraints, and if necessary, re-constrain it;

[0121] Step 6: If not, determine whether the updated particle position meets the constraint conditions of the objective function. If so, update the global optimal fitness value corresponding to the particle. If not, set the objective function F=F+100 corresponding to the particle that does not meet the conditions, and then update the global optimal fitness value corresponding to the particle.

[0122] Step 7: Determine whether the iteration is finished. If not, return to step 3. If the iteration is finished, output the particle population corresponding to the global optimal fitness value and obtain the optimal solution of the binary state vector.

[0123] The battery control module controls the operating state of the battery based on the optimal solution.

[0124] The battery control module sequentially connects the batteries whose binary state vector is 1 in series to form a battery string that meets the required voltage, then connects the battery strings in parallel, and connects the batteries whose binary state vector is 0 to the bypass.

[0125] The above embodiments are only descriptions of the preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A reconfigurable battery energy storage energy optimization method for achieving SOC balance, characterized in that: The following steps are involved: Collect the voltage of n batteries U=[u1,u2,...,u n ], current I=[i1,i2,...,i n ] and SOC value SOC=[soc1,soc2,...,soc n ] and randomly initialize the binary state vector of the battery; constraining the binary state vector of the battery using a population constraint; In combination with the optimization objective function and the constraint conditions, a particle swarm optimization algorithm is used to solve the optimal solution of the binary state vector; The operating state of the battery is controlled based on the optimal solution.

2. A reconfigurable battery energy storage energy optimization method for achieving SOC balance according to claim 1, characterized in that: The binary state vector is X=[x1, x2, ..., x n ],have: Wherein, n represents the number of batteries.

3. A reconfigurable battery energy storage energy optimization method for achieving SOC balance according to claim 1, characterized in that: The population constraints include: When the battery pack is discharged, x max =1,x min =0; When the battery pack is charging, x max =0,x min =1; Among them, x max Indicates the working state of the battery when the SOC value in the battery pack is the highest, x min Indicates the working status of the battery when the SOC value in the battery pack is the lowest.

4. A reconfigurable battery energy storage energy optimization method for achieving SOC balance according to claim 2, characterized in that: The optimization objective function includes: min F=ω1f1+ω2f2 Wherein, f1 represents the first sub-objective function, ω1 represents the weight coefficient of the first sub-objective function, f2 represents the second sub-objective function, ω2 represents the weight coefficient of the second sub-objective function, n represents the number of batteries, P represents the number of parallel branches, p represents the pth parallel branch, and U p Indicates the voltage of each parallel branch, U req Indicates the required voltage; The constraints include: IN min ≤U p ≤U max in off in j in max i disc ≤i j ≤i c Among them, U max and U min Indicates the upper and lower limits of the battery parallel branch voltage, u max and u off Indicates the upper and lower limits of the working battery voltage, i c and i disc Indicates the upper and lower limits of the current during charging and discharging.

5. A reconfigurable battery energy storage energy optimization method for achieving SOC balancing according to claim 4, characterized in that: The method for solving the optimal solution includes: Step 1: Initialize the binary state vector X and the algorithm parameters of the particle swarm optimization algorithm; Step 2: Determine the difference between the maximum SOC value and the minimum SOC value in the battery pack and the balance threshold value. When the difference is greater than the balance threshold value, use the population constraint to re-constrain the binary state vector X. When the difference is less than the balance threshold value, do not constrain the binary state vector X. Step 3: Calculate the initial individual optimal value and the global optimal fitness value of the objective function according to the binary state vector X, and enter the iteration to update the inertia weight and speed of the particles in the population: oh = oh max -(oh max -oh min )*k / G Among them, ω represents the updated inertia weight, ω max and ω min represents the upper and lower limits of the inertia weight, k represents the current number of iterations, G represents the number of iterations, represents the velocity of particle j in d-dimensional space, c1 and c2 represent acceleration factors, is the optimal position of individual particles, represents the position of particle j in d-dimensional space, represents the global optimal position of the particle, and rand() represents a random number between [0, 1]; Step 4: Map the updated particle velocity and binarize the mapped particles using random numbers to obtain the updated particle position: Sometimes, there is when Sometimes, there is Step 5: Determine whether the updated particle position requires the population constraint, and if necessary, re-constrain it; Step 6: If not, determine whether the updated particle position meets the constraint conditions of the objective function. If so, update the global optimal fitness value corresponding to the particle. If not, set the objective function F=F+100 corresponding to the particle that does not meet the condition, and then update the global optimal fitness value corresponding to the particle. Step 7: Determine whether the iteration is finished. If not, return to step 3. If the iteration is finished, output the particle population corresponding to the global optimal fitness value to obtain the optimal solution of the binary state vector.

6. A reconfigurable battery energy storage energy optimization method for achieving SOC balance according to claim 4, characterized in that: The method for controlling the working state of the battery based on the optimal solution includes: connecting batteries whose binary state vector is 1 in series to form a battery string that meets the required voltage, then connecting the battery strings in parallel, and connecting batteries whose binary state vector is 0 to a bypass.

7. A reconfigurable battery energy storage energy optimization system for achieving SOC balancing, wherein the optimization system applies the optimization method according to any one of claims 1 to 6, characterized in that: include: State vector initialization module, population constraint module, optimal solution solving module and battery control module; The state vector initialization module is used to collect the voltages of n batteries U = [u1, u2, ..., u n ], current I=[i1,i2,...,i n ] and SOC value SOC=[soc1,soc2,...,soc n ] and randomly initialize the binary state vector of the battery; The population constraint module constrains the binary state vector of the battery using population constraints; The optimal solution solving module is used to combine the optimization objective function and the constraint conditions to solve the optimal solution of the binary state vector using a particle swarm optimization algorithm; The battery control module controls the operating state of the battery based on the optimal solution.

Images