Battery parameter identification method and system based on tenoning predation algorithm

Abstract

The invention discloses a battery parameter identification method and system based on a tenoning predation algorithm, and relates to the field of battery parameter identification methods, and the method comprises the steps: constructing a second-order equivalent circuit model of a battery; constructing a fitness function based on the second-order equivalent circuit model; determining an initial tenon population; the initial tenon population comprises a plurality of tenons, and the position of each tenon represents a group of to-be-identified parameters of the second-order equivalent circuit model; calculating the fitness value of the position of each tenon in the initial tenon population by adopting the fitness function; according to the fitness value of the position of each tenon in the initial tenon population, the initial optimal position of the tenon is determined; determining a secondary optimal position of the tenon according to the initial optimal position of the tenon and the fitness function; determining the optimal position of the tenon according to the second optimal position of the tenon and the fitness function; the optimal position of the tenon is the optimal value of the to-be-identified parameter. According to the invention, the identification efficiency and precision are improved.

Description

Technical Field

[0001] The present application relates to the field of battery parameter identification methods, and in particular to a battery parameter identification method and system based on a peregrine falcon hunting algorithm. Background Art

[0002] Battery parameter identification plays an important role in the development and application of battery technology. It is directly related to the health status monitoring and safety analysis of the battery. Through battery parameter identification, we can deeply understand the internal state of the battery, accurately predict the remaining life of the battery, and formulate a maintenance plan accordingly to ensure the stable and reliable operation of the battery system.

[0003] In the process of battery parameter identification, commonly used methods include intelligent optimization algorithms such as particle swarm optimization and fitting algorithms. Fitting identification algorithms are widely used because of their simple principles and small amount of calculation. However, the battery model has many parameters, which increases the complexity and difficulty of fitting. At the same time, the performance of the fitting algorithm is often affected by the initial parameter values. Inappropriate initial values ​​may cause the algorithm to converge more slowly or fall into a local optimal solution, making it impossible to obtain accurate identification results. As a global search algorithm, the particle swarm algorithm shows good robustness and global search capabilities when identifying battery models, but it also has some disadvantages. First, the performance of the particle swarm algorithm is sensitive to parameter settings, and inappropriate parameters may lead to a decrease in algorithm performance. Secondly, the particle swarm algorithm may converge to a local optimal solution rather than a global optimal solution too early during the search process, which affects the identification accuracy. Summary of the invention

[0004] The purpose of this application is to provide a battery parameter identification method and system based on the peregrine falcon hunting algorithm, which can improve the efficiency and accuracy of identification.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a battery parameter identification method based on the peregrine falcon hunting algorithm, comprising:

[0007] Construct a second-order equivalent circuit model of the battery;

[0008] Constructing a fitness function based on the second-order equivalent circuit model;

[0009] Determine an initial peregrine falcon population; the initial peregrine falcon population includes a plurality of peregrine falcons, and the position of each peregrine falcon represents a set of parameters to be identified of the second-order equivalent circuit model;

[0010] The fitness function is used to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population;

[0011] Determining a preliminary optimal position for the peregrine falcon according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population;

[0012] Determining a sub-optimal position of the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function;

[0013] The optimal position of the peregrine falcon is determined according to the suboptimal position of the peregrine falcon and the fitness function; the optimal position of the peregrine falcon is the optimal value of the parameter to be identified.

[0014] Optionally, the second-order equivalent circuit model includes a voltage source, an ohmic internal resistor and two RC parallel networks; the RC parallel network includes a first capacitor, a second capacitor, a first resistor and a second resistor, the first capacitor is connected in series with the first resistor, the second capacitor is connected in series with the second resistor, the first capacitor and the first resistor are connected in parallel after the series connection, and the second capacitor and the second resistor are connected in parallel after the series connection; constructing a fitness function based on the second-order equivalent circuit model specifically includes:

[0015] Using formula determining voltages of the first capacitor and the second capacitor;

[0016] Among them, U 1,k+1 represents the voltage of the first capacitor at time k+1, U 1,k represents the voltage of the first capacitor at time k, U 2,k+1 represents the voltage of the second capacitor at time k+1, U 2,k represents the voltage of the second capacitor at time k, Δt is the sampling time, R1 represents the resistance value of the first resistor, R2 represents the resistance value of the second resistor, I k represents the total current of the battery branch at time k;

[0017] According to the voltage of the first capacitor and the voltage of the second capacitor, the formula Determine the simulated voltage value of the battery at time k;

[0018] in, represents the simulated voltage value of the battery at time k, U OCV (k) represents the open circuit voltage of the voltage source at time k, and R0 represents the resistance value of the ohmic internal resistance;

[0019] The fitness function is constructed based on the simulated voltage value of the battery at time k:

[0020]

[0021] Wherein, x0 is the parameter to be identified, and the parameter to be identified includes the resistance value of the ohmic internal resistance, the capacitance value of the first capacitor, the resistance value of the first resistor, the capacitance value of the second capacitor and the resistance value of the second resistor, Ud (k) is the actual experimental value of the battery at time k, and m is the number of parameters to be identified.

[0022] Optionally, determining the initial optimal position of the peregrine falcon according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population specifically includes:

[0023] Determine the position of the peregrine falcon with the smallest fitness value, the position of the peregrine falcon with the second smallest fitness value, the weight of the smallest fitness value, and the weight of the second smallest fitness value in the initial peregrine falcon population;

[0024] Adopting the parallelogram diagonal method, the position of each peregrine falcon in the initial peregrine falcon population is updated according to the weight of the minimum fitness value and the weight of the second minimum fitness value;

[0025] Calculate the fitness value of the position of each peregrine falcon in the updated initial peregrine falcon population, return to and execute the steps of determining the position of the peregrine falcon with the smallest fitness value, the position of the peregrine falcon with the second smallest fitness value, the weight of the smallest fitness value and the weight of the second smallest fitness value in the initial peregrine falcon population, until the number of iterations reaches the set maximum number of iterations, and then use the peregrine falcon position with the smallest updated fitness value as the initial optimal position of the peregrine falcon.

[0026] Optionally, the weight of the minimum fitness value and the weight of the second minimum fitness value are:

[0027]

[0028] Among them, W1 represents the weight of the minimum fitness value, W2 represents the weight of the second minimum fitness value, and F min Indicates the minimum fitness value, F min-1 Indicates the second minimum fitness value.

[0029] Optionally, the parallelogram diagonal method is used, and the formula for updating the position of each peregrine falcon in the initial peregrine falcon population according to the weight of the minimum fitness value and the weight of the second minimum fitness value is:

[0030] X new =(W1*(X min -X i )+W2*(X min-1 -X i ))*step_size+X i ;

[0031] Among them, X new is the position of each peregrine falcon in the updated initial peregrine falcon population, X i is the position of the i-th peregrine falcon in the initial peregrine falcon population, i ranges from (0, num], num represents the number of peregrine falcons in the initial peregrine falcon population, X minis the position of the peregrine falcon with the smallest fitness value, X min-1 is the position of the peregrine falcon with the second smallest fitness value, and step_size is the step size of each move.

[0032] Optionally, determining the sub-optimal position of the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function specifically includes:

[0033] Using the initial optimal position of the peregrine falcon as the spiral center to draw a spiral curve, and using the horizontal coordinate value on the spiral curve as the first search point;

[0034] Calculating the fitness values ​​of all first search points using the fitness function;

[0035] The first search point with the smallest fitness value is taken as the initial optimal position of the peregrine falcon, and the step of drawing a spiral curve with the initial optimal position of the peregrine falcon as the spiral center is returned and executed until the number of iterations reaches the set maximum number of iterations, and the initial optimal position of the peregrine falcon obtained with the maximum number of iterations is taken as the sub-optimal position of the peregrine falcon.

[0036] Optionally, the formula for drawing a spiral curve using the initial optimal position of the peregrine falcon as the spiral center is:

[0037] x = r * cos (theta) + h;

[0038] y=r*sin(theta)+b;

[0039] Where theta is an angle value uniformly distributed in [0, n*π], n is a set value, r is the first search radius, ranging from (0, R], and R decreases with the set number of iterations. The radius attenuation factor decreases; h is the horizontal coordinate of the spiral center, b is the vertical coordinate of the spiral center, and b=0, x is the horizontal coordinate on the spiral curve, and y is the vertical coordinate on the spiral curve.

[0040] Optionally, determining the optimal position of the peregrine falcon according to the suboptimal position of the peregrine falcon and the fitness function specifically includes:

[0041] Calculating the gradient of the suboptimal position of the peregrine falcon, and locally searching for a second search point with the suboptimal position of the peregrine falcon as the core position;

[0042] Adopting the momentum gradient descent method, updating the suboptimal position of the peregrine falcon according to the gradient;

[0043] The fitness function is used to calculate the fitness value of the updated sub-optimal position of the peregrine falcon;

[0044] Calculating the fitness values ​​of all second search points using the fitness function;

[0045] Determine the minimum fitness value of the second search point;

[0046] Comparing the updated fitness value of the suboptimal position of the peregrine falcon with the minimum fitness value of the second search point;

[0047] When the fitness value of the updated sub-optimal position of the peregrine falcon is less than or equal to the minimum fitness value of the second search point, the updated sub-optimal position of the peregrine falcon is used as the sub-optimal position of the peregrine falcon, otherwise the second search point corresponding to the minimum fitness value is used as the sub-optimal position of the peregrine falcon;

[0048] Return and execute the step of calculating the gradient of the suboptimal position of the peregrine falcon, and locally searching for the second search point with the suboptimal position of the peregrine falcon as the core position, until |gradf(X) is satisfied j -gradf(X) j-1 |≤1*10 -10 When , the suboptimal position of the peregrine falcon is the optimal position of the peregrine falcon;

[0049] Among them, gradf(X) j is the gradient calculated for the jth time, gradf(X) j-1 is the gradient calculated for the j-1th time.

[0050] Optionally, the formula for locally searching for the second search point with the suboptimal position of the peregrine falcon as the core position is:

[0051] X search =X1+(2*rand(numsearch,dim)-1)*R*RD;

[0052] Among them, X search represents the second search point, X1 represents the suboptimal position of the peregrine falcon, numsearch represents the number of the second search points, dim represents the dimension of X1, one dimension represents a parameter to be identified, (2*rand(numsearch,dim)-1) represents the generation of a matrix with numsearch rows and dim columns ranging from [-1,1], R represents the second search radius, and RD represents the radius attenuation factor.

[0053] In a second aspect, the present application provides a battery parameter identification system based on the peregrine falcon hunting algorithm, comprising:

[0054] Battery model construction module, used to construct the second-order equivalent circuit model of the battery;

[0055] A fitness function construction module, used to construct a fitness function based on the second-order equivalent circuit model;

[0056] An initial peregrine falcon population determination module is used to determine an initial peregrine falcon population; the initial peregrine falcon population includes a plurality of peregrine falcons, and the position of each peregrine falcon represents a set of parameters to be identified of the second-order equivalent circuit model;

[0057] A fitness value calculation module, used to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population using the fitness function;

[0058] A module for determining the initial optimal position of peregrine falcons, used for determining the initial optimal position of peregrine falcons according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population;

[0059] A sub-optimal position determination module for a peregrine falcon, used to determine a sub-optimal position for the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function;

[0060] The module for determining the optimal position of the peregrine falcon is used to determine the optimal position of the peregrine falcon according to the sub-optimal position of the peregrine falcon and the fitness function; the optimal position of the peregrine falcon is the optimal value of the parameter to be identified.

[0061] According to the specific embodiments provided in this application, this application has the following technical effects:

[0062] The present application provides a battery parameter identification method and system based on the peregrine falcon hunting algorithm, establishes a second-order equivalent circuit model of the battery, constructs a fitness function based on the second-order equivalent circuit model, and then determines the initial peregrine falcon population. By simulating the peregrine falcon's search behavior and tracking behavior, an optimization method is proposed, which uses the fitness function to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population; determines the initial optimal position of the peregrine falcon according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population; determines the sub-optimal position of the peregrine falcon according to the initial optimal position of the peregrine falcon and the fitness function; determines the optimal position of the peregrine falcon according to the sub-optimal position of the peregrine falcon and the fitness function, and the optimal position of the peregrine falcon is the optimal value of the parameter to be identified, thereby improving the efficiency and accuracy of parameter identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0064] Figure 1 A flowchart of a battery parameter identification method based on a peregrine falcon hunting algorithm provided in an embodiment of the present application;

[0065] Figure 2 A diagram of a second-order RC model of a battery provided in one embodiment of the present application;

[0066] Figure 3 A schematic diagram of a voltage and current curve of an intermittent constant current discharge experiment provided in an embodiment of the present application;

[0067] Figure 4 A schematic diagram of an OCV (Open circuit voltage)-SOC (State Of Charge) fitting curve used in an intermittent constant current experiment provided in an embodiment of the present application;

[0068] Figure 5 A schematic diagram of a curve showing changes in the fitness function during the iteration of the peregrine falcon hunting algorithm and the PSO (Partical Swarm Optimization) algorithm provided in one embodiment of the present application;

[0069] Figure 6 A schematic diagram of the resistance parameter identification result of the peregrine falcon hunting algorithm provided in one embodiment of the present application;

[0070] Figure 7 A schematic diagram of the capacitor parameter identification result of the Peregrine Falcon hunting algorithm provided in one embodiment of the present application;

[0071] Figure 8 A schematic diagram of a terminal voltage estimation curve of the Peregrine Falcon hunting algorithm and the particle swarm algorithm under an intermittent constant current discharge test provided in an embodiment of the present application;

[0072] Fig. 9 A schematic diagram of a relative error curve of terminal voltage estimation between the Peregrine Falcon hunting algorithm and the particle swarm algorithm under an intermittent constant current discharge test provided in an embodiment of the present application;

[0073] Fig.10 A schematic diagram of a voltage and current curve under dynamic conditions provided by an embodiment of the present application;

[0074] Fig.11 A schematic diagram of a terminal voltage estimation curve of a peregrine falcon hunting algorithm and a particle swarm algorithm under dynamic conditions provided in an embodiment of the present application;

[0075] Fig.12 A schematic diagram of a terminal voltage estimation error curve of a peregrine falcon hunting algorithm and a particle swarm algorithm under dynamic conditions provided in an embodiment of the present application;

[0076] Fig.13 A schematic diagram of functional modules of a battery parameter identification system based on the peregrine falcon hunting algorithm provided in another embodiment of the present application. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0078] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0079] In an exemplary embodiment, Figure 1 As shown, a battery parameter identification method based on the peregrine falcon hunting algorithm is provided, comprising the following steps S1-S7:

[0080] S1: Construct a second-order equivalent circuit model of the battery.

[0081] S2: Constructing a fitness function based on the second-order equivalent circuit model.

[0082] S3: Determine an initial peregrine falcon population; the initial peregrine falcon population includes a plurality of peregrine falcons, and the position of each peregrine falcon represents a set of parameters to be identified of the second-order equivalent circuit model.

[0083] S4: using the fitness function to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population.

[0084] S5: Determine a preliminary optimal position for the peregrine falcon according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population.

[0085] S6: Determine the sub-optimal position of the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function.

[0086] S7: Determine the optimal position of the peregrine falcon according to the suboptimal position of the peregrine falcon and the fitness function; the optimal position of the peregrine falcon is the optimal value of the parameter to be identified.

[0087] This example uses the INR18650-30P lithium-ion battery of DEMGC as the research object. The rated voltage of the battery is 3.7V and the capacity is 3000mAh. The battery is charged in a constant current charging mode (1C) until the cut-off voltage is reached. After standing for two hours, the battery reaches a fully charged state.

[0088] S1: Construct a second-order equivalent circuit model of the battery, such as Figure 2 As shown, the specific steps are as follows:

[0089] S1-1: Establish a second-order equivalent circuit model of the battery. The second-order equivalent circuit model includes a voltage source, an ohmic internal resistor and two RC parallel networks; the RC parallel network includes a first capacitor, a second capacitor, a first resistor and a second resistor, the first capacitor is connected in series with the first resistor, the second capacitor is connected in series with the second resistor, the first capacitor and the first resistor are connected in parallel, and the second capacitor and the second resistor are connected in series; the voltage source U OCV represents the open circuit voltage of the battery, which is a nonlinear function related to the state of charge (SOC); the ohmic internal resistance R0 represents the barrier effect of the electrode, electrolyte, and diaphragm on ions; R1C1 and R2C2 represent the polarization effect of the battery; where U d * represents the analog voltage value of the lithium battery (terminal voltage analog value), I represents the total current of the battery branch, and U1 and U2 are the voltages across the first capacitor C1 and the second capacitor C2 respectively.

[0090]

[0091] Discretizing equations (1) and (2) yields:

[0092]

[0093] Among them, U 1,k+1 represents the voltage of the first capacitor at time k+1, U 1,k represents the voltage of the first capacitor at time k, U 2,k+1 represents the voltage of the second capacitor at time k+1, U 2,k represents the voltage of the second capacitor at time k, Δt is the sampling time, Δt=1s, R1 represents the resistance value of the first resistor, R2 represents the resistance value of the second resistor, I k Represents the total current of the battery branch at time k.

[0094] Using formula Determine the simulated voltage value of the battery at time k; where, represents the simulated voltage value of the battery at time k, U OCV (k) represents the open circuit voltage of the voltage source at time k, and R0 represents the resistance value of the ohmic internal resistance.

[0095] S1-2: According to the ampere-hour integration method, the discrete expression of SOC can be obtained:

[0096]

[0097] Among them, SOC k and SOC k+1 represents the SOC value at the kth and k+1th moments, η represents the coulomb efficiency, Q nRepresents battery capacity, η=1, Q n =3AH.

[0098] S2: Construct a fitness function based on the second-order equivalent circuit model. Determine the open circuit voltage U through intermittent constant current discharge conditions OCV The functional relationship with SOC and the fitness function constructed with the output results of the S1 model are as follows:

[0099] S2-1: OCV test discharge conditions: (1) Select a 3000mAh Dongmei lithium-ion battery INR18650-30P, charge the battery at 1C constant current until it reaches 4.5V, and then let it stand for 2 hours. At this time, the SOC of the battery is 100%; (2) Discharge at 1C constant current for 6 minutes, discharging 10% of the SOC each time, and then let it stand for 2 hours; (3) Repeat (2) until the discharge cut-off voltage is reached.

[0100] like Figure 3 As shown, the process is to discharge at a constant current of 1C (3A) for 6 minutes, discharging 10% SOC each time, and then leaving it to stand for 2 hours.

[0101] S2-2: The corresponding voltage after standing for 2 hours for every 10% change in SOC is the open circuit voltage, and SOC-U is obtained. OCV Fitting curve, such as Figure 4 As shown, the fitting curve equation is as follows:

[0102] U OCV (k) = 157.076719519916SOC k 9 -332.354323051874SOC k 8 -96.9715961410876SOC k 7 +860.647682818301SOC k 6 -988.830656908515SOC k 5 +521.135116316975SOC k 4 -133.401098899786SOC k 3 +11.7651014705788SOC k 2 +2.19725487549047SOC k +3.02172287341135 (7)

[0103] Among them, SOC k t Represents the t-th power of the SOC value at the k-th moment.

[0104] S2-3: The goal of optimal parameter identification is to obtain the optimal circuit parameters based on experimental data, so that the sum of the errors between the output analog voltage value and the actual experimental value is minimized. Therefore, the fitness function is established by combining formula (3), formula (4), formula (5), formula (6) and formula (7):

[0105]

[0106] Wherein, x0 is a parameter to be identified, and the parameter to be identified includes the resistance value of the ohmic internal resistance, the capacitance value of the first capacitor, the resistance value of the first resistor, the capacitance value of the second capacitor and the resistance value of the second resistor, namely [R0R1C1R2C2], is the simulated voltage value of the battery at time k, U d (k) is the actual experimental value of the battery at time k, and m is the number of parameters to be identified.

[0107] S3: Determine the initial peregrine falcon population. The steps are as follows:

[0108] Define the upper and lower limits ub and lb of the search space of the peregrine falcon population, and randomly generate the initial solution, that is, the initial position of the peregrine falcon population:

[0109] X=[x1 x2 x3 x4 x5]=[R0 R1 C1 C2 R2]=(ub-lb)*rand+lb (9)

[0110] Here, rand is a random number in [0,1].

[0111] Define other parameters of the algorithm, including the initial number of peregrine falcons num, dimension dim, maximum number of iterations maxIter, search radius R, radius attenuation factor RD, etc.

[0112] S4: Calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population using the fitness function:

[0113] F=F(X)(10)

[0114] S5: According to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population, the initial optimal position of the peregrine falcon is determined, that is, the population cooperates to search for prey, and the individuals of the population move their positions according to the fitness to find the initial individual optimal position, including the following steps:

[0115] S5-1: Determine the position of the peregrine falcon with the smallest fitness value, the position of the peregrine falcon with the second smallest fitness value, the weight of the smallest fitness value, and the weight of the second smallest fitness value in the initial peregrine falcon population.

[0116] The weight of the minimum fitness value and the weight of the second minimum fitness value are:

[0117]

[0118] W2=1-W1 (12)

[0119] Among them, W1 represents the weight of the minimum fitness value, W2 represents the weight of the second minimum fitness value, and F min Indicates the minimum fitness value, F min-1 Indicates the second minimum fitness value.

[0120] S5-2: Use the parallelogram diagonal method to update the position of each peregrine falcon in the initial peregrine falcon population according to the weight of the minimum fitness value and the weight of the second minimum fitness value. The specific update formula is as follows:

[0121]

[0122] Simplifying, we get:

[0123] X new =(W1*(X min -X i )+W2*(X min-1 -X i )) * step_size+X i (14)

[0124] Among them, X new is the position of each peregrine falcon in the updated initial peregrine falcon population, X i is the position of the i-th peregrine falcon in the initial peregrine falcon population, i ranges from (0, num], num represents the number of peregrine falcons in the initial peregrine falcon population, X min is the position of the peregrine falcon with the smallest fitness value, X min-1 is the position of the peregrine falcon with the second smallest fitness value, step_size is the step size of each move and its range is [0,2].

[0125] S5-3: Calculate the fitness value of the position of each peregrine falcon in the updated initial peregrine falcon population:

[0126] F new =F(X new ) ( 15)

[0127] S5-4: Return to step S5-1, X new Reassign to X iRepeat until the number of iterations reaches the set maximum number of iterations maxIter, and output the peregrine falcon position with the smallest updated fitness value as the peregrine falcon's initial optimal position Best_peregrinePos and the fitness value Best_Fitness at this time.

[0128] S6: Determine the suboptimal position of the peregrine falcon according to the initial optimal position of the peregrine falcon and the fitness function, that is, the individual at the initial optimal position circles and searches for prey through a retractable spiral strategy, and find the suboptimal position according to the fitness, including the following steps:

[0129] S6-1: Draw a spiral curve using the initial optimal position of the peregrine falcon Best_peregrinePos as the spiral center, and use the horizontal coordinate value on the spiral curve as the first search point; the formula is as follows:

[0130] x=r*cos(theta)+h (16)

[0131] y=r*sin(theta)+b (17)

[0132] Where theta is an angle value uniformly distributed in [0, n*π], evenly divided into 1000 values, n is a set value, r is the first search radius, ranging from (0, R], and R decreases with the set number of iterations. The radius attenuation factor decreases; h is the horizontal coordinate of the spiral center, that is, the value corresponding to each dimension of Best_peregrinePos, b is the vertical coordinate of the spiral center, and b=0, x is the horizontal coordinate on the spiral curve, and y is the vertical coordinate on the spiral curve.

[0133] In formula (16) and formula (17), (x, y) is drawn as a spiral curve, and the value corresponding to each dimension of the search point is mapped to the spiral curve. The x value is selected as the value corresponding to each dimension of the first search point point.

[0134] S6-2: Calculate the fitness values ​​of all first search points using the fitness function:

[0135] Fit=F(point) (18)

[0136] S6-3: Find the minimum fitness value of the first search point Fit min And the corresponding first search point value Best_point, the first search point with the smallest fitness value is taken as the initial optimal position of the peregrine falcon.

[0137] S6-4: Return to step S6-1 and update the initial optimal position of the peregrine falcon Best_peregrinePos to the first search point Best_point with the minimum fitness value, and continue to execute until the number of iterations reaches the set maximum number of iterations. At this time, the sub-optimal position of the peregrine falcon Best_point and the fitness value Fit of the position are output. min .

[0138] S7: Determine the optimal position of the peregrine falcon according to the suboptimal position of the peregrine falcon and the fitness function, that is, the suboptimal position individual tracks the prey and dives at high speed to catch the prey. Finding the optimal position includes the following steps:

[0139] S7-1: Calculate the gradient of the sub-optimal position of the peregrine falcon, and locally search for a second search point with the sub-optimal position of the peregrine falcon as the core position, that is, substitute the sub-optimal position Best_point of the peregrine falcon output in step S6 into the gradient calculation formula to calculate the gradient of the sub-optimal position of the peregrine falcon:

[0140] The gradient is a vector that indicates that the directional derivative of a function at a certain point has a maximum value along that direction, that is, the function is along that direction at that point. The direction of the gradient is the direction in which the function grows fastest at that point, and the length of the gradient is this maximum rate of change.

[0141] In the more general case, for an n-dimensional function f(x1,x2,…,x n ), the gradient is an n-dimensional vector:

[0142]

[0143] Especially when using partial derivative formulas is not feasible or too complicated, numerical methods are often used to approximate the gradient based on the value of the function at nearby points - finite difference methods.

[0144] For X i The partial derivative of the direction (i.e. the gradient of X i Quantity):

[0145]

[0146] in:

[0147]

[0148] h0 is a small positive number, indicating that iThe step size of the movement in the direction can be adjusted to control the accuracy of the approximation by adjusting the value of h0. A smaller h0 value usually leads to a more accurate approximation. iter refers to the current number of iterations. The gradient gradf(X) of the point can be calculated by combining formula (19), formula (20) and formula (21). S7-2: Using the momentum gradient descent method, the suboptimal position of the peregrine falcon is updated according to the gradient:

[0149] v=β*v+(1-β)*gradf(X1) (22)

[0150] X 1new =X-α*v (23)

[0151] Where v represents momentum, with an initial value of 0, β represents the decay rate of momentum, which is usually set to a value close to 1 (such as 0.9), and gradf(X1) represents the gradient of the fitness function with respect to parameter X1. 1new represents the updated sub-optimal position of the peregrine falcon, X1 represents the sub-optimal position of the peregrine falcon, and α represents the learning rate, which is set according to the specific situation, but cannot be set too large or too small. If α is set too large, the algorithm may oscillate near the optimal solution and fail to converge; if it is set too small, although the algorithm may eventually converge to the optimal solution, the convergence speed may be very slow.

[0152] S7-3: Calculate the fitness value of the updated suboptimal position of the peregrine falcon using the fitness function:

[0153] F1=F(X 1new ) (twenty four)

[0154] S7-4: At the same time, the sub-optimal position of the peregrine falcon is used as the core position to locally search for the second search point, that is, the sub-optimal position Best_point of the peregrine falcon output by S6 is used as the core position to locally search for the second search point:

[0155] X searcj =X1+(2*rand(numsearch,dim)-1)*R*RD (25)

[0156] Among them, X search represents the second search point, X1 represents the suboptimal position of the peregrine falcon, numsearch represents the number of the second search points, dim represents the dimension of X1, one dimension represents a parameter to be identified, (2*rand(numsearch,dim)-1) represents the generation of a matrix with numsearch rows and dim columns ranging from [-1,1], R represents the second search radius, and RD represents the radius attenuation factor.

[0157] S7-5: Calculate the fitness values ​​of all second search points using the fitness function:

[0158] F2=F(X search ) (26)

[0159] S7-6: Find the minimum fitness value F of the fitness values ​​F2 of all second search points 2min And its corresponding second search point X search The value of X 2min .

[0160] S7-7: Compare the updated fitness value F1 of the suboptimal position of the peregrine falcon with the minimum fitness value F1 of the second search point 2min The size of determines the update position:

[0161] When the fitness value F1 of the updated suboptimal position of the peregrine falcon is less than or equal to the minimum fitness value of the second search point, the updated suboptimal position F 2min When the next cycle of the peregrine falcon's optimal position Best_point is initialized to X 1new ; On the contrary, the initial value of the optimal position Best_point of the peregrine falcon in a cycle is the second search point X corresponding to the minimum fitness value 2min .

[0162] S7-8: Return to step S7-1, and bring in the Best_point obtained in S7-7 to continue the optimization.

[0163] S7-9: When it reaches the optimal value, the search range becomes smaller, and the termination condition is met, it is determined that it can dive at high speed to prey:

[0164] |gradf(X) j -gradf(X) j-1 |≤1*10 -10 (27)

[0165] At this point, the iteration stops, and the best point the_Best_point that captures the sub-optimal position of the peregrine falcon is the optimal position of the peregrine falcon, and the optimal fitness value F_best=F(the_Best_point). Where gradf(X) j is the gradient calculated for the jth time (current gradient), gradf(X) j-1 is the gradient calculated for the j-1th time (the gradient calculated last time).

[0166] S8: Bring the optimal position of the peregrine falcon output by S7 into the battery discharge model to verify the accuracy of the parameters and estimate the SOC.

[0167] S8-1: Under different SOC conditions, the lithium-ion battery parameter identification results will have deviations, so the entire discharge process is divided into 10 stages for identification. In each stage, the battery undergoes static, pulse discharge, and static again. The SOC before discharge is defined as the initial SOC. Therefore, the fitness function is established by combining formula (3), formula (4), formula (5), formula (6) and formula (7):

[0168]

[0169] Where x = [R0, R1, C1, R2, C2], is the simulated voltage value of the battery at time k (terminal voltage analog value); U d (k) is the experimental true value of the battery at time k (actual value of the terminal voltage); k is the discrete time, and m is the number of parameters to be identified.

[0170] In this embodiment, the upper limit Ub of the search space of the parameters to be identified is [0.1, 0.1, 100000, 0.1, 100000], the lower limit Lb is [0.001, 0.001, 1, 0.001, 1], the initial number of peregrine falcons num is 30, the dimension dim is 5, and the maximum number of iterations maxIter is 1000. The peregrine falcon predation algorithm and the particle swarm optimization (PSO) algorithm are used to perform parameter identification on the second-order RC model of lithium-ion batteries. The fitness function change curve during the iteration process is as follows: Figure 5 As shown in the figure, the comparison shows that the Peregrine Falcon Predator Algorithm has certain advantages in terms of convergence speed and accuracy, which can effectively prevent the algorithm from falling into the local optimum and converge faster. The Peregrine Falcon Predator Algorithm is used to identify each battery discharge process, and the identification results are shown in Table 1 below:

[0171] Table 1. The identification results of the peregrine falcon hunting algorithm for each battery discharge process

[0172]

[0173] S8-2: Plot the result of each segment identification with the corresponding initial SOC to obtain the variation curve of the battery model parameters under different SOCs, such as Figure 6 and Figure 7 As shown. Using the obtained parameters, the estimated value of the battery terminal voltage can be estimated. By comparing it with the actual value, the accuracy of the identification algorithm can be evaluated. The comparison results are shown in Figure 8 and Fig. 9 The results show that the parameters identified by the Peregrine Falcon predation algorithm are more accurate, and the error between the terminal voltage prediction value and the actual value is smaller, especially at the beginning and end of the discharge, which shows that the Peregrine Falcon predation algorithm has better globality.

[0174] S8-3: In order to continue to verify the global performance of the peregrine falcon hunting algorithm, a dynamic working condition was set using the battery. Fig.10 As shown in the figure, it is a small cycle of dynamic working condition. The battery starts to discharge from the full charge state, first discharges at 2A for 5 seconds, then charges at 1A for 5 seconds, and finally discharges at 0.5A for 10 seconds. This is a small cycle. The total charge discharged in this process is: Q = 2×5-1×5+0.5×10 = 10 coulombs. In order to continue to discharge 10% SOC each time, and then let it stand for 2 hours, it needs to go through each rest before This cycle repeats 10 times until the cut-off voltage is reached.

[0175] S8-4: Continue to repeat the process of S8-1 and S8-2 to identify the parameters to be identified [R0, R1, C1, R2, C2] of each SOC segment, as shown in Table 2 below:

[0176] Table 2 Parameter results of the parameters to be identified in each SOC segment

[0177] SOC <![CDATA[R0(Ω)]]> <![CDATA[R1(Ω)]]> <![CDATA[C1(F)]]> <![CDATA[R2(Ω)]]> <![CDATA[C2(F)]]> 0.1 0.023075534 0.1 10868.87681 0.1 28481.11095 0.2 0.020050379 0.085936571 968.7677156 0.066299495 31792.51354 0.3 0.018849538 0.038992656 4207.00124 0.057896118 41055.93377 0.4 0.019398595 0.036846609 981.7048909 0.020536579 29913.08844 0.5 0.019000386 0.015144608 902.0829577 0.069472994 100000 0.6 0.018100476 0.019256824 752.3826143 0.001 45456.38328 0.7 0.018474552 0.034801431 1355.510271 0.061158952 100000 0.8 0.018774562 0.017516436 787.5631655 0.022721791 9222.863847 0.9 0.019024704 0.021681113 1849.613004 0.009090669 63249.66378 1 0.021227832 0.010473792 1074.877433 0.001 97847.80353

[0178] Using the obtained parameters, the estimated value of the battery terminal voltage can be estimated. By comparing it with the actual value, the accuracy of the identification algorithm can be evaluated. The comparison results are as follows: Fig.11 and Fig.12 The results show that the parameters identified by the Peregrine Falcon predation algorithm are more accurate, and the error between the terminal voltage prediction value and the actual value is smaller, especially in the dynamic change process of discharge to charge, which shows that the Peregrine Falcon predation algorithm has stronger adaptability.

[0179] The peregrine falcon hunting algorithm proposed in this application is a new type of heuristic algorithm. It provides a new method for battery parameter identification by imitating the peregrine falcon's hunting behavior, including four processes: population cooperation in searching for prey, individual circling in searching for prey, discovery and tracking of prey, and high-speed diving to hunt prey.

[0180] Compared with the prior art, the beneficial effects of this application are:

[0181] (1) This application establishes a second-order RC model of a lithium-ion battery based on Kirchhoff's law, and measures intermittent constant current discharge data through experiments to fit the OCV-SOC relationship.

[0182] (2) This application proposes an optimization method by simulating the search and tracking behavior of peregrine falcons. The algorithm can coordinate global search and local search in the entire search space to avoid falling into the local optimal solution, and has strong global search capabilities. The algorithm can perform a detailed search in the area close to the optimal solution, so as to accurately obtain the optimal solution. The algorithm can adjust parameters and strategies according to the characteristics of the problem to adapt to different types of optimization problems. It has high flexibility and strong stability. It solves the problems of premature convergence to the local optimum and slow convergence speed, and the identification effect is better than the traditional heuristic algorithm.

[0183] Based on the same inventive concept, the embodiment of the present application also provides a battery parameter identification system based on the peregrine falcon hunting algorithm for implementing the battery parameter identification method based on the peregrine falcon hunting algorithm mentioned above. Fig.13 As shown, including:

[0184] The battery model construction module is used to construct a second-order equivalent circuit model of the battery.

[0185] The fitness function construction module is used to construct a fitness function based on the second-order equivalent circuit model.

[0186] The initial peregrine falcon population determination module is used to determine the initial peregrine falcon population; the initial peregrine falcon population includes multiple peregrine falcons, and the position of each peregrine falcon represents a set of parameters to be identified of the second-order equivalent circuit model. The initial values ​​of the parameters to be identified of the battery are obtained as the initial values ​​of the peregrine falcon positions.

[0187] The fitness value calculation module is used to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population by using the fitness function.

[0188] The module for determining the initial optimal position of peregrine falcons is used to determine the initial optimal position of peregrine falcons according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population.

[0189] The module for determining the sub-optimal position of the peregrine falcon is used to determine the sub-optimal position of the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function.

[0190] The module for determining the optimal position of the peregrine falcon is used to determine the optimal position of the peregrine falcon according to the sub-optimal position of the peregrine falcon and the fitness function; the optimal position of the peregrine falcon is the optimal value of the parameter to be identified.

[0191] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0192] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A battery parameter identification method based on the peregrine falcon hunting algorithm, characterized in that: The battery parameter identification method based on the peregrine falcon hunting algorithm includes: Construct a second-order equivalent circuit model of the battery; Constructing a fitness function based on the second-order equivalent circuit model; Determine an initial peregrine falcon population; the initial peregrine falcon population includes a plurality of peregrine falcons, and the position of each peregrine falcon represents a set of parameters to be identified of the second-order equivalent circuit model; The fitness function is used to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population; Determining a preliminary optimal position for the peregrine falcon according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population; Determining a sub-optimal position of the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function; The optimal position of the peregrine falcon is determined according to the suboptimal position of the peregrine falcon and the fitness function; the optimal position of the peregrine falcon is the optimal value of the parameter to be identified.

2. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 1 is characterized in that: The second-order equivalent circuit model includes a voltage source, an ohmic internal resistor and two RC parallel networks; the RC parallel network includes a first capacitor, a second capacitor, a first resistor and a second resistor, the first capacitor is connected in series with the first resistor, the second capacitor is connected in series with the second resistor, the first capacitor and the first resistor connected in series are connected in parallel with the second capacitor and the second resistor connected in series; Constructing a fitness function based on the second-order equivalent circuit model specifically includes: Using formula determining voltages of the first capacitor and the second capacitor; Among them, U 1,k+1 represents the voltage of the first capacitor at time k+1, U 1,k represents the voltage of the first capacitor at time k, U 2,k+1 represents the voltage of the second capacitor at time k+1, U 2,k represents the voltage of the second capacitor at time k, Δt is the sampling time, R1 represents the resistance value of the first resistor, R2 represents the resistance value of the second resistor, I k represents the total current of the battery branch at time k; According to the voltage of the first capacitor and the voltage of the second capacitor, the formula Determine the simulated voltage value of the battery at time k; in, represents the simulated voltage value of the battery at time k, U OCV (k) represents the open circuit voltage of the voltage source at time k, and R0 represents the resistance value of the ohmic internal resistance; The fitness function is constructed based on the simulated voltage value of the battery at time k: Wherein, x0 is the parameter to be identified, and the parameter to be identified includes the resistance value of the ohmic internal resistance, the capacitance value of the first capacitor, the resistance value of the first resistor, the capacitance value of the second capacitor and the resistance value of the second resistor, U d (k) is the actual experimental value of the battery at time k, and m is the number of parameters to be identified.

3. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 1, characterized in that: Determining the initial optimal position of the peregrine falcon according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population includes: Determine the position of the peregrine falcon with the smallest fitness value, the position of the peregrine falcon with the second smallest fitness value, the weight of the smallest fitness value, and the weight of the second smallest fitness value in the initial peregrine falcon population; Adopting the parallelogram diagonal method, the position of each peregrine falcon in the initial peregrine falcon population is updated according to the weight of the minimum fitness value and the weight of the second minimum fitness value; Calculate the fitness value of the position of each peregrine falcon in the updated initial peregrine falcon population, return to and execute the steps of determining the position of the peregrine falcon with the smallest fitness value, the position of the peregrine falcon with the second smallest fitness value, the weight of the smallest fitness value and the weight of the second smallest fitness value in the initial peregrine falcon population, until the number of iterations reaches the set maximum number of iterations, and then use the peregrine falcon position with the smallest updated fitness value as the initial optimal position of the peregrine falcon.

4. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 3 is characterized in that: The weight of the minimum fitness value and the weight of the second minimum fitness value are: Among them, W1 represents the weight of the minimum fitness value, W2 represents the weight of the second minimum fitness value, and F min Indicates the minimum fitness value, F min-1 Indicates the second minimum fitness value.

5. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 4 is characterized in that: The formula for updating the position of each peregrine falcon in the initial peregrine falcon population according to the weight of the minimum fitness value and the weight of the second minimum fitness value using the parallelogram diagonal method is: X new =(W1*(X min -X i )+W2*(X min-1 -X i ))*step_size+X i ; Among them, X new is the position of each peregrine falcon in the updated initial peregrine falcon population, X i is the position of the i-th peregrine falcon in the initial peregrine falcon population, i ranges from (0, num], num represents the number of peregrine falcons in the initial peregrine falcon population, X min is the position of the peregrine falcon with the smallest fitness value, X min-1 is the position of the peregrine falcon with the second smallest fitness value, and step_size is the step size of each move.

6. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 1, characterized in that: Determining the sub-optimal position of the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function specifically includes: Using the initial optimal position of the peregrine falcon as the spiral center to draw a spiral curve, and using the horizontal coordinate value on the spiral curve as the first search point; Calculating the fitness values ​​of all first search points using the fitness function; The first search point with the smallest fitness value is taken as the initial optimal position of the peregrine falcon, and the step of drawing a spiral curve with the initial optimal position of the peregrine falcon as the spiral center is returned and executed until the number of iterations reaches the set maximum number of iterations, and the initial optimal position of the peregrine falcon obtained with the maximum number of iterations is taken as the sub-optimal position of the peregrine falcon.

7. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 6 is characterized in that: The formula for drawing a spiral curve using the initial optimal position of the peregrine falcon as the spiral center is: x = r*cos(theta)+h; y = r*sin(theta)+b; Where theta is an angle value uniformly distributed in [0, n*π], n is a set value, r is the first search radius, ranging from (0, R], and R decreases with the set number of iterations. The radius attenuation factor decreases; h is the horizontal coordinate of the spiral center, b is the vertical coordinate of the spiral center, and b=0, x is the horizontal coordinate on the spiral curve, and y is the vertical coordinate on the spiral curve.

8. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 1, characterized in that: Determining the optimal position of the peregrine falcon according to the suboptimal position of the peregrine falcon and the fitness function specifically includes: Calculating the gradient of the suboptimal position of the peregrine falcon, and locally searching for a second search point with the suboptimal position of the peregrine falcon as the core position; Adopting the momentum gradient descent method, updating the suboptimal position of the peregrine falcon according to the gradient; The fitness function is used to calculate the fitness value of the updated sub-optimal position of the peregrine falcon; Calculating the fitness values ​​of all second search points using the fitness function; Determine the minimum fitness value of the second search point; Comparing the updated fitness value of the suboptimal position of the peregrine falcon with the minimum fitness value of the second search point; When the fitness value of the updated sub-optimal position of the peregrine falcon is less than or equal to the minimum fitness value of the second search point, the updated sub-optimal position of the peregrine falcon is used as the sub-optimal position of the peregrine falcon, otherwise the second search point corresponding to the minimum fitness value is used as the sub-optimal position of the peregrine falcon; Return and execute the step of calculating the gradient of the suboptimal position of the peregrine falcon, and locally searching for the second search point with the suboptimal position of the peregrine falcon as the core position, until |gradf(X) is satisfied j -gradf(X) j-1 |≤1*10 -10 When , the suboptimal position of the peregrine falcon is the optimal position of the peregrine falcon; Among them, gradf(X) j is the gradient calculated for the jth time, gradf(X) j-1 is the gradient calculated for the j-1th time.

9. The battery parameter identification method based on the peregrine falcon hunting algorithm according to claim 8, characterized in that: The formula for locally searching for the second search point with the suboptimal position of the peregrine falcon as the core position is: X search =X1+(2*rand(numsearch,dim)-1)*R*RD; Among them, X search represents the second search point, X1 represents the suboptimal position of the peregrine falcon, numsearch represents the number of the second search points, dim represents the dimension of X1, one dimension represents a parameter to be identified, (2*rand(numsearch,dim)-1) represents the generation of a matrix with numsearch rows and dim columns ranging from [-1,1], R represents the second search radius, and RD represents the radius attenuation factor.

10. A battery parameter identification system based on the peregrine falcon hunting algorithm, characterized in that: The battery parameter identification system based on the peregrine falcon hunting algorithm includes: Battery model construction module, used to construct the second-order equivalent circuit model of the battery; A fitness function construction module, used to construct a fitness function based on the second-order equivalent circuit model; An initial peregrine falcon population determination module is used to determine an initial peregrine falcon population; the initial peregrine falcon population includes a plurality of peregrine falcons, and the position of each peregrine falcon represents a set of parameters to be identified of the second-order equivalent circuit model; A fitness value calculation module, used to calculate the fitness value of the position of each peregrine falcon in the initial peregrine falcon population using the fitness function; A module for determining the initial optimal position of peregrine falcons, used for determining the initial optimal position of peregrine falcons according to the fitness value of the position of each peregrine falcon in the initial peregrine falcon population; A sub-optimal position determination module for a peregrine falcon, used to determine a sub-optimal position for the peregrine falcon according to the preliminary optimal position of the peregrine falcon and the fitness function; The module for determining the optimal position of the peregrine falcon is used to determine the optimal position of the peregrine falcon according to the sub-optimal position of the peregrine falcon and the fitness function; the optimal position of the peregrine falcon is the optimal value of the parameter to be identified.

Images