Parameter identification method and device for equivalent circuit model of liquid metal battery

Abstract

The invention discloses a parameter identification method and device for an equivalent circuit model of a liquid metal battery, belongs to the technical field of battery management, and considers that all particles share a global optimal value in a single-population particle optimization algorithm, and once the global optimal value falls into a local optimal point, the parameter identification method and device can identify the equivalent circuit model of the liquid metal battery. All particles of the whole population can be quickly pulled to the wrong direction, so that the algorithm is converged too early; according to the scheme, a multi-population particle optimization algorithm is adopted to estimate parameters therein, the two populations are allowed to independently explore in a search space, when the population A falls into local optimum, the other population B may find a solution with lower fitness, the population A can receive a better solution found by the population B through communication among the populations, and the optimal solution is found by the population B; and therefore, the user jumps out of the local optimal trap where the user is located and turns to a more desirable search area. According to the algorithm, the probability of finding a global optimal parameter combination instead of a local optimal parameter combination is greatly improved, and the accuracy and reliability of a parameter identification result are ensured.

Description

Technical Field

[0001] This invention belongs to the field of battery management technology, and more specifically, relates to a method and apparatus for parameter identification of an equivalent circuit model of a liquid metal battery. Background Technology

[0002] Liquid metal batteries are a novel electrochemical energy storage technology that has seen significant development in recent years. These batteries consist of positive and negative liquid metal electrodes and a molten salt electrolyte. They offer advantages such as long cycle life, high energy density, and low energy storage cost under high-temperature operating conditions, making them promising for large-scale energy storage applications in power systems. Using battery models to estimate battery parameters and predict the battery's health status can effectively ensure the safe and stable operation of liquid metal batteries.

[0003] Among numerous battery models, the equivalent circuit model is widely used due to its advantages of low computational cost, simple and quick parameter identification process, and strong engineering applicability. However, how to accurately identify the parameters in the equivalent circuit model of liquid metal battery has always been a research challenge. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement needs of the prior art, the present invention provides a method and apparatus for parameter identification of equivalent circuit model of liquid metal battery, the purpose of which is to solve the technical problem that the parameters in the equivalent circuit model of liquid metal battery are difficult to identify accurately.

[0005] To achieve the above objectives, according to one aspect of the present invention, a method for parameter identification of an equivalent circuit model of a liquid metal battery is provided, comprising: S1: Limit the parameter search range corresponding to the combination of parameters to be identified in the equivalent circuit model of liquid metal battery; S2: Create population A and population B corresponding to the parameter search range, initialize the initial position and initial velocity of each particle in each population within the parameter search range, and each particle corresponds to a combination of identification parameters; S3: Calculate the current fitness of all particles in the current state; compare the current fitness of all particles, and take the current position and current fitness value of the particle with the smallest fitness as the current global optimum; wherein, the current fitness is: the deviation between the output voltage of the liquid metal battery equivalent circuit model generated according to the combination of parameters to be identified for the particle and the measured voltage; S4: Determine if the iteration stopping condition is met. If it is met, proceed to S5. If not, update the current velocity of each particle in each population based on the current global optimal value, update the current position based on the current velocity, and return to S3. S5: Exchange the current global optimum values ​​of population A and population B, calculate the current fitness of all particles in the current state, and find the current global optimum values ​​of population A and population B respectively; output the identification parameter combination of the particles corresponding to the smaller current global optimum value as the target identification result.

[0006] Furthermore, the combination of parameters to be identified includes: battery ohmic internal resistance R0, battery polarization internal resistance R1, and battery polarization capacitance C1.

[0007] Furthermore, the current fitness is: the output voltage U of the liquid metal battery equivalent circuit model generated based on the combination of parameters to be identified corresponding to the particles. sim With the measured voltage U test The root mean square error between them.

[0008] Furthermore, the current fitness is represented as: ; Where N is the number of sample data points involved in the calculation, and RMSE represents the root mean square error.

[0009] Furthermore, the exchange of the current global optimal values ​​of population A and population B includes: replacing the globally optimal particle in population A with the particle with the highest global fitness in population B, and replacing the globally optimal particle in population B with the particle with the highest global fitness in population A.

[0010] Furthermore, updating the current position based on the current velocity includes: updating the current position of the particle based on the updated current velocity of the particle, while ensuring that the position of the particle does not exceed the parameter boundary.

[0011] Furthermore, the initial position and initial velocity of each particle in each population within the parameter search range include: randomly setting the initial position and setting the initial velocity to 0.

[0012] According to another aspect of the present invention, a parameter identification device for an equivalent circuit model of a liquid metal battery is provided, comprising: The range limiting module is used to limit the parameter search range corresponding to the combination of parameters to be identified in the equivalent circuit model of liquid metal battery; The population creation module is used to create population A and population B corresponding to the parameter search range, and initialize the initial position and initial velocity of each particle in each population within the parameter search range, with each particle corresponding to a combination of identification parameters. The update iteration module is used to calculate the current fitness of all particles in the current state; compare the current fitness of all particles, and take the current position and current fitness value of the particle with the smallest fitness as the current global optimum; wherein, the current fitness is: the deviation between the output voltage of the liquid metal battery equivalent circuit model generated according to the combination of parameters to be identified for the particle and the measured voltage; The condition judgment module is used to determine whether the iteration stopping condition is met. If it is met, the module enters the target output module. If it is not met, the module updates the current velocity of each particle in each population based on the current global optimal value, updates the current position based on the current velocity, and then enters the update iteration module. The target output module is used to exchange the current global optimum values ​​of population A and population B, calculate the current fitness of all particles in the current state, and thus find the current global optimum values ​​of population A and population B respectively; the identification parameters of the particles corresponding to the smaller current global optimum value are combined as the target identification result and output.

[0013] According to another aspect of the present invention, an electronic device is provided, including a memory and a processor, the memory storing a computer program, wherein the processor executes the computer program to implement the steps of the parameter identification method for the equivalent circuit model of the liquid metal battery.

[0014] According to another aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the parameter identification method for the equivalent circuit model of the liquid metal battery.

[0015] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: (1) The parameter identification method for the equivalent circuit model of a liquid metal battery provided by this invention considers that in a single-population particle optimization algorithm, all particles share a global optimum. Once the global optimum falls into a local optimum, all particles in the entire population will be quickly pulled in the wrong direction, causing the algorithm to converge prematurely. This scheme uses a multi-population particle optimization algorithm to estimate the parameters. The multi-population particle optimization algorithm allows two populations to explore the search space independently. When population A falls into a local optimum, another population B may find a solution with lower fitness. Through inter-population communication, population A can receive the better solution found by population B, thereby escaping its local optimum trap and turning to a more promising search area. This algorithm greatly improves the probability of finding the global optimum parameter combination rather than the local optimum parameter combination, ensuring the accuracy and reliability of the parameter identification results.

[0016] (2) The parameter identification method for the equivalent circuit model of liquid metal battery proposed in this invention can effectively improve the speed of parameter identification in the equivalent circuit model of liquid metal battery. Since the evolutionary process of multiple populations is independent of each other within a certain interval, the algorithm can be computed in parallel on multiple CPUs or computer clusters. For single-population particle optimization algorithm, even if the particle evaluation in the population can be computed in parallel, the algorithm logic itself is serial and needs to wait for all particle evaluations to be completed before the global optimal solution can be updated. However, the operation of each population in multi-population particle optimization algorithm is independent, and the running efficiency is higher.

[0017] (3) The present invention provides a parameter identification method for the equivalent circuit model of liquid metal battery, which effectively solves the problems of easily getting trapped in local optimal solutions, low computational efficiency and poor accuracy when using traditional algorithms to estimate the parameters of the equivalent circuit model of liquid metal battery. It is of great significance for the health status management of liquid metal battery. Attached Figure Description

[0018] Figure 1 This is a flowchart of a parameter identification method for an equivalent circuit model of a liquid metal battery provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart of another parameter identification method for an equivalent circuit model of a liquid metal battery provided in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the second-order Thevenin equivalent circuit model of a liquid metal battery provided in Embodiment 1 of the present invention; Figure 4 This is a discharge curve of a liquid metal battery provided in Embodiment 1 of the present invention; Figure 5 This is a comparison chart of the simulated discharge curve and the actual discharge curve generated by the equivalent circuit model parameters identified by the multi-swarm particle optimization algorithm in Embodiment 1 of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0020] Example 1 This embodiment provides a parameter identification method for an equivalent circuit model of a liquid metal battery, such as... Figure 1 and Figure 2 As shown, it includes the following steps.

[0021] S1: Limit the parameter search range corresponding to the parameter combination to be identified in the equivalent circuit model of liquid metal battery; the parameter combination to be identified includes: battery ohmic internal resistance R0, battery polarization internal resistance R1 and battery polarization capacitance C1. Figure 3 This is a schematic diagram of the second-order Thevenin equivalent circuit model of a liquid metal battery.

[0022] S2: Create population A and population B corresponding to the parameter search range, initialize the initial position and initial velocity of each particle in each population within the parameter search range, and each particle corresponds to a combination of identification parameters.

[0023] It should be noted that the multi-objective swarm algorithm can be initialized first. The main parameters of the algorithm include population size, number of particles in each population (swarmSize), total dimension (dim), maximum iteration (maxIter), communication interval (exInter), and PSO parameters (including inertia weight w, individual learning factor c1, and social learning factor c2).

[0024] Furthermore, the initial position and initial velocity of each particle in each population within the parameter search range include: randomly setting the initial position and setting the initial velocity to 0.

[0025] S3: Calculate the current fitness of all particles in the current state; compare the current fitness of all particles, and take the current position and current fitness value of the particle with the smallest fitness as the current global optimum; wherein, the current fitness is: the deviation between the output voltage of the liquid metal battery equivalent circuit model generated according to the combination of parameters to be identified for the particle and the measured voltage.

[0026] Furthermore, the current fitness is: the output voltage U of the liquid metal battery equivalent circuit model generated based on the combination of parameters to be identified corresponding to the particles. sim With the measured voltage U test The root mean square error between them. Specifically, the current fitness is expressed as: ; Where N is the number of sample data points involved in the calculation, and RMSE represents the root mean square error.

[0027] Furthermore, updating the current position based on the current velocity includes: updating the current position of the particle based on the updated current velocity of the particle, while ensuring that the position of the particle does not exceed the parameter boundary.

[0028] S4: Determine if the iteration stopping condition is met. If it is met, proceed to S5. If not, update the current velocity of each particle in each population based on the current global optimal value, update the current position based on the current velocity, and return to S3.

[0029] S5: Swap the current global optimum values ​​of populations A and B. Specifically, replace the particle with the highest global fitness in population B with the global optimum particle in population A, and vice versa. Then, calculate the current fitness of all particles in the current state to find the current global optimum value for each population. Output the combination of identification parameters of the particle with the smaller current global optimum value as the target identification result.

[0030] It should be noted that the experimental parameters in Example 1 used a 20Ah Li||Sb-Sn liquid metal battery. The entire battery assembly is encapsulated inside a stainless steel casing, and the positive and negative current collectors are physically separated by a ceramic insulator. Preferably, the battery assembly encapsulation structure includes an extension rod to fix the insulator to the outside of the high-temperature furnace. This design effectively eliminates the influence of external factors on battery aging, thereby ensuring that the aging process is dominated only by the internal mechanisms of the battery.

[0031] Specifically, this experiment used the LANHE testing system (model: CT5001A) to conduct charge-discharge cycle tests on the liquid metal battery. The 20Ah Li||Sb-Sn liquid metal battery was placed in an electric furnace to avoid the influence of temperature changes on the experimental results. The charge-discharge cycle experiment of the liquid metal battery was conducted according to the following steps: (1) Charge the battery to full charge using constant current and constant voltage mode (cutoff voltage: 1.2 V, cutoff current: 0.2 A); (2) Let the battery stand for 1 h; (3) Discharge the battery with a constant current of 2.0 A until the voltage drops to 0.7 V; (4) Let it stand for 0.5 h; (5) Charge at a constant current of 2.0 A until the cutoff voltage reaches 1.2 V; (6) Let it stand for another 0.5 h; (7) Repeat steps (3) to (6).

[0032] Figure 4 The discharge curve of the liquid metal battery provided by this invention is shown. The parameter identification method for the equivalent circuit model of the liquid metal battery provided in this embodiment includes the following steps: (1) Initialize the algorithm parameters, including the number of particles in each population swarmSize, the maximum number of iterations maxIter, the communication interval exInter, the inertia weight w, the individual learning factor c1, and the social learning factor c2. (2) Set upper and lower limits for parameters (3) Initialize the population (4) Calculate the objective function value (5) Update particle velocity and position (6) Determine whether the communication interval has been reached. If yes, proceed to step (7); otherwise, proceed to step (4). (7) Inter-population communication, replacing the worst-performing particle. (8) Repeat steps (4) through (7). (9) Determine if the maximum number of iterations has been reached. If yes, proceed to step (10); otherwise, proceed to step (8). (10) Output results.

[0033] For example, the number of particles in each population (swarmSize) can be set to 50, the maximum iteration (maxIter) to 100, the communication interval (exInter) to 20, the inertia weight (w) to 0.9, the individual learning factor (c1) to 2, and the social learning factor (c2) to 1. Furthermore, the parameter range can be set as follows: , , .

[0034] Figure 5 A comparison of simulated discharge curves and actual discharge curves generated for the equivalent circuit model parameters is shown in the figure. As can be seen, the simulated voltage curve generated by the parameters identified by the multi-swarm particle optimization algorithm basically coincides with the measured voltage curve. This result verifies the effectiveness and accuracy of the multi-swarm particle optimization algorithm in estimating the parameters of the equivalent circuit model of liquid metal batteries.

[0035] Example 2 This embodiment provides a parameter identification device for an equivalent circuit model of a liquid metal battery, including: The range limiting module is used to limit the parameter search range corresponding to the combination of parameters to be identified in the equivalent circuit model of liquid metal battery; The population creation module is used to create population A and population B corresponding to the parameter search range, and initialize the initial position and initial velocity of each particle in each population within the parameter search range, with each particle corresponding to a combination of identification parameters. The update iteration module is used to calculate the current fitness of all particles in the current state; compare the current fitness of all particles, and take the current position and current fitness value of the particle with the smallest fitness as the current global optimum; wherein, the current fitness is: the deviation between the output voltage of the liquid metal battery equivalent circuit model generated according to the combination of parameters to be identified for the particle and the measured voltage; The condition judgment module is used to determine whether the iteration stopping condition is met. If it is met, the module enters the target output module. If it is not met, the module updates the current velocity of each particle in each population based on the current global optimal value, updates the current position based on the current velocity, and then enters the update iteration module. The target output module is used to exchange the current global optimum values ​​of population A and population B, calculate the current fitness of all particles in the current state, and thus find the current global optimum values ​​of population A and population B respectively; the identification parameters of the particles corresponding to the smaller current global optimum value are combined as the target identification result and output.

[0036] Example 3 This embodiment provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the parameter identification method for the equivalent circuit model of the liquid metal battery.

[0037] The electronic device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The memory can be used to store computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory.

[0038] Example 4 The present invention also relates to a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0039] Specifically, the memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0040] Example 5 This invention provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the method described in the above embodiments of this invention.

[0041] The technical features of the embodiments described above can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. It should be noted that the terms "in one embodiment," "for example," and "again" in this invention are intended to illustrate the invention and are not intended to limit the invention.

[0042] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A parameter identification method of a liquid metal battery equivalent circuit model, characterized by, The method comprises the following steps: S1: limiting a parameter search range corresponding to a to-be-identified parameter combination of an equivalent circuit model of a liquid metal battery; S2: creating a population A and a population B corresponding to the parameter search range, initializing an initial position and an initial speed of each particle in each population within the parameter search range, and each particle corresponding to an identified parameter combination; S3: calculating a current fitness of all particles in a current state; comparing the current fitness of all particles, and taking the current position and the current fitness value of the particle corresponding to the minimum fitness as a current global optimal value; wherein the current fitness is a deviation between an output voltage of the equivalent circuit model of the liquid metal battery generated according to the to-be-identified parameter combination of the particle and a measured voltage; S4: judging whether an iteration stop condition is met, if yes, entering S5; if not, updating the current speed of each particle in each population based on the current global optimal value, updating the current position according to the current speed, and returning to S3; S5: exchanging the current global optimal values of the population A and the population B, calculating the current fitness of all particles in the current state, and finding the current global optimal values of the population A and the population B; and outputting the identified parameter combination of the particle corresponding to the smaller current global optimal value as a target identified result.

2. The parameter identification method of a liquid metal battery equivalent circuit model of claim 1, wherein, The to-be-identified parameter combination comprises a battery ohmic resistance R0, a battery polarization resistance R1 and a battery polarization capacitance C1.

3. The parameter identification method of a liquid metal battery equivalent circuit model of claim 1, wherein, The current fitness is the root mean square error between the output voltage U of the liquid metal battery equivalent circuit model generated according to the parameter combination to be recognized corresponding to the particle sim and the measured voltage U test .

4. The parameter identification method of the equivalent circuit model of a liquid metal battery according to claim 3, characterized in that, The current fitness is expressed as: ； wherein N is a sample data point number participating in calculation, and RMSE represents a root mean square error.

5. The parameter identification method of a liquid metal battery equivalent circuit model of claim 1, wherein, The exchanging the current global optimal values of the population A and the population B comprises: replacing the global optimal particle in the population A with the global fitness maximum particle in the population B, and replacing the global optimal particle in the population B with the global fitness maximum particle in the population A.

6. The parameter identification method of a liquid metal battery equivalent circuit model of claim 1, wherein, The updating the current position according to the current speed comprises: performing current position updating on the particle based on the current speed of the particle obtained after updating, under the condition that the position of the particle is not more than a parameter boundary.

7. The parameter identification method of a liquid metal battery equivalent circuit model of claim 1, wherein, The initializing the initial position and the initial speed of each particle in each population within the parameter search range comprises: randomly setting the initial position, and setting the initial speed as 0.

8. A parameter identification device for a liquid metal battery equivalent circuit model, characterized by The method comprises the following steps: a range limiting module, configured to limit a parameter search range corresponding to a to-be-identified parameter combination of an equivalent circuit model of a liquid metal battery; a population creating module, configured to create a population A and a population B corresponding to the parameter search range, and initialize an initial position and an initial speed of each particle in each population within the parameter search range, and each particle corresponding to an identified parameter combination; an updating iteration module, configured to calculate a current fitness of all particles in a current state; compare the current fitness of all particles, and take the current position and the current fitness value of the particle corresponding to the minimum fitness as a current global optimal value; wherein the current fitness is a deviation between an output voltage of the equivalent circuit model of the liquid metal battery generated according to the to-be-identified parameter combination of the particle and a measured voltage; A condition judging module is configured to judge whether an iteration stopping condition is met, and if yes, enter a target output module; if not, update a current speed of each particle in each population based on a current global optimal value, update a current position according to the current speed, and enter an iteration updating module; The target output module is configured to exchange the current global optimal values of the population A and the population B, calculate current fitness of all particles in a current state, so as to find the current global optimal values of the population A and the population B, and output a combination of identification parameters of a particle corresponding to a smaller current global optimal value as a target identification result. 9.An electronic device comprising a memory and a processor, the memory storing a computer program, wherein, The processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by the processor, implements the steps of the method of any one of claims 1 to 7.

Images