Method, device, equipment and storage medium for estimating remaining battery life

By acquiring the battery status signal, particle filtering sampling and posterior probability density calculation, combined with CEEMD decomposition and resampling technology, the particle degradation problem in the remaining service life estimation of the battery is solved, and prediction accuracy and performance are improved.

CN117825971BActive Publication Date: 2024-11-29TOPAK POWER TECH CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202410165244.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-05
Publication Date
2024-11-29
Estimated Expiration
2044-02-05

AI Technical Summary

Technical Problem

In the existing battery residual service life estimation method, the prediction accuracy is insufficient during particle filtering length, and there are serious particle degradation problems.

Method used

By obtaining the battery status signal, performing particle filtering sampling, calculating the posterior probability density, combining CEEMD decomposition and resampling technology, optimizing particle weights, improving the accuracy of particle filtering, and improving the accuracy of battery capacity estimation.

Benefits of technology

It improves the prediction accuracy of the remaining service life of the battery, reduces the problem of particle degradation, and improves the prediction performance of the model.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117825971B_ABST
    Figure CN117825971B_ABST
Patent Text Reader

Abstract

The present invention discloses a method, device, equipment and storage medium for estimating the remaining service life of a battery, and belongs to the field of batteries. The present invention obtains a battery status signal to obtain a particle filter; selects a cycle period, samples the particle filter, and obtains a plurality of particles; obtains an observation value based on the relationship between the particle state and the observation value of the collected particles; corrects the observation value to obtain a posterior probability density; obtains a battery capacity estimation value based on the cycle period and the posterior probability density; and calculates the remaining service life of the battery based on the battery capacity estimation value. The RUL prediction of the present invention can improve the particle degradation problem, and the result has higher accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of batteries, and in particular to a method, device, equipment and storage medium for estimating the remaining service life of a battery. Background Art

[0002] Currently, the remaining service life of batteries is usually estimated using the PF algorithm. The PF algorithm has extremely strong modeling capabilities for non-Gaussian nonlinear problems, but the PF algorithm has serious particle degradation problems.

[0003] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of the present invention is to provide a method, device, equipment and storage medium for estimating the remaining service life of a battery, aiming to solve the problem of insufficient long-term prediction accuracy of particle filtering.

[0005] To achieve the above object, the present invention provides a method for estimating the remaining service life of a battery, obtaining a battery status signal and obtaining a particle filter;

[0006] Select a cycle period, sample the particle filter, and obtain multiple particles;

[0007] The sampled particles are used to obtain observation values ​​based on the relationship between the particle states and the observation values;

[0008] Correcting the observed value to obtain a posterior probability density;

[0009] Obtaining a battery capacity estimation value according to the cycle period and the posterior probability density;

[0010] The remaining service life of the battery is calculated according to the estimated value of the battery capacity.

[0011] Optionally, before selecting a cycle period and sampling the particle filter to obtain a plurality of particles, the method further includes:

[0012] Obtain process noise and observation noise in particle filtering;

[0013] According to the process noise and the functional relationship between the current state and the previous state, the system state equation is obtained;

[0014] The system observation equation is obtained according to the particle state and observation noise;

[0015] A particle filter system state model is constructed according to the state equation and the observation equation.

[0016] Optionally, the modifying the observed value to obtain a posterior probability density includes:

[0017] Get the current state of the sampled particles;

[0018] Predicting a priori probability density according to the particle state through the system state model;

[0019] According to the current cycle state of the particle and the functional relationship between the observation noise and the observation value, the current observation value is obtained;

[0020] The observed value is combined with the prior probability density for correction to obtain the posterior probability density.

[0021] Optionally, after the observation value is corrected to obtain the posterior probability density, the method further includes:

[0022] Perform sequence importance sampling on multiple particles;

[0023] Calculate the number of valid sample particles and resample to obtain valid particles.

[0024] Optionally, the performing sequence importance sampling on the plurality of particles comprises:

[0025] A plurality of collected particles are randomly selected from a known probability density distribution;

[0026] The cycle period of the particle is calculated by weight, and the importance weights of multiple particles are calculated and normalized.

[0027] Optionally, the calculating the number of valid sample particles and performing resampling to obtain valid particles includes:

[0028] Delete particles whose particle weights are lower than a preset threshold, and obtain a particle weight sequence consisting of the remaining particles;

[0029] Adding a set of white noise sequences in the form of positive and negative pairs of equal length to the particle weight sequence to obtain two sets of new signals;

[0030] Decomposing the two groups of new signals by CEEMD to obtain different intrinsic mode components;

[0031] Superimposing multiple layers of intrinsic mode components on the decomposed new signal to obtain a particle weight signal;

[0032] The particle weight signal is normalized and then resampled in layers to obtain particles with high weights.

[0033] Optionally, obtaining a battery capacity estimation value according to the cycle period and the posterior probability density includes:

[0034] Set multiple preset parameter points;

[0035] Substituting the state value and capacitance of the previous cycle of the selected cycle and the posterior probability density into the state equation and observation equation of the double exponential battery degradation model to obtain the state value of the selected cycle;

[0036] Obtain the state value and observation noise corresponding to the preset parameter point;

[0037] The battery capacity prediction value is obtained according to the state value corresponding to the preset parameter point and the observed noise.

[0038] In addition, to achieve the above object, the present invention also provides a battery remaining service life estimation device, the device comprising:

[0039] An acquisition module is used to acquire a battery status signal and obtain a particle filter;

[0040] The acquisition module is also used to select the cycle period, sample the particle filter, and obtain multiple particles;

[0041] The acquisition module is also used to input the sampled particles into the system observation equation to obtain the observed value;

[0042] The acquisition module is also used to correct the observed value to obtain the posterior probability density;

[0043] The result module is further used to obtain a battery capacity estimation value according to the cycle period and the posterior probability density;

[0044] A calculation module is used to calculate the remaining service life of the battery according to the estimated value of the battery capacity.

[0045] In addition, to achieve the above-mentioned purpose, the present invention also provides a battery remaining life estimation device, which includes: a memory, a processor, and a battery remaining life estimation program stored in the memory and executable on the processor, and the battery remaining life estimation program is configured to implement the steps of the battery remaining life estimation method.

[0046] In addition, to achieve the above-mentioned purpose, the present invention also provides a storage medium, on which a battery remaining service life estimation program is stored, and when the battery remaining service life estimation program is executed by a processor, the steps of the battery remaining service life estimation method are implemented.

[0047] The present invention acquires a battery status signal to obtain a particle filter; selects a cycle period, samples the particle filter, and obtains a plurality of particles; obtains an observation value based on the relationship between the particle state and the observation value of the sampled particles; corrects the observation value to obtain a posterior probability density; obtains a battery capacity estimation value based on the cycle period and the posterior probability density; and calculates the remaining service life of the battery based on the battery capacity estimation value. The RUL prediction of the present invention can improve the particle degradation problem, and the result has higher accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a structural schematic diagram of a battery remaining service life estimation device in a hardware operating environment involved in an embodiment of the present invention;

[0049] Figure 2 A schematic diagram of a flow chart of a first embodiment of estimating the remaining service life of a battery according to the present invention;

[0050] Figure 3 A schematic diagram of a flow chart of a second embodiment of estimating the remaining service life of a battery according to the present invention;

[0051] Figure 4 A schematic diagram of a flow chart of a third embodiment of estimating the remaining service life of a battery according to the present invention;

[0052] Figure 5 A schematic diagram of the recursive prediction performance of estimating the remaining service life of a battery according to the present invention;

[0053] Figure 6 A schematic diagram of the long-term prediction performance of estimating the remaining service life of a battery of the present invention;

[0054] Figure 7 It is a structural block diagram of the first embodiment of the battery remaining service life estimation device of the present invention.

[0055] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0056] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0057] Reference Figure 1 , Figure 1 The present invention is a schematic diagram of the structure of a device for estimating the remaining battery life in a hardware operating environment according to an embodiment of the present invention.

[0058] like Figure 1As shown, the battery remaining service life estimation device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0059] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the battery remaining service life estimating device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0060] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a data storage module, a network communication module, a user interface module, and a battery remaining service life estimation device program.

[0061] exist Figure 1 In the battery remaining life estimation device shown, the network interface 1004 is mainly used for data communication with other devices; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the battery remaining life estimation device of the present invention can be set in the battery remaining life estimation device, and the battery remaining life estimation device calls the battery remaining life estimation program stored in the memory 1005 through the processor 1001, and executes the battery remaining life estimation method provided by the embodiment of the present invention.

[0062] The present invention provides a method for estimating the remaining service life of a battery. Figure 2 , Figure 2 The figure is a flow chart of a first embodiment of a method for estimating the remaining service life of a battery according to the present invention.

[0063] In this embodiment, the remaining battery life estimation method includes:

[0064] Step S10: Obtain a battery status signal and obtain a particle filter.

[0065] Step S20: Select a cycle period, sample the particle filter, and obtain a plurality of particles.

[0066] It is important to understand that according to Bayesian estimation theory, the state estimation problem can be solved based on the existing data series. Estimate the current system status The credibility of Solve for the posterior probability density This is achieved through recursive calculation of prediction step and update step. The prediction step is to use the state equation to predict the prior probability density of the system state. , that is, by estimating the known system state Estimated status The update step uses the latest observation value to correct the prior probability density and obtain the posterior probability density .

[0067] It is understandable that since the Bayesian estimate of the posterior probability The integral needs to be calculated, but the integral calculation is too difficult, which may result in no analytical solution for the posterior probability. Therefore, Monte Carlo sampling is used to approximate the integral calculation, that is, the calculated value is used to approximate the integral, thereby obtaining an approximate solution for the posterior probability.

[0068] Specifically, assuming that Medium Sampling The method further comprises the following steps: obtaining process noise and observation noise in the particle filter; obtaining a system state equation according to the process noise and the functional relationship between the current state and the previous state; obtaining a system observation equation according to the particle state and the observation noise; and constructing a particle filter system state model according to the state equation and the observation equation.

[0069] Step S30: Obtain observation values ​​of the sampled particles according to the relationship between the particle states and the observation values.

[0070] Step S40: Correct the observed value to obtain the posterior probability density.

[0071] Specifically, the observation value is corrected to obtain the posterior probability density, including: obtaining the current state of the sampled particle; predicting the prior probability density according to the particle state through the system state model; obtaining the current observation value according to the current cycle state of the particle and the functional relationship between the observation noise and the observation value; and using the observation value combined with the prior probability density for correction to obtain the posterior probability density.

[0072] Step S50: Obtaining a battery capacity estimation value according to the cycle period and the posterior probability density.

[0073] Step S60: Calculate the remaining service life of the battery according to the estimated value of the battery capacity.

[0074] It should be noted that, whether the battery capacity estimate is less than a preset failure threshold of the battery is judged. If it is less than the preset failure threshold, the remaining service life of the battery is calculated, including: when the battery capacity estimate does not exceed the preset failure threshold, the remaining service life of the battery is calculated; when the battery capacity estimate exceeds the preset failure threshold, the cycle period plus the preset cycle is selected as a new cycle period, and the battery capacity estimate of the new cycle period is obtained through the CEPF algorithm, and so on, until the battery capacity estimate of the new cycle period does not exceed the preset failure threshold.

[0075] This embodiment obtains a battery status signal to obtain a particle filter; selects a cycle period, samples the particle filter, and obtains a plurality of particles; obtains an observation value based on the relationship between the particle state and the observation value of the sampled particles; corrects the observation value to obtain a posterior probability density; obtains a battery capacity estimation value based on the cycle period and the posterior probability density; and calculates the remaining service life of the battery based on the battery capacity estimation value. The RUL prediction of the present invention can improve the particle degradation problem, and the result has higher accuracy.

[0076] refer to Figure 3 , Figure 3 FIG. 4 is a flow chart of a second embodiment of a method for estimating remaining battery life according to the present invention.

[0077] After step S40 described in the first embodiment, the method further includes:

[0078] Step S401: performing sequence importance sampling on multiple particles.

[0079] It is important to understand that according to Bayesian estimation theory, the state estimation problem can be solved based on the existing data series. Estimate the current system status The credibility of Solve for the posterior probability density This is achieved through recursive calculation of prediction step and update step. The prediction step is to use the state equation to predict the prior probability density of the system state. , that is, by estimating the known system state Estimated status The update step uses the latest observation value to correct the prior probability density and obtain the posterior probability density .

[0080] It is understandable that since the Bayesian estimate of the posterior probability The integral needs to be calculated, but the integral calculation is too difficult, which may result in no analytical solution for the posterior probability. Therefore, Monte Carlo sampling is used to approximate the integral calculation, that is, the calculated value is used to approximate the integral, thereby obtaining an approximate solution for the posterior probability. Specifically, assuming that Medium Sampling samples, then for:

[0081]

[0082] in is the Dirac function. Assume that there is a function for solving specific practical problems. The expected solution formula is:

[0083]

[0084] The core idea of ​​the above expression is obtained by sampling The state values ​​calculated by the samples are averaged as the posterior filtering result. However, there is still a difficulty in Monte Carlo sampling, that is, the posterior probability is unknown, and it is obviously impossible to sample from an unknown probability distribution. Therefore, importance sampling needs to be introduced to solve this problem.

[0085] Specifically, the sequential importance sampling of multiple particles includes: randomly selecting multiple collected particles from a known probability density distribution; calculating the cycle period of the particles through weights, and calculating and normalizing the importance weights of the multiple particles.

[0086] Therefore, in order to reduce the computational complexity of weights, a weight derivation method needs to be introduced. Specifically, assuming is the importance probability density distribution, then the recursive form of the posterior probability distribution is:

[0087]

[0088] Step S402: Calculate the number of valid sample particles and perform resampling to obtain valid particles.

[0089] It should be understood that the calculation of the number of valid sample particles and resampling to obtain valid particles include: deleting particles whose particle weights are lower than a preset threshold to obtain a particle weight sequence consisting of the remaining particles; adding a set of white noise sequences in the form of positive and negative pairs of equal length to the particle weight sequence to obtain two groups of new signals; using CEEMD to decompose the two groups of new signals to obtain different intrinsic mode components; superimposing multiple layers of intrinsic mode components on the decomposed new signal to obtain a particle weight signal; normalizing the particle weight signal and then resampling it in layers to obtain high-weight particles.

[0090] It is understandable that in this embodiment, after the particle filter has been iterated for many times, some particles will have a few extremely large weights, while many particle weights become extremely small, and the variance between the weights is too large, which can be explained by the decrease in the number of effective particles in the particle swarm. A large number of invalid or inefficient particles have no effect on the posterior probability estimation. The traditional method is to resample the particles by breeding high-weight particles and discarding low-weight particles, thereby suppressing the particle degradation problem to a certain extent. According to this idea, the particles are resampled. After resampling, the particle set Updated to The method may cause the same high-weight particle to be reproduced multiple times. The traditional method weakens the particle diversity and affects the PF prediction performance. Therefore, the CEEMD algorithm can effectively solve the problem of modal aliasing, reduce the number of iterations required for modal decomposition, and reduce the computational cost.

[0091] In this embodiment, sequential importance sampling is performed on multiple particles; the number of effective sample particles is calculated, and resampling is performed to obtain effective particles, which can reduce the number of required iterations and reduce the calculation cost.

[0092] refer to Figure 4 , Figure 4 FIG. 4 is a flow chart of a third embodiment of a method for estimating remaining battery life according to the present invention.

[0093] Step S50 of the first embodiment includes:

[0094] Step S501: setting a plurality of preset parameter points.

[0095] It should be noted that, the selected cycle period is based on the relationship between the cycle period and the capacity. When the cycle period exceeds the preset starting point cycle period, the selected cycle period is substituted into the state equation and observation equation of the double exponential battery degradation model in the CEPF model to obtain the predicted value of the battery capacity.

[0096] In this embodiment, four parameter points a, b, c, and d are set, and then the double exponential battery degradation model is input for calculation according to the parameter point data.

[0097] Step S502: Substitute the state value and capacitance of the previous cycle of the selected cycle and the posterior probability density into the state equation and observation equation of the double exponential battery degradation model to obtain the state value of the selected cycle.

[0098] Step S503: Obtain the state value and observed noise corresponding to the preset parameter point.

[0099] In this embodiment, the particle weight signal after resampling is obtained; the state value and observation noise corresponding to the parameter point are obtained according to the particle weight and the relationship between the selected cycle period and the battery capacity through the state estimation equation of the CEPF algorithm.

[0100] Step S504: obtaining the battery capacity prediction value according to the state value corresponding to the preset parameter point and the observed noise.

[0101] Specifically, in this embodiment, the state equation can be expressed as:

[0102]

[0103] The observation equation is:

[0104]

[0105] in is the battery capacity value, The battery capacity value is obtained based on the state of the parameter point and the observed noise.

[0106] In this embodiment, four lithium-ion batteries (B5, B6, B7 and B18) of NASA's battery data set are selected for the experimental data set. For the B5 battery, data from the initial capacity of the battery to the decay to 85% of the capacity are selected as training samples, the remaining life cutoff point is the decay to 72% of the capacity, and the actual RUL is 65; the B6 battery selects data from the decay to 85% of the capacity as training samples, the remaining life cutoff point is the decay to 70% of the capacity, and the actual RUL is 50; the B7 battery selects data from the decay to 90% of the capacity as training samples, the remaining life cutoff point is 80%, and the actual RUL is 58; the B18 battery selects the first 90% as training samples, the remaining life cutoff point is 80%, and the actual RUL is 42.

[0107] Table 1 Evaluation of RUL recursive prediction performance based on CEPF

[0108]

[0109] First, according to Table 1 and Figure 5 It can be seen that in the first 25 cycles of the recursion, there is a large deviation between the predicted capacity and the actual capacity of the two algorithms. This is the initial value of the state parameter. It is determined comprehensively based on the initial values ​​of four groups of batteries. However, as the recursion continues, the influence of the initial value on the algorithm gradually weakens until it disappears. Secondly, for the PF and CEPF algorithms, by recursively predicting the RUL, that is, predicting the next battery capacity from the current step, the predicted results are extremely close to the actual results. The main error is at some battery capacity rebound points, such as the 20th and 31st cycles of the B5 battery. The actual normalized capacity is 0.9949 and 0.9975, respectively, and the PF predicted capacity is 0.9772 and 0.9780, respectively, while the CEPF predicted capacity is 0.9772, 0.9773 and 0.9781, respectively, which is slightly different from the actual capacity. Finally, the battery capacity prediction schematics of the PF and CEPF algorithms are not significantly different, but it can be observed that CEPF is closer to the actual capacity. As can be seen from the table, the prediction performance indicators of the CEPF algorithm are better than those of the PF algorithm.

[0110] Table 2 Evaluation of RUL long-term prediction performance based on CEPF

[0111]

[0112] From Table 2 and Figure 6 It can be seen that the CEPF estimate is close to the EF estimate, but the CEPF prediction is significantly better than the EF prediction. From the average values ​​of the four groups of batteries, the RMSE, MAE, MAPE, Er and PEr indicators of the CEPF algorithm are reduced by 22.52%, 23.44%, 22.19%, 30.11% and 33.03% respectively compared with the PF algorithm. The R2(%) difference is basically ignored because the actual envelopes of CEPF and PF are significantly different, and the R2(%) indicator has a weak evaluation ability.

[0113] The results of recursive prediction and long-term prediction show that the CEPF algorithm designed on the basis of the CEEMD algorithm can effectively correct the particle weight. The CEEMD method can improve the particle degradation problem, and the performance indicators predicted by the CEPF model are significantly better than those of the traditional PF.

[0114] In this embodiment, a plurality of preset parameter points are set; the state value and capacitance of the previous cycle of the selected cycle and the posterior probability density are substituted into the state equation and the observation equation of the double exponential battery degradation model to obtain the state value of the selected cycle; the state value and the observation noise corresponding to the preset parameter points are obtained; the battery capacity prediction value is obtained according to the state value and the observation noise corresponding to the preset parameter points, which improves the problem of particle degradation and further improves the model test performance.

[0115] refer to Figure 7 , Figure 7 It is a structural block diagram of the first embodiment of the battery remaining life estimation device of the present invention.

[0116] like Figure 7 As shown, the battery remaining life estimation device proposed in the embodiment of the present invention includes:

[0117] An acquisition module 10 is used to acquire a battery status signal and obtain a particle filter;

[0118] The acquisition module 10 is also used to select a cycle period, sample the particle filter, and obtain multiple particles;

[0119] The acquisition module 10 is further used to input the sampled particles into the system observation equation to obtain the observed value;

[0120] The acquisition module 10 is also used to correct the observed value to obtain the posterior probability density;

[0121] The result module 20 is further used to obtain a battery capacity estimation value according to the cycle period and the posterior probability density;

[0122] The calculation module 30 is used to calculate the remaining service life of the battery according to the estimated value of the battery capacity.

[0123] This embodiment obtains a battery status signal to obtain a particle filter; selects a cycle period, samples the particle filter, and obtains a plurality of particles; obtains an observation value based on the relationship between the particle state and the observation value of the sampled particles; corrects the observation value to obtain a posterior probability density; obtains a battery capacity estimation value based on the cycle period and the posterior probability density; and calculates the remaining service life of the battery based on the battery capacity estimation value. The RUL prediction of the present invention can improve the particle degradation problem, and the result has higher accuracy.

[0124] In one embodiment, the acquisition module 10 is also used to acquire process noise and observation noise in particle filtering; obtain the system state equation based on the process noise and the functional relationship between the current state and the previous state; obtain the system observation equation based on the particle state and the observation noise; and construct a particle filter system state model based on the state equation and the observation equation.

[0125] In one embodiment, the acquisition module 10 is also used to obtain the current state of the sampled particles; predict the prior probability density based on the particle state through the system state model; obtain the current observation value based on the current cycle state of the particle and the functional relationship between the observation noise and the observation value; use the observation value combined with the prior probability density for correction to obtain the posterior probability density.

[0126] In one embodiment, the acquisition module 10 is further used to perform sequential importance sampling on multiple particles; calculate the number of valid sample particles, and perform resampling to obtain valid particles.

[0127] In one embodiment, the acquisition module 10 is further used to randomly select a plurality of collected particles from a known probability density distribution; calculate the cycle period of the particles through weight calculation, and calculate and normalize the importance weights of the plurality of particles.

[0128] In one embodiment, the acquisition module 10 is also used to delete particles whose particle weights are lower than a preset threshold value to obtain a particle weight sequence composed of the remaining particles; add a set of white noise sequences in the form of positive and negative pairs of equal length to the particle weight sequence to obtain two sets of new signals; use CEEMD to decompose the two sets of new signals to obtain different intrinsic mode components; superimpose multiple layers of intrinsic mode components on the decomposed new signals to obtain particle weight signals; normalize the particle weight signals and then resample them in layers to obtain high-weight particles.

[0129] In one embodiment, the result module 20 is also used to set multiple preset parameter points; substitute the state value and capacitance of the previous cycle of the selected cycle and the posterior probability density into the state equation and observation equation of the double exponential battery degradation model to obtain the state value of the selected cycle; obtain the state value and observation noise corresponding to the preset parameter point; and obtain the battery capacity prediction value according to the state value and observation noise corresponding to the preset parameter point.

[0130] In addition, to achieve the above-mentioned purpose, the present invention also proposes a battery remaining life estimation device, which includes: a memory, a processor, and a battery remaining life estimation program stored in the memory and executable on the processor, and the battery remaining life estimation program is configured to implement the steps of the battery remaining life estimation method described above.

[0131] Since the battery remaining life estimation device adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.

[0132] In addition, an embodiment of the present invention further proposes a storage medium, on which a battery remaining life estimation program is stored. When the battery remaining life estimation program is executed by a processor, the steps of the battery remaining life estimation method described above are implemented.

[0133] Since the storage medium adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.

[0134] It should be understood that the above is only an example and does not constitute any limitation on the technical solution of the present invention. In specific applications, technicians in this field can make settings as needed, and the present invention does not limit this.

[0135] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present invention. In practical applications, technicians in this field can select part or all of them according to actual needs to achieve the purpose of the present embodiment, and no limitation is made here.

[0136] In addition, for technical details that are not described in detail in this embodiment, reference can be made to the engineering mode control method provided in any embodiment of the present invention, and will not be repeated here.

[0137] In addition, it should be noted that, in this article, the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of more restrictions, an element defined by the sentence "comprising a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0138] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0139] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0140] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for estimating remaining battery life, It is characterized in that The method comprises: Obtain battery status signal and obtain particle filter; Select a cycle period, sample the particle filter, and obtain multiple particles; The sampled particles are used to obtain observation values ​​based on the relationship between the particle states and the observation values; Correcting the observed value to obtain a posterior probability density; After the observed value is corrected to obtain the posterior probability density, the method further includes: Perform sequence importance sampling on multiple particles; Calculate the number of valid sample particles, perform resampling, and obtain valid particles; The calculating the number of valid sample particles and performing resampling to obtain valid particles includes: Delete particles whose particle weights are lower than a preset threshold, and obtain a particle weight sequence consisting of the remaining particles; Adding a set of white noise sequences in the form of positive and negative pairs of equal length to the particle weight sequence to obtain two sets of new signals; Decomposing the two groups of new signals by CEEMD to obtain different intrinsic mode components; Superimposing multiple layers of intrinsic mode components on the decomposed new signal to obtain a particle weight signal; Normalizing the particle weight signal and then resampling it in layers to obtain particles with high weight; Obtaining a battery capacity estimation value according to the cycle period and the posterior probability density; The remaining service life of the battery is calculated according to the estimated value of the battery capacity.

2. The method according to claim 1, It is characterized in that The selecting cycle period, sampling the particle filter, and obtaining a plurality of particles also includes: Obtain process noise and observation noise in particle filtering; According to the process noise and the functional relationship between the current state and the previous state, the system state equation is obtained; The system observation equation is obtained according to the particle state and observation noise; A particle filter system state model is constructed according to the state equation and the observation equation.

3. The method according to claim 2, It is characterized in that The step of correcting the observed value to obtain a posterior probability density includes: Get the current state of the sampled particles; Predicting a priori probability density according to the particle state through the system state model; According to the current cycle state of the particle and the functional relationship between the observation noise and the observation value, the current observation value is obtained; The observed value is combined with the prior probability density for correction to obtain the posterior probability density.

4. The method according to claim 1, It is characterized in that The sequential importance sampling of multiple particles includes: A plurality of collected particles are randomly selected from a known probability density distribution; The cycle period of the particle is calculated by weight, and the importance weights of multiple particles are calculated and normalized.

5. The method according to claim 1, It is characterized in that The step of obtaining a battery capacity estimation value according to the cycle period and the posterior probability density includes: Set multiple preset parameter points; Substituting the state value and capacitance of the previous cycle of the selected cycle and the posterior probability density into the state equation and observation equation of the double exponential battery degradation model to obtain the state value of the selected cycle; Obtain the state value and observation noise corresponding to the preset parameter point; The battery capacity prediction value is obtained according to the state value corresponding to the preset parameter point and the observed noise.

6. A device for estimating the remaining service life of a battery, It is characterized in that The device comprises: The acquisition module is used to acquire the battery status signal and obtain the particle filter; select the cycle period, sample the particle filter, and obtain multiple particles; input the sampled particles into the system observation equation to obtain the observation value; correct the observation value to obtain the posterior probability density; after correcting the observation value to obtain the posterior probability density, it also includes: sampling the multiple particles by order importance; calculating the number of effective sample particles, resampling, and obtaining effective particles; calculating the number of effective sample particles, resampling, and obtaining effective particles, including: deleting particles whose particle weights are lower than a preset threshold, and obtaining a particle weight sequence composed of the remaining particles; adding a group of white noise sequences in the form of positive and negative pairs of equal length to the particle weight sequence to obtain two groups of new signals; using CEEMD to decompose the two groups of new signals to obtain different intrinsic mode components; superimposing multiple layers of intrinsic mode components on the decomposed new signal to obtain a particle weight signal; normalizing the particle weight signal and then resampling in layers to obtain high-weight particles; A result module, used for obtaining a battery capacity estimation value according to the cycle period and the posterior probability density; A calculation module is used to calculate the remaining service life of the battery according to the estimated value of the battery capacity.

7. A battery remaining service life estimation device, It is characterized in that The device comprises: a memory, a processor, and a battery remaining service life estimation program stored in the memory and executable on the processor, wherein the battery remaining service life estimation program is configured to implement the steps of the battery remaining service life estimation method as described in any one of claims 1 to 5.

8. A storage medium, It is characterized in that The storage medium stores a battery remaining service life estimation program, and when the battery remaining service life estimation program is executed by the processor, the steps of the battery remaining service life estimation method according to any one of claims 1 to 5 are implemented.

Citation Information

Patent Citations

  • Lithium ion battery residual life prediction method based on fusion of improved particle filtering and double-exponential recession empirical physical model

    CN110457789A

  • Lithium ion battery remaining service life prediction method

    CN112415414A

  • Lithium ion residual life prediction method based on improved particle filter model

    CN116680983A

  • Method and System for Predicting Useful Life of a Rechargeable Battery

    US20150349385A1