Search

CN-122017642-A - Lithium battery health state estimation and correction method and system based on layered fusion

CN122017642ACN 122017642 ACN122017642 ACN 122017642ACN-122017642-A

Abstract

The invention discloses a lithium battery health state online estimation and correction method and system based on layered fusion, and belongs to the technical field of battery management systems. The method comprises the steps of inquiring a life table according to historical operation data of a battery by an offline priori layer to calculate initial maximum available capacity, cooperatively estimating SOC and maximum available capacity by a main-auxiliary double-expansion Kalman filter by an online estimation layer to output an online health state estimation value, calculating actual capacity by a two-point method when a high confidence coefficient condition is met by a strong correction layer to output a strong correction health state value, carrying out weighted fusion by a fusion decision layer according to confidence coefficient to output a final health state value, and feeding back the actual capacity to the auxiliary-expansion Kalman filter to carry out state reset. The invention solves the problems of difficult coupling estimation of the charge state and the maximum available capacity, long-term drift estimation on line and insufficient working condition adaptability, and has high precision, high robustness and good engineering practicability.

Inventors

  • QIAN XINXIN

Assignees

  • 孝感楚能新能源创新科技有限公司

Dates

Publication Date
20260512
Application Date
20260128

Claims (10)

  1. 1. The lithium battery health state estimation and correction method based on layered fusion is characterized by comprising the following steps of: step S1, an offline priori layer queries a pre-stored life table according to battery historical operation data to calculate initial maximum available capacity The initial maximum available capacity Transmitting to an online estimation layer; step S2, the auxiliary extended Kalman filter of the online estimation layer uses the initial maximum available capacity For initial value, estimate maximum available capacity The master extended Kalman filter utilizes the maximum available capacity Estimating state of charge (SOC), updating the maximum available capacity by the auxiliary extended Kalman filter by using the SOC According to the maximum available capacity And rated capacity Outputs the ratio of the on-line health state estimated value ; Step S3, the strong correction layer monitors the battery operation condition, and when the high confidence coefficient condition is met, the actual capacity is calculated through a two-point method According to the actual capacity And the rated capacity Outputs a strongly corrected state of health value ; S4, the fusion decision layer is used for estimating the value according to the online health state And the strongly corrected state of health value The confidence coefficient of the (B) is weighted and fused to output the final health state value And the actual capacity is calculated Feeding back to the auxiliary extended Kalman filter, resetting the maximum available capacity 。
  2. 2. The method of claim 1, wherein in step S1, the battery historical operating data includes a historical average temperature, a historical average state of charge interval, and a cumulative equivalent number of cycles, and the life table includes a calendar life table and a cycle life table.
  3. 3. The method of claim 1, wherein the state equation of the main extended kalman filter is: ; ; Wherein, the Is the SOC value at the kth time, The period of the sampling is indicated and, The coulomb efficiency is indicated as such, The current at the time of the k-th moment is indicated, For the maximum capacity at time k-1, For the terminal voltage of the RC loop at time k, Represents the internal polarization resistance, τ represents the time constant, Representing process noise.
  4. 4. The method according to claim 1, wherein in step S2, the secondary extended kalman filter applies the maximum available capacity Modeling is a random walk process, and the state equation is as follows: ; Wherein, the And the observation equation of the auxiliary extended Kalman filter is based on an ampere-hour integration principle, and the state of charge SOC output by the main extended Kalman filter is utilized to construct a virtual observed quantity.
  5. 5. The method of claim 4, wherein the observation equation for the secondary extended kalman filter is: ; Wherein, the And The state of charge estimates output by the main extended kalman filter at the kth time and the kth-1 time respectively, The current value at the kth time is, For the sampling period, η is the coulombic efficiency, For the maximum capacity at time k-1, The observed noise covariance of the secondary extended Kalman filter is associated with the state of charge estimation error covariance of the primary extended Kalman filter.
  6. 6. The method according to claim 1, wherein in step S3, the high confidence condition comprises at least one of the following conditions: condition A, constant current charging stage, initial state of charge is lower than preset low threshold Charged to a cut-off voltage ; A constant-current discharge stage, in which the initial charge state is full charge state, and the discharge is carried out to cut-off voltage Or the state of charge is below the preset low threshold 。
  7. 7. The method according to claim 6, wherein in step S3, the two-point method calculates the actual capacity The formula of (2) is: ; Wherein, the Indicating the total throughput of the charge or discharge segment, And The states of charge at the start and end of the segment are indicated, respectively.
  8. 8. The method according to claim 1, wherein in step S4, the formula of the weighted fusion is: ; ; Wherein, the Indicating the value of the final state of health, The fusion weights are represented as such, Representing strongly corrected state of health values Is used to determine the confidence coefficient of the (c) in the (c), Representing an online health state estimate Confidence coefficient of (c).
  9. 9. The method according to claim 1, wherein in step S4, the resetting comprises moving the state variable of the secondary extended Kalman filter from the current maximum available capacity Reset to the actual capacity And resetting the error covariance matrix of the auxiliary extended Kalman filter to a preset minimum value.
  10. 10. A lithium battery state of health estimation and correction system based on hierarchical fusion, adapted to the method of any one of claims 1 to 9, comprising: the offline priori layer module is used for inquiring a pre-stored life table according to the historical operation data of the battery and calculating the initial maximum available capacity And the initial maximum available capacity Transmitting to an online estimation layer module; An online estimation layer module comprising a primary extended Kalman filter and a secondary extended Kalman filter, wherein the secondary extended Kalman filter has the initial maximum available capacity Estimating maximum available capacity for initial value The main extended Kalman filter utilizes the maximum available capacity Estimating a state of charge, SOC, the secondary extended Kalman filter updating the maximum available capacity with the state of charge, SOC According to the maximum available capacity And rated capacity Is used for outputting the on-line health state estimated value ; The strong correction layer module is used for monitoring the battery operation condition and calculating the actual capacity by a two-point method when the high confidence coefficient condition is met According to the actual capacity And the rated capacity Is a ratio of the output of the (d) to the (d) output of the (d) output ; A fusion decision layer module for estimating the value according to the online health state And the strongly corrected state of health value The confidence coefficient of the (B) is weighted and fused to output the final health state value And the actual capacity is calculated Resetting the maximum available capacity by feeding back the auxiliary extended Kalman filter to the online estimation layer module 。

Description

Lithium battery health state estimation and correction method and system based on layered fusion Technical Field The invention relates to the technical field of battery management systems, and discloses a lithium battery health state estimation and correction method and system based on layered fusion. Background Accurate estimation of the State of Health (SOH) of lithium batteries is a key technical basis for Battery management systems (Battery MANAGEMENT SYSTEM, BMS) to implement State monitoring, life prediction and safety management. The state of health is generally defined as the ratio of the current maximum available capacity of the battery to the factory rated capacity, reflecting the age and remaining useful life of the battery. Along with the rapid development of electric vehicles and energy storage systems, accurate estimation of the state of health of the battery has important significance for guaranteeing safe operation of the system, optimizing energy management strategies and realizing gradient utilization of the battery. Existing state of health estimation methods face mainly the following technical challenges. First, the state of charge, SOC, and maximum available capacity of the batteryThe traditional single-model filtering method is easy to generate mutual interference when simultaneously estimating the two parameters, so that the estimated result is slow in convergence speed and even divergent. Secondly, the estimation method based on filtering is severely dependent on an accurate battery model and an accurate initial state, and initial value deviation or model parameter mismatch directly influences estimation accuracy and system convergence. Again, the pure online filtering algorithm lacks an absolute reference correction mechanism, and estimation errors caused by sensor noise and model mismatch can accumulate over time, resulting in drift of long-term estimation results. In addition, many high-precision algorithms rely on complete standard charge-discharge curves that are difficult to trigger or effectively apply under intermittent, dynamic load conditions of actual vehicle operation. In the prior art, although research attempts are made to adopt a dual Kalman filtering architecture, most schemes are only applied to joint estimation of a charge state and model parameters, or a complete technical scheme for carrying out systematic hierarchical fusion on online estimation and offline reference and high-precision anchor point events is lacking, so that the problems of coupling estimation, long-term drift and working condition adaptability are not fundamentally solved. Therefore, a health state estimation scheme capable of realizing advantage complementation, having self-adaptive capability and combining high precision and high robustness is needed. Disclosure of Invention The invention provides a lithium battery health state estimation and correction method and system based on layered fusion, which are used for solving the technical problems of difficult coupling estimation of a state of charge and a maximum available capacity, long-term drift caused by lack of absolute reference in online estimation, insufficient adaptability of an algorithm to actual dynamic working conditions and the like in the prior art. In order to solve the technical problems, the invention provides a lithium battery health state estimation and correction method based on layered fusion, which comprises the following steps: step S1, an offline priori layer queries a pre-stored life table according to battery historical operation data to calculate initial maximum available capacity The initial maximum available capacityTransmitting to an online estimation layer; step S2, the auxiliary extended Kalman filter of the online estimation layer uses the initial maximum available capacity For initial value, estimate maximum available capacityThe master extended Kalman filter utilizes the maximum available capacityEstimating state of charge (SOC), updating the maximum available capacity by the auxiliary extended Kalman filter by using the SOCAccording to the maximum available capacityAnd rated capacityOutputs the ratio of the on-line health state estimated value; Step S3, the strong correction layer monitors the battery operation condition, and when the high confidence coefficient condition is met, the actual capacity is calculated through a two-point methodAccording to the actual capacityAnd the rated capacityOutputs a strongly corrected state of health value; S4, the fusion decision layer is used for estimating the value according to the online health stateAnd the strongly corrected state of health valueThe confidence coefficient of the (B) is weighted and fused to output the final health state valueAnd the actual capacity is calculatedFeeding back to the auxiliary extended Kalman filter, resetting the maximum available capacity。 Preferably, in step S1, the battery historical operation data includes a historical average tempe