Battery SOC Estimation via Single Parameter Matrix Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the state of charge of a battery cell are complex, particularly in adjusting the matrices Qk and Rk, due to difficulties in quantifying modeling errors, which complicates the precision of the estimation process.
Innovation Solution
The method simplifies the adjustment of matrices Qk and Rk by setting them based on known matrices Fk and Hk, with the user only needing to choose an integer N0, rather than adjusting each coefficient, allowing for easier recalibration when Fk and Hk are modified, thereby increasing the precision of the state of charge estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to adjust matrices Qk and Rk by quantifying modeling errors, then measurement precision may be improved, but device complexity and ease of operation deteriorate due to the difficulty of adjusting each coefficient
Solution Approach 1:
The patent transforms the complex multi-parameter adjustment problem into a single-parameter adjustment by changing the formulation approach. Instead of adjusting each coefficient of matrices Qk and Rk individually based on difficult-to-quantify modeling errors, the invention introduces a single parameter N0 that automatically determines both matrices through established relationships with known matrices Fk and Hk. This parameter transformation resolves the contradiction by maintaining estimation precision while dramatically simplifying the adjustment process.
Solution Approach 2:
The system performs self-adjustment through the automatic calculation mechanism. When parameter N0 is set, the matrices Qk and Rk are automatically computed using the formulas Qk = (N0*Ik - Fk)*PFk and Rk = N0*(Hk*PHkT), eliminating the need for manual quantification of modeling errors and manual adjustment of multiple coefficients. This self-service approach resolves the operational complexity while maintaining precision.
2Measurement precision
If manual adjustment of each matrix coefficient is performed, then measurement precision may be improved, but ease of operation deteriorates due to the difficulty of recalibration when models are modified
Solution Approach 1:
The invention changes the operational paradigm from adjusting multiple coefficients to adjusting a single parameter N0. The matrices Qk and Rk are derived automatically from N0 and the known system matrices through deterministic formulas. This parameter transformation makes recalibration trivial - when Fk or Hk are modified, the user only needs to re-specify N0, and the system automatically regenerates the appropriate covariance matrices, greatly improving ease of operation while preserving precision.
Solution Approach 2:
The single parameter N0 serves multiple functions simultaneously: it controls the state noise covariance Qk, the measurement noise covariance Rk, and adapts to different modeling scenarios. This universal parameter approach eliminates the need for separate manual adjustments for different matrix coefficients and enables easy recalibration when models change, resolving the ease of operation contradiction.
3Measurement precision
If complex matrix adjustment procedures are used, then measurement precision may be improved, but productivity deteriorates due to the time required for adjustment and recalibration
Solution Approach 1:
The invention transforms a time-consuming multi-parameter adjustment process into an instantaneous single-parameter setting. By establishing direct mathematical relationships between N0 and the covariance matrices, the system eliminates iterative adjustment and manual quantification steps. The formulas Qk = (N0*Ik - Fk)*PFk and Rk = N0*(Hk*PHkT) enable immediate computation of optimal matrices once N0 is specified, dramatically improving productivity while maintaining precision.
Solution Approach 2:
The patent performs preliminary establishment of the mathematical relationships between N0 and the covariance matrices before actual operation. The formulas linking N0 to Qk and Rk are pre-derived and stored in the system, enabling instantaneous computation during operation without requiring real-time iterative adjustment or recalculation, thus improving implementation speed while preserving precision.
Data Source
Figure 1~3
Figure 4~9
Figure 10~13
AI summary
The invention relates to a method of estimating the charge state of a battery cell involving the following steps: calculating (116) a prediction of the charge state SOCk of the cell using a state model comprising a state-transition matrix Fk, this prediction being contaminated with a state noise of which the covariance is given by a matrix Qk; calculating (124) a prediction ŷk of a measured value yk using an observation model comprising an observability matrix Hk, this prediction ŷk being contaminated with a measurement noise of which the covariance is given by a matrix Rk; and adjusting (102, 120) the values of the matrices Qk and Rk by means of the following relations: Qk=[N0G0, k(N0)]-1 and Rk=I, where N0 is a predetermined integer strictly greater than one, G0, k(N0) is given by the following relation: I is the identity matrix.