Advances in Battery Manufacturing and Operating Status Analysis
Filtering and Artificial Intelligence Strategy
1st Edition
Will be published approx. on 16. December 2026
Book
Hardback
240 pages
978-1-394-43716-0 (ISBN)
Description
Advanced filtering and AI algorithms for battery system analysis
Advances in Battery Manufacturing and Operating Status Analysis details zonotopic and particle filtering methods for robust real-time estimation of critical battery parameters, alongside hybrid models combining filters with long short-term memory networks for remaining useful life prediction. Coverage of genetic algorithms and Q-learning addresses intelligent battery grouping and manufacturing capacity forecasting. Technical case studies walk through problem definitions, data preprocessing, model selection, implementation, and interpretation of results.
Key topics also include:
Zonotopic and particle filtering approaches for achieving robust, real-time estimation of critical battery state parameters in operational environments
Hybrid filter and long short-term memory network models designed to predict remaining useful life with improved accuracy
Genetic algorithm and Q-learning strategies applied to intelligent battery grouping and manufacturing capacity forecasting
Technical case studies covering problem definitions, data preprocessing, model selection, implementation, and real-world result interpretation
Data-driven strategies for optimizing battery lifecycle stages from manufacturing through operation and sustainable energy storage
Researchers and industry professionals in energy storage, power electronics, and electrical engineering R&D will find targeted algorithmic strategies for battery system management. Graduate students studying energy storage and related disciplines gain exposure to filtering and AI methods applied directly to manufacturing and operational analysis challenges.
Advances in Battery Manufacturing and Operating Status Analysis details zonotopic and particle filtering methods for robust real-time estimation of critical battery parameters, alongside hybrid models combining filters with long short-term memory networks for remaining useful life prediction. Coverage of genetic algorithms and Q-learning addresses intelligent battery grouping and manufacturing capacity forecasting. Technical case studies walk through problem definitions, data preprocessing, model selection, implementation, and interpretation of results.
Key topics also include:
Zonotopic and particle filtering approaches for achieving robust, real-time estimation of critical battery state parameters in operational environments
Hybrid filter and long short-term memory network models designed to predict remaining useful life with improved accuracy
Genetic algorithm and Q-learning strategies applied to intelligent battery grouping and manufacturing capacity forecasting
Technical case studies covering problem definitions, data preprocessing, model selection, implementation, and real-world result interpretation
Data-driven strategies for optimizing battery lifecycle stages from manufacturing through operation and sustainable energy storage
Researchers and industry professionals in energy storage, power electronics, and electrical engineering R&D will find targeted algorithmic strategies for battery system management. Graduate students studying energy storage and related disciplines gain exposure to filtering and AI methods applied directly to manufacturing and operational analysis challenges.
More details
Language
English
Place of publication
United States
Publishing group
John Wiley & Sons Inc
Target group
Professional and scholarly
ISBN-13
978-1-394-43716-0 (9781394437160)
Copyright in bibliographic data is held by Nielsen Book Services Limited or its licensors: all rights reserved.
Schweitzer Classification
Persons
Ziyun Wang is a Professor and Doctoral Supervisor at Jiangnan University's School of Automation and Intelligent Science and Deputy Director of the Engineering Center for the Application of Internet of Things Technology. His research is focused on advanced manufacturing, battery operation analysis, and filter design.
Yan Wang is a Professor and Doctoral Supervisor at Jiangnan University's School of Automation and Intelligent Science and Yangtze River Distinguished Professor of the Ministry of Education. Her research spans artificial intelligence, advanced control and system optimization, and industrial internet technology.
Zhicheng Ji is a Professor and Doctoral Supervisor at Jiangnan University's School of Automation and Intelligent Science, and Director of Jiangsu Engineering Research Center for Intelligent Optimization Manufacturing of Industrial Internet, and former Vice Chancellor of Jiangnan University. His research is focused on energy system design, state estimation, and fault diagnosis.
Yan Wang is a Professor and Doctoral Supervisor at Jiangnan University's School of Automation and Intelligent Science and Yangtze River Distinguished Professor of the Ministry of Education. Her research spans artificial intelligence, advanced control and system optimization, and industrial internet technology.
Zhicheng Ji is a Professor and Doctoral Supervisor at Jiangnan University's School of Automation and Intelligent Science, and Director of Jiangsu Engineering Research Center for Intelligent Optimization Manufacturing of Industrial Internet, and former Vice Chancellor of Jiangnan University. His research is focused on energy system design, state estimation, and fault diagnosis.
Content
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Research background and significance . . . . . . . . . . . . . . . . 1
1.2 Battery manufacturing . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Battery operating status analysis . . . . . . . . . . . . . . . . . . . 4
1.4 Principle of filter design methods . . . . . . . . . . . . . . . . . . 6
1.4.1 Linear filters . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.2 Nonlinear filters . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.3 Robust filters . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4.4 The integration of filters with AI . . . . . . . . . . . . . . . 7
1.5 Artificial intelligence strategy . . . . . . . . . . . . . . . . . . . . 8
1.6 Outlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 Advances in battery grouping by artificial intelligence strategy . . . . 11
2.1 Introduction to battery grouping . . . . . . . . . . . . . . . . . . . 11
2.2 Problem formulation and model establishment . . . . . . . . . . . 12
2.2.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . 12
2.2.2 Model establishment . . . . . . . . . . . . . . . . . . . . . 12
2.3 Chromosome encoding . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 Crossover modes designing . . . . . . . . . . . . . . . . . . . . . 13
2.5 Population mutations . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.6 Convergence analysis . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.7 Case study on battery grouping . . . . . . . . . . . . . . . . . . . 18
2.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 20
3 Advances in forecasting of battery manufacturing capacity by artificial
intelligence strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 Introduction to battery manufacturing capacity . . . . . . . . . . . 21
3.2 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Q-learning principle . . . . . . . . . . . . . . . . . . . . . . . . . 24
Iii
3.3.1 Markov decision process . . . . . . . . . . . . . . . . . . . 25
3.3.2 Bellman optimality equation and Q-function . . . . . . . . 25
3.3.3 Tabular Q-learning algorithm . . . . . . . . . . . . . . . . . 26
3.3.4 Formulation for parameter optimization . . . . . . . . . . . 26
3.4 Variational mode decomposition . . . . . . . . . . . . . . . . . . . 27
3.4.1 Constrained variational formulation . . . . . . . . . . . . . 27
3.4.2 Iterative solution by ADMM . . . . . . . . . . . . . . . . . 28
3.4.3 Remarks on parameter selection . . . . . . . . . . . . . . . 29
3.5 Long short-term memory . . . . . . . . . . . . . . . . . . . . . . . 29
3.6 Design of the forecasting method . . . . . . . . . . . . . . . . . . 30
3.7 Case study on forecasting of battery manufacturing capacity . . . . 34
3.7.1 Model parameter setting . . . . . . . . . . . . . . . . . . . 34
3.7.2 Algorithm comparative analysis . . . . . . . . . . . . . . . 35
3.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 39
4 Advances in battery operating status analysis by zonotopic filtering . 41
4.1 Introduction to battery operating status analysis . . . . . . . . . . . 41
4.2 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.3 Zonotopic filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.2 Zonotopic filtering-based operating state estimation . . . . . 48
4.3.2.1 Prediction step . . . . . . . . . . . . . . . . . . . 48
4.3.2.2 Update step . . . . . . . . . . . . . . . . . . . . 50
4.3.2.3 Operating state estimation . . . . . . . . . . . . . 52
4.3.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.4 Zonotope and Gaussian Kalman filters . . . . . . . . . . . . . . . . 53
4.4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.4.2 Main results . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.4.2.1 Prediction . . . . . . . . . . . . . . . . . . . . . 55
4.4.2.2 Filtering . . . . . . . . . . . . . . . . . . . . . . 56
4.4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.5 Orthotope-search-expansion-based extended zonotopic Kalman filter 61
4.5.1 Preliminaries and problem description . . . . . . . . . . . . 61
4.5.1.1 Preliminaries . . . . . . . . . . . . . . . . . . . . 61
4.5.1.2 Problem description . . . . . . . . . . . . . . . . 63
4.5.2 Main Results . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.5.2.1 OSE-algorithm-based time-varying parameter identification
. . . . . . . . . . . . . . . . . . . . . . 64
4.5.2.2 Optimal OSE-EZKF design for the discrete-time
LPV system . . . . . . . . . . . . . . . . . . . . 69
4.5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.6 Constrained zonotopic Kalman filtering . . . . . . . . . . . . . . . 73
4.6.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.6.2 Main results . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.6.2.1 Bounding the linearization error . . . . . . . . . . 74
4.6.2.2 Prediction step . . . . . . . . . . . . . . . . . . . 76
4.6.2.3 Update step . . . . . . . . . . . . . . . . . . . . 77
4.6.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.7 Case study on battery operating status analysis . . . . . . . . . . . 79
4.7.1 Case1:Zonotopic filter . . . . . . . . . . . . . . . . . . . . 79
4.7.1.1 State of charge simulation . . . . . . . . . . . . . 79
4.7.1.2 State of health simulation . . . . . . . . . . . . . 81
4.7.2 Case2:Zonotope and Gaussian Kalman filters . . . . . . . . 82
4.7.3 Case3:Orthotope-search-expansion-based extended zonotopic
Kalman filter . . . . . . . . . . . . . . . . . . . . . . 85
4.7.4 Case4:Constrained zonotopic Kalman filtering . . . . . . . 88
4.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 92
5 Advances in battery operating status analysis by filtering and artificial
intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.2 Introduction to particle filter and particle swarm optimization . . . 95
5.2.1 Particle Filter (PF) . . . . . . . . . . . . . . . . . . . . . . 95
5.2.1.1 Problem setting . . . . . . . . . . . . . . . . . . 95
5.2.1.2 Bayesian recursion . . . . . . . . . . . . . . . . . 96
5.2.1.3 Particle approximation and importance weighting. 96
5.2.1.4 Degeneracy and resampling . . . . . . . . . . . . 96
5.2.1.5 Canonical PF recursion . . . . . . . . . . . . . . 97
5.2.2 Particle Swarm Optimization (PSO) . . . . . . . . . . . . . 97
5.2.2.1 Background and motivation . . . . . . . . . . . . 97
5.2.2.2 Algorithmic formulation . . . . . . . . . . . . . . 97
5.2.2.3 Initialization and stopping . . . . . . . . . . . . . 98
5.3 Particle swarm optimization based orthometric hyperparallel space
filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.2 Orthometric hyperparallel space filtering based on particle
swarm optimization . . . . . . . . . . . . . . . . . . . . . . 99
5.3.2.1 Construction of the orthometric hyperparallel particle
search space . . . . . . . . . . . . . . . . . 99
5.3.2.2 Particle swarm optimization update based on orthometric
hyperparallel space . . . . . . . . . . . 103
5.3.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4 Improved particle filter algorithm based on the parallelotope . . . . 107
5.4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4.2 Design of the improved particle filter based on the parallelotope
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.4.2.1 Prediction . . . . . . . . . . . . . . . . . . . . . 108
5.4.2.2 Update . . . . . . . . . . . . . . . . . . . . . . . 112
5.4.2.3 Summary of the P-IPF . . . . . . . . . . . . . . . 117
5.4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 117
5.5 Projected particle-confinement-based zonotopic space filtering . . . 119
5.5.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.5.2 Design of projected particle-confinement-based zonotopic
space filtering algorithm . . . . . . . . . . . . . . . . . . . 120
5.5.2.1 Establishment of the zonotopic search space . . . 121
5.5.2.2 Projected particle-confinement-based zonotopic
space update . . . . . . . . . . . . . . . . . . . . 122
5.5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.6 Zonotopic feasible set optimized filter based on differential evolution 130
5.6.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.6.1.1 Differential evolution algorithm . . . . . . . . . . 131
5.6.2 Main conclusion . . . . . . . . . . . . . . . . . . . . . . . 132
5.6.2.1 Design of the Zonotopic Kalman Filter . . . . . . 132
5.6.2.2 DE-ZKF based generator matrix optimization . . 133
5.6.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 135
5.7 Case study on battery operating status analysis . . . . . . . . . . . 136
5.7.1 Case 1: PSO-OHSF . . . . . . . . . . . . . . . . . . . . . 136
5.7.2 Case 2: P-IPF . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.7.3 Case 3: PPC-ZSF . . . . . . . . . . . . . . . . . . . . . . . 144
5.7.4 Case 4: DE-ZKF . . . . . . . . . . . . . . . . . . . . . . . 148
5.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 151
6 Advances in battery remaining useful life analysis by filtering and artificial
intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.1 Introduction to battery remaining useful life analysis . . . . . . . . 153
6.1.1 Definition and significance of remaining useful life . . . . . 153
6.1.2 Challenges and current research limitations in RUL prediction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.1.3 Principal technical routes and chapter focus . . . . . . . . . 155
6.2 Dynamic complexity reduction zonotopic Kalman filter . . . . . . . 156
6.2.1 Preliminaries and problem formulation . . . . . . . . . . . 157
6.2.1.1 Preliminaries . . . . . . . . . . . . . . . . . . . . 157
6.2.1.2 Problem formulation . . . . . . . . . . . . . . . . 158
6.2.2 Design of the DCR-ZKF for state estimation of lithium battery
degradation . . . . . . . . . . . . . . . . . . . . . . . 160
6.2.2.1 SOC estimation on the short time scale . . . . . . 160
6.2.2.2 SOH estimation on the long time scale . . . . . . 161
6.2.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.3 Double exponential empirical particle filter . . . . . . . . . . . . . 163
6.3.1 Main results . . . . . . . . . . . . . . . . . . . . . . . . . . 164
6.3.1.1 DEE-PF model . . . . . . . . . . . . . . . . . . . 164
6.3.1.2 Prediction steps . . . . . . . . . . . . . . . . . . 167
6.3.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 168
6.4 Anti-aliasing filter and LSTM . . . . . . . . . . . . . . . . . . . . 169
6.4.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . 169
6.4.1.1 High-Frequency component prediction model . . 169
6.4.1.2 Monte Carlo-based optimization of interval thresholds
. . . . . . . . . . . . . . . . . . . . . . . . 172
6.4.1.3 SOH and RUL prediction . . . . . . . . . . . . . 172
6.4.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 172
6.5 Case study on battery remaining useful life analysis . . . . . . . . . 173
6.5.1 Case1:DCR-ZKF . . . . . . . . . . . . . . . . . . . . . . . 173
6.5.2 Case2:DEE-PF . . . . . . . . . . . . . . . . . . . . . . . . 176
6.5.3 Case3:MC-R-LSTM . . . . . . . . . . . . . . . . . . . . . 179
6.5.3.1 SOH decomposition . . . . . . . . . . . . . . . . 180
6.5.3.2 Main trend prediction . . . . . . . . . . . . . . . 181
6.5.3.3 Fluctuation component prediction . . . . . . . . . 181
6.5.3.4 Prediction of SOH and RUL . . . . . . . . . . . . 183
6.6 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 186
7 Summary and future outlook . . . . . . . . . . . . . . . . . . . . . . . 189
7.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
7.2 Future outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Research background and significance . . . . . . . . . . . . . . . . 1
1.2 Battery manufacturing . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Battery operating status analysis . . . . . . . . . . . . . . . . . . . 4
1.4 Principle of filter design methods . . . . . . . . . . . . . . . . . . 6
1.4.1 Linear filters . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.2 Nonlinear filters . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.3 Robust filters . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4.4 The integration of filters with AI . . . . . . . . . . . . . . . 7
1.5 Artificial intelligence strategy . . . . . . . . . . . . . . . . . . . . 8
1.6 Outlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 Advances in battery grouping by artificial intelligence strategy . . . . 11
2.1 Introduction to battery grouping . . . . . . . . . . . . . . . . . . . 11
2.2 Problem formulation and model establishment . . . . . . . . . . . 12
2.2.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . 12
2.2.2 Model establishment . . . . . . . . . . . . . . . . . . . . . 12
2.3 Chromosome encoding . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 Crossover modes designing . . . . . . . . . . . . . . . . . . . . . 13
2.5 Population mutations . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.6 Convergence analysis . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.7 Case study on battery grouping . . . . . . . . . . . . . . . . . . . 18
2.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 20
3 Advances in forecasting of battery manufacturing capacity by artificial
intelligence strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 Introduction to battery manufacturing capacity . . . . . . . . . . . 21
3.2 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Q-learning principle . . . . . . . . . . . . . . . . . . . . . . . . . 24
Iii
3.3.1 Markov decision process . . . . . . . . . . . . . . . . . . . 25
3.3.2 Bellman optimality equation and Q-function . . . . . . . . 25
3.3.3 Tabular Q-learning algorithm . . . . . . . . . . . . . . . . . 26
3.3.4 Formulation for parameter optimization . . . . . . . . . . . 26
3.4 Variational mode decomposition . . . . . . . . . . . . . . . . . . . 27
3.4.1 Constrained variational formulation . . . . . . . . . . . . . 27
3.4.2 Iterative solution by ADMM . . . . . . . . . . . . . . . . . 28
3.4.3 Remarks on parameter selection . . . . . . . . . . . . . . . 29
3.5 Long short-term memory . . . . . . . . . . . . . . . . . . . . . . . 29
3.6 Design of the forecasting method . . . . . . . . . . . . . . . . . . 30
3.7 Case study on forecasting of battery manufacturing capacity . . . . 34
3.7.1 Model parameter setting . . . . . . . . . . . . . . . . . . . 34
3.7.2 Algorithm comparative analysis . . . . . . . . . . . . . . . 35
3.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 39
4 Advances in battery operating status analysis by zonotopic filtering . 41
4.1 Introduction to battery operating status analysis . . . . . . . . . . . 41
4.2 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.3 Zonotopic filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.2 Zonotopic filtering-based operating state estimation . . . . . 48
4.3.2.1 Prediction step . . . . . . . . . . . . . . . . . . . 48
4.3.2.2 Update step . . . . . . . . . . . . . . . . . . . . 50
4.3.2.3 Operating state estimation . . . . . . . . . . . . . 52
4.3.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.4 Zonotope and Gaussian Kalman filters . . . . . . . . . . . . . . . . 53
4.4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.4.2 Main results . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.4.2.1 Prediction . . . . . . . . . . . . . . . . . . . . . 55
4.4.2.2 Filtering . . . . . . . . . . . . . . . . . . . . . . 56
4.4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.5 Orthotope-search-expansion-based extended zonotopic Kalman filter 61
4.5.1 Preliminaries and problem description . . . . . . . . . . . . 61
4.5.1.1 Preliminaries . . . . . . . . . . . . . . . . . . . . 61
4.5.1.2 Problem description . . . . . . . . . . . . . . . . 63
4.5.2 Main Results . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.5.2.1 OSE-algorithm-based time-varying parameter identification
. . . . . . . . . . . . . . . . . . . . . . 64
4.5.2.2 Optimal OSE-EZKF design for the discrete-time
LPV system . . . . . . . . . . . . . . . . . . . . 69
4.5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.6 Constrained zonotopic Kalman filtering . . . . . . . . . . . . . . . 73
4.6.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.6.2 Main results . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.6.2.1 Bounding the linearization error . . . . . . . . . . 74
4.6.2.2 Prediction step . . . . . . . . . . . . . . . . . . . 76
4.6.2.3 Update step . . . . . . . . . . . . . . . . . . . . 77
4.6.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.7 Case study on battery operating status analysis . . . . . . . . . . . 79
4.7.1 Case1:Zonotopic filter . . . . . . . . . . . . . . . . . . . . 79
4.7.1.1 State of charge simulation . . . . . . . . . . . . . 79
4.7.1.2 State of health simulation . . . . . . . . . . . . . 81
4.7.2 Case2:Zonotope and Gaussian Kalman filters . . . . . . . . 82
4.7.3 Case3:Orthotope-search-expansion-based extended zonotopic
Kalman filter . . . . . . . . . . . . . . . . . . . . . . 85
4.7.4 Case4:Constrained zonotopic Kalman filtering . . . . . . . 88
4.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 92
5 Advances in battery operating status analysis by filtering and artificial
intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.2 Introduction to particle filter and particle swarm optimization . . . 95
5.2.1 Particle Filter (PF) . . . . . . . . . . . . . . . . . . . . . . 95
5.2.1.1 Problem setting . . . . . . . . . . . . . . . . . . 95
5.2.1.2 Bayesian recursion . . . . . . . . . . . . . . . . . 96
5.2.1.3 Particle approximation and importance weighting. 96
5.2.1.4 Degeneracy and resampling . . . . . . . . . . . . 96
5.2.1.5 Canonical PF recursion . . . . . . . . . . . . . . 97
5.2.2 Particle Swarm Optimization (PSO) . . . . . . . . . . . . . 97
5.2.2.1 Background and motivation . . . . . . . . . . . . 97
5.2.2.2 Algorithmic formulation . . . . . . . . . . . . . . 97
5.2.2.3 Initialization and stopping . . . . . . . . . . . . . 98
5.3 Particle swarm optimization based orthometric hyperparallel space
filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.2 Orthometric hyperparallel space filtering based on particle
swarm optimization . . . . . . . . . . . . . . . . . . . . . . 99
5.3.2.1 Construction of the orthometric hyperparallel particle
search space . . . . . . . . . . . . . . . . . 99
5.3.2.2 Particle swarm optimization update based on orthometric
hyperparallel space . . . . . . . . . . . 103
5.3.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4 Improved particle filter algorithm based on the parallelotope . . . . 107
5.4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4.2 Design of the improved particle filter based on the parallelotope
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.4.2.1 Prediction . . . . . . . . . . . . . . . . . . . . . 108
5.4.2.2 Update . . . . . . . . . . . . . . . . . . . . . . . 112
5.4.2.3 Summary of the P-IPF . . . . . . . . . . . . . . . 117
5.4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 117
5.5 Projected particle-confinement-based zonotopic space filtering . . . 119
5.5.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.5.2 Design of projected particle-confinement-based zonotopic
space filtering algorithm . . . . . . . . . . . . . . . . . . . 120
5.5.2.1 Establishment of the zonotopic search space . . . 121
5.5.2.2 Projected particle-confinement-based zonotopic
space update . . . . . . . . . . . . . . . . . . . . 122
5.5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.6 Zonotopic feasible set optimized filter based on differential evolution 130
5.6.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.6.1.1 Differential evolution algorithm . . . . . . . . . . 131
5.6.2 Main conclusion . . . . . . . . . . . . . . . . . . . . . . . 132
5.6.2.1 Design of the Zonotopic Kalman Filter . . . . . . 132
5.6.2.2 DE-ZKF based generator matrix optimization . . 133
5.6.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 135
5.7 Case study on battery operating status analysis . . . . . . . . . . . 136
5.7.1 Case 1: PSO-OHSF . . . . . . . . . . . . . . . . . . . . . 136
5.7.2 Case 2: P-IPF . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.7.3 Case 3: PPC-ZSF . . . . . . . . . . . . . . . . . . . . . . . 144
5.7.4 Case 4: DE-ZKF . . . . . . . . . . . . . . . . . . . . . . . 148
5.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 151
6 Advances in battery remaining useful life analysis by filtering and artificial
intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.1 Introduction to battery remaining useful life analysis . . . . . . . . 153
6.1.1 Definition and significance of remaining useful life . . . . . 153
6.1.2 Challenges and current research limitations in RUL prediction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.1.3 Principal technical routes and chapter focus . . . . . . . . . 155
6.2 Dynamic complexity reduction zonotopic Kalman filter . . . . . . . 156
6.2.1 Preliminaries and problem formulation . . . . . . . . . . . 157
6.2.1.1 Preliminaries . . . . . . . . . . . . . . . . . . . . 157
6.2.1.2 Problem formulation . . . . . . . . . . . . . . . . 158
6.2.2 Design of the DCR-ZKF for state estimation of lithium battery
degradation . . . . . . . . . . . . . . . . . . . . . . . 160
6.2.2.1 SOC estimation on the short time scale . . . . . . 160
6.2.2.2 SOH estimation on the long time scale . . . . . . 161
6.2.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.3 Double exponential empirical particle filter . . . . . . . . . . . . . 163
6.3.1 Main results . . . . . . . . . . . . . . . . . . . . . . . . . . 164
6.3.1.1 DEE-PF model . . . . . . . . . . . . . . . . . . . 164
6.3.1.2 Prediction steps . . . . . . . . . . . . . . . . . . 167
6.3.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 168
6.4 Anti-aliasing filter and LSTM . . . . . . . . . . . . . . . . . . . . 169
6.4.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . 169
6.4.1.1 High-Frequency component prediction model . . 169
6.4.1.2 Monte Carlo-based optimization of interval thresholds
. . . . . . . . . . . . . . . . . . . . . . . . 172
6.4.1.3 SOH and RUL prediction . . . . . . . . . . . . . 172
6.4.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 172
6.5 Case study on battery remaining useful life analysis . . . . . . . . . 173
6.5.1 Case1:DCR-ZKF . . . . . . . . . . . . . . . . . . . . . . . 173
6.5.2 Case2:DEE-PF . . . . . . . . . . . . . . . . . . . . . . . . 176
6.5.3 Case3:MC-R-LSTM . . . . . . . . . . . . . . . . . . . . . 179
6.5.3.1 SOH decomposition . . . . . . . . . . . . . . . . 180
6.5.3.2 Main trend prediction . . . . . . . . . . . . . . . 181
6.5.3.3 Fluctuation component prediction . . . . . . . . . 181
6.5.3.4 Prediction of SOH and RUL . . . . . . . . . . . . 183
6.6 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . 186
7 Summary and future outlook . . . . . . . . . . . . . . . . . . . . . . . 189
7.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
7.2 Future outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195