AI for Status Monitoring of Utility Scale Batteries

Contents
About the Authors
Foreword
Preface
List of contributors
1 Introduction
1.1 Motivation for utility-scale battery deployment
1.2 Definition of AI in the context of battery management
1.3 Advantages of using AI for battery management
2 Utility-scale lithium-ion battery system characteristics
2.1 Overview of lithium-ion batteries
2.1.1 Battery working principle
2.1.2 Principles of status monitoring of utility-scale batteries
2.2 Lithium-ion batteries
2.2.1 Lithium iron phosphate batteries
2.2.2 Lithium cobaltate oxide batteries
2.2.3 Lithium manganese oxide batteries
2.3 Large capacity lithium-ion batteries
2.3.1 Application areas of utility-scale batteries
2.3.2 Characteristics of utility-scale battery systems
2.3.3 Operational challenges of utility-scale battery systems
3 AI-based equivalent modeling and parameter identification
3.1 Overview of battery equivalent circuit modeling
3.2 Modeling types and concepts
3.3 Equivalent circuit modeling methods
3.3.1 Basic equivalent circuit modeling methods
3.3.2 Second-order RC ECM
3.3.3 Compound equivalent circuit modeling
3.3.4 Voltage matching equivalent circuit modeling
3.3.5 Improved second-order equivalent circuit modeling
3.4 Introduction of complex testing conditions
3.4.1 HPPC test
3.4.2 Beijing bus dynamic stress test
3.4.3 Dynamic stress test
3.5 Offline parameter identification
3.5.1 Point calculation strategies
3.5.2 Curve-fitting methods
3.5.3 Model parameter validation and evaluation
3.5.4 Double exponential fitting strategies
3.6 AI-based online parameter identification
3.6.1 Recursive LS (RLS) method
3.6.2 Bias compensation RLS method
3.6.3 Forgetting factor RLS method
3.6.4 Improved multi-innovation LS (MILS) method
3.7 Conclusion
4 Use of artificial intelligence for utility-scale battery systems
4.1 Selection criteria for choice of AI techniques
4.1.1 Common AI methods for utility-scale battery systems
4.1.2 AI technology evaluation index
4.2 Monitoring of utility-scale batteries development with AI
4.2.1 Development status
4.2.2 Development prospect
4.3 Basic parameters in AI-based status monitoring
4.3.1 Voltage for input and correction
4.3.2 Capacity for internal state parameters
4.3.3 Internal resistance for state parameters
4.3.4 Polarization resistance and capacitance for internal parameters
4.3.5 Energy density for correction
4.4 Conclusion
5 AI-based state-of-charge estimation
5.1 Overview of SOC estimation
5.1.1 Definition of SOC
5.1.2 Main affecting factor analysis
5.1.3 Traditional estimation method
5.2 Backpropagation-based SOC estimation method
5.2.1 Backpropagation network model structure
5.2.2 Network information capacity optimization
5.2.3 Training sample collection design
5.2.4 Initial parameter weight design
5.3 Radial basis function-based SOC estimation
5.3.1 Radial basis function neural Network
5.3.2 Neural network parameter learning method
5.3.3 Improvement of RBFNN model for estimating battery SOC
5.4 Elman neural network-based SOC estimation
5.4.1 Elman neural network structure
5.4.2 Adaptive learning process design
5.4.3 Learning process optimization
5.5 Nonlinear autoregressive neural network-based SOC estimation
5.5.1 Nonlinear autoregressive neural network structure
5.5.2 Network creation and training
5.6 Improved genetic BP neural network-based SOC estimation
5.6.1 Genetic BP neural network structure
5.6.2 BP calculation procedure
5.6.3 GA fitness correction
5.7 Experimental verification for large-capacity batteries
5.7.1 Experimental design and data set establishment
5.7.2 BP-based SOC estimation analysis
5.7.3 Radial basis function-based SOC estimation analysis
5.7.4 Elman neural network-based SOC estimation analysis
5.7.5 Nonlinear autoregressive-based SOC estimation analysis
5.7.6 Genetic BP-based SOC estimation analysis
5.8 Conclusion
6 AI-based battery parameter estimation
6.1 AI-based battery state of health estimation
6.1.1 Definition of state of health
6.1.2 Main affecting factor analysis
6.1.3 Summary of AI-based SOH estimation
6.1.4 Modeling methods
6.1.5 Data-driven methods
6.1.6 Statistical law method
6.2 AI-based battery state of power determination
6.2.1 Definition of state of power
6.2.2 Main affecting factor analysis
6.2.3 Summary of AI-based SOP estimation
6.3 AI-based battery state of energy calculation
6.3.1 Definition of state of energy
6.3.2 Main affecting factor analysis
6.3.3 Summary of AI-based SOE estimation
6.4 AI-based battery remaining useful life prediction
6.4.1 Definition of remaining useful life
6.4.2 Main affecting factor analysis
6.4.3 Summary of AI-based RUL prediction
7 Examples and case studies for utility-scale battery operation
7.1 Battery system location and characteristics
7.1.1 Battery management system
7.1.2 Battery pack
7.2 SOC estimation based on genetic algorithm and BP neural network
7.2.1 BP neural network
7.2.2 Genetic algorithm
7.2.3 Experimental analysis
7.3 Prediction model of battery health degradation based on multi-scale depth neural networks
7.3.1 Definition of SOH
7.3.2 Multiscale decomposition based on EEMD and CA
7.3.3 A prediction model based on DBN
7.3.4 A prediction model based on LSTM
7.3.5 Combined model prediction framework
7.4 NARX neural network model based on filter fusion for SOC estimation
7.4.1 NARX neural network
7.4.2 LSTM neural network
7.4.3 Hybrid NARX and LSTM model
7.5 SOC estimation based on drosophila algorithm and BP neural network
7.5.1 BP neural network
7.5.2 Drosophila optimization algorithm
7.5.3 Drosophila algorithm to optimize BP neural network
7.5.4 Conclusion
7.6 SOH estimation based on sparrow search optimization-extreme learning machine network
7.6.1 Extreme learning machine network
7.6.2 Sparrow search optimization
7.6.3 Sparrow search optimization-extreme learning machine network
8 Summary and prospect for AI-based battery status monitoring
8.1 Lessons learned
8.2 Open questions
8.3 Opportunities and challenges
8.4 Development trend of AI in BMS
8.5 Conclusion
References
Index