BMS System Based On Artificial Intelligence Analysis

With the profound changes in the international energy landscape and the transformation of China's energy development strategy,the proportion of renewable energy in national energy has continued to increase.At the same time,the problem of the consumption of renewable energy has become a fundamental problem that restricts the development of renewable energy at present and in the longer term.Energy storage technology has found a solution to solve problems such as instability,difficulty in digestion,and high waste rate encountered in the development of renewable energy such as wind and solar energy.Promoting the rapid development of energy storage technology is to improve the efficiency of energy use and then reach China The only way for energy development strategy.Among them,the lithium-ion battery energy storage system is a kind of energy storage system with a large installed capacity and the greatest development potential.Its core competence lies in the accurate estimation of the state of charge by the battery management system.In this thesis,the NB-IOT Internet of Things is used to transmit the real-time battery running status data obtained by the lithium ion battery energy storage system BMS,including the training platform built by this laboratory and the operating data of the energy storage site of the partner company.A large number of actual operating conditions data were used to design and implement a lithium battery SOC estimation model based on LSTM network suitable for energy storage scenarios.Based on this,considering the dependence of the model's iteration speed on network parameters and model hyperparameters,a BN algorithm is introduced to optimize it,which makes the iteration speed of the estimated model faster and the error performance improved.Finally,according to the model's predicted noise phenomenon and many operating conditions,a denoising adaptive decoder is designed to perform feature extraction and denoising adaptively on the input data,which optimizes the model's robustness and generalization ability,and improves error performance.further optimization.Finally,the actual measurement data of the energy storage lithium battery is used for actual measurement,which verifies the effectiveness of the algorithm optimization and obtains a good estimation effect.In addition,this paper designs and implements the BMS intelligent measurement and control system for energy storage systems,and integrates the SOC artificial intelligence estimation algorithm of the energy storage lithium battery based on the operating state of the battery.The system adopts the design concept of front-end separation,which is divided into business layer,algorithm layer,and data layer.It has the characteristics of low coupling,high cohesion,and high scalability.Functions such as real-time estimation of battery SOC and visual display of energy storage battery operation were performed,and functional and performance tests were performed.