Temperature Control and Life Extension Scheme for Energy Storage System with Enhanced Network Security Based on Deep Learning

Abstract

Existing water cooling systems often adjust battery packs based on fixed temperature thresholds, which can easily lead to reduced battery efficiency and insufficient battery life. At the same time, the interconnection of multiple devices makes the temperature control and management of energy storage systems vulnerable to external DDoS (Distributed Denial of Service) attacks, resulting in data tampering. This study applies Bi-LSTM (Bidirectional Long Short-Term Memory) and PPO (Proximal Policy Optimization) algorithms to optimize temperature control and extend the life of lithium-ion batteries, and combines blockchain technology and AES-GCM (Advanced Encryption Standard with Galois/Counter Mode) algorithm to achieve network encryption protection. First, the battery pack temperature and environmental data are collected through sensors; a Bi-LSTM network is designed to capture the time series characteristics of battery temperature changes; the backpropagation algorithm is used to adjust the network parameters. Then, a learning environment is constructed based on factors such as the predicted temperature of the battery pack, cooling efficiency, and external environmental changes, and the system actions are optimized through the PPO algorithm to maximize the temperature control benefits. At the same time, a decentralized blockchain network is built to store and transmit energy storage system control instructions and temperature data; data is encrypted using AES-GCM and decrypted using a symmetric key at the receiving end to verify data integrity. The research results show that the final capacity of a 3000mAh lithium-ion battery after 150 charge and discharge cycles under the temperature control strategy is 2817.92mAh; the average response time of 10 rounds of tests is only 1.29 seconds; the success rate of defense against DDoS attacks is 96.3%; no data tampering occurs. The method used can significantly enhance the temperature control stability of the system and extend the battery life, achieving a high defense capability against network attacks and data tampering.