Multi-step ahead thermal warning network for energy storage system based on the core temperature detection.

Area of Science:

• Energy Storage Systems
• Battery Thermal Management
• Artificial Intelligence in Engineering

Background:

• Lithium-ion batteries are crucial for energy storage but are limited by temperature-dependent performance and safety.
• Core temperature, a key indicator of battery health, is difficult to measure directly, unlike surface temperature.
• Significant temperature differences between the core and surface of batteries pose a risk to energy storage systems.

Purpose of the Study:

• To develop a multi-step ahead thermal warning network for energy storage systems based on core temperature detection.
• To enhance the thermal safety of lithium-ion battery energy storage systems by predicting critical temperature events.
• To provide an early warning system that utilizes core temperature to prevent thermal runaway.

Main Methods:

• Development of a thermal warning network utilizing the measurement difference between core and surface temperatures.
• Integration of a long and short-term memory (LSTM) network to process input time series data for temperature prediction.
• Using core temperature as the primary criterion for issuing multi-step ahead warning signals.

Main Results:

• The developed thermal warning network accurately predicts whether the core temperature will reach critical values in a future time window.
• The network provides advance warning signals, enabling proactive measures to maintain thermal safety.
• Extensive testing verified the accuracy and effectiveness of the proposed thermal early warning model.

Conclusions:

• The core temperature-based thermal warning network is effective in enhancing the safety of lithium-ion battery energy storage systems.
• The system's ability to predict critical temperatures in advance allows for timely intervention, mitigating risks.
• This approach offers a viable solution for commercializing core temperature estimation and improving overall energy storage reliability.