Performance prediction, optimal design and operational control of thermal energy storage using artificial intelligence methods

Capable of storing and redistributing energy, thermal energy storage (TES) shows a promising applicability in energy systems. Recently, artificial intelligence (AI) technique is gradually playing an important role in automation, information retrieval, decision making, intelligent recognition, monitoring and management. With the assistance of AI techniques, TES systems can become more and more reasonable and intelligent, which paves a new path for the researches on TES. In the present review, a comprehensive literature summarization and analysis on the application of AI techniques to TES is presented. Performance prediction, optimal design, control and operation by means of AI for the TES systems with various applications are discussed and compared. This review shows that AI-based prediction models, like artificial neural network and support vector machine, can accurately estimate the TES performance and the properties of TES materials in a very fast fashion. AI-based optimization algorithms, such as genetic algorithm, particle swarm optimization, and teaching-learning-based optimization are able to optimize the design and operation of the TES systems towards the objectives like higher system efficiency, cost savings, more renewable energy utilization and less environmental impacts. Fuzzy logic can be utilized to properly design and control the TES systems where uncertain and imprecise factors are inevitably present. General strategies of the AI-based TES performance modelling and the completely AI-based design and control of the TES are summarized, while the main limitations are that AI cannot be used to directly unveil the unknown physical mechanism of the TES and that the lack of the comprehensive TES performance database hinders the real-world implementation. On the way to completely intelligent TES systems, further investigations on the enhancement of adaptation and self-improvement capability are necessary. In addition, the potential research topics are pointed out for the future development and deployment referring to the needs of the future smart energy system, intelligent and zero energy building, and smart home.