Cyber-Physical System-Based Optimization for Multiple Energy Systems with Imperfect State-of-Charge Knowledge of Storage Devices

Zhanbo Xu,Qing-Shan Jia,Xiaohong Guan

DOI: https://doi.org/10.23919/chicc.2018.8483004

2018-01-01

Abstract:The optimization of the multiple energy systems consisting of renewable energy resources, storage devices, various loads, etc., could significantly facilitate the energy efficiency and has thus attracted lots of attention. Since a perfect knowledge of the state of charge (SOC) of the storage devices may be difficult or costly in practice, in this paper we consider the optimization problem of the multiple energy systems with imperfect SOC knowledge of storage devices. The problem is formulated as a partially observable Markov decision process (POMDP). A cyber-physical simulation-based optimization algorithm is developed to solve the problem. The effectiveness of the developed method under uncertainties in the supply and demand and the performance degradation compared to the case with perfect SOC knowledge are demonstrated based on numerical case studies.