Reinforcement learning based adaptive power pinch analysis for energy management of stand-alone hybrid energy storage systems considering uncertainty

Abstract

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Energy Management of Microgrids Using Power Pinch Analysis in the Presence of Network Constraints

“…A 4-bus radial distribution test system is used as a case study to demonstrate this new methodology. The focus of the study is to investigate the effects of the proposed control methodology on actual distribution networks while considering network constraints as compared to previous studies done in [10]- [13] where PoPA through the PGCC, was used in energy interactions between different energy storage technologies, and a diesel generator only on an isolated energy system where network constraints were not considered. In this study, the initial focus will be on a much larger system and network line losses calculated through a load flow simulation are considered in the problem formulation.…”

“…The PGCC helped to identify the desired operational profile of energy carriers in the form of energy targets during a prediction horizon to meet certain system operational goals, and these energy targets are later implemented through a series of control actions that enabled the precise matching of the PGCC under both perfect and uncertain weather conditions. Nyong-Bassey et al [13] have explored the synergy between multiple energy storage technologies aimed at enhancing the reliability of intermittent RES of an isolated hybrid renewable energy system by proposing PMS inspired by PoPA approach through the PGCC method. To account for uncertainties in load demands and weather forecast, three adaptive PoPA-based PMS as against the Day-Ahead (DA) PoPA approach that assumes a perfect weather and load forecast were proposed in the study to significantly reduce the effect of forecast error while shaping the PGCC.…”

Probabilistic adaptive model predictive power pinch analysis (PoPA) energy management approach to uncertainty

“…Furthermore, energy management of BAT in the event that the battery becomes fully charged and the utilisation of the excess energy were not discussed. Thus, this paper presents an adaptation of the works of [4] and [7, 9] by defining the robust adaptive energy management algorithm in a probabilistic chance constrained framework. Furthermore, the excess energy in the system, represented by overcharging the BAT and energy recovered as well as over discharging the BAT , is considered in the chance constraints evaluated with the MCS.…”

“…The receding horizon model predictive control (RHMPC) PoPA, hence, employs a state feedback loop to adapt the model to the HESS. Thus, a closed loop is utilised and the estimated and real states are compared for discrepancy so as to achieve robustness to compensate for the weather/load uncertainty [9]. Hence, minimising the effect of uncertainty Δ H 1,2 between the real and the estimated state of charge and the state feedback error due to uncertainty is re‐computed as follows; where y ( k ) is the output state measured at time k , and superscripts m , n refer to the real and the estimated state of charge, respectively.…”

“…We compare three algorithms. The DA algorithm where uncertainty is not considered, the RHMPC‐PoPA algorithm where the prediction error is corrected [9] and the MC‐PoPA algorithm where uncertainty is considered.…”