A data-driven probabilistic evaluation method of hydrogen fuel cell vehicles hosting capacity for integrated hydrogen-electricity network

Abstract

The growing demand for hydrogen in hydrogen fuel cell vehicles (HFCVs) will present challenges for the safe operation of the integrated distributed hydrogen supply network (DHSN) and power distribution network (PDN). This paper proposes a data-driven probabilistic evaluation method for determining the hosting capacity of hydrogen fuel cell vehicles (HVHC). Firstly, a directional mapping method is proposed to model the maximum safety boundaries of critical electrolyzers. It integrates the thermal-electrical and dynamic operating characteristics while maintaining the full boundaries in dimension reduction to ensure accuracy and efficiency. A probabilistic HVHC evaluation model is developed to consider uncertain factors of high dimensions despite data deficiency, such as HFCV refueling demand and renewable power. The proposed model determines the maximum network tolerance for the HFCV number, constrained by the safety constraints of the PDN integrated with DHSN, on-site and off-site hydrogen supplies coupled by tube trailers. Finally, a cross-term decoupled data-driven polynomial chaos expansion is proposed to efficiently solve the developed probabilistic HVHC model. It is established based on raw and small samples without extracted probability distribution information. The solution approach also combines the cross-terms in expansions using Taylor expansion, making it efficient for high-dimensional problems. Furthermore, the accuracy and scale reduction effect of the solution algorithm are proven based on Wasserstein ambiguity sets. Numerical studies on three systems show that the proposed method has only a 0.01% error in HVHC results and an 8.38% computation time of the Monte Carlo method.