Event-triggered model predictive control for the cascade filling process of a hydrogen fuelling station

Abstract

Efficient filling strategies for hydrogen fuel cell vehicles are critical for hydrogen utilization efficiency at hydrogen fuelling stations. A novel event-triggered model predictive control framework is proposed in this article for the filling process of a hydrogen fuelling station, which involves multiple compressors, cascade storage tanks, and multiple dispensers. The filling process is formulated as a Mixed-Integer NonLinear Programming (MINLP) problem with the objective of minimizing the vehicle filling times and maximizing the hydrogen utilization efficiency. A solution approach that combines the mixed-integer linear programming and genetic algorithm is designed for solving the resulting MINLP problem. In addition, an event-triggered mechanism is proposed to increase the computational efficiency and to update the control inputs only when needed. Different sets of computational experiments are carried out to demonstrate the effectiveness of the mathematical formulation and the solution approach.