Algorithm Aversion-Aware Optimization of Integrated Electricity Charging and Hydrogen Refueling Pricing

The operation and scheduling of taxi fleets remain a key application area of Internet of Things (IoT) technologies. However, the operational strategies for new energy taxis—particularly hydrogen-powered and hydrogen-electric hybrid vehicles—are still underexplored. This paper investigates the deployment of plug-in hybrid hydrogen and electric taxis (PH2ETs) in ride-hailing services and proposes a bilevel optimization framework that jointly addresses PH2ETs’ hybrid charging decisions and the pricing strategy of integrated electricity and hydrogen refueling stations (IEHSs). PH2ETs determine charging options based on energy prices and time costs, while IEHSs dynamically adjust pricing in response to aggregate charging behaviors. To reduce the prediction error caused by uncertain behavior of drivers in actual scenarios, the model incorporates algorithm aversion, capturing drivers’ reluctance to follow platform-generated recommendations. An improved whale optimization algorithm (WOA) is developed to solve the multi-objective problem. Simulation results validate the effectiveness of the proposed charging strategy and demonstrate that incorporating algorithm aversion significantly improves the pricing performance of IEHSs over conventional approaches.