Predictive Ecological Cooperative Control of Electric Vehicles Platoon on Hilly Roads

Abstract The integration of eco-driving and cooperative adaptive cruise control (CACC) with platoon cooperative control (eco-CACC) has emerged as a pivotal approach for improving vehicle energy efficiency. Nonetheless, the prevailing eco-CACC implementations still exhibit limitations in fully harnessing the potential energy savings. This can be attributed to the intricate nature of the problem, characterized by its high nonlinearity and non-convexity, making it challenging for conventional solving methods to find solutions. In this paper, a novel strategy based on a decentralized model predictive control (MPC) framework, called predictive ecological cooperative control (PECC), is proposed for vehicle platoon control on hilly roads, aiming to maximize the overall energy efficiency of the platoon. Unlike most existing literature that focuses on suboptimal coordination under predefined leading vehicle trajectories, this strategy employs an approach based on the combination of a long short-term memory network (LSTM) and genetic algorithm (GA) optimization (GA-LSTM) to predict the future speed of the leading vehicle. Notably, a function named the Notch-Filter function ( NF ( $$\varphi$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>φ</mml:mi> </mml:math> )) is introduced to transform the hard state constraints in the eco-CACC problem, thereby alleviating the burden of problem-solving. Finally, through simulation comparisons between PECC and a strategy based on the common eco-CACC modifications, the effectiveness of PECC in improving platoon energy efficiency is demonstrated.