Novel Control Algorithm for Dynamic Emulation of Mechanical Loads of the Three-motor Test Bench of the Electric Vehicle

This paper presents a novel dynamic load emulation methodology applied in the three-motor test bench of the electric vehicle (EV), which can reproduce the motion states of the electric drive/brake system (EDBS) with high precision in longitudinal dynamics processes. First of all, the dynamics models of a front axis drive electric vehicle and a three-motor test bench are established. On the basis of the existed model, a novel dynamic load emulation strategy, which utilizes an adaptive sliding mode controller (ASMC) with an extended state observer (ESO) for speed tracking of each loading electric dynamometer and adopts the cross-coupling control scheme and an adaptive sliding mode speed synchronization controller to attenuate the load emulation error resulting from the unsynchronized rotational angular velocities between the two loading machines. The entire load emulation system has strong robustness and can reject the external disturbances coming from the vehicle motor and system parameters vibrations. Simulations are carried out in MATLAB/Simulink to verify the validity of the newly proposed control law, and results show that the novel load emulation strategy can reduce the load emulation error and eliminate the synchronous error significantly compared with the conventional parallel PI control.