Benchmark and Verification of Control Algorithm for Flywheel with Active Magnetic Bearing on Electric Vehicle and Proposal of New SAC Algorithm ($\epsilon$1 modification and bias variable $\Gamma$p approach)

This paper presents the requirements which a good rotor controller of flywheel energy storage system (FESS) with Active Magnetic Bearing (AMB) on Electric Vehicle (EV) should satisfy. AMB-FESS is used for energy regeneration. The dynamics of the flywheel rotor on EV is very complex and to compare various features a benchmark is necessary. Controllers are evaluated according to certain criteria as used by many motor companies. Controllers are selected to satisfy the design requirements prescribed by many motor companies. Thus the robust adaptive controllers were chosen and benchmarked using simulations and experiments. The investigation revealed that a controller with high rigidity gain and a reference model can not only suppress disturbance effects but also lower power consumption. Therefore Simple Adaptive Control with epsilon1 modification and bias variable gamma P approach is proposed. The utility of the new controller experimentally was proven.