A novel approach to suspension control in electric vehicles with in-wheel motors: Finite horizon stochastic LQR with NSGA-II algorithm optimization

A robust suspension-control strategy is presented for electric vehicles equipped with in-wheel motors and dynamic vibration absorbers (DVAs). A quarter-car model incorporating an in-wheel motor and an adjustable DVA is developed to mitigate excess unsprung mass. A finite-horizon stochastic linear quadratic regulator (FHS-LQR) that explicitly accounts for control delays and parameter uncertainties is coupled with an NSGA-II multi-objective optimizer to tune both DVA mechanics and controller weights. Compared with passive and conventional LQR baselines, the proposed framework reduces the root-mean-square values of sprung-mass acceleration, suspension deflection, tire dynamic load, and motor dynamic force by up to 40% under ISO 8608 road profiles and maintains performance for input delays up to 100 ms, demonstrating real-time feasibility and improved ride comfort.