This paper proposes a lateral control strategy whose weight parameters of both the upper-level Linear Quadratic Regulator (LQR) and the lower-level torque distribution controller are simultaneously optimized for improving the lateral stability of multi-axial in-wheel-motors-driven vehicles. Considering the severe tire wear on multi-axial vehicles, the optimization objective of the lower-level torque distribution controller includes the generalized force tracking and the longitudinal tire slip energy. Moreover, to address the complexity of tuning LQR parameters as well as enhance control effectiveness, this study employs a genetic algorithm to optimize the weight parameters of both the LQR and the torque distribution controller. Simulations of emergency lane-change scenarios at different speeds on a low-adhesion road were conducted to demonstrate the strategy. The simulation results show that the proposed strategy effectively improves the vehicle's lateral stability and achieves a good overall control performance.