Robust neural fixed-time sliding mode control for active suspension systems with heave–pitch dynamics

Most existing studies on vehicle suspension systems primarily focus on ride comfort and road holding performance, while the influence of pitch dynamics is often insufficiently addressed. This issue becomes more significant under real driving conditions, where coupled heave–pitch motions strongly affect vehicle stability and passenger comfort. To overcome these limitations, this article proposes a robust control framework capable of simultaneously regulating both heave and pitch dynamics. The proposed strategy combines smooth fixed-time sliding mode control with an active disturbance rejection control to enhance disturbance compensation and closed-loop convergence. In addition, a radial basis function neural network is incorporated to adaptively adjust the control gain online, thereby enhancing the controller's ability to accommodate nonlinearities and varying operating conditions. Furthermore, a detailed nonlinear suspension model is developed that accounts for suspension nonlinearities, hydraulic actuator dynamics, external disturbances, and lumped uncertainties. Numerical simulations are conducted under both step and stochastic road excitations. The obtained results show that the proposed method provides superior vibration suppression and pitch stabilization performance compared with conventional controllers. Under stochastic excitation, the root mean square body displacement is reduced by up to 72.7%, while the root mean square pitch angle is reduced by more than 90.1%. These results show the potential effectiveness of the proposed approach in improving ride comfort and vehicle stability while maintaining a reasonable level of control effort.