Adaptive LQR Active Control of Pantograph Based on MDO Algorithm

To enhance the dynamic performance and current collection quality of the high-speed railway pantograph-catenary system, this paper proposes an adaptive LQR active control strategy based on the Multi-strategy integrated Dandelion Optimization (MDO) algorithm. First, the standard dandelion algorithm is enhanced through multiple strategies to improve its global search capability and convergence speed, which is then used to optimize the weighting matrices of the adaptive LQR controller. The controller employs a displacement-scheduled approach, applying distinct optimized weighting matrices to different displacement intervals and achieving smooth switching via linear interpolation, thereby effectively adapting to the system’s nonlinear time-varying characteristics. Simulation results based on a coupled dynamic model demonstrate that the proposed ALQR-MDO strategy significantly improves current collection quality and exhibits strong anti-interference capability.