High-fidelity-model-driven Co-optimization Method for the Design and Control of Tracked Vehicles

Tracked vehicles, as key platforms in the land combat system, are confronted with the challenges of high performance requirements and short development cycles. The current development of tracked vehicles mostly adopts the first-design-then-control workflows. During the design stage, the rough-road excitation is insufficiently considered the longitudinal and vertical vehicle dynamics are often decoupled, and the coupling effects of design, control and road conditions on the maneuverability and smoothness of tracked vehicles, thus limiting the improvement of mobility and ride comfort. To address this issue, a high-fidelity-model-driven co-optimization method is proposed for tracked-vehicle design and control. A multi-objective and multi-constraint optimization problem is formulated for washboard roads and Class A to F road profiles. The dynamic constraints are described by a full-vehicle-test-validated 22-degree-of-freedom Lagrangian vehicle model coupling the vehicle body, suspension system and superstructure payload. The suspension stiffness and damping, motor characteristics, transmission ratio, center-of-gravity position and drive torque control parameters are optimized simultaneously using direct collocation. Simulated results show that the travel time is reduced by 11.09 and the pitch stability is improved by 73.66 on the washboard-road. On Class A to F roads, the mobility and ride comfort are improved by at least 1.62 and 29.05, respectively. The optimized design parameters provide references for vehicle parameter matching and real-time drive control strategy design.