| Real-time active suspension control remains a very popular and challenging topic in academia and industry today. Ground vehicle systems have complex dynamics characterized by nonlinearities, vehicle pitch and roll, vehicle flexibility effects, static and dynamic load effects, unknown frictions, deadzones and high amplitude disturbances. The result is excess vibrations that contribute to artificial vehicle speed limitations, reduced vehicle-frame life, biological effects on passengers, and detrimental consequences to cargo. Active suspension control aims to ameliorate these undesirable vibration effects. Ultimately, it should provide improvements to the criteria of passenger comfort, road handling and load carrying. Unfortunately these criteria are incompatible with each other such that current automobile suspension systems using passive components can only offer a compromise between them by providing fixed-rate spring and damping coefficients.; This dissertation presents an active suspension control approach for a ground vehicle system. It consists of inner ride control loops that reject terrain disturbances, outer attitude control loops that stabilize heave, pitch and roll responses, and an input decoupling transformation which blends the inner and outer control loops in a rigorous mathematical formulation. The input decoupling transformation is formally determined as the right inverse of the transpose of the full-vehicle Jacobian matrix. This approach overcomes the problem of a system such as a ground vehicle system that has reduced control effectiveness. The overall active control scheme is similar to the stability augmentation system (SAS). Feasibility of such an active suspension control approach is becoming more apparent with the advent of embedded real-time control systems like those developed at ARRI.; We have developed a real-time control system that is crucial to implementation of active suspension controllers on ground vehicle systems. The “PC-PC” controller simplifies implementation of advanced control algorithms and provides unprecedented levels of realtime system monitoring and user interaction. A variety of systems that traditionally use classical control methods potentially benefit from advanced control algorithms. |
