Semi-active fuzzy control of a powerpack with magneto-rheological dampers at full operating conditions

Abstract

Powerpacks are commonly used as the power source in diesel multiple units (DMUs) due to their flexibility and economy. In order to enhance the ride comfort of DMUs at full operating conditions, it is a great challenge to design a multiple objectives semi-active control system to reduce the transmission force to the DMU, vibration intensity at start-stop conditions, and structural vibration in stable operating conditions. A fuzzy semi-active control system incorporating magneto-rheological (MR) dampers and vibration isolators with small stiffness coefficients is designed. Firstly, a rigid-flexible coupling powerpack model is established to reflect the rigid body and structural vibration characteristics. Next, an MR damper’s dynamic behavior at full operating conditions is investigated experimentally, direct and inverse models are established using a genetic algorithm backpropagation (GA-BP) neural network. Then, a variable universe hybrid fuzzy controller is designed for the multiple objectives requirements, including quick response during low-frequency start-stop conditions and effective suppression of structural vibration during high-frequency stable operating conditions. Finally, the vibration isolation performance of the powerpack is verified by simulation and experiments, demonstrating the effectiveness of the semi-active control system with MR dampers and the variable universe hybrid fuzzy controller at full operating conditions.