Network based control with compensation of time-varying delay and modeling error

This paper describes a control method which compensates two factors in a network based control system: one is a time-varying delay on a network communication and the other is a modeling error due to uncertainty of the plant. For the time-varying delay, we model it with an accurate modeling by measuring the individual time delays, and utilize them for the model-based compensator. Because the time delay on the network is not completely uncertain but measurable value. If the values are known and explicitly used for the control, performance improves. In contrast, the perfect dynamics modeling of the plant is very difficult and the real plant may vary its property due to the variation of loads or deterioration through aging, therefore the modeling error is inevitable. The modeling error is considered to be a loop gain variation (multiplicative uncertainty) with phase delay due to the time delay. Therefore, we propose a compensator which is designed based on robust control theory for the uncertainty of the plant. To evaluate the performance of the proposed system, we carried out simulations and experiments on a motor control system. From the results, we verified the stability and satisfactory performance of the system with the proposed methods.