Improving control loop performance using dynamic rate adaptation in networked control systems

The dynamic rate adaptation was initially proposed to improve the operational flexibility of networked control systems at a global level. In a system with several control loops the sampling rate of each loop would be decreased in order to free bandwidth that will be used to schedule pending messages or to overcome possible overload situations during a short period of time. This paper proposes the use of dynamic rate adaptation to improve the control performance of individual loops when these loops are subject to long sampling to actuation delays. This paper presents preliminary simulation results obtained for different plants. It shows that, although counter intuitive, the decrease of the sampling rate of the loops can improve the control performance and the usefulness of the dynamic rate adaptation in this context.