MPC Based Power Allocation for Reliable Wireless Networked Control Systems

This paper considers a wireless networked control system (WNCS) where the controller and controlled object (plant) are connected via an unreliable wireless network. The control system should be designed considering the reliability of the given network while its network reliability is controllable by changing transmission power. Thus, this paper provides a joint optimization of transmission power and control-input for the focused WNCS, which is a kind of cross-layer design of communication and control layers. This work assumes that the total transmission power for the feedforward (from the controller to plant) and feedback (from the plant to controller) channels is limited and can be adaptively allocated to each wireless channel. If more transmission power is allocated to the feedforward channel, the control-input can be more reliably transmitted, while the control-state is less reliably transmitted over the feedback channel. Furthermore, the appropriate power allocation depends on the control-state and the previous communication results. Considering the above facts, this paper proposes a model-predictive control (MPC) based joint optimization and shows that the proposed system can provide appropriate power allocation and control-input and thus can enhance the quality of control.