Multi-variable networked predictive control with predefined decoupling

Networked Predictive Control (NPC) scheme has been proposed to tackle the disturbance caused by the network uncertainties in the networked control systems (NCS). In NPC scheme, at the controller side future control sequences for every possible time delay are generated, which are then packed into a single packet and transmitted to the plant side. The plant side selects the proper control signal based on the time delay measurement. The NPC can compensate for the effect of network uncertainties effectively. However, in NPC the control systems are normally divided into two parts: the controller side and plant side. That structure is not suitable for the multi-variable control in large-scale industrial applications (power plants and chemical factories etc.). In order to solve that problem, a Multi-variable NPC scheme is proposed in this paper. In this scheme, the controller, sensors and actuators are allowed to be located diversely, which fits the situation in large-scale industrial processes. The algorithms in the controller, sensors and actuators are redesigned to compensate for the random network delay and data dropout. To verify the effectiveness of the new method, a simulation of a boiler control system is given at the end of the paper.