Modeling and control of closed-loop networked PLC-systems

This paper presents a novel approach for modeling and control of Programmable Logic Controller based Networked Control Systems (PLC-based NCS). The approach allows control design of NCS using formal discrete modeling and simulation. This involves three steps. The first step is to design a structure-conserving hierarchal timed model for the whole PLC-based NCS using Colored Petri Nets (CPNs). The aim is to generate by simulation extensive sampled time delay data records that are used to analyze and model delay behavior. The second step is the bridge between formal modeling and simulation and NCS control design. It uses Direct and/or Hidden Markov Models (DMM, HMM) to describe mathematical models of NCS delays. The paper introduces a new concept of Mutual Markov modeling to analyze the interaction between the two induced delays, namely, the sensor-to-controller time delay τ_sc and the controller-to-actuator time delay τ_ca. In the third step, the proposed mutual Markov models are used to design a single mode dependent state feedback controller using jump linear systems (JLS) approach. The control design solves a stability condition using Linear/Bilinear Matrix Inequalities (LMI/BMI). A numerical example is provided to demonstrate the proposed procedure.