An Experimental Networked Control System with Fractional Order Delay Dynamics

Networked control systems gain recently more attention because of the possibility of providing remote control of industrial processes (with the usage of, e.g., TCP/IP protocol) efficiently and cost-effectively. One of the problems that control engineers has to face while developing network control systems is the spiky nature of the network random delays, which can be better characterized by so-called α-stable processes which are described by fractional lower-order statistics (closely linked to fractional order calculus). This paper presents an implementation example of a networked control system for the temperature control of a Peltier module platform developed by Mechatronics, Automation, and Embedded Systems Laboratory (MESALab) of the University of California, Merced. The system provides a user the possibility to remotely run a PID control algorithm executed by JavaScript code in the user's browser, performing the real-time control of the temperature of the Peltier module. Additionally the analysis of the system communication delays is performed. Obtained results show that the communication delays fits into a α-stable probability distribution with a heavy tail, indicating a fractional order behavior which effect can influence the control system performance.