PID robust control using Taguchi Method

The control of FOPD (first order plus delay) systems with high normalized time (ratio of time delay and time constant) is a very hard task, especially when the system parameters are subjected to variations. A great number of physical systems show such model behavior. In process control today, proportional integral derivative (PID) (continuous and digital) controllers still predominate and are sufficient for most needs with more than 95 per cent of control loops being of the PID type. In this work, we present a method of determination of an optimal robust PID controller based on the ISE (integral squared-error), using Taguchi method. This approach is very useful when the process parameters are prone to variations in a given range. Once the criteria of optimization posed, a statistical analysis is needed to determine if the control parameters are significant or not. The design of experiments method is certainly the most appropriate to tackle this type of problem. Indeed it allowed us to quantify the weight of the factors that affect the output and evaluate their interactions. The analysis of the results led us to define the most appropriate set up that minimizes the negative effect of noise factors.