Fuzzy genetic control for linear speed in multi-machine systems

Abstract

In today’s fast-moving industrial sectors which include paper, textile, and plastic manufacture the core of production quality is in the precise coordination of multi-drive systems. While PI controllers are the mainstay of the industry, they do have issues in that they struggle with the nonlinearity and dynamics of large-scale windings, which in turn causes instability and product integrity issues. To that end, this paper presents an optimized fuzzy-genetic controller (FLC-GA), which we put forward as a better linear speed synchronization solution. We used genetic algorithms in the tuning of fuzzy logic parameters, which also takes out the very time-consuming task of manual calibration, and at the same time sees a great increase in the system’s ability to deal with process variability. We put our FLC-GA through its paces in a head-to-head comparison with the classic PI and PI-PSO controllers. What we found was that our proposed controller did very well; we saw zero overshoot, a quick 0.5 s settling time, and the total elimination of tension ripples. Also, we saw from a 13.2% change in system inertia that the FLC-GA did a 65% better job in terms of speed accuracy and stability than what we see from standard PI control. We present the FLC-GA not only as a theoretical improvement but as a very robust, high-performance solution in the very tough field of continuous industrial synchronization.