Hardware Implementation of PID Autotuning with Efficient Particle Swarm Optimization

Abstract

Severa1 intelligent control systems these days are utilized by the concept of automatically tuning, especially in proportional-integral-derivative (PID) controller. Furthermore, increasing sensors and actuators disrupt the conventional computing system, which has limited resources and is difficult to meet the timing requirement. This paper develops a hardware implementation of PID auto-tuning based on the particle swarm optimization (PSO) with the parallel architecture of particles and variables so that latency is heuristically minimal. The result shows that the proposed hardware performs 1000x and 800x compared to conventional microprocessors. From the evaluation process, performance and resource utilization are found to be satisfactory compared to conventional PID.