Optimization Of Pid Controller For Double-Link Flexible Robotic Manipulator Using Metaheuristic Algorithms

This paper investigates the optimization approach of PID controller for double-link flexible robotic manipulator using metaheuristic algorithm. This research focus on population-based metaheuristic that is particle swarm optimization (PSO) and artificial bees algorithm (ABC) to tune the PID control parameters of the system. In the tuning process, the number of iteration was set and the number of particles was varied. The tuning process was interrupted once the convergence value of Mean Square Error (MSE) was achieved. For PSO, it was found that when the number of iteration increased, or the number of particles were set to higher values, there were no significant improvement of MSE. Results showed that 25 iterations were required for MSE to converge for hub angle and 20 iterations were required for MSE to converge for endpoint acceleration. Meanwhile, it was discovered that ABC portrayed the same pattern with PSO whereby when the number of iteration increased or the number of colony sizes were set to higher values, there were no significant improvement of MSE. From the results, 15 iterations were required for MSE to converge for hub angle and 25 iterations were required for MSE to converge for end-point acceleration. The performance of the algorithm was validated by evaluating the performance of the controllers in comparison with the conventional controller that is Ziegler Nichols (ZN) in term of input tracking capability and vibration suppression for both links. The system managed to reach desired angle for both hub angle 1 and 2. Besides, vibration reduction shows great improvement for both link 1 and 2. This signifies that, the PSO and ABC algorithm are very effective in optimizing the PID parameters