Optimization of super twisting sliding mode control gains using Taguchi method

• Main Authors: Jamaludin, Zamberi, Chiew, Tsung Heng, Bani Hashim, Ahmad Yusairi, Rafan, Nur Aidawaty, Abdullah, Lokman
• Format: Article
• Language: English
• Published: Korean Institute of Industrial Engineers 2018
• Online Access: http://eprints.utem.edu.my/id/eprint/21825/2/IEMS.pdf; http://eprints.utem.edu.my/id/eprint/21825/; http://www.iemsjl.org/journal/article.php?code=60302

This paper focuses on optimization of super twisting controller gains using Taguchi method with objective to minimize tracking error and the chattering effect. Two gain parameters in super twisting algorithm, that is L and W were identified as two factors with three levels respectively. The optimization method applied a L9 orthogonal array and the performance index used was root mean square of tracking error and Fast Fourier Transform of control inputs. The optimized super twisting controller with traditional sliding surface and the continuous control action laws was validated on a single axis direct driven linear motor. Analyses of variance and main effect plots were performed on the effect of gains variation on performance index. Values of L and W were chosen as 0.00002 and 0.08 respectively and were confirmed through confirmation test based on calculated confidence interval. Experimental results with 95% confidence level identified gain L as the significant factor in minimizing chattering effect while both gains L and W were responsible in minimizing tracking error in optimum condition. Optimized algorithm achieved 9.3% of reduction in root mean square of tracking error and 38.4% of reduction in chattering experimentally.