PI/PID Fine Tuning Method Based on Experimental Design for the First Order System with Time Delay

• Main Authors: Ming-hong Shieh, 謝旻宏
• Other Authors: Chin-min Hsiung
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/39360317713430666855

碩士 === 國立屏東科技大學 === 機械工程系 === 92 === The PID controllers are wildly applied in industry for its mature technology and easy to use. Mathematical measures are utilized in most controllers’ design; however, practical methods are more eagerly needed in factories. Based on experimental Design three PID controllers’ processes are provided in this research. These three experimental methods are Response Surface Methodology, Steepest Descent Method, and Fractional Factorial Design. There are steps of Response Surface Methodology search: through systemic experiment to improve statistics, so as to attain the final reflective point. The result shows that the effective PID controllers’ statistics which is different from theoretical method are provided in this research and are suitable for industry application. Steepest Descent Method is basically similar to Response Surface Methodology Search, but its searching implement is Steepest Descent Method on the second step. The conception of Fractional Factorial Design is to replace the Full Factorial design with Partial Factorial design in fine-tuning experiment. The design of treatment could avoid the necessity of using strict precision and the difficulty of discriminate treatment from inaccuracy. By exercising treatment, the better controlling parameter could be successfully attained. Furthermore, the efficiency increased two times when compared with Full Factorial design. PID controller could be adjusted to the best status through the three fine tuning processes which mentioned above. In short, this article provides PID controller’s steps in fine tuning and design method which is different from other theoretical analysis.