本論文係針對熱循環儀，設計與實現具有即時基因演算法調校參數之類PID模糊控制器。熱循環儀通常包含兩個以上目標溫度，並且需要快速地升溫與降溫，並穩定地保持在目標溫度，以達成系統需求。因此，精確快速的溫度控制，可以提升熱循環儀的性能。然而，元件公差與老化，都會影響控制器的穩定性而造成控制失準。因此，本論文利用模糊控制器調整類PID控制器的輸出變量。模糊控制器的輸入為溫度誤差與誤差變量，而輸出則作為類PD控制器的輸入及類PI控制器的輸入變量。而類PI控制器的輸入變量，加上其上一時刻之輸出即可作為類PI輸出量。最後、整合類PI控制器的輸出量與類PD控制器的輸出量成為類PID控制器的輸出量。除此之外，本論文更整合基因演算法，用於調整控制器輸入與輸出的增益，其中包含誤差增益、誤差變化量之增益、類PD控制器輸出的增益以及類PI控制器輸出變量的增益。並以時間乘絕對誤差積分值(ITAE)、熱循環時間以及最大誤差等參數之乘積的倒數，當作適應度函數，降低參數的單位不同所造成的影響，以演化最佳控制參數。本論文將此控制器實現於熱循環儀之嵌入式系統中。為了避免異常狀況發生，針對量測溫度、最大誤差與測試時間等不同參數設定邊界值。實驗結果顯示此方法不僅可以縮短熱循環時間，降低過衝量，亦可以自動調校控制器系統，取代人工調校與節省人力資源。

This thesis proposes a real-time GA based PID-Like fuzzy controller to tune the parameters of the controller for a thermal cycling system. The thermal cycling system usually has two or more target temperatures, and the controller has to quickly heat up and cool down current temperature to the desired values while keep the target temperatures for a predefined period stably. Accurate and fast temperature control is the most important issue for the thermal cycler. However, component tolerances and aging may reduce the stability of the controller and result in a failure control. Therefore, we propose a control design that possesses the real-time self-tuning capability. The thermal cycler system is controlled by a PID-like controller, and a fuzzy controller is added to adjust the output value. The inputs of the fuzzy controller are temperature error and its change of error, whereas the outputs of the fuzzy controller are the inputs of the PD-like controller and the PI-like controller. Finally, the outputs of the PD-like and PI-like controllers are added together as a total output of the PID-like controller. Besides, we integrate a genetic algorithm to tune several gains of the controller, including the gain of errors, the gain of the change of errors, and the gains of the outputs of PD-like and PI-like controllers. The fitness function of the genetic algorithm includes the integral of time-weighted absolute error (ITAE), thermal cycle time, and maximum error. Furthermore, we implement the proposed controller in an embedded system and set some boundaries of the parameters for avoiding the abnormal conditions. The experimental results show that this method can not only reduce the thermal cycle time and the overshoot but also real-time tune the parameters of the controller to replace the manual turning.

[1]C.-C. Kao, C.-W. Chuang, R.-F. Fung, “The self-tuning PID control in a slider–crank mechanism system by applying particle swarm optimization approach,” Mechatronics, Vol. 16, pp 513–522, October 2006
[2]J.H. Holland, “Genetic algorithms,” Scientific American, vol. 267, pp. 66–72, 1992.
[3]D. E. Goldberg, “Genetic Algorithms in Search Optimization and Machine Learning, ” Addison Wesley, pp. 41, ISBN 0-201-15767-5, 1989.
[4]Thermocouple, https://en.wikipedia.org/wiki/Thermocouple
[5]F. -J. Lin; W. -D. Chou, and P. -K. Huang, “Adaptive sliding-mode controller based on real-time genetic algorithm for induction motor servo drive,” IEE Proceedings - Electric Power Applications, vol. 150, no. 2, pp. 1–13, Jan. 2003.
[6]C.-T. Wu, J.-P. Tien, T.-H. S. Li, “Integration of DNA and Real Coded GA for the Design of PID-Like Fuzzy Controllers,” in Proceedings of 2012 IEEE International Conference on Systems, Man, and Cybernetics (SMC 2012), Seoul, Korea, pp. 2809–2814, Oct. 2012
[7]Polymerase chain reaction, https://en.wikipedia.org/wiki/Polymerase_chain_reaction
[8]Genetic Algorithm, http://en.wikipedia.org/wiki/Genetic_algorithm
[9]Thermoelectric effect, https://en.wikipedia.org/wiki/Thermoelectric_effect
[10]TMS320F28335, http://www.ti.com/product/TMS320F28335/description
[11]ITS-90 Table for type K thermocouple, https://srdata.nist.gov/its90/download/type_k.tab
[12]ADS1220, http://www.ti.com/product/ADS1220
[13]ADS1118, http://www.ti.com/product/ADS1118
[14]L. A. Zadeh, “Fuzzy sets,” Information and Control, vol. 8, pp. 338–353, 1965.
[15]H. R. Li and H. B. Gatland, “Conventional Fuzzy Controller and Its Enhancement,” IEEE Trans. Systems, Man, and Cybernetics-Part B, vol. 26, no. 5, pp. 791–797, Oct. 1996.
[16]M. A. Sahib, “A novel optimal PID plus second order derivative controller for AVR system,” Engineering Science and Technology, an International Journal, Vol. 18, pp. 194–206, June 2015
[17]A. Rubaai and P. Young, “EKF-Based PI-/PD-Like Fuzzy-Neural-Network Controller for Brushless Drives,” IEEE Transactions on Industry Applications, pp. 2391 - 2401, Oct. 2012
[18]A. Rubaai and P. Young, “Hardware/Software Implementation of Fuzzy-Neural-Network Self-Learning Control Methods for Brushless DC Motor Drives,” IEEE Transactions on Industry Applications, pp. 414 – 424, Aug. 2015
[19]J. Kennedy and R. Eberhart, “Particle Swarm Optimization,” in Proc. IEEE Int. Conf. Neural Networks, vol. 4, pp. 1942–1948, 1995
[20]Y. Wang, B. Li, T. Weise, J. Wang, B. Yuan, and Q. Tian, “Self-adaptive Learning Based Particle Swarm Optimization,” Information Sciences, vol.181, no. 20, pp. 4515-4538, 2011.