铝基覆铜板导热系数测试仪温控系统设计

Design of Temperature Control System for Thermal Conductivity Coefficient of Aluminum Clad Laminate

温控系统设计的优劣直接影响着铝基覆铜板导热系数测试平台的测量准确性。但其数学模型建立难,受控调节器件多,对该系统的设计提出了很高的要求。分析常规PID控制效果不佳的同时,针对该温控系统非线性、小时滞的特点,结合Q-leaming的扩张寻优策略与Metropolis的动作选择准则,提出基于Q-leaming(Metropolis准则下)的PID控制算法(简称Qm-PID)。与已有控制方法仿真对比后发现,该方法超调小、调节快、精度高、可行性大,为PID控制方法的延伸提供了一定程度的借鉴意义。

The performance of temperature control system directly influences the measurement accuracy of the thremal conductivity coefficient test platform of aluminum base clad laminate. However, the difficult construction of mathematic model and the numerous quantities of controlled cells make it difficult to design this system. Based on the analysis of the bad control effectiveness of regular PID and the non-linear and the small time-de- lay features of the temperature control system, a PID algorithm based on Q-learning under the Metropolis rule is proposed by combining the search optimizing strategy and the motivation choosing rule of Metropolis. Compared with the existing control approaches, the proposed approach is flexible to some extent due to its small overshoot, quick adjustment and high accuracy, therefore, it is significant in the field of PID control approaches.