基于遗传算法自整定和Smith预估的电子倍增电荷耦合器件温控系统设计

EMCCD Temperature Control System Based on Adaptive Genetic Algorithm and Smith Predictor

针对电子倍增电荷耦合器件(EMCCD)温度控制过程的时间滞后特性,根据热力学理论建立整个系统的数学模型,利用有限元仿真方法得到其传递函数,在此基础上采用遗传算法实现PID控制参数的在线自适应整定,同时结合Smith预估控制补偿由滞后环节产生的时间迟滞。采用Matlab对控制算法进行仿真分析,结果表明该方法较常规PID控制方法具有调整时间短、超调量小、抗干扰能力强等优点。设计基于单片机和TEC驱动模块的硬件电路将控制算法应用于EMCCD温度控制系统,相机实际制冷温度可达(-12±0.1)℃,图像本底噪声方差仅为18。实际应用结果表明温度控制系统是提升EMCCD相机成像性能的有效方法。

Referring to the time-delay of EMCCD temperature control object,the mathematical model based on the thermodynamic theory of the system is established and the transfer function is acquired by the method of finite element simulation. The parameters of PID controller are self-turned on line by genetic algorithm. And the Smith predictor is adopted to compensate the time delay which is generated by the inertia hysteric. Simulation results show that compared with conventional PID the algorithm has short regulating time,small overshoot,strong anti-interference and is convenient in parameters turning. The circuit based on MCU and TEC driving module is designed to apply the algorithm in EMCCD imaging system. The refrigerating temperature of system reaches(- 12 ± 0. 1) ℃ and the background noise is only 18. The application examples illustrates that the temperature control system is effective for improving performance of EMCCD imaging system.