采用预测控制进行纵向塞曼幅值热稳频的研究

Study of a Thermally Compensated He-Ne Zeeman Laser Frequency Stabilization System with Model Predict Control Algorithm

产生纵向塞曼频率分裂的He—Ne激光器输出左、右旋圆偏振光,其光强差和激光频率呈稳定的调谐曲线。装置利用单片机采集幅值信号,采用预测控制理论进行数字调节,输出PWM(PulseWidth Modulation)信号补偿激光管的温度以实现热稳频控制。为实现对热稳频大滞后、非线性这一复杂系统的高质量控制,提出了预测控制方法,该方法克服了传统PID(Proportional IntegralDifferential)调节中存在的调整工作量大,控制后果不能及时反馈的不足,适用于各种热稳频过程。数字仿真和实验验证预测控制是稳定的,对模型误差的适应能力强。对10s采样周期,估算得系统频率稳定度达6.7×10^(-9)。

A thermally compensated frequency stabilization system is nonlinear with large time delay. Its control process is too complex to describe with the traditional model. As a result, the conventional PID algorithm designed by experiment has the disadvantages of difficulty in adjustment and control performance. This paper presents a new algorithm; model predict control (MPC) algorithm, which uses the Hammerstein model for forecasting the process output over a certain horizon in the future, optimizing the control in long-range predictive control strategy. The concept is implemented in a real industry scale single-chip-processor controlled thermally compensated frequency stabilization system. The result indicates that the algorithm is effective. The laser tube reached a frequency stability of 6. 7 ×10-9 in the sampling period of 10 s.