System Identification and Control of Environment inside Greenhouse

本文采用系统辨识的方法对温室内的热环境进行了分析，探讨了系统辨识的适用性。在此基础上，设计了具有ovserver的最优调节器作为温室内热环境的控制系统，并对控制性能进行了检验，在塑料大棚内进行了系统辨识实验。首先，作者用MA模型（滑动平均模型）和阿尔马模型（自回归滑动平均模型）分析了一个以室外辐射为输入，室内气温为输出的系统。研究了系统阶次与辨识误差的关系。对根据室外辐射和通风机的开关来确定室内空气温度的系统进行了辨识，并在一个带有水-空气热交换器的塑料温室模型上进行了温室内热环境控制实验。观测器由阿尔马模型参数估计值计算。基于Riccati方程的解和状态反馈增益设计最优调节器。在LQ控制（线性二次型控制）中，采用一定的加权因子值对控制性能进行了实验研究。

In this study, the author analyzed the thermal environment inside greenhouse by means of system identification and investigated the applicability of system identification. Furthermore, as control system for the thermal environment inside greenhouse, optimal regulator with ovserver was disigned and control performance was examined.The experiment of system identification was carried out in a vinyl house. At first, the author analyzed a system with an input of outside radiation and an output of inside air temperature by means of MA model (moving average model) and ARMA model (auto-regressive moving average model). Relationship between order and error in identification of the system was also investigated. Identification of system which determines inside air temperature based on outside radiation and on-off of ventilating fan was also carried out.The experiment of control of the thermal environment inside greenhouse was carried out in a vinyl model greenhouse with a water-to-air heat exchanger for heating/cooling. Observer was calculated by the estimated parameters of ARMA model. Optimal regulator was designed based on solution of Riccati equiation and state feedback gain. Control performance was examined experimentally with some values of weighting factor in LQ control (linear quadratic control).