Research Of The Control Problems In Embeded Greenhouse Monitoring System Based On WSN

The current overall modernization level of facility agriculture is still very low in our country. Most domestic greenhouses are using simple greenhouse control system, even totally controlled by hand. With the rapid development of economy and the raising of agricultural production level, the proportion of greenhouse with environment control facilities has increased year by year, the demand of trancition from traditional facility agriculture to modern facility agriculture is increasingly urgent.The applicability problem of the greenhouse control system is hampering the development of greenhouse industry. The intelligent greenhouse control systems are needed by large amounts of greenhouse, which systems are supposed to have the advantage of low-cost, high reliability, convenient operation, and high applicability. This research adopted embedded system instead of IPC, chose wireless sensor network instead of wired network, so the applicability problem of the traditional greenhouse control system was fundamentally eliminated.The main subjects this research project completed are listed below:the hardware architecture and system functions of the embedded greenhouse monitoring system was designed based on wireless sensor networks; A fusion algorithm of the greenhouse sampling data was studied according to the characteristics of WSN, which algorithm combined with the consistency test, batch estimation and aptive weighted averaging together; For the unstable problem of greenhouse control system caused by the discreteness and delay of WSN sampling feed back data, the study proposed a new method to eliminate the discreteness and delay, which method adopted the recursive grey prediction and spline interpolation algorithm to handle the feedback data; The dynamic greenhouse mass and energy balance equation was studied and a greenhouse simulation model in Simulink environment was constructed based on the equation. The rationality of the simulation model and the effectiveness of the grey prediction and interpolation method were verified through a number of simulation experiments; A practical decoupling control method was studied, which method could make the greenhouse temperature and humidity decoupled and linearized by feedback and feedforward. The control law that made the system completely decoupled was easily deduced by this method, and consequently the nonlinear MIMO system was transformed into two individual linear first-order SISO systems.The methods this study proposed solved several great problems that hamper the application of WSN in embedded greenhouse real time monitoring system. The grey prediction and interpolation method presented in the paper could ensure good real-time performance of the WSN sampling data without increasing the sampling frequency. With the greenhouse simulation model constructed in this study, greenhouse simulation experiments under various conditions could be conveniently carried out. Through the adopted data fusion method, WSN sampling precision could be effectively increased. The feedback plus feed forward linearization decoupling method could be easily applied in the greenhouse control system equipped with an embedded platform because of its simplicity.