Research And Design Of Greenhouse Soil Moisture Control System Based On CPS

With the development of agricultural technology,facility agriculture has become an important part of modern agriculture.Greenhouse as an important agriculture facility,which is developing rapidly.Greenhouse control system as the core of the greenhouse,the precision of environmental control will directly affect the healthy growth of plants and crop yield in greenhouse.Therefore,greenhouse control is bound to be refined.In this paper,the greenhouse control system based on CPS(Cyber-physical system)was studied,with emphasis on the fine control of greenhouse soil moisture.Because of the strict demand for soil moisture in the process of growing seedlings with Feijoa,the greenhouse was chosen as the research object.Through extensive reference to domestic and foreign typical literature,the CPS space-time event model was used as the technical basis.Aiming at the correctness and predictability of environmental perception,the soil moisture sensor placement algorithm and soil moisture dynamic prediction model were studied.This paper has carried on the meaningful exploration for the research and the promotion of greenhouse fine control,mainly carries on the research from the following several aspects:The development and application of greenhouse control system and CPS were studied.Based on the CPS space-time event model,the greenhouse sprinkler irrigation event model was analyzed.Aiming at the difficulties of greenhouse soil moisture control system,the whole scheme of greenhouse soil moisture control system based on CPS was designed.Aiming at the spatial variability of greenhouse soil,a point placement algorithm combining regular mesh and Delaunay triangulation was designed,and the variation coefficient was used to constrain the triangulated grid.The distribution characteristics of soil water content in greenhouse are determined by a few sensors,which not only guarantee the perception accuracy,but also reduce the cost and data redundancy of the traditional placement method.Aiming at the strong time-delay of greenhouse soil moisture change,a neural network dynamic prediction model based on Delaunay triangulation optimization algorithm was designed,andspace-time factor is introduced into the model.The change of surface soil water potential was used to characterize the internal water flow of soil,and finally to predict the soil water content at the root system in the future,which provides the decision basis for the control system.The hardware design of soil moisture control system in CPS greenhouse was completed.Aiming at the central computing unit(DSP),the software platform of SYS/BIOS real-time operating system was built,and the management and distribution of multi-thread were realized.The man-machine interface based on TCP/IP was designed by using Lab VIEW.Finally,the software and hardware of the greenhouse soil moisture control system were tested.The test result show that,the control system can effectively control the soil water content in the root system within the preset range.It is helpful to improve the survival rate of Feijoa cuttage breeding.