Research And Implementation Of Intelligent Household Greenhouse System Based On Multi-sensor Fusion

With the development of urbanization and the increase of urban population,people's demands for quality of life are also increasing.Green environment is more and more important to people's healthy life,growing plants in home or working environments has become a trend.The current smart greenhouse technology is generally used in the cultivation of agricultural greenhouses,but it has never been introduced into the home or work environment.With the concept of IOT is proposed to the rapid development of artificial intelligence nowadays,more and more of these technologies are gradually applied to society and life.And thus triggered a profound social change and the rapid development of science and technology.Along with the people's yearning for the beautiful home environment and the increasing demand for the household greenhouse,an intelligent household greenhouse system is designed and developed by combining the related technology of the IOT and artificial intelligence skillfully.The concept of IOT is used in the intelligent household greenhouse system.The system is designed with three-tier architecture,it includes the perceptual layer,the network layer and the application layer.A multi-sensor data level fusion algorithm is designed for Data collection in greenhouse environment,which aims to get the more accurate data values of greenhouse environment.And the control method under the automatic control mode of the household greenhouse system is controlled intelligently by using the T-S fuzzy neural network controller which is trained and then the control strategy is obtained.The main contents of this paper are as follows:(1)The demand of intelligent household greenhouse system is analyzed,and four kinds of functions are designed according to requirement of the system,including greenhouse environment parameter detection,greenhouse environment adjustment and control,environment data transmission and system man-machine interaction.The architecture of intelligent household greenhouse system based on IOT is divided into three layers:perceptual layer,network layer and application layer.(2)The structure of household greenhouse is designed.In the greenhouse system,the hardware of the ZigBee network in the perceptual layer and the STM32 intelligent gateway in the network layer is also designed.(3)The design of system communication protocol and various application software are presented in this paper.A complete multilayer communication protocol is designed to meet the requirement of bidirectional transmission of data and control information between layers.The software design of the ZigBee network of the system perception layer,the STM32 intelligent gateway of the network layer,the computer software of the network layer and the App of the application layer is carried out respectively.(4)The purpose and function of the data level fusion in the intelligent household greenhouse are explained briefly,and the effect of improving environment is obtained by improving the accuracy of sensing environmental data.An adaptive weighted fusion algorithm based on support degree is designed.Through the simulation experiment,the fusion result obtained by this method is proved to be little variation,high accurate and stable.This method is suitable for data level fusion of intelligent household greenhouse system.(5)The T-S fuzzy neural network model is selected to obtain the control strategy for the automatic control mode in the intelligent household greenhouse system.The model of T-S fuzzy neural network in temperature regulation control and humidity regulation control is established.The sample data are trained and the fuzzy neural network is tested by the data.The intelligent control strategy of the intelligent household greenhouse system can be given.(6)The intelligent household greenhouse system is comprehensively tested,and the host computer function,the mobile terminal App function and the greenhouse equipment control are tested respectively.Through the above tests,the design requirements of the intelligent household greenhouse system have been reached,and the feasibility of the system has been verified.