Research On Data Analysis And Processing Method For Intelligent Led Plant Growth Cabinet Based On Internet Of Things

LED plant growth cabinet(PGC) is the facility with high precise environmental control and efficient agricultural system to realize crops annual continuous production. As easy-operational production mode to make the crops and vegetables not or rarely limited by the natural conditions, the PGC uses computer to control the environmental elements, such as temperature, humidity, light, concentrations of CO2, nutrition, automatically. The PGC adopts LED as the light resource, which has the advantages of less heat, monochrome with combination, energy conservation, environmental protection, long service life. The LED light resource is suitable for intensive production mode of hermetic PGC. It can narrow the cultivation layer space substantially, and greatly improve the utilization rate of light and space. The application of Internet of things in PGC has become the researchful focus in recent years. Combining Internet of things(IoT) technology and PGC, the Intelligent IoT and PGC system has been built by using RFID, sensors to acquire the information in real time. Then, the system proceeds with real-time communication and sharing of information in a reliable way through various communication networks. Finally, through analyzing and processing acquisitive data information, the plant growth state model has been established to predict plant growth trend and realize the intelligent decision-making and remote monitoring for PGC. Therefore it intelligently adjusts necessary growth environment of plant to make the plant always in the best state, finally makes the PGC product green pollution-free vegetables and crops, intelligently, efficiently, and in high yield. In order to achieve the high accuracy data analysis and processing method for intelligent PGC of Internet of Things, this research focuses on the following aspects:1. The intelligent LED PGC based on Internet of Things has been designed and established, including the hardware control side, Android control software and Internet of Things server, in order to achieve the acquisition, real time transmission and storage of the growing environmental information and outline images in the crops’ whole life cycle, provide research data and the experimental platform for data analysis and processing in the PGC.2. According to plant characteristics information measurement process, image enhancement and denoising method, image threshold segmentation method, and plant image feature extraction methods have been compared. The appropriate methods have been chosen to measure the crop canopy leaf area and plant height of effective feature information, and then use reference to proportionally transform to get the actual growth characteristic data, in order to lay the foundation for the data analysis and processing.3. Plant growth state prediction model, based on the BP neural networks improved by Kalman filtering, has been proposed with the use of environmental elements and growth characteristic information. By using Kalman filtering to improve the BP neural network model, the precision of prediction has been improved, and the validity of the model has been verified. The Plant growth state prediction model based on ARMAX model or ARX model has been proposed to establish the plant canopy leaf area and plant height prediction model. The prediction and the actual measurement of canopy leaf area or plant height are the same trends. The verification methods of models show the validity of the models.