Study On Fuzzy Neural Network And Knowledge Modle Based Assistant Decision System In Protected Grape Production

Precise and intelligent management of grape production is an indispensible part of the modern fruit industry which features"high-yield, high quality, efficient, ecological. "To achieve the goal of smart decision support for the protected grape production, this study combed the application of artificial intelligence and other information technologies in the agricultural production areas, explored the framework of knowledge rules under grape root restriction, probed into key links of the production site, web server and the information exchange, established the structure design layout and realization approaches of the assistant decision system, built up a knowledge model based grape root restriction assistant decision system and fuzzy neural network based grape diseases intelligent diagnosis system, realized the online and offline, Web+SP co-assisted decision support service model.The main results are as follows:(1) Took the smart phone as a terminal device, Windows Mobile operating system, embedded software and hardware environment have been used in grape production and management, an expert decision-making system have been developed in portable service terminal, which is capable of knowledge querying, decision supporting, and providing expert level rational plan to grape producers, the system not only can be used offline alone , but also can be more powerful and comprehensive through online accessibility to the background sever in real time; (2) Accoding to the theories and methods of software development,the key technologies in developing smartphone software based on Windows Mobile, and information exchange with greenhouse sensors, web server have been studied. Using bluetooth technology to collect data, and exchange information with background server through 3G, GPRS, the"Web+SP" collaborative model of decision supporting services have been profoundly represented;(3) In combining the advantages of software component technology and network of Web Services, the decision support models coupled with web services, a backup platform for online services , system upgrade and data maintenance of the system have been constructed, which contains real-time monitoring, experts online support and other additional functions. The platform enriched the management functions of the whole system greatly, enhanced the self-learning ability, intelligence and scalability of the system network, and laid a solid foundation for the establishment of a precise,intelligent online management system in grape production;(4) To realize smart diagnosis of grape diseases, artificial neural network and fuzzy theory have been applied to developed a dynamic, intelligent grape diseases online diagnosis system.The system consists of a five sub-networks parallel topology neural network based on optimal BP algorithm devoleped by C# language. Analysis showed that the system is stable, responsive and smoothly, and reached a relatively high diagnostic accuracy, a continuous optimization through debugging and self-learning can improve the network's performance and accuracy, and provide real-time, reliable disease diagnosis decision support for mobile terminals and registered users.