Design Of Expert System For Western Superior Crops

In the 21st century, the information technology has been widely applied in various fields of human production and life, and agricultural production in China is in the crucial period of the transition from traditional agriculture to modern agriculture. To realize agricultural modernization and achieve a qualitative leap in productivity, information technology as a means of management and dissemination must be developed with great effort. As an important content of the information technology, intelligent expert system has become a hot spot in the research field of precision agriculture. It holds wide application prospect and great development potential, and has important significance to improving the quality of laborers and realizing agricultural modernization.This paper mainly studies artificial intelligence, knowledge engineering and the application and development of expert system in the agricultural informationization field, thoroughly analyzes several kinds of typical expert systems, and proposes a design scheme of the expert system according to the issue, namely the western superior crops' production management. On the basis of the research and analysis of the classical expert system's design methods, the development direction of the expert system is with the characteristic of multiple knowledge libraries and multiagent. The expert system discussed in this paper mainly includes three modules which are greenhouse control, plant diseases diagnosis and intelligent adjustment.In order to achieve precision management of the greenhouse, the growth model library and rule library based on generating rules are bulid, and the greenhouse management subsystem is developed. According to the crops' growth model and parameter control indices, the developed system can predict the crops' growing conditions, and make accurate adjustment to the greenhouse's environment parameters.For the intelligent diagnosis of plant diseases, the knowledge representation method based on fuzzy theory and the credibility model based on inexact reasoning are applied to the construction of the plant diseases diagnosis subsystem, which is a better solution to the complex relationship and unaccurate descriptiom of the plant diseases. The developed system can make nearly the same judgment as an expert on the information of plant diseases, and give treatment measures for the diseases.In order to similate the experts' production experience more precisely, and achieve intelligent adjustment to the greenhouse's parameters, a neural network algorithm is introduced to build an expert system with self-learning and reasoning mechanism, so that the greenhouse's parameter optimization subsystem is designed. The developed system can independently learn the laws concealed in huge data obtained from years of experiments by agricultural experts, and dynamically adjust unreasonable parameters to make crops grow under the optimal conditions.