Intelligent Algorithm Applied Research In Expert Systems

An expert system is an important direction of artificial intelligence research , the expert system is a intelligence explanation software system that solving and gives a kind of specific direction of the problem. It simulates a reallity domain experts methods to solve the problems and give reasoning. The reality of experts can solve problems of a particular field, puts forward the optimal solution, thus can promote social development and progress in many aspects of application.Human expert reasoning is a kind of intelligent form of the more mature, Application of the intelligence method in the expert system in process of reasoning, made experts system can solve more complex problem, can think more,this is an important challenge. Reasonable organizations and use more popular intelligence method, made these algorithms involved used in computer in the decision-making process of expert system is a key point in the research of expert system. The traditional expert systems has many difficult aspects such as widespread knowledge expression,the accurate of reasoning steps.This article make sure the ability of the neural network learning and reasoning by traditional knowledge and neural network operation mechanism. The article has given an method named RE-BP by analysi the neural network and the respective features of expert system. When proposed the algorithm, give a efficient and easy algorithm to make the data to be ordered. By experiment to improve the efficient of this algorithm.The following is the main work:1 Analysis and experiment of existing artificial neural network and the combination of expert knowledge and its application.2 combines expert system and the characteristics of neural network algorithm proposed RE - BP algorithm.3 in the proposed algorithm from 2 to large amounts of data to a sorting, in research existing data sorting method is proposed on the basis of new prority method - end positioning sorting method. 4 By the analysis based on the genetic algorithm, let its application in the neural network, the weights of neural network as the adjustment, and the improvement and application of the algorithm proposed in experiments proves its feasibility.