Research On The Decision-Making Of Space Situation Assessment Based On Expert System

In the past twenty years,the development of artificial intelligence theory and technology have provided a great impetus for the development of situational awareness technology.as a battlefield support decision-making,Situation estimation technology is more and more widely used in artificial intelligence and other fields.In the field of situation estimation,scientists have gradually developed a well-developed theoretical system,but due to difficulties such as knowledge representation and reasoning,they still can not be solved.At present,the theoretical framework of the situation assessment system is not unified,which makes it encounter many constraints and obstacles in the development process.With the continuous development and improvement of Bayesian network technology in the fields of knowledge representation and inference algorithms,the method of situation estimation based on Bayesian network technology has drawn more and more attention.In this thesis,the space battlefield is the background,the space satellite is the research object of the situation estimation system,and do research on how to quickly and effectively determine the enemy's battle intention in the battlefield.The space satellite attitude estimation system based on Bayesian network was established based on the situation conclusion obtained by reasoning.In order to respond quickly to the enemy's situation.Use rules-based expert system,and establish expert system architecture and reasoning mechanism to make the appropriate decisions on the situation.In order to grab the first chance in a rapidly changing battlefield and win the war.The main contents of this thesis are as follows:The function model of situation assessment is studied,and the level function of situation assessment is analyzed.In depth learning and analysis of situational awareness,situational awareness and situational prediction,and then achieve a more thorough understanding of the function of situation assessment and the application of the estimation process.The basic theory of Bayesian network and Bayesian network are introduced,and the directed edges and conditional probability tables of Bayesian network in space battlefield are discussed.By analyzing the characteristics of Bayesian network model of space battlefield and Bottom-up reasoning to complete the process of situation estimation.For subjective Bayesian methods,expert knowledge is used to set the conditional probability of each node.As well as the Bayesian reasoning is applied to the space battlefield situation estimation.By calculating the posterior probability of each node to indicate the change of its confidence,so as to determine whether this situation is possible.This article briefly introduces the Matlab software BNT toolbox,and use the BNT toolbox to complete the calculation of the posterior probability of each node in the network.Simplify the calculation process,saving manpower and material resources.In this thesis,the model used in the space battlefield situation estimation is instantiated,and a reasonable Bayesian network structure and conditional probability table are constructed.Taking some nodes as an example,Bayesian network reasoning is carried out,and the reasoning results are analyzed and compared.Verifying that the method we proposed is feasible and effective.When the Bayesian network is used for situation estimation,the system parameters can't be adjusted in time and the future moment can't be predicted.In order to solve this problem,the application of dynamic Bayesian network in battlefield situation assessment has been studied.By introducing the time factor,a network model has been established,and the process of probability parameter and reasoning has been analyzed.Through the inferential simulation of dynamic Bayesian model,the experimental results of static Bayesian network are reasoned by continuous reasoning.The static Bayesian results are optimized to make the inference results accurate and practical.The dynamic Bayesian network can effectively utilize the time information in the reconnaissance data and deal with the various factors affecting the analysis and decision in real time dynamically,which is of great reference value to the commander's combat decision.This thesis analyzes the construction of knowledge base of decision system from knowledge representation and inference engine design respectively.The result of Bayesian network reasoning is input into the rule-based expert system.With the improvement of the knowledge base and the design of the inference engine,the decision-making result can be promptly obtained when the situation is known,so that a quick response can be achieved.Through the analysis and research on the expert system of space battlefield situation expert decision-making system,the decision-making system designed and constructed is feasible and effective.