Reasoning And Algorithm Uncertainty Interval Number Causality Diagram

Real-life phenomenon and objective things, the prevalence of diversity,imprecision and uncertainty caused by people in various fields of knowledge and understanding of the information there is uncertainty, and one of the areas of artificial intelligence is the use of the machine simulate the human brain to such uncertainty knowledge representation and reasoning. Uncertain knowledge representation method comprising: belief network, fuzzy logic, set theory, the theory of evidence and dynamic causal diagrams. These methods in dealing with uncertain information, there will be a corresponding advantages and disadvantages, there is some commonality between different methods. This paper focuses on the causal diagram of Knowledge Representation and Reasoning, build complex models, analytical reasoning difficulties and basic event for the application of interval numbers, combined with rough set reduction as well as evidence of the theory of complex graphics in obtaining the advantages of the data areas to discuss and study. The main contents are as follows:(1)Introduces the causality diagram of the knowledge expression and reasoning,with the nodes represent events or variables, cause and effect diagram is a diagram with a ring-type structure, which means that the parent and child tree nodes will derive mutual process occurs. Causal probability map as a knowledge representation method for accurate calculation of the probability of an event derived value, but the actual case, but since the initial data errors, deletions, and other reasons can not get exact value, the exact value of the proposed extension of this situation this article for the interval number, interval number may indicate the use of the range of features exact value will be converted into a number of intervals, can effectively deal with ambiguity and uncertainty of events, but also reduce the difficulty of obtaining accurate values. Wherein the upper and lower bounds of the interval number, using non-probability methods to obtain evidence theory, probability interval for the upper and lower bounds based on the calculated likelihood function and reliability function,better reliability, and then analyzed according to the theoretical calculation interval demand events interval probability, simplify causal diagram parsing algorithm difficult.(2) In the causal diagram of the process, according to researchers have each elementary event in the entire causality diagram in the case of basic event probabilities derived importance, structure importance and critical degree. Based on the numerical size and range of basic events, obtained under different conditions probability interval number in descending order based on the basic events in the engineering application can be based on the size of the range of probability values,obtained under certain conditions easiest failure the basic events that are causing system failure cause event, and to strengthen safeguards.(3)Once the system is too complicated, in practice it is difficult to identify the source of the problem, this paper combined with rough set, rough set introduces the basic theory and use of knowledge Rough set of causal diagrams simplified, easy to find fault in Engineering source, and fault diagnosis based on the path set causality diagram and minimal cut sets fault diagnosis method also conducive to maintenance of the investigation to shorten the time in the actual project.In this article, the expression system described method for uncertain knowledge,reasoning causality diagram model, research, introduced the rough set theory and DS advantage in dealing with uncertain knowledge of, and take advantage of the knowledge of the uncertain evidence theory, interval analysis, rough set theory and causality diagram combined reach to avoid get an accurate value of the difficulties basic events, as well as the degree of difficulty logical complexity, causal diagram troubleshooting ago some reduction, will help shorten the troubleshooting time, in line with the actual situation, diagnosis quickly, the better.