Research On Reasoning And Fuzzy Bayesian Network For Rotary Kiln Fault Diagnosis Method

Because the cement rotary kiln is one of the main equipment in the new dry process cement production line, its running status is normal or not directly related to the cement quality and production, environmental pollution and the consumption of resources.Firstly, based on the cement rotary kiln calcine system as the research object, analyzing and comparing the advantages and disadvantages of all kinds of fault diagnosis technology, combining the advantages of bayesian network in fault diagnosis, each link of the system shows the uncertainty and polymorphism of problem, based on bayesian network for polymorphic calcining cement rotary kiln system for fault diagnosis analysis. With the traditional bayesian network calcining cement rotary kiln system fault diagnosis model is established, and the network is calculated parameter learning and reasoning.Secondly, parts due to the system failure probability under different states are often difficult to obtain, presents the fuzziness, For this kind of situation, this article combines the fuzzy theory and bayesian network is put forward, using fuzzy bayesian networks, the combination of fuzzy bayesian networks for polymorphic calcining cement rotary kiln system for fault diagnosis analysis.Afterwards, because the Bayesian networks for application is the first step in Bayesian network inference, however both in precise is precise inference algorithm of Bayesian network is one of the most commonly used, but because it is not in the process of transformation of the structure of the uniqueness, therefore, to find the optimal joint tree is also a NP hard problem. For this kind of situation, this paper presents a joint tree algorithm based on minimum short edge search algorithm, and to construct a new algorithm, and its application to the mass of a tree structure, thus to solve the Bayesian network node delete order problem in the process of triangulation, then find the optimal joint tree structure. By comparing with the old algorithm, the new algorithm shows better performance, in the interest of time, so I did not apply new improves reasoning algorithm to cement rotary kiln.Finally, I pointed out the elimination order and the similarities and differences between the traveling salesman problem, just to verify the feasibility in theory.