Research On The Optimization And Generation Techniques For Diagnostic Strategy Based On Multi-signal Model

As an innovative diagnostic conception and mode, Integrated Diagnostic Technology has important significance in enhancing the efficiency of maintenance and reducing the life cycle cost of equipment. It is a more efficient way to meet the demand of the reliability, maintainability and supportability of equipment in complex application environment. As a key technology of Integrated Diagnosis, the research objective of the diagnostic strategy optimization is to improve the condition monitor and fault isolation ability of a complex system, in which way, the skill requirements of the user will be reduced. Unfortunately, most of the currently existing diagnostic strategies are static, which have less flexibility and poor ability in multi-fault diagnosis and inaccuracy diagnosis. Referring to the present situation and existing problems, the optimization techniques for diagnostic strategy was studied in this thesis, and a new diagnostic strategy was proposed. A Diagnostic Strategy Optimization System was constructed based on this new methodology at last.The main contents and conclusions are as follows:(1)The research status of integrated diagnosis and diagnostic strategy was discussed on the basis of survey and research enough. It was pointed out that the main problems of the currently existing diagnostic strategies were poor multi-fault diagnosis ability, less flexibility, poor inaccuracy diagnosis ability and unreasonable assumption of single-fault, which indicated the research direction of diagnostic strategy optimization.(2)An innovative diagnostic strategy was proposed based on the multi-signal flow graph model. The problems of how to construct a Bayesian Network automatically, sequential fault diagnosis, multi-fault diagnosis, the adaptation to the environment and coupling fault diagnosis were all discussed from theories and methods.(3)Diagnostic strategy optimization software named Diagnostic Strategy Optimization System was constructed based on the theoretical study in this thesis.(4)Taking the TBD234 marine diesel as an object, its failure modes and test methods were summarized and analyzed on the basis of full investigation in the producing and using units. A multi-signal model was set up based on the idea of hierarchical modeling. The diesel engine's testability such as fault detection rate and fault isolation rate was improved by means of testability analysis and optimization tool.(5)The functions of the Diagnostic Strategy Optimization System were verified based on the multi-signal model of the diesel engine. The result of practical application shows that the Diagnostic System achieves prospective functions such as the construction of a Bayesian Network automatically, do sequential fault diagnosis and multi-fault diagnosis, etc, which indicates that the diagnostic strategy proposed in this thesis is correct and scientific.