Research On Lagrange Relaxation Algorithms And Its Application In Circle Diagnostic Method

The modern avionics is a high-tech intensive equipment, with the rapid development of technology and its more and nore complex structure, the higher and stricter requirements were put forward for equipment testing and diagnosis. The traditional diagnostic methods have been unable to meet the demand of testing, at the same time, resulting in a large waste of human and money resource. Aiming at these problem, relevant experts do extensive researches and practices and find the only way to solve the testing problem is to strengthen the design for testability.Design for testability, means testing problem must be considered concurrently with functional design of equipments. It improve testing controllable of internal circuit with the optimization design and obtain sufficient diagnostic information with minimum cost, which through the comprehensive application of Built In Test, the automatic testing resources.There are a lot of testability design technology, a key and core technology among these is the testability modeling. To establish testability modeling of equipment, then obtain the state information of equipment and design diagnostic strategy based on testability model. This technology has been studied for more than 30 years and some achievements have been made in foreign countries, but in our country the study of design for testability started late and there are many problems to be solved.This dissertation studied testability modeling for normal systems. First this paper discussed and summarized the theory of testability modeling, then studied on the optimization design technologies for diagnostic strategy in typical single fault situations. Based on above studies, an improvement to AO algorithm has been put forward on search strategy, which promated efficiency of search, guarantee the average test cost no greater than 1+ ε multiples of the former. Finally, use an example to demonstrate the effectiveness of this method.In the situation of multiple faults, the diagnostic problem is transformed into a set covering problem according to a thought of the maximum posteriori probability. At the same time, consideration for available test sequence provided by single-fault diagnostic strategy, Lagrange relaxation and subgradient optimization is used to solve multiple-fault diagnosis problem. Fanally, use an example to demonstrate the effectiveness of the method which can avoid missed diagnosis and misdiagnosis.