Abstract:

In order to obtain the optimal test set for a real time monitoring and fault diagnosis system based on dependency model with high efficiency and low cost, a modified particle swarm optimization (PSO) algorithm of multi objects is proposed. Taking the reasons why a test should be left in the optimal test set into consideration, two specified updating methods are introduced to find the set with optimal multi criteria, such as number of fault detected and isolated, number of test and cost, object functions, and the updating method for optimal particles is also modified specially. The simulation results show that the proposed algorithm has better computation efficiency and precision, which can give effective direction to the test selection of the realtime monitoring and fault diagnosis system.