Genetic algorithms have demonstrated considerable success in providing good solutions to many NP-Hard optimization problems. For such problems, exact algorithms that always find the optimal solution are only useful for small toy problems, so heuristic algorithms such as the genetic algorithm must be used in practice. In this paper, we apply the genetic algorithm to the the NP-Hard problem of Multiple Fault Diagnosis (MFD). We compare a pure genetic algorithm with several variants that include local improvement operators. These operators which are often domain-specific are used to accelerate the genetic algorithm in converging on optimal solutions. Our empirical results indicate that by using the appropriate local improvement operator, the genetic algorithm is able to find an optimal solution in all but a tiny fraction of the cases, and at a speed orders of magnitude faster than exact algorithms.

Engineered Conditioning (EC) is a Genetic Algorithm operator that works together with the typical genetic algorithm operators: mate selection, crossover, and mutation, in order to improve convergence toward an optimal multiple fault diagnosis. When incorporated within a typical genetic algorithm, the resulting hybrid scheme produces improved reliability by exploiting the global nature of the genetic algorithm as well as "local" improvement capabilities of the Engineered Conditioning operator.We show the significance of the Engineered Conditioning operator during Multiple Fault Diagnosis (i.e., finding the collection of simultaneously occurring disorders that best explains the observed symptoms or disorder manifestations). Within the Multiple Fault Diagnosis domain, we show the improvement of diagnostic reliability when using the engineered conditioning operator with the genetic algorithm compared to results from the genetic algorithm without the new operator. Reliability is based on the number of diagnostic trials for which the two versions of the genetic algorithm find the optimal diagnosis. For comparison purposes, optimal diagnoses have been computed using a search method that is guaranteed to find the optimal solution.

No abstract