Fault Localization Method Based on Enhanced GA- BP Neural Network

In the process of software development and maintenance, software debugging is the most complicated and expensive part. In recent years, automated software fault localization technology has attracted many scholars’ attention, various approaches have been proposed. In this paper, a technique named EGA-BPN is proposed which can provide suspicious locations for fault localization automatically without requiring any prior information of program structure or semantics. EGA-BPN is a software fault localization method based on enhanced Genetic Algorithm-Back Propagation neural network. Firstly, through processing running traces of the program, coverage information of test cases is converted to the training samples of neural network; secondly, the initial weights and thresholds of the neural network are computed by GA, the training data are substituted in neural network in training orderly, and then use orthogonal experimental design helping to adjust the parameters of the neural network; finally, test matrix is calculated by the neural network to count the suspiciousness of each statement, and the fault is located at the statements with higher suspicious value. Through comparative experiments between the proposed method, GA-BPN and BPN, the experiment results show that the enhanced GA-BP neural network-based fault localization technology has certain validity.