Multi-objective fruit fly optimization algorithm for test point selection

This paper presents a multi-objective fruit fly optimization algorithm (MOFOA) to solve test point selection problem. In the MOFOA, a binary string is used to represent a location of fruit fly, the number of 1s and the different position of 1s in the binary string represent the distance and direction of FOA respectively. The iteration search of MOFOA is based on smell search and vision search. Both the number of isolated faults and selected test points compose a multidimensional fitness function to enhance the global exploration ability. More than one possible optimal solution is searched by the approach. The accuracy and the efficiency of the proposed algorithm are proven by experiments. The results show that the MOFOA is more accurate and more efficient than other algorithms.