Enhanced convergence of Bat Algorithm based on dimensional and inertia weight factor

Heuristic optimisation method typically hinges on the efficiency in exploitation and global diverse exploration. Previous research has shown that Bat Algorithm could provide a good exploration and exploitation of a solution. However, Bat Algorithm can be get trapped in a local minimum in some multi-dimensional functions. Thus, the phenomenon of slow convergence rate and low accuracy still exits. This paper aims to modify the exploitation of Bat Algorithm in optimising the solution by modifying dimensional size and providing inertia weight. Benchmark test function is then performed for the basic Bat Algorithm and the modified Bat Algorithm (MBA) for comparison. The result is analysed according to the number of iteration needed for a convergence toward the objective. From simulations, it is found that the modified dimension and additional inertia weight factor of Bat Algorithm proves to be more effective than the basic Bat Algorithm in terms of searching for a solution while improving quality of results in all cases or significantly improving convergence speed.