A Kalman filter approach for solving unimodal optimization problems

In this paper, a new population-based metaheuristic optimization algorithm, named Simulated Kalman Filter (SKF) is introduced. This new algorithm is inspired by the estimation capability of the Kalman Filter. In principle, state estimation problem is regarded as an optimization problem, and each agent in SKF acts as a Kalman Filter. Every agent in the population finds solution to optimization problem using a standard Kalman Filter framework, which includes a simulated measurement process and a best-so-far solution as a reference. To evaluate the performance of the SKF algorithm in solving unimodal optimization problems, it is applied to unimodal benchmark functions of CEC 2014 for real-parameter single objective optimization problems. Statistical analysis is then carried out to rank SKF results to those obtained by other metaheuristic algorithms. The experimental results show that the proposed SKF algorithm is a promising approach in solving unimodal optimization problems and has a comparable performance to some well-known metaheuristic algorithms