Research On Multi-Objective Cat Swarm Optimization Algorithm Base On Neighborhood Rough Set Model

Lots of engineering and scientific problems can be summarized as Multi-Objective Optimization Problems(MOP),The essential difference between Multi-Objective Optimization and Single-Objective Optimization is that the solution of Multi-Objective Optimization is not unique,and there is a set of solutions,called Pareto-optimal set or Nondominated Set.How to solve such problems effectively has always been the focus of attention of academia and industry.Cat Swarm Optimization(CSO)is a swarm intelligence optimization algorithm proposed by Chu et al.in 2006.It originates from the observation of the life habits of cats.Not easy to fall into local optimum,CSO has the characteristics of simple principle.It has been applied in many fields such as function optimization,machine learning,image processing and so on.Existing CSO has the shortcomings of slow convergence speed and low optimization accuracy.For this reason,our research combines the theory and method of neighborhood rough set,and carries out the following two aspects of work:(1)To overcome the shortcomings of slow convergence in CSO tracking mode,an improved cat swarm optimization algorithm based on neighborhood rough set(NRS-CSO)is proposed.In the tracking mode of CSO,combining the position information of cat group at a certain time,two adaptive parameters are calculated by using neighborhood rough set theory,which is used to improve the speed update formula to optimize the convergence speed of the algorithm.The experimental results show that the NRS-CSO algorithm is superior to the standard CSO algorithm and some existing improved algorithms in terms of convergence speed and optimization accuracy.(2)For multi-objective optimization problems,a multi-objective cat swarm optimization algorithm based on neighborhood rough set(NRSMO-CSO)is proposed.The theoretical analysis and experimental results show that the algorithm not only has a fast convergence rate for multi-objective optimization problems,but also has good uniformity and broad distribution of solutions,which provides a new idea for solving multi-objective problems.