An improved Ant Colony Optimization for continuous domains
An improved Ant Colony Optimization for continuous domains(ACOR)was proposed to overcome the shortcomings of ACOR,such as falling into local optimization and poor convergence accuracy.Firstly,the diversity of the population was enhanced by improving the way the guidance solution selected by ACOR.Secondly,the inertia weight was used to improve the mean value updating formula,and the parameter ζ,which was a parameter in ACOR algorithm,was linearly decreased.Finally,in order to help the algorithm jump out of the local optimum as soon as possible when it was stuck,a differential mutation strategy was used to monitor the algorithm's search,new solutions were generated when it was stuck,and better solutions were filtered out to jump out of the local optimum.The experiments were designed in two dimensional environments,including 16 variable dimensional functions and 4 fixed dimensional functions,and the optimization results of the improved algorithm and other algorithms were shown.The results indicated that the above defects in the search process were effectively alleviated by the improved algorithm.
Ant Colony Optimization for continuous domainsconvergence accuracyselection of guidance solutioninertia weightdifferential mutation
国家科技期刊平台
NSTL
万方数据
