Three parameter Weibull distribution estimation based on particle swarm optimization

Abstract

Through the simulation data, maximum likelihood estimation (M-L), ant colony algorithm (ACO) and particle swarm optimization (PSO) are used to compare the parameters estimation of three parameter Weibull distribution, the shortage of the maximum likelihood estimation in the three parameter solution process is analyzed, For the defects of ACO algorithm optimization result that it is easy to cause the local optimum (producing “premature” phenomenon), a solution based on the whole area iteration method proposed. By comparing the performance of the three algorithms under the index of fitness, efficiency, correlation, AD test and so on, the conclusion that ant colony algorithm and particle swarm optimization algorithm have better applicability in parameter estimation of three parameter Weibull distribution and particle swarm optimization algorithm is superior to ant colony algorithm and maximum likelihood estimation is drawn.