Differential evolution algorithm based on wavelet basis function and optimal mutation strategy and application

Abstract

The invention relates to the technical field of differential evolution algorithms and particularly relates to a differential evolution algorithm based on a wavelet basis function and an optimal mutation strategy and application. In the differential evolution algorithm based on the wavelet basis function and the optimal mutation strategy, the wavelet basis function is adopted to control a parameterF and normal distribution to control CR, so that the diversity of solutions is guaranteed. The algorithm convergence is accelerated, and algorithm performance is improved. Upon the selection of mutation policies, advantages are complemented based on five variation strategies. An optimal variation strategy method is provided. The method is used as a variation strategy of a differential evolution algorithm, and is used for improving the local search capability of the algorithm and ensuring the global search characteristic of the algorithm. Meanwhile, the method can efficiently complete effective solution of a test function and an airport gate position allocation problem, and experimental results also prove that the proposed method has competitiveness compared with a comparison method.

Description

technical field [0001] The invention relates to the technical field of differential evolution algorithms, in particular to a differential evolution algorithm based on wavelet basis functions and an optimal mutation strategy and its application. Background technique [0002] Differential evolution (DE) algorithm is a random search algorithm based on population evolution proposed by Storn and Price. The DE algorithm simulates the process of population evolution in nature through mutation, recombination and selection operations. DE adopts real number coding, which reduces the complexity of genetic operations. Its principle is simple, easy to understand and implement, and it has the characteristics of strong robustness, few control parameters and strong search ability. It is a simple and efficient evolutionary algorithm. In recent years, although good results have been achieved in many fields such as engineering design, operations research, and biomedicine, many shortcomings ha...

AI Technical Summary

Problems solved by technology

Although the CoDE algorithm adopts a multi-operator mutation strategy, it lacks the necessary learning process and cannot perform adaptive selection of related strategies according to the evolution process.

The SaDE algorithm can timely adjust the mutation strategy and parameter selection probability based on previous experience, but the learning process of the algorithm is complicated, and the parameter setting in the actual application process is difficult

[0004] The above-mentioned improved algorithm has achieved certain effects by improving the mutation, crossover operation or control parameters, but it is difficult to set the parameters in the actual application process, and it lacks the necessary learning process, and it cannot automatically implement related strategies according to the evolution process. Adaptive selection, resulting in the algorithm easily falling into local optimum, poor optimization ability, and low convergence accuracy

Images