Natural Computing Method And Application Research Based On Fusion Strategy

With the development of the times,information technology has developed rapidly,and the development of science and technology has led human beings to move forward.Natural computing has been widely used in the field of scientific and technological research,especially in the fields of optimization problem-solving,pattern recognition,and machine learning.Natural computing has powerful global search and complex information processing capabilities and has high accuracy and robustness in practical applications.However,the traditional natural computing method has the defects of slow convergence speed in the later stage of the algorithm,reduced population diversity,and easily falling into local optimum.To address these issues,a fusion strategy-based natural computation method is proposed.The research focus of this paper is:(1)A natural computing method that combines quadratic reverse and Cauchy perturbation is proposed.This method divides the algorithm into two parts through probability conversion,one is the quadratic reverse strategy,and the other is the original natural computing Strategy.The quadratic inversion strategy has the characteristics of two inversions,and the individual can be obtained through the second inversion to obtain a better individual,which improves the convergence speed and convergence accuracy of the algorithm.After the position is updated,the optimal value is disturbed by the strategy of Cauchy disturbance to improve the ability of the algorithm to jump out of the local optimum.The effectiveness of the strategy is tested by eighteen benchmark functions.The experimental results show that this strategy has a significant improvement in the convergence speed and optimization ability.(2)A natural computing method that integrates multi-population evolutionary strategies is proposed,which groups populations into three sub-populations by fitness.These three sub-populations are divided into the development sub-population,exploration sub-population,and moderate sub-population.According to the parity of the number of iterations,the algorithm is divided into two stages: population movement and population evolution.In the stage of population movement,the development sub-population has better fitness and is suitable for local search.The exploration subpopulation has poor fitness and is suitable for global search.The moderate population with medium fitness randomly chooses local search or global search.During the population evolution stage,individuals in the exploitative population evolve their positions by performing a deep local search around the current optimal solution.Exploring individuals in the subpopulation expands the search range and enhances the global exploration ability by increasing the absolute value of the position vector.The medium subpopulation improves the diversity of the population through the strategy of reverse learning.Apply it to the butterfly optimization algorithm,and compare and analyze the effectiveness of the strategy through experiments.Experimental results show that this strategy improves the convergence speed and optimization ability of the algorithm.The above two natural calculation methods have a certain degree of universality but are applicable to most intelligent algorithms.And the two strategies are applied to five kinds of engineering optimization problems and robot path planning.This is to test the practicality of the strategy.Through multiple benchmark functions and application testing,experiments have shown that both natural calculation methods have significantly improved convergence speed and optimization ability.