Harris Hawks Optimization Algorithm Based On Hybrid Strategy And Its Application

Intelligent optimization algorithms are a class of algorithms that can automatically search for the best solution.They usually use heuristic strategies to optimize the solution of complex problems.These algorithms have been widely used in many fields to solve various optimization problems.Harris hawks algorithm(HHO)is a swarm intelligence algorithm designed by simulating Harris hawks group cooperative predation.Although it has some advantages,such as fewer parameters and simple principles,it also has many problems,such as low convergence accuracy,prematurity,poor balance between search and mining.This paper proposes two improved algorithms for HHO.It is used to optimize the optimal parameter combination of BP neural network and the resulting model is applied to predict the net power output.Specific research contents are as follows:(1)A multi-tactic Harris hawks optimization algorithm(MTHHO)was proposed.A new nonlinear convergence factor was designed to replace the original linear convergence factor to balance the search and mining of the algorithm.In view of the poor effectiveness of the late mining strategy,t distribution was used to disturb the optimal individual term of the original formula,so as to improve the effectiveness of the strategy.The variable spiral strategy was used to replace the original levy flight strategy to improve the convergence accuracy of the algorithm.The validity of MTHHO algorithm was verified by 15 benchmark functions in CEC2005.(2)Harris hawks optimization algorithm based on transformation of parabola(TPHHO)was proposed.The Logistic-Tent chaotic mapping was used to initialize the population and improve the diversity of the population.Inspired by genetic algorithm,the optimal individual of the population was randomly modified to prevent the population from being trapped in a state of stagnation.In order to solve the problem of poor convergence accuracy,the parabola bounding method in TSO algorithm was introduced to correct the algorithm flow.The effectiveness of TPHHO algorithm was verified by using 17 benchmark functions in CEC2005.(3)BPNN model based on MTHHO optimization was proposed.The improved MTHHO algorithm was used to optimize the combination of weight and bias parameters of BP neural network,and the final result model was obtained and used to predict the net power output of a power plant.The experimental results verify the effectiveness and practicability of the MTHHO algorithm in application.