Improvement On Grasshopper Optimization Algorithm And Its Applications In Chemical Engineering Process Modelling

The grasshopper optimization algorithm(GOA)is a kind of novel biomimetic algorithm inspired by the different behaviors in the grasshopper lifecycle.As a member of swarm intelligent optimization algorithms,the advantages of GOA are simple structure,easy operation and integration with the other algorithms.The disadvantages of the basic grasshopper optimization algorithm are prematurity,slow convergence rate and poor search accuracy.Based on the previous research work,this thesis studies some improvements on GOA and several improved grasshopper optimization algorithms are presented.The proposed algorithms are used for solving the problems of chemical engineering process modelling.The main contents of the thesis are as follows:(1)The improved grasshopper optimization algorithm with parameter adaptation(IGOA)is presented.In this algorithm,the cosine adaptive strategy is used for adjusting the control parameter.The coevolution strategy is used for declining the dependence on initialization of the population.Gaussian mutation is added to improve the diversity of the population.The performance of the algorithm is compared with the other swarm optimization algorithms with some benchmark functions and the results show the effectiveness of IGOA.(2)The enhanced grasshopper optimization algorithm with feedback strategy(EGOA)is proposed.In this algorithm,cosine adaptive strategy is kept.The feedback strategy is used for adjusting the control parameter according to the evolution rate.The elite individuals of grey wolf algorithm and operation of evolutionary population dynamics are used for improving the search ability of the algorithm.Results of typical benchmark functions show the effectiveness of EGOA.(3)The hybrid grasshopper optimization algorithm(HGOA)by combining IGOA and EGOA is proposed.The effectiveness of HGOA is verified by the experiment results of some benchmark functions.HGOA is used for solving the nonparametric modelling problems of delayed coking process and the reactor-regenerator system of fluid catalytic cracking unit.The simulation results show the effectiveness of HGOA.