Improvement Of Grey Wolf Optimizer And Its Application In Operational Amplifier Design

Grey Wolf Optimizer(GWO)is an emerging swarm intelligent optimization method which imitates the natural Grey Wolf.It is inspired by the leadership level and hunting mechanism of Wolf pack.GWO has the advantages of simple structure,clear concept,few parameters setting and small amount of calculation.However,the research and application of GWO is still in the initial stage,and there are still some shortcomings,such as slow convergence speed,weak local search ability,insufficient utilization of population information,easy convergence to local optimal,and low convergence accuracy when facing complex problems.In view of this,this paper analyzes and improves the standard gray Wolf optimization algorithm,proposes two new gray Wolf optimization algorithms with better performance,and applies the improved gray Wolf optimization algorithm to the design field of the standard two-stage operational amplifier.The main research work of this paper is as follows:Grey Wolf algorithm is easy to fall into local optimum at the later optimization stage,and the probability of falling into local optimum is higher when solving highdimensional functions due to its higher complexity.In order to solve the above problems,this paper proposes a hybrid grey Wolf optimization algorithm(DGWO)based on drunken walk and reverse learning.In the iterative process,the dominant Wolf and the worst Wolf in each generation of the population were reverse learned and compared,and the dominant Wolf was partially retained,while the Wolf would lead the drunken walk.Ten standard test functions(100D,500 D and1000D)and10D CEC2013 test functions are used to verify the performance of the proposed algorithm.The simulation results show that the hybrid gray Wolf algorithm has advantages in accuracy and convergence speed compared with the contrast algorithm.In the gray Wolf algorithm,the leading Wolf plays an absolute leading role,while other wolves are completely dominated,which leads to the underutilization of the population information in the Wolf group.If the algorithm falls into local optimum for a long time,the Wolf group will concentrate near the leading Wolf,and the leading wolves will be close to each other,and the population will completely lose diversity.In order to solve these problems,an improved grey Wolf algorithm,called stochastic bestworst-average grey Wolf algorithm(RBWM?GWO),is proposed in this paper.The main difference between this algorithm and the classic gray Wolf algorithm lies in that the leading Wolf introduces the random walk process,which strengthens the exploration and development ability,adds the three laws in the late iteration,initializes part of the Wolf group with certain probability in the late iteration,and avoids completely falling into the local optimum in the late iteration.The general performance of RBWM?GWO was evaluated using different dimensions of CEC 2017 baseline function(10D,30 D and 50D),and the results were compared with the five algorithms.The results show that the RBWM?GWO algorithm is superior to the other algorithms in accuracy,robustness and convergence speed.In order to simplify the parameter design of analog integrated circuit,this paper takes two operational amplifiers as examples,and abstract the mathematical optimization model with the goal of maximizing the open-loop low-frequency gain.Two improved gray Wolf algorithms and five other swarm intelligence algorithms are applied to the practical engineering optimization problem,and the effectiveness of the proposed algorithm is verified by comparing the simulation results.