Application Of Improved Salp Swarm Algorithm In Engineering Design And Path Planning

With the development of human society and people’s cognition,various optimization problems become more and more complex,and the traditional solution methods are difficult to produce results when solving complex optimization problems.In recent years,the meta-heuristic intelligent optimization algorithm with simple mechanism and good performance has attracted the attention of many scholars.In 2017,Australian scholar S.Mirjalili put forward a new kind of heuristic intelligent Algorithm—Salp Swarm Algorithm(SSA).Due to its simple structure,fast convergence speed and few control parameters,many scholars began to study it and applied it to passive time difference positioning,image segmentation,hybrid power system,and many other fields.However,like other intelligent algorithms,salp swarm algorithm also has the disadvantage of easily falling into local extremum and the solution result is not very stable.To solve this problem,this paper makes a deep analysis of the algorithm mechanism,and puts forward a new improvement idea to improve the optimization performance of the algorithm,and the improved algorithm is applied to engineering optimization design and robot path planning.The research work of this paper mainly includes:(1)An adaptive dynamic role salp swarm algorithm(ERDSSA)with effective scaling and random crossover strategy is proposed to solve the problem of optimal engineering design.Firstly,the pareto distribution and chaos mapping are introduced into the leader position update formula in the global search phase,so that the leaders of salps can carry out global search more effectively and play their rolebetter.At the same time,leader—follower adaptive adjustment strategy is introduced in the selection of global search and local search,which makes the algorithm focus more on global breadth exploration in the early stage and dig more deeply near the optimal value in the later stage,so as to improve the convergence accuracy of the algorithm.Then in the local search stage,the random crossover strategy is introduced to expand the randomness of the position update of the follower,so as to increase the diversity of the population.Finally,the pseudo-code of the algorithm flow is given,the time complexity is proved by theoretical analysis,and the algorithm in this paper is used to solve typical engineering design problems of different difficulty such as welding beam,pressure vessel and three-bar truss.By comparing and analyzing the optimization design results of five test comparison algorithms and the solution results of more than ten other algorithms in the literature,the experimental results show that the improved algorithm is effective and superior in solving engineering optimization design problems.(2)To better solve the problem of robot path planning,a parasitic salp swarm algorithm(DPSSA)with differential evolution strategy is proposed.Firstly,the position of salps of the previous generation was added into the updating formula of the leader position of the algorithm,which enhanced the adequacy of global search and effectively prevented the algorithm from falling into local extremum.Second,inertia weight was added into the leader position update formula,which reasonably adjusted the balance between breadth search and depth mining of salp leaders in different iteration periods,and improved the solution accuracy of the algorithm.Thirdly,the parasitic and host double populations with different evolutionary mechanisms and their parasitic behaviors as well as the idea of survival of the fittest are introduced to increase the diversity of the population and improve the ability of the algorithm to jump out of local extremum.Finally,the pseudo-code of the algorithm flow is given,andthe time complexity of the improved algorithm is proved to be the same as that of the basic algorithm by theoretical analysis,and the simulation experiment is carried out on 15 different feature test functions through seven comparison algorithms,and the test results show that the optimization accuracy and stability of the improved algorithm are significantly improved.(3)By combining the parasitic salp swarm algorithm and the cubic hermite interpolation method,the individual coding is defined,and the fitness function is constructed to solve the robot path planning problem.Under three different complexity of the scene,through seven different contrast algorithm,simulation experiment and test the path planning problem,the experimental results show that the improved algorithm path length of the optimal value,the worst value,average value and variance are better than the other six comparison algorithm,and shows the superiority of the improved algorithm for solving path planning problem and stability.