Theoretical Research And Application Of Swarm Intelligence Algorithm Based On Supervision Mechanism

In today’s academic research field,the research content is more and more close to practical problems.When such practical problems are optimization problems and have the characteristic of being inefficient,the use of meta-heuristic algorithms to solve such problems has become a hot topic in the field of academic research.Meta-heuristic algorithm is a swarm intelligence optimization algorithm,which is not only based on the aggregation of swarm members,but also shows the independent intelligence of swarm members.The swarm intelligence optimization algorithm is a process of continuously searching for the optimal value in the iterative process,but in the later stage of the algorithm iteration,the diversity of the population decreases,the exploration function becomes poor,and the algorithm is easy to fall into a local optimum.In order to make the swarm intelligence optimization algorithm maintain the diversity of the population in the iterative search process,a supervision mechanism is introduced,according to a certain rule,the algorithm will be replayed.This paper mainly introduces the supervision mechanism in the lion swarm optimization(LSO).Aiming at the problem that LSO is easy to fall into the local optimum,this paper proposes two improved LSO.The first one is based on the LSO,and introduces the sine cosine algorithm and the elite opposition-based learning to improve the LSO and improve the local search ability of the LSO.In addition,the introduction of a supervision mechanism enhances the ability of lions to jump out of local optimization.To verify the performance of the improved LSO,it is applied to the fuzzy C-means(FCM)clustering algorithm.And for a series of shortcomings of the FCM algorithm,such as being sensitive to the initial cluster center and noise data and easy to fall into local minima and so on.The initial cluster center of the FCM algorithm is the optimal solution obtained by the improved LSO,and then the FCM algorithm optimizes the initial cluster center to obtain the global optimal value of the FCM algorithm,thus effectively solving a series of problems existing in the FCM algorithm.In the experimental part,the classical data set is used for verification.The results show that the FCM clustering algorithm introduced with the improved LSO has improved the optimization ability of the algorithm compared with the FCM clustering algorithm introduced with the original LSO and the original FCM clustering algorithm.And the clustering effect is better.The second one is also based on the LSO,and introduces the interactive search algorithm to modify the LSO.An interaction mechanism is introduced in the position update of the lioness,so as to ensure the correctness of the lioness in the search direction and improve the search efficiency of the lioness.In the position update of lion cubs,a tracking and crossover mechanism is introduced to improve the convergence speed of the algorithm.Besides,a supervisory mechanism is introduced to avoid getting stuck in local optima.In order to verify the performance of the modified lion swarm optimization,the modified LSO,LSO and particle swarm algorithm were applied to the robot path planning problem of body welding production line respectively.The result of optimization is more ideal,which proves the effectiveness of the improved algorithm,and the use of the improved LSO can improve the efficiency of simulation work,which has certain guiding significance for the simulation work of industrial robots in automatic production lines.Finally,the paper summarizes the current work progress,and points out the deficiencies and follow-up research directions.