Research And Application Of Lion Swarm Optimization And Improved Algorithm

With the continuous expansion of the modern engineering field,many new engineering technology application scenarios are constantly emerging.New application scenarios are developing in the direction of complexity and intelligence.The optimization problem is an unavoidable problem in many engineering applications.The necessity of solution to these problems gives rise to swarm intelligence algorithms.Because of its self-organization,flexibility and robustness,swarm intelligence algorithms have unique advantages over traditional optimization algorithms in solving complex optimization problems.As an emerging group intelligence algorithm in recent years,the lion group algorithm simulates the population composition and behavior patterns of lion groups in nature,so that the agents cooperate with each other to complete the optimization process.The algorithm model simulates the different behavior patterns of the three individuals in the lion group:lion king,lioness and lion cub,and different individuals cooperate with each other to create a way to solve the global optimization problem.Among the three kinds of individuals,the lion king guards the territory and preferentially enjoys food.The lionesses hunt for food through collaboration.The lion cubs are divided into three types:learning to hunt,eating when hungry,and being expelled by the time they reach their new year.The mutual cooperation among the agents existing in the lion group algorithm mechanism is a relatively distinctive feature in the group intelligence algorithm that has been developed.The update strategy of the algorithm plays a vital role in the efficiency of an swarm intelligence algorithm.This paper focuses on improving the update strategy of the algorithm in order to pursue better algorithm performance.Reinforcement learning is essentially solving the Markov decision process about the problem,and finally gets the strategy to maximize the value function.Reinforcement learning is suitable for the cooperation of multi-agents,and its ideas can be widely used in the research of group intelligence algorithms.The financial crisis of listed companies is an important factor that has affected the domestic financial market in recent years.Early warning of financial crisis for listed companies can help decision makers improve business conditions at the beginning of the crisis and prevent future risks.For the company's interests and financial markets The stability of is of great significance.The parameter design involved in the prediction model is a typical optimization problem,and its main solution is to apply group intelligence algorithms including the lion algorithm.This article takes the lion optimization algorithm and improved algorithm,as well as the application of the algorithm as the research goal,reviews the research process of the group intelligence algorithm,and the domestic and foreign experts and scholars'research results on the lion algorithm,and introduces the lion algorithm model and process;The problems and deficiencies of the traditional lion algorithm are discussed,the focus is on the deficiencies in the update strategy,combined with the reinforcement learning idea to improve the lion algorithm,and under a series of standard test functions,the traditional algorithm,improved algorithm and other group intelligence algorithms are compared and analyzed to discuss the design of some parameters in this process.Subsequently,the improved algorithm is used as a tool to implement and analyze the early warning model of financial crisis of listed companies.We compare the performance of the improved algorithm with that of multiple swarm intelligence algorithms under multiple standard test functions,combined with the analysis of the effect in the application scenario,the improved algorithm proposed in this paper has excellent performance,overcomes the defects of the traditional algorithm,and improves the algorithm's stability and flexibility,and can be well applied in practice.