Research On Firefly Algorithm Based On Chaotic Mapping

As a relatively novel heuristic group bionic intelligent Algorithm,the Firefly Algorithm(FA)has simple concept,clear process,few parameters and is easy to implement.At the same time,it has faster convergence speed and better optimization accuracy.It is one of the effective algorithms in optimization Algorithm.Like other intelligent algorithms,the firefly algorithm also has some deficiencies.For example,the FA algorithm depends on the distribution of the initial solution,and it converges slowly at the later stage of optimization,and is prone to oscillations and falls into local optimum.Chaos optimization algorithm is a stochastic optimization algorithm based on the ergodicity and randomness of chaotic motion.In recent years,with the rapid development of computer science and technology and the rise of heuristic algorithm,many scholars embedded chaos optimization method into heuristic algorithm to optimize the algorithm's optimization performance.Based on analyzing and summarizing the mechanism of chaos mapping and firefly algorithm,this paper proposes an adaptive firefly optimization algorithm based on chaos mapping and applies it to the prediction of stock prices in real financial problems.The main research results are as follows:1.Aiming at the shortcomings of the firefly algorithm,the applications of chaos optimization ideas to firefly algorithm,the first use of chaotic map initialization of population,then the optimal solution in the population in the late iterations design local search operator,at the same time in the whole process of iterative adaptive adjustment of step length factor,put forward a kind of adaptive firefly optimization algorithm based on the Tent chaotic maps(SATC-FA).The test results of the 8 benchmark functions show that the SATC-FA algorithm is improved in terms of convergence speed and search accuracy,and the SATC-FA algorithm is also superior to the cuckoo optimization algorithm based on chaos theory.2.The improved algorithm is applied to the prediction of stock price in financial market.BP neural network is very powerful in forecasting stock price and other non-linear systems,but it also has inherent defects.In order to improve the prediction accuracy,the SATC-FA algorithm proposed in this paper is combined with BP to build a BP neural network(SATC-FA-BP)model based on adaptive firefly algorithm,and then the stock price prediction is made.Through the selected data of 4 stocks,the simulation comparison analysis of the prediction accuracy of BP,FA-BP and SATC-FA-BP network models was conducted.The results show that the SATC-FA-BP network model is superior to the other two models and can predict stock price effectively...