The Realization Of Swarm Intelligence Optimization Algorithm Base On MapReduce Programming Model

The era of big data has come, following a series of emerging technologies, such as Mobile Internet, Internet of Things, cloud computing and so on. At present, how to mine valuable information from massive data quickly and accurately is an important research topic around the whole world. The traditional single swarm intelligence optimization algorithms cannot satisfy the requirement of the massive data processing, cloud computing technology development both in the efficiency and computational complexity, which has derived a new research direction in data analysis and processing.As a parallel programming model, MapReduce is skilled in dealing with big data and large calculation. According to the characteristics of swarm intelligence optimization algorithm, combined with the MapReduce programming model, we can effectively use the computing advantages of cluster to deal with the large amount of calculations, and shorten the time of task execution. This article studies and analyzes several classical swarm intelligence optimization algorithms, improving Rain Forest Algorithm’s defects of slow convergence and low accuracy in high dimensional optimization problem. And then, the article proposes an idea to implement it based on MapReduce programming model, referring to the properties of Rain Forest Algorithm,The main researches of this paper are arranged as following:Firstly, the article briefly introduces the principle, process of the Map Reduce programming model and several classic swarm intelligence optimization algorithms; Secondly, it studies on the principle and process of rainforest algorithm, improving the rain forest algorithm from the two aspects of the parameter setting and the iterative method, which brings in a Rain forest algorithm based on adaptive population and a Rain forest algorithm based on sun vector; Thirdly, this paper analysis the feasibility of the rainforest algorithm of MapReduce parallel scheme by integrating rainforest algorithm with the MapReduce programming mode; At last, it proposes a coarse-grained Rain forest algorithm based on MapReduce, and makes cluster experiments on the Hadoop platform. The results show that, This algorithm can not only maintain the accuracy of the rainforest algorithm, but also can utilize the advantages of the MapReduce programming model, which would shorten the execution time of the algorithm and improve the efficiency of the implementation.