Parallel Programming Of Particle Swarm Optimi Zation Algorithm Based On Spark And Its Application In Reservoir Scheduling

Reservoir scheduling is an important link in the management of reservoir group, and the level of reservoir scheduling has a direct impact on the comprehensive benefits of hydropower stations.The Heihe River Basin has a large amount of water resources and complex data relations.It is difficult to optimize the scheduling using conventional techniques.Using big data processing and evolutionary computation techno logy,and based on the big data calculation framework Spark and the parallel particle swarm optimization algorithm,this paper has developed a multi-objective optimal scheduling system for the Heihe River Basin reservoir based on the current water resources scheduling problem in the Heihe River Basin.In the process of research and development,the author analyzes the characteristics of multiple objectives of the Heihe River Basin water resources optimal scheduling system,transforms multiple targets into individual targets,and obtains the solution model of the problem.Then we study the parallel algorithm programming model,the particle swarm optimization algorithm for solving problem and its parallelization strategy,and the particle swarm algorithm parallelization method based on Spark big data calculation framework.Based on the study of theory and technology,Hadoop2.7.1,Sparkl.5.2,Spark on Yarn software as the foundation, to build a big data platform;in HDFS distributed file system of the platform is stored in the Heihe Basin water resources data.Then,on the Ubuntu Linux operating system environment and the Spark platform,the Scala language was used to develop the multi-objective optimal scheduling parallel program of the reservoir group based on the particle swarm optimization algorithm,and the multi-objective optimal scheduling system of reservoir group which can deal with the big data of the reservoir and the high-speed operation scheduling optimization program can be realized.The data loading,pro gram operation,the results of inquiries and other operations of the system are performed on the Ubuntu Linux operation system and Spark platform,but for ordinary users who are not familiar with the Spark operation mechanism,there are lots of difficulties.In order to solve this problem,we also developed a reservoir group of...