Swarm Intelligence Optimization Algorithm And The Application Under The Big Data Environment

Big data is the inevitable product of modern information technology. To realize the analysis and application of big data will bring immeasurable economic value and boost social development. Under the big data environment, the scale, generation speed and handling difficulties of data have tremendously complicated the aspects to be optimized. Big data often feature a large scale, multi dimensions, powerful confinement, quick update and multiple objectives. Concerning these characteristics, it is extremely difficult to seek the globally optimal solution through the traditional optimization method.Swarm intelligence can achieve self-organization, paralleled, distributional and decentralized control. This provides the basis for seeking the solution plan for the complex optimization problems. However, the traditional swarm intelligence model is not in a position to solve the complex problems in the practical big data environment, including multi-dimensional, highly-restrictive and multi-objective optimization. Therefore, it is of vital theoretical and practical significance to design a new intelligent optimization algorithm targeted at the optimization problems under the big data environment.This paper succeeds in exploring a swarm intelligence optimization algorithm based on the dynamic characteristics concerning problems such as large scale, multiple dimensions, and powerful confinement. The major research content and innovational points of this paper are shown below:1. An in-depth analysis and discussion is conducted concerning the characteristics and solutions of the optimization problems under the big data environment. The current swarm intelligence algorithms and their improvement and application are studied.2. A self-adaption control of the swarm intelligence model based on the Logistics swarm dynamic model is put forward to realize the self-adapting dynamic changes of swarm model during the evolution process. The realization of the strategy does not rely on the specific steps of the algorithm evolution operation, so it is applicable to natural algorithms targeted at various swarm optimizations. The strategy is applied to the particle swarm optimization algorithm with the typical trial function adopted to test its performance. The experiment result shows that the new algorithm put forward based on the self-adapting dynamic control strategy of the swarm model registers a great improvement in terms of solution precision and convergence rate compared with the original algorithm.3. The bio-dynamic optimization algorithm is a new biological evolution dynamic optimization algorithm which integrates the concept and theory of the swarm dynamics into the typical swarm intelligence algorithm. This paper adopts the standard test as an example to qualitatively analyze the characteristics of the algorithm. Through the analysis of the large scale and multiple dimensions, it can be seen that the swarm intelligence algorithm based on the Lagrangian dynamics has better swarm diversity and global optimization capacity compared with the traditional algorithm.4. Vehicle routing is a typical combinational optimization problem, which has been proved as an NP difficult problem. Under the big data environment, vehicle routing problem is characterized by powerful confinement and large scale. Concerning the problem, this paper applies the biodynamic optimization algorithm to solving vehicle routing and capacitated vehicle routing and compares it with the algorithm. The result shows that the algorithm is competitive in solving combiantional optimization problems.