基于GPU的并行优化算法研究

该文介绍了GPU概念及发展、并行计算的概念以及与串行计算相比而具有的优势,指出智能优化算法具有天然的并行性和分布性,在基础理论和工程应用中具有很高的研究价值,该文对智能优化算法中的模拟退火算法、遗传算法、禁忌搜索算法、人工神经网络算法及蚁群算法的原理和实际应用进行了深入研究,提出了基于GPU的并行优化算法。

This paper introduces the concept and development of GPU, the concept of parallel computation and the advantage of the serial calculation, and points out that the intelligent optimization algorithm has the natural parallelism and the distribution, and has very high research value in the basic theory and engineering application. In this paper, the principle and practical application of the simulated annealing algorithm, the genetic algorithm, the tabu search algorithm, the artif icial neural network algorithm and the ant colony algorithm in the intelligent optimization algorithm are deeply studied, and a parallel optimization algorithm based on the GPU is proposed.