基于扩展神经网络的网格资源调度优化算法

Optimal scheduling algorithm for grid resource based on extensive neural network

针对具有时间约束和若干资源约束的网格资源调度问题,提出了一种基于扩展神经网络的求解新方法GRSENN。资源调度问题首先被分解为一系列多维背包问题并提出相应的数学模型,然后通过把问题的各种约束表示为Hopfield神经网络的能量函数项,进而导出神经网络模型。实验仿真结果表明该方法的有效性,并可避免通常神经网络所具有的容易陷入局部极小点的缺陷。

A neural network approach （GRSENN） is proposed to solve the problems related to grid resource scheduling, which has time and resource constraints. The original problem is first decomposed into a series of multidimensional knapsack models and a mathematical model is established at the same time. Then by means of expressing the various kinds of problem with the energy function of Hopfield neural network, the neural network model is derived.The simulation results show that GRSENN works effectively for this kind of problems and can effectively avoid some typical shortcomings in this field, such as local minima.