ABSTRACT

Energy consumption prediction is a hot issue, which is of great significance to regional energy security. In the existing prediction research with small samples, the time delay characteristic of an energy consumption system in itself is often ignored. To reflect the time delay characteristic of an energy consumption system and accurately grasp its development trend, a novel nonlinear time delay grey model with fractional order accumulation is presented. The new model is utilized to forecast and analyze Guangdong’s primary energy consumption, in which the time delay parameter is ascertained by the grey correlation analysis method, and the other parameters are determined via particle swarm optimization. The results show the simulation accuracy of the new model is higher than those of the other 3 grey models, and the predicted results in the next three years can provide decision-making and theoretical reference for the relevant departments of Guangdong province.

摘要

能耗预测是一个热点问题，对区域能源安全具有重要意义。在现有的小样本预测研究中，能源消耗系统本身的时延特性常常被忽略。为了反映能耗系统的时延特性并准确掌握其发展趋势，提出了一种具有分数阶累加的非线性时滞灰色模型。利用新模型对广东省一次能源消费进行预测和分析，通过灰色关联分析法确定了时延参数，并通过粒子群算法确定了其他参数。结果表明，新模型的仿真精度高于其他三种灰色模型，