输变电工程耐热导线材料价格预测仿真  

Simulation of power transmission project heat-resistant conductor material price

作　　者：朱晓虎[1] 王道静[1] 刘士李[1] 殷敏[1] 

机构地区：[1]国网安徽省电力公司电力经济技术研究院,安徽合肥230071

出　　处：《计算机仿真》2017年第8期425-428,456,共5页Computer Simulation

摘　　要：对于输变电工程耐热导线材料的价格预测,可以有效的掌握当前导线材料价格的变化规律。对导线材料价格进行预测,需要将历史价格分解为若干价格分量,对各个分量的价格进行预测,并将各个预测分量进行叠加输出,完成对材料价格的预测。传统方法通过优化搜索获取的粒子的位置向量分量作为网络的权值,但忽略了将预测分量进行叠加,导致预测精度低。提出基于集合经验模态分解的输变电工程耐热导线材料价格预测方法。利用EDD分解原理将历史价格分解为若干价格分量,对各个分量的价格进行预测,将各个预测分量进行叠加作为价格预测的期望输出值,融合于神经网络理论思想给出价格预测的期望输出和实际输出的差值,利用双极性Sigmoid函数完成对输变电工程耐热导线材料价格预测。实验仿真证明,所提方法预测精确度高,可以为电网工程扩容改造提供有力的依据。In this paper, a method for price expectation of heat -resistant conductor material in power transmis- sion and transformation is proposed based on ensemble empirical mode decomposition. Historical price was decom- posed into several price components using EDD decomposition theory. Then, price of each component was predicted and each prediction component was superposed to be desired output of price expectation. Integrated with neural net- work theory,difference value between desired output and actual output was provided. Thus, using bipolar Sigmoid function ,the price expectation was completed. From experimental simulation, the conclusion can be drawn that the method has high predic tion precision and can provide powerful gist for extension and renovation of power grid pro- ject.

关 键 词：输变电工程 耐热导线材料 价格预测 

分 类 号：TU22[建筑科学—建筑设计及理论]