基于基因表达式编程的电力消费预测模型--武汉市的实证

Study on the Electricity Consumption Forecasting Model Based on Gene Expression Programming

本文运用基因表达式编程(GEP)方法,基于武汉市1990年至2005年的相关统计数据,构建了电力消费预测模型。研究显示,GEP法能够较好地克服有限样本数据序列的缺陷,并具有精度高、计算过程较为简化等优势。研究结果表明,武汉市电力消费主要影响因素的影响程度依次排序为:产业结构、能源效率改进、社会消费总额、人口、地区生产总值指数和电力价格指数。基于上述结论我们提出了解决电力短缺应该以加强电力能源效率改进为主导的政策建议。

This paper focuses on establishing the electricity consumption-forecasting model based on the Gene Expression Programming that was brought forward by a Portuguese researcher, named Ferreira C. in 2001. The result indicates that the methodology that can overcome barriers such as limited data, complex calculating process is feasible and effective with higher accuracy.