基于FI-GA-NN融合的区域能源安全外生警源分级预警研究

发布时间：2018-10-05 18:31

【摘要】：区域能源安全预警研究对于解决中国现阶段区域能源安全突发事件频现问题,保障区域经济与区域安全协调发展具有重要现实意义。本文以区域能源安全外生警源为研究对象,通过对区域能源安全事件案例收集及整理,构建了能源安全外生警源预警指标和数据集。融合模糊积分(Fuzzy Integral)、遗传算法(Genetic Algorithm)和神经网络(Neural Network)等方法的基本原理,设计了区域能源安全外生警源分级预警的FI-GA-NN模型,该模型首先利用模糊积分方法评估出区域能源安全外生警源样本分级预警的期望值,然后利用训练样本对遗传神经网络进行训练,最后对外生警源测试样本进行分级预警。实验测试结果表明:(1)利用FI-GA-NN模型对外生警源训练样本(1999-2006年)进行拟合训练,模型收敛速度快,训练到第717步时,模型误差平方和小于期望值。经过大约60代的搜索后模型的拟合度趋于稳定,模型训练的实际输出值与期望输出值较接近;(2)利用FI-GA-NN模型对能源安全外生警源测试样本(2007-2015年)进行分级预警,预警准确率较高,能有效提高区域能源安全外生警源预警的正确性,降低预警风险,模型体现出了较强的应用价值。
[Abstract]:The study of regional energy security early warning is of great practical significance in solving the problem of the frequency of regional energy security emergencies and ensuring the coordinated development of regional economy and regional security. This paper takes the regional energy security exogenous warning source as the research object and constructs the early warning index and data set of the energy security exogenous alarm source by collecting and sorting out the regional energy security incident cases. Based on the basic principles of fuzzy integral (Fuzzy Integral), genetic algorithm (Genetic Algorithm) and neural network (Neural Network), the FI-GA-NN model of regional energy security exogenous alarm is designed. The model firstly uses fuzzy integral method to evaluate the expectation value of regional energy security exogenous warning source samples, then uses training samples to train genetic neural network, and finally uses external source test samples to carry out classification early warning. The experimental results show that: (1) the model convergence rate is fast and the square sum of the model error is less than the expected value when the training sample of FI-GA-NN model is used for the training of external sources (1999-2006). After about 60 generations of searching, the fitting degree of the model tends to be stable, and the actual output value of the model training is close to the expected output value. (2) using the FI-GA-NN model to grade the energy security exogenous alarm source test samples (2007-2015), the warning accuracy is high. It can effectively improve the correctness of regional energy security exogenous warning and reduce the risk of early warning. The model reflects a strong application value.
【作者单位】：重庆理工大学管理学院;
【基金】：国家自然科学基金项目(71301181) 重庆市社会科学规划项目(2015YBSH051) 重庆市教委人文社科项目(15SKG134)
【分类号】：F224;F426.2

【参考文献】