基于GA-BP神经网络的配电网工程造价预测
|
摘要
为了解决前馈(BP)神经网络在配电网工程建设工程造价预测时,容易陷入局部极小而导致预测精度降低的问题,提出了一种GA-BP神经网络的配电网工程造价预测模型。模型试算与分析结果表明:除了个别样本数据外,GA-BP模型预测数据的相对误差小于BP模型预测数据的相对误差。其中,GA-BP模型的预测数据的相对误差整体最小,BP模型的相对误差整体最大,BP整体的相对误差要稍小于GA-BP。GA-BP和BP的模型平均相对误差数值更小,GA-BP模型的平均相对误差最小,说明该模型的预测稳定性最强。此外,GA-BP和BP的模型稳定性和预测的精度上都要优于GA-BP和BP。其中,GA-BP的预测模型最好,BP预测模型最差。该基于GA-BP神经网络的配电网工程造价预测模型为提高配电网工程造价预测精度提供了一定的理论基础。
In order to solve the problem that the BP neural network is easy to fall into the local minimum when forecasting the construction cost of distribution network project leads to the decrease of prediction accuracy,a GA-BP neural network project cost prediction model is proposed to improve the prediction accuracy of BP neural network. The results show that,in addition to individual sample data,the relative error of GA-BP model prediction data is smaller than that of BP model. Among them,the overall relative error of GA-BP model is the minimum,the overall relative error of BP model is the maximum,and the overall relative error of BP is slightly smaller than GA-BP. The average relative error values of GA-BP and BP models are smaller,and the average relative error of GA-BP model is the smallest,indicating that the prediction stability of this model is the strongest. In addition,the stability and prediction accuracy of GA-BP and BP models are better than GA-BP and BP,among which GA-BP model is the best and BP model is the worst. This study provides a theoretical basis for improving the forecasting accuracy of distribution network engineering cost.
In order to solve the problem that the BP neural network is easy to fall into the local minimum when forecasting the construction cost of distribution network project leads to the decrease of prediction accuracy,a GA-BP neural network project cost prediction model is proposed to improve the prediction accuracy of BP neural network. The results show that,in addition to individual sample data,the relative error of GA-BP model prediction data is smaller than that of BP model. Among them,the overall relative error of GA-BP model is the minimum,the overall relative error of BP model is the maximum,and the overall relative error of BP is slightly smaller than GA-BP. The average relative error values of GA-BP and BP models are smaller,and the average relative error of GA-BP model is the smallest,indicating that the prediction stability of this model is the strongest. In addition,the stability and prediction accuracy of GA-BP and BP models are better than GA-BP and BP,among which GA-BP model is the best and BP model is the worst. This study provides a theoretical basis for improving the forecasting accuracy of distribution network engineering cost.
引文
[1]黄兵,赵晋泉.基于连续潮流的配电网供电能力评估[J].电力工程技术,2019,38(1):14-18.
[2]李园园,张平朗.±800 k V特高压直流输电线路工程造价结构及投资水平分析[C]//电网工程造价管理优秀论文,2011.
[3]秦博,张平朗. 1 000 k V特高压交流输电线路工程投资研究[C]//电网工程造价管理优秀论文,2011.
[4] DYBOWSKI R,ROBERTS S J. Confidence intervals and prediction intervals for feed-forward neural networks[C]//In Clinical Plications of Artificial Neural Network. Cambridge University Press.2008:298-327.
[5] YAMILLE D,VALLE E. Particle swarm optimization:basic concepts,variants and applications in power systems[J]. IEEE Transactims on Evolutionary Computation,2008,12(2):171-195.
[6]郭新菊,邵永刚,李旭阳,等.基于BP神经网络的配电网工程造价预测模型建立与应用[J].电子设计工程,2017,25(23):63-68.
[7]阮伟兵.配电网建设项目中工程管理问题策略分析[J].科技视界,2018(35):234-235.
[8]向乃瑞,闫海,王炜,等. GA-BP神经网络预测金属腐蚀速率[J].电力学报,2018,33(1):48-54.
[9]陈征,黎青青,肖乃松,等.基于GA-BP神经网络的柴油喷雾贯穿距预测[J].中南大学学报(自然科学版),2018,49(1):247-252.
[10]宋云亭,吴俊玲,彭冬,等.基于BP神经网络的城网供电可靠性预测方法[J].电网技术,2008(20):56-59.
[11]顾瑶媛,谢红,吴旭波.基于MIV特征筛选和BP神经网络的三维人体参数转换[J].上海工程技术大学学报,2012,26(4):361-364.
[12]张永礼,陆刚,武建章.基于MIV和GA-BP模型的农业机械化水平影响因素实证分析[J].农业现代化研究,2015,36(6):1026-1031.
[13]罗建春,晁勤,罗洪,等.基于LVQ-GA-BP神经网络光伏电站出力短期预测[J].电力系统保护与控制,2014,42(13):89-94.
[14]姚仲敏,潘飞,沈玉会,等.基于GA-BP和POS-BP神经网络的光伏电站出力短期预测[J].电力系统保护与控制,2015,43(20):83-89.
[15]岳衡,韩翔宇,张继春,等. GA改进BP神经网络在抗滑桩孔爆破开挖中的应用[J].工程爆破,2016,22(2):28-33.
[2]李园园,张平朗.±800 k V特高压直流输电线路工程造价结构及投资水平分析[C]//电网工程造价管理优秀论文,2011.
[3]秦博,张平朗. 1 000 k V特高压交流输电线路工程投资研究[C]//电网工程造价管理优秀论文,2011.
[4] DYBOWSKI R,ROBERTS S J. Confidence intervals and prediction intervals for feed-forward neural networks[C]//In Clinical Plications of Artificial Neural Network. Cambridge University Press.2008:298-327.
[5] YAMILLE D,VALLE E. Particle swarm optimization:basic concepts,variants and applications in power systems[J]. IEEE Transactims on Evolutionary Computation,2008,12(2):171-195.
[6]郭新菊,邵永刚,李旭阳,等.基于BP神经网络的配电网工程造价预测模型建立与应用[J].电子设计工程,2017,25(23):63-68.
[7]阮伟兵.配电网建设项目中工程管理问题策略分析[J].科技视界,2018(35):234-235.
[8]向乃瑞,闫海,王炜,等. GA-BP神经网络预测金属腐蚀速率[J].电力学报,2018,33(1):48-54.
[9]陈征,黎青青,肖乃松,等.基于GA-BP神经网络的柴油喷雾贯穿距预测[J].中南大学学报(自然科学版),2018,49(1):247-252.
[10]宋云亭,吴俊玲,彭冬,等.基于BP神经网络的城网供电可靠性预测方法[J].电网技术,2008(20):56-59.
[11]顾瑶媛,谢红,吴旭波.基于MIV特征筛选和BP神经网络的三维人体参数转换[J].上海工程技术大学学报,2012,26(4):361-364.
[12]张永礼,陆刚,武建章.基于MIV和GA-BP模型的农业机械化水平影响因素实证分析[J].农业现代化研究,2015,36(6):1026-1031.
[13]罗建春,晁勤,罗洪,等.基于LVQ-GA-BP神经网络光伏电站出力短期预测[J].电力系统保护与控制,2014,42(13):89-94.
[14]姚仲敏,潘飞,沈玉会,等.基于GA-BP和POS-BP神经网络的光伏电站出力短期预测[J].电力系统保护与控制,2015,43(20):83-89.
[15]岳衡,韩翔宇,张继春,等. GA改进BP神经网络在抗滑桩孔爆破开挖中的应用[J].工程爆破,2016,22(2):28-33.