一种改进油田产量预测算法的研究
|
摘要
针对油田产量预测算法存在的收敛速度慢、精度不高的缺陷,论文提出一种基于BP网络算法的改进油田产量预测算法。该算法通过在BP网络预测算法中引入自适应的权值调整算子,在偏差逆向传导的过程中优化了网络连接权值的调整,加快预测算法的收敛速度,使得算法摆脱在样本数据不足的情况下可能出现的局部最优缺陷,提高了算法预测精度。最后通过算例验证了该算法的有效性及其比传统BP算法更好的性能。
In view of the slow convergence speed and low precision of oilfield production prediction algorithm,a prediction algorithm of oilfield production based on BP network algorithm is proposed in this paper. The algorithm through the forecast weight adjustment operator into the adaptive algorithm in the BP network,in the process of optimization of reverse bias to adjust the connection weights of the network,accelerates the prediction speed of convergence of the algorithm,and makes the algorithm to get rid ofpossible defects in the local optimal sample data under the condition of insufficient,reduces the prediction results of deviation thatmakes up the disadvantage of the traditional BP network prediction algorithm. Finally,a numerical example is given to demonstratethe effectiveness of the proposed algorithm and the advantages of the traditional BP network algorithm.
In view of the slow convergence speed and low precision of oilfield production prediction algorithm,a prediction algorithm of oilfield production based on BP network algorithm is proposed in this paper. The algorithm through the forecast weight adjustment operator into the adaptive algorithm in the BP network,in the process of optimization of reverse bias to adjust the connection weights of the network,accelerates the prediction speed of convergence of the algorithm,and makes the algorithm to get rid ofpossible defects in the local optimal sample data under the condition of insufficient,reduces the prediction results of deviation thatmakes up the disadvantage of the traditional BP network prediction algorithm. Finally,a numerical example is given to demonstratethe effectiveness of the proposed algorithm and the advantages of the traditional BP network algorithm.
引文
[1]丁冠阳,黄世军,张雪娇.低渗透油藏油井递减后产量递减不确定性评价方法[J].大庆石油地质与开发,2017,36(4):47-51.DING Guanyang,HUANG Shijun,ZHANG Xuejiao. Uncertainty evaluation method of production decline in lowpermeability reservoir after decreasing oil wells[J]. Daqingpetroleum geology and development,2017,36(4):47-51.
[2]严禛,伍星蓉.基于BFO-BP神经网络的储层预测研究[J].能源与环保,2017,39(7):210-213.YAN Zhen,WU Xingrong. Study on energy and environmental protection[J]. Reservoir prediction based onBFO-BP neural network,2017,39(7):210-213.
[3]杨志浩,李治平.基于BP神经网络的底水油藏控水压裂选段新方法[J].地质与勘探,2017,53(4):818-824.YANG Zhihao,LI Zhiping. The bottom water reservoir BPneural network control method of fracturing water from[J]. Geology and prospecting based on,2017,53(4):818-824.
[4]常丽,陈冬.基于改进BP神经网络的管外测量原油含水率研究[J].仪表技术与传感器,2016(9):123-126.CHANG Li,CHEN Dong. Research on water content measurement of crude oil based on improved BP neural network[J]. Instrument technology and sensor,2016(9):123-126.
[5]田亚鹏,鞠斌山.基于遗传算法改进BP神经网络的页岩气产量递减预测模型[J].中国科技,2016,11(15):1710-1715.TIAN Yapeng,JU Binshan. Improved genetic algorithmBP neural network prediction model of shale gas production decline Chinese technology based on[J]. China Science and Technology,2016,11(15):1710-1715.
[6]陈志海,董广为,廉培庆.穿透非均质储层的复杂轨迹井产量计算新方法[J].石油与天然气地质,2016,37(3):444-449.CHEN Zhihai,DONG Guangwei,LIAN Peiqing. Throughcomplex trajectory wells heterogeneous reservoir calculation method of[J]. Oil and gas geology,2016,37(3):444-449.
[7]李昌盛,宋海,肖莉,等.基于遗传算法优化BP神经网络的地层破裂压力预测方法[J].西安石油大学学报(自然科学版),2015,30(5):75-79.LI Changsheng,SONG Hai,XIAO Li,et al. Genetic algorithm BP neural network formation fracture pressure prediction method based on[J]. Journal of Xi'an PetroleumUniversity(Natural Science Edition),2015,30(5):75-79.
[8]汪雷,林亮,李晶晶,等.基于测井信息的煤储层渗透率BP神经网络预测方法[J].煤炭科学技术,2015,43(7):122-126.WANG Lei,LIN Liang,LI Jingjing,et al. coal reservoirpermeability logging information BP neural network prediction method based on[J]. Coal science and technology,2015,43(7):122-126.
[9]王卫江,史玥婷,刘箭言,等.基于神经网络的电参数反演载荷算法[J].北京理工大学学报,2015,35(7):706-710.WANG Weijiang,SHI Wei,LIU Jianyan,et al. Load inversion algorithm of electric parameters based on neuralnetwork[J]. Journal of Beijing Institute of Technology,2015,35(7):706-710.
[10]杜睿山,唐世伟,王辉,等.基于BP的油田产能成功度综合后评价模型研究[J].计算机与数字工程,2014,42(8):1325-1328,1426.DU Du,TANG Shiwei,WANG Hui,et al. Research oncomprehensive post evaluation model of oilfield productivity based on BP.[J]. Computer and digital engineering,2014,42(8):1325-1328,1426.
[11]吴财芳,姚帅,杜严飞.基于时间序列BP神经网络的煤层气井排采制度优化[J].中国矿业大学学报,2015,44(1):64-69.WU Caifang,YAO Shuai,DU Yanfei. CBM time seriesBP neural network drainage system optimization basedon[J]. Journal of China University of Mining and Technology,2015,44(1):64-69.
[12]熊健,曾山,王绍平.不对称垂直裂缝井产量递减规律[J].西安石油大学学报(自然科学版),2014,29(2):74-77.XIONG Jian,ZENG Shan,WANG Shaoping. Productiondecline law of asymmetric vertical fractured well[J].Journal of Xi'an Petroleum University(Natural ScienceEdition),2014,29(2):74-77.
[13]熊健,曾山,王绍平.低渗透油藏变导流垂直裂缝井产能模型[J].岩性油气藏,2013,25(6):122-126.XIONG Jian,ZENG Shan,WANG Shaoping. Productivity model of variable conductivity vertical fractured wellin low permeability reservoir[J]. Lithologic reservoir,2013,25(6):122-126.
[14]樊灵,赵孟孟,殷川,等.基于BP神经网络的油田生产动态分析方法[J].断块油气田,2013,20(2):204-206.FAN Ling,ZHAO Mengmeng,YIN Chuan,et al. Dynamic oil production BP neural network analysis methodbased on[J]. Fault block oil&gas field,2013,20(2):204-206.
[15]艾敬旭,单学军,侯天江.五点井网注水井压裂裂缝参数对油井产量的影响[J].断块油气田,2011,18(5):649-652.AI Jingxu,SHAN Gang,HOU Gang. Influence of fracture parameters on oil well production in five well patternnet injection well[J]. Fault block oil and gas field,2011,18(5):649-652.
[2]严禛,伍星蓉.基于BFO-BP神经网络的储层预测研究[J].能源与环保,2017,39(7):210-213.YAN Zhen,WU Xingrong. Study on energy and environmental protection[J]. Reservoir prediction based onBFO-BP neural network,2017,39(7):210-213.
[3]杨志浩,李治平.基于BP神经网络的底水油藏控水压裂选段新方法[J].地质与勘探,2017,53(4):818-824.YANG Zhihao,LI Zhiping. The bottom water reservoir BPneural network control method of fracturing water from[J]. Geology and prospecting based on,2017,53(4):818-824.
[4]常丽,陈冬.基于改进BP神经网络的管外测量原油含水率研究[J].仪表技术与传感器,2016(9):123-126.CHANG Li,CHEN Dong. Research on water content measurement of crude oil based on improved BP neural network[J]. Instrument technology and sensor,2016(9):123-126.
[5]田亚鹏,鞠斌山.基于遗传算法改进BP神经网络的页岩气产量递减预测模型[J].中国科技,2016,11(15):1710-1715.TIAN Yapeng,JU Binshan. Improved genetic algorithmBP neural network prediction model of shale gas production decline Chinese technology based on[J]. China Science and Technology,2016,11(15):1710-1715.
[6]陈志海,董广为,廉培庆.穿透非均质储层的复杂轨迹井产量计算新方法[J].石油与天然气地质,2016,37(3):444-449.CHEN Zhihai,DONG Guangwei,LIAN Peiqing. Throughcomplex trajectory wells heterogeneous reservoir calculation method of[J]. Oil and gas geology,2016,37(3):444-449.
[7]李昌盛,宋海,肖莉,等.基于遗传算法优化BP神经网络的地层破裂压力预测方法[J].西安石油大学学报(自然科学版),2015,30(5):75-79.LI Changsheng,SONG Hai,XIAO Li,et al. Genetic algorithm BP neural network formation fracture pressure prediction method based on[J]. Journal of Xi'an PetroleumUniversity(Natural Science Edition),2015,30(5):75-79.
[8]汪雷,林亮,李晶晶,等.基于测井信息的煤储层渗透率BP神经网络预测方法[J].煤炭科学技术,2015,43(7):122-126.WANG Lei,LIN Liang,LI Jingjing,et al. coal reservoirpermeability logging information BP neural network prediction method based on[J]. Coal science and technology,2015,43(7):122-126.
[9]王卫江,史玥婷,刘箭言,等.基于神经网络的电参数反演载荷算法[J].北京理工大学学报,2015,35(7):706-710.WANG Weijiang,SHI Wei,LIU Jianyan,et al. Load inversion algorithm of electric parameters based on neuralnetwork[J]. Journal of Beijing Institute of Technology,2015,35(7):706-710.
[10]杜睿山,唐世伟,王辉,等.基于BP的油田产能成功度综合后评价模型研究[J].计算机与数字工程,2014,42(8):1325-1328,1426.DU Du,TANG Shiwei,WANG Hui,et al. Research oncomprehensive post evaluation model of oilfield productivity based on BP.[J]. Computer and digital engineering,2014,42(8):1325-1328,1426.
[11]吴财芳,姚帅,杜严飞.基于时间序列BP神经网络的煤层气井排采制度优化[J].中国矿业大学学报,2015,44(1):64-69.WU Caifang,YAO Shuai,DU Yanfei. CBM time seriesBP neural network drainage system optimization basedon[J]. Journal of China University of Mining and Technology,2015,44(1):64-69.
[12]熊健,曾山,王绍平.不对称垂直裂缝井产量递减规律[J].西安石油大学学报(自然科学版),2014,29(2):74-77.XIONG Jian,ZENG Shan,WANG Shaoping. Productiondecline law of asymmetric vertical fractured well[J].Journal of Xi'an Petroleum University(Natural ScienceEdition),2014,29(2):74-77.
[13]熊健,曾山,王绍平.低渗透油藏变导流垂直裂缝井产能模型[J].岩性油气藏,2013,25(6):122-126.XIONG Jian,ZENG Shan,WANG Shaoping. Productivity model of variable conductivity vertical fractured wellin low permeability reservoir[J]. Lithologic reservoir,2013,25(6):122-126.
[14]樊灵,赵孟孟,殷川,等.基于BP神经网络的油田生产动态分析方法[J].断块油气田,2013,20(2):204-206.FAN Ling,ZHAO Mengmeng,YIN Chuan,et al. Dynamic oil production BP neural network analysis methodbased on[J]. Fault block oil&gas field,2013,20(2):204-206.
[15]艾敬旭,单学军,侯天江.五点井网注水井压裂裂缝参数对油井产量的影响[J].断块油气田,2011,18(5):649-652.AI Jingxu,SHAN Gang,HOU Gang. Influence of fracture parameters on oil well production in five well patternnet injection well[J]. Fault block oil and gas field,2011,18(5):649-652.