磷石膏充填体强度GA-BP神经网络预测模型
|
摘要
在测试磷石膏充填材料物理化学性能的基础上,测定了不同配比充填体的强度,用分形理论揭示了磷石膏充填体强度特征。研究表明,磷石膏充填体强度与水泥含量、料浆浓度、粒径分形维数、孔隙分形维数及分维数相关率相关。根据充填体强度影响因素,建立了磷石膏加粉煤灰作为充填料的充填体强度预测的BP神经网络模型。同时利用遗传算法优化BP神经网络的学习过程,验算结果显示,GA-BP神经网络模型预测误差在4%以内,具有较高的预测精度。
After testing physical and chemical properties of phosphogypsum backfill material,the strengths of backfill with different proportioning were obtained from experiments,and the strength characteristics of phosphogypsum backfill were revealed by fractal theory.Researches show that the strengths of phosphogypsum backfill are related to the content of cement,the concentration of slurry,the fractal dimension of particle,the fractal dimension of porosity and the correlation coefficient of fractal dimension.Based on the influencing factors of backfill strength,the BP neural network model was established to predict strengths of backfill by using phosphogypsum and coal ash as backfill materials.A genetic algorithm was adopted to optimize the BP neural network learning process.The calculation result shows GA-BP neural network model has higher prediction precision,with errors less than 4%.
After testing physical and chemical properties of phosphogypsum backfill material,the strengths of backfill with different proportioning were obtained from experiments,and the strength characteristics of phosphogypsum backfill were revealed by fractal theory.Researches show that the strengths of phosphogypsum backfill are related to the content of cement,the concentration of slurry,the fractal dimension of particle,the fractal dimension of porosity and the correlation coefficient of fractal dimension.Based on the influencing factors of backfill strength,the BP neural network model was established to predict strengths of backfill by using phosphogypsum and coal ash as backfill materials.A genetic algorithm was adopted to optimize the BP neural network learning process.The calculation result shows GA-BP neural network model has higher prediction precision,with errors less than 4%.
引文
[1]古德生,李夕兵.现代金属矿床开采科学技术[M].北京:冶金工业出版社,20063
[2]李夕兵,刘志祥,古德生.矿业固体尾废与采空区互为资源战略的思考[J].矿冶工程,2005,25(6):1-15.
[3]赵才智,周华强,柏建彪,等.膏体充填材料强度影响因素分析[J].辽宁工程技术大学学报,2006,25(6):904-906.
[4]CHANG Qing-liang,ZHOU Hua-qiang,HOU Chou-jiong.Using par-ticle swarm optimization algorithm in an artificial neural network toforecast the strength of paste filling material[J].Journal of china uni-versity of mining&technology,2008,18(4):551-555.
[5]邓建,李夕兵,古德生.无间柱连续采矿法的充填体可靠性研究(Ⅰ)[J].矿冶工程,2001,21(1):21-23.
[6]常庆粮,周华强,秦剑云,等.膏体充填材料配比的神经网络预测研究[J].采矿与安全工程学报,2009,26(3):74-77.
[7]刘志祥,李夕兵.尾砂级配的混沌优化[J].中南大学学报(自然科学版),2005,36(4):683-688.
[8]刘志祥,李夕兵.尾砂分形级配与胶结强度的知识库研究[J].岩石力学与工程学报,2005,24(10):1789-1793.
[9]特科特D L.分形与混沌[M].北京:地震出版社,1993.
[10]Prosperini N,Perugini D.Particle size distributions of some soilsfrom the Umbria Region[J].Geoderma,2008,145:185-195.
[11]TAO Gao-Liang,ZHANG Ji-Ru.Two categories of fractal models ofrock and soil expressing volume and size-distribution of pores andgrains[J].Chinese Science Bulletin,2009,54(23):4458-4467.
[12]刘志祥.深部开采高阶段尾砂充填体力学与非线性优化设计[D].长沙:中南大学资源与安全工程学院,2005.
[13]Mkrtchian T,Dight P.Experience in electroactivation backfill hard-ening[J].Australasian Institute of Mining and Metallurgy Publica-tion Series,1998(n1):333-335.
[14]李一帆,张建明,邓飞,等.深部采空区尾砂胶结充填体强度特性试验研究[J].岩土力学,2005,26(6):865-868.
[15]彭续承.充填理论及应用[M].长沙:中南工业大学出版社,1998.
[16]何玉斌,李新忠.神经网络控制技术及其应用[M].北京:科学出版社,2000.
[17]朱凯,王正林.精通MATLAB神经网络[M].北京:电子工业出版社,2010.
[18]郭海湘,诸克军,胡杰,等.GA-BP嵌套算法的理论及应用[J].数学的实践与认识,2008,38(1):116-125.
[2]李夕兵,刘志祥,古德生.矿业固体尾废与采空区互为资源战略的思考[J].矿冶工程,2005,25(6):1-15.
[3]赵才智,周华强,柏建彪,等.膏体充填材料强度影响因素分析[J].辽宁工程技术大学学报,2006,25(6):904-906.
[4]CHANG Qing-liang,ZHOU Hua-qiang,HOU Chou-jiong.Using par-ticle swarm optimization algorithm in an artificial neural network toforecast the strength of paste filling material[J].Journal of china uni-versity of mining&technology,2008,18(4):551-555.
[5]邓建,李夕兵,古德生.无间柱连续采矿法的充填体可靠性研究(Ⅰ)[J].矿冶工程,2001,21(1):21-23.
[6]常庆粮,周华强,秦剑云,等.膏体充填材料配比的神经网络预测研究[J].采矿与安全工程学报,2009,26(3):74-77.
[7]刘志祥,李夕兵.尾砂级配的混沌优化[J].中南大学学报(自然科学版),2005,36(4):683-688.
[8]刘志祥,李夕兵.尾砂分形级配与胶结强度的知识库研究[J].岩石力学与工程学报,2005,24(10):1789-1793.
[9]特科特D L.分形与混沌[M].北京:地震出版社,1993.
[10]Prosperini N,Perugini D.Particle size distributions of some soilsfrom the Umbria Region[J].Geoderma,2008,145:185-195.
[11]TAO Gao-Liang,ZHANG Ji-Ru.Two categories of fractal models ofrock and soil expressing volume and size-distribution of pores andgrains[J].Chinese Science Bulletin,2009,54(23):4458-4467.
[12]刘志祥.深部开采高阶段尾砂充填体力学与非线性优化设计[D].长沙:中南大学资源与安全工程学院,2005.
[13]Mkrtchian T,Dight P.Experience in electroactivation backfill hard-ening[J].Australasian Institute of Mining and Metallurgy Publica-tion Series,1998(n1):333-335.
[14]李一帆,张建明,邓飞,等.深部采空区尾砂胶结充填体强度特性试验研究[J].岩土力学,2005,26(6):865-868.
[15]彭续承.充填理论及应用[M].长沙:中南工业大学出版社,1998.
[16]何玉斌,李新忠.神经网络控制技术及其应用[M].北京:科学出版社,2000.
[17]朱凯,王正林.精通MATLAB神经网络[M].北京:电子工业出版社,2010.
[18]郭海湘,诸克军,胡杰,等.GA-BP嵌套算法的理论及应用[J].数学的实践与认识,2008,38(1):116-125.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心