基于泛函网络的周期来压预测方法研究

计算机科学 ›› 2013, Vol. 40 ›› Issue (Z6): 243-246.

基于泛函网络的周期来压预测方法研究

崔铁军,马云东

辽宁工程技术大学安全科学与工程学院阜新123000;大连交通大学辽宁省隧道与地下结构工程技术研究中心大连116028

出版日期:2018-11-16 发布日期:2018-11-16
基金资助:
本文受国家自然科学基金项目(51050003),辽宁省自然科学基金(201202022)资助

Prediction of Periodic Weighting Based on Optimized Functional Networks

CUI Tie-jun and MA Yun-dong

Online:2018-11-16 Published:2018-11-16

摘要/Abstract

摘要： 为预测周期来压,构建了基于小波和混沌优化的泛函网络(FN)预测方法。该方法利用小波分解技术将所选的样本集数据分解成不同频率的分量。基于混沌理论对分量相空间进行重构。各重构分量分别使用FN模型进行训练。最后,将各个FN模型得到的预测分量进行小波重组,得到完整的周期来压荷载预测波形。通过在重构时的计算发现,荷载的时序序列有一定的混沌性。通过模拟并与3种其它模型进行比较发现,基于小波和混沌优化FN的预测模型得到的最终周期来压荷载波的精度更高,收敛性也较好,但是,时间成本较大。

关键词: 周期来压预测,小波处理,混沌优化,泛函网络

Abstract: In order to improve the prediction the periodic weighting load,this paper proposed a new periodic weighting prediction method based on Functional networks(FN) optimized wavelet analysis and chaos．In the method,selected sample set data is decomposed using wavelet decomposition technique to obtain the different frequency components．The phase space of the components is reconstructed by chaos theory．The reconstructed components are respectively transmitted into FN model to carry on prediction．Finally,all the predicted components got by FN models are reconstructed by wavelet to get complete prediction waveform．The result of the calculation in the reconstruction shows that in a certain period,the load temporal sequence has some chaotic property．Through the comparison between the result of the method and three other models,the final load ware of the method is higher accuracy and convergence,but the cost of time is large.

Key words: Prediction of periodic weighting,Wavelet analysis,Chaos,Functional networks

崔铁军,马云东. 基于泛函网络的周期来压预测方法研究[J]. 计算机科学, 2013, 40(Z6): 243-246. https://doi.org/

CUI Tie-jun and MA Yun-dong. Prediction of Periodic Weighting Based on Optimized Functional Networks[J]. Computer Science, 2013, 40(Z6): 243-246. https://doi.org/

参考文献

[1] 贺超峰,华心祝,杨科,等．基于BP神经网络的工作面周期来压预测[J]．安徽理工大学学报:自然科学版,2012,2(1):59-63
[2] 潘一山等.深部开采覆岩应力变化规律模拟试验研究[J].中国地质灾害与防治学报,2011,22(3):61-66
[3] 刘正和,赵阳升,弓培林,等.大采高放顶煤工作面矿压显现规律的研究[J].太原理工大学学报,2011,42(5):524-527
[4] 许家林,鞠金峰.特大采高综采面关键层结构形态及其对矿压显现规律的研究[J].岩石力学与工程学报,2011,30(8):1547-1556
[5] 潘岳,顾士坦,戚云松．周期来压前受超前隆起分布荷载作用的坚硬顶板弯矩和挠度的解析解[J]．岩石力学与工程学报,2012,1(10):2053-2063
[6] 卢国志,汤建泉,宋振骐．传递岩梁周期裂断步距与周期来压步距差异分析[J]．岩土工程学报,2010,32(4):538-541
[7] Castello E.Functional networks[J].Neural Proc Lett,1998,7:151-159
[8] Castello E,Gutierrez J M.Nonlinear time series modeling andprediction using functional networks,extracting information masked by chaos[J].Phys,Lett A,1998,244:71-84
[9] Castello E,Cobo A,Gutierrez J M.Working with differentialfunctional and difference equations usingfunctional networks[J].Apply Math Model,1999,23:89-107
[10] Castillo E,Cobo A,Gutierrez J M.Functional networks with applications[M].[S.l.]:Kluwer Academic Publishers,1999
[11] 陈民铀,杨艳利,何为．等基于粒子群优化算法的电阻抗图像重建[J]．重庆大学学报,2011,34(1):82-87
[12] 席剑辉．混沌时间序列的长期预测方法研究[D]．大连:大连理工大学,2005
[13] 孙韩林,金跃辉,崔毅东,等.粗粒度网络流量的灰色模型预测[J].北京邮电大学学报,2010,3(1):7-11
[14] 谢高岗,闵应骅,张大方,等.一个基于实际测试的网络流量模型[J].计算机工程与科学,2001,3(5):51-54
[15] Kim H S,Eykholt R,Salas J D.Nonlinear dynamics,delaytimes,and embedding windows[J].Physica D,1999,7:48-60
[16] 李占英,王科俊,徐亮,等．混沌对角递归神经网络的船舶横摇预报方法[J]．控制与决策,2012,7(11):1681-1693

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed