摘要
研究目的:桩基的作用系提高地基承载力,有效地减少构筑物的沉降。桩基形式多样,能应用于各种不同的地层,可广泛地应用于对沉降有严格要求的建筑物、活载占总荷载比例较大的建(构)筑物和其他对基础有特殊要求的构筑物。桩基检测工作的质量和检测评定结果的可靠性是确保桩基工程的质量安全的重要环节。本文将桩基承载力的有关理论及试验数据与层次分析法和神经网络理论有效结合,建立预测单桩极限承载力神经网络模型,使桩基承载力的求解更加快捷、便利、准确。研究结论:(1)工程实例中站房桩基单桩竖向抗压静载试验达1 500 t属国内少有,成功的现场试验及准确的试验结果是进行试验数据与神经网络理论有效结合的前提条件;(2)人工智能神经网络在岩土工程领域的应用使许多复杂的、模糊的非线性问题得到有效解决,对该方面付诸更多的改进和完善及研究工作具有借鉴意义;(3)桩基承载力的有关理论及试验数据与层次分析法和神经网络理论有效结合是解决大吨位桩基承载力试验和计算工作的一种有效新途径;(4)本研究结论可为类似大吨位桩基承载力试验和计算工作提供一定的指导和参考。
Research purposes： The effect of pile foundation is improving effectively reducing the settlement of structures. Pile foundation the bearing capacity of foundation and is in a variety of forms, and it can be applied to a variety of different stratums, can be widely applied to buildings which have strict request for settlement, buildings in which live load is big proportion of the total load, and other structures with special requirements for foundations. The quality of pile foundation inspection work and the reliability of evaluation results is an important link to ensure the safety and the quality of the pile foundation engineering. This paper effectively integrates the theory of pile foundation bearing capacity and experimental data with analytic hierarchy process and neural network theory, establishes neural network model for prediction of pile ultimate bearing capacity, to make the solution of bearing capacity of pile more fast, convenient and accurate. Research conclusions： （ 1 ） In engineering stations, the test of single pile vertical static load up to 1 500 t is rare, so successful field tests and accurate test results is the prerequisite for the effective combination of test data and neural network theory. （2） The application of artificial neural networks in geotechnical engineering makes effective solution of complex and fuzzy nonlinear problems. It has valuable reference significance on putting on more improvement, perfection and research work in this field. （3） The effective combination of pile bearing capacity theories and experimental data with analytical hierarchy process and neural network theory is an effective new way to solve the test and calculation of large tonnage pile bearing capacity. （4） This research conclusion can provide guidance and reference for similar test and calculation of large tonnage pile bearing capacity.
出处
《铁道工程学报》
EI
北大核心
2016年第4期65-69,共5页
Journal of Railway Engineering Society
关键词
BP神经网络
桩基
承载力
BP neural network
pile foundation
bearing capacity
