Piled embankments,which offer many advantages,are increasingly popular in construction of high-speed railways in China.Although the performance of piled embankment under static loading is well-known,the behavior under...Piled embankments,which offer many advantages,are increasingly popular in construction of high-speed railways in China.Although the performance of piled embankment under static loading is well-known,the behavior under the dynamic train load of a high-speed railway is not yet understood.In light of this,a heavily instrumented piled embankment model was set up,and a model test was carried out,in which a servo-hydraulic actuator outputting M-shaped waves was adopted to simulate the process of a running train.Earth pressure,settlement,strain in the geogrid and pile and excess pore water pressure were measured.The results show that the soil arching height under the dynamic train load of a high-speed railway is shorter than under static loading.The growth trend for accumulated settlement slowed down after long-term vibration although there was still a tendency for it to increase.Accumulated geogrid strain has an increasing tendency after long-term vibration.The closer the embankment edge,the greater the geogrid strain over the subsoil.Strains in the pile were smaller under dynamic train loads,and their distribution was different from that under static loading.At the same elevation,excess pore water pressure under the track slab was greater than that under the embankment shoulder.展开更多
To study ice-induced vibration of a compliant conical structure, a series of model tests were performed from 2004 to 2005. In the tests, the ice sheet before the compliant conical structure was found to fail in two-ti...To study ice-induced vibration of a compliant conical structure, a series of model tests were performed from 2004 to 2005. In the tests, the ice sheet before the compliant conical structure was found to fail in two-time breaking. From 2005 to 2006, this type of ice failure was studied through more groups of tests. The tests show that two-time breaking is the typical failure of ice before steep conical structures, and is controlled by other factors at the same time, such as ice speed and the angle of the cone.展开更多
In this paper, the maneuvering characteristics of a low speed submersible are investigated. First, the captive model tests are carried out to obtain the hydrodynamic forces acting on the submersible using a Planar Mot...In this paper, the maneuvering characteristics of a low speed submersible are investigated. First, the captive model tests are carried out to obtain the hydrodynamic forces acting on the submersible using a Planar Motion Mechanism (PMM). For the hydrodynamic forces within a wide range of attack angles, the hydrodynamic coefficients which are usually used in the conventional maneuvering motion are quite difficult to be applied. In this case, a Fourier series is adopted to represent the hydrodynamic forces and it fits the experimental data well. Then, based on the experimental results the simulation calculations are made to predict some of the maneuvering performance of the low speed submersible.展开更多
In order to improve the dynamic response bandwidth of three-stage electro-hydraulic servo valve,a new method,speed-feedback control is presented in this paper.The construction and principle of three-stage electro-hydr...In order to improve the dynamic response bandwidth of three-stage electro-hydraulic servo valve,a new method,speed-feedback control is presented in this paper.The construction and principle of three-stage electro-hydraulic servo valve are explained,and the mathematical model of three-stage electro-hydraulic servo valve is built in frequency domain.Experimental and simulation results show that the bandwidth compared with proportional control is improved under speed-feedback control.Moreover,the research results play an important role in developing high performance three-stage electro-hydraulic servo valve.展开更多
准确识别车辆当前驾驶行为模式是自动驾驶领域亟待解决的技术问题。为实现驾驶行为模式精准解析,提高模型识别精度和可靠性,通过开展自然驾驶试验,采集高速工况下20名驾驶人的驾驶行为数据及视觉特性数据等多源参数信息,分析4类典型驾...准确识别车辆当前驾驶行为模式是自动驾驶领域亟待解决的技术问题。为实现驾驶行为模式精准解析,提高模型识别精度和可靠性,通过开展自然驾驶试验,采集高速工况下20名驾驶人的驾驶行为数据及视觉特性数据等多源参数信息,分析4类典型驾驶行为模式(自由行驶、跟车、左换道、右换道)运行规律及多源参数耦合特性。基于主成分分析法确定4类驾驶行为模式表征指标集,使用支持向量机、随机森林决策树算法建立驾驶行为模式识别模型,通过学习训练,分析比较模型识别结果,对识别效果较好的模型进一步优化,分析优化模型对4类驾驶行为模式识别精度的时序性变化特征。研究结果表明:支持向量机模型、随机森林决策树模型、基于多层感知器神经网络的随机森林优化模型总体识别精度分别为89.4%、90.5%、91.9%;4类驾驶行为模式的AUC(area under the curve)值均高于0.93,可较好地识别车辆当前驾驶行为模式。此外,随机森林优化模型对4类驾驶行为模式的识别精度均随时间推移,呈现先增长后趋于稳定的变化态势,且同一时刻的自由行驶及跟车模式识别精度高于向左及向右换道模式。研究结果可为高级别自动驾驶系统决策及控制策略的制定提供理论基础和技术支持。展开更多
基金National Natural Science Foundation of China under Grant Nos.51622803,51378177 and 51420105013the 111 Project under Grant No.B13024
文摘Piled embankments,which offer many advantages,are increasingly popular in construction of high-speed railways in China.Although the performance of piled embankment under static loading is well-known,the behavior under the dynamic train load of a high-speed railway is not yet understood.In light of this,a heavily instrumented piled embankment model was set up,and a model test was carried out,in which a servo-hydraulic actuator outputting M-shaped waves was adopted to simulate the process of a running train.Earth pressure,settlement,strain in the geogrid and pile and excess pore water pressure were measured.The results show that the soil arching height under the dynamic train load of a high-speed railway is shorter than under static loading.The growth trend for accumulated settlement slowed down after long-term vibration although there was still a tendency for it to increase.Accumulated geogrid strain has an increasing tendency after long-term vibration.The closer the embankment edge,the greater the geogrid strain over the subsoil.Strains in the pile were smaller under dynamic train loads,and their distribution was different from that under static loading.At the same elevation,excess pore water pressure under the track slab was greater than that under the embankment shoulder.
基金This project is financially supported by the National High Technology Research Development Program of China(863Program) by the National Natural Science Foundation of China(Grant No.50609015)
文摘To study ice-induced vibration of a compliant conical structure, a series of model tests were performed from 2004 to 2005. In the tests, the ice sheet before the compliant conical structure was found to fail in two-time breaking. From 2005 to 2006, this type of ice failure was studied through more groups of tests. The tests show that two-time breaking is the typical failure of ice before steep conical structures, and is controlled by other factors at the same time, such as ice speed and the angle of the cone.
基金Pre-Research Found of Shipbuilding Industry of China
文摘In this paper, the maneuvering characteristics of a low speed submersible are investigated. First, the captive model tests are carried out to obtain the hydrodynamic forces acting on the submersible using a Planar Motion Mechanism (PMM). For the hydrodynamic forces within a wide range of attack angles, the hydrodynamic coefficients which are usually used in the conventional maneuvering motion are quite difficult to be applied. In this case, a Fourier series is adopted to represent the hydrodynamic forces and it fits the experimental data well. Then, based on the experimental results the simulation calculations are made to predict some of the maneuvering performance of the low speed submersible.
基金Ministry of Science and Technology of China(No.2006BAF01B12-03)
文摘In order to improve the dynamic response bandwidth of three-stage electro-hydraulic servo valve,a new method,speed-feedback control is presented in this paper.The construction and principle of three-stage electro-hydraulic servo valve are explained,and the mathematical model of three-stage electro-hydraulic servo valve is built in frequency domain.Experimental and simulation results show that the bandwidth compared with proportional control is improved under speed-feedback control.Moreover,the research results play an important role in developing high performance three-stage electro-hydraulic servo valve.
文摘准确识别车辆当前驾驶行为模式是自动驾驶领域亟待解决的技术问题。为实现驾驶行为模式精准解析,提高模型识别精度和可靠性,通过开展自然驾驶试验,采集高速工况下20名驾驶人的驾驶行为数据及视觉特性数据等多源参数信息,分析4类典型驾驶行为模式(自由行驶、跟车、左换道、右换道)运行规律及多源参数耦合特性。基于主成分分析法确定4类驾驶行为模式表征指标集,使用支持向量机、随机森林决策树算法建立驾驶行为模式识别模型,通过学习训练,分析比较模型识别结果,对识别效果较好的模型进一步优化,分析优化模型对4类驾驶行为模式识别精度的时序性变化特征。研究结果表明:支持向量机模型、随机森林决策树模型、基于多层感知器神经网络的随机森林优化模型总体识别精度分别为89.4%、90.5%、91.9%;4类驾驶行为模式的AUC(area under the curve)值均高于0.93,可较好地识别车辆当前驾驶行为模式。此外,随机森林优化模型对4类驾驶行为模式的识别精度均随时间推移,呈现先增长后趋于稳定的变化态势,且同一时刻的自由行驶及跟车模式识别精度高于向左及向右换道模式。研究结果可为高级别自动驾驶系统决策及控制策略的制定提供理论基础和技术支持。
