Improved calibration of simulation models in railway dynamics: application of a parameter identification process to the multi-body model of a TGV train

This paper aims at estimating the vehicle suspension parameters of a TGV (Train à Grande Vitesse) train from measurement data. A better knowledge of these parameters is required for virtual certification or condition monitoring applications. The estimation of the parameter values is performed by minimising a misfit function describing the distance between the measured and the simulated vehicle response. Due to the unsteady excitation from the real track irregularities and nonlinear effects in the vehicle behaviour, the misfit function is defined in the time domain using a least squares estimation. Then an optimisation algorithm is applied in order to find the best parameter values within the defined constraints. The complexity of the solution surface with many local minima requires the use of global optimisation methods. The results show that the model can be improved by this approach providing a response of the simulation model closer to the measurements.     

海水源热泵技术在港口建设中的应用

结合海水源热泵在大连港的应用情况介绍其技术原理、设备选型及其与传统空调系统相比较的优越性。     

流场中周期环肋圆柱壳辐射声压的理论和实验研究

对流场中一无限长周期环肋圆柱壳受激振动的辐射声压进行了理论和实验研究。分别采用Flugge方程和Helmholtz方程建立壳体及声场的波运动方程,考虑肋骨的4种作用力和力矩,利用周期理论求得壳体的振动响应,利用耦合边界条件最终可以求得无限长周期环肋圆柱壳的声场压力,并同实验结果作了比较。     

基于改进稀疏正则化的扩展卡尔曼滤波损伤识别研究

基于传统扩展卡尔曼滤波方法（EKF）的损伤识别过程是一个典型的反问题求解，反问题的不适定性使EKF识别结果容易受噪声干扰，导致EKF算法收敛困难、识别精度下降。基于l₁正则化的EKF算法在一定程度上可以缓解此问题，但其不能获得足够准确的稀疏解。为此，提出了一种改进稀疏正则化的EKF方法。该方法采用Arctangent罚函数代替l₁范数以施加损伤稀疏性约束；通过伪测量技术将稀疏性约束嵌入到EKF中，获得施加约束方程后的递推解。采用3层剪切结构和悬臂梁2个试验算例验证了该方法的有效性。研究结果表明：即使在噪声干扰下，所提方法也可以准确识别出结构损伤；相比传统l₁正则化EKF方法，该方法需要的观测信息更少，能获得更准确的稀疏解。     

基于相控阵列的噪声源识别及仿真研究

介绍了基于麦克风阵列进行噪声源识别的基本原理,采用相位延迟对噪声源空间进行扫描,确定声源的波达方向,进而识别声源的位置。利用计算机仿真技术模拟了空间两点声源产生的噪声,并利用该算法进行了声源位置识别。理论与仿真研究表明采用相控阵列能够识别噪声源的位置,同时能够获得不同噪声源的噪声大小和噪声贡献量,利用该方法可以对列车的运行噪声源进行分析。     

摆式列车倾摆伺服控制系统辨识方法仿真研究

针对摆式列车倾摆伺服控制系统的原理及特点,应用不同的辨识方法对倾摆伺服控制系统进行辨识,并对各种方法辨识结果进行比较。仿真试验结果表明:当考虑倾摆控制系统的非线性时,神经网络方法能很好地辨识出倾摆伺服控制系统的动态特性,解决倾摆伺服控制系统辨识中的非线性问题,而其它方法较差。     

频率法测量索力中的参数分析与索抗弯刚度的识别

通过有限元法及样条拟合技术获取斜拉索索力与频率的对应关系,并讨论这些关系在频率法测量索力中的应用。在此基础上,以崖门大桥斜拉桥索力测试为研究背景详细分析了频率法测量索力中诸如索的抗弯刚度、支座条件和斜度等参数对索力测试精度的影响,并提出一种基于多阶频率测试结果的斜拉索抗弯刚度识别方法。     

基于一维卷积神经网络的驾驶人身份识别方法

近年来，智能网联汽车（ICV）已成为智能工业时代最有前景的发展方向。作为现代移动的重要模式，ICV的设计和开发越来越强调个性化需求。提出一种仅使用车载CAN总线行车状态数据，基于深度学习的驾驶人身份识别通用框架。首先采集20名驾驶人在固定试验路线下，包括不同道路类型、不同交通条件下的自然驾驶行车状态数据集；其次对9种类型的CAN信号行车数据进行数据清洗与重采样，构建数据样本集。搭建了由卷积层、池化层、全连接层、SoftMax层构成的一维卷积神经网络（1-D CNN）驾驶人身份识别模型，并且使用Adam算法、L2正则化、Dropout、小批量梯度下降等方法对模型性能进行优化。算法验证过程中，探讨了模型卷积核占比、卷积核数量、卷积层层数、全连接层节点规模对模型识别准确率的影响，进而对模型结构参数进行优选。进一步地，将该算法与K近邻（KNN）、支持向量机（SVM）、多层感知器（MLP）等传统机器学习方法及深度学习算法长短时记忆网络（LSTM）进行对比分析，同时探究样本时间窗口大小、样本数据重叠度、驾驶人数量对模型识别结果的影响。在数据时间窗口为1 s、数据重合度80%的条件下，对20名驾驶人进行识别，评价指标宏观F₁分数可达99.1%，表明该模型表现明显优于其他对比模型算法，其对驾驶人身份识别表现稳定，鲁棒性强。     

Vehicle identification sensors location problem for large networks

Finding the optimal location for sensors is a key problem in flow estimation. There are several location models that have been developed recently for vehicle identification (ID) sensors. However, these location models cannot be applied to large networks because there are many constraints and integer variables. Based on a property of the location problem for vehicle ID sensors, given the initial vehicle ID sensors that are pre-installed and fixed on the network, this article presents a solution that greatly reduces the size of this location problem. An applied example demonstrates that when 8% of the arcs from a real network that are randomly selected have a vehicle ID sensor, the reductions are as large as 97% for the number of remaining constraints in the location model and 84% for the adjusted diameter of the feasible region of target flow. Using these two indices as target functions, two greedy algorithms are presented for solving the vehicle ID sensor location problem. These two algorithms were applied to an example in Mashhad city with 2,526 arcs, 7,157 origin-destination pairs and 121,627 paths. Using these algorithms, installing vehicle ID sensors on 8% of the network arcs results in satisfaction of 99.82% of the constraints in the location model and 97.6% reduction in the adjusted maximum possible error index. This means that deploying a low number of vehicle ID sensors on a real large network, with these greedy algorithms, yields a high level of observability.     

消磁脉冲交流发电机定子电流的数值计算

论文首先阐述了基于同步发电机不控整流的消磁脉冲电源的原理及组成,接着研究了消磁工况下同步发电机的工作特性,求得了同步发电机定子电流的解析表达式,并采用数值逼近法计算出了定子电流瞬时值及相关参数,最后某消磁站的实验数据表明,理论计算结果和实验测量值相符,证明了数值计算方法的正确性。