To improve the ability of the deep learning model to handle imbalanced data, a fault diagnosis method based on improved gated convolutional neural network (IGCNN) is proposed. Firstly, an improved gated convolution layer is proposed for feature extraction, with the batch normalisation (BN) layer applied to adjust the data distribution and enhance the generalisation performance of the model. Then, the feature learned by multiple gated convolution layers and pooling layers is fed to the fully connected layer for fault type identification. Finally, the label-distribution-aware margin (LDAM) loss function is employed to adjust the model being more sensitive to the minority class and mitigate the influence of imbalanced data on the model. Experimental validation is conducted using two bearing datasets. Results show that the proposed method is more robust than other fault diagnosis methods, with higher recognition accuracy in severely imbalanced dataset.
To analyze the seismic response characteristics of unequal-span subway station structures in saturated sites, a three-dimensional numerical model of an unequal-span subway station structure is established, based on the finite element analysis software MIDAS-GTS. The elastic modulus, cohesion, Poisson’s ratio, and friction angle are selected as the sensitivity parameters. Moreover, combined with the saturated two-phase medium dynamic analysis method, the orthogonal test method is also utilized, to obtain the corresponding seismic response range. The results show that, the lower end of the shear wall and the vicinity of the cantilever span are prone to bending failure, and that the central columns are prone to shear failure. Under the action of a horizontal ground motion, or under the combination of horizontal and vertical ground motions, the influence of the elastic modulus is the largest, the influence of the Poisson’s ratio and the friction angle is the second largest, and the influence of cohesion is the smallest. This procedure of seismic response characteristics for unequal-span subway station structures can provide a reference for the seismic design of these structures.
以西安某拟建数据中心为例,开展地铁列车运行对数据中心机房设备的振动影响评价.以"振源-传播途径-受振体"为基线,采用实测与数值计算相结合的研究方法.首先,开展场地自由衰减特性测试,获取地铁列车振动沿场地的传播规律;其次,在拟建建筑物场地边界距离地铁线路最近处布置测点,获取过车时的振动响应;然后,建立拟建建筑物结构有限元模型,获取其主要动力特性;最后,以场地实测加速度作为输入,采用一致激励法进行结构的车致振动响应计算,并对建筑物内精密设备的振动影响进行预测与评价.研究表明:(1)地铁列车运行时该数据中心机房设备振动达标;(2)地铁列车振动主频集中在31.5 Hz～80 Hz;在31.5 Hz以下的中低频段,场地表现出明显的整体振动特点;在31.5 Hz以上的中高频段,地面高频振动分量随距离迅速衰减;(3)会车工况与近轨工况地面监测结果接近,说明近轨列车振动能级显著大于远轨,会车时振动能量较近轨列车增加不明显.研究思路可为此类结构的振动分析和减隔振设计提供参考.
以某型号离心风机为研究对象,保留相关结构细节特征,建立三维有限元模型.基于CFD(Computational Fluid Dynamics)软件采用雷诺时均法对其内部流场开展稳态计算,并采用声类比积分法对风机的气动噪声进行预测.通过气动性能以及辐射噪声的试验测试和仿真结果对比来验证仿真模型的有效性.由宽频噪声模型分析可知,离心风机噪声源位置主要集中于叶片以及蜗舌附近.考虑到工程用离心风机结构改型成本较高,对叶片进行穿孔二次设计,提出一种优化的穿孔方案,探讨叶片穿孔对离心风机气动性能以及噪声声压级的影响.研究结果表明,优化后的叶片穿孔设计可降低叶片表面的压力脉动进而减少离心风机气动噪声,所提出的设计方案可使风机噪声总声压级降低2.64 dB.
描述中国航发商用航空发动机有限公司(商发公司)新研制的大涵道比风扇声学试验器(Fan Aeroacoustic Test Facility,FATF)的初步设计方案和声学调试结果.FATF配备尺寸36 m×33 m×15 m、截止频率100 Hz的全消声室,是目前国内最大的风扇声学试验全消声室.FATF还配备进气整流罩(Turbulence Control Screen,TCS)、远场声阵列等试验设备,满足风扇远场噪声试验需求.对FATF全消声室自由声场、频率范围和背景噪声进行调试和校准,结果显示,全消声室的自由声场频率范围、背景噪声等技术指标均完全满足设计要求,证明FATF的全消声室技术指标均已达到国际先进水平,可满足后续风扇气动声学试验需求,支撑航空发动机风扇的降噪设计与改进优化.
