Deep convolutional neural networks (CNNs) have presented amazing performance in the task of semantic segmentation. However, the network model is complex, the training time is prolonged, the semantic segmentation accuracy is not high and the real-time performance is not good, so it is difficult to be directly used in the semantic segmentation of road environment images of autonomous vehicles. As one of the three models of deep learning, the auto-encoder (AE) has powerful data learning and feature extracting capabilities from the raw data itself. In this study, the network architecture of auto-encoder and convolutional auto-encoder (CAE) is improved, supervised learning auto-encoder and improved convolutional auto-encoder are proposed, and a hybrid convolutional auto-encoder model is constructed by combining them. It can extract low-dimensional abstract features of road environment images by using convolution layers and pooling layers in front of the network, and then supervised learning auto-encoder are used to enhance and express semantic segmentation features, and finally de-convolution layers and un-pooling layers are used to generate semantic segmentation results. The hybrid convolutional auto-encoder model proposed in this paper not only contains encoding and decoding parts which are used in the common semantic segmentation models, but also adds semantic feature enhancing and representing parts, so that the network which has fewer convolutional and pooling layers can still achieve better semantic segmentation effects. Compared to the semantic segmentation based on convolutional neural networks, the hybrid convolutional auto-encoder has fewer network layers, fewer network parameters, and simpler network training. We evaluated our proposed method on Camvid and Cityscapes, which are standard benchmarks for semantic segmentation, and it proved to have a better semantic segmentation effect and good real-time performance.

Recognition of safety helmets wearing by construction workers is a common target detection topic in applications of deep learning-based image processing. This paper provides a study of an enhanced YOLOv5-based method, in which the challenges caused by complicated construction environment backgrounds, dense targets, and the irregular shape of safety helmets are addressed. In a trunk network, feature extraction is more based on the target shape by using the Deformable Convolution Net instead of the conventional convolution; in the Neck, a Convolutional Block Attention Module is introduced to weaken feature extraction of complex backgrounds by giving weights to enhance the characterization ability of target features; and the original network’s Generalized Intersection over Union Loss is replaced by Distance Intersection over Union Loss to overcome the problem of erroneous location when the population is dense. The dataset for the training network is created by mixing open-source datasets with autonomous collecting to evaluate the effectiveness of the algorithm. We observed that the improved model has a detection accuracy of 91.6%, up 2.3% over the original network model, and a detection speed of 29 frames per second, which is compliant with most security cameras’ capture frame rate.

为提高现代设备维修的效率和节省维修成本,提出了以可靠性为中心的维修策略.针对现代设备涉及多学科领域、技术密集及复杂程度高的特点,阐述了以可靠性为中心的维修策略,按照该维修策略实施的7大逻辑步骤,制定了针对现代设备维修的实施方案,制作出维修逻辑决策图和科学、完整的维修大纲.以南京某数控机床为例,用可靠性为中心的维修策略进行维修,结果表明,该策略可以减少维修时间,降低维修成本,对于提高企业的经济效益具有实际意义.

The wavelet transform theory is used to motor fault diagnosis in this paper, considering its characteristics of multi-resolution and stronger feature extraction ability than Fourier. The paper emphasizes de-noising and eliminating the singular value point of the wavelet transform in the non-stationary signal. And it makes a detailed and in-depth analysis about how to detect the frequency components of weak signal by using equivalent power spectrum of reconstruction signal, which is acquired by using the wavelet transform. Through the comparison analysis of the simulation signal and motor vibration signal’s experimental data, the corresponding energy of original signal’s equivalent power spectrum and reconstructing signal’s equivalent power spectrum are compared to determine the fault frequency, so as to accurately find out the motor fault. DOI: http://dx.doi.org/10.11591/telkomnika.v11i9.3288

This paper generally focuses on the study of Kinematics analysis and simulation of six degree of freedom rotational joint(6R) manipulator. To solve positive and inverse kinematics problems, manipulator D-H mathematical model is established. For 6R manipulator to achieve the desired trajectory, this paper strives to provide and prove a general elliptic curve in the space to the end of the manipulator’s trajectory. By applying this method, we can extrapolate, at the end of the manipulator can reach within the scope of those of more complex curve, moreover, we can identify the arbitrary trajectories in any given situations as well. For 6R manipulator trajectory planning, the authors choose to put into use the cubic polynomial interpolation method in the joint space trajectory planning. This paper makes a simulation effort on 6R manipulator movement by using Adams, and got the expected results, which verifies the correctness of this method.

Based on the development of National Instruments Company LabVIEW graphical programming development software platform,design turbine typical fault monitoring and diagnosis system,such as the turbine pressure pulsation,vibration and pendulum degree and cavitations erosion of software research system,with the help of LabVIEW powerful data acquisition analysis processing function,the paper designed the signal acquisition,signal analysis,signal processing,the results display and diagnosis signal of the whole system,established research signal spectrum galleries and met the demand of analysis of the diversity of research signal.The system can continuously online monitoring state of the characteristics of turbine signal,and real-time show detected signals change processes,according to the acquisition signal data research combined with the turbine failure mechanism and characters,quickly judge the faults of the type and parts,and then provide reliable basis and the necessary information for maintenance,and then save the enterprise cost and improve enterprise benefit.

In order to diagnose the motor fault accurately, a new diagnosis method of motor fault was proposed which is based on AR parameter model spectrum estimation theory. As to real-time acquisition motor signals, energy signals were firstly got from acquisition signals and then were extracted from new samples at 10 sampling interval, and finally AR parameter model spectrum estimation was done on the small sample signals, when the data were relatively short, AR parameter model spectrum estimation can also present signals contained all frequency components. Based on the different frequency characteristics of motor signals on the power spectra, the diagnosis cause of motor signals can be determined. The implementation of the theory can well describe the power spectrum diagram of the signals which has the characteristics of smooth spectral lines, sharp spectral peak and accurate frequency location; it also can improve the resolution of spectrum estimation and facticity.