查看更多>>摘要：A partial overlap between adjacent strips during airborne Light Detection And Ranging (LiDAR) data scanning is required to ensure data integrity. The overlap area is observed two times and contains richer target details. To reduce data density yet retaining target details in the overlap area, a method based on reducing the influence of repeated observation data is presented. The proposed method defines the repeated observation data as the multiple samples of the same location from two adjacent strips, identifies them by locating the pairwise closest points from two adjacent strips with their distance below a distance threshold, and eliminates unimportant points inside the repeated observation data according to the criterion that LiDAR points with low curvature or high incidence angle can be removed without affecting the sharp features and the overall quality of data represen-tation of the overlap area. The optimal distance threshold is adaptively determined using a Gaussian fitting model by modeling the mean distance between all the nearest neighbors in one original LiDAR strips. Finally, reduction rate and information entropy metric are put forward to quantitatively evaluate the effectiveness of the proposed method. The developed method is applied to two real airborne LiDAR datasets with urban and forestry scenes for experimental testing. The quantitative accuracy evaluation results show that reduction rate in urban and forestry area can reach to 25.1% and 12.8%, respectively. Furthermore, the data density of the overlap area could be reduced while the information entropy and DTM accuracy still can be maintained.
查看更多>>摘要：Small diameter inner wall surface reflective linear bearing sleeve, in the manufacturing, transportation, and other processes. The inner wall surface will inevitably produce minor height difference surface defects, resulting in gaps or wear problems in the use of products, affecting the overall life of the equipment and even causing a significant security risk. However, auto-detection is challenging to achieve because of the small space during the inspection and comprehensive inspection requirements. At the same time, it is required to have high accuracy and fast efficiency during the inspection, etc., to avoid causing the increase of the industrial production cycle and limit the yield. Based on the physical characteristics of the light reflection on the inner wall surface of such parts, this paper proposes a high-dynamic online detection method for small-diameter internal wall surface defects. The reflective fringes are used to enhance and magnify the minor surface defects of the inner wall, and the new spiral fringes are derived by modelling and simulation. The practical reflection of the characteristics of micron defects on the reflective inner wall is realized through the establishment of a mechanical inspection environment, and the use of image processing technology, online rapid and comprehensive detection of micron defects is achieved. Application and production in the actual industrial environment are realized, and the online high precision detection of defects of phi 22mm is realized. Experiments prove that the method can detect bulges with only a height of 5 mu m, and the detection accuracy rate can reach 99.31%.
查看更多>>摘要：The light-screen is widely used as a sensor in the measurement of projectile flight parameters of rapid-fire weapons due to the high sensitivity and the fast response. The current measurement theory regards the lightscreen as an ideal plane. therefore, it is necessary to arrange at least four light-screens in space to form a detection array to measure the flight velocity and coordinates of the projectile. However, the actual light-screen is not an ideal plane but has a certain thickness, and the spatial shape distribution of light-screen thickness should follow a certain law. Based on the formation principle of light-screen, the spatial shape distribution of light-screen thickness is analyzed in this paper. According to the corresponding relationship between the trajectory position and the thickness distribution of the light-screen the characteristic difference of the LSPTS (Light-screen Passing Through Signal) is studied, then the measurement model is built and the measurement formula is deduced. Finally, the live firing experiment is carried out, and the results show that the proposed measurement method can realize the simultaneous measurement of the flight velocity and impact coordinates with only two light-screens, which reduces the number of light-screens by half compared with the current measurement methods. And the experiment result shows that the relative error of the impact coordinate measurement results is 8.1% on the abscissa axis and 8.2% on the ordinate axis, and the relative error of flight velocity measurement error is 3.1%. The research in this paper can provide a new idea for measuring the flight parameters of projectile by photoelectric principle.
查看更多>>摘要：High-resolution (24-bit) ADC based sampling devices e.g. NI PXI 5922 and NI PXI 4461 are usually used for precise electric power measurement purpose so it is required to characterize them in a proper way. According to that, several measurement techniques are proposed to characterize ac amplitude and channel characteristics at the maximum frequency of up to 100 kHz. An automated test and calibration system based on a planar multi junction thermal voltage converter and a modular electronic instrumentation platform has been developed. Furthermore, two ADCs, i.e. NI PXI 5922 and NI PXI 4461 have been compared with reference standard (Swerlein algorithm) to investigate their precision in RMS amplitude measurement of ac signals. In addition, this paper presents the PXI digitizers' channel characterization in both sinusoidal and non-sinusoidal regime as the main contribution which results in comprehensive insight into digitizers' performances and their mutual comparison. Accordingly, relevant uncertainty analysis is presented as well.
查看更多>>摘要：Detecting and classifying the melt pool states in laser-based direct energy deposition (L-DED) is crucial for reducing defects and enhancing the mechanical properties of L-DED metal parts. Although physics-based modeling methods and traditional machine learning algorithms such as convolutional neural network have been introduced to monitor the melt pool states, improving the low accuracy of these methods remains to be challenging. To address this issue, we developed a DenseNet-39 model to classify the melt pool states. 80 singletrack samples were fabricated using a linear scan strategy by L-DED and using a coaxial high-speed camera to capture the melt pool images in-process. Experimental results have demonstrated the superior performance of DenseNet-39 in classifying the melt pool states with 99.3% accuracy, achieved a lower computation burden, and less processing time. We used CAM to explain the mechanism of classification by DenseNet-39. DenseNet-39 provides the potential applications of online process monitoring in L-DED.
查看更多>>摘要：Time-resolved temperature measurement of dynamic-compressed material is critical in the high-pressure research fields. For higher temporal resolution requirement of the dynamic experiments, a broadband optical pyrometer was developed, with a rise time of similar to 70 ps and a measurable temperature range of 2500 K to higher than 10000 K. Expanding the detecting bandwidth to 100 nm similar to 300 nm for each channel, sensitivity of the detection system increases by nearly one order of magnitude compared to the conventional configuration. An improved temperature calibration procedure and data processing method was proposed, and temperature results of the shock compression experiments using a light gas gun were identical to the simultaneous measurements by a narrowband pyrometer. The dynamic radiation from the laser shocked matter also can be perfectly captured by the developed pyrometer, and the temporal profile of temperature agreed well with the shock velocity profile in both the shape and the duration time.
查看更多>>摘要：Varying working condition leads to the data distributions offset between training (source domain) and testing (target domain), which results in insufficient capability of traditional intelligent recognition methods for rotating machinery. Aiming at this problem, a novel intelligent recognition method based on adversarial domain adaptation convolutional neural network (ADACNN) is proposed for bearing faults. First, a weight-sharing convolutional neural network (CNN) is constructed to map training data from both source and target domains to a feature and predicted label spaces, respectively. Then, adversarial learning and maximum mean discrepancy (MMD) are separately introduced in the feature and predicted label space for domain adaptation, and the ADACNN model is established. Finally, a method based on ADACNN, which can realize domain adaptation in two different spaces, is proposed for intelligent recognition of bearing faults. The proposed method is validated by various cross-domain bearing fault recognition tasks under variable speeds and loads. Compared with shallow models, CNN, and commonly used domain adaptation methods, the proposed method has more than a 4% fault recognition accuracy advantage under varying working conditions.
查看更多>>摘要：Unmodeled errors in the global navigation satellite system (GNSS) observation equations are inevitable and have an impact on parameter estimation accuracy. In this study, we focus on inverting unmodeled errors in relative positioning and analyse their characteristics to provide a theoretical foundation for processing. First, we present an inversion method for unmodeled errors and demonstrate its effectiveness using an example. Then, we systematically examine the characteristics of unmodeled errors. The findings indicate that (1) the unmodeled errors in the short baselines are eliminated more effectively, whereas the unmodeled errors in the long baselines are more significant and unpredictable; (2) the long-term components (200-2500 epochs) in unmodeled errors are significant, but the short-term components (below 100 epochs) are not; and (3) unmodeled errors between L1 and L2 frequencies have a clear correlation, with the correlation coefficient being & GE; 75%. Finally, we prove that atmospheric delays are the primary source of unmodeled errors in long baselines.
查看更多>>摘要：This paper aims at the development of a method for monitoring oil debris as well as oil viscosity, using the optical measurement. An on-line optical oil debris sensor is designed firstly, then a multi-target tracking algorithm based on Gaussian Mixture Model is custom-developed to compute the properties (velocity, diameter) of wear debris in micro-flow channel through images captured by sensor. A mathematical relationship between debris velocity and debris diameter, debris density, oil flow, oil viscosity, oil temperature is statistically established using a three-level-five-factor box behnken design approach based on response surface methodology analysis. Then the analysis of variance is used to confirm the significance of the variables and their interactions. Finally, to validate the mathematical model, an experiment is carried out to compare its performance to that of results obtained using a pin disc test rig. The results demonstrate the measurement error of debris density and oil viscosity is 6.07% and 7.97%, respectively. The proposed approach has excellent suitability to monitor the oil condition.
查看更多>>摘要：Increasing the productivity of high-precision manufacturing requires efficient product inspection, for example, using integrated scanning systems on CNC machine tools (MT). However, measuring the dimensional and shape accuracy of workpieces using a touch probe (TP) on MT is negatively affected by the geometric accuracy setting of the MT, thermal stability of the MT and its surroundings, type of scanning system.The aim of this paper was to examine the possibilities of assessing the capability of touch probe measurements on MT, to define the size of the permissible controlled tolerance of the workpiece and to assess the influence of the geometric accuracy of the machine on the size of the permissible tolerance parameter of the inspected workpiece. In the presented experiments, the calculation of the TP measurement capability is based on a new approach respecting the VDA5 standard extended by the minimum tolerance (TOLmin) conformity assessment procedure according to ISO 14253-1:2017.
NSTL
Elsevier