Abstract:Given the critical role of high-speed transportation facilities, non-destructive testing technology is required in all aspects of their production, deployment, use and maintenance. We discuss the necessity of non-destructive testing of high-speed transportation facilities, the status of testing, the development trend of testing technology, including new detection technology based on multi-physics principle, material stress, microstructure, mechanical properties and other status testing, as well as non-destructive testing and structural health that integrate artificial intelligence. Our research will lay the foundation for non-destructive testing of high-speed transportation facilities.

Abstract:The structure of GH4169 is so complex that it is difficult to be evaluated accurately with a single ultrasonic parameter. However, the multi-parameter ultrasonic evaluation (MUE) method has a problem in selecting a reasonable parameter set. Therefore, a new full-parameter ultrasonic evaluation method is proposed to introduce quasi-circular mapping into evaluation. All ultrasonic parameters are projected into the space of a two-dimensional circle. A projection polygon is constructed, and the second-order features with global ultrasonic information are extracted. Then the high-order polynomial fitting is carried out with the grain size, and the evaluation problem is transformed into an optimization problem with minimum fitting error and quasi-circular mapping parameters as design variables. Finally, the grey wolf optimization is used to solve and obtain the final all-parameter ultrasonic evaluation model. The experimental results show that compared with other ultrasonic evaluation methods, the new method has the characteristics of high accuracy and good robustness.

Abstract:In view of the disadvantages of the traditional theodolite surveying method applied to verticality detection for construction tower crane, a global navigation satellite system (GNSS) based verticality intelligent detection technique in rounds for construction tower crane is proposed. And then, a verticality intelligent detection method in rounds for construction tower crane is designed based on GNSS dynamic detection model and the GNSS-based verticality intelligent detection system in rounds (GNSS_VDS) is designed and developed based on Visual Studio 2017 platform using C# program language. The preliminary experimental results show that the accuracy of intelligent detection in rounds of GNSS_VDS system is better than 3 cm in the horizontal direction and 4 cm in the vertical direction, verifying the effectiveness of the proposed algorithm. The high-precision intelligent solution can be provided for real-time monitoring of anti-toppling stability for construction tower crane.

Abstract:Defects may sometimes develop within the interface of a diffusion bonded dual-alloy blisk and the ultrasonic evaluation is the best method of controlling the interface defects. However, the acoustic impedance difference and complex geometry bring great difficulty to the accuracy and reliability of ultrasonic evaluation. An evaluation method is proposed in this paper, in which the ultrasonic signal is introduced into the inner hole wall of the blisk with a 90° acoustic mirror, realizing the ultrasonic evaluation of bonding interface of dual-alloy blisk. Experiments are conducted to study the evaluation capability of the proposed method for bonding interface of dual-alloy blisk in terms of sensitivity and SNR, etc. Results show that the evaluation sensitivity of the proposed method can reach a flat-bottomed hole equivalent of Ф2.0 mm, and the ultrasonic evaluation results are in good agreement with those of metallographic analysis, verifying the reliability of the proposed method.

Abstract:Aiming at time-varying feature identification of important brain regions in brain functional network based on fMRI-BOLD signal, a method of constructing time-varying observation model of dynamic brain network Rich-club is proposed in this paper. By clustering the similarity of Rich-club set of all time sampling points for one people, the temporal and spatial information of the brain function network is fused, and then the dynamic Rich-club importance evaluation model for brain network is built, which can describe the spatiotemporal comprehensive importance of the important brain regions quantitatively. Our work presents an effective method for the dynamic feature observation of the important brain regions in brain dynamic function network, and it also provides a basis technology for analyzing the differences of the important brain regions between healthy people and autistic people.

Abstract:In order to solve the difficulty of segmenting the linear guide surface defects from image with a complex background, a method based on gray level co-occurrence matrix (GLCM)and non-negative matrix factorization (NMF) to suppress the texture background to realize defect feature enhancement was proposed. Firstly, the GLCM multi-feature statistics was used to reconstruct the background texture map of the linear guide surface to achieve a certain degree of texture background suppression. Then, the texture was divided into several sub-image blocks, and a certain sub-image block was randomly selected for NMF dimension reduction. Next, the basic matrix decomposed by NMF was traversed by the same size image block in the texture map to find its Euclidean distance, and the averaged distance was assigned to the center pixel of the corresponding image block in the texture image to realize texture background suppression and features enhancement. Finally, the defects were classified based on K-means clustering and support vector machine. In the experiment, the recognition accuracy of scratches, cracks and crash defects in the test set are 89.06%，88.46% and 95.12%, which shows that the proposed method can suppress the texture background effectively and enhance the defect features of the linear guide surface image, and it can separate the defects and identify their types accurately.

Abstract:Rapid detection of pavement cracks is important for road maintenance and rehabilitation, but the traditional crack detection method is time-consuming, labor-intensive and low accuracy. Therefore, an improved U-net neural network model is proposed in this study. By adjusting the model structure and fine-tuning parameters, the U-net model can accurately and automatically identify pavement cracks. In this paper, a new semi-automatic marking software is developed to label pavement cracks based on Canny edge detection and Otsu segmentation algorithms, and the labeled 2D laser images are used as the training dataset. In addition, data enhancement methods are used to augment the training database. In the experimental stage, the cross-entropy loss function is used to compute error differences between the predicted value and the true value based on Adam optimization algorithm. Findings show that the improved U-net model is better than the original U-net model and the fully connected neural network model in terms of detection accuracy and algorithm robustness. This study provides a solution for the rapid detection of pavement diseases, which will be beneficial to road maintenance management department which can rapidly take corrective measures to ensure road traffic safety.

Abstract:Structural health monitoring technology can provide methods for flight test load assessment of aircraft active surface to verify the effectiveness of load design method and analysis data, and to provide the most direct data reference for safety verification and optimization of the active surface structure. The instability of traditional strain gauges in low-temperature environments may cause abnormal data collection. It makes the real-time monitoring and assurance of accuracy of flight load a difficult problem. This paper proposes a method using fiber Bragg grating (FBG) sensors based on stable low-temperature performance, for flight load evaluation of the connecting panels by establishing the strain-load relationship model under the tensile and compression load conditions of the structure via theoretical analysis, finite element simulation and tests, and provides a FBG sensor optimized layout method for strain and load assessment requirements of typical connecting panels of aircraft. The results show that the strain-load inversion method based on FBG sensors is effective and can meet the strain and load monitoring requirements of typical actuator connecting panels.

Abstract:In the topic of damage detection of composite structures based on lamb wave technology, damage index is commonly used for damage identification. However, its threshold is largely of expertise-dependence and poor performance at knowledge generalization. Therefore, a method based on the concept of least margin is proposed, which integrates even machine learning models and outputs the identification result by polling all models’ decision. The proposed method avoids the shortage that damage recognition relies on a single but incomprehensive model, and puts the confidence on a number of most qualified models instead. Significantly higher accuracy of damage identification for composite structures is manifested through test verification.

Abstract:In order to non-destructively evaluate the microstructure and strengthening state of GH738 forging products during the heat treatment and brazing production cycle, the GH738 alloy is heat-treated under solution, stabilizing, ageing and braze welding treatment conditions. Metallographic and ultrasonic tests are carried out. The metallographic structure and ultrasonic characteristics are established. The mapping relationship between signal parameters is detected, and the influence mechanism of tissue changes on the changes of ultrasound signal parameters is analyzed. Results show that the effect of grain diameter on sound velocity is weak. During the hot working process, the sound velocity decreases after solution treatment, and the sound velocity continues to increase after stabilization and aging treatments. After the brazing process, the sound velocity starts to decrease and get minimum. The attenuation coefficient and non-linear coefficient change with the grain diameter. Both of them rise to the stabilization process and then slowly rise to flatten out. That means ultrasonic attenuation and non-linear coefficients can be used to characterize the grain diameter and texture discontinuity of GH738 alloy.

Abstract:Aiming at the problems of selective laser melting (SLM) processing of GH3536 alloy, the difference in acoustic impedance between the unfused small defect and the intact area is small, and the ultrasonic echo defect defect is not obvious,this paper presents a signal processing method based on variational modal decomposition. Ultrasonic testing is performed on different micro-sized defect specimens, and the db7 wavelet function is used to decompose the signal into three layers. Then, the minimum and minimum （Minimaxi）rules are used as thresholds to perform hard threshold noise reduction and reconstruct the signal to extract defect echoes in the signal mutation points to identify defects. Results show that the method has obvious noise reduction effect, can detect the smallest detectable defect size of 0.15 mm and significantly improves the sensitivity of defect detection.

Abstract:The application of machine learning method based on ultrasonic signals in tensile and shear performance evaluation of titanium steel explosive composite bars is studied. This paper proposes a probabilistic neural network (PNN) evaluation and classification method based on the eigenvalues of ultrasound signals. Firstly, 120 samples of workpiece are taken as the object to obtain the full sequence A-scan signals of water immersion ultrasonic testing. The signals are analyzed in time domain and improved covariance power spectral density estimation. Six characteristic values are used as PNN input: depth of the composite layer, reflection frequency of the upper composite layer, spectral energy, reflection frequency of the lower composite layer, spectral energy, and attenuation of the secondary reflected wave on the lower surface. Then, a tensile test is performed on the workpiece sample to obtain the tensile and shear strength values as the PNN output. Finally, a classification training model is established based on 96 sample characteristic signals and tensile and shear strength values. The remaining 24 samples are used as the test set, and the tensile and shear strength values of these samples are classified and predicted. Experimental results show that the accuracy rate of 24 consecutive predictions is 94.35%. This article finds new ideas for the fast and full coverage evaluation of the tensile and shear properties of titanium steel explosive composite bars.

Abstract:Node localization is one of the core technologies in wireless sensor network（WSN）. Different localization methods have different effects on localization results. In response to the deficiency in accuracy of localization during the process of searching for the unknown node localization with the least-square algorithm, a localization algorithm is proposed based on improved cuckoo search（CS） algorithm. The first step is to establish the mathematical model according to the optimization objective, and then we design the fitness function of cuckoo search algorithm. Finally, we modify the parameters of the step length and rejection probability and thus determine the coordinate position of the unknown nodes quickly. The simulation results show that the proposed algorithm is better than the distance vector hop(DV-HOP) and the self-adaption cuckoo search and distance vector hop(SACSDV-HOP) algorithms. The algorithm can effectively reduce the error of localization in WSN and improve the accuracy in locating nodes. Accordingly, it has high practicability in WSN node localization.

Abstract:The use of full digital coding in millimeter-wave massive MIMO systems requires a large number of radio frequency(RF) chains, resulting in excessive energy consumption. A beam selection scheme based on discrete cuckoo algorithm (DCS) is proposed to reduce the number of required RF chains without obvious performance loss. Firstly, the beam selection model of millimeter wave massive MIMO system is analyzed, and the DCS algorithm is used to solve the model. Owing to the the abnormal coding in the Levy flight discretization result,the heuristic greedy algorithm is proposed to repair it. To speed up the convergence of the algorithm, the replication in the genetic algorithm is introduced into the DCS algorithm, and the global optimal nest is copied to replace the discovered nests. The simulation result shows that the proposed beam selection scheme based on the improved DCS algorithm can obtain better rate performance than several existing schemes.

Abstract:A new radio frequency(RF) fingerprint extraction method is proposed for Airmax devices with proprietary protocols. Firstly, the construction of software and hardware experimental environment and the Airmax technology are introduced. Then the extraction method of the frame preamble signals is introduced, which is divided into two rough positioning and precise position. And the extraction method of Airmax RF fingerprint is expounded from theoretical analysis and experimental verification. A total of 14 dimensional features are extracted, in which 2 features are related to the frequency and 12 features are related to the amplitude. Finally, based on the 14-dimensional features, the K-means algorithm and the decision tree model are used to train and classify the data of features, and the classification precision is calculated. The precision of both models reach 100%. For the classification problem of four devices, the precisions of the K-means algorithm and the decision tree model are 92.4% and 100%, respectively.

Abstract:To deal with the limitation of boundary effect and improve tracking speed in the traditional object tracking based on correlation filter, a fast discriminative scale space with background-aware correlation filters (BACF)algorithm is proposed. The BACF algorithm effectively solves the boundary effect caused by cyclic shift, greatly increases the quality and quantity of training samples, and thus improves the performance of the tracker. However, because of its disadvantage in scale detection strategy, it seriously affects its tracking speed. The we design a one-dimensional scale filter. The scale filter and translation filter are trained and optimized independently. The proposed algorithm greatly improves tracking speed while ensuring scale and translation estimation. The experimental results show that compared with BACF algorithm, the proposed algorithm can improve the tracking speed by about 75% without losing the tracking accuracy.

Abstract:Focused on the problem that the signal sparsity is difficult to be predicted accurately in actual electromagnetic environment, we propose a distributed modulated wideband converter (DMWC) reconstruction algorithm based on regularized weak correlation, which does not rely on the sparsity as convergence condition. First, the atoms that satisfy the weak correlation are added to the index set. Then, the index set is regularized and the newly selected atoms are added to the support set. When the residual energy reaches the threshold condition, the iteration is stopped. Finally, the support set out-of-bounds condition is set, and the invalid atoms with less correlation are deleted to obtain the final support set. Simulation results show that the proposed algorithm can greatly improve the tolerance of DMWC to signal transmission attenuation. In addition, under the same conditions, the recovery performance of this algorithm is better than orthogonal matching pursuit (OMP) algorithm.

Abstract:The existing bidimensional empirical mode decomposition (BEMD) algorithms are inefficient in extrema searching, intrinsic mode functions sifting and iteration processing, moreover the adaptability needs to be further improved. Therefore, this paper proposes an improved BEMD method based on multi-scale extrema. Firstly, the concept and establishment method of bidimensional multi-scale local extrema binary tree are given, and then a new approach based on multi-scale extrema is presented to determine window sizes for order-statistics and smoothing filters. This method significantly improves the adaptability of multi-scale decomposition of two-dimensional signals, and also significantly improves the decomposition efficiency. Experimental results of natural image and synthetic texture image decomposition show that the proposed method has obvious advantages in adaptability and efficiency compared with the existing fast adaptive EBMD method.

Abstract:Three-dimensional laser scanning is a new technology for fast acquisition of high-precision point clouds. However, due to the influence of the structure, roughness, texture of the object itself and measurement environment, the acquired point clouds mostly have isolated noise points. In view of the difficulty of removing complex noise in the point cloud data model of cultural relics, a denoising method for point cloud with geometric feature preservation is proposed. Firstly, the large-scale noise is deleted by rasterizing the point cloud, then the curvature factor and density factor of the data points in the point cloud are defined, and the clustering objective function of fuzzy C-means clustering (FCM) is constructed by weighting the factors. Finally, the small-scale noise is deleted by using the feature-weighted FCM algorithm, thus the denoising of point cloud is realized. The experimental results show that the denoising method for point cloud with geometric feature preservation has good denoising effect on cultural relics point cloud data, which is an effective point cloud denoising algorithm.