Search

Your search keyword '"Zhu, Wenfa"' showing total 28 results
Start Over Author "Zhu, Wenfa"
28 results on '"Zhu, Wenfa"'

6. Sparse TFM imaging of different-scale phased array based on the DWSO algorithm.

Author

Wei, Zhengbo, Zhu, Wenfa, Zhang, Hui, Chai, Xiaodong, Qi, Weiwei, Fan, Guopeng, and Zhang, Haiyan

Subjects
*PARTICLE swarm optimization, *ULTRASONIC arrays, *PHASED array antennas, *SPARSE matrices, *NONDESTRUCTIVE testing
Abstract

Using sparse matrix for total focusing method (TFM) imaging can improve computational efficiency in non-destructive testing (NDT). In sparse matrix design, binary particle swarm optimisation (BPSO) and genetic algorithm (GA) often fall into local optimal solution, and the computational efficiency decreases significantly with the expansion of array scale. In this paper, a discrete war strategy optimisation (DWSO) is proposed to realise sparse array ultrasonic imaging. This method uses equidistant scatter mapping to real-number encode the array position and limit the search range. Then, the fitness function is constructed with low side lobe peak and narrow main lobe width to obtain the optimal array element distribution. Experiments on standard test block demonstrate the DWSO algorithm has a narrower main lobe width and lower side lobe peaks than BPSO and GA. The imaging time of the sparse array constructed by this algorithm is reduced by more than 65% compared with the full array. As array scale grows, the running time variation of the proposed method is reduced by 73.2% and 77.3% compared with GA and BPSO, which has good adaptability and provides theoretical support for real-time defect detection. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

8. Sparse Optimal Design of an Ultrasonic Sensor Array for Fast TFM Based on a Discrete Slime Mold Algorithm

Author

Zhu, Wenfa, Wei, Zhengbo, Xiang, Yanxun, Chai, Xiaodong, Liu, Sihao, Fan, Guopeng, Zhang, Haiyan, Zhang, Hui, and Qi, Weiwei

Abstract

The total focusing method (TFM) is an ultrasonic phased array imaging algorithm used in ultrasonic nondestructive testing (NDT) that processes large amounts of data from full matrix capture (FMC). This limits its application in some industrial fields with real-time requirements. To solve this problem, a sparse array optimization method is applied to FMC-TFM that can reduce time consumption and improve imaging efficiency. However, conventional intelligent optimization methods, such as genetic algorithm (GA), use binary encoding, which require intensive computation and are easily trapped in local optima. This article proposes a discrete slime mold algorithm (DSMA), in which the slime mold position is coded in real numbers instead of binary. In the optimization process, a mapping model between the slime mold and ultrasonic array is established. A fitness function with a narrow main lobe and low sidelobe is constructed to obtain the sparse array position with the best performance. In experiments, the proposed method reduces the imaging time by more than 50% compared with conventional TFM, without affecting imaging quality. Compared with a GA and binary particle swarm optimization (BPSO), the proposed method improves array performance indicator (API) and signal-to-noise ratio (SNR) performance.

Published
2024
Full Text
View/download PDF

23. Research on calibration technology in rail abrasion detection based on machine vision

Author

Zhu Wenfa and Ma Huizhen

Subjects
Pixel, Computer science, Machine vision, business.industry, Coordinate system, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Astrophysics::Instrumentation and Methods for Astrophysics, Rail profile, Feature (computer vision), Camera auto-calibration, Computer Science::Computer Vision and Pattern Recognition, Calibration, Computer vision, Artificial intelligence, business, Camera resectioning
Abstract

When using machine vision system to detect rail abrasion, we need to calibrate camera and light plane in order to obtain geometrical relationship between camera and light plane along with coordinates of points on rail profile in world coordinate system. This paper uses camera calibration based on 2D planar calibration target to realize camera calibration. The experiment indicates that it has many advantages such as high accuracy, simple procedures, et al. Using indirect calibration method to conduct light plane calibration, this approach uses target curve with single-pixel width to get the pixel coordinate of feature points so as to reduce the effect on accuracy caused by the width of light.

Published
2015

24. Train positioning based on strap-down inertial system

Author

Hu Guo and Zhu Wenfa

Subjects
Inertial frame of reference, Computer science, Multiple integral, In vehicle, Average filter, Simulation, DC bias, Slip (vehicle dynamics), Beacon
Abstract

The positioning method based on beacon combination with speed sensor in vehicle is widely adopted in train operation. In this method, however, there is large erro due to the interval of two adjacent beacons and slip/slide of speed sensor. A train positioning method based on SINS technology used to calculate the running trajectory of train is present in this paper. In order to work out the trajectory in real-time, the average filter was utilized to remove the DC component of acceleration signals in SINS and the double integral filter based Simpson's rule was applied to transform the acceleration signals to displacement signals. Experiment shows that this method is capable of getting the trajectory of a running train with high precision and good repeatability.

Published
2013

Catalog

Books, media, physical & digital resources
See catalog results