At present, aluminum plates are widely used in aerospace, machinery, and other fields because of their light weight, good manufacturability, and strong corrosion resistance. However, during long-term use, defects that will affect the structural strength such as cracks are easy to appear. Therefore, it is necessary to conduct automatic non-destructive testing of large metal plates to determine the location, size, and structural damage to their defects. Compared with manual detection, automated detection has many advantages such as high detection efficiency, high repeatability, strong flexibility, and low labor intensity. Based on the spinor theory, this paper uses the product of exponentials (POE) to model the kinematics of the Aubo-i5 manipulator and obtain the positive kinematics equations, which are compiled by analytic geometry and algebraic methods combined with Paden-Kahan subproblems and matrix theory. The inverse kinematics algorithm of the 6R manipulator is based on this model. An automated non-destructive testing system including a robotic arm and ultrasonic phased array testing equipment was built. At the same time, a robotic arm simulation platform was built based on ROS, and the inverse kinematics algorithm was used to control the robotic arm through the MoveIt!. The angle error can be controlled within 2.5°.

Aiming at the problem of inaccurate measurement results caused by factors such as gas leakage during the calibration of the standard meter gas flow standard device of the same type and small diameter, this paper proposes a similarity assessment method of single flow point energy consumption mapping based on Fréchet distance (FD). First, through the statistical analysis of the historical energy consumption data of gas flow standard device, a discrete standard energy consumption mapping is formed. Next, the energy consumption mapping to be tested is standardized and the similarity of the energy consumption mapping is determined by the binary matrix accelerated search of the FD between the energy consumption mapping to be tested and the standard energy consumption mapping. Finally, the similarity threshold is calculated by combining the standard deviation of the FD and the reliability of the gas flowmeter calibration process is evaluated accordingly. The test shows that the method in this paper can effectively determine the reliability of the gas flowmeters calibration results.

The translation of images is expected to resolve the issue of acquiring a substantial number of images. Aiming at the problems of unstable training and weak feature learning ability of image translation network model, a new Self-Attention Cycle Generative Adversarial Network (SA-CycleGAN) is proposed. The model is based on the CycleGAN framework, but with several differences. It introduces a self-attention module in the generator and a residual network structure to guide it to focus on the most discriminative region between the source and target domains. Concurrently, LPIPS loss is incorporated into the loss function to facilitate the training of the generator to produce higher-quality images. The experiment demonstrated the efficacy of this method in preserving the image background while accurately translating the colour and texture nuances present in the standard data set. Concurrently, the translated images exhibit significant advantages in terms of the KID index in comparison with alternative translation methodologies.