Detection of Defects in Adhesive Coating Based on Machine Vision
Dong Guotao, Hanjun Gao, Qiong Wu*, et.al.

• A machine vision detection method for detecting adhesive coating defects after lining spraying has been proposed.
• The problem of coating integrity under high aspect ratio and unstable lighting conditions has been effectively resolved .
• In the design of the hardware system, full consideration has been given to the deflection compensation of the detection equipment

Expanding the Defect Image Dataset of Composite Material Coating with Enhanced Image-to-image Translation
Dong Guotao, Hanjun Gao*, Kai Yang, Qiong Wu.

• A method named E-Pix2Pix has been proposed to enhance the defect image dataset of composite coatings in solid rocket engines.
• The depth research has been conducted on the influence of parameters during the image generation process.
• The impact of generating image content on classification models has been studied, and the potential negative impact is discussed.

Formation Mechanism and Evolution of Interfacial Residual Stress between Composite Propellant and Rubber Liner during Curing Process
Hanjun Gao, Dong Guotao , Hao Yuan, et.al.

• A finite element model has been established to simulate the stress changes throughout the curing process.
• A vision based device has been designed to monitor deformation changes during the curing process.
• The detected deformation is combined with the simulation results to optimize the key parameters of the simulation in order to achieve accurate stress assessment.

Analyze the Influence of Carburized Layer Thickness and Assembly Error on the Reliability of Thin-walled Gear Ring System based on ABAQUS
Dong Guotao , Jing Guo, Minghui Lin, et al.

• Parametric modeling technology to establish a three-dimensional model for the gear ring system
• Finite element analysis and solution of various assembly errors that may occur during gear assembly process.
• Evaluate the performance of gear rings under different carburization layer thicknesses.
• A software based on Abaqus secondary development has been designed to achieve parameterized finite element simulation

Review of Control Strategy for Steady Movement of Pneumatic Manipulator
Dong Guotao, Chunyan Gao*, Xuan Chen, et al.

• A review has been conducted on three directions: fuzzy control, neural network control, and robust control.
• Fuzzy control and neural network control have developed rapidly and achieved good integration with other control theories. However, the development of robust control is slow and its application is limited.
• Based on existing research results, it is expected that the stable motion control strategy of pneumatic robotic arms will continue to develop along the trend of integrating and supplementing control theories.

Patent Name: A Machine-vision-based Creep Deformation Detection Device for Complex Components
Inventor: Hanjun Gao,Dong Guotao, Kun Yan, et al.

• Public account: CN117890205A
• The present invention provides a machine vision based creep deformation detection device for complex components, which can monitor the creep deformation of complex components throughout the entire process while ensuring the heat treatment effect, providing reliable data for exploring the creep evolution law of complex components.

Patent Name: A method for expanding the composite material coating defect dataset based on improved Pix2Pix
Inventor: Hanjun Gao,Dong Guotao, Qiong Wu, et al.

• Public account: CN117079752A
• The attention mechanism is introduced to improve the performance of the generator. The discriminator is designed to evaluate the generated defect sample images. The adversarial control design of the generator and discriminator is used to balance the abilities of the generator and discriminator, enabling the generator to better generate images containing defect features. Through the present invention, the problem of small and uneven sample sizes in defect recognition tasks has been solved.