Auxiliary Positioning System Of Nuclear Power Plant Refueling Based On Machine Vision

The development of human society is inseparable from energy,and nuclear energy is valued by all countries for its cleanliness,environmentally friendliness,high-efficiency and low-consumption.The current method of using nuclear energy is mainly to carry out nuclear fission reactions in the reactor core to generate electricity.The fuel assembly is the key component of the nuclear reactor core,whose accurate installation and unloading is the prerequisite for safe and stable operation of the reactor.At present,the fuel assembly refueling is still operated by manual.There are many problems such as numerous operators,complicated process,low-degree automation,and the exposure to strong nuclear radiation.With the development of computer and image processing technology,the machine vision technology which stands out with high precision,efficiency,safety and reliability,has been widely used in various industrial fields.In view of the existing problems in fuel assembly refueling and loading operations,and the requirement of the special working environment and accuracy standard,this thesis develops a machine vision-based research on the auxiliary positioning system for the refueling of nuclear power plants.The main content of the thesis is as follows:(1)This thesis proposes a design scheme of the machine vision-based auxiliary positioning system for refueling.Aiming at the principle of the refueling auxiliary positioning system and the working environment of the fuel assembly,the systematic hardware and software design scheme is proposed.The core of the hardware design lies in the underwater imaging system.Based on the optical imaging principles,combined with the specific size and positioning requirements of the nuclear power plant refueling equipment and fuel assembly,the selection scheme of the camera parameters of the underwater imaging system is given to ensure the acquisition high-quality images.According to the control software requirements of the refueling auxiliary positioning system,the software of the refueling auxiliary positioning system is designed in detail.(2)This thesis proposes an underwater image enhancement algorithm based on convolutional neural networks.In order to solve the problems of color shift,unclear edges,as well as low brightness,contrast,and resolution in the underwater images taken by underwater imaging systems,this thesis proposes an underwater image enhancement algorithm based on convolutional neural networks.Started from the underwater imaging model,the algorithm makes the images clear and achieves super-resolution through detail enhancement and color correction.The experiment results prove the effectiveness of the proposed algorithm.(3)This thesis proposes a sub-pixel edge detection algorithm based on partial area effect.In order to meet the high requirements of the refueling auxiliary positioning,which the traditional edge detection positioning algorithm cannot satisfy,this paper proposes a sub-pixel edge detection algorithm based on the local area effect.Under the principle of local area effect,the algorithm effectively improves the positioning accuracy of the fuel assembly position through image denoising,sub-pixel edge extraction,and edge fitting.Experiments show that the detection precision of the algorithm is 0.1mm,which meets the accuracy requirements of the system.(4)This thesis realizes the refueling auxiliary positioning system.According to the principle and design of the refueling auxiliary positioning system,an experimental platform for the refueling auxiliary positioning system was built,and the corresponding software was developed and tested on the test platform.The test results show that the system meets the designed accuracy requirements and realizes the high-precision fuel assembly refueling operations.