The Development Of The High-Power LED Intelligent Module Micro-Coating System

Nowadays,the LED industry is in a period of steady growth,and the global LED market keeps growing rapidly.So,LED packaging area growth is also very fast.Because of the importance and professionalism of LED packaging equipment,market demand for LED packaging equipment,especially coating equipment is increasing.But in the current domestic market,the LED is coated by the traditional dispensing process,which haves some problems like uneven thickness and uneven consistency.Therefore,it is necessary to develop the new coating equipment using a new spraying process to break the technological blockade of foreign countries and achieve a high degree of automation and intelligence.The main worker of this paper is as follows:(1)Determine the overall framework structure of the system,such as the relationship between the software control part and the hardware equipment part,clarify the requirements of each part of the system,and select the coding language framework for the control software and equipment's hardware equipment.This step laid the working foundation of integral system.(2)Design the hardware structures,such as circulating mixing structure and multi-motion axis structure,to solve practical process problems like precipitation of phosphor powder glue mixture and the high-precision coating.This step guaranteed the stable and efficient of coating equipment.(3)Design and develop the control software under the Qt programming framework based on C++.determined that the structure of the control software is composed of three layers,the bottom layer-the functional layer-the interface layer.Then develop and implement functional modules like user management,path planning,history management.Ensure the reliability and stability of the control software.(4)Implement the calibration and coordinate transformation of industrial CCD camera,which ensures that the template matching algorithm can be applied to accurately locate the coating substrate.This step meets the requirement of high precision coating.(5)Used the deep learning algorithm based on transfer learning and convolutional neural network to train the neural network suitable for the classification of substrates waited to be coated.This neural network can recognize and classify different substrates under harsh conditions.This step realized the automation and intelligence of the system.