木板钉子缺陷检查系统设计(3)
MATLAB: matrix laboratory, a multi-paradigm numerical computing environment and fourth-generation programming language.
1 Introduction
In this part will describe some information and problems of the company AvenPallen, and will also introduce the scopes and some limitations.
1.1 Background
The company AvenPallen is mainly responsible for the production of different kinds of wood products, and these products are usually nailed in a robot cell. However due to the unevenness of the density of wood and other issues, which will cause the nail cannot be completely driven into the wood, tilt or twist etc. Resulting in uneven painting affect the smoothness of the pallet, unable to meet the requirements of customers and cause waste. To solve this problem the company AvenPallen manages to seek some systems to achieve quality assurance process in nails, at the same time they can meet the maximum benefit
1.2 Problem description
In the manufacturing process it is desired to obtain the quality assurance of nailed pallets automatically. Different solutions are possible. On the one hand, the tolerances for the height of nails are no more than 1mm and it is a relatively small detection data, which require extremely high sensitivity of the system. On the other hand, each board on the pallet also has tolerances from -1mm to 3mm in height, which also increases the complexity of the system design. This project keeps focus on suggesting and evaluating different solutions for quality assurance of nailed pallets automatically.
1.3 Scope
The goal of the project is to suggest an evaluated system able to detect un-knocked nails in a wood pallet. The requirement is to make sure the tolerances for the height of nails are within 1mm, but the limitations are that tolerance is too small to measure and each pallet has a permissible tolerance from -1mm to 3mm in height, which is difficult to find a suit sensor to meet the height.
2 Methodology
The aim of this report is try to find some systems to make quality assurance of the nailing process. There are three methods have been considered, these are height identify, shadow identify and figure identify.
For height identify, the manual measurement and infrared inspection system can be considered. But the manual measurement usually has low efficiency and accuracy, therefore it is better to use the mechanical inspection system replaced the manual inspection. When talked to the supervisor, the infrared inspection system is hard to achieve. The infrared light is harm to people’s health and it is expensive. Thus, the optical light instead of the infrared light, which is cheaper than infrared light and for this system do not need to pay for the protective cover from infrared light.
For the shadow identify, vision technique has been thought out. After discussing with the manager of the company AvenPallen, the vision technique system has been improved. This system sets a light and a CCD camera to get the figures of the shadow, then uses software to analyse them in order to distinguish products between qualified and substandard.
For the figure identify, 3D laser scanning system is a good choice. 3D scanning system can get a full 3D image, thus the exact position of nail in the space can be obtained to improve system accuracy. And the high-precision and high-quality products can be produced.
However compare with the cost performance of these four systems, 3D laser scanning has the most expensive price and the highest requirement of the software and hardware. Therefore, the optical inspection system and the vision technique system have been improved. In the end, the optical system and the improve system shadow inspection system have been tested.
3 Theory
This chapter will introduce some theories which can help the assumed systems to come out.
3.1 Mechanical inspection