Auto-Identification Of Defective Product In Shock Absorber Production Process

In 2010, China's Vehicle production and sales both extend 18 million which leads the world. More and more Chinese families have their own cars. Vehicle safety and comfort have always been the focus of research at home and abroad. Shock absorber is an important part of the suspension of the car, and it has the material effect to car safety and comfort. Shock absorber manufacturer must show the product features 100% in damper test rig to ensure the product meet design requirements. Indicator diagram of shock absorber plays an important role in identifying whether it is qualified. At present, shape identification of the indicator diagram of shock absorber depends heavily on experience. The topic uses BP neural networks which based on MATLAB, a system which can be used for auto-identification of the defective product is designed. The system can effectively determine whether the shock absorber is qualified and can be used at home and abroad. The topic divides the main research content into the following four parts:1,All kinds of defect types which possibly appears in the production process is discussed, the main reason about each defect is analyzed and the appropriate solution is given. At the same time, each of the indicator diagram type of shock absorber are classified by the defect type, the main features of indicator diagram of shock absorber is analyzed.2,A method for extracting the feature parameter, which can show the feature of each indicator diagram of shock absorber, in indicator diagram is designed. It is designed by the data collection methods in data acquisition system of the damper test rig and the requirements of extracting feature parameter, and the feature parameter is processed with the method of normalization.3,The main parameters of BP neural networks was determined as to ensure the effectively and rapidity of identification system through the BP neural networks and its realization in MATLAB, also, the rules of BP networks training are analyzed.4,An system used for automatically identification the defective shock absorber was designed ,through which the example of indicator diagram of shock absorber examination, the validity of the system is proven, and the system has the superiority which can effectively identify the defect types that can't be identified by currently method on the market. This research solves the main defect which hides in the identification of defective production in shock absorber production process. The new identification system has a huge vendibility, which can effectively improve the efficiency of the shock absorber test, also can be used in all the shock absorber company.