Error Analysis And Compensation Of On-line Inspection System

With the development of science and technology, the application of complex surface parts are using more and more widely in the areas of aerospace, shipbuilding, mold, automotive and other industrials. Inspection of complex surface parts, has become increasingly related to the level of a country’s equipment manufacturing industry. At present, most are used in the form of off-line inspection for complex surface parts, such as the use of the coordinate measuring machine testing. The off-line inspection exists several shortcomings including a secondary clamping error, expensive equipment and strict requirement to the operating environment. Using the technology on-line inspection can directly inspect the workpiece without changing its position and development of a closed-loop processing machining-inspection-compensation system. Recently, more and more attentions have been paid to direct inspecting on numerical contral machines. In order to tackle the difficult problems of an on-line inspection system, this paper focuses on the error analysis and compensation method of on-line inspection system, which includes the analysis of pretrip error prediction method in the probe system to improve the detection accuracy through error compensation. Finally, the uncertainty evaluation to the test results of the inspection system is studied. The main contents and conclusions of this dissertation are listed as follows.(1) Investigate and overview the development status on the topics of error analysis and compensation methods of on-line inspection system all over the world,and point out the difficult issues to be solved in this area.(2) Analyse the various error sources in the on-line inspection system, find out the main factors which affect the inspection accuracy, and study the error acquisition and compensation methods. For the pre-trip error of the contacting probe, this paper proposed two predicting modeling approaches for the pre-trip errors through the genetic algorithm BP neural network and the regular RBF network, which can be used to predict the pre-trip errors in different normal directions.(3) Study the measurement error and uncertainty assessment methods for the on-line spection system, including applying the methods of the uncertainty A-type evaluation and uncertainty B-type evaluation to the system. Through the use of uncertainty evaluation, the test results of the free-form surface parts are assessed.(4) Based on the proposed prediction model and evaluation method, the modules are implemented in the inspection system..The experimental work are performed with examples of the workpieces to verify the pre-trip error prediction and compensation effect and detection accuracy uncertainty.With this study, the online inspection system developed has the functions of error modeling, error compensation and accuracy uncertainty evaluation, etc. The system can be widely applied to the relevant enterprises to improve the machining accuracy of the surface parts.