几何量多传感器多点测量关键技术的分析及应用-analysis and application of key technologies for multi-sensor multi-point measurement of geometric quantities.docxVIP

AbstractQuality process control is the remarkable characteristic of mechanical manufacturing industry, in order to do 100% on-line detection to the products, high detection efficiency is required as well as high detection accuracy, which puts forward a new challenge to the precision detection technologies. Geometrical multi-sensor multi-probe measurement technology provides a powerful solution for the manufacturing informationization and quality control requirement of the large-scale manufacturing process.In this dissertation, the key technologies of geometrical multi-sensor multi-probe measurement technology are explored on the basis of analyzing the latest developments in this field at home and abroad, these key technologies include the design of the multi-channel Gaussian measurement system, the multi-channel sensor signal anti-jamming technology, the self-calibration technology of the complicated masters, the technologies of system integration and applications, etc.The major work and innovative researches are as following:Taking the system which has the minimum product of time duration and frequency bandwidth and perfect time-frequency characteristic as the optimal measuring system model, and the realization methods of the Gaussian system are explored. Taking section of Gaussian prototype function expansion as the approximation function of the Gaussian function, for the first-order system and the second-order system, the central limit theorems that cascaded finite terms of the first-order or the second-order system can be an approximation of the Gaussian system are proved to be true, which provides a theoretical guidance for the realization of the Gaussian system. Theoretical analysis shows that the maximum deviation of Gaussian approximation with 16 cascaded first-order systems is 1.4%, and the maximum deviation of the Gaussian approximation with 16 cascaded second-order systems is 0.1%.For the corsstalk of the alternating voltage driven multi-channel sensor si