| With the improvement the steel market share of cold-rolled strip, high-tech, high-quality plate and strip products are more and more popular in users. line detection and pattern recognition techniques for improving the quality plate and strip products important technical aspects, it has important significance. As an important technology of improving product quality of plate strip, on-line detection of Strip shape and flatness pattern recognition has important research significance. Under the support of the National Natural Science Fund Project(Serial number: 51027003), based on a closed piezoelectric shapemeter(patent number: CN101497084A) and a comprehensive testing platform of piezoelectric shapemeter(patent number: CN101518787A) by our group independent research, this paper analyzed the influence on the test results by various factors in the process of flatness detection through experiments, researched the mechanism of the plate with the wave shape defects and the method of flatness pattern recognition.On the basis of dynamic calibration of plate shape roller,this paper researched the influence of the speed of strip on the strip shape detection by adjusting the speed of plate shape roller, the experimental analysis showed that, while the strip speed reaches a certain value the flatness detection signal will be stable, and it will not change with the increase of stripe speed; researched the influence of the strip steel tension on flatness detection by adjusting the strip steel tension, and this paper carried on the test and analysis of the generator of plate with shape defects for the first time, by changing the local strip tension, studied the the change of strip steel tension distribution and the flatness defect caused by it.This paper also studied the method of the flatness pattern recognition, established the flatness pattern recognition system based on the Elman neural network.The recognition system was tested by the test data of the generator of plate with shape defects,the rusult proved the feasibility of the pattern recognition system. This paper designed the flatness pattern recognition user interface based on the experimental platform with the Matlab GUI development tool. |
PDF Full Text Request