Study On Modeling And Prediction Algorithm Of The Rolling Mill’s Torsional Vibration

It had been a keypoint of increasing production efficiency and updating product quality that predicting the buckling mechanics induced by low-frequency torsional mutational load acting on rotational spindle of the rolling mill’s main driving systems. The project was supported by the National Significiant Tackle Key Problem for 10th-5-year Plan of China (Research on torsional vibration of the rolling mill.Grant No.ZZ02-13B-02-03-1). By means of modeling and studing torsional vibration’s prediction algorithm, the buckling and evolvement mechanics was revealed which describing the course of gradually happening or paroxysmal torsional vibration of the rolling mill. The prediction algorithm was formed that realized distinguishing property and character of torsional vibration. The main contents are as follows:A systematic research means had been designed under the frame of continuous topology set-valued and differential systems. In the course of from topology set-valued to differential inclusion and again to topology set-valued, the prediction algorithm of the rolling mill’s torsional vibration had been gradually formed and perfected.Firstly, under the frame of the continuous topology set-valued, global coupling uniform model was established including complex nonlinear operating mode and complex configuration about rolling mill’s main driving systems, utilizing systematic dynamics and topology dynamical system theory , which as well the torque and torsional angular displacement as systematic input and output signal. And the calculation equation of system’s and local’s torsional vibration sensitivity was gained. Under the frame of differential inclusion, quantitative calculation prediction algorithm of the rolling mill’s torsional vibration was achieved. Furthermore the evoluting mechanics of mutation torsional vibration was revealed, through studying response characteristic of torsional vibration of the rolling mill in different outside motivation nature and different boundary initial condition in the rolling processing. It realized fuzzy evaluation and estimate identify of the torsional vibration topology network system’s output of the rolling mill. Secondly, the bifurcation and it’s balance and stability position using qualitative methods was studied. The torsional vibration condition and it’s form of stability solution was explored after the balance position buckling, by tracing bifurcation solution. It solved the stable convergence of several suloutions on identical frequency band in nonlinear periodic system of torsional vibration system, and gained kinematic stability criterion and structure stability criterion of the rolling mill’s torsional vibration system. The qualitative methods of torsional vibration stability predicition was gradually formed in the course of the research.Thirdly, by associativing the statistical characteristics extracting and topology network fusion idea, the study estabilished network integration predicition algorithm on the rolling mill’s torsional vibration, including torsional vibration topology network systematic predicition algorithm, single shaft sector local predicition algorithm , speed fluctuation series decoupling decomposed local predicition algorithm, extracting signal characteristic and it’s quickly calculation local predicition algorithm. These algorithms could realize quantitative predicition and synthesis evaluating the torsional vibration in different level and location of the rolling mill.Finally, in order to make the predicition algorithms of the rolling mill’s torsional vibration to pravtice and application, the research designed and manufacture fault source test workingpieces, and layout measurement system. And the dynamic experiment was finished on the torsional vibration analogue platform. The feasibility and effectivity of studied were verified about mentioned above related theory or algorithm.