【作者】 王宁；

【导师】 高金吉； 江志农；

【作者基本信息】 北京化工大学 ， 控制理论与控制工程， 2008， 硕士

【摘要】 现代旋转机械正朝着高速化、重载化、智能化的方向发展,透平、风机、压缩机等高速旋转设备已经成为过程工业中的核心装置。然而,由于质量不平衡而引起的同频振动影响了设备的正常运转,因此保证转子平衡成为解决问题的关键。本文研究的主动平衡技术,寻求在线完成转子的平衡过程,力图避免传统动平衡方法中设备停车给企业带来的经济损失。本文的主要研究内容如下:(1)总结了国内外自动平衡技术的研究现状和发展趋势,分析了自动平衡的主要难点和关键技术,阐述了本文研究的意义(2)研究了动平衡和自动平衡的动力学原理和基本理论(3)本文还研究了电磁型主动平衡装置基于影响系数法的单平面自校正自适应控制算法,最后在Jeffcott转子系统上针对正确和错误的影响系数初始值,仿真对比研究了传统的影响系数法、固定参数和时变参数影响系数法的平衡过程,仿真结果表明了时变参数影响系数法的优越性。(4)本文采用了简单的模型,设计了针对持续注液平衡头的低层控制算法,并对它做了仿真实验研究。更多还原

【Abstract】 Modern rotating machinery developed on the trend of high speed、heavy load and intelligence. The high speed rotary device are the key equipment in the process industry, such as turbomachinery、pan、compressor and so on. Resonance vibration caused by unbalance influences the ordinary operation of equipments, so rotor’s balancing is the critical factor for solving this problem. In order to complete the balancing process on line, the active balancing technique is studied in this paper. This skill avoids the broken down of the equipment on tradition balancing method, which decreases the company’s loss. The main works could be summarized as follow:(1) The state of the art and tendency of the auto-balancing technology are reviewed and the significance is discussed.(2) The dynamics principle and basic theory of dynamic balance and auto-balance are studied.(3) Single-plane self-correction controlling algorithm based on the influencecoefficient method for electromagnetic active balancing device is also studied in this paper. Aiming atcorrect and false initial value of the influence coefficient, contrast simulations among balancing process rates of conventional influence coefficient method,fixed parameter and time-varying parameter influence coefficient method arecarried out on the Jeffcott rotor system. Simulation results indicate superiority of time-varying parameter influence coefficient method.(4)Used the simple model, which designs the control algorithm of continuously-dripping balancing head is also studied in this paper.更多还原