叶片—转子系统振动特性与参数辨识方法研究

发布时间：2017-12-30 18:15

本文关键词：叶片—转子系统振动特性与参数辨识方法研究　出处：《大连理工大学》2013年博士论文　论文类型：学位论文

【摘要】：高速旋转机械作为一类被广泛使用的机械设备,在许多行业(如航空航天、电力、冶金等)中都发挥着重要的作用,转子系统作为高速旋转机械的重要组成部分,其研究一直受到广泛关注,近年来由于转子系统的故障产生的恶性事故给企业造成过极大的损失,因此对转子系统进行准确的建模以及故障和参数的辨识,逐渐受到关注。本文以转子系统为主要研究对象,建立了叶片-转子-轴承耦合系统的非线性动力学模型,并将改进的差分进化算法应用到失谐叶盘系统的优化排序以及转子系统的结构参数和故障参数的辨识中。为了进一步提高运算效率,提出了多点加点Kriging代理模型和改进的差分进化算法相结合的参数辨识方法,高效准确的实现了线性和非线性转子系统的参数辨识工作。本文的主要工作有以下几个方面： 1、建立了能够全面考虑叶片弯曲振动的叶片-转子-轴承系统的非线性动力学模型。首先,利用集中质量法将转子系统离散。其次,为分析叶片对转子系统的惯性效应并考虑系统的时变性,将叶片模化为悬臂梁结构,利用假设模态法对叶片进行离散分析,同时根据叶片的循环对称性对系统的微分方程进行简化降阶。对于叶片数量较多的叶片-转子-轴承系统可以实现明显的维数降低的效果。利用数值方法对系统动力学方程进行求解,并通过分岔图、最大Lyapunov指数曲线、相图、Poincar6截面映射、时域波形、幅值谱图等分析了该系统在非线性油膜力作用下的弯扭耦合振动特性。最后,讨论了叶片的存在以及叶片长度对系统非线性演化过程的影响。 2、建立了叶片-圆盘系统的集中参数化模型,分析了叶盘系统自由振动和受迫振动下的振动特性,并讨论了失谐叶盘系统各阶模态的局部化程度以及失谐叶盘系统响应影响的一般规律。结合所得规律,提出了一种评价失谐叶盘系统振动优劣的新的评价参数。最终将差分进化算法应用到失谐叶盘系统叶片排序的优化研究中。在兼顾降低振幅和平衡叶片间振幅大小的情况下,使各叶片较均匀的分担系统的整体振动能量,以达到降低疲劳、延长寿命的目的。 3、在基本差分进化算法的基础上提出了适用于转子系统参数辨识的遗传-自适应混合差分进化算法。由于转子系统待辨识参数过大或者过小同时待辨识区间范围较大,因此首先引入具有全局搜索能力的遗传算法以缩小问题的寻优区间,其次为了防止问题陷入局部最优,提出了自适应Cauchy变异和自适应Caussian变异策略,用以修正差分进化算法原有的变异策略。并以考虑不平衡量的线性转子模型和考虑油膜力和碰摩力共同作用下的非线性转子模型为对象进行仿真分析,验证了所提出方法在转子系统参数辨识中的可行性和准确性。最后与基本差分进化算法和遗传算法进行比较,结果显示本文提出的GA-AHDE优化算法能够快速有效的逼近全局最优解。 4、在原有Kriging代理模型的基础上,结合改进的自适应混合差分进化算法设计了新的转子系统参数辨识方法,在每次更新Kriging代理模型时,增加当前由差分进化算法得到的最优设计点,以提高模型的全局预测精度。通过数值算例和实验,验证了该方法在转子系统参数辨识中的高效性和准确性。在此基础上又将新的多点加点准则引入到Kriging代理模型中,即在每次更新模型时除了增加当前最优设计点外,还根据搜索进程加入相关度较大或较小的点,从而进一步提高了Kriging代理模型的精度,更大程度的提高了搜索效率。最后通过完成线性算例和非线性算例的辨识工作,讨论了多点加点的Kriging代理模型与改进的差分进化算法相结合的参数辨识方法在不同情况下的的辨识结果,并分析了该方法的适用条件。
[Abstract]:High speed rotating machinery as a mechanical equipment widely used, in many industries (such as aerospace, electric power, metallurgy etc.) have played an important role, as an important part of the high-speed rotor system of rotating machinery, its research has been widespread concern, accidents in recent years due to the fault of rotor system to produce the enterprise had caused a great loss, so the accurate identification and fault modeling and parameters of the rotor system, has attracted increasing attention.
In this paper, the rotor system is the main research object, established the nonlinear dynamic model of blade rotor bearing system, the identification and the improved differential evolution algorithm is adopted to optimize the mistuned bladed disk system sort and structure parameters of rotor system and fault parameters. In order to further improve the computational efficiency, and puts forward the method of parameter identification of multi point sampling Kriging model and the improved differential evolution algorithm combining, realize the efficient work of parameter identification of linear and nonlinear rotor system. The main works of this paper are as follows:
1, established can comprehensively consider the nonlinear dynamics model of blade blade bending vibration of the rotor bearing system. Firstly, using the lumped mass method to discrete the rotor. Secondly, for the analysis of the inertial effect of blades on the rotor system considering time-varying system, Ye Pianmo as the cantilever beam structure, by using assumed mode method the discrete analysis on the blades, and according to the differential equation of circular symmetry of leaves on the system are simplified and reduced order. For the number of leaf blade rotor bearing system can achieve obvious dimension reduction effect. Using numerical method for system dynamics equations, and the bifurcation diagram, the largest Lyapunov exponent curve, phase diagram section Poincar6, mapping, time domain waveform, amplitude spectrum analysis of coupled bending and torsional vibration characteristics in the nonlinear oil film force of the system. Finally, discuss the blade The influence of the existence of the blade and the length of the blade on the nonlinear evolution of the system.
2, a concentrated parametric model of blade - disc system, analysis of the free vibration of bladed disk system and forced vibration vibration, and discusses the localization of mistuned bladed disk system modal and response of mistuned bladed disk system of general rules. Combined with the law, put forward a evaluation of mistuned bladed disk system vibration the merits of the new evaluation parameters. Finally the differential evolution algorithm is adopted to optimize system of vanes of mistuned bladed disk. The decrease amplitude and amplitude balance between the blades under the overall vibration energy makes the blade sharing system more uniform, in order to achieve reduce fatigue, prolong life.
3, based on the basic differential evolution algorithm for genetic parameter identification of rotor system - adaptive hybrid differential evolution algorithm is proposed. The rotor system parameters is too large or too small to be identified at the same time range is larger, so the genetic algorithm has the global search ability is first introduced to reduce the problem of optimization the second interval, in order to prevent problems fall into local optimum, we propose adaptive Cauchy mutation and adaptive Caussian mutation strategy is used to revise the differential mutation strategy evolutionary algorithm. The original nonlinear rotor model and to consider the balance of linear model and considering the amount of rotor oil film force and the interaction of rub impact force as the object for simulation analysis and verify the feasibility and accuracy of the proposed method in parameter identification of rotor system. And finally the basic differential evolution algorithm and genetic algorithm are compared, the results show The GA-AHDE optimization algorithm presented in this paper can quickly and effectively approach the global optimal solution.
4, based on the original Kriging model, combined with the improved adaptive hybrid differential evolution algorithm design method for parameter identification of rotor system, each update Kriging agent model, the optimal design point of the differential evolution algorithm is obtained, in order to improve the prediction accuracy of the global model. Through the numerical example and experimental validation of the method in parameter identification of rotor system in high efficiency and accuracy. On the basis of the multi-point sampling criterion is introduced into the new Kriging model, namely each update model in addition to the optimal design point, according to the search process added to a greater or lesser degree. Thus, the Kriging model accuracy is further improved, to a greater extent to improve the search efficiency. Finally, complete the identification work is linear and nonlinear case examples, discussed the multi point plus point K The identification results of different parameter identification methods based on riging surrogate model and improved differential evolution algorithm are applied in different cases, and the applicable conditions of the method are analyzed.

【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TH113.1

【参考文献】