基于BiSNet的航天发动机转子智能健康监控

航天发动机某部件转子健康监控,对检测发动机能否正常工作进而影响航天发射任务的成败至关重要.传统健康监控方法需要依赖经验和专业知识,在对故障机理深入理解后才能构建相应的健康监控模型.针对传统转子健康监控手段效率低、适应性差、过度依赖专家经验等不足,结合海量转子振动健康监控历史数据,提出并设计采用神经网络BiSNet的人工...     

直升机旋翼桨叶气弹优化减振设计方法

从振源着手通过设计参数的最优选择设计直升机旋翼桨叶,使传递到机身的交变载荷最小达到降低振动水平的目的是旋翼桨叶设计思想的进步。本文在简单概述该方法的基础上,以某4桨摆振柔软的无铰复合材料桨叶为研究对象,将其大梁模拟为一单闭室复合材料盒形梁,研究了通过铺层角的优化选择,降低4次/转的桨毂交变力与力矩的情况。数值算例表明方法效果良好。      

构件振动可靠性设计方法初探

本文提出一种进行构件振动可靠性设计的方法,建立了激振力频率与构件固有频率干涉的概率模型,并利用Goodman曲线,给出了将导致构件损坏的强迫共振响应的概率计算公式及其使用条件。原则上,本文方法对其固有频率相对不太密集的构件更为适用,如航空发动机中的叶片。      

多维系统测量的误差分析计算

多维系统测量参数众多,系统误差分析比较复杂。本文归纳给出多维系统测量误差分析计算方法措施,并给出算例。     

导弹直接侧向力机动突防方案设计

针对敌方不同来袭方位的拦截弹,提出了一种可用于导弹末段侧向力机动的突防方案。在建立拦截弹和导弹的相对运动模型基础上给出了脱靶量与视线转率、相对距离和机动加速度的解析表达式。以拦截弹在视线系Z向上的脱靶量达到最大值作为性能指标,对机动发动机开启时刻进行优化确定。同时通过分析机动发动机与Y向舵偏的不同组合方式对导弹的速度增量及交会角的影响,确定机动发动机的推力方向应与气动力方向一致,并使交会角明显减小。仿真结果表明,此方案能使导弹最终落入拦截弹杀伤半径10 m之外,成功实现突防。     

Robust H∞ controller design for attitude stabilization of flexible spacecraft with input constraints

This paper addresses the attitude stabilization and vibration suppression problem for flexible spacecraft subject to model parameter uncertainty, controller perturbations, external disturbances and input constraints. The attitude model of flexible spacecraft is described and converted into a state space form in terms of passive and active vibration suppression schemes. A novel state feedback controller is proposed based on the exactly available expectation of a new variable, which is introduced to model a randomly occurring controller gain perturbation. Based on Lyapunov stability theory, sufficient conditions for the existence of the nonfragile H_∞ controller considering input constraints are given based on linear matrix inequalities (LMIs) in terms of additive perturbation and multiplicative perturbation. Then, the developed controller subject to required constraints can be obtained, where the nonfragile property is fully considered to improve the tolerance to uncertainties in the controller. Numerical simulations are performed to demonstrate the effectiveness and superiority of the proposed control strategy in attitude stabilization and vibration suppression, where it should be noted that the passive vibration suppression scheme is superior for high natural frequencies while the active vibration suppression scheme is superior for low natural frequencies. Moreover, the low natural frequencies have more influence on the performance of attitude stabilization and vibration suppression.     

基于Labview的振动冲击测量系统设计

根据振动冲击试验设备测试要求,设计并研制一套测量系统。系统基于虚拟仪器测量技术设计,通过采集振动冲击所产生加速度信号的频率和幅值,运用FFT技术进行信号处理和分析,完成对振动冲击试验设备的测量。实验效果表明,该系统具有测量功能完善、测量结果准确可靠、软件操作简单等特点,能够满足目前在用的所有振动冲击试验设备的测量需求,在实践中已得到应用。     

A novel stability prediction approach for thin-walled component milling considering material removing process

The milling stability of thin-walled components is an important issue in the aviation manufacturing industry,which greatly limits the removal rate of a workpiece.However,for a thin-walled workpiece,the dynamic characteristics vary at different positions.In addition,the removed part also has influence on determining the modal parameters of the workpiece.Thus,the milling stability is also time-variant.In this work,in order to investigate the time variation of a workpiece's dynamic characteristics,a new computational model is firstly derived by dividing the workpiece into a removed part and a remaining part with the Ritz method.Then,an updated frequency response function is obtained by Lagrange's equation and the corresponding modal parameters are extracted.Finally,multi-mode stability lobes are plotted by the different quadrature method and its accuracy is verified by experiments.The proposed method improves the computational efficiency to predict the time-varying characteristics of a thin-walled workpiece.     

压气机进气畸变诱导叶片振动数值研究

介绍了一种考察进气畸变对压气机叶片振动及结构强度影响的流固耦合分析方法。对压气机进行全周数值模拟分析进气总压畸变下转子流场的非定常响应;并结合有限元模型分析叶片模态及振动响应,通过气动模型与结构模型的不断信息迭代实现叶片振动过程的模拟。分析表明,周向总压畸变会引起叶片气动力在周向分布的不均匀及叶片应力波动,导致压气机叶片振动,对其结构稳定性产生重要影响。     

一种惯组外场标定振动干扰的分离方法

惯组在外场进行标定时常会受到地基振动的干扰,这种干扰会对惯组的测试精度造成极大的影响.为解决这一问题,根据工程实际需要,提出了一种基于经验模态分解和数字信号滤波技术相结合的惯组输出振动干扰分离方法.首先利用经验模态分解将受振动干扰的惯组输出分解成若干本征模态分量,然后采用低通滤波器对高频分量进行滤波,对低频分量相应区段做幅值修正,最后将处理过的各分量叠加,即重构出消除振动干扰的信号.相对于传统的剔除野值后的平滑滤波方法,新方法保证了数据信息的完整性,有效降低了振动干扰的不利影响,提高数据信息的可靠性和真实性.仿真实验和工程实践表明,该方法对外场的瞬时振动干扰有较强地分离能力.