【作者】 王召斌；

【导师】 翟国富；

【作者基本信息】 哈尔滨工业大学 ， 电机与电器， 2013， 博士

【摘要】 航天电磁继电器是国防武器系统中实现信号传递、负载切换、电路保护与控制、系统隔离等功能的关键机电元件。为了满足武器装备日益提高的战备需求,已有武器型号对航天电磁继电器提出了24年贮存期的指标要求。而目前对于航天电磁继电器这类高可靠长寿命机电元件的贮存可靠性关注较少,相关研究面临试验及检测水平不高、性能参数变化规律不明、失效模式与失效机理不清、评估与预测手段不足等一系列问题。因此,如何评估航天电磁继电器的贮存可靠性已成为亟待解决的重要问题。本文以“长期贮存,一次使用”的国防武器装备为应用背景,以型号产品中使用较为广泛的某航天电磁继电器为研究对象,提出一种基于性能退化的贮存可靠性试验及评价方法,通过对贮存退化试验过程特性参数变化情况分析与失效机理研究,对航天电磁继电器的贮存可靠性建模、贮存可靠性评估和相关退化参数的预测方法进行深入研究,可为武器装备系统的定寿与延寿提供技术支撑。首先,为了解决航天电磁继电器贮存可靠性退化数据的高精度、全自动获取问题,本文设计并实现了航天电磁继电器贮存退化试验系统。采用FPGA和ARM单片机为硬件核心架构,通过集散控制的方式建立多通道、高速的信号采集分析系统,可最多同步监测40支继电器触点压降、线圈电流,并设计了时间参数的软件算法。采用VC++面向对象的编程方法,设计实现继电器贮存退化试验的参数设置、退化参数的趋势显示、分析等功能的上位机软件系统。该测试分析系统时间参数的检测分辨率为1μs,触点接触电阻的测试分辨率为0.01mΩ。为航天电磁继电器贮存可靠性研究提供可靠退化数据与有效实验平台。然后,针对传统基于现场贮存的可靠性试验方法难以满足长寿命产品的贮存可靠性评价要求问题,本文对航天电磁继电器的贮存退化试验条件、试验方法、试验过程进行了探索性研究。采用图分析法及数理统计方法对试验获得的的继电器接触电阻及时间参数数据进行了分析,得到贮存条件下继电器各性能参数随时间的变化规律,在此基础上,将接触电阻与吸合时间确定为贮存退化的敏感参数。其次,从航天电磁继电器的功能、结构及贮存环境应力出发,从其传导特性的角度对继电器的贮存失效机理进行了分析,建立了退化物理模型及可靠性统计模型。采用SEM试验及EDX分析方法,定期对继电器触点进行微观形貌及化学成分变化分析,为继电器贮存失效机理分析提供依据。采用结合Larson-Miller经验公式建立了基于应力松弛的继电器吸合时间贮存退化模型。根据扩散理论,通过三维Fick方程建立了航天电磁继电器的触点贮存等效动力学模型,该模型将接触电阻的增长与环境温度、接触压降、接触压力等诸多因素进行了考虑和有效关联,为接触电阻的快速评估提供了理论指导。最后,针对现有的贮存可靠性评估方法多为静态、事后分析且默认产品符合某指定的分布类型并基于经验公式估算产品可靠性的问题,结合Holm电接触理论,在继电器触点贮存失效机理的研究基础上,建立了综合考虑接触电阻增长随机性与确定性的电磁继电器贮存可靠性评估模型。研究了评估模型的参数评估方法,并对航天电磁继电器的贮存可靠度进行了评估。综合人工神经网络非线性适应能力强,灰色系统理论弱化数据序列波动性的特点,建立了航天继电器贮存退化参数的G-ANN模型,并对接触电阻实测值进行了参数预测,结果表明短时预测精度较高。提出了时间序列分析与一维小波变换方法相结合的预测参数预处理方法,降低预测参数中随机信号对预测精度的影响。通过回归理论估计了贮存退化模型的参数,并通过所建模型对航天电磁继电器常温下的贮存寿命进行了预测。本研究为开展航天电磁继电器的贮存可靠性提供了必要的研究基础与研究手段,其关键技术与相关方法也可以推广应用到其它机电元件的贮存可靠性研究中。更多还原

【Abstract】 Aerospace electromagnetic relay (EMR) is one of the key electromechanicalcomponents used for signal transmission, load switching, circuit protection andcontrol and system isolation. To increase the weapon performance and meet theneed of “long-term storage, one shot”,24years storage life has been proposed insome weapon project. At present, concern for the storage reliability of AerospaceEMR is respectively light although it comes to the highly-dependable and long-lifeelectrical components. And the corresponding research still faces a series ofproblems, such as low level experiment and test detection, unclear law o fperformance parameters changing, ambiguous about failure mode and failuremechanism, and lack of effective evaluating and predicting means. Therefore, howto evaluate the storage reliability of aerospace EMR has become the importantsubject that demanding prompt solution at present.This dissertation with the background of “long-term storage, one shot” weaponequipment system and a type of widely used aerospace EMR as research object. Anew method for testing storage reliability is proposed by performance parametersdegradation. Based on the research of failure mechanism and the analysis ofparameters changing in storage degradation testing, the modeling storage reliability,storage reliability assessing method, and degradation parameters predictionmethods of aerospace EMR are extensively investigated. These studies will providethe support and reference for determining and extending the life of missile weaponsystem.First of all, in this dissertation: firstly, the test and analysis system for storagedegradation test is designed which uses FPGA and ARM SCM architecture assystem’s hardware core. Signal acquisition and analysis system of high speedmult i-channel is built by distributing control method. This test system provides thepossibility of measure synchronously up to40relays’ contact resistances ècontactvoltage drop and coil current. Software algorithms of relay time parameters aredesigned and implemented. The upper computer software is developed to carry outfunctions that test parameters setting, trends showing and data querying ofdegradation parameters by VC++object orient programming. The time parametersdetective resolution can achieve1s and the measurement resolution of contactresistance is no more than0.01m.This test and analysis system provides theresource of degradation data and effective testing platform for studying storage reliabilit y.Secondly, the traditional reliability testing methods based on field storagecan’t satisfy the needs in storage reliability evaluation of long-life products. Thisdissertation did the exploratory research on the storage degradation test method andtest process for aerospace EMR. The experimental data of contact resistance andtime parameter are analyzed by the methods of mathematical statistics and diagramanalysis. And the variat ion law of performance parameters changing with time isobtained, which provides the bases for determining degradation sensit iveparameters.Thirdly, based on the structure and function of aerospace EMR, the storagefailure mechanism is investigated by conductive properties themselves. Themicroscopic morphology and changes in chemical elements for relay contactsurface was analyzed by SEM and EDX regularly, which provide references for therelay storage failure mechanism of relay. The relay closing time storagedegradation model is established based on stress relaxation, and applied theLarson-Miller empiric formula. According to the diffusion theory, storage equivalentkinetic model of relay is built by three-dimensional Fick equation. This modelconnects resistance increasing of aerospace EMR with temperature, contact voltagedrop, contact stress and some other factors effectively. And, these studies providethe new ideas and references for rapid assessment of relay contact resistance.Finally, the exist ing storage reliabilit y assessment methods are mostly staticand post-mortem. They acquiesce in a fact that product distribution subject to aspecified distribution and estimate the product reliabilit y based on empiricalformula. Based on Holm electrical contact theory and contact storage failuremechanism, the storage reliabilit y assessment model of aerospace EMR isestablished which synthesizes the stochastic and deterministic of contact resistanceincrement. Meanwhile, the parameters estimation method of assessment model isstudied, and storage reliabilit y of aerospace EMR is assessed. Artificial neuralnetwork has strong nonlinear adaptive capacity and Grey system theory weakensdata sequence fluctuation. The G-ANN combination prediction model of relaydegradation parameters has been built up by combining above characteristics. Thisapproach forecasts the contact resistance degradation of aerospace relays in thestorage test, and the results show that this method has higher precision ofprediction. Prediction parameters preprocessing method is proposed whichcombines time series analysis with one-dimensional wavelet transform method. This reduces the random signal’s influence on the prediction precision in predictionparameters. Parameters of the storage degradation model are estimated throughregression theory. And the storage life of Aerospace EMR under normal stress ispredicted.The research results in this dissertation provide the necessary theoreticalmethod and basis for studying storage reliability of the aerospace EMR. Its keytechniques and methods could be also applied to other electromechanicalcomponents storage reliability studies.更多还原