Aero-engine fault diagnosis applying new fast support vector algorithm

A new fast learning algorithm was presented to solve the large-scale support vector machine ( SVM ) training problem of aero-engine fault diagnosis.The relative boundary vectors ( RBVs ) instead of all the original training samples were used for the training of the binary SVM fault classifiers.This pruning strategy decreased the number of final training sample significantly and can keep classification accuracy almost invariable.Accordingly , the training time was shortened to 1 / 20compared with basic SVM classifier.Meanwhile , owing to the reduction of support vector number , the classification time was also reduced.When sample aliasing existed , the aliasing sample points which were not of the same class were eliminated before the relative boundary vectors were computed.Besides , the samples near the relative boundary vectors were selected for SVM training in order to prevent the loss of some key sample points resulted from aliasing.This can improve classification accuracy effectively.A simulation example to classify 5classes of combination fault of aero-engine gas path components was finished and the total fault classification accuracy reached 96.1%.Simulation results show that this fast learning algorithm is effective , reliable and easy to be implemented for engineering application.     

应用快速多分类SVM的航空发动机故障诊断方法

提出了一种新的快速多分类SVM算法,用于解决大样本情况下航空发动机的多类故障诊断问题。首先,选用层次支持向量机(H-SVM)来实现多类分类,用各类数据中心代表该类数据,通过自组织特征映射神经网络(SOFM)进行聚类,把类中心之间距离较近的数据归为同一个子类进行训练,得到H-SVM层次结构。其次,在训练H-SVM中的二元分类器时,应用相对边界向量(RBV)代替全部训练样本,在保持分类精度几乎不变的条件下大幅度减少了训练样本数,使训练时间明显缩短;同时,由于支持向量的数量减小,分类时间也相应缩短。在分类数据混迭较为严重的情况下,新算法先剔除混迭的异类数据,再计算RBV,并且把与计算的RBV距离小于一定数值的样本都选择来训练SVM,保证了RBV的合理性,防止了关键数据的丢失,有效提高了分类精度。针对一个航空涡喷发动机5类复合故障的分类进行了实例仿真,总的故障分类正确率达到91.2%,二元SVM的训练时间最多只有原来的16.20%;当训练样本总数达到7500的大规模情况下,根据本算法,约减后的样本数量只有原来的3.05%。仿真结果表明,提出的算法有效、可靠,容易实现。     

正确选择数字示波器

以当前流行的数字示波器为例 ,介绍了其带宽、取样速度、存储深度、接口等一些基本概念。对如何更好的使用示波器和选择示波器都具有重要的参考借鉴价值。     

一种高准确度超低频电压测试标准——8801超低频电压及失真测试仪

8801超低频电压及失真测试仪是一种采用单片计算机、A/D变换器的采样-计算式仪器,它可以测试超低频信号的周期、电压、失真和谐波等各种参数。电压测试不确定度高达0.02％。文中介绍了主要技术指标、工作原理、技术特点、误差分析等,对软件作了简要说明。     

复合绝热管气囊成型工艺设计和实验

对两端收口复合绝热管进行了工艺方案分析,确定气囊成型工艺为优选工艺方案,在此基础上阐述了气囊成型工艺设计的原则,依据这个原则开展了典型样件的研制.结果表明：方案选择合理,工艺设计完整、可靠,产品满足各项指标要求.     

基于参数优化的截尾序贯检验法

截尾序贯概率比检验法和序贯网图检验法（SMT）是在序贯概率比检验法（SPRT）的基础上发展起来的2种方法,可进一步控制试验次数。但是这些截尾序贯方法的停止边界仍有待优化,本文以最常见的截尾SPRT为例,说明决定停止边界的参数的取值直接影响了最大试验样本量,且存在最佳取值。并以此为基础,对SMT和传统的SPRT进行了综合比较,发现截尾SMT并不优于传统的截尾SPRT,它不过是一种广义形式的截尾序贯检验法,SPRT的研究重点应该是截尾形式的构造。最后本文提供了一种曲线形式的截尾序贯检验法.     

直线度测量中采样间距与偏差形状关系的理论研究

从频域理论上对直线度测量中采样间距与偏差形状的吻合关系进行了较为深入的理论分析，提出在相同的采样误差条件下，随着采样点数的增加，测量误差将随之下降。因此，直线度测量采样间距的选择应按测量准确度要求来确定。     

A sample collector for robotic sample return missions I: Temperature effect on collected mass

Sample return is playing an increasingly important role in solar system exploration. Among the possible mission on the horizon, are sample return from asteroids, comets, the Moon and Mars. A collector initially intended for near-Earth asteroids is the touch-and-go-impregnable-pad (TGIP). Here we explore the effect of temperature on its collection capabilities. Temperatures expected on near-Earth asteroid mission targets range from −43 to 36 °C. Experiments were conducted at −75, −50, −25, 23, 65, and 105 °C. It was found that the mass of sample collected by the TGIP increased almost linearly to 23 °C and then leveled off at higher temperatures. We also found that the collector did not lose its ability to collect samples after being subjected to −75 °C temperatures (essentially frozen) and then thawed. These experiments have shown that the TGIP can operate effectively at temperatures expected on near-Earth asteroids, especially if collection is performed on the sunward side of the asteroid.     

A sample collector for robotic sample return missions II: Radiation tests

Sample return from small solar system objects is playing an increasingly important part in solar system exploration. Critical to such missions is a robust, simple, and economic sample collector. We have developed a collector such as this for near-Earth asteroid sample return missions that we have termed the Touch-and-Go Impregnable Pad (TGIP). The collector utilizes a silicone substrate that is pushed into the dust and gravel surface layer of the asteroid. As part of a systematic evaluation of the TGIP, we have investigated the resilience of this substrate to ionizing radiations. Several miniature versions of the collector, containing typically ∼3 g of the collection substrate, were exposed to 0.564 MeV beta particles from a ⁹⁰Sr source and a 6 MeV electron beam in a linear accelerator to simulate the wide range of energies of solar and galactic ionizing radiation. Various radiation levels up to eight times greater than expected on a six-year asteroid mission (in the case of beta radiation) and 50 times greater than expected (in the case of the 6 MeV electron radiation) were administered to the substrate. After irradiation, the efficiency of the substrate in collecting samples of mock regolith was compared with that of collectors that had not been irradiated. No difference beyond experimental uncertainty was observed and we suggest that the operational TGIP will not be affected adversely by radiation doses expected during a typical six-year inner solar system mission.     

深空探测器多次引力辅助转移轨道全局搜索

介绍行星引力辅助近似模型,包括近行星点无动力和有动力两种方式;采用开普勒轨道拼接法与双脉冲Lambert算法,建立多次引力辅助转移轨道的参数优化模型;通过广度优先搜索算法对引力辅助行星序列进行穷举,对发射窗口和天体间转移时间进行离散化网格搜索,每一步网格搜索后均采用合理的定界剪枝方法,减少后续计算量。这种全局搜索方法不需要提前指定引力辅助行星序列,并可得到对应不同飞行时间以及不同引力辅助次数的搜索结果,获得若干多天体引力辅助转移轨道初步结果,为进一步局部优化设计奠定基础。文章给出几组全局搜索算例,得到从地球到火星、木星和土星的多次引力辅助转移轨道,验证了全局搜索方法的有效性。