基于支持向量机的组合分类方法及应用

为了解决采用神经网络、决策树作为弱分类器的AdaBoost组合分类存在的不足,进一步改善组合分类效果,提出采用支持向量机(SVM)作为弱分类器的一种新的组合分类诊断方法——AdaBoost-SVM。该方法没有采用一个固定的SVM的核参数,而是自适应调整SVM中的核参数,从而得到一组有效的SVM弱分类器。通过对基准数据库的测试及航空发动机故障样本的诊断,结果表明,所提AdaBoost-SVM方法较好地解决了现有的Ada-Boost组合分类方法中存在的弱分类器本身参数选取困难问题及训练轮数的合理选取问题,并具有更好的泛化性能,更适合对分散程度较大、聚类性较差的航空发动机故障样本进行分类。     

航空发动机进口温度畸变参数和防喘控制系统设计

主要研究了航空发动机防喘控制系统的设计和研制，特别是研究了当战斗机发射导弹时温度畸变对发动机的影响。引入的温度畸变参数包括：进口相对平均温升δT；进口温升率dT／dt；当量热区角Φe；连续温度畸变时间tB。通过试验和数据分析的方法得出发动机防喘控制系统的设计准则，发动机防喘控制系统的研制和飞行试验证明了该设计准则的正确性。     

现代民用飞机的发动机火警探测系统设计分析

发动机火警探测系统是现代飞机一个很重要的组成部分，它能否准确及时的探测到发动机火警，不仅影响到飞机的正常营运，还直接危及到飞行安全。本文分析了现代飞机的发动机火警探测系统的总体设计，探测器原理和处理探测器信号的判断逻辑和系统设计特点。     

航空发动机结构参数和非结构参数不确定系统鲁棒H∞输出反馈控制

分析了航空发动机混合不确定性系统的鲁棒稳定性和鲁棒性能;引入结构奇异值μ综合法,研究了结构参数和非结构参数混合摄动下的鲁棒H∞输出反馈控制;对航空发动机不确定性系统设计的控制器在双转子涡喷发动机气动热力学非线性模型上。      

基于主元核相似度免疫机制的故障诊断方法及应用

1引言目前,国内外已发展了多种应用于航空发动机的智能故障诊断方法,如专家系统方法、神经网络方法,以及新近提出的基于支持向量机的诊断方法。但这些方法需要有足够的已知故障模式的训练样本才能发挥出优异的性能,而且算法实时性较差,难以满足快速在线诊断的要求[1]。传统的基     

基于FSVM改进隶属度的发动机振动性能分析

为了更有效地掌握航空发动机振动性能的影响因素,提出了改进FSVM信息熵的融合定量分析方法.首先,对模糊支持向量机(FSVM)的模糊隶属度函数进行改进,建立多类模糊隶属度计算模型.再将该方法应用到航空发动机整机振动性能评估,计算出振动故障模式与故障原因之间的权值,建立了一个多参数的发动机振动性能分析模型;并对各类振动原因对发动机整体性能的影响进行定量分析,为发动机的振动抑制提供量化参考指标.最后,通过与实际经验作比较,验证了该方法是可行和有效的.     

An extrapolation approach for aeroengine’s transient control law design

Transient control law ensures that the aeroengine transits to the command operating state rapidly and reliably. Most of the existing approaches for transient control law design have complicated principle and arithmetic. As a result, those approaches are not convenient for application. This paper proposes an extrapolation approach based on the set-point parameters to construct the transient control law, which has a good practicability. In this approach, the transient main fuel control law for acceleration and deceleration process is designed based on the main fuel flow on steady operating state. In order to analyze the designing feature of the extrapolation approach, the simulation results of several different transient control laws designed by the same approach are compared together. The analysis indicates that the aeroengine has a good performance in the transient process and the designing feature of the extrapolation approach conforms to the elements of the turbofan aeroengine.     

基于奇异值趋势分解和LS-SVR的航空发动机性能指数组合预测

发动机性能指数是衡量其性能优劣的重要指标之一。针对发动机性能指数具有非线性、非平稳的特征,引入多层次多尺度的思想,在此基础上提出一种基于奇异值趋势分解的组合预测方法。利用奇异值趋势分解提取原始数据的趋势项和波动项;以改进粒子群算法分别获取趋势项和波动项在最小二乘支持向量回归模型中的最佳参数组合(嵌入维数、延迟时间、惩罚因子、核参数),并引入回归移动的思想,在此基础上利用最佳的最小二乘支持向量回归模型进行预测。预测结果表明预测精度明显增加,计算时间也相对减少。提前预测步长在5步之内时,精度变化不大;步长超过10步,精度下降很快。与不同预测方法比较,证明了方法的有效性。      

Dynamic analysis of C/C composite finger seal

A seal device as an important component of aeroengines has decisive influence on per- formance, reliability, and working life of aeroengines. With the development of aeroengines, demands on the performance characteristics of seal devices are made strictly. Finger seal as a novel kind of sealing device, recently attracts more and more attentions in academic circles and engineer- ing fields at home and abroad. Research on finger seals has been extensively developed, especially on leakage and wear performances under dynamic conditions. However, it is a pity that the work on finger seals has been limited with a single approach that is improving the performance by structural optimization; in addition, the technology of dynamic analysis on finger seals is weak. Aiming at the problems mentioned above, a distributed mass equivalent dynamic model of finger seals considering the coupling effect of overlaid laminates is established in the present paper, the dynamic perfor- mance of 2.5 dimension C/C composite finger seal is analyzed with the model, and then the effects of fiber bundle density and fiber bundle preparation direction on finger seal＇s dynamic performance are discussed, as well as compared with those of Co-based alloy finger seal. The current work is about dynamic analysis of finger seals and application of C/C composite in this paper may have much academic significance and many engineering values for improving research level of finger seal dynamics and exploring feasibility of C/C composite being used for finger seals.     

航空发动机叶片关键制造技术研究进展

以金属和复合材料航空发动机叶片为研究对象,论述、分析了航空发动机叶片制造采用的加工工艺及关键制造技术,并对其优缺点及发展趋势进行了简要分析。