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.     

Partition method for improvement of piecewise linear model with full envelope covered

Since ambient conditions vary in a wide range within the full flight envelope, the existing piecewise linear model (PLM), which is based on sea-level static condition with use of corrected parameters for other points in the flight envelope, cannot meet the accuracy for replacing the nonlinear model. To obtain more accurate linear models, a method of partitioning the flight envelope over a grid of Mach number and altitude boxes was suggested. Then, a set of linear models for a given operating condition was selected by picking the nearest (Mach number, altitude) box in the flight envelope. Through the selected set of linear models, interpolating for power level based on a weighted sum of corrected rotor speeds can obtain a linear model with acceptable accuracy. Simulation results of different points within the full flight envelope showed that the maximum error between nonlinear model and the existing PLM was more than 50%, while the maximum error between nonlinear model and the improved PLM was within 8%. It is concluded that the improved PLM performs accurately, especially under the non-standard conditions. In addition, the improved PLM can satisfy the real-time requirement better than the existing PLM.      

入侵检测中拒绝服务攻击模式分析

基于特征的IDS为了检测到攻击,必须处理能与攻击表现相匹配的攻击描述,该过程可以简化为与网络报文部分匹配的模式描述,也可以复杂化为将多传感器输出映射到抽象攻击表现的状态机描述或神经网络描述。描述了入侵检测系统(IDS)中涉及的几个典型拒绝服务攻击(DoS)模式,对这些DoS攻击模式进行了详细的分析。DoS特征是书写检测特定攻击过滤器的必要知识,详细描述了这些DoS攻击的特征,并提出了抵御DoS攻击的一个应对措施———IDS。     

复合材料机翼结构形状/尺寸综合优化设计

提出了一种复合材料机翼结构的形状／尺寸综合优化设计方法,对形状／尺寸综合优化问题进行解耦处理,得到了一种双层循环优化策略。采用均匀设计法给出若干样本点,使用准则优化手段进行尺寸优化设计,构造了形状设计变量与尺寸优化结果之间的二次多项式响应面模型,采用复合形法完成了基于响应面的形状优化设计。最后,对某型大展弦比飞翼结构进行了优化设计。优化算例结果表明,所提方法可行且有效,有较高的优化效率。     

Computations of unsteady transonic flow over airfoil at low Reynolds number

Transonic flow over a thin airfoil at low Reynolds number was studied numerically by directly solving two-dimensional full Navier-Stokes equations through 5th order weighted essentially non-oscillatory(WENO) scheme without using any turbulence model.A series of distinguished unsteady phenomena for a thin 2-D transonic airfoil flow were presented.Due to continuous adverse pressure gradient in the subsonic flow downstream of the sonic line, the unsteady separated boundary layer with main vortex and secondary vortex was developed at the rear of the airfoil.At the trailing edge,the vortex-shedding was characterized by periodical connection of the main vortex and secondary vortex on the other side of the airfoil.The unsteady separation and vortex-shedding occurred with the same period.On the airfoil surface,the average pulse pressure related to the unsteady supersonic region was obviously smaller than that related to the vortex-shedding at the trailing edge.With the attack angle increasing from 0° to 2°, the frequency of vortex-shedding decreases about 4.2%.At last, the turbulence intensity and many second-order statistics in the wake region were investigated.