扭转载荷下带环状裂纹圆棒的J积分的评价

研究了扭转载荷下带环状裂纹圆棒试样的J积分的评价.首先得到了一系列带有不同深度环状切口的中碳钢圆棒试样的扭矩-扭转角关系,然后分别利用单试样的简单评价法和多试样的能量法计算J积分值.结果发现,当圆棒试样的切口深度足够深(即裂纹长度足够长)或扭转角足够小时,两种方法计算的J积分值相当一致.     

前向神经网络快速学习算法在发动机模型辨识中的应用

介绍一种采用前向神经网络辨识发动机非线性模型的快速学习方法,与现有同类方法相比,该方法采用矩阵的QR分解求解线性方程组,解决了维数过高的矩阵直接求逆带来的解失真问题,并把这种方法应用于航空发动机的非线性模型辨识。算例仿真验证了算法的可行性。      

Optimal assimilation for ionospheric weather – Theoretical aspect

Observation, specification and prediction of ionospheric weather are the key scientific pursuits of space physicists, which largely based on an optimal assimilation system. The optimal assimilation system, or commonly called data assimilation system, consists of dynamic process, observation system and optimal estimation procedure. We attempt to give a complete framework in this paper under which the data assimilation procedure carries through. We discuss some crucial issues of data assimilation as follows: modeling a dynamic system for ionospheric weather; state estimation for static or steady system in sense of optimization and likelihood; state and its uncertainty estimation for dynamic process. Meanwhile we also discuss briefly the observability of an observation system; system parameter identification. Some data assimilation procedures existed at present are reviewed in the framework of this paper. As an example, a second order dynamic system is discussed in more detail to illustrate the specific optimal assimilation procedure, ranging from modeling the system, state and its uncertainty calculation, to the quantitatively integration of dynamic law, measurement to significantly reduce the estimation error. The analysis shows that the optimal assimilation model, with mathematical core of optimal estimation, differs from the theoretical, empirical and semi-empirical models in assimilating measured data, being constrained by physical law and being optimized respectively. The data assimilation technique, due to its optimization and integration feature, could obtain better accurate results than those obtained by dynamic process, measurement or their statistical analysis alone. The model based on optimal assimilation meets well with the criterion of the model or algorithm assessment by ‘space weather metrics’. More attention for optimal assimilation procedure creation should be paid to transition matrix finding, which is usually not easy for practical space weather system. High performance computing hardware and software studies should be promoted further so as to meet the requirement of large storage and extensive computation in the optimal estimation. The discussion in this paper is appropriate for the static or steady state or transition process of dynamic system. Many phenomena in space environment are unstable and chaos. So space environment study should include and integrate these two branches of learning. This revised version was published online in August 2006 with corrections to the Cover Date.     

涡轮泵振动参数与统计特征量的线性相关性分析

为了剔除冗余的振动参数和统计特征量，提高涡轮泵故障检测的实时能力，建立了线性相关性假设检验模型。利用该假设检验模型和某型液体火箭发动机的历史试车数据，检验了7路涡轮泵振动参数以及13种常用统计特征量的线性相关性，并分析了统计特征量对涡轮泵故障的敏感性和稳定性，以及涡轮泵振动数据的正态性。数据统计分析表明，大部分涡轮泵振动参数以及统计特征量显著线性相关，幅值统计特征量的故障敏感性弱，但稳定性强，而无量纲统计特征量的情况恰好相反，并且正常涡轮泵振动数据服从正态分布，而异常涡轮泵振动数据不再服从正态分布。据此，为涡轮泵故障检测选择了线性相关性弱，故障特征反映能力强的3路振动参数和3种统计特征量。     

一种新的双重门限证实数字签名方案

在已有的(t,n)门限数字签名方案及证实数字签名的基础上,提出了新的双重门限证实数字签名方案(t1,n1;t2,n2)。新方案首次在签名证实阶段引入了门限方案,由一群符合一定条件的t2个人组成证实者(confirm er)。因为是由不同的各方参与进来,方案具有了更加广泛的可信任性,从而使得该方案适用于某些安全程度要求较高的情况。     

利用大迎角试飞数据辨识飞机气动导数

基于用局部线化代替非线性概念．探索了利用大迎角飞行试验数据辨识飞机空气动力参数问题。在准定常假设条件下，提出了迎角分割算法和时间分割算法，针对歼教七飞机的失速飞行试验数据，用最小二乘回归方法和最大似然法验证了上述大迎角参数辨识的思想，取得了较好的结果，为进一步开展大迎角参数辨识技术的工程应用奠定了基础。     

导弹导引头性能的模糊综合评判研究

利用导引头系统指标的相对性、模糊性,运用模糊数学方法提出了多级模糊综合评判法;并对关键技术——隶属函数、算子和权重的确定进行了研究。实验计算表明:采用模糊综合评判法可以解决导引头性能的评估问题。     

用Ti/Cu/Ni中间层二次部分瞬间液相连接Si3N4陶瓷的研究

采用Ti/Cu/Ni中间层对Si3N4陶瓷进行二次PTLP连接,研究Ti箔厚度、连接工艺参数对Si3N4/Ti/C/Ni连接强度和界面结构的影响.结果表明Ti箔厚度对连接强度的影响是通过对反应层厚度的影响体现的;在本文试验条件下,改变二次连接工艺参数对Si3N4/Ti/Cu/Ni二次PTLP连接界面反应层厚度无明显影响,其对室温强度的影响是由于连接接头残余应力的变化所导致的;Si3N4/Ti/Cu/Ni二次PTLP连接界面微观结构为Si3N4/反应层/Cu-Ni固溶体层(少量的Cu-Ni-Ti)/Ni.     

并行鲁棒观测器对发动机减小转速信号的故障分离

通过对减小转速故障信号产生机理的研究 ,提出了一种基于多重模型假设下的并行鲁棒观测器残差输出的故障分离方法 ,建立了实际系统的正常工作状态和 4种故障工作模式 ,将该方法应用于某型航空涡扇发动机导致减小转速信号产生的传感器的故障检测与分离。数字仿真和误差分析结果显示 ,文中提出的方法可靠性高 ,故障定位准确 ,对发动机减小转速信号的故障分离行之有效。      

直升机动力舱流场及温度场的模拟

应用非结构网格,建立了直升机发动机动力舱的有限元模型,并根据舱内实际冷却气体的流通情况,采用SIMPLE算法,求解了流动和传热的控制方程,模拟了动力舱内流场与温度场的分布,研究了流场、压力场及温度场之间的依变关系。在此基础上针对现有设计存在的问题,提出了动力舱冷却系统的改进方案,并对不同方案的改进效果进行了分析比较。