基于随机拟合的悬浮式深弹弹阵特征建模＊

通过分析悬浮式深弹发射后在空中的弹道特性,建立了深弹弹着点坐标的计算模型。采取蒙特卡洛方法,分别对舰艇六自由度状态下,单管和六联装火箭深弹的弹着点坐标进行了仿真计算,对弹着点的分布规律进行了研究,并得出弹着点的联合密度函数。结果表明,在发射参数存在正态扰动时,弹着点散布区域均呈椭圆形分布,弹着点坐标均仍服从正态分布。     

Deconsolidation and combustion performance of thermally consolidated propellants deterred by multi-layers coating

Both heating and solvent-spray methods are used to consolidate the standard grains of double-base oblate sphere propellants plasticized with triethyleneglycol dinitrate (TEGDN) (TEGDN propellants) to high density propellants. The obtained consolidated propellants are deterred and coated with the slow burning multi-layer coating. The maximum compaction density of deterred and coated consolidated propellants can reach up to 1.39 g/cm³. Their mechanic, deconsolidation and combustion performances are tested by the materials test machine, interrupted burning set-up and closed vessel, respectively. The static compression strength of consolidated propellants deterred by multi-layer coating increases significantly to 18 MPa, indicating that they can be applied in most circumstances of charge service. And the samples are easy to deconsolidate in the interrupted burning test. Furthermore, the closed bomb burning curves of the samples indicate a two-stage combustion phenomenon under the condition of certain thickness of coated multi-layers. After the outer deterred multi-layer coating of consolidated samples is finished burning, the inner consolidated propellants continue to burn and breakup into aggregates and grains. The high burning progressivity can be carefully obtained by the smart control of deconsolidation process and duration of consolidated propellants. The preliminary results of consolidated propellants show that a rapid deconsolidation process at higher deconsolidation pressure is presented in the dynamic vivacity curves of closed bomb test. Higher density and higher macro progressivity of consolidated propellants can be obtained by the techniques in this paper.     

比例规则后件的T-S模糊模型辨识方法

模糊划分应满足测量准则,这一点常被忽略。采用余弦平方函数作为隶属函数,解决了曲线型隶属函数在模糊化分后满足测量准则问题。T-S模糊模型后件由线性函数构成,以局部线性化方式实现全局非线性,通过分段性逼近辨识曲线,常用于系统辨识中,但其辨识参数多是其主要问题。比例规则后件的T-S模糊系统具有万能逼近特性,且构成参数少,用于非线性系统的辨识中,可以减少辨识参数,提高辨识速度。给出其在系统辨识中应用的方法,并通过仿真实例说明本方法的有效性。     

抑制成像数据噪声的模糊多目标算法

以模糊数学和决策理论为基础,通过建立成像数据模糊指数函数及平方误差 模糊指数函数,提出了一种称为模糊多目标优化的新的算法模型,对成像数据的噪声进行抑 制实现数据优化。实验结果表明,所提出的模糊多目标优化算法模型有较强的抑制噪声能 力,理论和实验有较好的一致性。     

计算机控制小工具抛光技术中磨盘材料对去除函数的影响

研究了计算机控制小工具抛光(CCOP)加工中三种常用的磨盘材料对去除函数特性的影响,进一步完善材料去除模型,用以指导光学零件的加工。利用自行研制的AOCMT光学加工机床及接触式测量系统进行实验和分析。从去除函数形状、去除效率及稳定性、表面形貌三个方面进行了研究。实验结果表明:聚氨酯材料适用于粗抛光阶段;阻尼布材料适用于精抛光阶段;沥青材料适用于最后的修形加工和表面处理。     

顾及地磁影响的GNSS电离层层析不等像素间距算法

突破传统全球导航卫星系统(Global Navigation Satellite System, GNSS)层析反演研究忽略地球磁场对电离层变化活动的影响,同时顾及不同电离层高度电子密度变化较大的影响,提出在地磁坐标系下建立电离层高度方向上不等像素间距的GNSS层析反演方法。在此基础上,通过建立新的电离层层析迭代松弛因子,提高GNSS电离层电子密度层析反演结果精度。利用IRI—2007电离层模型、GNSS实测数据与电离层测高仪数据,从模拟验证和实测数据对比两方面,反演统计了不同层析算法估算的峰值电离层电子密度误差、电子密度剖面结果平均绝对百分比误差及均方根误差,验证了顾及地磁影响的GNSS电离层层析不等像素间距算法的有效性。     

Cross-linking density and aging constitutive model of HTPB coating under prestrain thermal accelerated aging

In order to study the cross-linking density and aging constitutive relationship of HTPB coating during storage, the thermal accelerated aging tests at 0%, 3%, 6% and 9% prestrains were carried out. The cross-linking density of HTPB coating at different aging stages were tested using low-field ¹H NMR and the variation of cross-linking density was analyzed. The aging model of cross-linking density considering the chemical aging and the physical stretching factors was established. The uniaxial tensile tests were carried out on HTPB coating at different aging stages and the cross-linking density was introduced into Ogden hyperelastic constitutive model as a characterization parameter of correction coefficient. Combined with uniaxial tensile test results, a prestrain aging constitutive model of HTPB coating was established. The results show that the cross-linking density of HTPB coating increases rapidly at first and then slowly with the increase of thermal accelerated aging time without prestrain. Under prestrain conditions, the cross-linking density of HTPB coating decreases at the early stage, and increases rapidly at first and then slowly at the middle and late stages of thermal accelerated aging. The correlation coefficients of aging model of cross-linking density and aging constitutive model with test results are R > 0.9500 and R > 0.9900 respectively, which can be used to accurately describe the cross-linking density and aging constitutive relationship of HTPB coating under prestrain accelerated thermal aging conditions.     

超低照度下微光图像增强神经网络损失函数设计分析

超低照度下(环境照度小于2×10~(-3)lux)微光图像具有低信噪比、低对比度等特点,使目标难以辨识,严重影响观察效果。为了提高超低照度下微光图像质量,设计了一种用于微光图像增强的卷积自编码深度神经网络,并针对传统的均方误差损失函数不符合人类视觉感知特性等问题,结合现有的全参考图像质量评价指标,研究了包括感知损失在内的几种损失函数,并提出了一种新的可微分损失函数。实验结果表明,在网络结构不发生改变的情况下,所提损失函数具有更好的性能,在提高微光图像信噪比和对比度的同时,能够有效地增强图像内部细节信息。     

COSMIC掩星电子密度廓线的质量检核

利用COSMIC掩星任务自2010年至2014年的电离层电子密度廓线,使用只考虑廓线自身特性的4种参数进行质量检核,并对廓线质量的时空分布进行分析。发现在廓线质量的空间分布上,廓线不合格率在高纬地区最高,其次是低纬地区,在中纬地区最低,这可能与电子密度分布在磁赤道附近存在赤道异常、两极地区的磁场强度最大有关。廓线质量的季节变化较明显,在南、北半球,冬、春两季的廓线不合格率均显著高于夏、秋两季。另外,廓线质量具有一定昼夜分布特性,不合格率白天明显较夜晚低,且在晨昏分界线上变化较大。合格廓线的电子密度峰值和峰值高度分布在磁赤道附近明显高于其他区域,呈现"双驼峰"现象。     

基于相关性函数的多站微动特征分析与提取

针对多站雷达精度跨度大、难以有效进行融合识别的问题,提出了基于相关性函数的多站加权融合方法。首先建立了弹道目标滑动散射模型,通过时延相乘重构回波,并利用扩展Hough变换提取出距离像的曲线参数,从而建立方程组以求取微动信息。然后利用相关性函数对各雷达的支持度进行分析,最终对支持度高的观测数据进行融合识别。仿真结果表明该方法计算简单,能有效提高微动参数的估计精度,客观地反映各部雷达的可靠度。