第三代移动通信系统中非线性多用户检测技术

在CDMA系统中，多用户检测技术是提高系统容量，减少用户间干扰的关键。非线性多用户检测包括干扰消除和神经网络检测技术。其中干扰消除检测是对不同用户独立进行MAI估计，然后从接收信号中减去对该用户而言的部分或全部MAI；多用户检测从本质上看是一个组合优化问题，所有解决组合优化的算法原则上均可适用于多用户检测，其中基于神经网络的解组合优化方法当然也适用于多用户检测。本文论述了几种非线性多用户检测器的基本原理及其特点。     

径向基神经网络在优化导引律中的应用

应用神经网络的非线性的特点,在纯比例导引律的导引下,分析战机的运动轨迹,通过采样取得的一系列载机的控制输入数据,根据追踪导引控制的内在要求(目标视线角速度的绝对值减小并趋于零),设计载机导引控制器的相应的离散输出值,离线训练一个径向基神经网络模块,嵌入到载机控制回路中,作为战机导引的控制输入,实现载机在纯比例导引下的性能优化;仿真结果表明,在节能及省时方面,能达优化纯比例导引律的目的。     

一种适合飞行仿真应用的非定常非线性旋翼尾迹增强模型

基于被广泛应用的Peters—He有限状态动态尾迹模型,本提出了一种增强的非线性、非定常的旋翼尾迹模型,并混合王氏涡流理论建立了旋翼尾迹的旋转分量计算模型(桨尖轨迹平面处)。将原来的线性动态入流理论改造为非线性理论模型,并取消了计算影响系数短阵中的小角度假设。增强后的诱导速度模型适合计算低速飞行状态下旋翼平面内诱导速度,理论计算结果和实验值吻合较好。经优化的FORTRAN程序在常规单CPH微机系统中即可实现实时计算。     

利用Gaussian型RBF网络进行函数逼近的构造性估计

前馈人工神经网络有着极其广泛的应用，如何估计隐层神经元数及相应的逼近误差，一直是确定前馈网络结构的难点和关键。RBF网络是一种最重要的前馈网络，本文给出了利用Gaussian型RBF网络逼近连续函数或Lebesgue-可积函数时的构造性隐层单元数显式估算式及相应的显式逼近误差估算式。文中的结论也易于推广到离散样本的情形。这些结论对于提高Guassian型RBF在实际应用时的计算精度和减少计算量具有一定的指导意义。     

基于水平集方法的连续体结构拓扑优化方法研究

考虑到现有各种连续体结构拓扑优化方法的优缺点,提出了一种新的连续体结构拓扑优化新策略。该方法将无网格法和水平集方法结合起来,充分发挥各自的优点,使得该拓扑优化新策略具有良好的收敛性和计算精度。     

基于Log—Sigmoid型径向基神经网络的雷达辐射源识别

提出基于一种新的鲁棒性径向基 (RBF)神经网络的雷达辐射源识别方法。此网络应用Log -Sigmoid函数作为基函数 ,避免了学习过程中的不稳定状态 ,并且比传统的RBF有更好的学习性能和函数逼近能力。并介绍一种新的归一化函数 ,通过归一化函数把不同类型 ,不同量纲的原始评估数值转换到 [- 1,1]区间 ,该效用函数较好地体现了“奖优罚劣”的原则 ,同时又更有利于神经网络的训练。仿真实验证明了该方法的优越性。     

The Effect of Composite Flexures on Aeroelastic Stability of a Hingeless Rotor Blade

The effects of ply orientation angle of composite flexures on stability of hingeless rotor blade system are studied.The composite hingeless rotor blade system is simplified as a hub,a flap flexure and a lag flexure.pitch bearing and main blade.The kinematics formulations are inferred by employing the moderate deflection beam theory.The shear deformation and warping related to torsion are considered.The quasi-steady strip theory with dynamic inflow effects is applied to obtain the aerodynamic loads acting on the blade.Based on these.the set of finite element formulations of a hingeless rotor blade system is worked out.The numerical results show that the ply angle of the composite flexures has great effects on the aeroelastic stability of rotor blade.     

Reduced order model for unsteady aerodynamic performance of compressor cascade based on recursive RBF

Based on Recursive Radial Basis Function (RRBF) neural network, the Reduced Order Model (ROM) of compressor cascade was established to meet the urgent demand of highly efficient prediction of unsteady aerodynamics performance of turbomachinery. One novel ROM called ASA-RRBF model based on Adaptive Simulated Annealing (ASA) algorithm was developed to enhance the generalization ability of the unsteady ROM. The ROM was verified by predicting the unsteady aerodynamics performance of a highly-loaded compressor cascade. The results show that the RRBF model has higher accuracy in identification of the dimensionless total pressure and dimensionless static pressure of compressor cascade under nonlinear and unsteady conditions, and the model behaves higher stability and computational efficiency. However, for the strong nonlinear characteristics of aerodynamic parameters, the RRBF model presents lower accuracy. Additionally, the RRBF model predicts with a large error in the identification of aerodynamic parameters under linear and unsteady conditions. For ASA-RRBF, by introducing a small-amplitude and high-frequency sinusoidal signal as validation sample, the width of the basis function of the RRBF model is optimized to improve the generalization ability of the ROM under linear unsteady conditions. Besides, this model improves the predicting accuracy of dimensionless static pressure which has strong nonlinear characteristics. The ASA-RRBF model has higher prediction accuracy than RRBF model without significantly increasing the total time consumption. This novel model can predict the linear hysteresis of dimensionless static pressure happened in the harmonic condition, but it cannot accurately predict the beat frequency of dimensionless total pressure.     

基于分层思路的动态非线性气动力建模方法

很多非线性气动力模型难以精确预测系统的小扰动线性特征。针对这一局限,提出了一种非线性分层模型,用于辨识跨声速非线性非定常气动力。分层建模需要同时提供微幅振荡和大幅振荡两套训练样本,首先通过线性模型(如带外输入的自回归(ARX)模型)对微幅振荡样本进行建模,而后采用非线性模型(如径向基函数神经网络(RBFNN))对大幅振荡的样本与线性模型的差量进行建模,进而把线性模型和非线性模型的输出相叠加,得到分层非线性动力学模型。算例表明建立的分层气动力模型与单一自回归径向基函数(AR-RBF)神经网络模型相比不仅具有更高的数值精度,可以精确预测大幅运动中的非线性特征,而且在小扰动状态下自动退化为线性模型,能够精确刻画结构微幅振荡下的线性动力学特性。