基于分布参数线性化模型的分级燃烧循环液体火箭发动机频率特性计算

采用一维分布参数模型描述发动机的管路系统 ,将发动机各部件和管路的线性化模型表示为联系输入输出参数的传输矩阵形式。根据发动机各部件的连接关系 ,将发动机系统划分为若干模块 ,用矩阵变换的方法求得发动机各模块的传输矩阵 ,构成发动机系统方程组 ,用数值求解的方法计算了某分级燃烧循环液氧

基于线性化参数模型的三轴磁场传感器校准方法

针对三轴磁场传感器零偏误差、刻度因子不一致性误差和非正交误差等,建立了磁场传感器的三个敏感轴与参考正交坐标系的转换矩阵;基于标量校准思想,推导了传感器三轴测量值与磁场标量值之间的线性化参数模型;根据最小二乘原理,进行传感器模型参数辨识。选取两款三轴磁通门磁力仪进行了误差校准实验研究,校准后,传感器的测量误差可以降低一个数量级。在实验过程中,不需要高精度的实验设备,在稳定地磁场环境下即可进行。实验研究表明,此方法是一种方便易行、精度较高的传感器校准方法。

新安江模型参数的线性化率定

为了解决一般概念性水文模型参数率定结果不稳定的问题,以新安江模型为例,提出了新安江模型日模参数的线性化率定方法。首先通过理想模型将该方法与SCE-UA方法及单纯形方法进行对比研究。率定结果中3种算法所获得的平均目标函数值分别为0.02、0.10、8.39m3/s,平均循环次数分别为8、637、327,而且由线性化率定方法所获得的各参数值方差也要比其他2种方法小得多;说明线性化率定方法能够找到参数真值,计算精度和效率更高,率定结果更稳定。然后采用建阳和长滩河2个流域的实测资料对该方法进行应用检验,结果表明,同样可以较快地率定出稳定的模型参数优值。建阳流域的10组率定结果的目标函数值皆为100.35m3/s,循环次数也皆在8次以内,而且2个流域检验期的径流深相对误差皆在9.68%以内,确定性系数皆在0.819以上。因此,线性化参数率定方法确实能够解决非线性模型参数率定结果不稳定的问题,不会产生不相关的局部参数优值,并且不受参数初值影响,计算精度高,循环次数少,是一种可行有效的全局优值参数优选方法。

二次参数方程组的线性化模式比较

线性化（关于参数）是求解二次参数方程组（λ^2A＋λB＋C）x（λ）=f的有效途经．采用不同线性化模式，对计算会产生不同效果．本文就参数的取值，矩阵的结构与性质对线性化模式的计算所产生的可能影响进行了分析．数值实验显示了不同模式的不同计算效果．．

摩擦摆支座等效线性化参数及隔震设计方法研究

首先推导了摩擦摆支座的水平和竖向恢复力公式,讨论了水平位移对竖向恢复刚度的影响;其次针对目前摩擦摆支座隔震设计、分析中存在的不足,重点讨论了基于规范设计反应谱迭代求解其等效线性化参数的实施流程和需要注意的问题;提出了一种综合隔震周期目标和摩擦力影响粗定等效曲率半径,并根据静轴力分布情况选择摩擦系数大小的摩擦摆支座隔震方案布设方法;最后结合实际工程案例,对比分析了橡胶支座隔震方案和摩擦摆支座隔震方案在不同组合工况下对支座内力、位移的隔震效果及其分布情况。结果表明:两种隔震方案均能达到降度设计的隔震目标,后者由于受三向耦合影响,各支座上的内力分布较前者的离散性更大;摩擦摆支座有可能出现受拉现象,应用中需注意设置抗拉构造措施;摩擦摆支座隔震方案中应重视空间扭转问题和控制支座抬升量。

基于采样稀疏学习的非线性动力系统辨识

现有系统辨识方法常将非线性动力系统辨识问题转化为稀疏回归问题。针对现有方法应对噪声与小样本能力不足的问题,提出了一种基于采样稀疏学习的非线性动力系统辨识方法。首先,在复杂辨识问题被描述为稀疏回归问题的基础上,引入指示变量对稀疏结构进行编码;然后,基于嵌有Metropolis-Hastings算法的Gibbs采样器对稀疏结构进行学习;最后,通过简单的线性回归获得动力学方程的各项系数。实验结果表明,相较于LASSO算法,采样稀疏学习算法在较高的噪声水平下仍然保持较好的性能。此外,该算法无须进行阈值设定或正则化参数调整,具有更高的灵活性与鲁棒性。

经济计量非线性模型的一个注记

在非线性经济计量模型的估计中 ,当总体方差σ2 较大时 ,用参数线性化的逆变换代回而得的估计值不是原模型的参数估计值 ;

快速路入口匝道的的非参数自适应迭代学习控制

基于快速路交通系统重复性和周期性的特征,引入"拟伪偏导数"概念,给出了宏观交通流模型沿迭代轴的非参数动态线性化形式.进一步,提出了快速路入口匝道的非参数自适应迭代学习控制(NP-AILC)方案.该控制方法本质上是无模型的,并且学习增益可迭代调节.收敛性分析表明当系统初始状态随迭代次数随机变化时,该方法可实现几乎完全跟踪性能.仿真结果进一步验证了方法的有效性.

一类热传导方程非线性源项识别问题

利用单调性估计方法证明了一类热传导方程非线性源项识别问题的可识别性 ,并运用函数逼近理论 ,把对非线性源项的识别问题转化成一常系数参数识别问题 ;给出了一种对非线性源项识别的可实现的线性化迭代算法及相应的数值实例 .

饱和土体固结变形特征的一种非线性描述

基于非线性假定的固结方程及其标准化形式,给出一个描述饱和土体非线性固结变形特征的实用的理论模型。在假定孔隙比与有效应力关系的基础上,研究土层固结度随时间因素的演化过程,进而分析饱和土体固结变形的内在规律性,指出其与经典Terzaghi渗透固结变形理论的差异。其中,定义了一个表征土体非线性特征的重要参量λ,分析该参数的大小以及土层厚度的影响,从而较好地解释和描述不同厚度土层(或不同的排水条件)固结形态的差异性。利用理论分析结果,对一些典型试验结果进行对照分析,说明上述理论的合理性。

基于模糊神经网络的火电厂锅炉主蒸汽温度控制方法

现行方法在火电厂锅炉主蒸汽温度控制中应用效果不理想,阶跃响应性能较差,而且超调量较高,因此本文提出基于模糊神经网络的火电厂锅炉主蒸汽温度控制方法。利用线性化分布参数建模方法建立火电厂锅炉主蒸汽温度参数数学模型,利用模糊神经网络进行锅炉主蒸汽温度误差补偿控制,实现基于模糊神经网络的火电厂锅炉主蒸汽温度控制。经试验证明,本文方法阶跃响应时间较短,超调量在1℃以内,在工业控制领域具有良好的应用前景。

五自由度气浮平台的自适应控制方法研究

针对五自由度气浮微重力平台动力学参数难以确定、重心漂移这两大应用难点,该文推导了线性化的动力学模型,并且使用自适应控制方法完成了参数自整定以及误差的渐进收敛。在实物实验中针对某一给定的跟踪任务,对比控制力矩法和自适应方法的结果表明,带有参数自适应的控制率不仅使得系统参数收敛,同时也极大提高了控制品质。虽然参数自适应方法已经在众多领域得到了应用,但是大多控制对象待辨识的参数数量较少或具备天然的参数线性结构,因此,该文是自适应控制在高维参数辨识问题上的一次重要实践,同时也证明了该文所推导的动力学方程与自适应率的正确性。

熔化极气体保护焊中弧长系统的改进无模型自适应控制

针对熔化极气体保护焊中电流与弧长非线性时变系统难以建立精确的模型,且在焊接过程中,系统存在各类不确定因素并易受到内外界扰动的问题,提出一种改进的无模型自适应控制方法。首先,设计了带有可测与不可测干扰以及不确定因素的动态线性化泛模型,采用参数估计算法在线估计伪偏导数,引入干扰观测器实时补偿逼近内外界扰动和不确定因素;进而,通过构建准则函数设计出弧长系统的控制律通,在此基础上,得到改进的无模型自适应控制方案;最后,证明了该方法的鲁棒稳定性。仿真结果表明:与现有无模型自适应控制方案相比,所提出方法具有良好的跟踪性能和更快的响应速度;当焊接过程中遇到可测和不可测干扰以及不确定因素时,所提出的控制方法具有更强的鲁棒性,从而确保了焊接过程中弧长的稳定和较好的控制品质。

富春江水库区间洪水模拟站网优化研究

建于20多年前的富春江水库流域水文站网,暴露出了一些问题。为了提高洪水预报模型的精度,降低坝址区间洪水突发性的威胁,必须对站网进行优化。以富春江水库流域为例,主要考虑兰溪到富春江大坝区间区域,采用变动雨量站权系数的方法,设定不同的雨量站布设方案,代入新安江模型进行计算,采用线性化参数率定方法进行参数率定,以平均洪量误差最小和洪水场次合格率最高为目标函数,进行站网优化。结果表明,为了提高洪水预报精度,对于降雨空间变化方差大的洪水,应尽量多设站点。综合考虑富春江水库流域的实际情况,选择了区间布设15个雨量站的方案。

Double-Parameter Solutions of Projective Riccati Equations and Their Applications

The purpose of the present paper is twofold. First, the projective Riccati equations （PREs for short） are resolved by means of a linearized theorem, which was known in the literature. Based on the signs and values of coeffcients of PREs, the solutions with two arbitrary parameters of PREs can be expressed by the hyperbolic functions, the trigonometric functions, and the rational functions respectively, at the same time the relation between the components of each solution to PREs is also implemented. Second, more new travelling wave solutions for some nonlinear PDEs, such as the Burgers equation, the mKdV equation, the NLS^＋ equation, new Hamilton amplitude equation, and so on, are obtained by using Sub-ODE method, in which PREs are taken as the Sub-ODEs. The key idea of this method is that the travelling wave solutions of nonlinear PDE can be expressed by a polynomial in two variables, which are the components of each solution to PREs, provided that the homogeneous balance between the higher order derivatives and nonlinear terms in the equation is considered.

Distribution of Vertical Turbulent Mixing Parameter Caused by Internal Tidal Waves and Solitary Waves in the South Yellow Sea

Many observations show that in the Yellow Sea internal tidal waves (ITWs) possess the remarkable characteristics of internal Kelvin wave, and in the South Yellow Sea (SYS) the nonlinear evolution of internal tidal waves is one of the mechanisms producing internal solitary waves (ISWs), which is different from the generation mechanism in the case where the semidiurnal tidal current flows over topographic drops. In this paper, the model of internal Kelvin wave with continuous stratification is given, and an elementary numerical study of nonlinear evolution of ITWs is made for the SYS, using the generalized KdV model (GKdV model for short) for a continuous stratified ocean, in which the different effects of background barotropic ebb and flood currents are considered. Moreover, the parameterization of vertical turbulent mixing caused by ITWs and ISWs in the SYS is studied, using a parameterization scheme which was applied to numerical experiments on the breaking of ISWs by Vlasenko and Hutter in 2002. It is found that the vertical turbulent mixing caused by internal waves is very strong within the upper layer with depth less than about 30m, and the vertical turbulent mixing caused by ISWs is stronger than that by ITWs.

Identification of LPV Models with Non-uniformly Spaced Operating Points by Using Asymmetric Gaussian Weights

In this paper, asymmetric Gaussian weighting functions are introduced for the identification of linear parameter varying systems by utilizing an input-output multi-model structure. It is not required to select operating points with uniform spacing and more flexibility is achieved. To verify the effectiveness of the proposed approach, several weighting functions, including linear, Gaussian and asymmetric Gaussian weighting functions, are evaluated and compared. It is demonstrated through simulations with a continuous stirred tank reactor model that the oroposed aonroach nrovides more satisfactory aonroximation.

Design parameter optimization of beam foundation on soft soil layer with nonlinear finite element

Finite element method was performed to investigate the influences of beam stiffness, foundation width and cushion thickness on the beating capacity of beam foundation on underlying weak laminated clay. The comparison between numerical results and results from field test including plate-bearing test and foundation settlement observation shows reasonable agreement. According to the numerical results, the beam width, length, cross section and cushion thickness were optimized. The results show that the stresses in subgrade soil decrease greatly with increasing the cushion thickness and width of foundation. However, the foundation settlement and influencing depth of displacement also increase correspondingly under conditions of relatively thinner cushion thickness. For the foundations on underlying weak layer, increasing foundation width merely might be inadequate for improving the bearing capacity, and the appropriate width and cushion thickness depend on the response of subgrade. A comparison between rigid and flexible beams was also discussed. The influence of a flexible beam foundation on subgrade is relatively smaller under the same loading conditions, and the flexible beam foundation appears more adaptable to various subgrades. The proposed flexible beam foundation was adopted in engineering. According to the calculation results, beam width of 2.4 m and cushion thickness of 0.8 m are proposed, and a flexible beam foundation is applied in the optimized design, which is confirmed reasonable by the actual engineering.

Parameters identification of the compound cage rotor induction machine based on linearized Kalman filtering

A mathematical model has been built up for compound cage rotor induction machine with the rotor resistance and leakage inductance in the model identified through Kalman filtering method. Using the identified parameters, simulation studies are performed, and simulation results are compared with testing results.

Vibration transmissibility characteristics of smart spring vibration isolation system

The objective of this work was to study the vibration transmissibility characteristics of the undamped and damped smart spring systems. The frequency response characteristics of them were analyzed by using the equivalent linearization technique, and the possible types of the system motion were distinguished by using the starting and ending frequencies. The influences of system parameters on the vibration transmissibility characteristics were discussed. The following conclusions may be drawn from the analysis results. The undamped smart spring system may simultaneously have one starting frequency and one ending frequency or only have one starting frequency, and the damped system may simultaneously have two starting frequencies and one ending frequency. There is an optimal control parameter to make the peak value of the vibration transmissibility curve of the system be minimum. When the mass ratio is far away from the stiffness ratio, the vibration transmissibility is small. The effect of the damping ratio on the system vibration transmissibility is significant while the control parameter is less than its optimal value. But the influence of the relative damping ratio on the vibration transmissibility is small.