By this ,this paper provide three training methods using weight matrix to emphasize the inherent differen.
在语音识别系统中,都是通过提取特征向量来计算待识语音与模型之间的概率或距肉,然后根据最大概率或最小距离判断待识语音的类别,对大量实验数据的观察发现:特征向量的各维对语音的表达能力是不一样的,同时特征向量在语音的时间轴上表达能力也不一样,根据这种特性,提出了三种训练算法:在训练中计算出加权矩阵,以此来加强易混淆数字间的本质区分特征,减弱随机特征,在汉语数字串识别实验中,得到了比较理想的实验结果,错误下降40。
It is the key technology for design of the constant beamwidth array to calculate the weight matrix.
恒定束宽加权矩阵的获得方法是设计此类声纳基阵的技术关键。
A direct method to detecmine based on simple explicit relation among the optimal state feedback matix and the weighting matrices is proposed This method don′t need to do recursive calculation or complex transform for solving It is an effective approach to elminate the ″bottleneck″ problem that are caused by the application of LQ optimal theory to the engineerin
基于分析系统LQ最优控制解得到的最优状态反馈矩阵与加权矩阵之间的简单显式关系,提出一个直接确定[Q,R]的方法它不需要迭代计算或复杂变换,就能求解出[Q,R],为解决LQ最优控制理论在工程应用的“瓶颈”问题提供了一条有效途径
MPC solved a constrained convex quadratic optimization by defining reference trajectories, constraint limits, prediction horizon, control horizon and weighting matrices.
通过设定参考轨迹、输入输出约束、控制步长、预测步长及加权矩阵,解决了系统的凸二次型优化问题。
Based on the Hamiltonian system\'s theory,the relationship between closed-loop poles of system characteristic equation and weighting matrices was thoroughly investigated.
根据哈密尔顿系统理论,深入研究了系统特征方程的闭环极点和加权矩阵的关系,给出了希望加权矩阵的确定方法。
Approach to coordinate driving torque of redundant actuated parallel manipulator based on weighted matrix;
基于加权矩阵的过驱动并联机构驱动力矩调节法
In order to make the designed linear meet the practical production requirement better,this paper represents the basic principle and design ways of LQR,introduces some choice rules of the weighted matrix Q and R.
为了使线性系统更好地适应实际的需要,本文简述了线性二次型最优控制器原理及设计方法,介绍了加权矩阵Q和R的一些选择规则,通过Matlab仿真讨论了参数Q和R变化对最优控制系统的影响,证实了该设计所得到的控制器效果较好,而且便于实现,达到了设计目的。
The relations between the closed-loop and the weighting matrix Q in the quadratic performance index are analysed in this paper.
本文分析了闭环极点与二次型性能指标中的加权矩阵Q之间的关系。
The method of chaos optimization,which is based on the transient response of the system,designed for weighting matrix of LQR controller is proposed,in order to solve the difficult of the choice of weighting matrix.
针对线性二次最优控制中加权矩阵难以确定的问题,引入一种系统瞬态响应这一具有直接工程意义的品质指标,提出基于logistic混沌变量的LQR权矩阵优化设计方法,为改善混沌变量大范围寻优效果不明显的不足,引入了混沌局部细搜索,找出全局最优值。
Then,the parametric formula for calculating the weighting matrix Q is given.
为了认识Butterworth最优控制的本质,揭开Butterworth最优传递函数与加权矩阵Q,R的相互关系,本文研究Butterworth最优控制的逆问题。
In this article we study the relations among D1, D2, D3, D4, which are in the reflexive g-inverse matrixM-r=D1D2D3D4of the bordered matrixM=ABC0.
该文研究加边矩阵M =ABC 0的自反广义逆M-r =D1D2D3 D4中的子矩阵D1,D2 ,D3 和D4的关系 ,还研究了矩阵 A-r C-rB-r 0 和M-r 之间的关系。
In this article we study the singularity of the bordered matrix M=ABC0 and give the structure of its reflxive g-inverses M-r=D1D2D3D40 by applying the multiple quotient singular value decomposition QQ-SVD.
作者运用多个矩阵的商型奇异值分解QQ -SVD ,研究加边矩阵M =ABC 0的奇异性 ,并给出它的自反广义逆M-r =D1 D2D3 D4的结构。
A New Method of Constructing Weighted Matrix of WDOP
一种构造WDOP中加权矩阵的新方法
Numerical Algorithm to Calculate Weighting Matrix of Constant Beamwidth Array Using Chebyshev Window
利用Chebyshev窗函数获得恒定束宽加权矩阵的数值算法
Weighted Moore-Penrose Inverse of Fuzzy Matrix
Fuzzy矩阵的加权Moore-Penrose逆
The resulted judgment matrix is a weighted sum of the original judgment matrix and the completely consistent judgment matrix.
调整后的判断矩阵等于原判断矩阵和完全一致性矩阵的加权和。
The Algorithm of Approximating a Given Matrix with a Hankel Matrix of Lower Rank and Weighted Approximations;
低秩Hankel矩阵逼近及其加权逼近的算法
Kronecker product and singular value decomposition of weighted extended matrix;
Kronecker积与加权延拓矩阵的奇异值分解
On Existence of Weighted Generalized Moore-Penrose Inverse of Matrix over Ring
环上矩阵加权Moore-Penrose逆存在的条件
Reverse Order Law for the Weighted Drazin Inverse of a Matrix Right Semi-tensor Product
矩阵右半张量积的加权Drazin逆的反序律
Weighted Matrix Padé-type Approximation and Its Applications in Control Theory;
加权的矩阵Padé-型逼近及其在控制论中应用
An Approximation Method of Positive Semi-definite Matrix Based on Weighted F-norm;
一种基于加权F-范数的半正定矩阵的逼近方法
Reverse order law for the weighted Moore-Penrose inverse of a triple left matrix semi-tensor product
三矩阵左半张量积的加权Moore-Penrose逆的反序律
Reverse Order Law for the Weighted Moore-Penrose Inverse of a Matrix Right Semi-tensor Product
矩阵右半张量积的加权Moore-Penrose逆的反序律
An Element Error Compensation STAP Approach Based on Covariance Matrix Taper
一种基于协方差矩阵加权的阵元误差补偿STAP处理
The Weighted Least Squares Solutions of Several Classes of Linear Matrix Equations and It s Optimal Approximation Problems;
几类线性矩阵方程的加权最小二乘解及其最佳逼近问题
The extension of Optimally Weighted Ls and Matrix Schwarz inequality;
最优加权最小二乘估计解与矩阵希瓦兹不等式的推广
Steganalysis for JPEG images based on weighting fusion and Markov matrix
基于加权融合和马尔可夫矩阵的JPEG图像隐写分析
Weighted Least Square Solution of Symmetric Orthogonal Anti- symmetric Matrix on the Linear Manifold
线性流形上对称正交反对称矩阵的加权最小二乘解
Query Expansion of Pseudo Relevance Feedback Based on Matrix-Weighted Association Rules Mining
基于矩阵加权关联规则挖掘的伪相关反馈查询扩展
CopyRight © 2020-2024 优校网[www.youxiaow.com]版权所有 All Rights Reserved. ICP备案号:浙ICP备2024058711号