一种基于混沌寻优神经网络模型的语音识别矢量量化算法  

A Vector Quanization Algorithm in Speech Recognition Based on Chaotic Search Neural Network

作　　者：刘宇红[1] 刘桥[1] 任强[1] 

机构地区：[1]贵州大学信息与通信工程系,贵阳550025

出　　处：《计算机工程与应用》2003年第34期90-92,107,共4页Computer Engineering and Applications

基　　金：贵州省自然科学基金(编号:30262001)

摘　　要：矢量量化在语音识别中占有重要的地位,传统的LBG算法虽然收敛速度快,但极易陷入局部最优点。论文利用混沌运动固有的随机性与轨道遍历性等优良性质,提出了一种基于混沌寻优的Hopfield神经网络模型,并将其运用于语音识别中的矢量量化。该算法不仅收敛速度快,而且能够获得全局最优解,且初始解对算法的影响很小。实验结果表明该算法综合性能指标优于传统算法,具有较高的应用价值。Vector quanization plays an important role,traditional LBG algorithm owns the advantage of fast convergence, but it is easy to get the local optimal solution.In this paper,a model which is based on chaotic search Hopfield neural network is proprosed with the help of good properties of chaos motion,such as stochastic and ergodicity and so on,it is applied to the vector quanization in speech recognition.The algorithm not only owns the fast convergence,but it is easy to get the global optimal solution,the influence of the initial solution on the algorithm is small.The experimental results demonstrate that the algorithm is more effective than the traditional one,and possesses higher practicability.

关 键 词：HOPFIELD神经网络 矢量量化 混沌寻优 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]