A Speech Signal Coding Method Based On Chaotic Temporal Prediction Model

In the field of signal processing and analysis,speech signal coding is not only an important research topic,but also the basis of digital communication,voice recognition and communication confidentiality.At present,the linear predictive coding(LPC)based on the correlation of speech samples has been widely used in the field of speech signal processing.However,it is found that the speech signal and its time series have complex nonlinear characteristics,the speech signal is bounded,the fractal dimension is usually not an integer,and the maximum Lyapunov exponent that the speech signal with chaotic characteristicsin the previous study.This discovery provides a new idea for the research of speech signal,which makes the research direction from the linear to the nonlinear field,and began to try to use different methods to deal with voice signals,including speech signal encoding has become the main research contents.Based on the chaotic speech signal processing,the parameters of the chaotic time series prediction model of the existing speech signal are improved,and the model is optimized by DUPSO algorithm.The core work of this paper is as follows:1.Analyze the speech signal.In this paper,we use the knowledge of chaos theory and the maximum Lyapunov exponent to analyze the speech samples,and choose the speech with chaotic characteristics as the research object of this algorithm.2.The DUPSO algorithm has been improved.In order to make the DUPSO algorithm better predict the speech signal,this paper proposes to use the minimum sample maximum error as the fitness function to ensure the accuracy of the algorithm.3.Model the speech signal prediction.The phase space is considered in the construction of the solution,and no phase space reconstruction is needed.At the same time,the concept of frame length parameter is introduced,which breaks through the limitation of fixed frame in speech signal processing and selects the optimal data frame.By modeling the optimal frame,a normalized model with good generalization force is obtained.4.Simulation analysis.The results show that the algorithm proposed in this paper can fully excavate the time series information of chaotic speech signal and establish the chaotic speech signal prediction model more accurately,which reduces the error compared with the traditional algorithm and Which provides a new idea for speech signal processing.