A chaos theoretical exponent value calculation system

Abstract

The present invention provides a system for analyzing a time series signal by a method of Chaos Theory and calculating a chaos theoretical exponent value. It is a chaos theoretical exponent value calculation system comprising: a means for receiving an input of predetermined parameters, a means for reading a time series signal, a means for cutting out from the read time series signal a time series signal for each processing unit x=x(i), a means for calculating a chaos theoretical exponent value of the read time series signal, and a means for outputting a chaos theoretical exponent value of the calculated time series signal, wherein the means for calculating a chaos theoretical exponent value comprises: a means for calculating a chaos theoretical exponent value for a sampling time in a time series signal of the cut-out processing unit x=x(i), and a means for calculating, based on the chaos theoretical exponent value for the sampling time, a chaos theoretical exponent value of a time series signal for a predetermined time.

Description

TECHNICAL FIELD [0001] The present invention relates to a system for analyzing a time series signal by a method based on Chaos Theory and calculating a chaos theoretical exponent value thereof. BACKGROUND ART [0002] There are chaos theoretical exponent values in accordance with the Chaos Theory such as a correlative dimension, KS entropy, Lyapunov exponent and the like. The Lyapunov exponents, which are relatively easier to be calculated, are used for the assessment of chaoticity of a phenomenon which gives a time series signal. It is common to analyze a time series signal, in particular a signal having periodic characteristics such as a speech voice, by calculating a first Lyapunov exponent or a Lyapunov spectrum. [0003] The maximum Lyapunov exponent, or the first Lyapunov exponent in the Lyapunov spectra, is calculated in general in order to determine the chaoticity of a time series signal (here the chaoticity means the characteristics of fluctuations or the characteristics due to...

AI Technical Summary

Benefits of technology

[0021] As have been described above, the inventors of the present invention have invented a system which makes it possible to calculate a chaos theoretical exponent value that could not have been so far processed in a dynamics-changing system and to perform the process thereof at a high-speed and on a real-time basis. In addition, the inventors have invented a system which makes it possible to calculate a chaos theoretical exponent value even from a time series signal which includes noises.

[0033] In addition, by using the inventions set forth in claims 6 and 7, a high precision calculation of a chaos theoretical. exponent value can be made. When the invention set forth in claim 8 is used, because it is not needed to sort microscopic chaotic index values in the processing unit for calculating a macroscopic chaos theoretical exponent value, it makes it possible to perform a high speed calculation of a chaos theoretical exponent value.

[0036] In addition, the system in accordance with the invention described above, which generates the neighborhood points set (or the formal neighborhood points set) from the periodicity condition, is able to generate the neighborhood points set much faster than the conventional methods which generate the neighborhood points set as a set of points satisfying the neighborhood condition from the entire processing units, allowing to significantly shorten the time for calculating the chaos theoretical exponent value when compared with the time for calculating the first Lyapunov exponent by a conventional technique.

[0037] For example, when processing the speech voice signals sampled at 44.1 kHz, and when compared with the case of conventional techniques where the processing unit time is set to one second to calculate an average first Lyapunov exponent by a moving average processing with the processing unit starting at each sample, the system in accordance with the present invention calculates the average chaos theoretical exponent value in a shorter time of two decimal orders or more.

[0038] In the conventional techniques, the neighborhood points set is generated from the entire processing units as a set for satisfying the neighborhood condition, so that a stable processing result is not always obtained when the dynamics changes within a processing unit. The system in accordance with the present invention, in contrast, which generates the neighborhood points set (or the pro form a formal neighborhood points set) from the periodicity condition, and which also allows the application of the neighborhood condition or the convergent calculation continuity condition in addition to the periodicity condition, makes it possible to calculate the chaos theoretical exponent value when a stable dynamics is present and to obtain temporally local processing result far more stable than ever.

[0044] The calculated chaos theoretical exponent values can be visualized to be displayed by the inventions in accordance with claims 10 and 11. This allows not only to calculate and to output as numerical data, but also to visualize so as to facilitate the understanding of a decision maker using the chaos theoretical exponent value. In particular, the foregoing allows the visualization of chaos theoretical exponent values at a temporally higher resolution, thereby ultimately improving the precision.

Problems solved by technology

In addition, as one of problems, the system is characterized in that it cuts out the section where the signal waveforms (voice signal waveforms) and the like are locally stable, in the cutting out of data for a processing unit from a time series signal.

Images