Method for designing wavelet filter by means of switching current and adaptive chaotic differential evolution

Abstract

The invention relates to a method for designing a wavelet filter by means of switching current and adaptive chaotic differential evolution. The method comprises the following steps of: (1) structuring a time-domain approximating function of a basic wavelet function according to a network function approximating theory; (2) solving an optimal coefficient of the wavelet approximating function; (3) carrying out Laplace transform on the wavelet approximating function; (4) solving a network function of the wavelet filter; and (5) integrating a switch current technology to the wavelet approximating network function, that is to say, designing the wavelet filter by adopting the switch current technology, wherein an adaptive chaotic differential evolution optimized algorithm is adopted for solving the optimal coefficient of the wavelet approximating function. Compared with the prior art, in the invention, the structured time-domain wavelet approximating function has generality and is suitable for the approximation of any wavelet functions; the adaptive chaotic differential evolution optimized algorithm has the advantages of simple operating process, high convergence speed, strong global optimum capability and the like; and the wavelet filter designed by means of the switch current technology works in a current mode, has the characteristics of low voltage, low power consumption and high speed, and is completely compatible with a digital CMOS (Complementary Metal-Oxide-Semiconductor Transistor) VLSI (Very Large Scale Integrated Circuits) process.

Description

technical field [0001] The invention relates to a design method of a switching current self-adaptive chaotic differential evolution wavelet filter. Background technique [0002] Wavelet filters are widely used in signal analysis, image and speech processing, and nonlinear science, and the design of wavelet filters has received much attention. In the prior art, there are various wavelet filter design methods. In the reported design methods of switched capacitor wavelet filters, the control of the filter expansion coefficient is realized by adjusting the circuit clock frequency or capacitance ratio, but linear floating capacitors are required, which is incompatible with digital CMOS technology. At the same time, the non-ideality of the operational amplifier in the filter affects the design accuracy. In addition, due to the reduction of the size of the integrated process, the power supply voltage is continuously reduced, which will directly reduce the maximum voltage swing ac...

AI Technical Summary

Problems solved by technology

In the reported switch capacitor wavelet filter design method, the filter expansion coefficient control is realized by adjusting the circuit clock frequency or capacitance ratio, but linear floating capacitors are required, which is not compatible with digital CMOS technology

At the same time, the non-ideality of the operational amplifier in the filter affects the design accuracy

In addition, due to the reduction of the size of the integrated process, the power supply voltage is continuously reduced, which will directly reduce the maximum voltage swing acting on the switched capacitor, thereby reducing the maximum achievable dynamic range of the wavelet filter.

In the logarithmic domain wavelet filter design method, the frequency domain transfer function of the filter is often obtained by the Padé approximation method, but it is difficult to determine the polynomial degree of the numerator and denominator of the tr

Images