Analog circuit fault diagnosis method based on chaos cloud model adaptive firefly algorithm

Abstract

The present invention discloses an analog circuit fault diagnosis method based on a chaos cloud model adaptive firefly algorithm. The method comprises the steps of: applying a certain test stimulus toa tested circuit, and collecting output response signals of the tested circuit at a measurable node of the circuit; employing the wavelet fusion method to perform extraction of a circuit fault feature set from the output response signals; employing the CCAFA-LSSVM (Chaos Cloud Model Adaptive Firefly Algorithm-Least Squares Support Veotor Maohine) to perform fault diagnosis of the fault feature set to achieve classification and location operation of the circuit faults. The analog circuit fault diagnosis method employs a cloud model with a capacity of processing object fuzziness and randomnessto perform corresponding improvement of the firefly algorithm and allow the firefly algorithm to have a good generalization ability and a high robustness. Therefore, when the fault diagnosis is performed, the resolution ratio of the fault mode is high, the diagnosis performance is good, and the fault elements can be accurately located, and the diagnosis performance and the efficiency of the testedanalog circuit can be improved.

Description

technical field [0001] The invention relates to the field of analog circuit fault diagnosis, in particular to an analog circuit fault diagnosis method, in particular to an analog circuit fault diagnosis method using a chaotic cloud self-adaptive firefly algorithm. Background technique [0002] With the rapid development of integrated circuits (Integrated Circuit, referred to as IC), the complexity and density of circuits are constantly increasing. Now digital, analog and mixed signal circuits are integrated into ICs on the same substrate to form SoC (System -on-a-Chip, system-on-a-chip, referred to as SoC) and NoC (Network On Chip, network on chip, referred to as NoC). Digital circuits such as memories and controllers belong to digital IP (Intellectual Property, intellectual property, referred to as IP) cores, because they have complete test standards such as IEEE 1149.1 standards and IEEE 1500 standards, as well as their discrete output characteristics, making them Testabi...

AI Technical Summary

Problems solved by technology

Therefore, an intelligent algorithm is used to optimize its parameters, but the intelligent algorithm has a certain degree of randomness when the population is initialized, and it is easy to be limited to a local optimum or has problems such as repeated oscillations in the late iteration, which makes the searched parameters not good, resulting in Problems such as low fault diagnosis rate

Images