The Spectrum Analysis Optimization for Band-limited Discrete Spectrum Continuous Signal Based on PSO

Abstract

In the actual system the law of changing with time of continuous signal can be completely described by the signal frequency spectrum. Only choose the reasonable sampling frequency and sampling time zone width, signal spectrum can approximate continuous signal actual spectrum. Because of the frequency leakage and spectrum aliasing existence, the analysis result which has a minimum of spectrum component has the most close to the continuous signal spectrum. Through the particle swarm optimization algorithm is used to choose different sampling data amount, make sampling time length just be integer times of continuous signal sampling period (or approximate cycle), then it can completely eliminate (or partly eliminate) spectrum leakage of sampling data, thus make sampling data spectrum and the sampling continuous signal spectrum completely (or as) consistent, realize spectrum of band-limited discrete spectrum continuous signal accurate recognition.