外文翻译原文--一种基于混合粒子群优化算法的一维相检索.pdfVIP

A Hybrid Particle Swarm Optimization-Based Algorithm for 1-D Phase Retrieval Guangbin Liu, Yingfeng Qiu, and Feng Chao Xi’an Research Inst. of Hi-Tech Measurement and Control Department Xi’an City, P.R.China xaqiuyf@163.com Abstract plication. The 1-D phase retrieval problem is to reconstruct the The previous solutions for 1-D phase retrieval problem phase given the modulus of its Fourier transform. Equiv- usually employed an iterative algorithm applied on a dis- alently, the original signal is synthesized by applying the crete approximation of a signal. The utilization of these inverse Fourier transform. It can be described as follows. algorithms is seriously limited by the unpredictability of The original signal is represented by the complex function their convergence, stagnation and abundant computation. of a real variable: X (ω) = m(ω)exp(j φ(ω)), where m(ω) In this paper, a particle swarm optimization algorithm is in- denotes the modulus and φ(ω) the phase. Given m(ω), we troduced to solve this problem. Firstly, 1-D phase retrieval need to determine φ(ω). was regarded as a constrained nonlinear optimization prob- A 1-D signal cannot be uniquely restored only from its lem. Secondly, it was converted to an unconstrained nonlin- Fourier modulus. Indeed, there are infinite possibilities. ear optimization problem by using penalty function method. In order to determine a unique solution, more information Finally, a particle swarm optimizati