Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Improved stochastic parallel gradient descent algorithm-based optical fiber coupling method

A gradient descent algorithm and fiber coupling technology, applied in the field of optical communication, can solve the problems of unstable algorithm and slow convergence speed, and achieve the effect of improving the convergence speed and parameter stability.

Inactive Publication Date: 2019-05-21
INST OF OPTICS & ELECTRONICS - CHINESE ACAD OF SCI
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is to provide a fiber coupling method based on an improved stochastic parallel gradient descent algorithm to solve the problems of slow convergence speed and algorithm instability of the existing fiber coupling using stochastic parallel gradient descent algorithm

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Improved stochastic parallel gradient descent algorithm-based optical fiber coupling method
  • Improved stochastic parallel gradient descent algorithm-based optical fiber coupling method
  • Improved stochastic parallel gradient descent algorithm-based optical fiber coupling method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The specific implementation manners of the present invention will be described in detail below in conjunction with the accompanying drawings. However, the following examples are only used to illustrate the present invention in detail, and do not limit the scope of the present invention in any way.

[0025] Such as figure 1 As shown, in the following embodiments as an example, a fiber coupling method based on the improved stochastic parallel gradient descent algorithm, a simulation program for optimizing fiber coupling, and a comparison with the traditional fiber coupling method using the SPGD algorithm.

[0026] Step 1. First, apply a random disturbance {δx} to the control quantity x of all tilting mirrors coupled with the fiber, and calculate the objective function, that is, the change quantity δJ of the optical power meter, where the random disturbance {δx} needs to satisfy the Bernoulli distribution. The update formula is as follows:

[0027] δJ=J(u 1 +δu 1 , u ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides an improved stochastic parallel gradient descent algorithm-based optical fiber coupling method. The invention aims to solve the problems of low convergence speed and poor stability of a conventional method which adopts a parallel gradient descent algorithm in optical fiber coupling. According to the method, stochastic disturbance {deltaui} is applied to the control quantityof optical fiber coupling; the change quantity of the optical power J of an objective function, which is generated after the disturbance is applied, is calculated; and the approximate gradient of theobjective function is obtained, wherein the approximate gradient is expressed by an equation that gradl<(n)>= {deltJdeltul<(n)>; the first-order moment and second-order moment of the approximate gradient are calculated, so that the adaptive gain of the stochastic parallel gradient descent algorithm is solved; the adaptive gain is adopted to replace the fixed gain of a stochastic parallel gradientalgorithm, so as to be used to iteratively update the control quantity; and the parallel gradient descent algorithm jumps out of local optimum, and convergence can be speeded up, and the stability ofthe algorithm can be increased.

Description

technical field [0001] The invention belongs to the field of optical communication technology research, in particular to an optical fiber coupling method based on an improved stochastic parallel gradient descent algorithm. Background technique [0002] As an emerging communication technology, Free Space Optical Communication has become a research hotspot due to its advantages of high bandwidth, high security, and no frequency band license. With the continuous development of optical fiber technology, more and more optical fiber technology and devices are applied to the laser communication system, which greatly improves the performance and flexibility of the laser communication end machine. In actual field communication, if the receiving end platform The jitter or the jitter of the light beam due to factors such as atmospheric turbulence will cause a mismatch between the far-field spot and the fiber mode field, resulting in a decrease in fiber coupling efficiency and affecting...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G02B6/26G02B27/00
Inventor 胡钦涛毛耀周国忠乔琦聂康周翕李志俊张超段倩文何秋农
Owner INST OF OPTICS & ELECTRONICS - CHINESE ACAD OF SCI
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products