Particle filter weight processing and resampling method based on FPGA

A particle filter and resampling technology, applied in impedance networks, digital technology networks, electrical components, etc., can solve the problems of complex particle filter algorithm structure, expansion of calculation, and large amount of calculation, and achieve accurate calculation results and improve efficiency. performance, and the effect of meeting performance requirements

Inactive Publication Date: 2014-06-11
SHANGHAI UNIV +1
View PDF1 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Particle filter has great advantages in dealing with parameter estimation and state filtering problems of nonlinear non-Gaussian time-varying systems, but the particle filter algorithm has a complex structure and a large amount of calculation. When the dimension of the system state space increases, the calculation of particle filter The amount will expand several times, how to apply particle filter to real-time system is the biggest problem facing

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
  • Particle filter weight processing and resampling method based on FPGA
  • Particle filter weight processing and resampling method based on FPGA

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0024] see figure 1 and figure 2 , this FPGA-based particle filter weight processing and resampling method is characterized in that the following two steps:

[0025] 1) Parallel processing of particle weight sorting:

[0026] In the particle filter algorithm, the particle number M is selected as an even number, and the Barrier distance between the particle feature and the target feature is calculated as the weight; after obtaining the weight of each particle, the weight normalization operation is not performed, but the particle The weights are sorted; the basic method of the particle weight sorting algorithm is as follows:

[0027] The sorting algorithm requires the particle weights to be arranged in order from large to small. The hardware sorting algorithm includes two sorting state machines, state 1 and state 2: in state 1, the particle with an odd number is compared with the even-numbered particle with a sequence number one bit lower than it. If the weight of odd-number...

Embodiment 2

[0031] This embodiment is basically the same as Embodiment 1, and the special features are:

[0032]In the parallel processing of the sorting of weights, two state machines are included, including M / 2 comparators and controllers; the comparators are connected to adjacent particle weights, and their sizes are compared; the controllers are connected to compare The device exchanges the weight and position information of the two particles according to the result of the comparator; the particle weight sorting algorithm uses the parallel design of the hardware system, so that M / 2 comparison and exchange operations can be performed in each cycle, and Compared with the single cycle of the traditional PC sorting algorithm, there is only one comparison and exchange operation, which improves the efficiency of sorting operations.

[0033] In the operation process of adaptive resampling according to weight distribution, an N h threshold register ， a N l threshold registers and a FIFO bu...

Embodiment 3

[0037] This FPGA-based particle filter weight processing and resampling method includes two steps: parallel processing of particle weight sorting and adaptive resampling according to particle weight distribution:

[0038] 1) Refer to the parallel processing method of particle weight sorting figure 1 :

[0039] The particle weight sorting algorithm in the figure takes the particle number M as 6, and calculates its weight. The numbers in the box represent the particle weight, and the serial numbers of the particles are marked by Roman numerals Ⅰ~Ⅵ. A group of dotted arrows represent Perform compare and exchange operations.

[0040] The particle weight sorting algorithm contains two states: In state 1, the particle with the odd number is compared with the even-numbered particle whose serial number is one bit smaller than it with a comparator. If the weight of the odd-numbered particle is less than the weight of the even-numbered particle, then The weight and position informatio...

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 a particle filter weight processing and resampling method based on an FPGA. The particle filter weight processing and resampling method has the advantages that on one hand, a weight normalization process is abandoned, parallelization sorting processing is carried out on weights, all particles can be compared and exchanged by M / 2 comparators and controllers within one period, and the comparison and exchanging efficiency of a sort algorithm is improved; on the other hand, a threshold value of a resampling part is set according to permutation distribution of the weights, the large-weight particles, the middle-weight particles and the small-weight particles are determined through Nh and N1, and the threshold value is not a preset threshold value. The large-weight particles and the middle-weight particles are reserved in the resampling process, the small-weight particles are eliminated, the weights and the position information of the large-weight particles are input into an FIFO, one circuit of the weights and the position information are output through an FIFOOUT to replace a weight of a small-weight particle register, the other circuit of the weights and the position information are input into the FIFO, and the function of copying the large-weight particles is achieved. By means of the particle filter weight processing and resampling method, the accuracy of target position calculation, the diversity of particle samples and the high particle replacement efficiency in the hardware environment are guaranteed.

Description

technical field [0001] The present invention relates to the field of digital image processing, in particular, the present invention relates to an FPGA-based particle filter weight processing and resampling method, which belongs to the field of electronic information. Background technique [0002] Particle filter is a method based on Monte Carlo method and recursive Bayesian estimation, which is widely used in motion estimation problems in state space, especially in filtering problems of nonlinear and non-Gaussian systems. The basic idea of ​​particle filtering is: according to the empirical conditional distribution of the system state vector, a group of random sample collections called "particles" are generated by sampling in the state space, and the particle's state is continuously corrected in each recursion through the actual observation value. Weight size and sample position, and finally adjust the initial empirical condition distribution through the adjusted particl...

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): H03H17/02
Inventor 何康曹向明叶武郭旭雷陆小锋陆亨立
Owner SHANGHAI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap