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

Indoor Trajectory Prediction Method Based on Bidirectional Recurrent Neural Network

A neural network and two-way loop technology, applied in biological neural network models, measurement devices, surveying and navigation, etc., can solve problems such as inability to save too much context information, data sparsity of sampling points, indoor and outdoor space differences, etc.

Inactive Publication Date: 2019-07-05
JILIN UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] In order to solve the problems existing in the existing indoor trajectory prediction algorithm, such as the difference between indoor and outdoor space, the data sparsity of the sampling points used to predict the trajectory, and the inability to save too much context information, etc.

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
  • Indoor Trajectory Prediction Method Based on Bidirectional Recurrent Neural Network
  • Indoor Trajectory Prediction Method Based on Bidirectional Recurrent Neural Network
  • Indoor Trajectory Prediction Method Based on Bidirectional Recurrent Neural Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] Step 1. Indoor space pretreatment;

[0031] (1) Divide the indoor space into a reference system composed of grids of the same size according to the set value, and number CID(x, y) in the two directions of x and y, and use the historical data trajectory to obtain the target moving indoors The grid sequence that passes through time, to judge the connectivity between the grids; for historical data, there may be sampling errors, in order to ensure the validity of historical data, set the threshold of the number of data points, if the historical data points in the grid exceed the set threshold, the historical data in the grid is considered to be valid.

[0032] Definition 1. Spatial proximity: For the grid number CID(x,y), the grid with |x-x′|≤1,|y-y′|<1 or |x-x′|<1,|y-y′|≤1 is defined as close in space.

[0033] Definition 2. Spatial connectivity: In the historical trajectory, if there are directly connected historical trajectories for spatially close grids, it is defined...

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 discloses an indoor track prediction method based on a bidirectional circulation neural network. The method comprises the following steps: step one, preprocessing an indoor space; step two, clustering the destinations of concentrated tracks obtained by indoor positioning; step three, using an indoor positioning technology to obtain the sampling point sequence of a target to be predicted; step four, calibrating the sampling point sequence; and step five, according to the characteristics of indoor track prediction problems, improving the structure of a bidirectional circulation neural network. A self-adaption gradient regulator is used to carry out training to adjust the data of an embedding list so as to minimize the deviation. The calibrated sampling point sequence is input into an improved bidirectional circulation neural network to predict the destination. The provided algorithm can well predict the track of an indoor moving target. The space connectivity problem is solved through historical track collection. The sampling point sequence calibration can guarantee that the sampling points are not influenced by sparse data, and the prediction accuracy is improved.

Description

technical field [0001] The invention relates to bidirectional cyclic neural network and indoor positioning Background technique [0002] In people's daily life, a large number of trajectories are generated every day. Although there have been many analyzes on the trajectories of moving objects in recent years, most of them are aimed at outdoor behaviors and have produced good experimental results. In fact, people also generate a large number of trajectories indoors, but the measurement methods and spatial connectivity of indoor and outdoor spaces are very different. The outdoor space is a typical space network or Euclidean space, and the indoor space is defined by a grid. Composition, and the adjacent grids in the indoor space may not be physically connected, for example, the grids of two adjacent rooms in the room are connected in space, but due to the limitation of the wall, they are not connected physically. Trajectory analysis made a big difference. At the same time, GP...

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 Patents(China)
IPC IPC(8): G01C21/20G06N3/02
CPCG01C21/206G06N3/02
Inventor 王生生岳晴
Owner JILIN UNIV
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com