A Pedestrian Trajectory Prediction Method Combining Social Force Model and Kalman Filter

A social force model and Kalman filter technology, applied in the field of pedestrian trajectory prediction, can solve the error of trajectory prediction results, without considering the relationship between pedestrians and pedestrians, pedestrians and obstacles, etc., to achieve lower errors and meet prediction requirements Effect

Active Publication Date: 2021-12-24
CHINA UNIV OF MINING & TECH
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Although there have been many achievements in pedestrian trajectory prediction, the main problem in the prediction process of existing methods is that pedestrians are regarded as general moving objects for trajectory tracking and prediction, without considering the relationship between pedestrians and pedestrians and pedestrians and obstacles. The role of the relationship
In a multi-pedestrian environment, due to the active perception of people, pedestrians will actively change the direction of movement to avoid collisions with other people or obstacles when moving towards the destination, and this sudden change in the direction of movement will lead to trajectory prediction results. big error
At present, there is no prediction method that considers the active obstacle avoidance of pedestrians

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
  • A Pedestrian Trajectory Prediction Method Combining Social Force Model and Kalman Filter
  • A Pedestrian Trajectory Prediction Method Combining Social Force Model and Kalman Filter
  • A Pedestrian Trajectory Prediction Method Combining Social Force Model and Kalman Filter

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] The present invention will be further described below.

[0036] Concrete steps of the present invention are:

[0037] 1. Kalman filter initialization;

[0038] In the tracking process, because the sampling time is very short, the movement of the target within the sampling time can be regarded as uniform motion. Therefore, the motion equation of the pedestrian is obtained as

[0039]

[0040] because

[0041] x k =AX k-1 +W k-1 (1)

[0042] Z k =HX k +V k (2)

[0043] In the formula, X k is the system state at time k, A is an n×n-dimensional state transition matrix; H is an m×n-dimensional observation matrix; W k-1 Gaussian distribution with mean zero and covariance matrix Q; V k To measure the noise, satisfy the Gaussian distribution whose covariance matrix is ​​R;

[0044] Writing formula (7) in matrix form, the state transition matrix of the system can be obtained as: system state variables The measurement value extracted in the measurement upda...

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 a pedestrian trajectory prediction method that integrates a social force model and a Kalman filter. The Kalman filter is divided into two parts: time update and measurement update; Identify various parameters of the force model; use the estimated pedestrian trajectory obtained from the simulation in step 2, and calculate the position value of the pedestrian at the next moment according to the Kalman time update formula in step 1, and finally obtain the prior estimate value X (k|k‑1) ; Calculate the pedestrian's current position measurement value Z according to the Kalman measurement update formula k , combined with the prior estimate X (k|k‑1) Calculate the optimal estimated value; set the error threshold ψ, judge the error between the predicted position of the social force model and the optimal estimated value, and make corrections to complete the trajectory prediction work. It can have a more accurate predicted trajectory when pedestrians actively avoid, turn and walk in a straight line, and effectively reduce the error with the actual trajectory, so as to meet the required prediction requirements.

Description

technical field [0001] The invention relates to a pedestrian trajectory prediction method, in particular to a pedestrian trajectory prediction method integrating a social force model and a Kalman filter. Background technique [0002] Pedestrian trajectory prediction is to estimate the pedestrian's position at the next moment or the trajectory of the pedestrian in the future based on the current movement information and historical data of pedestrians. Although human motion has great randomness, in a structured environment such as a station, human long-term motion is usually regular, manifested as a continuous trajectory connecting various entrances and exits. Pedestrian trajectory prediction has great practical value in unmanned driving, robot obstacle avoidance planning, and urban traffic management. [0003] Existing methods for pedestrian trajectory prediction are mainly divided into two categories: one is data-based modeling methods, such as human motion prediction algor...

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): G06T7/277G06N3/00
CPCG06N3/006G06T7/277G06T2207/30241G06T2207/30196
Inventor 杨春雨汤瑶汉汪芸尤龙卢铁
Owner CHINA UNIV OF MINING & TECH
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