Method and apparatus for positioning vehicle

Patent metadata

Field	Value
Publication number	US-11009355-B2
Application number	US-201815876008-A
Country	US
Kind code	B2
Filing date	Jan 19, 2018
Priority date	Apr 17, 2017
Publication date	May 18, 2021
Grant date	May 18, 2021

Abstract

The present disclosure discloses a method and apparatus for positioning a vehicle. In some embodiments, the method comprises: acquiring an a priori position of a to-be-positioned vehicle at a current positioning moment determined by performing a strapdown calculation between a previous positioning moment and the current positioning moment; determining a map area for searching in a laser point cloud reflected value map; matching a reflected value characteristic of a projection area generated by projecting a real-time laser point cloud, to obtain a map matching position according to a matching result; positioning, using the a priori position in combination with observation data of a vehicle-mounted global navigation satellite system (GNSS) receiver of the vehicle, to obtain a satellite positioning position; and fusing the a priori position, the map matching position and the satellite positioning position to generate a positioning result of positioning the vehicle at the current moment.

First claim

What is claimed is:

1. A method for positioning a vehicle, the method comprising:
acquiring a priori position of a to-be-positioned vehicle at a current positioning moment determined by performing a strapdown calculation between a previous positioning moment and the current positioning moment by a vehicle-mounted inertial navigation system of the to-be-positioned vehicle;
determining a map area, having a predetermined range, for searching in a laser point cloud reflected value map by using the priori position, the laser point cloud reflected value map being a map generated based on reflected value characteristics of projection areas after projecting pre-collected laser point clouds;
matching a first reflected value characteristic of a projection area generated by projecting a real-time laser point cloud collected by scanning a surrounding environment through a vehicle-mounted Light Detection and Ranging (lidar) of the vehicle at the current positioning moment with a second reflected value characteristic of the map area for searching to obtain a matching result, and determining a map matching position according to the matching result;
positioning, using the priori position in combination with observation data of a vehicle-mounted global navigation satellite system (GNSS) receiver of the vehicle, to obtain a satellite positioning position at the current positioning moment; and
fusing the priori position, the map matching position and the satellite positioning position to generate a positioning result of positioning the vehicle at the current positioning moment,
wherein the method is performed by at least one processor.

2. The method according to claim 1 , further comprising:
time-updating a Kalman filter having a vehicle position deviation, a vehicle speed deviation, a vehicle attitude deviation, a zero bias deviation of an accelerometer and a gyroscope in the vehicle-mounted inertial navigation system of the vehicle at each positioning moment and a clock offset and a clock drift of the vehicle-mounted global navigation satellite system (GNSS) receiver as state variables; and
fusing the priori position, the map matching position and the satellite positioning position to generate a positioning result of positioning the vehicle at the current positioning moment comprising:
determining a difference between the map matching position and the priori position and/or a difference between the satellite positioning position and the priori position as an observation value of the Kalman filter, and measurement-updating the Kalman filter; and
amending the priori position by using a measurement-updated vehicle position deviation to generate the positioning result.

3. The method according to claim 2 , wherein matching the first reflected value characteristic of the projection area generated by projecting the real-time laser point cloud collected by scanning the surrounding environment through the vehicle-mounted lidar of the vehicle at the current positioning moment with the second reflected value characteristic of the map area for searching to obtain the matching result, and determining the map matching position according to the matching result comprises:
calculating measurement matching probabilities of matching the first reflected value characteristic of the projection area with the second reflected value characteristic including reflected value characteristics of map areas corresponding to sample positions of the map area for searching; and
generating a coordinate of the map matching position, based on coordinates of the sample positions and the corresponding measurement matching probabilities.

4. The method according to claim 3 , wherein matching the first reflected value characteristic of the projection area generated by projecting the real-time laser point cloud collected by scanning the surrounding environment through the vehicle-mounted lidar of the vehicle at the current positioning moment with the second reflected value characteristic of the map area for searching to obtain the matching result, and determining the map matching position according to a matching result further comprises:
for each sample position, overlapping a center of the projection area with the sample position, and determining an area in the laser point cloud reflected value map overlapping the projection area as the map area corresponding to the sample position.

5. The method according to claim 3 , wherein the method further comprises generating the laser point cloud reflected value map, the generating comprising:
dividing a ground plane of an earth surface in a world coordinate system into a plurality of map grids of a same size and shape, each map grid corresponding to a sample position; and
storing a reflected value characteristic of a sample laser point cloud collected for each sample position in a map grid corresponding to the sample position.

6. The method according to claim 5 , wherein the reflected value characteristic comprises a reflected value mean of the laser point cloud; and
calculating measurement matching probabilities of matching the reflected value characteristic of the projection area with reflected value characteristics of map areas corresponding to sample positions comprises:
determining differences of reflected value means between the projection area and the map areas corresponding to the sample positions on the map grids; and
weight-averaging the differences of the reflected value means on the map grids, by using a number of laser points corresponding to the projection area in the map grids as a weight, and determining the measurement matching probabilities of matching the reflected value characteristic of the projection area with the reflected value characteristics of the map areas corresponding to the sample positions based on the weight-averaged means.

7. The method according to claim 3 , wherein generating the coordinate of the map matching position, based on coordinates of the sample positions and the corresponding measurement matching probabilities comprises:
determining prediction probabilities of the vehicle at the sample positions at the current positioning moment, based on the priori position and an priori deviation distribution of the priori position obtained by time-updating the Kalman filter;
amending the measurement matching probabilities by using the prediction probabilities of the sample positions to obtain a posteriori matching probabilities; and
weight-averaging the coordinates of the sample positions, by using the posteriori matching probabilities corresponding to the sample positions as a weight, to obtain the coordinate corresponding to the map matching position.

8. The method according to claim 2 , wherein positioning, using the priori position in combination with observation data of a vehicle-mounted GNSS receiver of the vehicle, to obtain a satellite positioning position comprises:
setting a fused weight of the observation data obtained by observing each positioning satellite at the current positioning moment by the vehicle-mounted GNSS receiver of the vehicle, based on the priori position.

9. The method according to claim 8 , wherein setting the fused weight of the observation data obtained by observing each positioning satellite at the current positioning moment by the vehicle-mounted GNSS receiver of the vehicle, based on the priori position comprises:
calculating a distance between the vehicle and the each positioning satellite based on the priori position and a satellite position of the each positioning satellite, to obtain a calculated value;
acquiring an observation value of the distance between the vehicle and the each positioning satellite obtained by observing the each positioning satellite by the

Assignees

• Baidu online network technology beijing co ltd

Inventors

• Cai Renlan
• Yang Xiaolong
• Wan Guowei
• Lu Weixin
• Song Shiyu
• Xu Baoqiang

Classifications

• G06V20/56

  exterior to a vehicle by using sensors mounted on the vehicle · CPC title

• G01S17/89

  for mapping or imaging · CPC title

• G08G1/137

  the indicator being in the form of a map · CPC title

• G01S17/00

  Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems · CPC title

• G01C21/30 Primary

  Map- or contour-matching · CPC title