Global map based deep reinforcement learning for parking
A method for calculating an optimum route for a first vehicle to travel from a current location to a selected parking space includes receiving map data of a parking zone, the map data including local map data and global map data of the parking zone, sensor data, parking space data indicating availability of parking spaces within the parking zone, and vehicle dynamics and location data, selecting, as the selected parking space, at least one available parking space based on the parking space data, determining a plurality of target positions between the current location and the selected parking space, calculating, in response to the selecting of the selected parking space and based on the plurality of target positions, the optimum route, including calculating trajectories of the first vehicle along an entire route between the current location and the selected parking space, and controlling the first vehicle to travel from the current location to the selected parking space.
1 . A method for calculating an optimum route for a first vehicle to travel from a current location to a selected parking space, the method comprising:
receiving
map data of a parking zone, wherein the map data includes local map data and global map data of the parking zone,
sensor data,
parking space data indicating availability of parking spaces within the parking zone, and
vehicle dynamics and location data;
selecting, as the selected parking space, at least one available parking space based on the parking space data;
determining a plurality of target positions between the current location and the selected parking space;
calculating, in response to the selecting of the selected parking space and based on the plurality of target positions, the optimum route, wherein the calculating of the optimum route includes calculating trajectories of the first vehicle along an entire route between the current location and the selected parking space;
controlling the first vehicle to travel from the current location to the selected parking space;
receiving first image data sensed by a first sensor mounted on the first vehicle, wherein the first image data corresponds to the parking zone;
receiving second image data sensed by a second sensor mounted on a second vehicle, wherein the second image data corresponds to the parking zone;
generating a digital map of the parking zone based on the first image data and the second image data, wherein the global map data includes the first image data and the second image data;
sending the digital map from a remote server to a device associated with the first vehicle; and
displaying the digital map on the device, enabling a user of the device to see real-time information regarding availability of parking spaces in the parking zone.
2 . The method of claim 1 , wherein the global map data includes data obtained by at least one of (i) a second vehicle and (ii) a camera or sensor associated with the parking zone.
3 . The method of claim 1 , wherein the determining of the plurality of target positions includes calculating trajectory points for the first vehicle for each of the plurality of target positions.
4 . The method of claim 1 , wherein each of the plurality of target positions includes a target pose of the first vehicle.
5 . The method of claim 4 , wherein the calculating of the optimum route includes calculating, based on the target poses of the first vehicle at respective target positions of the plurality of target positions, a plurality of optimum route segments between the plurality of target positions.
6 . The method of claim 1 , wherein the calculating of the optimum route includes using deep reinforcement learning.
7 . The method of claim 6 , wherein the using of the deep reinforcement learning includes generating at least one path based on the global map data, generating a costmap, and calculating the optimum route based on the global map data and the costmap.
8 . The method of claim 1 , wherein the generating of the digital map includes using a visual simultaneous localization and mapping (SLAM) system.
9 . A system for calculating an optimum route for a first vehicle to travel from a current location to a selected parking space, the system comprising:
a plurality of image sensors configured to generate image data;
one or more processors; and
memory coupled to the one or more processors, the memory storing instructions that, when executed by the one or more processors, cause the one or more processors to:
receive
map data of a parking zone, wherein the map data includes local map data and global map data of the parking zone,
sensor data,
parking space data indicating availability of parking spaces within the parking zone, and
vehicle dynamics and location data;
select, as the selected parking space, at least one available parking space based on the parking space data;
determine a plurality of target positions between the current location and the selected parking space;
calculate, in response to the selecting of the selected parking space and based on the plurality of target positions, the optimum route, wherein the calculating of the optimum route includes calculating trajectories of the first vehicle along an entire route between the current location and the selected parking space;
control the first vehicle to travel from the current location to the selected parking space;
receive first image data sensed by a first sensor mounted on the first vehicle, wherein the first image data corresponds to the parking zone;
receive second image data sensed by a second sensor mounted on a second vehicle, wherein the second image data corresponds to the parking zone;
generate a digital map of the parking zone based on the first image data and the second image data, wherein the global map data includes the first image data and the second image data;
send the digital map from a remote server to a device associated with the first vehicle; and
display the digital map on the device, enabling a user of the device to see real-time information regarding availability of parking spaces in the parking zone.
10 . The system of claim 9 , wherein the global map data includes data obtained by at least one of (i) a second vehicle and (ii) a camera or sensor associated with the parking zone.
11 . The system of claim 9 , wherein the determining of the plurality of target positions includes calculating trajectory points for the first vehicle for each of the plurality of target positions.
12 . The system of claim 9 , wherein each of the plurality of target positions includes a target pose of the first vehicle.
13 . The system of claim 12 , wherein the calculating of the optimum route includes calculating, based on the target poses of the first vehicle at respective target positions of the plurality of target positions, a plurality of optimum route segments between the plurality of target positions.
14 . The system of claim 9 , wherein the calculating of the optimum route includes using deep reinforcement learning.
15 . The system of claim 14 , wherein the using of the deep reinforcement learning includes generating at least one path based on the global map data, generating a costmap, and calculating the optimum route based on the global map data and the costmap.
16 . A non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions which, when executed by one or more processors of an electronic device, cause the electronic device to:
receive
map data of a parking zone, wherein the map data includes local map data and global map data of the parking zone,
sensor data,
parking space data indicating availability of parking spaces within the parking zone, and
vehicle dynamics and location data;
select, as a selected parking space, at least one available parking space based on the parking space data;
determine a plurality of target positions between a current location and the selected parking space;
calculate, in response to the selecting of the selected parking space and based on the plurality of target positions, an optimum route from the current location of a first vehicle to the selected parking space, wherein the calculating of the optimum route includes calculating trajectories of the first vehicle along an entire route between the current location and the selected parking space;
control the first vehicle to travel from the current location to the selected parking space;
receive first image data sensed by a first sensor mounted on the first vehicle, wherein the first image data corresponds to the parking zone;
receive second image data sensed by a second sensor mounted on a second vehicle, wherein the second image data corresponds to the parking zone;
generate a digital map of the parking zone based on the first image data and the second image data, wherein the global map data includes the first image data and the second image data;
send the digital map from a remote server to a device associated with the first vehicle; and
display the digital map on the device, enabling a user of the device to see real-time information regarding availability of parking spaces in the parking zone.
17 . The non-transitory computer-readable storage medium of claim 16 , wherein the global map data includes data obtained by at least one of (i) a second vehicle and (ii) a camera or sensor associated with the parking zone.
18 . The non-transitory computer-readable storage medium of claim 16 , wherein at least one of (i) the determining of the plurality of target positions includes calculating trajectory points for the first vehicle for each of the plurality of target positions and (ii) each of the plurality of target positions includes a target pose of the first vehicle.
19 . The non-transitory computer-readable storage medium of claim 18 , wherein the calculating of the optimum route includes calculating, based on the target poses of the first vehicle at respective target positions of the plurality of target positions, a plurality of optimum route segments between the plurality of target positions.