On-board processing device, vehicle control device, and self-position estimation method

A vehicle control device comprising a processor configured to:

• record i) a traveling map generated based on sensor data acquired while a vehicle is traveling and ii) travel status data including a first travel status at a time of generation of a past generated map, wherein the travel status data represents a vehicle operating condition and an environmental condition that affect sensor accuracy;
• execute first self-position estimation to estimate a self-position of the vehicle based on a result of first collation to collate past point group data included in the past generated map with first point group data included in the traveling map, the first point group data including road surface point group data representing a road marking displayed on a road surface and high point group data representing an elevated feature having a height above the road surface;
• in response to a determination that the road surface point group data has higher accuracy compared with the high point group data, select a subset of the road surface point group data, as second point group data, from the first point group data based on the first travel status at the time of generation of the past generated map and a second travel status at a time of generation of the traveling map;
• estimate the self-position of the vehicle based on a result of second collation to collate the past point group data included in the past generated map with the second point group data;
• and
• control, based on the estimated self-position of the vehicle, at least one of a steering device, a driving device, or a braking device of the vehicle.

A self-position estimation method for estimating a self-position by an on-board processing device, the method comprising:

• recording a past generated map generated in the past and a traveling map generated based on sensor data acquired while a vehicle is traveling, the traveling map including first point group data, wherein the first point group data includes road surface point group data representing a road marking displayed on a road surface and high point group data representing an elevated feature having a height above the road surface;
• executing first collation to collate past point group data included in the past generated map with the road surface point group data and collate the past point group data with the high point group data;
• executing first self-position estimation to estimate a self-position of the vehicle based on a result of the first collation;
• in response to a determination that the road surface point group data has higher accuracy compared with the high point group data, estimating the self-position of the vehicle based on a result of second collation to collate the past point group data with second point group data, wherein the second point group data is a subset of the road surface point group data, the subset selected from the past point group data and the first point group data;
• and
• controlling, based on the estimated self-position of the vehicle, at least one of a steering device, a driving device, or a braking device of the vehicle.