公开(公告)号：US20230277027A1

公开(公告)日：2023-09-07

申请号：US17943634

申请日：2022-09-13

申请人： Megnath RAMESH ,  Francis Christopher Imeson ,  Stephen Smith ,  Baris Fidan

发明人： Megnath RAMESH ,  Francis Christopher Imeson ,  Stephen Smith ,  Baris Fidan

IPC分类号： A47L11/40 ,  G06F17/12 ,  G05D1/02

CPC分类号： A47L11/4011 ,  G06F17/12 ,  G05D1/0219 ,  G05D1/0274 ,  A47L2201/04 ,  G05D2201/0203

摘要： A system and method of minimum turn coverage of arbitrary non-convex regions. Coverage planning is the task of generating a path that ensures the tool carried by the robot covers all regions of interest. The number of turns in the path can affect the time to cover the region and the quality of coverage (tools like cameras and cleaning attachments commonly have poor performance around turns). In recent turn-minimizing coverage methods, the region is partitioned to be covered by the least number of rectangles of width equal to the tool's width. The partitioning problem is typically solved using heuristics that have no optimality guarantees. A linear programming (LP) approach is disclosed to generate an axis-parallel coverage plan that minimizes the number of turns taken by the robot. The LP method solves this problem optimally in polynomial time. Coverage plans are generated for real regions using the LP method.

公开(公告)号：US11131992B2

公开(公告)日：2021-09-28

申请号：US16206506

申请日：2018-11-30

申请人： DENSO International America, Inc. ,  Sriram Subramanian ,  Sushrut Bhalla ,  Jaspreet Sambee ,  Mark Crowley ,  Sebastian Fischmeister ,  Donghyun Shin ,  William Melek ,  Baris Fidan ,  Ami Woo ,  Bismay Sahoo

发明人： Zhiyuan Du ,  Joseph Lull ,  Rajesh Malhan ,  Sriram Subramanian ,  Sushrut Bhalla ,  Jaspreet Sambee ,  Mark Crowley ,  Sebastian Fischmeister ,  Donghyun Shin ,  William Melek ,  Baris Fidan ,  Ami Woo ,  Bismaya Sahoo

IPC分类号： G05D1/00 ,  G05D1/02 ,  G06N3/04 ,  G06N3/08

摘要： A RLP system for a host vehicle includes a memory and levels. The memory stores a RLP algorithm, which is a multi-agent collaborative DQN with PER algorithm. A first level includes a data processing module that provides sensor data, object location data, and state information of the host vehicle and other vehicles. A second level includes a coordinate location module that, based on the sensor data, the object location data, the state information, and a refined policy provided by the third level, generates an updated policy and a set of future coordinate locations implemented via the first level. A third level includes evaluation and target neural networks and a processor that executes instructions of the RLP algorithm for collaborative action planning between the host and other vehicles based on outputs of the evaluation and target networks and to generate the refined policy based on reward values associated with events.