To provide a reinforcement learning device capable of simple and efficient reinforcement learning in relation to a traveling aspect of a vehicle.SOLUTION: A reinforcement learning device for performing reinforcement learning in relation to a traveling aspect of a target vehicle uses a simulation image in which a first type symbol representing the target vehicle moves along a road on an abstract road map in an environment in which a second type symbol representing a moving object distinguished according to its attributes moves along a road on the abstract road map according to a scenario (S11-S15), and when the first type symbol falls into an error state (S14), causes processing related to the reinforcement learning to be continued by performing restoration to the simulation image in a state in which the first type symbol is located at a checkpoint (S18, S19).SELECTED DRAWING: Figure 5
【課題】車両の走行態様について、簡素で、効率的な強化学習が可能な強化学習装置を提供することである。【解決手段】対象車両の走行態様についての強化学習を行う強化学習装置であって、移動体をその属性に応じて区別して表した第2種記号がシナリオに従って抽象化道路地図における道路内を移動する環境下において対象車両を表す第1種記号が抽象化道路地図の道路内を移動するシミュレーション画像を用い(S11~S15)、第1種記号がエラー状態となったときに(S14)、第1種記号がチェックポイントに位置する状態のシミュレーション画像に復帰させて(S18、S19)、強化学習に係る処理を継続させる。【選択図】図5
REINFORCEMENT LEARNING DEVICE
強化学習装置
2024-08-16
Patent
Electronic Resource
Japanese
IPC: | G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion |
Minimax Reinforcement Learning
AIAA | 2003
|Reinforcement Learning‐Based Filter
Wiley | 2022
|