A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
MACHINE LEARNING DEVICE FOR CONSTRUCTION MACHINERY, CONSTRUCTION MACHINERY, AND MACHINE LEARNING METHOD FOR CONSTRUCTION MACHINERY
To achieve flexible control of construction machinery.SOLUTION: A machine learning device 20 is provided as a machine learning device for a construction machine that learns a target value for an overlapping length along a rolling compaction route for rolling compaction work performed in a rolling compaction area by a construction machine. The machine learning device 20 comprises: a state observation unit 21 which observes an actual overlapping length according to the rolling compaction work; and an overlapping length determination unit 22 which determines the target value from the overlapping length observed by the state observation unit 21. The overlapping length determination unit 22 decreases the target value when the observed overlapping length is larger than a predetermined first threshold value, and increases the target value when the observed overlapping length is smaller than the predetermined first threshold value.SELECTED DRAWING: Figure 5
【課題】建設機械の柔軟な制御を実現できる。【解決手段】建設機械による転圧エリアの転圧走行作業において、転圧路の重複長の目標値を学習する建設機械用機械学習装置としての機械学習器２０である。この機械学習器２０は、前記転圧走行作業による実際の重複長を観測する状態観測部２１と、状態観測部２１により観測された前記重複長から前記目標値を判定する重複長判定部２２と、を備える。重複長判定部２２は、観測される前記重複長が予め決められた第１閾値よりも多い場合に前記目標値を小さくし、観測される前記重複長が予め決められた第１閾値よりも少ない場合に前記目標値を大きくする。【選択図】図５
MACHINE LEARNING DEVICE FOR CONSTRUCTION MACHINERY, CONSTRUCTION MACHINERY, AND MACHINE LEARNING METHOD FOR CONSTRUCTION MACHINERY
To achieve flexible control of construction machinery.SOLUTION: A machine learning device 20 is provided as a machine learning device for a construction machine that learns a target value for an overlapping length along a rolling compaction route for rolling compaction work performed in a rolling compaction area by a construction machine. The machine learning device 20 comprises: a state observation unit 21 which observes an actual overlapping length according to the rolling compaction work; and an overlapping length determination unit 22 which determines the target value from the overlapping length observed by the state observation unit 21. The overlapping length determination unit 22 decreases the target value when the observed overlapping length is larger than a predetermined first threshold value, and increases the target value when the observed overlapping length is smaller than the predetermined first threshold value.SELECTED DRAWING: Figure 5
【課題】建設機械の柔軟な制御を実現できる。【解決手段】建設機械による転圧エリアの転圧走行作業において、転圧路の重複長の目標値を学習する建設機械用機械学習装置としての機械学習器２０である。この機械学習器２０は、前記転圧走行作業による実際の重複長を観測する状態観測部２１と、状態観測部２１により観測された前記重複長から前記目標値を判定する重複長判定部２２と、を備える。重複長判定部２２は、観測される前記重複長が予め決められた第１閾値よりも多い場合に前記目標値を小さくし、観測される前記重複長が予め決められた第１閾値よりも少ない場合に前記目標値を大きくする。【選択図】図５
MACHINE LEARNING DEVICE FOR CONSTRUCTION MACHINERY, CONSTRUCTION MACHINERY, AND MACHINE LEARNING METHOD FOR CONSTRUCTION MACHINERY
建設機械用機械学習装置、建設機械および建設機械用機械学習方法
AOKI HIROAKI (author) / KATAYAMA SABURO (author)
2018-11-01
Patent
Electronic Resource
Japanese
European Patent Office | 2017
| European Patent Office | 2016
European Patent Office | 2018
European Patent Office | 2018