Deeppipe: A hybrid model for multi-product pipeline condition recognition based on process and data coupling.
Chang WangJianqin ZhengYongtu LiangMiao LiWencai ChenQi LiaoHaoran ZhangPublished in: Comput. Chem. Eng. (2022)
Keyphrases
- hybrid model
- data collection
- data sets
- raw data
- image data
- database
- data analysis
- spatial data
- neural network
- computer vision
- data quality
- data sources
- input data
- redundant data
- recognition accuracy
- sufficient conditions
- data points
- object recognition
- pattern recognition
- training data
- feature extraction
- recognition rate
- nearest neighbor
- data processing
- end users
- process model
- long term
- human activities
- data structure
- recognition process
- product development