The cascade-correlation learning architecture
Advances in neural information processing systems 2
The Promise of Neural Networks
The Promise of Neural Networks
Finding correlations between 3-D surfaces: a study in asymmetric incremental sheet forming
MLDM'12 Proceedings of the 8th international conference on Machine Learning and Data Mining in Pattern Recognition
| Hi-index | 0.00 |
Automating the process of cold-bending steel reinforcing bars for concrete structures can save time, prevent serious injuries, and increase productivity. Neural network models can play a key role in the adaptive control of this process.Operational objectives such as just-in-time (JIT) production and delivery of materials have fostered the need for real-time control of construction processes. Intelligent, adaptive control models are needed to automate these construction tasks. One such task is the cold bending of steel reinforcing bars, called rebars. This task involves accurately deforming a material, with variable mechanical properties, into a specified shape for a reinforced concrete structure.The traditionally manual process, performed with a table bender, is tedious and can involve serious and sometimes permanent injuries. Moreover, frequent pauses for setup and recalibration of the rebar bending machine with each new bundle of rebars hampers productivity.An automated intelligent-control process, suitable for handling material with inherently variable properties, can alleviate some of these problems. Two options are available for modeling the material behavior: the more traditional empirical statistical approach and an approach that uses a machine-learning technology such as a neural network. In this article, we explain why a neural network is an appropriate choice and how it can be used to predict springback--the variable elastic recovery after rebar bending.