Machine Learning Approach to Determine Corrosion Potential of Friction Stir Welded Joints
The main objective of the research study is to create an Artificial Neural Network (ANN) architecture to predict corrosion resistance from the Friction Stir Welding experimental dataset given to us. This is also to find the mean squared error and mean absolute percentage error of the given model which will help to analyze the losses and efficiency of the model. In the Artificial Neural Architecture, Tool Rotational Speed (rpm), Axial force (kN) and Welding Speed (mm/min) are the inputs while Corrosion Potential is the output. It is observed that different line plots for loss and mean square error the train plot loss decreases as epoch is increased while for test, the relation between loss and epoch remains constant at 0.4109.
Keywords: Neural Networks, Friction Stir Welding, Corrosion Potential, Machine Learning
Cite this Article Akshansh Mishra, Adarsh Tiwari, Vaibhav, Nitin Kumar Dubey. Machine Learning Approach to Determine Corrosion Potential of Friction Stir Welded Joints. Journal of Image Processing & Pattern Recognition Progress. 2020; 7(1): 5–17p.
- There are currently no refbacks.