Loading...

The Representation of an Intelligent Process Planning Structure for Additive Manufacturing Systems Based on Cognitive Models

Rezaei, Mohammad Reza | 2023

88 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 55985 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Haji, Alireza; Houshmand، Mahmoud; Fatahi Valilai, Omid
  7. Abstract:
  8. In the traditional process planning, the expert decides upon the production process for a product by considering the previous cases and by decomposing, analysis and interpreting the shape of the product in terms of more familiar previous cases. In recent decades and by developments in the disciplines of the artificial intelligence, mass customization, cloud manufacturing and additive manufacturing, a necessity for process planning systems which need less human intervention by intelligent interpretation of the shape and consideration of the available data from previous productions, has emerged. Inspired by the traditional process planning by human expert, and considering the overall process of decomposition-interpretation-composition, the mental process which takes place when an expert wants to do the task, this study is intended to introduce a basis for a system which has the required flexibility and learnability to carry out the task of process planning automatically in additive manufacturing. The proposed system acts as: it first decomposes the shape into simpler and more interpretable parts based on the harmony of the points on the edges or by considering convexity of the shape in different points; In the next step the system classifies the created parts using neural networks and determines their current position in the shape. Then it compares the as-is conditions of the parts with more favorable conditions which are determined by the favorability of the previous similar parts, and considering some wholistic criteria, the framework decides upon the initial shape. In some case examples it would be illustrated that how the proposed framework, with such a mechanism, and by considering criteria like support volume as the wholistic criterion and by changing the utility space for different conditions, can help in making decisions in determining one of the most important parameters in additive manufacturing: fabrication orientation. This framework is a first step towards a system with more learnability and flexibility which can help to improve the agility of the manufacturing system, can reduce the role of human error in analyzing the details, and help in taking the factors which have been hidden from the human agent or been difficult to analyze, into account implicitly and through previous cases
  9. Keywords:
  10. Artificial Intelligence ; Computer Aided Process Planning ; Cognitive Modeling ; Additive Manufacturing ; Deep Neural Networks ; Artificial Intelligence in Additive Manufacturing ; Fabrication Orientation in Additive Manufacturing

 Digital Object List

 Bookmark

No TOC