Genetic Algorithm Object Packing for Additive Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current additive manufacturing techniques face challenges in efficiently optimizing object packing density within a build volume, as object packing problems are computationally intensive and often result in exponentially growing solution spaces, making it difficult to find near-optimal solutions.
Innovation Solution
A genetic procedure is employed to determine initial orientations of objects based on minimum bounding boxes, initializing a population and iteratively applying evaluation, selection, crossover, and mutation operations to prioritize solution spaces with higher probabilities of producing near-optimal packings, thereby reducing the computational complexity and improving packing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If exhaustive search methods are used to find optimal object packing arrangements, then packing density can be maximized, but computational time and complexity grow exponentially
Solution Approach 1:
The patent applies preliminary action by initializing the genetic algorithm with bounding box orientations determined before the optimization process. This pre-computation of initial orientations based on minimum bounding boxes provides a structured starting point that guides the search toward optimal solutions more efficiently, reducing the computational time required to achieve high packing density without exhaustive search
Solution Approach 2:
The patent employs parameter changes by transforming the object orientation representation into bounding box formats and using these as initial parameters for the genetic algorithm. This parameter transformation enables the algorithm to work with more manageable representations that reduce solution space complexity while maintaining the ability to achieve optimal packing arrangements
2Manufacturing precision
If the solution space is searched exhaustively to ensure optimal packing, then near-optimal solutions can be found, but the computational complexity grows exponentially
Solution Approach 1:
The patent applies segmentation by dividing the complex packing optimization problem into manageable components: determining bounding boxes for each object, initializing orientations based on these bounding boxes, and then applying genetic algorithms to optimize the arrangement. This segmentation reduces the overall computational complexity by breaking down the exponential solution space into structured sub-problems that can be solved more efficiently
Solution Approach 2:
By performing preliminary computation of bounding box orientations before initiating the genetic algorithm, the patent reduces the complexity of the subsequent optimization process. This pre-processing step structures the initial population in a way that guides the genetic algorithm toward optimal solutions more efficiently, avoiding exhaustive search of the entire solution space
Data Source
AI summary
Examples of methods for determining object packings are described herein. In some examples, a method includes determining, for each of a plurality of objects, an initial orientation corresponding to a minimum bounding box. In some examples, the method includes determining a packing of the plurality of objects using the initial orientations with a genetic procedure.