Ionic Crystal Similarity Analysis Using Conflict Graph Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for analyzing ionic crystals face challenges due to the large number of atoms in the unit cell, leading to increased computational bits required and difficulty in achieving accurate similarity analysis.
Innovation Solution
A crystal analysis method and device that employs a conflict graph to reduce the number of bits needed by using extension nodes and loop structures to handle the repeating structure of ionic crystals, allowing for accurate similarity determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of atoms in the unit cell is increased to represent ionic crystals accurately, then the structural representation becomes more complete, but the number of bits required for computation increases beyond the annealing machine's capacity
Solution Approach 1:
The patent segments the crystal structure analysis by separating cationic atoms from anionic atoms. Instead of treating all atoms uniformly, the method divides the computational problem into two parts: representing cationic atoms as nodes in a graph and representing anionic atoms as bonds between these nodes. This segmentation reduces the number of nodes (and thus bits) required while preserving the complete structural information through the bond relationships.
Solution Approach 2:
The patent extracts and removes redundant information from the computational model. By recognizing that anionic atoms in ionic crystals primarily serve as connectors between cationic atoms, the method extracts only the essential structural information (cationic atom positions and their connectivity) and represents the anionic atoms implicitly through bonds, rather than explicitly including all anionic atoms as separate nodes that would increase the bit count.
2Loss of information
If all anionic atoms are represented as nodes in the graph, then the complete bonding information is captured, but the computational burden and bit requirements increase significantly
Solution Approach 1:
The patent merges the representation of anionic atoms into the bonds between cationic atoms. Instead of treating anionic atoms as separate entities (nodes) that would individually consume computational resources, the method combines their representational function into the edges (bonds) of the graph, where each bond implicitly represents the presence and connectivity of anionic atoms without requiring separate node representations.
Solution Approach 2:
The patent inverts the conventional approach by not representing the majority species (anionic atoms) as nodes, but rather representing the minority species (cationic atoms) as nodes and using the bonds between them to encode the presence of anionic atoms. This inversion reduces the node count while maintaining bonding information completeness.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A crystal analysis method for a computer to execute a process includes creating a graph that indicates data of repeating unit cell in an ionic crystal and data of an adjacent repeating unit cell that is adjacent to the repeating unit cell; analyzing the ionic crystal based on the graph; and when a number of first intra-cell node that indicates data of an anionic atom bonded to a cationic atom in the repeating unit cell is n, setting a number of second intra-cell node that indicates data of the anionic atom in the repeating unit cell n-1 or less, wherein the data of repeating unit cell includes a plurality of intra-cell nodes that indicate data of atoms in the repeating unit cell, and the plurality of intra-cell nodes include the first intra-cell node and the second intra-cell node.