US 12,456,542 B2
Material representation in computational systems
Nathaniel H. Park, San Jose, CA (US); Dmitry Zubarev, San Jose, CA (US); Gowri Nayar, San Jose, CA (US); and James L. Hedrick, Pleasanton, CA (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Mar. 12, 2021, as Appl. No. 17/200,666.
Prior Publication US 2022/0293216 A1, Sep. 15, 2022
Int. Cl. G16B 25/20 (2019.01); G06F 16/903 (2019.01); G06F 16/9035 (2019.01); G06N 3/08 (2023.01); G16B 20/20 (2019.01)
CPC G16B 25/20 (2019.02) [G06F 16/90335 (2019.01); G06F 16/9035 (2019.01); G06N 3/08 (2013.01); G16B 20/20 (2019.02)] 20 Claims
 
1. A method comprising:
computationally generating graphs encoding data on at least one polymer compound, wherein the graphs comprise (i) nodes encoding architectural features of the at least one polymer compound, (ii) edges encoding chemical bonds between the nodes, topology of edge connectivity, and a direction of the chemical bonds, and (iii) edge weights encoding information on characteristics of the chemical bonds between the nodes including degree of polymerization;
computationally generating a predictive model comprising data identifying connections between the graphs, wherein directionality and connectivity of the chemical bonds of the at least one polymer compound encodes atomistic, architectural, and topological information comprising data modeling of user-defined polymeric materials with user-defined parameters based on user-defined queries that include architectural, topological and chemical-based features of a chemical structure, including topology of edge connectivity and degree of polymerization between nodes; and
producing the user-defined polymeric materials,
wherein the architectural features of the nodes include at least three of end groups, branching points, repeat units, crosslinks, and blocks.