US 11,781,972 B2
Techniques for selection of light source configurations for material characterization
Gearoid Murphy, Mountain View, CA (US); Diosdado Rey Banatao, Los Altos Hills, CA (US); Artem Goncharuk, Mountain View, CA (US); and Neil Treat, Los Gatos, CA (US)
Assigned to X Development LLC, Mountain View, CA (US)
Filed by X Development LLC, Mountain View, CA (US)
Filed on Nov. 5, 2020, as Appl. No. 16/949,601.
Prior Publication US 2022/0136959 A1, May 5, 2022
Int. Cl. G01N 21/25 (2006.01); G01N 21/3563 (2014.01); G01N 33/44 (2006.01); G06N 3/126 (2023.01)
CPC G01N 21/255 (2013.01) [G01N 21/3563 (2013.01); G01N 33/442 (2013.01); G06N 3/126 (2013.01)] 24 Claims
OG exemplary drawing
 
1. A method for selecting a spectroscopic sensor light source configuration, the method comprising:
obtaining, by a computer system, a light source dataset describing a plurality of light sources and a spectroscopic dataset describing a plurality of materials;
initializing, by the computer system, a genetic algorithm with an initial generation of solutions, an individual solution of the initial generation of solutions comprising a subset of light sources of the plurality of light sources;
selecting, by the computer system using the genetic algorithm, a first individual solution and a second individual solution from the initial generation of solutions, the first and second individual solutions respectively described by a first chromosome encoding and a second chromosome encoding;
generating, by the computer system using the genetic algorithm, a new individual solution from the first and second individual solutions by combining the first chromosome encoding and the second chromosome encoding;
evaluating, by the computer system using the genetic algorithm, a specificity of the new individual solution to a target material of the plurality of materials;
in accordance with the specificity of the new individual solution to the target material surpassing a specificity of the first individual solution to the target material or a specificity of the second individual solution to the target material: adding, by the computer system using the genetic algorithm, the new individual solution to a new generation of solutions;
populating, by the computer system using the genetic algorithm, the new generation of solutions with a plurality of new individual solutions;
generating, by the computer system using the genetic algorithm, one or more subsequent generations of solutions by iterating the genetic algorithm;
selecting, by the computer system using the genetic algorithm, one or more implementation individual solutions from a final generation of the one or more subsequent generations by identifying the one or more implementation individual solutions exhibiting a threshold specificity to the target material; and
outputting, by the computer system, the one or more implementation individual solutions, wherein an implementation individual solution of the one or more implementation individual solutions comprises a spectroscopic sensor light source configuration.