Sequential sensor placement for gas emission detection using record count

A computer readable medium storing computer executable code for automated sensor placement location selection, the computer executable code comprising:

• code for obtaining wind rose distribution data associated with a site for gas emission detection;
• code for generating a plurality of wind realizations from the wind rose distribution data;
• code for, for each candidate sensor location of a plurality of candidate sensor locations at the site, generating a plurality of records associated with predicted sensor measurements at the candidate sensor location for each of a plurality of potential gas emission locations at the site and subject to each of the plurality of wind realizations;
• code for, for each candidate sensor location of a plurality of candidate sensor locations at the site, generating a record count based on the plurality of records;
• code for, for each candidate sensor location of a plurality of candidate sensor locations at the site, ranking the candidate sensor location based on the record count;
• code for iteratively selecting candidate sensor locations, from the plurality of candidate sensor locations, based on the ranking, wherein the code for iteratively selecting comprises:
• code for, in each iteration, selecting for gas emission detection sensor placement a candidate sensor location, from the plurality of candidate sensor locations, having a highest ranking;
• and
• code for, in each iteration, removing at least the selected candidate sensor location from the plurality of candidate sensor locations;
• and
• code for outputting the selected sensor locations.

An apparatus for automated sensor placement location selection, the apparatus comprising:

• one or more memories storing computer executable code;
• and
• one or more processors configured to execute the computer executable code and cause the apparatus to:
• obtain wind rose distribution data associated with a site for gas emission detection;
• generate a plurality of wind realizations from the wind rose distribution data;
• for each candidate sensor location of a plurality of candidate sensor locations at the site:
• generate a plurality of records associated with predicted sensor measurements at the candidate sensor location for each of a plurality of potential gas emission locations at the site and subject to each of the plurality of wind realizations;
• generate a record count based on the plurality of records;
• and
• rank the candidate sensor location based on the record count;
• iteratively select candidate sensor locations, from the plurality of candidate sensor locations, based on the ranking, wherein the iteratively selecting comprises, in each iteration:
• selecting for gas emission detection sensor placement a candidate sensor location, from the plurality of candidate sensor locations, having a highest ranking;
• and
• removing at least the selected candidate sensor location from the plurality of candidate sensor locations;
• and
• output the selected sensor locations.