Method and system for calibrating measured values for ambient air parameters using trained models

A vehicle mounted data acquisition system (DAS) comprising:

• a processor;
• an analog to digital converter connected to the processor;
• at least one gaseous component concentration sensor connected to the analog to digital converter;
• memory connected to the processor, wherein the memory comprises instructions that, when read by the processor, cause the processor to:
• measure, by the at least one gaseous component concentration sensor, a concentration of gas;
• receive, via the analog to digital converter connected to a processor, an analog signal from the gaseous component concentration sensor that represents the measured concentration of gas;
• generate, by the analog to digital converter, an uncalibrated sensor value xm(k) that indicates the measured concentration of gas, wherein k is an integer time index;
• measure a speed of the vehicle mounted DAS that is time-paired with the uncalibrated sensor value xm(k) and include the speed of the vehicle mounted DAS in a vector of accuracy enhancing parameters p(k);
• receive, from a cloud server, reference unit location information indicating a location of a geographically fixed reference unit that is external to the vehicle mounted DAS, wherein the geographically fixed reference unit comprises at least one gaseous component concentration sensor and hardware that stabilizes vibration of the at least one gaseous component concentration sensor;
• determine a distance from the vehicle mounted DAS to the geographically fixed reference unit based on the received reference unit location information;
• determine whether or not the distance from the vehicle mounted DAS to the geographically fixed reference unit is less than a threshold distance;
• in circumstances where the distance from the vehicle mounted DAS to the geographically fixed reference unit is less than a threshold distance:
• receive a reference sensor value xs(k) from the geographically fixed reference unit, wherein the reference sensor value xs(k) corresponds with at least one measurement of the at least one gaseous component concentration sensor comprised by the geographically fixed reference unit that is time-paired with the uncalibrated sensor value xm(k);
• calculate a trust weight h(k) that is time-paired with the uncalibrated sensor value xm(k) such that the trust weight h(k):
• is inversely proportional to the speed of the vehicle mounted DAS, and
• is inversely proportional to the distance from the vehicle mounted DAS to the geographically fixed reference unit;
• generate a training dataset point according to ([xm(k)p(k)T],xs(k)], h(k));
• train a calibration model using the training dataset point;
• apply the calibration model to the uncalibrated sensor value xm(k) and the vector of accuracy enhancing parameters p(k) to generate a calibrated sensor value;
• and
• wirelessly transmit the calibrated sensor value to the cloud server;
• in circumstances where the distance from the vehicle mounted DAS to the geographically fixed reference unit is not less than a threshold distance:
• apply the calibration model to the uncalibrated sensor value xm(k) and the vector of accuracy enhancing parameters p(k) to generate a calibrated sensor value;
• and
• wirelessly transmit the calibrated sensor value to the cloud server.

A non-transitory computer readable medium that includes instructions that, when read by a processor of a vehicle mounted data acquisition system (DAS), cause the processor to:

• cause measurement, by at least one gaseous component concentration sensor comprised by the vehicle mounted DAS, a concentration of gas;
• receive, via an analog to digital converter connected to a processor, an analog signal from the gaseous component concentration sensor comprised by the vehicle mounted DAS that represents the measured concentration of gas;
• generate, by the analog to digital converter, an uncalibrated sensor value xm(k) that indicates the measured concentration of gas, wherein k is an integer time index;
• measure a speed of the vehicle mounted DAS that is time-paired with the uncalibrated sensor value xm(k) and include the speed of the vehicle mounted DAS in a vector of accuracy enhancing parameters p(k);
• receive, from a cloud server, reference unit location information indicating a location of a geographically fixed reference unit that is external to the vehicle mounted DAS, wherein the geographically fixed reference unit comprises at least one gaseous component concentration sensor and hardware that stabilizes vibration of the at least one gaseous component concentration sensor;
• determine a distance from the vehicle mounted DAS to a geographically fixed reference unit based on the received reference unit location information;
• determine whether or not the distance from the vehicle mounted DAS to the geographically fixed reference unit is less than a threshold distance;
• in circumstances where the distance from the vehicle mounted DAS to the geographically fixed reference unit is less than a threshold distance:
• receive a reference sensor value xs(k) from the geographically fixed reference unit, wherein the reference sensor value xs(k) corresponds with at least one measurement of the at least one gaseous component concentration sensor comprised by the geographically fixed reference unit that is time-paired with the uncalibrated sensor value xm(k);
• calculate a trust weight h(k) that is time-paired with the uncalibrated sensor value xm(k) such that the trust weight h(k):
• is inversely proportional to the speed of the vehicle mounted DAS, and
• is inversely proportional to the distance from the vehicle mounted DAS to the geographically fixed reference unit;
• generate a training dataset point according to ([xm(k)p(k)T],xs(k)], h(k));
• train a calibration model using the training dataset point;
• apply the calibration model to the uncalibrated sensor value xm(k) and the vector of accuracy enhancing parameters p(k) to generate a calibrated sensor value;
• and
• wirelessly transmit the calibrated sensor value to a cloud server;
• in circumstances where the distance from the vehicle mounted DAS to the geographically fixed reference unit is not less than a threshold distance:
• apply the calibration model to the uncalibrated sensor value xm(k) and the vector of accuracy enhancing parameters p(k) to generate a calibrated sensor value;
• and
• wirelessly transmit the calibrated sensor value to a cloud server.