Gas sensor array low-concentration
About
This dataset contains 6 gas responses collected by a sensor array consisting of 10 metal oxide semiconductor sensors, with gas concentrations at the ppb level (below the minimum detection limit of the sensors)
Variables Info:
In the gsalc. csv file, each row represents a gas sample, and the first and second columns of each row represent gas labels and concentration labels, respectively. Starting from the third column, the response of the sensor array to the sample is recorded. Each sample has 9000 response points, which are concatenated from the responses of 10 sensors, in the order of TGS2603, TGS2630, TGS813, TGS822, MQ-135, MQ-137, MQ-138, 2M012, VOCS-P, 2SH12 (i.e. the first 900th sample points are the responses of sensor TGS2603, and the 901-1800th sample points are the responses of sensor TGS2630...).
Class labels:
We collected response signals from the array for six gases including ethanol, acetone, toluene, ethyl acetate, isopropanol, and n-hexane. Each gas had three concentrations of 50ppb, 100ppb, and 200ppb, and five samples were collected for each gas and concentration, for a total of 90 samples. The collection process is divided into three stages: baseline (5 minutes), injection (10 minutes), and cleaning (15 minutes), with a sampling frequency of 1Hz. This dataset provides the response of the baseline and injection stages.
Subject Area
Computer Science
Instances
90
Features
–
Data Types
Multivariate
Tasks
Classification, Regression, Clustering
Feature Types
Continuous
Features
–
Introductory Paper
Feature Ensemble Learning for Sensor Array Data Classification Under Low-Concentration Gas
Leilei Zhao, Fengchun Tian, Junhui Qian, Hantao Li, Zhiyuan Wu. 2023.
IEEE Transactions on Instrumentation and Measurement