Cost-Effective Sensor Systems for Measuring Extracted Chlorophyll-a Concentration
Chlorophyll-a concentration is one of the most measured metrics in both water quality and plant health monitoring. It is an indicator of algal biomass and provides insight into stressors such as eutrophication and bloom risk. It is also a widely used metric in terrestrial ecosystems as an indicator of photosynthetic activity and nutrient limitation. Most currently used laboratory-based methods for measuring chlorophyll-a exploit spectroscopic techniques and require expensive instrumentation, like spectrophotometer or fluorometer. In addition, the readings are taken inside a black box to avoid optical noise. The purpose of this thesis is to propose a smart, low-cost, and portable sensor system to measure the concentration of chlorophyll-a in an extracted solution. The goals were achieved using two distinct spectral method. The first approach involves two consumer-grade spectral sensors that read the optical reflectance at 12 discrete wavelengths in visible and near-infrared spectra. The system was tuned for an optimal distance from the sensors to the solution and an enclosure was printed to maintain the distance, as well as to avoid natural light interference. Extracted chlorophyll-a solutions of 52 different concentrations were prepared, and at least 5 readings per sample were taken using the proposed smart sensor system. The ground truth values of the samples were measured in the laboratory using Thermo Nano 2000C. After cleaning the anomalous data, different machine learning models were trained to determine the significant wavelengths that contribute most towards chlorophyll-a measurement. Finally, a decision tree model with 5 important features was chosen based on the lowest Root Mean Square and Mean Absolute Error when it was tested on the validation set. The final model resulted in a mean error of ±0.9 μg/L when applied on the test set. The total cost for the device was around CAD 135. For the next approach, a rapid system has been proposed using electric impedance spectroscopy (EIS) to measure the concentration of chlorophyll-a, extracted into 95%(v/v) ethanol. Two electrodes accompanied with a high precision impedance converter from Analog Device was used for the development of the sensor. The system was tuned for a fixed electrode orientation, effective area, electrode to electrode distance and excitation voltage by studying different relevant experiments. The proposed sensor was calibrated using the impedance of 95%(v/v) ethanol. Extracted chlorophyll solutions of 60 different concentrations were prepared. At least 5 readings per sample were taken using the proposed system from 1.5 kHz to 7.5 kHz. Samples were then analyzed using standard methods by a spectrophotometer (Genesys20) from Thermo Scientific. Study of Pearson coefficient, principal component analysis, variance inflation factor and backward elimination were used to identify the significant features for chlorophyll-a measurement using EIS. Finally, a simple linear regression model with 11 important features in the range 2.3kHz to 4.7kHz was chosen based on the lowest Root Mean Square (RMS) and Mean Absolute (MA) Error. The coefficient of determination, R2 of the fitted model was 0.93. MAE for the final proposed model is ±0.904 μgL-1 when applied on the test set.
Chlorophyll-a, Electrical impedance spectroscopy, spectroscopy, water quality monitoring, rapid sensor
Master of Science (M.Sc.)
Electrical and Computer Engineering