RESUMO
Fourier transform near-infrared (FT-NIR) spectroscopy is a versatile and non-destructive analytical tool widely utilized in industries such as food, pharmaceuticals, and agriculture. While traditional FT-NIR instruments pose limitations in terms of cost and complexity, the advent of portable and affordable systems like NeoSpectra Scanners has broadened accessibility. Partial Least Squares Regression (PLSR) stands as an industry-standard method in Chemometrics for analyzing chemical compositions. This work addresses optimizing PLSR models in FT-NIR spectroscopy, focusing on enhancing accuracy and adaptability in material analysis. Unlike traditional PLSR models which often rely on grid searching a limited number of parameters, such as latent variables, the presented approach effectively expands the parameter space. A novel framework combining Bayesian search and stacking techniques is introduced to enable more customization while ensuring time and performance efficiency, along with automation in model development. Bayesian search efficiently explores hyperparameters space, enabling faster convergence to optimal model settings without exhaustive exploration. The proposed stacked model leverages learned knowledge from the top-performing PLSR models optimized through Bayesian methods, amalgamating a unified and potent body of knowledge. Bayesian-stacked models are compared with PLSR models that use grid search for a limited parameter set. Findings show a marked improvement in model performance: a 51.5% reduction in Root Mean Square Error (RMSE) for the training dataset and a 26.1% reduction for the testing dataset, alongside a 10.9% increase in the correlation coefficient square (R2) for the training dataset and a 10.4% increase for the testing dataset. Notably, Bayesian search reduces the model optimization time by approximately 90% compared with the grid search. Furthermore, when addressing instrumental variations, the models demonstrate an additional improvement, evident in the average reduction of 24.1% in the mean range of prediction. Overall, results demonstrate that the presented approach not only increases the prediction accuracy but also offers a more efficient, automated and robust solution for diverse spectroscopic applications.
RESUMO
Optical accelerometers are popular in some applications because of their better immunity to electromagnetic interference, and they are often more sensitive than other accelerometer types. Optical fibers were employed in most previous generations, making micro-fabrication problematic. The optical accelerometers that are suitable for mass manufacture and previously mentioned in the literature have various problems and are only sensitive in one direction (1D). This study presents a novel optical accelerometer that provides 3D measurements while maintaining simple hybrid fabrication compatible with mass production. The operating concept is based on a power change method that allows for measurements without the need for complex digital signal processing (DSP). Springs hold the proof mass between a light-emitting diode and a quadrant photo-detector, allowing the proof mass to move along three axes. Depending on the magnitude and direction of the acceleration affecting the system, the proof mass moves by a certain amount in the corresponding axis, causing some quadrants of the quadrant detector to receive more light than other quadrants. This article covers the design, implementation, mechanical simulation, and optical modeling of the accelerometer. Several designs have been presented and compared. The best simulated mechanical sensitivity reaches 3.7 µm/G, while the calculated overall sensitivity and resolution of the chosen accelerometer is up to 156 µA/G and 56.2 µG, respectively.