RESUMO
The transition of electrochemical sensors from lab-based measurements to real-time analysis requires special attention to different aspects in addition to the classical development of new sensing materials. Several critical challenges need to be addressed including a reproducible fabrication procedure, stability, lifetime, and development of cost-effective sensor electronics. In this paper, we address these aspects exemplarily for a nitrite sensor. An electrochemical sensor has been developed using one-step electrodeposited (Ed) gold nanoparticles (EdAu) for the detection of nitrite in water, which shows a low limit of detection of 0.38 µM and excellent analytical capabilities in groundwater. Experimental investigations with 10 realized sensors show a very high reproducibility enabling mass production. A comprehensive investigation of the sensor drift by calendar and cyclic aging was carried out for 160 cycles to assess the stability of the electrodes. Electrochemical impedance spectroscopy (EIS) shows significant changes with increasing aging inferring the deterioration of the electrode surface. To enable on-site measurements outside the laboratory, a compact and cost-effective wireless potentiostat combining cyclic and square wave voltammetry, and EIS capabilities has been designed and validated. The implemented methodology in this study builds a basis for the development of further on-site distributed electrochemical sensor networks.
RESUMO
A suitable framework for the development of artificial neural networks is important because it decides the level of accuracy, which can be reached for a certain dataset and increases the certainty about the reached classification results. In this paper, we conduct a comparative study for the performance of four frameworks, Keras with TensorFlow, Pytorch, TensorFlow, and Cognitive Toolkit (CNTK), for the elaboration of neural networks. The number of neurons in the hidden layer of the neural networks is varied from 8 to 64 to understand its effect on the performance metrics of the frameworks. A test dataset is synthesized using an analytical model and real measured impedance spectra by an eddy current sensor coil on EUR 2 and TRY 1 coins. The dataset has been extended by using a novel method based on interpolation technique to create datasets with different difficulty levels to replicate the scenario with a good imitation of EUR 2 coins and to investigate the limit of the prediction accuracy. It was observed that the compared frameworks have high accuracy performance for a lower level of difficulty in the dataset. As the difficulty in the dataset is raised, there was a drop in the accuracy of CNTK and Keras with TensorFlow depending upon the number of neurons in the hidden layers. It was observed that CNTK has the overall worst accuracy performance with an increase in the difficulty level of the datasets. Therefore, the major comparison was confined to Pytorch and TensorFlow. It was observed for Pytorch and TensorFlow with 32 and 64 neurons in hidden layers that there is a minor drop in the accuracy with an increase in the difficulty level of the dataset and was above 90% until both the coins were 80% closer to each other in terms of electrical and magnetic properties. However, Pytorch with 32 neurons in the hidden layer has a reduction in model size by 70% and 16.3% and predicts the class, 73.6% and 15.6% faster in comparison to TensorFlow and Pytorch with 64 neurons.
Assuntos
Redes Neurais de Computação , Numismática , Coleta de Dados , Aprendizado de MáquinaRESUMO
Modern production equipment is based on the results of quality control as well as process parameters. The magnetic anisotropy of materials is closely connected to internal mechanical stress by the Villari effect, and also to hardening effects due to plastic deformations, and could therefore provide an interesting basis for process control. Nevertheless, the analysis of anisotropic properties is extremely sensitive to sensor and workpiece misalignments, such as tilting. In this work, a novel eddy current sensor system is introduced, performing a non-contact measurement of the magnetic anisotropy of a workpiece and realizing a separation and correction of tilting effects. The measurement principle is demonstrated with the example of two samples with different magnetic anisotropy values induced by cold forming. Both samples are analyzed under different tilt angles between the sensor axis and the surface of the workpiece. In this work, digital signal processing is demonstrated on the acquired raw data in order to differentiate the effects of tilt and of anisotropy, with the use of preliminary results as an example of two prepared samples.
