Optical multiparameter platform for the quality monitoring of liquids

• Autor(a) principal: Duarte, Daniel André Pires
• Data de Publicação: 2020
• Idioma: eng
• Título da fonte: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Texto Completo: http://hdl.handle.net/10773/29203
• topic: Optic sensors; Multiparameter sensors; Liquid sensors; Color and turbidity sensor; Refractive index sensor; Machine learning; Data fusion; Internet of things

The next decade will be established with the maturation and massification of the internet of things, known as IoT. The IoT not only will be used by the domestic consumer, with machines and objects having the ability to be permanently connected to the internet, but also will be the backbone for the new industrial revolution 4.0, already underway, where several installed devices in the production chains will promote a great interaction with the user, in permanent communication. It is within this context that this work was done. In this work optical sensors for the monitoring of liquids were developed, mainly for the parameters monitoring of color, turbidity and refractive index. These sensors are developed in a low-cost perspective, having wireless communication capability to send and visualize data to an internet server based platform. This work is initiated with the search of the state of art sensors for liquids. With the survey concluded, the best technologies that suffice the initial sensor objectives are selected, this is, in-line measurement, low-cost, low complexity and best cost-resolution ratio possible. For the color and turbidity, sensors were developed based in the transmission and scattering at 90º of light between plastic optic fibers in an air-gap configuration. In this configuration, the liquid goes through the measurement zone that is situated between the fibers. Depending on the light absorption/scattering profile, the two parameters are measured. Due to the multivariate sensor nature, machine learning and data fusion algorithms were developed and applied to the raw data to achieve more precise and robust measurements of the parameters. For the refractive index, it was developed a sensor based in the total reflection of light in a prism where the refractive index is measured by the profile of the reflected light detected by a charge-coupled device. The sensors were applied in case studies. The color and turbidity sensor was tested in the monitoring of wine pressing, fermentation and maturation. The refractive index sensor was tested to identify the possibility to be used in the monitoring of cleaning solutions for industrial machinery.