Titulo Estágio
Multisensor data fusion learning models for the NanoSen-AQM project
Áreas de especialidade
Sistemas Inteligentes
Local do Estágio
Laboratório de Redes Neuronais (LARN-CISUC)
Enquadramento
This internship will take place in the context of the NanoSen-AQM project. The challenge of the NanoSen-AQM project is to monitor ambient air pollution and inform the public of air quality in real time in a sustainable way. The goal is to develop an electronic system based on low cost and low consumption sensors and validate the system at different locations in the Sudoe territory, based on certified instruments for measuring air pollutants.
The electronic system uses gas sensors based on nanotechnology and microelectronics, computer learning techniques to discriminate and quantify toxic gases in the air, and cloud computing technology for managing and visualizing air quality. Small in size, lightweight and easy to use, the system is easily integrable into stations, mobile units and personal air pollution measurement equipment and thus suitable for use in sensor networks. These provide high spatial and temporal resolution data, which allow the validation of predictive models of air quality.
The main outputs are high performance nanosensors for the detection of toxic gases in the air; multi-sensor systems adaptable to a wide variety of platforms for monitoring air quality; and a cloud computing system to monitor and predict air quality, and inform and raise public awareness about air quality.
The project involves universities, R&D centers, SMEs and public administrations in Spain, France and Portugal. The transnational nature of the partnership allows the value chain to be covered and addresses the transboundary nature of air pollution.
Objetivo
In this internship the student should conceive and develop methods and architectures for sensor data pre-processing and data fusion. This activity includes the following actions:
1) Sensor data pre-processing:
• Standardization and sizing.
• Reduction of size.
• Sensor data filtering: Spurious data and noise.
• Management of lost data.
2) Development of the aggregate model for data fusion:
• Elaboration of self-organized maps to carry out the visualization of the relevant data of 1).
• Design of support vector machines for the detection of air pollutants.
• Test and validate the model.
Plano de Trabalhos - Semestre 1
- Analyse data and define representation structure (1 month);
- Define the architecture of the pre-processing module (1 months);
- Implementation and testing of the pre-processing module (2 months)
- Write intermediate report (1 month);
Plano de Trabalhos - Semestre 2
- Analyse data sources, define representation and fusion architecture (2 months);
- Implementation and testing of the aggregate model for data fusion (3 months)
- Write final report (1 month);
Condições
This work should take place in the context of an european research project. A 3-month scholarship of 745 euros per month is foreseen for this work, renewable for another 3 months.
Orientador
Orientador: Joel Arrais (jpa@dei.uc.pt); Coorientadoras: Bernardete Ribeiro e Catarina Silva
jpa@dei.uc.pt 📩