Design of algorithms for improving resilience of sensors in space exploration
Tipus de documentProjecte Final de Màster Oficial
Data2021-05-26
Condicions d'accésAccés obert
Tots els drets reservats. Aquesta obra està protegida pels drets de propietat intel·lectual i
industrial corresponents. Sense perjudici de les exempcions legals existents, queda prohibida la seva
reproducció, distribució, comunicació pública o transformació sense l'autorització del titular dels drets
Abstract
The main goal of this thesis is to design and implement software, based on artificial neural networks, capable of predicting data values for failures in the wind sensor aboard NASA's rovers. To achieve this objective, it is necessary to understand the context and operation of the wind sensors, designed by the UPC-ETSETB. Distinct conditions and situations are described to characterize the variables and neural networks proposed. This project demonstrates that it is possible to obtain variables, which entails that the behavior of a particular sensor can be predicted by the conduct of the rest. In the process of the project, different machine learning models are constructed according to different data sets and target results. Then, the performance of the model is evaluated by such indicators as error rate. The result of data recovery is measured by root mean square error. At last, the model results are compared with the data interpolation results, and the conclusion is drawn. The result of model prediction is better than that of data interpolation. These models show promising results with different input data and might provide robustness to the measurements of wind sensors. At the same time, the more input data, the higher the accuracy of output results.
MatèriesNeural networks (Computer science), Machine learning, Xarxes neuronals (Informàtica), Aprenentatge automàtic
TitulacióMÀSTER UNIVERSITARI EN ENGINYERIA DE TELECOMUNICACIÓ (Pla 2013)
Col·leccions
Fitxers | Descripció | Mida | Format | Visualitza |
---|---|---|---|---|
shengyi_thesis.pdf | 3,503Mb | Visualitza/Obre | ||
Design of algor ... s in space exploration.pdf | 3,503Mb | Visualitza/Obre |