Publication
Railway signal monitoring - an algorithmic approach for improved. Maintenance strategies
| Summary: | Abstract This study is part of Solvit’s project SIGRail Monitoring and it aims to develop and optimise algorithms that will enhance railway monitoring and maintenance processes. The work is divided into four sections. The initial stage of the project entails the formulation of an ensemble method for the identification of the railway line on which the train is situated, utilising geographical coordinates and machine learning techniques. Subsequently, the algorithm for obtaining the kilometric point (PK) is significantly enhanced combining an artificial neural network (ANN) and the golden section method for identifying the optimal distance between two points in space. The third phase of the study requires defining the direction of the train, then separating and counting the journeys within a given file. This is a crucial step, as it forms the basis for the subsequent phase, the fourth stage, in which a dynamic algorithm is employed to generate key performance indicators (KPIs), evaluating the quality of each journey based on a number of factors including the level of signal, the quality of the signal, the handover point, and the serving cell connected. The first, second and third parts have been incorporated into the project, so some of the evidence presented here is the result of this implementation. |
|---|---|
| Main Authors: | Henriques, Inês Figueiredo Leão |
| Subject: | Railway monitoring Machine learning Signal evaluation Anomaly detection Optimization Monitorização ferroviária Aprendizagem automática Avaliação de sinal Deteção de anomalias Otimização |
| Year: | 2024 |
| Country: | Portugal |
| Document type: | master thesis |
| Access type: | open access |
| Associated institution: | Universidade Politécnica de Lisboa |
| Language: | English |
| Origin: | Repositório Científico da Universidade Politécnica de Lisboa |
| Summary: | Abstract This study is part of Solvit’s project SIGRail Monitoring and it aims to develop and optimise algorithms that will enhance railway monitoring and maintenance processes. The work is divided into four sections. The initial stage of the project entails the formulation of an ensemble method for the identification of the railway line on which the train is situated, utilising geographical coordinates and machine learning techniques. Subsequently, the algorithm for obtaining the kilometric point (PK) is significantly enhanced combining an artificial neural network (ANN) and the golden section method for identifying the optimal distance between two points in space. The third phase of the study requires defining the direction of the train, then separating and counting the journeys within a given file. This is a crucial step, as it forms the basis for the subsequent phase, the fourth stage, in which a dynamic algorithm is employed to generate key performance indicators (KPIs), evaluating the quality of each journey based on a number of factors including the level of signal, the quality of the signal, the handover point, and the serving cell connected. The first, second and third parts have been incorporated into the project, so some of the evidence presented here is the result of this implementation. |
|---|
Funded activities
No funded projects found.