Modelización y análisis de políticas de mantenimiento en sistemas sometidos a múltiples procesos de degradación

• Autores: Lucía Bautista Bárcena
• Directores de la Tesis: María Inmaculada Torres Castro (dir. tes.), Luis Landesa Porras (codir. tes.)
• Lectura: En la Universidad de Extremadura ( España ) en 2023
• Idioma: español
• Tribunal Calificador de la Tesis: Rosa Elvira Lillo Rodríguez (presid.), Carlos Javier Pérez Sánchez (secret.), Massimiliano Giorgio (voc.)
• Programa de doctorado: Programa de Doctorado en Modelización y Experimentación en Ciencia y Tecnología por la Universidad de Extremadura
• Materias:
  • Matemáticas
    • Investigación operativa
      • Fiabilidad de sistemas
    • Probabilidad
      • Aplicación de la probabilidad
      • Procesos estocásticos
• Enlaces
  • Tesis en acceso abierto en:  TESEO  Dehesa 
• Resumen

  • As systems have become more sophisticated, there has been an increase in the number and nature of their components. Some of them show continuous degradation and others fail suddenly without warning. Another approach to take into account is that the degradation processes affecting the system may appear at different times, triggered by an external process.

    A maintenance strategy for systems subject to multiple degradation processes is proposed. Such systems can refer to systems with multiple components or systems subject to different degradation processes. Both points are equivalent in the following sense: systems subject to different degradation processes can be seen as a system of multiple components where not all the components start to deteriorate at the same instant of time.

    Since the introduction of continuous degradation processes, such as the gamma or the Wiener process, several extensions of these models have been proposed in order to offer a closer approach to real-life situations. A random effects model in which a parameter of the main degradation process follows a random variable is analysed. This allows the variability between the different degradation processes to be represented. In addition, dependencies that arise between components are addressed by studying the Pearson¿s correlation coefficient in the case of imperfect repairs in a system subject to bivariate degradation.