MANUEL SOLER NICOLAU: DOCTOR OF INDUSTRIAL ENGINEERING

CITEF’s main assets are its human resources, comprising doctors, engineers and technicians who are totally dedicated to performing industrial R+D+i projects and who are supported by the teaching staff of the Escuela Técnica Superior de Ingenieros Industriales of the Universidad Politécnica de Madrid (ETSII-UPM) [Madrid Technical University’s School of Engineering].



Below is an example of this approach, which is encompassed in Manuel Soler Nicolau’s Doctoral Thesis entitled “Intelligent Multi-objective optimisation of Energy Dimensioning in Signed Railway Systems and Real Time Software Platforms“, through which he obtained his Doctorate in Industrial Engineering under the instruction of Professor José Manuel Mera.

The Thesis was presented on 3 February at ETSII-UPM, gaining a “cum laude” classification. While working on his Thesis, Dr Nicolau developed an effective methodology to improve the Railway Traction Electric System design process. The process employed to improve the designs of the electric systems used to power traction is usually based on a repetitive procedure that, according to the experience of the designer and the limits set by regulations, is adjusted depending on analyses through simulations and continuous changes to the initial design. The Methodology considered in this Thesis involves automating the process by directly applying Artificial Intelligence, represented in multi-objective optimisation algorithms (NSGA-II, AMGA-II and ɛ-MOEA).

The automatic process takes into account issues such as the cost of environmental impact, difficulty connecting to the general distribution network, critical maintenance zones and the cost of the elements used in the power system. Ultimately, the Methodology will obtain a set of optimum solutions that, as well as having the lowest overall cost, will contain automatic filters to validate the power system in terms of minimum voltages, rail-earth return voltage and power capacity supplied by substations. It is also able to exclude any designs that exceed previously established budget limits.

The guided metaheuristic search explores a large number of possibilities automatically so that the user can choose the one that is most suited to his/her needs from among the best ones found. All this can be done in a relatively short overall time.