Description

This master's thesis deals with machine tools characterized by a periodic behaviour controlled with binary signals and high-power peaks. This require an oversized electrical network which causes high costs because electrical utilities charge greatly the power peaks. To solve this issue, a new peak-shaving methodology is proposed based on model predictive control, to reduce the power-consumption peaks height in machine tools with periodic behaviour. An study is made on which sampling time, model and solver ts better this kind of problem. A test-bench that emulates the electrical behaviour of a machine tool has been build in order to test the proposed methods online and with real data. In the scenarios simulated, the peak height has been reduced between 35% and 15%. The development of this dissertation has been part of the IKERCON project from the Institut de Robòtica i Informatica Industrial in cooperation with Ikergune and pursues reducing machine tools production cost.

UPCommons link.

The work is under the scope of the following projects:

• IKERCON: Control avanzado de procesos complejos de manufactura (web)