Multi-objective optimisation of the operation schedule of POWER plants using artificial intelligence

This study describes a methodology for multi-objective optimisation of the day-ahead operation schedule of power plants. The methodology is aligned to systems with operational flexibility, such as combined heat and power plants and heat pumps that use heat storages. When optimising the operation schedule, the storage is used as a source of flexibility, taking into account the boundary conditions of the system, such as the fulfilment of the heat load and the compliance with minimum running times. As an algorithmic base, the multi-dimensional archive of phenotypic elites (MAP-Elites) optimisation algorithm is used. The optimisation goals considered for multi-objective optimisation are the reduction of plant wear, the minimisation of CO
₂
emissions in the electricity mix and the maximisation of the revenue on the electricity market.