Improving the performance of non-linear multi-objective optimization problems

Within optimization, the added accuracy in using non-linear relationships often comes with the trade-off of longer run times. A software solution was developed which allows modellers to easily import a seed to an RTC-Tools optimization run. Combined with time-dependent linearization, performance gains of 60% regarding run-times were observed in complex, non-linear multi-objective operational systems. With improved run-times, non-linear optimization problems become more accessible. Clients can solve more accurate models of their systems, allowing them to make more informed decisions about how they use energy. Additionally, the advice of the model is far more valuable when run times are fast and advice can be implemented more quickly before conditions change.