Energy optimal control of thermal comfort in trams

Using linearised models, cascaded control loops are established. A model-predictive controller regulates the thermal comfort inside the tram as a master controller and outputs an auxiliary control variable to the slave controller. A heuristic, realised as mixed-integer optimisation, converts the auxiliary control variable in an energy-optimal way into switching commands for the HVAC. For validation, the method is implemented on a rapid controller prototyping platform. Testing of the control approach is done during several climatic wind tunnel measurements. The annual energy consumption is extrapolated using these results.