Enhancing energy system operation through Model Predictive Control with disturbance compensation

Model Predictive Controllers have demonstrated their effectiveness in optimizing system management across various domains. These controllers operate by predicting the future behavior of the system and computing the optimal control inputs accordingly. To achieve this, they make use of optimization algorithms which rely on forecasts of external disturbances that influence the performance of the system. Typically, these forecasts are assumed to be accurate and predetermined, enabling the controller to provide the optimal set-points for the system over a specified future time horizon. However, discrepancies between the forecasted and actual disturbances affecting the system can lead to suboptimal system performance. When such mismatches occur, the system must adapt its operation despite the precomputed set-points. Without an appropriate mechanism for adjustment, this can result in a significant deviation from the optimal operational trajectory. In the field of energy systems, such a mismatch could lead to an increase in energy consumption or operational cost, despite the attempt to minimize them. For instance, if there is an increase in a user’s energy demand compared to the forecasted one, additional energy conversion units may need to be activated to meet the demand. This reactive adjustment, without a proper rule, often leads to non-optimal system operation. To address this challenge, this study introduces an innovative predictive control approach with a disturbance compensation mechanism. In addition to determining the optimal set-points, the proposed controller estimates a marginal optimal trajectory for the set-points within predefined disturbance ranges (i.e. ± 30 %). The results obtained are used to control the system in case of discrepancies between the forecasted and actual disturbances. This additional capability ensures that the system can dynamically adapt to the actual conditions while maintaining close-to-optimal performance. The efficacy of the proposed controller is demonstrated in a simulation environment within the framework of energy systems. The case study comprises a cogeneration plant, a thermal energy storage unit, and a boiler, which are used to meet the thermal demand of an end-user. The results show its ability to manage the system operation effectively, even in the presence of variations in external disturbances, thereby enhancing overall system efficiency and reducing operational costs.