Robust Nonlinear Model Predictive Control with Model Predictive Sliding Mode for Continuous-Time Systems

This paper presents a robust, tube-based nonlinear model predictive controller for continuous-time systems with additive disturbances which cascades two sampled-data model predictive controllers: the first creates a desired path using nominal dynamics and the second maintains the true state close to the nominal state by regulating a sliding variable designed on the error between the true and nominal states. The sampled-data model predictive approach permits easy incorporation of continuous-time sliding mode dynamics, allowing a dynamic boundary layer and tube design to be included. In this way, the control applied to the system capitalizes on the robustness properties of traditional sliding mode control while incorporating system constraints. Stability analysis is presented in the context of input-to-state stability for continuous-time systems. The proposed controller is implemented on two case studies, is compared to benchmark tube-based model predictive controllers, and is evaluated using average root mean square values on the state and input variables, in addition to average integral square and integral absolute error values on the position states. Results reveal the proposed technique responds to higher levels of disturbance with significant increases in control effort; eliminates constraint violation by using of constrained SMC as the secondary controller; and maintains similar tracking performance to benchmark controllers at lower levels of control effort.

Can Deep Models Help a Robot to Tune Its Controller? A Step Closer to Self-tuning Model Predictive Controllers

Motivated by the difficulty roboticists experience while tuning model predictive controllers (MPCs), we present an automated weight set tuning framework in this work. The enticing feature of the proposed methodology is the active exploration approach that adopts the exploration–exploitation concept at its core. Essentially, it extends the trial-and-error method by benefiting from the retrospective knowledge gained in previous trials, thereby resulting in a faster tuning procedure. Moreover, the tuning framework adopts a deep neural network (DNN)-based robot model to conduct the trials during the simulation tuning phase. Thanks to its high fidelity dynamics representation, a seamless sim-to-real transition is demonstrated. We compare the proposed approach with the customary manual tuning procedure through a user study wherein the users inadvertently apply various tuning methodologies based on their progressive experience with the robot. The results manifest that the proposed methodology provides a safe and time-saving framework over the manual tuning of MPC by resulting in flight-worthy weights in less than half the time. Moreover, this is the first work that presents a complete tuning framework extending from robot modeling to directly obtaining the flight-worthy weight sets to the best of the authors’ knowledge.

Development of a Grey Box Thermal Dynamic Model without Building Construction Knowledge

Model Predictive Controllers (MPC) in building Heating Ventilation and Air Conditioning (HVAC) systems have demonstrated significant energy savings when compared to typical on/off controllers. MPCs require information about the building’s thermal dynamics which is challenging to model, especially for older structures without buildings specifications. This research investigates the ability to develop a grey box thermal dynamic model that can determine the net thermal dynamics, without any building construction information. Sensors were installed within a test cell to monitor the building automation system (BAS) points, and collect building element surface temperature data. The simulation program Simulink was used to develop and test iterations of grey box models. The final model, that relies solely on BAS points, is able to predict the ambient temperature for a 3-hour Prediction Window to within 1.7% accuracy. This model demonstrates the potential for more buildings to implement HVAC MPC systems with grey box thermal dynamic modeling

Development of a Grey Box Thermal Dynamic Model without Building Construction Knowledge

Model Predictive Controllers (MPC) in building Heating Ventilation and Air Conditioning (HVAC) systems have demonstrated significant energy savings when compared to typical on/off controllers. MPCs require information about the building’s thermal dynamics which is challenging to model, especially for older structures without buildings specifications. This research investigates the ability to develop a grey box thermal dynamic model that can determine the net thermal dynamics, without any building construction information. Sensors were installed within a test cell to monitor the building automation system (BAS) points, and collect building element surface temperature data. The simulation program Simulink was used to develop and test iterations of grey box models. The final model, that relies solely on BAS points, is able to predict the ambient temperature for a 3-hour Prediction Window to within 1.7% accuracy. This model demonstrates the potential for more buildings to implement HVAC MPC systems with grey box thermal dynamic modeling

Virtual Inertia Control Methods in Islanded Microgrids

Although the deployment and integration of isolated microgrids is gaining widespread support, regulation of microgrid frequency under high penetration levels of renewable sources is still being researched. Among the numerous studies on frequency stability, one key approach is based on integrating an additional loop with virtual inertia control, designed to mimic the behavior of traditional synchronous machines. In this survey, recent works related to virtual inertia control methods in islanded microgrids are reviewed. Based on a contextual analysis of recent papers from the last decade, we attempt to better understand why certain control methods are suitable for different scenarios, the currently open theoretical and numerical challenges, and which control strategies will predominate in the following years. Some of the reviewed methods are the coefficient diagram method, H-infinity-based methods, reinforcement-learning-based methods, practical-swarm-based methods, fuzzy-logic-based methods, and model-predictive controllers.