Network Based Model Predictive Control Research Articles

Advancing Sustainable Building Practices: Intelligent Methods for Enhancing Heating and Cooling Energy Efficiency

Our work is dedicated to enhancing sustainability through improved energy efficiency in buildings, with a specific focus on heating and cooling control and the optimization of thermal comfort of occupants. With an energy consumption of more than 60% in buildings, HVAC systems are the biggest energy users. By integrating advanced technology, data algorithms, and digital twins, our study aims to optimize energy performance effectively. We have developed a Neural Network-based Model Predictive Control (NNMPC) to achieve this goal. Leveraging technologies such as MQTT communication, Wi-Fi modules, and field-programmable gate arrays will enhance scalability and flexibility. Our findings demonstrate the efficacy of the NNMPC system deployed on the PYNQ board for reducing sensible thermal energy usage for both cooling and heating purposes. Compared to traditional On/Off control systems, the NNMPC achieved an impressive 40.8% reduction in heating energy consumption and a 37.8% decrease in cooling energy consumption in 2006. In comparison to the On/Off technique, the NNMPC demonstrated a 25.6% reduction in annual heating energy consumption and a 28.8% drop in annual cooling energy consumption in the simulated year of 2017. We observed that, across all strategies and platforms, there were no instances where the Predicted Mean Vote (PMV) fell below −0.5. However, a significant proportion of PMV values (ranging from 65% to 83%) were observed between −0.5 and 0.5, signifying a high level of occupant comfort. Additionally, for PMV values between 0.5 and 1.0, percentages ranged from 16% to 33% for both years. Importantly, the NNMPC exhibited notable efficiency in maintaining occupants’ comfort within this range, requiring less energy while ensuring highly satisfactory environments.

Energy flexibility quantification of a tropical net-zero office building using physically consistent neural network-based model predictive control

Building energy flexibility plays a critical role in demand-side management for reducing utility costs for building owners and sustainable, reliable, and smart grids. Realizing building energy flexibility in tropical regions requires solar photovoltaics and energy storage systems. However, quantifying the energy flexibility of buildings utilizing such technologies in tropical regions has yet to be explored, and a robust control sequence is needed for this scenario. Hence, this work presents a case study to evaluate the building energy flexibility controls and operations of a net-zero energy office building in Singapore. The case study utilizes a data-driven energy flexibility quantification workflow and employs a novel data-driven model predictive control (MPC) framework based on the physically consistent neural network (PCNN) model to optimize the building energy flexibility. To the best of our knowledge, this is the first instance that PCNN is applied to a mathematical MPC setting, and the stability of the system is formally proved. Three scenarios are evaluated and compared: the default regulated flat tariff, a real-time pricing mechanism, and an on-site battery energy storage system (BESS). Our findings indicate that incorporating real-time pricing into the MPC framework could be more beneficial to leverage building energy flexibility for control decisions than the flat-rate approach. Moreover, adding BESS to the on-site PV generation improved the building self-sufficiency and the PV self-consumption by 17% and 20%, respectively. This integration also addresses model mismatch issues within the MPC framework, thus ensuring a more reliable local energy supply. Future research can leverage the proposed PCNN-MPC framework for different data-driven energy flexibility quantification types.

Multi-Time Scale Optimal Dispatch for the Wind Power Integrated System With Demand Response of Data Centers Based on Neural Network-Based Model Predictive Control

Multi-Time Scale Optimal Dispatch for the Wind Power Integrated System With Demand Response of Data Centers Based on Neural Network-Based Model Predictive Control

Adaptive model predictive control based on neural networks for Hover attitude tracking with input saturation and unknown disturbances

The anti-disturbance control of quadrotor attitude tracking under saturation constraints is a difficult problem. In this paper, a neural network-based model predictive controller for quadrotor systems with input saturation and external disturbances is developed. The unmodeled dynamics and external disturbances of the system are simplified to the disturbance superimposed on the nominal system, and the gradient descent neural networks are used to complete the estimation and compensation of the disturbance. The adaptive model predictive controller is designed based on the nominal system. The disturbance value estimated by the neural network adaptively adjusts the control constraints in the model predictive controller. The robustness and anti-disturbance of the designed controller are analyzed. The experiments show that, compared to the robust model predictive control, the algorithm proposed in this paper reduces the steady-state mean errors of the yaw, pitch, and roll attitude channels of the Hover system. Specifically, the algorithm results in a decrease of 2.622%, 2.292%, and 1.192% without external disturbances and 2.056%, 4.17%, and 0.956% with outside disturbances. Experimental results confirm the effectiveness of the proposed method.

LSTM-Based Model Predictive Control of Piezoelectric Motion Stages for High-Speed Autofocus

In this paper, we proposed a neural network-based model predictive control of piezoelectric motion stages for autofocus. Rather than using an internal controller to account for the problematic hysteresis effects of the piezoelectric motion stage, we use the long short-term memory unit to integrate the hysteresis effects and the focus measurement into a single learning-based model. Subsequently, a model predictive control method is developed based on this long short-term memory model that successfully finds the optimal focus position using a series of focus measurements derived from a sequence of images. To further improve the speed of the long short-term based model predictive control, an optimized backpropagation algorithm is proposed that optimizes the model predictive control cost function. Experiments verified our proposed method reduces at minimum 30&#x0025; regarding autofocus time when compared to well-known ruled-based autofocus methods and other learning-based methods. Videos of the experiments are available at https:&#x002F;&#x002F; youtu.be&#x002F;AyvnMIq48Vc and <uri>https://youtu.be/SQN3ETbuf2g</uri>.

Collision Avoidance Strategy for Unmanned Surface Vessel Considering Actuator Faults Using Kinodynamic Rapidly Exploring Random Tree-Smart and Radial Basis Function Neural Network-Based Model Predictive Control

Path planning and tracking are essential technologies for unmanned surface vessels (USVs). The kinodynamic constraints and actuator faults, however, bring difficulties in finding feasible paths and control efforts. This paper proposes a collision avoidance strategy for USV by developing the kinodynamic rapidly exploring random tree-smart (kinodynamic RRT*-smart) algorithm and the fault-tolerant control method. By utilizing the triangular inequality and the intelligent biased sampling strategy, the kinodynamic RRT*-smart shows its advantages in terms of path length, cost and running time. With consideration of kinodynamic constraints, a feasible and collision-free trajectory can be provided. Then, a radial basis function neural network-based model predictive control (RBF-MPC) method was designed that compensates for the model’s uncertainties by developing the radial basis function neural network (RBF-NN) approximator and by constructing a feedback-state training dataset in real time. Furthermore, two types of fault situation were analyzed considering the thruster failure. We established the faults’ mathematical models and investigated the fault-tolerant strategies for different fault types. The simulation studies were conducted to validate the effectiveness of the proposed strategy. The results show that the proposed planning and control methods can avoid obstacles in faulty conditions.

Control of the Multi-Timescale Process Using Multiple Timescale Recurrent Neural Network-Based Model Predictive Control

Control of the Multi-Timescale Process Using Multiple Timescale Recurrent Neural Network-Based Model Predictive Control

Dynamic Obstacle Avoidance for USVs Using Cross-Domain Deep Reinforcement Learning and Neural Network Model Predictive Controller

This work presents a framework that allows Unmanned Surface Vehicles (USVs) to avoid dynamic obstacles through initial training on an Unmanned Ground Vehicle (UGV) and cross-domain retraining on a USV. This is achieved by integrating a Deep Reinforcement Learning (DRL) agent that generates high-level control commands and leveraging a neural network based model predictive controller (NN-MPC) to reach target waypoints and reject disturbances. A Deep Q Network (DQN) utilized in this framework is trained in a ground environment using a Turtlebot robot and retrained in a water environment using the BREAM USV in the Gazebo simulator to avoid dynamic obstacles. The network is then validated in both simulation and real-world tests. The cross-domain learning largely decreases the training time (28%) and increases the obstacle avoidance performance (70 more reward points) compared to pure water domain training. This methodology shows that it is possible to leverage the data-rich and accessible ground environments to train DRL agent in data-poor and difficult-to-access marine environments. This will allow rapid and iterative agent development without further training due to the change in environment or vehicle dynamics.

ResNet and PolyNet Based Identification and (MPC) Control of Dynamical Systems: A Promising Way

This paper deals with model predictive control synthesis which take benefits from artificial neural networks to model (non-linear) dynamical system. More precisely, thanks to a systematic and rigorous methodology, it is shown that residual networks (ResNet) and PolyInception networks (PolyNet) neural network architectures, developed initially for image recognition, are very good candidate for i) identification of dynamical systems, ii) being used as embedded model in a model predictive control laws. Concretely, the widely used non-linear dynamical system quadruple tank process is used as a benchmark. The neural network architectures studied are i) feedforward networks as a reference point, and the two other linked to Euler integration method ii) residual networks and iii) PolyInception networks. Networks training is performed by mixing classical back-propagation algorithm and hyperparameters optimisation through heuristics. The identification results provided show that neural networks of types ii) and iii) perform better than the classical one i), with a better generalisation capability. Finally, model predictive controllers are synthesized based on the various networks trained. The simulation results obtained for controlling water levels of a 4 tanks system benchmark give interesting insights. They show that residual networks based model predictive control is better suited than feedforward networks and PolyInception networks based ones, both taking into account computation time and set point errors.

Model Predictive Control Based Type-3 Fuzzy Estimator for Voltage Stabilization of DC Power Converters

The stability of power electronic devices is prone to instability when supplying constant power loads (CPL) due to their negative impedance specifications. This necessitates an efficient control methodology to overcome the destabilization issue and regulate the response of the output voltage. In particular, this article proposes a novel interval type-III fuzzy neural network-based model predictive control (IT3FNNMPC) to mitigate the destructive effects of CPL in dc/dc boost converters implemented in the microgrid (MG) and power system distribution. In a typical MGs configuration, the IT3FNNMPC controller is formulated and designed in a constraint control problem to achieve a fast-transient response of system voltage. The feasibility and applicability of the designed IT3FNNMPC controller have been validated by hardware-in-the-loop real-time simulations.

Artificial neural network-based model predictive control for optimal operating conditions in proton exchange membrane fuel cells

In the large-scale commercialization of proton exchange membrane fuel cells (PEMFC), efficient control of the dynamic operation requires the consideration of complex nonlinearities in current, temperature, pressure, and membrane hydration. In this study, an artificial neural network (ANN) model was developed and integrated with model predictive control (MPC) to provide optimal conditions for operating the PEMFC system under different current load changes. For this purpose, after modeling a 1.2 kW PEMFC in a MATLAB/Simulink environment, the accuracy of the model was validated within a maximum relative error of 2.30% compared to the experimental data. Next, the power production, temperature, and internal relative humidity of the model were extracted under all feasible operating conditions. The extracted data were used to train the nonlinear autoregressive network with exogenous inputs (NARX), and hyperparameters of neural networks were selected through Bayesian optimization. The NARX model trained with optimal parameters had a mean square error (MSE) of 2.14 × 10−4 and was integrated into the MPC to obtain neural network model predictive control (NNMPC). The developed NNMPC provides the optimal operating temperature and pressure for the PEMFC model at 10-s intervals through iterative simulation of the NARX under Bayesian optimization. To verify the performance of the control system, the developed NNMPC and fixed-setpoint scenario were compared. Each control scenario consisted of three current load profiles: low, medium, and high current. The NNMPC yielded improved system power by maintaining the optimal stack temperature, cathode pressure and membrane hydration for current changes at a low computational cost. The optimized power of the PEMFC system showed an average increase of 10.9% compared to fixed-setpoint condition.

A Customized Artificial Pancreas System with Neural Network based Model Predictive Control for Type 1 Diabetic Rats

Improvements in technology have allowed researchers to develop artificial pancreas systems (APS) that mimic glucose regulation function of a healthy pancreas. This study aims to develop a closed-loop APS to control blood glucose level (BGL) of diabetic rats in vivo. To this end, we developed a customized APS with a control algorithm based on model predictive control (MPC).

Neural Network Based Model Predictive Control for a Quadrotor UAV

A dynamic model that considers both linear and complex nonlinear effects extensively benefits the model-based controller development. However, predicting a detailed aerodynamic model with good accuracy for unmanned aerial vehicles (UAVs) is challenging due to their irregular shape and low Reynolds number behavior. This work proposes an approach to model the full translational dynamics of a quadrotor UAV by a feedforward neural network, which is adopted as the prediction model in a model predictive controller (MPC) for precise position control. The raw flight data are collected by tracking various pre-designed trajectories with PX4 autopilot. The neural network model is trained to predict the linear accelerations from the flight log. The neural network-based model predictive controller is then implemented with the automatic control and dynamic optimization toolkit (ACADO) to achieve real-time online optimization. Software in the loop (SITL) simulation and indoor flight experiments are conducted to verify the controller performance. The results indicate that the proposed controller leads to a 40% reduction in the average trajectory tracking error compared to the traditional PID controller.

Neural network-based model predictive control for thin-film chemical deposition of quantum dots using data from a multiscale simulation

Recently, thin-film deposition of quantum dot (QDs) to manufacture solar cells and displays have received significant attention due to the lucrative optoelectronic properties of these devices. Unfortunately, (a) the existing macroscopic thin-film deposition models in the literature do not consider the surface-level interactions; (b) the detailed surface-level models do not consider the entire thin-film deposition ensemble; and (c) multiscale modeling studies considering both the scales are not tailored for QD systems. Thus, to address this knowledge gap, in this work, a multiscale thin-film deposition model is developed. First, the droplet distribution and evaporation dynamics during the thin-film deposition of QDs are described using heat and mass balance equations. Second, a microscopic discrete-element method (DEM)-based particle aggregation model that describes the surface-level particle interactions is developed and combined with the macroscopic dynamics. Furthermore, a model predictive controller (MPC) is designed to regulate the film thickness and minimize the film roughness by manipulating key process variables. To design a feasible MPC, a computationally efficient artificial neural network (ANN) model of the thin-film deposition model is constructed, and it is incorporated within the MPC. The closed-loop simulation results showcase the capability of the MPC to achieve the required film thickness and minimize the roughness.

Improved Active Damping Stabilization of DAB Converter Interfaced Aircraft DC Microgrids Using Neural Network-Based Model Predictive Control

The stability problem induced by constant power loads (CPLs) is becoming more prominent in more electric aircraft. In this article, the stabilization issue of dc distribution bus with dual-active-bridge (DAB) converters was investigated. On the basis of voltage regulation, an active damping solution based on model predictive control (MPC) was proposed by explicitly incorporating a stabilization term into a cost function. In this solution, to ensure the output voltage performance with a smaller steady-state error and a dynamic voltage overshoot, an adaptive weighting factor was adopted with a stray resistor taken into account. A theoretical design of the weighting factors was performed using the artificial neural network (ANN) method. The proposed approach could supply stiff load voltage and provide good dc-link voltage stabilization, as well as load voltage overshoot clamping. These findings were verified through the computer simulations and practical experiments of an 800-W converter prototype with a switching frequency of 20 kHz, an input voltage of 270 V, and an output voltage of 270 V.

Neural network-based model predictive control system for optimizing building automation and management systems of sports facilities

Sports facilities are considered complex buildings due to their high energy demand and occupancy profiles. Therefore, their management and optimization are crucial for reducing their energy consumption and carbon footprint while maintaining an appropriate indoor environmental quality. This work is part of the SportE3.Q project, which aims to manage and optimize the operation of sports facilities. A neural network (NN)-based model predictive control (MPC) management and optimization system is proposed for the heating, ventilation, and air conditioning (HVAC) system of a sports hall in the sports and events complex of Qatar University (QU). The proposed approach provides an integrated dynamic optimization method that accounts for future system behavior in the decision-making process, consisting of a prediction element and an optimizer. A NN is used to implement the dynamic prediction element of the MPC system and is compared with other machine learning (ML)-based models, which are support vector regression (SVR), k-nearest neighbor (kNN), and decision trees (DT). The NN-based model outperforms the other ML models with an average root mean squared error (RMSE) of around 0.06 between the actual and the predicted values, and an average R of 0.99 as NNs are popular for their high accuracy and reliability. Two schemes of the proposed NN-based MPC system are investigated for managing and optimizing the operation of the hall’s HVAC system for enhanced energy use and indoor environment quality, as well as for providing occupancy profile recommendations to aid the facilities’ managers in handling their operation. In alignment with the objective of the SportE3.Q project, up to 46% energy reduction was achieved while jointly optimizing the thermal comfort and indoor air quality. In addition, Scheme 2 of the proposed system provided productive occupancy recommendations for a healthier indoor environment.

Neural network-based model predictive control with fuzzy-SQP optimization for direct thrust control of turbofan engine

A nonlinear model predictive control method based on fuzzy-Sequential Quadratic Programming (SQP) for direct thrust control is proposed in this paper for the sake of improving the accuracy of thrust control. The designed control system includes four parts, namely a predictive model, rolling optimization, online correction, and feedback correction. Considering the strong nonlinearity of engine, a predictive model is established by Back Propagation (BP) neural network for the entire flight envelope, whose input and output are determined with random forest algorithm and actual situation analysis. Rolling optimization typically uses SQP as the optimization algorithm, but SQP algorithm is easy to trap into local optimization. Therefore, the fuzzy-SQP algorithm is proposed to prevent this disadvantage using fuzzy algorithm to determine the initial value of SQP. In addition to the traditional three parts of model predictive control, an online correction module is added to improve the predictive accuracy of the predictive model in the predictive time domain. Simulation results show that the BP predictive model can reach a certain degree of predictive accuracy, and the proposed control system can achieve good tracking performance with the limited parameters within the safe range.

Model Predictive Control With a Cascaded Hammerstein Neural Network of a Wind Turbine Providing Frequency Containment Reserve

This article presents an application of neural network-based Model Predictive Control (MPC) to improve the wind turbine control system’s performance in providing frequency control ancillary services to the grid. A closed-loop Hammerstein structure is used to approximate the behavior of a 5MW floating offshore wind turbine with a Permanent Magnet Synchronous Generator (PMSG). The multilayer perceptron neural networks estimate the aerodynamic behavior of the nonlinear steady-state part, and the linear AutoRegressive with Exogenous input (ARX) is applied to identify the linear time-invariant dynamic part. Using the specific structure of the Cascade Hammerstein design simplifies the online linearization at each operating point. The proposed algorithm evades the necessity of nonlinear optimization and uses quadratic programming to obtain control actions. Eventually, the proposed control design provides a fast and stable response to the grid frequency variations with optimal pitch and torque cooperation. The performance of the MPC is compared with the gain-scheduled proportional-integral (PI) controller. Results demonstrate the effectiveness of the designed control system in providing Frequency Containment Reserve (FCR) and frequency regulation in the future of power systems.

Artificial Neural Network-Based Model Predictive Control Using Correlated Data

This work addresses the problem of implementing model predictive control (MPC) in situations where the training data available for modeling contains possible correlations, and an artificial neural network (ANN)-based model is being used. In particular, we consider a problem where data sets are collected from a process that operates under the closed-loop condition in which correlations are induced between several input and output variables. In this situation, if the correlation problem is not addressed, manipulated inputs (calculated by MPC without considering the specific correlation in the input space) and independently prescribed set-points may require predictions in regions where the model is not trained, resulting in a poor closed-loop performance. To address this issue, principal component analysis (PCA)-based strategies are applied to both the input and output spaces in a way that maintains model validity. To that end, a new constraint on the squared prediction error (SPE) is incorporated into the ANN-based MPC optimization problem to make control actions follow the PCA model built using the training input data. Next, a PCA model is developed using the training output data, and then an optimization problem subject to the SPE constraint is defined to calculate set-points which are achievable. The effectiveness of the proposed ANN-based MPC to track these set-points is demonstrated using a chemical reactor example. Finally, a new autoencoder-based strategy is proposed to compute the achievable set-points. This is performed by replacing the PCA-based constraint with the autoencoder-based constraint in the optimization problem to calculate the set-points. The results indicate that the ANN-based MPC performance is improved when the autoencoder-based set-points are used.

Design of Deep Neural Network Based Model Predictive Controller for a Car-like Mobile Robot

Design of Deep Neural Network Based Model Predictive Controller for a Car-like Mobile Robot