This article presents a computationally efficient and high performing approximate long-horizon model predictive control (MPC) for permanent magnet synchronous motors (PMSMs). Two continuous control set MPC (CCS-MPC) formulations are considered: the classical current tracking delta MPC (Del-MPC) and the torque tracking economic MPC (EMPC). To achieve offset-free torque tracking under model uncertainties and in all regions of operation, a disturbance observer and a dq -current reference generator are used. To enable real-time implementation of the long-horizon CCS-MPC, the development of a real-time capable solver is not required, since MPC approximation based on deep neural networks (DNNs) is considered and utilized for controller’s evaluation at run time. The approximation is done by training the DNN to learn the MPC functionality based on offline-generated training data and in an open-loop manner. The robust and offset-free tracking performance of the proposed DNN-based approximate long-horizon Del-MPC and EMPC has been validated through simulation and real-time implementation at test bench and compared to the state-of-the-art field oriented control (FOC) using internal model controller with field-weakening (FW) part and to the exact short-horizon MPC based on the fast gradient method (FGM-MPC). Results show that the long-horizon MPC can achieve significantly faster torque transient responses in comparison with the short-horizon FGM-MPC and the conventional FOC, especially in FW region.

The present study aims at analyzing and assessing the performance of grid-connected photovoltaic (PV) systems, where the considered arrangement is the two-stage PV system. Normally, the maximum power point tracking (MPPT) process is utilized in the first stage of this topology (DC-DC). Furthermore, the active and reactive power control procedure is accomplished in the second stage (DC-AC). Different control strategies have been discussed in the literature for grid integration of the PV systems. However, we present the main techniques, which are considered the commonly utilized and effective methods to control such system. In this regard, and for MPPT, popularly the perturb and observe (P&O) and incremental conductance (INC) are employed to extract the maximum power from the PV source. Moreover, and to improve the performance of the aforementioned methods, an adaptive step can be utilized to enhance the steady-state response. For the inversion stage, the well-known and benchmarking technique voltage-oriented control, the dead-beat method, and the model predictive control algorithms will be discussed and evaluated using experimental tests. The robustness against parameters variation is considered and an extended Kalman filter (EKF) is used to estimate the system’s parameters. Future scope and directions for the research in this area are also addressed.

Maximum power point tracking (MPPT) techniques extract the ultimate power from the photovoltaic (PV) source. Therefore, it is a fundamental control algorithm in any PV configuration. The research in this area is rich and many MPPT methods have been presented in the literature. However, in the current study, we focus on the PV model-based MPPT algorithms. In this regard, the classification of this category can be mainly divided into curve fitting methods and techniques based on the mathematical model or characteristics of the PV source. The objective of the PV model-based MPPT algorithm is to allocate the position of the maximum power point (MPP). Thus, no searching efforts are required to capture that point, which makes it simple and easy to implement. Consequently, the aim of this study is to give an overview of the most commonly utilized model-based MPPT methods. Furthermore, discussion and suggestions are also addressed to highlight the gap in this area. The main methods from the literature are compared together. The comparison and evaluation are validated using an experimental hardware-in-the-loop (HIL) system, where high efficiency (more than 99%) can be obtained with a simple calculation procedure and fast convergence speed.

In this paper, a comparative review for maximum power point tracking (MPPT) techniques based on model predictive control (MPC) is presented in the first part. Generally, the implementation methods of MPPT-based MPC can be categorized into the fixed switching technique and the variable switching one. On one side, the fixed switching method uses a digital observer for the photovoltaic (PV) model to predict the optimal control parameter (voltage or current). Later, this parameter is compared with the measured value, and a proportional–integral (PI) controller is employed to get the duty cycle command. On the other side, the variable switching algorithm relies on the discrete-time model of the utilized converter to generate the switching signal without the need for modulators. In this regard, new perspectives are inspired by the MPC technique to implement both methods (fixed and variable switching), where a simple procedure is used to eliminate the PI controller in the fixed switching method. Furthermore, a direct realization technique for the variable switching method is suggested, in which the discretization of the converter’s model is not required. This, in turn, simplifies the application of MPPT-based MPC to other converters. Furthermore, a reduced sensor count is accomplished. All conventional and proposed methods are compared using experimental results under different static and dynamic operating conditions.

This article proposes a dual-vector finite-control-set model predictive control (FCS-MPC) with reduced complexity for a novel nine-level active neutral point clamped (ANPC) converter. This topology considerably reduces the used number of power switches compared to other topologies. Only nine power switches and two flying capacitors (FCs) are used to generate nine voltage levels. The proposed MPC scheme notably reduces the computational burden by directly locating the best two vectors without the need for multiple evaluations of the cost function as in the conventional method. Using one weighting factor in the cost function, three objectives are considered, namely, current tracking, FCs voltage control, and dc-link stabilization, reducing the heavy effort of coordinating weighting factors. Mathematical analyzes were carried out to determine the optimal duration of the selected voltage vectors. While the sequence of the two voltage vectors is identified based on the total harmonic distortion (THD) definition to minimize its value. Compared with standard FCS-MPC, lower steady-state errors, lower THDs, better harmonic distribution, and shorter execution times are achieved. The proposed MPC method is validated and compared with other prior-art control methods through experimental implementation.

In this article, an improved finite-control-set model predictive control (FCS-MPC) is presented for an active neutral point clamped (ANPC) topology. The considered converter significantly reduces the required power electronics components compared with other common dc-link converters, where only seven active switches, one bidirectional switch, and two floating capacitors (FCs) are employed to produce nine levels in the phase voltage. The developed FCS-MPC handles three control objectives with only one weighting factor, namely, current control, FC balancing, and NP potential stabilization, which reduces the cumbersome effort required for weighting factors coordination. In addition, the number of iterations required to identify the optimal vector is significantly reduced, which, in turn, reduces the execution time of the algorithm. The proposed control method empowers the considered converter to operate in different modes under the faulty condition of the bidirectional switch without any structure modification, which guarantees continuous operation of the converter while ensuring the balancing of FCs and dc-link capacitors in all operating modes. The sensitivity of the proposed FCS-MPC to parameter mismatch, which is a basic issue of MPC-based techniques, is tackled by employing an extended Kalman filter (EKF) to online estimate the system parameters. The proposed FCS-MPC algorithm is experimentally validated and compared with the conventional FCS-MPC method under different operating conditions.

In this paper, a new proposal for the implementation of the well-known direct power control (DPC) technique in grid-connected photovoltaic (PV) systems is suggested. Normally, the DPC is executed using a look-up table procedure based on the error between the actual and reference values of the active and reactive power. Thus, the structure of the DPC is simple and results in a fast transient behavior of the inner current loop (injected currents). Therefore, in the current study, the DPC is reformulated using a dead-beat function. In this formulation, the reference voltage vector (RVV) is obtained in the α−reference frame. Consequently, the switching states for the inverter can be obtained based on the sign of the components of the RVV. The suggested DPC is compared with the conventional one and other switching tables, which are intended for performance enhancement. Furthermore, an extended Kalman filter (EKF) is utilized to eliminate all grid-voltage sensors. Moreover, the switching frequency of the proposed technique is minimized without any need for weighting factors or cost function evaluation. The overall control technique is validated using a hardware-in-the-loop (HIL) experimental set-up and compared with other schemes under different operating conditions.

The study focuses on addressing the growing concern of noise pollution resulting from 
increased transportation. Effective strategies are necessary to mitigate the impact of noise pollution. 
The study utilizes noise regression models to estimate road-traffic-induced noise pollution. However, 
the availability and reliability of such models can be limited. To enhance the accuracy of predictions, 
optimization techniques are employed. A dataset encompassing various landscape configurations 
is generated, and three regression models (regression tree, support vector machines, and Gaussian 
process regression) are constructed for noise-pollution prediction. Optimization is performed by fine- 
tuning hyperparameters for each model. Performance measures such as mean square error (MSE), 
root mean square error (RMSE), and coefficient of determination (R2 ) are utilized to determine the 
optimal hyperparameter values. The results demonstrate that the optimization process significantly 
improves the models’ performance. The optimized Gaussian process regression model exhibits the 
highest prediction accuracy, with an MSE of 0.19, RMSE of 0.04, and R2 reaching 1. However, this 
model is comparatively slower in terms of computation speed. The study provides valuable insights 
for developing effective solutions and action plans to mitigate the adverse effects of noise pollution.