An improved data-driven based model predictive control for zero-sequence circulating current suppression in paralleled converters

The parallel three-phase grid-connected converters (GCCs) with common dc and ac buses can increase the system power level effectively. However, zero sequence circulating current (ZSCC) will occur inevitably. In this paper, the mechanisms and characteristics of ZSCC in two parallel three-phase GCCs are first investigated based on three typical cases. From the analysis, it was found that not all ZSCC can be suppressed. Generally speaking, ZSCC caused by inconsistent modulation voltages between modules cannot be effectively handled. And ZSCC caused by different injected zero sequence voltages (ZSVs) or asynchronous carriers can be suppressed. Based on the theoretical analysis, identical ZSV injection and carrier synchronization methods without communication are, thus, proposed to suppress ZSCC at the source, which are easily implemented without adding any hardware. Finally, the theoretical analysis and the effectiveness of the proposed methods on ZSCC suppression are well-verified by experiments.

The Zero Sequence Circulating Current (ZSCC) flows inevitably in parallel converters sharing common DC and AC sources. The currents flowing commonly in all converters increase losses and decrease the overall capacity of parallel converters. This paper proposes a simple and effective ZSCC suppression method based on the Space Vector PWM (SVPWM) method with the ZSCC controller. The zero sequence voltage for the proposed SVPWM is calculated based not on the phase voltage references, but on the grid voltage. The limit of linear modulation index of the converters with the proposed method is analyzed compared to other methods, and it is proven that the limit of linear modulation index can be maximized with the proposed method. The effectiveness of the proposed method has been verified through the experimental set-up consisted of the four parallel three-level converters. It has been confirmed that the ZSCC is well suppressed and larger linear modulation index is achieved at the same time with the proposed method. Moreover, the proposed control method does not require any communication between converters to suppress the ZSCC unlike the conventional methods.

Data-driven model predictive control (MPC) has been regarded as an attractive control technology for nonlinear dynamic processes. However, as a model-based approach, data-driven MPC usually suffers from an inaccurate model, process uncertainty, and calculation burden. To provide an efficient and accurate controller for nonlinear dynamic processes, an event-based data-driven adaptive MPC (EDAMPC) scheme is proposed. First, a prediction model employing a self-organizing long short-term memory (SOLSTM) neural network is developed to obtain a compact model structure and improve the generalization ability. The structure of the SOLSTM neural network is constructed dynamically by integrating the activity and significance of the neurons. Second, an error-triggered online learning mechanism is developed to update model parameters adaptively based on prediction errors. The nonlinear process dynamics can be captured in the presence of process uncertainty. Third, an event-based control strategy is introduced to reduce communication and computation resources. The convergence of the SOLSTM neural network and the nominal stability of the EDAMPC scheme is analyzed. Finally, the effectiveness and superiority of the EDAMPC scheme are demonstrated via a numerical case and an industrial application.

This paper proposes a finite set model predictive control (FS-MPC) based thrust maximization technique for linear induction machines used in linear metros. For modeling of the proposed control method, the end effect is taken into consideration. The proposed control method is used to achieve maximum thrust per ampere and to reduce the thrust ripples. It differs from the FS-MPC methods, where the cost function consists of the thrust and angle errors. The thrust error is calculated from the difference between the reference thrust and the predicted thrust, and the angle error is calculated from the difference between the angle of predicted primary current and the angle of the predicted secondary flux in one side and π/4 on the other side. A comparison between the proposed method and the finite set model predictive direct thrust control (FS-MPDTC) is presented to illustrate the superiority of the proposed method. Both simulation and experimental analysis are conducted to validate the effectiveness of the proposed finite set model predictive direct angle control (FS-MPDAC). A prototype test platform is developed in the laboratory with two 3 kW arc induction motors. The simulation model, experimental test platform, and test results are presented in this paper.

In this paper, the switching frequency reduction in two paralleled Uninterruptible Power Supplies (UPS) is studied in order to obtain maximum global system efficiency. Typically, by reducing the switching frequency of a converter, its efficiency increases. However, when connected in parallel, UPSs are subjected to a Zero Sequence Circulating Current (ZSCC) that compromises a correct system operation and increases the losses of the global system. Experimental results show that to achieve the maximum system efficiency, the ZSCC must be effectively eliminated. Finite Control Set Model Predictive Control (FCSMPC) is used as the control scheme.

In this paper, a new data-driven model predictive control (MPC) is considered based on bilinear subspace identification. The system's nonlinear behavior is described with a bilinear subspace predictor structure in MPC framework. Thus, the MPC formulation results in a fixed structure objective function with constraints regardless of the underlying nonlinearity. Therefore, a bilinear predictive control is implemented by exploiting the structural properties of the identified bilinear subspace predictor model. The open-loop optimization problem of MPC that is nonlinear in nature is solved with series quadratic programming (SQP) without any approximations. These improvements and closely integration of modeling and control also eliminate the intermediate design step, which provides a means for data-driven controller design in generalized predictive controller (GPC) framework. Finally, the proposed control approach is illustrated with a simulation of a nonlinear continuously stirred tank reactor (CSTR) system.

This paper proposes a new control scheme for parallel three-phase pulse width modulation converters under generalized unbalanced operating conditions. An average model of the parallel system in positive-sequence synchronous reference frame is derived to analyze the influence of generalized unbalanced operating conditions in ac side. It is seen that the unbalance factors in filter inductance will not only give rise to negative-sequence circulating current, but also contribute to generating zero-sequence circulating current (ZSCC). The negative-sequence circulating current can be inhibited by suppressing the negative-sequence components in ac output currents of parallel modules with a proportional integral resonant (PIR) controller. An improved feed-forward strategy and a PIR controller for ZSCC control are proposed for unbalanced operating conditions. The disturbances in ZSCC caused by unbalance factors in filter inductance can be rejected with feed-forward strategy. Because the disturbance in ZSCC is the fluctuation in grid frequency which can be suppressed by a resonance controller, therefore, a PIR controller is adopted in ZSCC controller. The proposed scheme can effectively suppress the circulating currents between the parallel modules and as a result, the distortions in output currents can be greatly reduced. Experimental results confirm the performance and effectiveness of the proposed method.

Parallel inverters are widely used in the micro-grid systems to meet the demand for increased power capacity and enhanced reliability. However, unbalanced connections between the parallel inverters may cause a spike in the zero-sequence circulating current (ZSCC), leading to performance degradation. To address this issue, a virtual voltage vector-based sequential model-free predictive control method is proposed in this paper for parallel converters. Specifically, extended state observer (ESO) and virtual voltage vectors are introduced into the finite control-set model predictive control (FCS-MPC) architecture to cope with the multi-objective including ZSCC inhibition and robustness performance improvement. Next, to prevent the adjusting of weighting factors, a sequential model-free predictive control solution is constructed. The main feature of the proposed control methodology is that the reliability and robustness of the controlled system can be significantly enhanced without knowing the accurate mathematical models and parameters. Finally, the effectiveness of the suggested approach is verified by performance evaluation.

This work proposes a novel data-driven robust model predictive control (DDRMPC) framework for automatic control of greenhouse temperature and CO2 concentration level. The essential concept is to combine dynamic models of greenhouse temperature and CO2 concentration level with data-driven models that identify uncertainty in weather forecast error. By leveraging a machine learning approach, support vector clustering with weighted generalized intersection kernel, data-driven uncertainty sets for ambient temperature and solar radiation are constructed from historical weather data. A training-calibration procedure that tunes the size of uncertainty sets is implemented to ensure that data-driven uncertainty sets attain appropriate performance guarantee. In order to solve the optimization problem in DDRMPC, an affine disturbance feedback policy that provides tractable approximations of optimal control is utilized. A case study of controlling temperature and CO2 concentration level in a greenhouse is carried out. The results show that the proposed DDRMPC framework can prevent the greenhouse climate from becoming harmful to plant and fruit. DDRMPC approach ends up with 20 % less total economic cost than rule-based control strategy. The proposed DDRMPC approach also gives better control performance comparing to certainty equivalent MPC and robust MPC.

This work presents the design and implementation of a data-driven Nonlinear Model Predictive Control (NMPC) framework for an Unmanned Aerial Vehicle (UAV) equipped with a 3-DOF robotic arm. Real-world datawas collected using the Matrice 100 platform and Dynamixel MX-28AR actuators to identify a high-dimensional linear model via Dynamic Mode Decomposition with Control (DMDc), capturing the interactions between the aerial vehicle and the manipulator across 21 state variables. This DMDc-based model is embedded within the NMPC formulation to predict system behavior over finite horizons. The UAV’s orientation is represented using quaternions, enabling continuous and singularity-free attitude control. Additionally, the redundancy of the UAV-manipulator system allows for the integration of secondary objectives into the cost function, supporting flexible task execution. To meet real-time requirements, the control problem is solved using the ACADOS solver. The resulting controller achieves high-precision tracking while managing internal constraints, demonstrating the potential of data-driven NMPC in aerial manipulation tasks.

Converters parallel operation can increase capacity of the system, but it may lead to potential zero-sequence circulating current, so the control of circulating current was an important goal in the design of parallel inverters. In this paper, the Real Time Digital Simulator (RTDS) is used to model the converters parallel system in real time and study the circulating current restraining. The equivalent model of two parallel converters and zero-sequence circulating current(ZSCC) were established and analyzed, then a strategy using variable zero vector control was proposed to suppress the circulating current. For two parallel modular converters, hardware-in-the-loop(HIL) study based on RTDS and practical experiment were implemented, results prove that the proposed control strategy is feasible and effective.

Strict effluent quality standards and physical limitations of the equipment make accurately controlling dissolved oxygen (DO) concentration in wastewater treatment processes (WWTPs) a challenging task. In this paper, a data-driven model predictive control strategy with performance constraint (PC-MPC) is proposed to ensure accurate tracking of DO concentration set-point. First, a data-driven model is established using the fuzzy neural network, effectively addressing the problem of complex nonlinear modeling of WWTPs and providing accurate predictive output for the control process. Second, a model predictive control (MPC) strategy with performance constraint is designed, in which tracking error is considered as a performance constraint index within the MPC framework, enabling accurate tracking of the DO concentration to the desired set-point. Third, an adaptive constraint optimization algorithm based on the Levenberg-Marquardt (L-M) method is developed. By utilizing the exponential penalty function to reformulate the system constraints, the finite horizon optimization control problem can be directly solved to determine the optimal control sequence. Finally, the stability of the proposed PC-MPC strategy is proven through theoretical analysis and its effectiveness is validated on the benchmark simulation model No. 1.

This work introduces the control concept of variable sampling time finite control-set model predictive control (FCS-MPC). The new control concept is introduced in theory based on a review of the conventional FCS-MPC concepts performed with a constant sampling time. Based on the partitioning of the sampling instant in multiple smaller sampling instants it is possible to optimize, besides the switching states, the switching states turn-on times. Therefore, the proposed variable sampling time FCS-MPC sets both: the switching state and the related turn-on times. To utilize the available calculation power for longer sampling instants an adaptation of the control- and prediction horizon to the sampling time is proposed. The theoretical control concepts are applied to the control of a grid connected two-level voltage-source converter where a simple L-type line-filter is used to demonstrate the control performance of the variable sampling time FCS-MPC algorithms in the laboratory environment.

In this paper, a model predictive control (MPC) framework for building energy system setpoint optimization is developed and tested. The performance of the MPC framework is presented in comparison to a baseline case, where a fixed setpoint schedule is used. To simulate the MPC procedure, an EnergyPlus building model is used to represent the actual building that the optimal setpoints are applied to, and a Gaussian process (GP) regression meta-model is used in the MPC procedure that generates the optimal setpoints. The performance outputs that are used for evaluation are total heating, ventilation and air conditioning (HVAC) energy usage and the Fanger predicted mean vote (PMV) thermal comfort measure. The inputs for the GP regression meta-models are selected to be representative of data points that could be obtained by modern supervisory control and data acquisition (SCADA) systems to support data-driven building models. The supervisory MPC framework is capable of reducing the total energy usage with minor adjustments in thermal comfort.

The parallel operation of converters is increasingly being adopted to meet higher power demands and improve reliability. However, open-circuit faults in semiconductor switches must be located quickly to prevent unnecessary derating and costly offshore interventions. In parallel topologies, zero-sequence circulating current (ZSCC) distorts phase current residuals at fault inception, making conventional residual-based fault diagnosis methods unreliable. This paper proposes a fast fault diagnosis method that integrates phase current residual analysis with real-time ZSCC polarity. The method monitors which residuals exceed a fixed threshold and compares the polarity of the dominant residual with that of the ZSCC. In this way, it can distinguish and locate single-switch, same-converter dual-switch, and cross-converter dual-switch faults. A three-stage timing window mechanism captures residual and ZSCC signatures within only a few sampling periods, enabling precise fault location without additional sensors. Experimental and simulation results verify the accuracy and robustness of the proposed method, demonstrating its potential for practical deployment in offshore wind turbines.