Conventional Model Predictive Control Research Articles

Robust Model Predictive Control for AFE-Inverter Drives With Common Mode Voltage Elimination

This paper proposes a novel and robust version of Model Predictive Control scheme for AC drives based on Voltage Source Inverter (VSI) with Active Front End (AFE). The main feature of the proposed MPC is elimination of Common Mode Voltage (CMV) without imposing a penalty on the corresponding term in the cost function, but rather by a smart utilisation of the restricted set of switching states in a computationally efficient algorithm. Furthermore, the paper proposes to split the conventional MPC scheme into separate Control and Modulation stages, and to enhance the Control stage by integral action, and the Modulation stage - by a Feedback Quantizer.

Modulated Model Predictive Torque Control for Interior Permanent Magnet Synchronous Machines

Thanks to the advancements in the processor industry, the popularity in the industrial applications of Finite control set model predictive control (FCS-MPC) is increasing. FCS-MPC has several advantages, such as high closed-loop bandwidth, the inclusion of the control constraints, and nonlinearities. However, the control signals are directly produced by the predictive controller since no modulator is used. Hence, the system has non-fixed switching frequency, and the maximum achievable switching frequency is limited by the half of the sampling frequency. However, the control goals may suffer from the undesired ripples in case of a noticeable low switching frequency. To eliminate these ripples the sampling period of the system can be reduced. But this increases the computational burden on the processor. To overcome the unwanted oscillations in the control variables and decrease the computational burden on the processor, a modulated model predictive control (M2PC) strategy is proposed in this paper. The M2PC combines the space vector pulse width modulator (SVPWM) and FCS-MPC. Torque of the interior permanent magnet synchronous motor (IPMSM) is controlled with M2PC method. The motor is controlled in a constant torque region with the combination of the M2PC method and maximum torque per ampere (MTPA) control strategy. The comparative results of the conventional MPC method and M2PC method are reported in the paper and the superiority of the M2PC strategy is validated by simulation works. The results demonstrate that the M2PC method significantly reduces total harmonic distortion (THD) in stator currents. Based on the results, the M2PC method provides a better control performance for IPMSMs with significantly reduced torque ripples.

An improved long‐horizon model predictive control for DFIG in WECS with variable sampling‐time

This paper presents an improved finite-control-set model predictive control (MPC) strategy for the doubly fed induction generator (DFIG). The MPC-based controller's main drawback is high computational burden, which requires a long time to perform online optimization. In this regard, real-time implementation becomes impossible in some cases. Thereupon, this paper suggests a variable sampling time in the prediction process to achieve a balance between the prediction horizon and computational burden. In other words, compared to the conventional MPC-based techniques, longer horizon is predicted with fewer samples. In this method, there are two sampling times including a fixed sampling time for the system discretization structure, and a variable sampling time used in the prediction process. The proposed technique also utilizes a comparative incremental algorithm to prevent checking all possible control inputs over the objective function in order to reduce the computational time further. In addition, a penalty term, which includes the number of switch changes, is added to the cost function to reduce the switching frequency. Both the simulation and experimental results show that the proposed MPC can significantly improve the DFIG control system's performance by increasing the accuracy of tracking response and reducing the computational time. The experimental test is carried out by the hardware in loop system by using TMS320F28338 and PC-Lab card. The computational delay caused by the digital signal processor and other influential factors are compensated by using the prediction plus prediction approach. The results confirm that the proposed controller has a lower overshoot and power ripple compared to the vector controller and direct torque controller. It also reduces the computational time up to 40% and reduces the average switching frequency by 12% in comparison to the conventional MPC.

Economic Model Predictive Control for Post-Combustion CO2 Capture System Based on MEA

For the post-combustion CO2 capture (PCC) system, the time variability of the economic performance is key to the production process of such an actual industrial process. However, the performance index used by the conventional model predictive control (MPC) does not reflect the economy of the production process, so the economic cost function is used instead of the traditional performance index to measure the economy of the production process. In this paper, a complete dynamic model of the PCC system is constructed in Aspen Plus Dynamics. The effectiveness of the model is verified by dynamic testing; subspace identification is carried out using experimental data, a state-space equation between flue gas flow and lean solvent flow; the CO2 capture rate is obtained; and dynamic models and control algorithm models of accused objects are established in Matlab/Simulink. Under the background of the environmental protection policy, an economic model predictive control (EMPC) strategy is proposed to manipulate the PCC system through seeking the optimal function of the economic performance, and the system is guaranteed to operate under the economic optimal and excellent quality of the MPC control strategy. The simulation results verify the effectiveness of the proposed method.

Model Predictive Torque Control of Five-Phase PMSM by Using Double Virtual Voltage Vectors Based on Geometric Principle

The conventional model predictive torque control (MPTC) suffers from the heavy computation burden and the uncertainty of the weighting factor. This article proposed a new MPTC scheme for a five-phase permanent magnet synchronous motor by using double virtual voltage vectors (VVVs), in which no weighting factor is required. The VVVs that can restrain the harmonics work as the candidate vectors. The deadbeat principle is adopted to predict the reference voltage, which can be used to select the optimal VVV directly in avoid of testing all candidate VVVs, and hence, the computation burden is reduced. The geometric principle is applied to select the second voltage vector and calculate the vector duration. The second vector can be a VVV or null vector. Therefore, double VVVs or a VVV and a null vector can be synthesized for the next period. It should be noted that no weighting factor is required because of the introduced geometric principle. Due to the selection of the second voltage vector, the torque and flux ripples are reduced and the steady-state performance is improved. The proposed MPTC has the same fast dynamic response with the conventional MPTC. The experiments are presented to verify the proposed MPTC scheme.

Indirect Shared Control for Cooperative Driving Between Driver and Automation in Steer-by-Wire Vehicles

It is widely acknowledged that drivers should remain in the control loop before automated vehicles completely meet real-world operational conditions. This paper presents an “indirect shared control” framework for steer-by-wire vehicles, which allows the control authority to be continuously shared between the driver and automation through an weighted-input-summation method. A “best-response” driver steering model based on model predictive control (MPC) for indirect shared control is proposed. Unlike any conventional driver model for manual driving, this model assumes that drivers can learn and incorporate the controller strategy into their internal model for predictive path following. The analytic solution to the driver model is provided to enable off-line simulations. A driving-simulator experiment was conducted to demonstrate the advantages of the indirect shared control system in a highway lane-keeping task. The result showed that the proposed indirect shared control method was effective to improve the subjects’ lane-keeping performance and reduce steering control effort. The proposed driver steering model was also validated by the experiment data, which produced a smaller prediction error than the conventional MPC driver model.

Three-Dimension Space Vector Based Finite Control Set Method for OW-PMSM With Zero-Sequence Current Suppression and Switching Frequency Reduction

The zero-sequence current (ZSC) of open-end winding permanent magnet synchronous motor (OW-PMSM) with common dc bus mainly comes from zero-sequence voltage (ZSV) and third harmonic back electromotive force, so the ZSV levels of voltage vectors in model predictive control (MPC) need to be considered to suppress the ZSC. Different from the conventional MPC methods for OW-PMSM, which usually select the optimal voltage vector under αβ0 coordinate system, this article proposes a three-dimension space vector (TDSV) based MPC method under ABC three-phase coordinate system for OW-PMSM with ZSC suppression and switching frequency optimization strategy. Benefiting from the three-phase voltages decoupling feature of OW-PMSM, TDSVs in ABC coordinate system are regularly distributed in a space cube. A simple optimal TDSV selection method is proposed based on subcubes division. TDSV contains the predicted α-,β-,0 -axis voltages information, so ZSC suppression is achieved automatically. Furthermore, utilizing the switch states redundancy feature of dual-inverter system, a cost function is designed to optimize the switch states change and reduce the switching frequency without affecting the optimal TDSV selection. The performance and switching frequency reduction effect of the proposed method are verified and compared with the existing methods through experiments.

A weighted voltage model predictive control method for a virtual synchronous generator with enhanced parameter robustness

To address the problem of insufficient system inertia and improve the power quality of grid-connected inverters, and to enhance the stability of the power system, a method to control a virtual synchronous generator (VSG) output voltage based on model predictive control (MPC) is proposed. Parameters of the inductors, capacitors and other components of the VSG can vary as the temperature and current changes. Consequently the VSG output voltage and power control accuracy using the conventional MPC method may be reduced. In this paper, to improve the parameter robustness of the MPC method, a new weighted predictive capacitor voltage control method is proposed. Through detailed theoretical analysis, the principle of the proposed method to reduce the influence of parameter errors on voltage tracking accuracy is analyzed. Finally, the effectiveness and feasibility of the proposed method are verified by experimental tests using the Typhoon control hardware-in-the-loop experimental platform.

Performance Analysis of Multi-Level Inverter-Fed Position Sensorless PMSM Drive Using Modified MPTC

In this article, the performance of multilevel inverter-fed position sensorless permanent magnet synchronous motor (PMSM) drive is analyzed using modified model predictive torque control (MPTC) method using two-level and three-level inverters, respectively. The MPTC method helps to control electromagnetic torque directly, which is not possible in model predictive current control, whereas the conventional model predictive torque and flux control method requires weighing factors that are similar to PI controllers used in field-oriented control or direct torque and flux control method. To simplify the control structure, most of the researchers are looking forward to weighing factorless predictive control strategies. The multilevel inverters required complex control algorithms such as space vector modulation. However, the model predictive control methods made it simple by using an embedded control structure. Furthermore, position sensorless control strategies are much needed because speed sensors are much prone to failures. Therefore, in this article, a model reference adaptive system (MRAS)–based speed sensorless approach is used to eliminate the requirement of a position sensor. The proposed multilevel inverter-fed position sensorless PMSM drive using modified MPTC is simulated in MATLAB/SIMULINK environment under different operating conditions such as under no-load, load, and change in speed. The simulation results are validated with the experimental results.

Equal weighted cost function based weighting factor tuning method for model predictive control in power converters

An equal-weighted cost function-based weighting factor tuning method for model predictive control (MPC) in power converters is presented here. In conventional MPC methods developed for power converters, the use of weighting factors is a necessity to determine the importance of sub-terms in multi-objective cost functions. The proposed method is based on equalizing the importance of cost terms by using normalized values rather than the actual values. As a consequence of normalized values, all sub-terms are transformed into a similar range independent from their actual values. Hence, the weights of cost terms are equalized, leading to the elimination of weighting factor tuning necessity. The proposed equal-weighted cost function not only offers simplicity in the MPC design but also guarantees the desired controller performance with the dynamically weighted sub-terms. The proposed control method is validated on a grid-connected single-phase three-level T-type inverter. Experimental results are carried out to demonstrate the effectiveness of the proposed method under steady-state and dynamic conditions. Moreover, the comparative results with the conventional MPC are presented. The results reveal that the proposed approach exhibits excellent performance against the variations in the operating point of the inverter.

Comparative assessment of alternative MPC strategies using real meteorological data and their enhancement for optimal utilization of flexibility-resources in buildings

Model predictive control (MPC) method shows superiority in enhancing system performance. But its actual performance in operation is hampered by forecast uncertainties. Different methods have been proposed to mitigate the impacts of these uncertainties, such as shrinking horizon MPC (SHMPC) and stochastic MPC (SMPC). However, the year-round performance of these methods has rarely been investigated, and little is known about their relative performance, particularly in utilization of building flexibility-resources. In this study, the year-round performance of conventional MPC (CMPC), SHMPC and SMPC strategies when used to optimize utilization of flexibility-resources in buildings with PV power generation and battery storage systems are compared using real-time optimal control (RTC) as the benchmark, and their enhancement is ultimately proposed. Forecast uncertainties are quantified based on real meteorological data. Different optimization horizon start times and shrinking intervals are tested. Results show that there is no one control strategy which is superior to the other strategies under all operating conditions. The SMPC strategy has a higher probability of achieving daily cost saving than CMPC and SHMPC strategies, but still has higher daily costs than the RTC strategy in the winter. Hybrid control strategies with proper control schemes implemented under different operating conditions are therefore recommended.

Robust output feedback model predictive control of time‐delayed systems using interval observers

AbstractThis work addresses the problem of robust output feedback model predictive control (MPC) of constrained, discrete‐time, time‐delayed linear systems subject to (bounded) additive disturbances. The proposed predictive controller incorporates an interval observer, that exploits the available measurement to update the set‐membership of the system states, and an interval predictor, used in the prediction step of the MPC. The proposed algorithm is similar to conventional MPC for linear systems, offering guarantees on recursive feasibility and robust constraint satisfaction, with low computational complexity. Furthermore, the delay is explicitly taken into account in the design, and the obtained scheme easily encompasses delays on the state, on the measurement or the control input.

A Low-Complexity Three-Vector-Based Model Predictive Torque Control for SPMSM

In conventional model predictive torque control (MPTC), only one voltage vector (VV) is applied during a whole control period, thus causing a large torque ripple. To improve the steady-state performance, some two-vector-based control schemes have been proposed. However, the selection of the optimal VV pair is complex as well as has a large computational burden, and the improvement of performance is still limited by the direction and amplitude of the output VV. This article proposes a low-complexity three-vector-based MPTC for SPMSM drives, which can precisely determine the appropriate active voltage vectors (AVVs) with the predicted torque error. Then, a modified switching table is developed to directly select the AVVs, thus greatly reducing the complexity and computational burden of the algorithm. To obtain a better steady-state performance, a duty cycle calculation method based on torque and stator flux difference parameters is newly proposed to achieve the deadbeat control of torque and stator flux. And then, the experimental comparisons with the double-vector-based MPTC are conducted. The results show that the proposed MPTC can effectively reduce the steady-state torque ripple while maintaining a good dynamic performance as well as almost a fixed switching frequency for all speed ranges.

Two-Stage Series Model Predictive Torque Control for PMSM Drives

A two-stage series model predictive torque control for permanent magnet synchronous motor drives is proposed in this article. First, the control of torque and flux linkage is converted to double-torque control (i.e., the control of active torque and reactive torque) based on the instantaneous power theory to eliminate the weighting factor in conventional model predictive torque control. Then, the double-torque prediction trajectory and double-torque reference are extrapolated to two control periods. The concept of two-stage series predictive control is proposed in this article, in which the cost functions of two prediction instants are designed to form series control structure. In the proposed method, the first-stage prediction is used to select candidate voltage vectors and the second-stage prediction is used to evaluate the optimal voltage vector. Since the cost function of two control periods in the future is considered when selecting the voltage vectors, the possibility that the same voltage vector is applied to adjacent control periods increases, thereby reducing the switching frequency of the system. Finally, the effectiveness of the proposed method is verified by experiments.

Benchmarking of data predictive control in a real-life apartment during heating season

Model Predictive Control is an energy efficient climate control strategy in buildings. However, the effort associated with physics-based modelling seems to prevent widespread application in residential buildings. Applying machine-learning algorithms on historical data promises efficient generation of predictive models for control. In a recent experimental study, Data Predictive Control based on random forests and linear models outperformed a baseline controller during cooling season. In this paper, the approach is benchmarked against hysteresis control and conventional Model Predictive Control based on an RC-network model during heating season. Data Predictive Control shows promising results in terms of energy consumption and thermal comfort.

Model Predictive Current Control With Low Complexity for Single-Phase Four-Level Hybrid-Clamped Converters

This article presents a simplified model predictive control (MPC) method for single-phase four-level hybrid-clamped converters (1P-4L-HCCs), achieving high-quality current control and strong capacitor voltage balancing. Different from a conventional MPC, the proposed method aims to dramatically reduce switching states (control voltage vectors) in the cost function optimization of MPC. It utilizes only 16 switching states, compared to 64 in conventional methods, resulting in reduced execution time and significantly increased control bandwidth. Steady-state and dynamic performances are evaluated by simulated and experimental test, and the main results are provided to verify the effectiveness of the proposed MPC method.

Disturbance Observer and Tube-Based Model Predictive Controlled Electric Vehicles for Frequency Regulation of an Isolated Power Grid

This paper presents a frequency regulation approach applied to a renewable penetrated isolated power grid in the presence of a large number of Electric Vehicles (EVs). A disturbance observer is initially designed for a reduced-order model of the system to produce a supplementary frequency control signal for the aggregated EVs. The reduced order model is obtained by combining the variations in the load, wind, photovoltaic systems, and aggregated EVs in order to yield a lumped disturbance to be estimated by the proposed disturbance observer. A tube-based Model Predictive Control (MPC) is then proposed to provide efficient control signals for improving the response of the aggregated EVs. The control signals are produced so as to obtain the least value of frequency deviation error with a minimum control effort while taking a variety of operational and physical constraints into consideration. The proposed control strategy offers the capability to cope with the uncertainties caused by the external disturbances by coordinating the charging and discharging of aggregated EVs. The impact of the delay in communication links is also investigated by conducting a stability analysis to obtain the delay margin. Simulations are conducted on an isolated system to illustrate the effectiveness of the designed observer, and also to examine the advantage of the presented tube-based MPC over a conventional MPC, a fuzzy proportional-integral control, and a linear quadratic regulator control.

Integration of genetic algorithm tuned adaptive fading memory Kalman filter with model predictive controller for active fault-tolerant control of cement kiln under sensor faults with inaccurate noise covariance

Reliable control of cement kiln under sensor faults is vital to ensure cement clinker quality. To accomplish this objective, an active fault-tolerant control (FTC) scheme is proposed for burning zone temperature control of cement kiln even under inaccurate measurements. FTC is achieved by integrating adaptive fading memory Kalman filter (AFMKF) in an already existing model predictive controller (MPC) framework. Residue from the Kalman filter is used to detect faults, and the model prediction replaces sensor measurements. Hence, there is a need for an accurate model and its associated noise covariance. Model accuracy is ensured by the online estimation of sensor noise covariance using the proposed AFMKF. Hyperparameters governing the AFMKF are tuned using the multi-objective genetic algorithm (GA), aiming to minimise the modelling and tracking error. Sensor measurement failure, bias, and stuck are considered for performance validation of the proposed FTC. Percentage recovery from fault is evaluated in terms of Integral square error (ISE) and energy utilisation compared with a conventional MPC. Experimental results illustrate the significant improvement in model accuracy and fault-tolerant behaviour using the proposed GA tuned AFMKF.

A Novel Control Approach to Hybrid Multilevel Inverter for High-Power Applications

This paper proposes a hybrid control scheme for a newly devised hybrid multilevel inverter (HMLI) topology. The circuit configuration of HMLI is comprised of a cascaded converter module (CCM), connected in series with an H-bridge converter. Initially, a finite set model predictive control (FS-MPC) is adopted as a control scheme, and theoretical analysis is carried out in MATLAB/Simulink. Later, in the real-time implementation of the HMLI topology, a hybrid control scheme which is a variant of the FS-MPC method has been proposed. The proposed control method is computationally efficient and therefore has been employed to the HMLI topology to mitigate the high-frequency switching limitation of the conventional MPC. Moreover, a comparative analysis is carried to illustrate the advantages of the proposed work that includes low switching losses, higher efficiency, and improved total harmonic distortion (THD) in output current. The inverter topology and stability of the proposed control method have been validated through simulation results in MATLAB/Simulink environment. Experimental results via low-voltage laboratory prototype have been added and compared to realize the study in practice.

Comparative Assessment of Alternative Mpc Strategies Using Real Meteorological Data and Their Enhancement for Optimal Utilization of Flexibility-Resources in Buildings

Model predictive control (MPC) method shows superiority in enhancing system performance. But its actual performance in operation is hampered by forecast uncertainties. Different methods have been proposed to mitigate the impacts of these uncertainties, such as shrinking horizon MPC (SHMPC) and stochastic MPC (SMPC). However, the year-round performance of these methods has rarely been investigated, and there is little knowledge on their relative performance, particularly in optimal utilization of flexibility-resources in buildings. In this study, the year-round performance of conventional MPC (CMPC), SHMPC and SMPC strategies when used to optimize utilization of flexibility-resources in buildings are compared using real-time optimal control (RTC) as the benchmark, and their enhancement is ultimately proposed. Forecast uncertainties are quantified based on real meteorological data. Different optimization horizon start times and shrinking intervals are tested. Results show that there is no one control strategy which is superior to the other strategies under all operating conditions. The SMPC strategy has a higher probability of achieving daily cost saving than CMPC and SHMPC strategies, but still has higher daily costs than the RTC strategy in the winter. Hybrid control strategies with proper control schemes implemented under different operating conditions are therefore recommended.