Effects and compensations of computational delay in finite set-model predictive control in renewable energy system

Techno-enviro-socio-economic design and finite set model predictive current control of a grid-connected large-scale hybrid solar/wind energy system: A case study of Sokhna Industrial Zone, Egypt

This paper presents an enhanced model predictive voltage and current control of a three phase two level four leg inverter. Commercially trusted PI control strategy experience the arbitration of steady state and transient performance. Also the PI controller gain parameters are not optimal for the whole control operation space. So as to upgrade the control resilience of the converter, a new Model predictive control(MPC) is proposed. MPC is well acknowledged for its superior dynamic performance, intuitiveness, multi objective and multi constrain treatment. However, the basic FCS-MPC (Finite Control Set MPC) develops a variable switching frequency which leads to difficulties in filter design and poor THD performance. So a fixed switching frequency algorithm is implemented by including a virtual modulator based on a predefined switching sequence. Simultaneously, the switch redundancy has been enforced to eliminate the weight factor coupled between load voltage and load current. Simulation results are presented and compared to highlight the performance of the proposed MPC.

This article proposes a unique control strategy for a deadbeat multiple vector finite-set model predictive current control with an embedded integral action (MV-FMPC) for permanent magnet synchronous motor drives. Torque ripple and phase current distortions in permanent magnet synchronous motor (PMSM) drives are minimized with the proposed controller by adopting the multiple vector approach to the finite-set model predictive control. The controller uses a hexagonal co-ordinate system to simplify the location and identification of the virtual vectors created, thereby eliminating the use of large look-up tables and reducing computational burden. When used with the proposed deadbeat prediction model, the overall steady-state performance, system robustness, and quality of disturbance rejection are improved compared to the state-of-the-art finite-set model predictive current control (FS-MPC) methods with pulsewidth modulation. The improvements are due to the modified deadbeat prediction model with integral action, the algorithm used for multiple virtual voltage identification and the retention of the cost function in the proposed method. The proposed deadbeat MV-FMPC method and its improvements over the conventional FS-MPC have been verified through simulation and experiments with an interior-type permanent magnet synchronous machine.

Model predictive control (MPC) is one of the control strategies used recently to control power converters because of its advantages of handling nonlinear functions, fast dynamic response, and eliminating the modulators and PI regulators. However, it has one disadvantage: it requires high processing power to operate correctly due to the high computational burden on the controller. This paper presents a new method of controlling the performance of grid-tied multilevel inverter functioned as STATCOM based on finite control set model predictive current control (FCS-MPCC). This control strategy is employed to control a five-level asymmetric stacked multilevel inverter (ASMLI) and coupled with the network through LCL-filter. Multilevel inverters suffer from capacitors voltages unbalance problem, and LCL-filters require an efficient damping method to function appropriately; these two challenges are included in the cost function of the FCS-MPCC. The FCS-MPCC strategy is constructed in two methods. The first method is the traditional method, which requires 729 calculations per sampling instant. An effective calculation method has been proposed as a second method, as it requires 27 calculations at each sampling instant. These two methods are compared to ensure that the proposed method does not affect STATCOM performance. MATLAB / Simulink is used to simulate these two models. The result demonstrates that the proposed method reduces implementation time five times than the other method with no loss of reference traceability, balanced capacitors voltage, and damped filter performance.

Disturbance rejection strategies are very useful for the robustness improvement of the predictive control method. But they can only be used in the modulated-based predictive control methods such as continuous set model predictive control (CS-MPC) and deadbeat control. This paper presents a robust current prediction model based on total disturbance observer (TDO), which is applicable in the finite set model predictive current control (FS-MPCC). In the proposed method, the disturbance is directly used as a part of the prediction model instead of the disturbance rejection loop. So, the proposed method has two advantages over the disturbance rejection-based CS-MPC schemes. The first advantage is no need for a controller, which is an essential part of the disturbance rejection-based CS-MPC. Therefore, the proposed method is simpler and has fewer control parameters. The second feature is that the proposed model is in the stationary frame. In this way, the frame transformation is avoided in the prediction model. Moreover, to guarantee zero steady-state error in the current prediction model, this paper proposes a complete designing process for TDO based on the convergence analysis. The performance of the proposed control system is evaluated through simulations and experimental tests.

This paper presents the modeling and predictive control of permanent magnet toroidal motor (PMTM). Based on the spatial structure characteristics and working principle of toroidal motor, the time-varying inductance parameters are derived. The nonlinear mathematical model of toroidal motor is established in rotating coordinate system. The output fluctuation of the toroidal motor is analyzed. To improve the output performance of toroidal motor, the full predictive cascade speed and current control (F-MPC) strategy for toroidal motor is proposed. For F-MPC of toroidal motor, the finite control set model predictive current control (FCS-MPCC) strategy with delay compensation is adopted in current inner loop, and the deadbeat model predictive speed control (DB-MPSC) strategy is adopted in speed outer loop. Meanwhile, to achieve DB-MPSC, the reduced-order Luenberger observer is designed to obtain load torque in real time. F-MPC is the combination of current-speed prediction and Luenberger observer; it can overcome the limitation of excessive parameter adjustment of traditional cascade PI (Proportional Integral) controller. The simulation results show that compared with FOC (Field Oriented Control) and FCS-MPCC, F-MPC can improve the dynamic response and anti-interference performance of toroidal motor, and it can also reduce the fluctuation of output speed and torque significantly.

Due to model predictive control can better deal with the constraints problem of nonlinear systems and improve the dynamic performance of the controlled system, therefore, this technology has attracted much attention in the field of motor drive. This article first introduced the basic principle of model predictive control, continuous control set model predictive control and finite control set model predictive control. Secondly, summarized the research status of generalized predictive control, explicit model predictive control, model predictive current control, model predictive torque control and commonly used improved model predictive control in motor drive systems. Thirdly, prospected the future development trend based on the current research status of model predictive control in motor drive system. Finally, the advantages and disadvantages of the continuous control set model predictive control and the finite control set model predictive control were comprehensively compared, and the ways in which the two algorithms act on the motor drive system were summarized.

Finite control set model predictive current control (FCS-MPCC) has attracted the attention of a large number of researchers due to its intuitive idea, fast dynamic response and control objective flexibility. However, FCS-MPCC has a high steady-state current ripple. Three-vector-based model predictive current control (TV-MPCC) is an effective method to improve steady-state performance. However, it results in high motor parameter dependency. In this paper, the effect of motor parameters mismatch in FCS-MPCC has been analyzed. The control performance of FCS-MPCC is impacted when the motor parameters are mismatched. Then based on the disturbance estimation technique, a disturbance feedforward compensation based generalized proportional integral observer is used to reduce the disturbances and improves the robustness performance of the system. Experimental results show that the proposed method effectively improve the robustness of the system.

The ultralocal model of a plant expresses the behavior of the output variable in relation to the input without requiring any specific information about the plant. Replacing the conventional model of the system with an ultralocal model can effectively improve the robustness of the model predictive control (MPC) because all types of uncertainties that correlate with the system modeling are removed in the ultralocal model. This article proposes an integral sliding mode observer (ISMO)-based ultralocal model for the finite-set model predictive current control (FS-MPCC) of an induction motor (IM). The ISMO is constructed based on the Lyapunov theory to guarantee the stability of the proposed control method. Moreover, an analytical comparison is made between the proposed ISMO and the conventional extended state observer (ESO) in the application of the ultralocal model. This comparison shows that the stability of ESO depends on the unknown term of the ultralocal model. In contrast, the stability of the proposed ISMO does not rely on any terms of the ultralocal model. Experimental tests confirm that the proposed FS-MPCC method is practically applicable to the IM drive system. Also, it improves the robustness of the predictive method against model uncertainty.

Model predictive control (MPC) is a new trend and capitative solution for controlling power converters. A finite control set model predictive current control (FCS-MPCC) technique is proposed to control output currents and voltages of flying capacitors for a five-level nested neutral point-clamped (5L-NNPC) inverter in Traction system. The 5L-NNPC converter works over the extensive range of voltages with interesting qualities such as less components in contrast to other conventional five-level converters. Converter model is driven in discrete-time concerning the switching stats of converter. Output currents and voltages of flying capacitor are predicted using each possible switching state for every sample. Cost function compares these predicted variables with their references and choose the switching state for next sample that corresponds to the minimum cost. Suggested FCS-MPCC technique is simulated and analyzed for NNPC converter using MATLAB.

This paper presents a heuristic Finite-Set Model Predictive Control (FS-MPC) strategy for current control of induction machines. FS-MPC methods are very intuitive and easy to understand; furthermore, they in general offer the possibility to control any nonlinear system with arbitrary user-defined terms in the cost function and do not require a modulator. One of their most important drawbacks is the necessary calculation effort which rises exponentially with the prediction horizon. Especially for multilevel inverters this can be a crucial factor for a successful real-time implementation. The proposed heuristic strategy for reducing the calculation effort can also be extended to multilevel inverters. The algorithm is verified experimentally for FS-MPC current control of an induction machine with three prediction steps.

ABSTRACTAs a result of the increasing use of electric vehicles, ensuring high‐performance speed and torque control of brushless direct current (BLDC) motors has become of great importance for energy efficiency. In order to prevent the torque ripple of the finite control set model predictive current control (FCS‐MPCC), commutation moments are detected by Hall effect sensors in conventional methods. However, this method cannot exhibit a long‐life structure because of physical strain damaging the sensors and electrical connections. In this study, commutation moments and durations are captured and determined with a new approach. Commutation moments are captured with zero crossing detectors and commutation durations are determined by using the position information obtained from the encoder. Moreover, three‐phase back electromotive forces (EMFs) of the BLDC motor applied to FCS‐MPCC to predict the stator phase currents are estimated with a novel adaptive extended Kalman filter (AEKF) which has the estimation capability without any speed sensor. Furthermore, another improvement is implemented in the calculation of the cost function of FCS‐MPCC by taking into account the difference between the predicted and the reference torque of the BLDC motor different from the conventional MPCC methods. The proposed drive system is tested under different scenarios at various speeds under load torque, stator resistance, and leakage inductance variations in simulation. It is proven by simulation results that phase commutations can be achieved stably with the proposed phase commutation determination method. In addition, the simulation results show that the proposed novel AEKF estimator and the FCS‐MPCC in which the cost function is calculated by regarding not only the current error but also the moment error have impressive prediction and control performance, respectively.

Model predictive control, especially model predictive current control, has received a great deal of attention in the motor drive field in recent years, due to its ability to render fast dynamic response and to handle multiple variables, nonlinearities, and system constraints in an intuitive way. However, the conventional single-vector-based model predictive hysteresis current control brings about some problems, such as high sampling frequency and poor steady-state control performance. In this paper, a novel double-vector-based model predictive hysteresis current control (DV-MPHCC), which utilizes an arbitrary vector and a zero vector to form a voltage vector combination, is proposed. To verify the improved performance over a short predictive horizon, a series of comparisons are conducted between the two control algorithms by simulation and experiment. Both simulation and experiment results validate that the DV-MPHCC proposed has advanced steady-state control performance, and a lower sampling frequency.

This paper proposes a current model predictive control strategy for the permanent magnet synchronous motor (PMSM) based on a novel sliding mode observer to reduce the cost of PMSM and ensure good tracking performance. A super twisting sliding mode observer (STSMO) is designed to address the issues of high-frequency chattering and noise sensitivity caused by the large positive gain of traditional SMO. The discontinuous effect of the traditional SMO switching function is introduced into the derivative of the control rate, and a smooth estimate of the back electromotive force (EMF) is obtained through integration. Replace the sign function with a sigmoid function with smooth continuity to further reduce the chattering effect. To enhance the dynamic performance of the PMSM current loop, a finite control set model predictive control (FCS-MPC) strategy is employed in place of the conventional PI controller. Within each sampling period, all possible switching states are evaluated, and the optimal one is selected and directly applied to the inverter. Additionally, a dual-vector model predictive current control (DVMPCC) method is adopted to reduce current ripple. This approach synthesizes a voltage vector with arbitrary magnitude and direction by combining two voltage vectors within each sampling period. Numerical results demonstrate that the proposed sensorless PMSM predictive current control method achieves high accuracy in speed estimation and excellent dynamic response performance.

The paper presents a new implementation of direct flux and current vector control of an induction motor drive using the techniques of model predictive control. The advantages offered by predictive control are used to enhance the dynamics of direct flux vector control. To minimize the problems of variable switching frequency inherent to finite control set predictive control, an alternative approach using pulse width modulation is studied for command execution as occurs in the so-called modulated model predictive control. A comparison between finite control set and modulated model predictive control is presented and the results are also compared with the control implementation through traditional proportional-integral regulators to highlight the advantages and drawbacks of predictive control based strategies. Apart from a greater harmonic content in stator currents, the predictive control can offers control dynamics comparable with proportional-integral control while maintaining immunity against machine parameter variations and excluding the need for controller tuning.