劉 康志

In this paper, the speed tracking problem of the interior permanent magnet synchronous motor (IPMSM) of an electric vehicle is studied. A cascade speed control strategy based on active disturbance rejection control (ADRC) and a current control strategy based on improved duty cycle finite control set model predictive control (FCSMPC) are proposed, both of which can reduce torque ripple and current ripple as well as the computational burden. First of all, in the linearization process, some nonlinear terms are added into the control signal for voltage compensation, which can reduce the order of the prediction model. Then, the dq-axis currents are selected by maximum torque per ampere (MTPA). Six virtual vectors are employed to FCSMPC, and a novel way to calculate the duty cycle is adopted. Finally, the simulation results show the validity and superiority of the proposed method.

The control method based on the equivalent-input-disturbance (EID) estimator and the Luenberger state observer has received much attention in recent years. However, the property of EID-based control systems is still not well investigated. A number of design procedures were proposed but lacked sufficient theoretic justification. In this article, the two-degree-of-freedom nature of an EID-based control system is revealed. Specifically, the reference-tracking is determined by an outer loop, while the disturbance rejection and robustness is determined by an inner loop consisting of the EID estimator and state observer. The bandwidth constraints on the inner loop are analyzed for plants having zeros or poles in the open right-half plane by using the Bode and the Poisson integral formulas. These analyses provide a theoretic justification for conventional design procedures. Further, a coordinated design algorithm is provided for the EID-based uncertain control systems. In addition, a comprehensive comparison of the EID-based and the disturbance observer (DOB) based control systems is conducted in the system design aspect. Lastly, comparative studies of the EID-based and DOB-based control methods are given for various types of plants to validate the developed design algorithm.

In this paper, a time series model based on hybrid-kernel least-squares support vector machine (HKLSSVM) with three processes of decomposition, classification, and reconstruction is proposed to predict short-term wind power. Firstly, on the basis of the maximal wavelet decomposition (MWD) and fuzzy C-means algorithm, a decomposition method decomposes wind power time series and classifies the decomposition time series components into three classes according to amplitude- frequency characteristics. Then, time series models on the basis of least-squares support vector machine (LSSVM) with three different kernels are established for these three classes. Non-dominated sorting genetic algorithm II optimizes the parameters of each forecasting model. Finally, outputs of forecasting models are reconstructed to obtain the forecasting power. The proposed model is compared with the empirical-mode-decomposition least-squares support vector machine (EMD-LSSVM) model and wavelet-decomposition least-squares support vector machine (WDLSSVM) model. The results of the comparison show that proposed model performs better than these benchmark models.(c) 2020 ISA. Published by Elsevier Ltd. All rights reserved.

This paper proposes a novel optimization method for energy storage systems (ESSs) to smooth wind farm output to satisfy the technical requirements and reduce the rated power (rated energy capacity) and charge/discharge loss of the ESS. The state of charge of the ESS operated by the proposed method can be regulated, guaranteeing some constraints. The proposed method has a very simple structure, does not require the wind farm output forecast and numerical optimization, such as particle swarm optimization. Therefore, a high-grade functional computation device is not needed. The effectiveness of the proposed method is verified by comparative analysis with conventional approaches through simulations.

The paper presents a multiaspect analysis of multivalues and the broadband nature of system oscillation. By analyzing the ambient signal caused by random small disturbances during the normal operation of interconnected power grids, many system operation characteristics can be obtained. The traditional signal processing method cannot extract the information from ambient signals effectively. Aiming at the problem of broadband oscillation mode superposition and the difficulty of extracting information from ambient signals, an iterative adaptive variational mode decomposition (IA-VMD) method is proposed based on frequency domain analysis and signal energy. Additionally, the IA-VMD method, combined with a bandpass filter and the Prony algorithm, is used to realize the modal identification of broadband oscillation and ambient signals. Simulation experiments show that the IA-VMD method has good adaptability, antinoise characteristics, and a certain significant engineering application value as well.

This paper proposes a novel and easily implementable over-current suppression control for voltage-controlled virtual synchronous generators (VSGs). The VSG is vulnerable to grid voltage disturbances such as voltage dip/recovery because it consists of voltage source inverters (VSIs). In the worst case, the over-current caused by the voltage disturbance damages the VSI. The proposed VSG control can suppress the over-current caused by not only symmetrical but also asymmetrical voltage dip/recovery. The key idea is to use novel virtual grid voltages for the VSG reference output voltage. The proposed VSG control requires no complicated analysis/tuning for the implementation and no additional equipment such as a fault current limiter. Because of these advantages, the over-current in the VSG can be suppressed without increasing the cost and the tuning effort. The effectiveness of the proposed method is verified through simulations and experiments.

As a high-precision control method for periodic signals, repetitive control has been widely investigated from both theoretical and practical sides. Since a repetitive control system contains two completely different actions: continuous control within each repetition period and discrete learning between periods, exploring the best combination of these two actions may provide us with a potential to achieve higher levels of performance. For this reason, we devised a two-dimensional repetitive control method that features preferential adjustment of control and learning actions. On the other hand, there are aperiodic disturbances in a repetitive control system. It is a challenge to reject such kinds of disturbances. To solve this problem, the equivalent-input-disturbance approach is integrated into a two-dimensional repetitive control system to improve disturbance-rejection performance. This section explains the concepts of two-dimensional repetitive control and an equivalent input disturbance and then shows how these two concepts are combined to construct a high-precision control system.

This article explores a successive convex optimization method for solving a class of nonconvex programming problems subject to bilinear matrix inequality (BMI) constraints. In particular, many control issues can be boiled down to BMI problems, which are typically NP-hard. To get out of the predicament, we propose a more relaxed feasible set to approximate the original one, based on which a local optimization algorithm is developed and its convergence is analyzed. As a case of application, we consider static output-feedback control for uncertain systems with disturbances in restricted frequency intervals. In order to strengthen the disturbance-rejection capability over the given frequency range, we establish sufficient and necessary analysis conditions via the generalized Kalman-Yakubovich-Popov lemma, under which the homogeneous polynomially parameter-dependent technique is adopted to facilitate the design. Finally, several examples are given to demonstrate the efficiency of the results.

In recent years, it is required to extend the operation regions of induction motors (IMs) even at low-speed in regenerating while maintaining a high control performance. Conventional rotor speed estimation methods using an adaptive flux observer are unstable at low-speed in regenerating, particularly near the zero frequency. In this paper, we propose a simultaneous estimation method of the rotor speed and the stator resistance to improve the control performance of the speed-sensorless vector control of IMs. In addition, we prove that the proposed estimation method is stable even at low-speed in regenerating. Furthermore, the effectiveness of the proposed method is demonstrated experimentally.

Rate of penetration (ROP) optimization is crucial for drilling processes due to its vital role in increasing efficiency. This article proposes a new hybrid bat algorithm (HBA) to achieve the maximum ROP accurately. A data-driven ROP model is established by combining the wavelet filtering and optimized support vector regression according to the drilling characteristics. After that, five modifications, namely, iterative local search, stochastic inertia weight, pulse rate and loudness improvement, directional echolocation, and modified local random walk are combined to further improve the global optimization performance of the bat algorithm. Extensive experiment results on the IEEE Congress on Evolutionary Computation 2005 standard benchmark problems show that the proposed algorithm has successful outcomes compared with ten conventional algorithms. Additionally, in the application of the HBA to a real-world drilling process in Shennongjia area, Central China, the ROP has been improved by 34.84%, which is the largest compared with the conventional popular ROP optimization methods.

Spatial 3D formation drillability distribution is crucial for the drilling trajectory planning, collision detection, and drilling optimization in the natural gas and petroleum fields. Conventional geostatistical methods are usually used to establish the 3D formation drillability field model. However, few or no machine learning methods, which have a powerful fitting capability, are used to build that model. This paper proposes a new modeling framework for establishing the spatial 3D formation drillability model. Four methods, one geostatistical (Kriging), one non stochastic (ScatteredInterpolant), and two machine learning methods (random forest and support vector regression) are analyzed in this modeling framework. Well logging data such as acoustic and formation density are introduced as the input parameters. Moreover, the mutual information analysis is introduced to measure the correlations between the 3D coordinates and formation drillability. Finally, comparisons are explored in the 10 fold cross-validation, 3D modeling, and final test experiments using data from Xujiaweizi area, Northeast China. The results indicate that SI has the best 10-fold cross-validation performance while RF achieves the best prediction accuracy in the final test among the four compared methods. The proposed new modeling framework for 3D formation drillability model provides a platform and is applicable to other modeling methods and data.

The distributed consensus control is talked about for general multi-agent networks perturbed by exogenous disturbances including matched and mismatched ones. Incorporating the equivalent-input-disturbance (EID) concept, a new protocol is devised for the rejection of disturbances including both matched and mismatched ones. Specifically, the outputs of the multi-agent network attain asymptotically consensus regardless of the mismatched disturbances; and the states achieve asymptotically consensus in spite of matched ones. Moreover, the developed protocol does not increase extra restraints on system design compared with the original protocol without considering disturbance rejection, and only a few requirements on the disturbances are made. The information shared among the multi-agent network is only the relative output between neighboring agents. Finally, numerical simulation validates the developed method.

Virtual synchronous generators composed of an inverter are affected by an over-current caused by voltage dip and recovery. This letter addresses the current analysis under a symmetrical voltage dip and recovery. (c) 2020 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

Wind power generation provides an attractive method for tackling global environmental issues. However, the power grid cannot accommodate large amount of wind farms (WFs) because the fluctuation of WF output degrades the power quality (frequency and voltage) in the power grid. Technical requirements that are related to WF power fluctuation are issued in many countries in order to introduce the WF without degrading power quality. Therefore, it is essential to smooth the WF output in order to satisfy the technical requirements. This paper proposes an operation methodology for a system that is composed of energy storage systems (ESSs) and WF by kinetic energy (KE) control. Moreover, an optimal KE control is presented. The economical aspect and the advantage of the proposed system are verified through scenario simulations.

This article presents an improved disturbance estimation and rejection method based on equivalent input disturbance (EID) with H-infinity robust performance for a time-varying uncertain system. The analysis and synthesis are conducted in the frequency domain. Different from conventional two-degree-of-freedom (2-DOF) H-infinity control, some related weighting functions are introduced into a new devised structure to separate the rejection of the parameter uncertainty and disturbance from reference tracking. Compared with the existent EID-based approaches designed in the time domain of which the robust performance is not addressed, H-infinity robust performance of the closed-loop system is guaranteed in this design. Treating the difference between the real plant and an ideal plant, namely, the overall effect of the parameter uncertainty and disturbance, as an EID, an EID estimation and compensation strategy is then established to cancel out the effect of the EID in system output. Furthermore, a constant scaling matrix is introduced into system design to reduce the conservativeness of a quadratic stabilization condition. Finally, a comparison with conventional 2-DOF robust H-infinity control illustrates the advantages of the developed method.

Networked control systems (NCSs) are control systems closed through communication network and applied to remote control of robots, tele-surgery and numerous industrial applications. In NCS, data have to coarsely quantised when the communication network is restricted due to the traffic limit. Under the traffic limitation, the design of an appropriate quantiser is highly important in realising a desired control performance. Further, the coordination of quantisation and control input also plays a decisive role. Towards an optimal balance between the traffic capacity of network and the control performance, this study proposes an MPC based approach for quantised state feedback control systems with dynamic quantisers. In the proposed approach, the control input and the quantisation interval of the quantiser are optimised online. The proposed method can explicitly treat input/state constraints, the parameter of quantiser and other physical restriction. The effectiveness of the proposed method is demonstrated by simulations and experiments.

This paper proposes a novel position sensorless positioning servo system using feedback error learning (FEL). Recently, a position sensorless positioning servo system of interior permanent magnet synchronous motors (IPMSMs) has been studied for realizing the high reliability and cost reduction. For the fast response of positioning servo systems, accurate parameters of IPMSMs are required. However, the parameters vary due to temperature and/or aging, which causes degradation of the positioning control performance. Thus, this paper proposes a novel position-sensorless positioning servo system for IPMSMs using the FEL algorithm for motor parameters identification. Besides, this paper employs the comb filters, which is well known as the fast rotor position estimation method. Experiments are demonstrated to validate the effectiveness and feasibility of the proposed method.

This paper proposes a novel adaptive flux observer for the q-axis inductance identification of position sensorless controlled interior permanent magnet synchronous motor (IPMSM). At high load operation, q-axis inductance of IPMSM decreases due to the magnetic saturation phenomenon, which makes the position estimation performance of flux observer that estimates the rotor position using the IPMSM mathematical model degrade. In addition, in the worst case, some motors cause a loss- synchronism phenomenon due to the difference between pre-measured q-axis inductance and actual one. Therefore, compensation of q-axis inductance fluctuation is necessary to realize the stable and high-performance position estimation. Generally, to realize the parameter identification, the adaptive controller is constructed based on Certainty Equivalence (CE) principle. Since this principle does not have the phase margin in the low- speed region, resulting in may be instability. In addition, this principle has the feature that the identification speed is slow. Thus, this paper proposes a novel compensation method to realize the stable and fast q-axis inductance identification by using Dynamic Certainty Equivalence (DyCE) principle. This paper also demonstrates the feasibility of the proposed method through experiments.

With the rapid increase of multi-terminal and multi-infeed VSC-HVDC lines, the dimension and scale of AC/DC system equation and Jacobian matrix (JM) increase sharply, which makes the load flow (LF) analysis very complicated. In this paper, a new hybrid method is proposed to address this issue. This hybrid algorithm is composed of a 1 + root 2 order Newton-Raphson (NR) method and a simplified Newton (SN) method. Specifically, the 1 + root 2 order NR method is executed up to a preset number of iterations first, then the algorithm is switched to the SN until the system accuracy is met. A good compromise between the solution accuracy and computation burden can be achieved once a suitable iteration number is preset. To this end, a unified iterative form suitable for solving the LF of AC/DC systems with VSC-HVDC is derived in detail: a dimension reduction treatment is proposed to solve the problem of increasing dimension of JM; a rule of thumb is provided for the determination of the iteration number of 1 + root 2 order NR. Then, the proposed hybrid method is validated on the modified IEEE-14, 30, 57 and 118 test systems. Comparisons and analyses are made for numerous scenarios, including the basic test of LF, different control modes and control objectives, influence of active and reactive powers on VSC parameters, algorithm performance under various initial states and heavy loading, etc. Compared with Newton and 1 + root 2 order NR methods, the hybrid method shows a significantly shortened convergence time, 30% of 1 + root 2 order NR and 47% of Newton on average.

In many real-world rotary machines, there is an inherent need to attenuate time-varying but position-dependent periodic disturbances. This article proposes an improved repetitive controller, in which a spatial internal model, a spatial low-pass filter, and a frequency alignment are synthesized. The controller is digitally implemented in a uniform time-sampling system. Specifically, a universal internal model of arbitrary periodic signal in spatial domain is derived for the attenuation of the position-dependent disturbances in the rotary machine, and a spatial low-pass filter is designed for the stability of the closed-loop system. Furthermore, a frequency alignment strategy is designed to compensate the bias resulted from the filter. Moreover, the cascaded utilization of the proposed controllers is introduced to improve the performance of a position-feedback servo system. Experimental verification and comparisons validate the feasibility and effectiveness of the proposed method in both single and cascaded configurations.

Rate of penetration (ROP) prediction is crucial for the optimization and control in drilling process due to its vital role in maximizing the drilling efficiency. This paper proposes a novel intelligent model to predict the drilling ROP considering the process characteristics. First, the geological background and the drilling process of the case study are described. Based on the mechanism and frequency spectrum analysis, the strong nonlinearity and different low-frequency and high-frequency data noises between the data variables are detected. After that, the intelligent model is established via three stages. In the first stage, a wavelet filtering method is introduced to reduce these noises in the drilling data. In the next stage, the model inputs are determined by the mutual information method, which significantly decreased the model redundancy. In the last stage, a hybrid bat algorithm is proposed to optimize the hyper-parameters of the support vector regression model. Finally, the proposed model is validated by using the data from a drilling site in the Shennongjia area, Central China. The results demonstrate that the proposed method outperforms eight well-known methods and another three methods without different data preprocessing procedures in prediction accuracy.

The rate of penetration (ROP) model is of great importance in achieving a high efficiency in the complex geological drilling process. In this paper, a novel two-level intelligent modeling method is proposed for the ROP considering the drilling characteristics of data incompleteness, couplings, and strong nonlinearities. Firstly, a piecewise cubic Hermite interpolation method is introduced to complete the lost drilling data. Then, a formation drillability (FD) fusion submodel is established by using Nadaboost extreme learning machine (Nadaboost-ELM) algorithm, and the mutual information method is used to obtain the parameters, strongly correlated with the ROP. Finally, a ROP submodel is established by a neural network with radial basis function optimized by the improved particle swarm optimization (RBFNN-IPSO). This two-level ROP model is applied to a real drilling process and the proposed method shows the best performance in ROP prediction as compared with conventional methods. The proposed ROP model provides the basis for intelligent optimization and control in the complex geological drilling process. (C) 2019 Elsevier B.V. All rights reserved.

In order to realize energy-saving motor drive systems, the energy consumed by the motor drive must be reduced as more as possible. To this end, we propose a new and integrated model for motor drives including both the bending vibration and the torsional vibration as well as the gyroscopic effect in this paper. Special attention is paid to the coupling of these two kinds of vibrations. The approach is based on Lagrangian mechanics and the precise computation of the energies in the bending vibration and the torsion vibration from the viewpoint of flexible material mechanics.

This paper proposes a new filter design method that can smooth the wind power generation output while keeping the energy capacity and power rating of the energy storage system (ESS) small. Wind power generation causes frequency fluctuations in power systems. Therefore, in many cases, the ESS is needed for smoothing the fluctuations of wind generator output. In order to improve the performance of the ESS controlled with the conventional low-pass filter (LPF), this paper investigates other types of filters. First, the relations between filter and energy capacity/power rating of the ESS are disclosed through the worst-case study. Then, the frequency fluctuation analysis of the power system with respect to the wind power generation is carried out in order to realize a pin-point smoothing. Finally, a metaheuristic optimization method is proposed for the multiobjective design of the filter based on these relations so as to achieve an optimal tradeoff between the output smoothing and energy capacity/power rating of the ESS. It is shown via simulation analysis with real wind data that the proposed method can suppress the frequency fluctuation of the power system effectively with the ESS having smaller energy capacity and power rating.

This paper addresses the real-time control of transmission power for small cell base stations (SBSs) exploiting energy- harvesting sources. We employ model predictive control and optimize an objective function that contains the number of users with a given quality of experience and the average state of charge. We first determine the number of active SBSs in the viewpoint of energy efficiency and then approximate the objective function. Finally, we illustrate the proposed method through a numerical example, comparing it with a static method based on statistical information.

A robust tracking control method is presented in this paper for an uncertain plant with an unknown state delay and an exogenous disturbance. The effects of the uncertainties, the delay, and the exogenous disturbance are treated as a total disturbance thus, the construction of the observer does not need the delay information. The system design is divided into the design of the gains of the state-feedback controller as well as the design of the gains of the observer and the improved equivalent-input-disturbance (EID) estimator. The pole-assignment method is used to design the gains of the state-feedback controller of a simplified system. A robust-stability condition in the form of a linear matrix inequality is derived to determine the gains of the observer and the improved EID estimator. Since the devised Lyapunov functional is of a more general form than those in existing EID-based methods and the restrictive commutative condition is avoided in this design, the developed design method is less conservative. Finally, comparisons of the developed method with a sliding-mode control method for a matched disturbance and a conventional EID-based method for an unmatched disturbance illustrate the validity and superiority of the developed method.

In the robust control of systems with dynamic uncertainty, there aremainly two methodologies: one based on the gain bound of uncertainty and the other based on the phase bound. The well-known small-gain approach belongs to the former and is able to conduct nonconservative robust performance design by using μ-synthesis. Meanwhile, the study on the latter case is still under way, and only a positive-real method was proposed by the authors recently to handle the robust performance synthesis. This paper aims at improving the positive-real method. The key idea is to model a passive uncertainty not only by its phase bound but also utilizing its largest gain. This is achieved by applying the notion of output strict passivity well known in nonlinear control. Furthermore, the negative-imaginary uncertainty can be treated in this framework more naturally. A case study on a vibration system validates the advantage of the present method.

An improved equivalent-input-disturbance (EID) approach is devised to enhance the disturbance-rejection performance for a strictly proper plant with a state delay in a modified repetitive-control system. A gain factor is introduced to construct an improved EID estimator. This increases the flexibility of system design and enables the adjustment of the dynamical performance of disturbance rejection. Moreover, the commutative condition, which is widely used for the conventional EID estimator, is avoided. Thus, it reduces the conservativeness of design by removing the constraints imposed by the commutative condition. The system is divided into two subsystems, and the separation theorem is applied to simplify the design. For one subsystem, the delay information on both the modified repetitive controller and the plant is used to reduce the conservativeness of stability condition. The resulting linear matrix inequality (LMI) is used to find the gain of the state-feedback controller. Another LMI is derived to design the gains of the state observer and the improved EID estimator for the other subsystem. A case study on a metal-cutting system validates the superiority of the developed method.

<p>This letter addresses reduction of the energy capacity of energy storage systems (ESSs) using H_∞ control theory. The ESSs have been widely used for smoothing fluctuations of wind generator outputs. We propose a direct and efficient method to design controllers that achieve the minimization considering the trade-off between the energy capacity of the ESSs and frequency variation in power systems.</p>

This paper addresses the robust performance synthesis problem for systems with positive-real or negative-imaginary uncertainty. The conventional robust performance design methods are mostly based on the gain information of uncertainty, which is not suitable for uncertainties characterized by the phase information. The approach taken here is based on the input-output relation of systems in the frequency domain, which enables a less conservative yet numerically tractable solution to the robust performance synthesis. The effectiveness is validated through a detailed case study on the head positioning control of hard disk drives. The results and comparison demonstrate that the proposed method can achieve a much higher robust performance than existing methods. (C) 2017 Elsevier Ltd. All rights reserved.