Conventional grid-connected voltage-source inverters (VSI) with power sources such as photovoltaic (PV) arrays or energy storage units usually require a two-stage power transfer. The first is a DC-DC stage where the energy flows from the primary source to the DC-link or vice versa. The second stage is the grid-connected stage, where the inverter is synchronized with the grid to achieve the desired power flow. In both stages, inductance is an essential component that can become considerably expensive and bulky as the power rating of the system increases. This paper introduces a novel method that eliminates the set of inductors used in the DC-DC stage in grid-connected systems with LCL filter configuration. This is achieved by integrating the DC-DC stage in the common mode (CM) of the LC inverter-side circuit while controlling the power flow in dual-active bridge (DAB) operation. The proposed method can reduce the cost associated with the inductance and decrease system footprint. This paper demonstrates the method, using Matlab simulations, in a single-phase application and verifies the results in an equivalent experimentation.

Eliminating common-mode (CM) current often necessitates costly galvanic isolation. Semiconductor-based active isolation can fix the issue at a far lower cost. This paper proposes an active-isolated non-inverting buck-boost converter and compares its cost to an equivalent galvanically isolated solution. The proposed topology integrates series-connected switching elements across both input and output interfaces, thereby attenuating voltage stress on individual semiconductors. An experimental prototype of a 30V to 10V-75V, 100W bipolar active-isolated non-inverting buck-boost converter has been developed to demonstrate the effectiveness of active isolation in minimizing CM leakage current. Together, these results demonstrate the technical and economic feasibility of semiconductor-based active isolation as an alternative to galvanic isolation.

The rise of new disease variants, such as COVID-19, influenza, and others, highlights the critical need for advanced epidemiological modeling to guide early-stage outbreak management, especially when vaccine options are not available or reliable. This paper presents a novel, hybrid, data-driven model that integrates Agent-Based Modeling (ABM) with an extended SEIHRD (Susceptible, Exposed, Infectious, Hospitalized, Recovered, and Dead) framework, enhanced by N-step Deep Q Reinforcement Learning (N-Step DQRL). This model merges ABM’s behavioral insights with the SEIHRD model’s progression dynamics, utilizing DQRL for adaptive, data-informed decision-making. It is particularly focused on enhancing non-pharmaceutical interventions, such as lockdown policies, which are crucial in managing outbreaks in the absence of vaccines. This approach strikes a balance between detailed analysis and scalability, vital for policymakers in responding to emerging disease variants. The model’s efficacy, as evidenced by an analysis of recent COVID-19 data, highlights its potential to significantly improve global pandemic preparedness and response, merging behavioral analysis with disease progression trends through the use of advanced deep learning techniques.

Accurate comprehensive parameter estimation and analysis are essential for high performance modeling and control strategy of interior permanent magnet synchronous motors (IPMSMs). This article proposes an improved electrical machine (EM) model considering cross coupling effects and rotor position compensation to accurately estimate parameters, including stator winding resistance, inductances and permanent magnet (PM) flux linkage. In this article, the stator winding resistance is estimated separately by decoupling from other unknown parameters through removing common elements, using basic measurements including speed, voltage and current. Furthermore, dq-axis inductances and mutual inductances are investigated and constructed into mapping over various currents through decoupling coefficients in the proposed mathematical model. Meanwhile, rotor position compensation is considered to reduce the effects of rotor position error on parameter estimation. The proposed approach can improve the accuracy of estimation which is validated by the experiment on a laboratory prototype IPMSM and compared with other estimation methods ignoring either mutual inductances or rotor position compensation.

Future electric vehicle (EV) traction motors require high power density, high efficiency, wider speed range, lower torque ripple, and lower weight and volume. Hairpin (HP) windings are a favored option for traction motor windings; however, their efficiency tends to degrade due to AC losses at high operating speeds. Therefore, HP winding designs that offer higher EM performances, including higher efficiencies in the full operational region, and lower winding weight are necessary for future EVs. In this regard, this paper investigated the arrangement of the HP windings within the stator slot, considering different phases, and utilized an improved winding function-based model and analytical AC loss estimation to propose an optimal concentric winding (CW) configuration for a commercially available 140 kW, 15000 rpm induction machine (IM). According to the results, except for the similar torque capacity in the maximum torque per ampere (MTPA) region, the proposed optimal CW IM configuration demonstrated superior overall performance and characteristics, including output power, torque, and efficiency, across a wide speed range; both MTPA and field weakening regions compared to the IM with distributed windings (DW). Additionally, the proposed CW configuration reduces winding weight by 12.96% compared to the IM with conventional DW, which is a significant advantage in terms of weight and volume reduction.

Multi–motor electric vehicles (MMEVs) have been identified as a solution to the inherent disadvantages of current electric vehicles (EVs) in energy consumption and driving range. These improvements are achieved by operating each motor in its peak efficiency region, as appropriate for the driving scenario. However, an MMEV’s increased efficiency is highly dependent on the sizes of the motors. Existing studies focus mostly on improving overall system efficiency through improved energy management control strategies. The few studies that do seek to size the powertrain, introduce additional components, and alter the power capabilities of the vehicle, often resulting in increased cost and reduced dynamic performance. This paper investigates electric motor sizing methodologies used in these studies and restructures their implementation without adding new components to the existing powertrain. The optimal torque split ratio is determined using a Non–Dominated Sorting Genetic Algorithm (NSGA–II) for achieving improved overall system efficiency while preserving dynamic performance. Furthermore, using the proposed methodology, the optimal torque–split ratio is determined for a 2021 Ford Mustang Mach–E EV case study. Improved system efficiency through the optimization does not require redesign of the Mach–E’s powertrain as no additional components are introduced thus avoiding additional manufacturing costs.

As a computationally efficient tool for the machine’s steady–state performance prediction, equivalent circuit model (ECM) of induction machines (IMs) has been an established option in literature. The results and performance predictions obtained from ECM are significantly affected by: (i) the leakage reactances as the function of the geometry of the machine’s rotor and stator slots (ii) the skin, proximity, and slotting effects and (iii) the saturation of the core. Simultaneous consideration of these effects has been ignored in conventional ECM of IMs for simplicity. In this paper, a novel permeance-based ECM is proposed and developed based on the dimensions of the stator and rotor slots to simultaneously incorporate leakage zig–zag, tooth top and overhang reactances into modeling steps. Saturation, slotting, proximity and skin effects are also fully taken into account to improve the accuracy of the modeling and performance prediction compared to the conventional ECM. Finite element analysis is used to verify the accuracy of the proposed ECM when compared to the conventional ECM based on the steady state performance characteristics such as torque, electromagnetic loss, and efficiency predictions.

Asymmetrical star-delta winding interior permanent magnet synchronous motor (IPMSM) is introduced as a capable option in supporting higher torque and lower torque ripple characteristics compared to the symmetrical star-delta winding IPMSM. This is at the expense of having higher radial electromagnetic (EM) force and potential vibration-related concerns including high acceleration. EM force and vibration reduction in the asymmetrical star-delta winding IPMSMs have not been well addressed in the literature so far. In order to improve the EM force and acceleration characteristics of the asymmetrical star-delta winding IPMSM while keeping its developed torque within the desired ranges, various stator and rotor geometrical parameters are defined and are changed individually. A sensitivity analysis is utilized to introduce the most effective geometrical design variables for the highlighted objectives. The reduction in the EM force, through the Maxwell-Stress tensor method, and the acceleration on the outer surface of the motor housing of the improved asymmetrical star-delta winding IPMSM structure over a base asymmetrical star-delta winding IPMSM model are investigated and reported via EM and structural simulations.

Induction machines are a popular choice for tractive applications due to inherent cost savings and performance benefits driving industry to search for an optimal rotor bar design. Induction machines suffer from low torque densities due to larger size and increased losses incurred in the rotor bars making these the performance objectives to be improved through optimization. Communicating through objective functions (OFs), a performance model to rapidly evaluate design parameters coupled with genetic algorithm (GA) can be used to produce an optimal rotor bar; however, conventional OF modeling may introduce function bias or complex coefficient calculations leading to dominated objectives, stalling and premature convergence leading to an unoptimized solution. In this paper, the rotor bar of a squirrel cage induction machine (SCIM) is modeled by a permeance based equivalent circuit model (ECM) creating a link between the rotor slot geometry and equivalent circuit parameters. The model considering skin and slotting effect as well as slot, zigzag, tooth top and overhang leakage reactance effects coupled with a multi- objective GA through novel hyperbolic tangent based OFs to optimize the rotor bar geometry. The optimal rotor bar shape proposed offers increased output torque and reduced total machine losses resulting in a higher operating efficiency.

Weight, torque, and efficiency of induction machines (IMs) are directly affected by their stator winding configuration. To evaluate this, this paper presents the electromagnetic (EM) performance comparison of four squirrel cage IMs using different winding configurations, namely, integral slot distributed winding (ISDW), fractional slot concentrated winding (FSCW), integral slot concentrated winding (ISCW) and integer slot concentric winding (ISCW2). The same active volumes with identical electric and magnetic loading constraints were assigned for all IMs with the same materials. The weight, torque, EM losses, and efficiency values in both maximum torque per ampere (MTPA) and field weakening regions were assessed. The results indicate that the ISCW2 IM possessed the best overall performance and characteristics in terms of torque, torque density and efficiency in a wide speed range among the investigated IMs; except its lower developed torque compared to the ISDW IM in MTPA region. To resolve this issue, a winding layout optimization was carried out via a winding function-based analysis to improve the torque performance of the ISCW2 IM in MTPA region as well.

Link prediction and node classification in social networks remain open research problems with respect to Artificial Intelligence (AI). Innate representations about social network structures can be effectively harnessed for training AI models in a bid to predict ties; and detect clusters via classification of actors with regard to a given social network. In this paper, we have proposed a distinct hybrid model: Representation Learning via Knowledge-Graph Embeddings and Convolution Operations (RLVECO), which hybridizes the strengths of Knowledge-Graph Embeddings (VE) and Convolution Operations (CO) in extracting and learning meaningful features from social graphs via Representation Learning (RL). RLVECO utilizes an edge sampling approach for exploiting features of a social graph via learning the context of each actor with respect to its neighboring actors.

This paper presents an implementation of wide bandgap devices in soft-switching inverter architectures to fully exploit the benefits of high switching frequencies and fast switching transients without the adverse effects of motor insulation degradation and elevated levels of electromagnetic interference. Circuit simulation tools and analytical loss models are utilized to validate the advantages of the auxiliary resonant commutated pole soft-switching converter architecture against the implementation of wide bandgap devices in a conventional two-level hard-switching inverter where techniques are employed to control the dV/dt of the semiconductor devices. In addition, it is found that high-frequency operation is possible without the penalty in efficiency since switching losses are minimized in soft-switching inverter topologies. This characteristic leads to increased inverter efficiency and reduced harmonic losses in the motor which makes the architecture attractive for future electric vehicle traction applications.

In recent times, Social Network Analysis (SNA) has become a very important and interesting subject matter with regard to Artificial Intelligence (AI) in that a vast variety of processes, comprising animate and inanimate entities, can be examined by means of SNA. As a result, prediction tasks within social network structures have become significant research problems in SNA. Hidden facts and details about social network structures can be effectively and efficiently harnessed for training AI models with the goal of predicting missing links/ties within a given social network. Thus, important factors such as the individual attributes of spatial social actors, and the underlying patterns of relationship binding these social actors must be taken into consideration because these factors are relevant in understanding the nature and dynamics of a given social network structure. In this paper, we have proposed an interesting hybrid model: Representation Learning via Knowledge-Graph Embeddings and ConvNet (RLVECN). Our proposition herein is designed for examining, extracting, and learning meaningful facts for resolving link prediction problems about social network structures. RLVECN utilizes an edge sampling approach for exploiting the representations of a social graph, via learning the context of each actor with respect to its neighboring actors, with the goal of generating vector-space embeddings per actor which are further harnessed for innate representations via a Convolutional Neural Network (ConvNet) sublayer. Successively, these relatively low-dimensional representations are fed as input features to a downstream classifier for solving link prediction problems in a given social network. Our proposition, RLVECN, has been trained and evaluated on six (6) real-world benchmark social graph datasets.

Node classification and link prediction problems in Social Network Analysis (SNA) remain open research problems with respect to Artificial Intelligence (AI). Inherent representations about social network structures can be effectively harnessed for training AI models in a bid to detect clusters via classification of actors as well as predict ties with regard to a given social network. In this paper, we have proposed a unique hybrid model: Representation Learning via Knowledge-Graph Embeddings and ConvNet (RLVECN). Our proposition is designed for analyzing and extracting expressive feature representations from social network structures to aid in link prediction, node classification and community detection tasks. RLVECN utilizes an edge sampling technique for exploiting features of a given social network via learning the context of each actor with respect to its associate actors.

Convection coefficient in the end-winding is a critical thermal parameter in Lumped Parameter Thermal Network (LPTN) model solution for an accurate motor winding temperature prediction. However, it is a challenging task to determine this convection coefficient due to complex heat and air circulation characteristics in the end-winding region. Until now, all researches focus on Totally Enclosed Fan-cooled (TEFC) Aluminum Rotor Induction Motor (ARIM) with a rotor having fins on its end-rings. But Copper Rotor Induction Motor (CRIM) has a rotor that does not have any fins on its end-rings. Hence, this research will determine convection coefficient in the end-region of Copper Rotor Induction Motor (CRIM) that has smooth rotor end. A Computational Fluid Dynamic (CFD) technique along-with a Lumped Parameter Thermal Network (LPTN) model is proposed to determine this convection coefficient. Thermal experiments on a 20-hp Copper Rotor Induction Motor (CRIM) are conducted to validate this proposed approach.

In this paper, supercapacitors and grid-tie inverters are compared as means of integrating regenerative braking functionality into elevator systems. In dynamic braking, a conventional motor drive is unable to utilize energy during braking periods because it is powered by a rectifier in which energy flows in only one direction. Typically, braking resistors are used to dissipate any excess energy generated that can result in a breakdown of the device if left uncontrolled. Alternatively, supercapacitors or grid-tie inverters can be installed to the DC-link and the energy either stored for future use or fed to the power grid, thus preventing energy waste. Models for these two regenerative braking methods are constructed and simulated in the Simulink environment in order to compare their energy efficiencies under several unique elevator usage patterns. In usage patterns that allow the motor to generate energy within a long duration, the simulation results show the supercapacitor option is less of an improvement in energy efficiency than the grid-tie method due to energy flowing through two energy conversion devices. On the contrary, in usage patterns that limit regeneration within short bursts, the grid-tie inverter option is less efficient than the supercapacitor method due to output current limitations.

Wide-band gap (WBG) power devices are gaining significant interest over conventional Si insulated gate bipolar transistor (IGBT) power devices for inverters in electric vehicle (EV) propulsion applications. Understanding motor-drive performance on a drive-cycle due to the influence of such emerging inverters is essential to design an optimal electric powertrain system which is superior in terms of cost, weight and efficiency when compared to the state-of-the-art. Specifically, this paper introduces a novel, hybrid approach that includes numerical and analytical simulations and analysis to consider drive-cycle load and switching characteristics of various power switches and harmonics generated by the inverter. A comprehensive list of performance indices including inverter and motor losses, current harmonics and torque ripple are initially selected and used in this approach. The impact of two-level IGBT, GaN, and SiC inverters on a surface permanent magnet machine are determined for a drive-cycle.

Wide band-gap (WBG) devices operating at higher switching frequencies have become attractive for EV traction inverter application in the recent years as they can be used to reduce harmonic distortion and switching loss. On the other hand, 800 V EV battery drivetrain systems have been recently deployed in commercially available in premium electric vehicles in order to achieve higher power for propulsion and faster charging. Given this scenario, this paper puts an effort to investigate and compare a 2-level inverter, a 3-level hybrid inverter and a nine-switch inverter for such an application. Motivation for investigating each of the inverter types is provided in the respective sections. This paper utilizes a novel hybrid modelling approach to get deeper insights into the power loss of inverter with device non-linearity and the harmonics, torque ripple and magnetic saturation in the motor due to different inverter types and their outputs. Cosimulation has been performed on each of the inverter types driving the same 100 kW surface-mounted permanent magnet motor on a WLTC drive cycle and results are discussed.

Due to high torque and power density, permanent magnet synchronous motor usually faces significant challenges for thermal stability. Moreover, it becomes difficult to keep thermal stability within the motor because the magnets that are embedded in the rotor or mounted on the rotor surface are susceptible to heat. Therefore, thermal investigation is one of the mandatory procedures during design phase of the permanent magnet synchronous motor to ensure thermal stability. In this paper, a universal lumped parameter thermal network model has been proposed for thermal design of all types of permanent magnet synchronous motors which can be used for wide range of winding configurations and any position of magnets in the rotor. The proposed analytical thermal model has been used to check the thermal performance of both newly designed interior and surface mounted permanent magnet synchronous motor prototypes for future electric vehicle application. Further, a numerical thermal model has been proposed also to check temperature distribution throughout all parts of both motors and for comparison.

This paper investigates permanent magnet temperature (PMT) modeling and estimation for permanent magnet synchronous machines (PMSMs) by using the measured speed harmonic. First, a linear temperature model is derived to demonstrate that the magnitude of the speed harmonic decreases linearly with the increase of PMT. To achieve this linear model, the speed harmonic is induced by the injected harmonic currents satisfying certain conditions developed in this paper. To improve the estimation performance, PMT estimation is represented in a state-space model based on the derived temperature model, and the Kalman filter is applied to estimate the PMT from the measured speed harmonic. Compared with existing methods, the proposed approach has advantages in terms of simplicity in estimation and robustness to the variation of machine resistance and inductances. The proposed Kalman filter based modeling and estimation approach is evaluated with extensive experiments on a laboratory PMSM drive system under different speed and load conditions.

This paper proposes a novel gradient descent based maximum torque per ampere (MTPA) control algorithm for interior permanent magnet synchronous machines (IPMSMs) by using the measured speed harmonics. The proposed approach does not require machine parameters and thus is not influenced by the machine and drive nonlinearities. Hence, the proposed approach can ensure a robust MTPA control under different loading conditions. Specifically, in the proposed approach, a small q-axis harmonic voltage is injected into the machine to induce a small harmonic component in the machine speed. Based on the PMSM torque equation, the mathematical relation between the induced speed harmonic and the output torque is derived, which shows that the magnitude of the induced speed harmonic is proportional to the output torque of an IPMSM. Therefore, the speed harmonic is explored to seek the MTPA angle, in which the MTPA angle is found when the speed harmonic magnitude is maximized. In particular, the gradient descent algorithm is employed to detect the MTPA angle, which is computationally efficient and converges quickly. The proposed approach is evaluated with both simulations and experiments based on a laboratory IPMSM drive system.

Permanent magnet synchronous motor (PMSM) drives include several design and control variables that affect power losses. The improvement in system efficiency through control techniques will reduce the size of inverter and motor cooling system as well as the operating costs of the system. This paper proposes a comprehensive analysis of loss controllability using various control parameters to reduce the motor as well as inverter losses. Power losses considered include switching and conduction losses in the inverter and fundamental and harmonic losses in the stator winding, core and rotor magnets. The inverter losses have been calculated using an analytical model as functions of DC link voltage, load current and switching frequency. A 2-D electromagnetic coupled loss model is developed for calculating the losses in a laboratory interior PMSM (IPM) for varying operating conditions. An analytical model is used to obtain a comprehensive understanding of the behavior of motor losses that were obtained from the electromagnetic model with respect to control parameters. The dependence of motor, inverter and the system level losses are tested and validated on a laboratory IPM drive system. Consequently, suggestions are made regarding selection of control variables in order to reduce system-level losses in PMSM drives.

Accurate information of permanent magnet (PM) flux linkage is of significance to high-performance control and condition monitoring of permanent magnet synchronous machines (PMSMs). During machine operation, the PM flux linkage can vary due to temperature rise and saturation. Thus, this paper investigates how temperature rise and saturation influence the PM flux linkage under different operation conditions. Under no-load condition, the PM flux linkage is estimated from the back-EMF test. Under load condition, a speed harmonic based PM flux linkage estimation approach is proposed, in which the PM flux linkage is estimated from the speed harmonic without requiring machine parameters. Thus, the proposed estimation approach is not affected by the machine and drive nonlinearities and thus can guarantee the estimation performance. The proposed approach is applied for PM flux linkage estimation under various loads and temperatures to investigate the influence of temperature rise and saturation on the PM flux linkage. Experimental results demonstrate that the PM flux linkage under no-load is larger than that under load due to magnetic saturation, the PM flux linkage under load decreases as the saturation level increases, and the one under both load and no-load decreases linearly as the PM temperature increases.

The equivalent circuit parameters of an interior permanent magnet synchronous machine (PMSM) vary under different operating conditions due to magnetic saturation and temperature rise. Knowledge of accurate machine parameters is of paramount importance for high performance PMSM control and condition monitoring. Therefore, this paper presents a novel parameter estimation method, which employs both the electrical and mechanical model of an interior PMSM (IPMSM) for simultaneously estimating three parameters; namely, the d- and q-axis inductances and permanent magnet flux linkage. Specifically, the proposed parameter estimation approach is mathematically justified at first. Thereafter, by using the machine mechanical model with the speed measurement, a third equation in addition to the dq-axis voltage equations is developed to make the estimation model full-rank for simultaneously estimating three parameters. Lastly, the proposed approach is implemented with the recursive least square algorithm to estimate machine parameters and it is validated with simulations and experimental tests on a laboratory IPMSM drive system.

Integrated charging (IC) technology in electric vehicles (EVs) employing conventional power electronics and motor drivetrain components facilitates level 3 fast charging capabilities with reduction in overall weight and cost of the vehicle. However, when the winding inductances of 3-phase interior permanent magnet synchronous machines (IPMSMs) are realized as line inductors for battery charging, due to machine saliency, the magnetic fields produced by the sinusoidal AC supply results in 1) asymmetrical voltages in the air-gap as a function of rotor position and 2) relatively high magnitudes of oscillating torques causing harmful noise and vibrations. This can lead to significant AC losses with risk of permanent magnet demagnetization in the machine. Since the same motor is employed for traction application as well, it is of significance to optimally design the machine for IC operation. Thus, this paper exclusively investigates three IPMSM rotor configurations to be employed for IC operation in EV. This paper firstly presents a conventional dq-axis circuit model based damper design approach implemented for mitigating the saliency effect during IC. Then, a comparative performance analysis of the rotor configurations with damper bars during IC operation on machine saliency; asymmetrical voltage waveforms; oscillating electromagnetic torque; permanent magnet operating point and magnet losses is performed using finite-element analysis (FEA). Results obtained are analyzed and discussed.

In permanent magnet synchronous machine, high-frequency (HF) signal injection has been extensively investigated for permanent magnet temperature (PMT) estimation, in which PMT is estimated from the temperature-dependent HF resistance. Existing studies require prior knowledge on the HF resistance and neglect the fact that PMT is temporally correlated. This paper proposes a state-space model for PMT estimation, in which PMT is modeled with a piecewise linear equation to explore the temporal correlation. The state-space model is nonlinear due to unknown model parameters, which is required to be known in existing studies. This paper proposes to use expectation maximization particle filter (EM-PF) for simultaneous PMT and model parameter estimation. After EM-PF estimation, the state-space model becomes linear, so Kalman filter is employed for online PMT estimation. The proposed EM-PF along with a Kalman-filter-based approach can explore the temporal correlation among PMTs to improve the estimation performance, which can be hardly achieved in existing studies regarding PMT as a time-independent parameter. It should be noted that EM-PF is for initial PMT and model parameter estimation, while Kalman filter is for online PMT estimation ensuring computation efficiency and real-time capability. Our approach is validated with both numerical and experimental investigations.

For permanent magnet synchronous machines (PMSMs), accurate magnet flux linkage information is critical for permanent magnet condition monitoring and drive performance improvement. This paper proposes a novel particle-filter-based magnet flux linkage estimation approach by using the harmonics in the machine speed. In the proposed approach, the harmonic current is first injected to change the speed harmonics, and the particle filter is then applied to estimate the magnet flux linkage from the speed harmonics. With a proper selection of injected harmonic current, it is capable of simultaneously estimating the magnet flux linkage and reducing the torque ripples as well as the speed ripples. The proposed approach is based on the machine mechanical equation, so it is not influenced by the magnetic saturation, the resistance variation, and the inverter nonlinearity. Specifically, at first, a novel state-space model is developed based on the machine mechanical equation, which models the relation between the magnet flux linkage and the speed harmonic. The state-space model is nonlinear, so the particle filter is employed for a magnet flux linkage estimation. Our particle-filter-based estimation approach is validated on a laboratory PMSM drive system under different loads, speeds, and temperatures.