Due to the increasing performance and decreasing price of microcontrollers in recent years, applying a high sampling frequency has become more feasible in modern control that is known as multisampling technology. The first motivation to use multisampling is emulating the analog control, thereby reducing the control delay and improving the stability of power electronics controllers. On the other hand, more information can be acquired from the multisampled current/voltage, which helps to save the cost and improve the reliability. In this article, a review of the multisampling application in power electronics converters is provided. Starting from the control delay analysis in single/double-sampling control, the modeling, implementation, and related control strategies are given when using multisampling pulsewidth modulation. Then, based on the multisampled data, the applications in condition monitoring and parameter estimation are discussed. Finally, perspectives on challenges and future trends are discussed.

Accurate estimation of the state of health (SOH) of batteries is essential for maximizing the lifetime of the battery and improving the safety and economy of any energy storage system. Data-driven methods can use measurement data to effectively estimate the SOH, but the estimation performance depends on the relevance between the selected feature and SOH. In this paper, fuzzy entropy (FE) of battery voltage, is proposed as a new feature for SOH estimation and validated on Li-ion batteries. Compared with the traditional sample entropy, the FE can capture the variation of voltage during the battery degradation more efficiently in terms of the parameter selection, data noise, data size and test condition. Moreover, the aging temperature variation is involved in the established SOH estimator as the temperature is a disturbance variable in the real applications. The FE-SOH is used as the input-output data pair of the support vector machine, and a single-temperature model, a full-temperature model, and a partial-temperature model are established. As a result, the FE-based method has better estimation accuracy under aging temperature variation. The FE-based method also decreases the dependence on the size of the required training data. Finally, the effectiveness of the proposed method is verified by experimental results.

Research on the modeling and fault diagnosis of rotor eccentricities has been conducted during the past two decades. A variety of diagnostic theories and methods have been proposed based on different mechanisms, and there are reviews following either one type of electric machines or one type of eccentricity. Nonetheless, the research routes of modeling and diagnosis are common, regardless of machine or eccentricity types. This article tends to review all the possible modeling and diagnostic approaches for all common types of electric machines with eccentricities and provide suggestions on future research roadmap. The paper indicates that a reliable low-cost non-intrusive real-time online visualized diagnostic method is the trend. Observer-based diagnostic strategies are thought promising for the continued research.

Multisampling control provides an attractive way to reduce the control delays in LCL-filtered grid-connected inverters. Thereby, the bandwidth and stability margin will be improved. However, high-frequency switching harmonics (SHs) are introduced in the control loop when the inverter-side current is sampled. To investigate the effect of multisampled high-frequency SHs, the relationship between the double-update pulsewidth modulation (PWM) and multiupdate PWM is deduced through geometric deduction. It is shown that the multiupdate PWM is equivalent to double-update PWM with sampling instant shift, and the equivalent Nyquist frequency is equal to the switching frequency. Moreover, the nonaveraged value of current is sampled within one switching period and aliased low-order harmonics will appear in the grid-side current. Hence, filtering the multisampled SHs is necessary, and an improved repetitive filter (IRF) is proposed to remove all the sampled SHs and keep the advantage of phase boost using the multisampling control. The method is evaluated with a single-loop inverter-side current control, and its effectiveness is verified through the simulation and experiment.

Multisampling control provides an attractive way to improve the dynamics and stability of high-power three-phase grid-connected inverters. Beyond the desired performance in current control, more information can be acquired through multisampling. However, such an assumption is rarely discussed in grid-connected inverters. In order to fill this gap, a line voltage sensorless control scheme is proposed for LCL -filtered inverters. In this article, a multisampling is performed on the inverter-side current, and the linear regression of the multisampled current during zero vectors of inverter-modulation voltage is employed to estimate the filter capacitor voltage. Consequently, the cost of line voltage sensors is reduced and the failure of line voltage sensors is prevented when using multisampling current control. Furthermore, to address start-up transients, the line voltage can still be estimated by temporarily locking the upper arms of the inverter, and dc-link voltage can be precharged to the target value at the same time. Experimental results validate the effectiveness of the method on a down-scale inverter.

Machine Learning is widely studied in battery state of health (SOH) estimation due to its advantage in establishing the non-linear mapping between measurements and SOH. However, the requirement of a big dataset and the lack of robustness limit the practical application, especially in small sample learning. To tackle these challenges, an optimal ensemble framework called BaggELM (bagging extreme learning machine) is proposed for battery SOH estimation. Specifically, the required dataset is reduced by optimizing the input voltage and the hyperparameters of the BaggELM algorithm. Moreover, a statistical post-processing method is used to aggregate multiple ELMs, and the final estimate is determined by the maximum probability density value. As a result, the effects of random parameterization of ELM and the training data size on SOH estimation are suppressed, thus improving the robustness and accuracy of the conventional BaggELM. Compared to other classic machine learning methods, the proposed method reduces the required data size, maintains strong robustness and high estimation accuracy, making it a promising solution for small-sample-based SOH estimation. Finally, the effectiveness of the proposed estimation framework is verified using the accelerated aging dataset from Lithium Iron Phosphate (LFP) and Nickel Manganese Cobalt (NMC) batteries.