Switched-capacitor (SC) converters are a type of variable structure systems.SC converters have been typically furnished with open-loop control or some linear feedback controls. However, these approaches provide slow dynamic response, small operation region, and large voltage ripple. Some new nonlinear control methods have been proposed in the literatures to overcome these barriers, however, such control methods are only useful for applications involving low power SC converters because the limitation of theses control methods. This paper presents a nonlinear Constant On Time Variable Frequency One-Cycle Control (CVFOCC) method based on dynamic capacitor ampere-second balance transient calculation (CASBTC) modeling method and one-cycle control (OCC) technique for high power SC converters. This method successfully addresses the regulation issue in high power SC converters. Using high power two-switch boosting switched-capacitor (TBSC) converters as examples, simulation and experiment results demonstrated that CVFOCC method significantly improved the dynamic response for TBSC converter and achieved excellent line and load regulation in a wide operation range.

The effects of off-state breakdown on characteristics of power AlGaAs/InGaAs pseudomorphic HEMTs (PHEMTs) are investigated in detail. While the gate leakage current is substantially decreased after breakdown stress, no obvious changes in drain-to-source current and transconductance are observed. Prior to breakdown stress, gate leakage current shows a nearly ideal 1/f noise characteristic with an Ig/sup 2/ dependence, suggesting a surface generation-recombination current from the interface of the passivation layer. After stress, the gate current noise can be drastically reduced. The results suggest an alternative for alleviating the gate leakage current in PHEMTs.

An adaptive machine learning (ML) based smart manufacturing interactive cyber physical human system (ICPHS) is conceptualized, designed, and implemented. One of its significant properties is an ML model that during deployment self-evolves with the streaming data in a self-labeling manner. This automated adaptive ML system is realized by leveraging the underlying causality during human machine interactions, and initialized by a priori domain knowledge and preprocessed public datasets. The system defines a causal and temporal mapping of worker and machine states where one side can label the other automatically. A case study in machines with different automation levels is conducted in a multiuser facility by using finite state machine representations for machine energy states, worker activity states, and interaction transition functions. The fully automated adaptive ML system improves its accuracy for human machine interaction detection by up to 12.5% and shows potential to recognize more fine-grained actions.