Machine learning (ML) has emerged as one of the most powerful tools transforming all areas of science and engineering. The nature of molecular dynamics (MD) simulations, complex and time-consuming calculations, makes them particularly suitable for ML research. This review article focuses on recent advancements in developing efficient and accurate coarse-grained (CG) models using various ML methods, in terms of regulating the coarse-graining process, constructing adequate descriptors/features, generating representative training data sets, and optimization of the loss function. Two classes of the CG models are introduced: bottom-up and top-down CG methods. To illustrate these methods and demonstrate the open methodological questions, we survey several important principles in constructing CG models and how these are incorporated into ML methods and improved with specific learning techniques. Finally, we discuss some key aspects of developing machine-learned CG models with high accuracy and efficiency. Besides, we describe how these aspects are tackled in state-of-the-art methods and which remain to be addressed in the near future. We expect that these machine-learned CG models can address thermodynamic consistent, transferable, and representative issues in classical CG models.

This paper, the first of two companion papers, mainly reports theoretical considerations on working principle and structural design of a passive MEMS dc current sensor. The proposed sensor is capable of measuring dc current flowed in a two-wire appliance cord without using any cord separator like that used in Hall-effect based sensor. The crucial component of the proposed sensor is a cantilever made up of a piezoelectric film and a neodymium micro-magnet integrated on the end. The theoretical considerations are emphasized on the effect of partitioning piezoelectric thin film on the total output voltage, the magnetic force acted on the micro-magnet in a magnetic field generated by the dc current, and the cantilever stiffness subject to the sensing principle. A prototype of the proposed MEMS dc current sensor is thus micro-fabricated, with ten PZT partition plates and a 1-mm ³ micro-magnet, in accordance with the theoretical consideration of the design principle. The prototype device provides a physical support for further experimental verifications to be discussed in Part II.

This paper, the second part of two companion papers, reports experimental methodology and verification results based on the prototype device of the proposed passive MEMS dc current sensor in Part I. The relationship between output voltages and crucial factors, such as the relative position of the sensor to the two-wire appliance cord, different connection modes of the PZT partition plates, as well as the applied dc current has been characterized. The experimental methodology mainly includes those of how to position the sensor device, how to determine the measurement range for sensitivity characterization, and how to electrically connect the PZT partition plates in different modes. Sensitivity characteristics were studied in ac current from 8 to 400 mA and in dc current from 0.5 to 3 A, respectively. Two typical connection modes, i.e., in series and parallel, were investigated with ten PZT partition plates. In the case of dc current, the waveforms of the output voltage in the individual PZT plate were studied in detail. First, the initial peak voltages corresponding to ON-OFF switching of dc current can be clearly characterized with a higher amplitude (relatively lower frequency) for turning ON and a lower one (relatively higher frequency) for turning OFF. Second, the measured peak voltages are smaller than the analytical ones, but showing similar increasing tendency with the increasing of the dc current. In light of the above-mentioned experimental verifications, the proposed passive MEMS dc current sensor was proved to be applicable to direct measurements of the two-wire appliance cords without using cord separator. The proposed sensor devices are, therefore, more convenient than Hall-effect-based sensors in electricity end-use measuring and monitoring of dc power supply.

Elastomers are made of chain-like molecules to form networks that can sustain large deformation. Rubbers are thermosetting elastomers that are obtained from irreversible curing reactions. Curing reactions create permanent bonds between the molecular chains. On the other hand, thermoplastic elastomers do not need curing reactions. Incorporation of appropriated filler particles, as has been practiced for decades, can significantly enhance mechanical properties of elastomers. However, there are fundamental questions about polymer matrix composites (PMCs) that still elude complete understanding. This is because the macroscopic properties of PMCs depend not only on the overall volume fraction ($\varphi$) of the filler particles, but also on their spatial distribution (i.e., primary, secondary, and tertiary structure). This work aims at reviewing how the mechanical properties of PMCs are related to the microstructure of filler particles and to the interaction between filler particles and polymer matrices. Overall, soft rubbery matrices dictate the elasticity/hyperelasticity of the PMCs while the reinforcement involves polymer–particle interactions that can significantly influence the mechanical properties of the polymer matrix interface. For $\varphi$ values higher than a threshold, percolation of the filler particles can lead to significant reinforcement. While viscoelastic behavior may be attributed to the soft rubbery component, inelastic behaviors like the Mullins and Payne effects are highly correlated to the microstructures of the polymer matrix and the filler particles, as well as that of the polymer–particle interface. Additionally, the incorporation of specific filler particles within intelligently designed polymer systems has been shown to yield a variety of functional and responsive materials, commonly termed smart materials. We review three types of smart PMCs, i.e., magnetoelastic (M-), shape-memory (SM-), and self-healing (SH-) PMCs, and discuss the constitutive models for these smart materials.

A position and orientation correction scheme for current sensing based on magnetic piezoelectric cantilevers is proposed in this letter, offering an insight for the improvement of measurement accuracy. In this scheme, an array of magnetic piezoelectric cantilevers, converting mechanical vibration generated by current induced magnetic force applied on magnets at the ends of cantilevers into output voltages, is utilized. A system of nonlinear equations is established in terms of amplitude of objective current, position, and orientation errors. They are derived by solving the system of equations treated as variables, instead of parameters. As a result, the measurement error caused by position and orientation errors decreased substantially. Experiments of passive current sensing for two-wire appliances prove the validity of the proposed position and orientation correction scheme. The application of the proposed scheme can be further extended to sensors based on the magnetic field sensing mechanism, for achieving higher measurement accuracy.

This paper proposes a method for passive sensing a two-wire DC current without using any cord separator. In this method, a piezoelectric thin-film cantilever with a micro-magnet on its end is positioned near a power cord. The DC current in the power cord induces a magnetic field, which generates a magnetic force acting on the micro-magnet. Consequently, the piezoelectric cantilever is bent and induces an output voltage that is sensitive to small variations of DC currents. A prototype device has been fabricated and experimentally studied. It was found that the initial peak amplitude of the piezoelectric output voltage increases linearly with the DC current value upon “ON-OFF” switching. Compared to the Hall-effect based sensing, this method has the advantages of no need of driving voltage or cord separators even for two-wire appliance cords. Apart from the conventional consumer electronics, this method may have a wide application foreground in the new emerging fields, such as energy vehicles, lithium ion battery, or smart power grid.