In the process of finding high-performance materials for organic photovoltaics (OPVs), it is meaningful if one can establish the relationship between chemical structures and photovoltaic properties even before synthesizing them. Here, we first establish a database containing over 1700 donor materials reported in the literature. Through supervised learning, our machine learning (ML) models can build up the structure-property relationship and, thus, implement fast screening of OPV materials. We explore several expressions for molecule structures, i.e., images, ASCII strings, descriptors, and fingerprints, as inputs for various ML algorithms. It is found that fingerprints with length over 1000 bits can obtain high prediction accuracy. The reliability of our approach is further verified by screening 10 newly designed donor materials. Good consistency between model predictions and experimental outcomes is obtained. The result indicates that ML is a powerful tool to prescreen new OPV materials, thus accelerating the development of the OPV field.
In the electricity market, customers have many choices to reduce electricity cost if they can economically schedule their power consumption. Renewable hybrid system, which can explore solar or wind sources at low cost, is a popular choice for this purpose nowadays. In this paper optimal energy management for a grid-connected photovoltaic-battery hybrid system is proposed to sufficiently explore solar energy and to benefit customers at demand side. The management of power flow aims to minimize electricity cost subject to a number of constraints, such as power balance, solar output and battery capacity. With respect to demand side management, an optimal control method (open loop) is developed to schedule the power flow of hybrid system over 24 hours, and model predictive control is used as a closed-loop method to dispatch the power flow in real-time when uncertain disturbances occur. In these two kinds of applications, optimal energy management solutions can be obtained with great cost savings and robust control performance.
Aspect-based sentiment classification (ABSC) aims to predict sentiment polarities of different aspects within sentences or documents. Many previous studies have been conducted to solve this problem, but previous works fail to notice the correlation between the aspect’s sentiment polarity and the local context. In this paper, a Local Context Focus (LCF) mechanism is proposed for aspect-based sentiment classification based on Multi-head Self-Attention (MHSA). This mechanism is called LCF design, and utilizes the Context features Dynamic Mask (CDM) and Context Features Dynamic Weighted (CDW) layers to pay more attention to the local context words. Moreover, a BERT-shared layer is adopted to LCF design to capture internal long-term dependencies of local context and global context. Experiments are conducted on three common ABSC datasets: the laptop and restaurant datasets of SemEval-2014 and the ACL twitter dataset. Experimental results demonstrate that the LCF baseline model achieves considerable performance. In addition, we conduct ablation experiments to prove the significance and effectiveness of LCF design. Especially, by incorporating with BERT-shared layer, the LCF-BERT model refreshes state-of-the-art performance on all three benchmark datasets.
Overhead cranes are widely used in industrial applications for material displacing. Many linear or nonlinear control schemes have been proposed for overhead cranes and implemented on electronic systems, but energy efficiency of transportation has seldom been considered in motion planning. This paper aims at finding an optimal solution of motion planning in terms of energy efficiency for overhead cranes. Using the optimal control method an optimal trajectory is obtained with less energy consumption than the compared trajectories and is also satisfying physical and practical constraints such as swing, acceleration, and jerk. Besides the energy optimal model, we also propose two other models to optimize time efficiency and safety during transportation. The results obtained have been compared with some existing motion trajectories, and have been shown to be superior to these benchmarks in terms of energy efficiency, time efficiency and safety respectively.
Model predictive control (MPC) has been successfully applied to many transportation systems. For the control of overhead cranes, existing MPC approaches mainly focus on improving the regulation performance, such as tracking error or steady-state error. In this paper, energy efficiency as well as safety is newly considered in our proposed MPC approach. Based on the system model designed, the MPC approach is applied to minimize an objective function that is formulated as the integration of energy consumption and swing angle. In our approach, promising results in terms of low energy consumption and small swing angle can be found, whilst the solutions obtained can satisfy all practical constraints. Our test results indicate that the MPC approach can ensure stability and robustness of improving energy efficiency and safety.
SummaryVarious in situ synthesis methods have been developed for the polymerization of 3,4-ethylenedioxythiophene monomers, such as electropolymerization, oxidative chemical vapor deposition, and vapor phase polymerization. Meeting industrial requirements through these techniques has, however, proven challenging. Here, we introduce an alternative method to fabricate highly conductive poly(3,4-ethylenedioxythiophene) (PEDOT) films in situ by solution means. The process involves sequential deposition of oxidants (V2O5 in this case) and monomers. Excess reactants and by-products can be completely removed from the PEDOT film by MeOH rinsing. The obtained PEDOT films possess good crystallinity and high doping level, with carrier concentration three orders of magnitude higher than that of the commercial product (PH1000, Heraeus GmbH). The electrical conductivity of the as-cast PEDOT film reaches up to 1,420 S/cm. In addition, this method is fully compatible with large-scale printing techniques. These PEDOT conducting films enable the realization of flexible touch sensors, which demonstrate superior flexibility and sensitivity.
It is acknowledged that the structure of a material determines its activity or property. During the development of organic photovoltaic (OPV) materials, it is vitally important to build the relationship between chemical structures and photovoltaic properties. However, the conventional way based on trial-and-error experiments requires a significant amount of time and resources. Here, it is demonstrated that deep learning can be employed to quickly evaluate the performance of new OPV materials. The deep learning model allows direct use of pictures of chemical structures as input, possesses an excellent nonlinear analyzing capability, and has a low demand for computing power. After training the model with a database from the Harvard Clean Energy Project, it is able to predict the photovoltaic performance based on a given chemical structure of an OPV donor material. The prediction accuracy reaches 91.02% using a verification set of 5000 molecules. The codes are converted into visual pictures to understand how features are extracted by the model. In addition, the influence of database size on prediction accuracy is discussed. The model is further tested by using experimentally verified OPV materials and received positive results. Together, the results suggest that deep learning is promising for the quick evaluation of new OPV materials.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.