Search

Organic electrode materials have exhibited good electrochemical performance in batteries, but their voltages and rate capabilities still require improvement to meet the increasing demand for batteries with high energy and power density. Herein, we design and synthesize a branched dihydrophenazine-based polymer (p-TPPZ) as a cathode material for dual-ion batteries (DIBs) through delicate molecular design. Compared with the linear dihydrophenazine-based polymer (p-DPPZ, with a theoretical capacity of 209 mAh g–1), p-TPPZ possessed a higher theoretical capacity of 233 mAh g–1 and lower highest occupied molecular orbital energy levels,which resulted in a high actual capacity (169.3 mAh g–1 at 0.5 C), an average discharge voltage of 3.65 V (vs. Li+/Li) and a high energy density (618.2 Wh kg–1, based on the cathode materials). The branched structure of p-TPPZ led to a larger specific surface area than that of p-DPPZ, which was beneficial for the electrolyte infiltration and fast ionic transport, contributing to the high power density. Due to the fast reaction kinetics, even at a power density of 23,725 W kg–1 (40 C), the energy density still reached 474.5 Wh kg–1. We also made a detailed investigation of the p-TPPZ cathode's charge storage mechanism. This work will stimulate the further molecular design to develop organic batteries with both high energy and power density.

Both non-equal trail lengths and non-zero initial errors are practical challenges to learning control of robotic and mechatronic systems. Iterative learning to update input is still desired, because of the repetitive motion nature of the controlled objects. This paper concerns with the adaptive iterative learning control method for performing non-identical tasks. The time scaling technique is applied to normalize non-equal trail lengths, while the error-tracking approach is adopted for coping with initial errors. Theoretical results for performance analysis are presented in detail. The uniform convergence of the tracking error is examined, while boundedness of the variables in the closed-loop is characterized. It is shown that the fully-saturated learning algorithm plays an important role in assuring uniform boundedness of the control input. The proposed control design method does not require the magnitude transformation, and removes the assumption of identical initial conditions. The time scaling technique is verified to be effective in assuring the expected performance, for tracking tasks with non-equal trail lengths and initial errors.

As a class of familiar nonlinear systems, nonaffine systems are frequently encountered in practical applications. Currently, in the context of learning control, there is a lack of research results about such general class of nonlinear systems, especially for the case of performing infinite interval tasks. This article focuses on the incremental adaptive control for nonlinear systems in nonaffine form, without requiring periodicity or repeatability. Instead of using the integral adaptation, incremental adaptive mechanisms are developed and the corresponding control schemes are presented, by which the numerical integration for implementation can be avoided. With the proposed incremental adaptation, the implicit function theorem is introduced to solve the intractability problem of the nonaffine structure. The robustness (robust convergence) of the tracking error is characterized, with the aid of a proposed key lemma, while the boundedness of all the variables is examined. Numerical results are presented to verify the effectiveness of the proposed control design.

AI fairness is tasked with evaluating and mitigating bias in algorithms that may discriminate towards protected groups. This paper examines if bias exists in AI algorithms used in disaster management and in what manner. We consider the 2017 Hurricane Harvey when flood victims in Houston resorted to social media to request for rescue. We evaluate a Random Forest regression model trained to predict Twitter rescue request rates from social-environmental data using three fairness criteria (independence, separation, and sufficiency). The Social Vulnerability Index (SVI), its four sub-indices, and four variables representing digital divide were considered sensitive attributes. The Random Forest regression model extracted seven significant predictors of rescue request rates, and from high to low importance they were percent of renter occupied housing units, percent of roads in flood zone, percent of flood zone area, percent of wetland cover, percent of herbaceous, forested and shrub cover, mean elevation, and percent of households with no computer or device. Partial Dependence plots of rescue request rates against each of the seven predictors show the non-linear nature of their relationships. Results of the fairness evaluation of the Random Forest model using the three criteria show no obvious biases for the nine sensitive attributes, except that a minor imperfect sufficiency was found with the SVI Housing and Transportation sub-index. Future AI modeling in disaster research could apply the same methodology used in this paper to evaluate fairness and help reduce unfair resource allocation and other social and geographical disparities.

Hurricane Harvey in 2017 marked an important transition where many disaster victims used social media rather than the overloaded 911 system to seek rescue. This article presents a machine-learning-based detector of rescue requests from Harvey-related Twitter messages, which differentiates itself from existing ones by accounting for the potential impacts of ZIP codes on both the preparation of training samples and the performance of different machine learning models. We investigate how the outcomes of our ZIP code filtering differ from those of a recent, comparable study in terms of generating training data for machine learning models. Following this, experiments are conducted to test how the existence of ZIP codes would affect the performance of machine learning models by simulating different percentages of ZIP-code-tagged positive samples. The findings show that (1) all machine learning classifiers except K-nearest neighbors and Naïve Bayes achieve state-of-the-art performance in detecting rescue requests from social media; (2) using ZIP code filtering could increase the effectiveness of gathering rescue requests for training machine learning models; (3) machine learning models are better able to identify rescue requests that are associated with ZIP codes. We thereby encourage every rescue-seeking victim to include ZIP codes when posting messages on social media. This study is a useful addition to the literature and can be helpful for first responders to rescue disaster victims more efficiently.

In this paper, a barrier Lyapunov function (BLF)-based backstepping control design is proposed for uncertain rigid spacecraft with both input and output constraints. A modified BLF (MBLF) is constructed to extend the application scope of the traditional logarithmic BLF. Through using the MBLFs in each step of the backstepping design, an adaptive constrained control scheme is presented to guarantee the tracking performance and the constraint requirement of spacecraft systems, and the differentiation of the virtual control is avoided with the employment of the tracking differentiator. The uncertainty bounds are estimated by designing adaptive update laws, such that no prior knowledge is required on the bound of the lumped uncertainty including input saturation and faults. Numerical simulations demonstrate the effectiveness of the proposed scheme.

In this paper, an adaptive finite-time fault-tolerant control scheme is proposed for the attitude stabilization of rigid spacecrafts. A first-order command filter is presented at the second step of the backstepping design to approximate the derivative of the virtual control, such that the singularity problem caused by the differentiation of the virtual control is avoided. Then, an adaptive fuzzy finite-time backstepping controller is developed to achieve the finite-time attitude stabilization subject to inertia uncertainty, external disturbance, actuator saturation, and faults. Through using an error transformation, the prescribed performance boundary is incorporated into the controller design to guarantee the prescribed performance of the system output. Numerical simulations demonstrate the effectiveness of the proposed scheme.

Collaborative filtering (CF) approaches, which provide recommendations based on ratings or purchase history, perform well for users and items with sufficient interactions. However, CF approaches suffer from the cold-start problem for users and items with few ratings. Hybrid recommender systems that combine collaborative filtering and content-based approaches have been proved as an effective way to alleviate the cold-start issue. Integrating contents from multiple heterogeneous data sources such as reviews and product images is challenging for two reasons. Firstly, mapping contents in different modalities from the original feature space to a joint lower-dimensional space is difficult since they have intrinsically different characteristics and statistical properties, such as sparse texts and dense images. Secondly, most algorithms only use content features as the prior knowledge to improve the estimation of user and item profiles but the ratings do not directly provide feedback to guide feature extraction. To tackle these challenges, we propose a tightly-coupled deep network model for fusing heterogeneous modalities, to avoid tedious feature extraction in specific domains, and to enable two-way information propagation from both content and rating information. Experiments on large-scale Amazon product data in book and movie domains demonstrate the effectiveness of the proposed model for cold-start recommendation.