Data-driven rogue waves solutions for the focusing and variable coefficient nonlinear Schrödinger equations via deep learning.

In this paper, we investigate the data-driven rogue waves solutions of the focusing and the variable coefficient nonlinear Schrödinger (NLS) equations by the deep learning method from initial and boundary conditions. Specifically, first- and second-order rogue wave solutions for the focusing NLS equation and three deformed rogue wave solutions for the variable coefficient NLS equation are solved using physics-informed memory networks (PIMNs). The effects of optimization algorithm, network structure, and mesh size on the solution accuracy are discussed. Numerical experiments clearly demonstrate that the PIMNs can capture the nonlinear features of rogue waves solutions very well. This is of great significance for revealing the dynamical behavior of the rogue waves solutions and advancing the application of deep learning in the field of solving partial differential equations.

Manoeuvre Target Tracking in Wireless Sensor Networks Using Convolutional Bi-Directional Long Short-Term Memory Neural Networks and Extended Kalman Filtering.

Aiming at the problem that traditional wireless sensor networks produce errors in the positioning and tracking of motorised targets due to noise interference, this paper proposes a motorised target tracking method with a convolutional bi-directional long and short-term memory neural network and extended Kalman filtering, which is trained by using the simulated RSSI value and the actual target value of motorised targets collected from the convolutional bi-directional neural network to the sensor anchor node, so as to obtain a more accurate initial value of the manoeuvre target, and at the same time, the extended Kalman filtering method is used to accurately locate and track the real-time target, so as to obtain the accurate positioning and tracking information of the real-time target. Through experimental simulation, it can be seen that the algorithm proposed in this paper has good tracking effect in both linear manoeuvre target tracking scenarios and non-linear manoeuvre target tracking scenarios.

[Research and Implementation of 433 MHz-based Wireless Implantable Pacing System].

In order to solve the problem of communication interference and communication distance caused by the rapid pacing system when establishing the rapid atrial fibrillation model, a low-power implantable pacing system based on 433 MHz communication frequency to form a star network is designed. The system includes an implantable pacemaker, a programmer head, and programmer software. The pacemaker is composed of a wireless communication module, a pacing module, an ECG monitoring module, and a power management module. The programmer head acts as an intermediate node in the star network and is controlled by PC programmer software to program each pacemaker. This article introduces the hardware design and software flow of each part of the system, and describes the results of in vivo simulation and in vivo animal models of the system. The results show that the designed system and application method are effective and feasible for the rapid atrial pacing atrial fibrillation model. 433 MHz wireless communication, implantable, pacemaker system, low-power, ECG monitoring.

Stability and spectroscopic analysis of CsPbBr3 quantum dots modified with 2-n-octyl-1-dodecanol.

Due to their excellent optical and electrical properties, all-inorganic metal halide perovskite CsPbBr3 quantum dots (QDs) have become one of the most promising materials in the field of optoelectronics during recent years. However, the stability of CsPbBr3 QDs limits their practical application and further development to a certain extent. In order to improve their stability, CsPbBr3 QDs were modified with 2-n-octyl-1-dodecanol for the first time in this paper. The 2-n-octyl-1-dodecanol-modified CsPbBr3 QDs were prepared by the ligand-assisted reprecipitation (LARP) method at room temperature in an air environment. Then the stability of the samples was tested at different temperatures and humidity. When the humidity was 80%, the photoluminescence (PL) intensity of both unmodified and modified CsPbBr3 QDs increased to different degrees because the appropriate amount of water changed the crystallization environment. The PL intensity of the modified QDs increased, and the peak positions were basically not shifted, proving that they did not agglomerate. Thermal stability test results showed that the PL intensity of the 2-n-octyl-1-dodecanol-modified QDs could still maintain 65% of the original intensity at 90 °C, which is 4.6 times that of the unmodified CsPbBr3 QDs. Experimental results show that the stability of CsPbBr3 QDs is significantly improved after 2-n-octyl-1-dodecanol modification, which demonstrates the excellent surface passivation of CsPbBr3 QDs by 2-n-octyl-1-dodecanol.

Autonomous Binarized Focal Loss Enhanced Model Compression Design Using Tensor Train Decomposition.

Deep learning methods have exhibited the great capacity to process object detection tasks, offering a practical and viable approach in many applications. When researchers have advanced deep learning models to improve their performance, the model derived from the algorithmic improvement may itself require complementary increases in computational and power demands. Recently, model compression and pruning techniques have received more attention to promote the wide employment of the DNN model. Although these techniques have achieved a remarkable performance, the class imbalance issue during the mode compression process does not vanish. This paper exploits the Autonomous Binarized Focal Loss Enhanced Model Compression (ABFLMC) model to address the issue. Additionally, our proposed ABFLMC can automatically receive the dynamic difficulty term during the training process to improve performance and reduce complexity. A novel hardware architecture is proposed to accelerate inference. Our experimental results show that the ABFLMC can achieve higher accuracy, faster speed, and smaller model size.

The Short-Term Load Forecasting for Special Days Based on Bagged Regression Trees in Qingdao, China.

There are many factors that affect short-term load forecasting performance, such as weather and holidays. However, most of the existing load forecasting models lack more detailed considerations for some special days. In this paper, the applicability of the bagged regression trees (BRT) model combined with eight variables is investigated to forecast short-term load in Qingdao. The comparative experiments show that the accuracy and speed of forecasting have some improvements using the BRT than the artificial neural network (ANN). Then, an indicator variable is newly proposed to capture the abnormal information during special days, which include national statutory holidays, bridging days, and proximity days. The BRT model combined with this indicator variable is tested on the load series measured in 2018. Experiments demonstrate that the improved model generates more accurate predictive results than BRT model combined with previously variables on special days.

Effect of Traditional Chinese Exercise on Gait and Balance for Stroke: A Systematic Review and Meta-Analysis.

OBJECTIVE: A systematic review is conducted to determine the effect of traditional Chinese exercise for patients with stroke. METHODS: Studies are obtained from PubMed, Embase, Cochrane Library, EBSCO, Web of Science, and CNKI. Only randomized controlled trials were left to evaluate the effects of traditional Chinese exercise for patients with stroke, and with no limits on study data or language. The primary outcome was the Berg balance score (BBS), Functional walking scale. And a random-effects model was used to calculate the pooled mean difference (MD) with 95% confidence interval (CI). RESULTS: A total of 9 studies on 820 participants conform to the inclusion criteria, whereas eight studies on 704 participants are used as data sources for the meta-analysis, all trials were published between 2004 and 2013. The BBS indicates that the efficacy of traditional Chinese exercise on balance of patients with stroke is better than that of other training or no training in short term [MD (95%CI) = 11.85 [5.41, 18.30], P < 0.00001]. And the short physical performance battery, Functional walking scale, limit of stability were observed significant differences on balance (p<0.05) and gait (p<0.05) between traditional Chinese exercise and other exercises or no exercise. In addition, there is an article showed that some other form (physiotherapy exercises focused on balance) significantly improved balance ability for stroke patients compared to tai chi chuan practice (Berg test = 0.01, Romberg, and standing on one leg). CONCLUSION: In our meta analysis, the positive findings of this study suggest traditional Chinese exercise has beneficial effects on the balance ability in short term. However, we drew the conclusion according to the extreme heterogeneity, and evidence of better quality and from a larger sample size is required. Because of the inconsistent outcomes, there are short of enough good evidence for patients with stroke to prove the effects of traditional Chinese exercise on gait. SYSTEMATIC REVIEW REGISTRATION: http://www.crd.york.ac.uk/PROSPERO PROSPERO registration number: CRD42013006474.