The cyan-emitting BaSi₂O₂N₂:Eu²⁺ phosphor is a promising narrow-band and high-efficiency luminescent material used in wide-color-gamut white light-emitting diodes (wLEDs). However, its serious degradation under thermal attacks hinders its practical applications and needs to be improved. Herein, we proposed to deposit a nano-sized Al₂O₃ film around each BaSi₂O₂N₂:Eu²⁺ particle through atomic layer deposition (ALD) in a fluidized bed reactor to improve its thermal stability. Thermal gravimetric analysis results showed that the Al₂O₃ layer with a thickness of only 11 nm had an obvious anti-oxidization effect, by which the oxidation temperature in air of the Al₂O₃ coated phosphor was largely increased from ∼550 to ∼750 °C. Moreover, the Al₂O₃ coated phosphor remained 93% of its luminescence intensity in comparison to 73% of the uncoated one when degraded under water-steam at 200 °C for 24 h. The oxidization of both the BaSi₂O₂N₂ host matrix and the doped Eu²⁺ ions was reduced by the Al₂O₃ layer. Meanwhile, the wLEDs fabricated with the Al₂O₃ coated phosphor showed a luminous flux of 3 times higher than that of the uncoated one when aged under 100 mA for 300 h. The greatly improved thermal degradation property of BaSi₂O₂N₂:Eu²⁺ phosphor and the reliability of the wLEDs indicate that the ALD approach could be a feasible route to produce uniform and nano layers on phosphors and enhance their stability.

@en

Population ageing has become a severe problem worldwide. Human activity recognition (HAR) can play an important role to provide the elders with in-time healthcare. With the advantages of environmental insensitivity, contactless sensing and privacy protection, radar has been widely used for human activity detection. The micro-Doppler signatures (spectrograms) contain much information about human motion and are often applied in HAR. However, spectrograms only interpret magnitude information, resulting in suboptimal performances. We propose a radar-based HAR system using deep learning techniques. The data applied came from the open dataset “Radar signatures of human activities” created by the University of Glasgow. A new type of hybrid map was proposed, which concatenated the spectrograms amplitude and phase. After cropping the hybrid maps to focus on useful information, a convolutional neural network (CNN) based on LeNet-5 was designed for feature extraction and classification. In addition, the idea of transfer learning was applied for radar-based HAR to evaluate the classification performance of a pre-trained network. For this, GoogLeNet was taken and trained on the newly-produced hybrid maps. These initial results showed that the LeNet-5 CNN using only the spectrograms obtained an accuracy of 80.5%, while using the hybrid maps reached an accuracy of 84.3%, increasing by 3.8%. The classification result of transfer learning using GoogLeNet was 86.0%.

@en

While chemical industrial development in China is growing rapidly, the corresponding safety training resources remain inadequate, which may often lead to increased risk of chemical accidents. These accidents are often associated with the negligence of safety management, poor safety hazard awareness, and lack of safety practice. In order to alleviate these prominent risk factors in chemical industries in China, our study develops a talent training model related to chemical process safety. First, we propose an approach for establishing the “talent training model” related to chemical process safety, consisting of three steps: analyzing the current status and existing problems of talent training related to chemical process safety, determining the theoretical basis and training objectives for developing interdisciplinary talents, and designing a new talent training model. Second, we establish a talent training model using the proposed method, which includes a comprehensive curriculum system, a diversified teaching pattern, and a quintuple evaluation method. Furtherly, we determine the expected outcomes of the talent training model. The research results provide an innovative chemical process safety training method that is applicable nationwide, also it works as a reference for other rapidly developing countries in the chemical process industry to improve safety within the chemical industry.

@en

A novel fault tolerant control algorithm is proposed in this paper based on model reference reinforcement learning for autonomous surface vehicles subject to sensor faults and model uncertainties. The proposed control scheme is a combination of a model-based control approach and a data-driven method, so it can leverage the advantages of both sides. The proposed design contains a baseline controller that ensures stable tracking performance at healthy conditions, a fault observer that estimates sensor faults, and a reinforcement learning module that learns to accommodate sensor faults using fault estimation and compensate for model uncertainties. The impact of sensor faults and model uncertainties can be effectively mitigated by this composite design. Stable tracking performance can also be ensured even at both the offline training and online implementation stages for the learning-based fault tolerant control. A numerical simulation with gyro sensor faults is presented to demonstrate the efficiency of the proposed algorithm.@en

Industrial bifacial n-type front and back contact (nFAB) silicon solar cells, consisting of a boron-doped p+ emitter and a phosphorus-doped n+ back surface field (BSF), are known to give good bifaciality, high and stabilized efficiency. One possible approach to further enhance the cell efficiency is to convert conventional passivated emitter and rear totally diffused (PERT) into rear locally diffused (PERL) structure. Herein, bifacial nFAB PERT and PERL cells are fabricated by combining atmospheric pressure chemical vapor deposition (APCVD) of phosphosilicate glass (PSG) as doping source and laser processing. For PERL cells, two approaches are studied to locally form phosphorus-doped BSF: 1) laser doping, and 2) laser ablation of a diffusion barrier layer. For ablation approach, an alkaline treatment is introduced immediately after laser process, which leads to the formation of locally textured BSF. Due to this locally textured contact, the resultant fill factor (FF) and series resistance (Rs) loss of the PERL cells are even less than that of the reference PERT cells. As a result, the champion cell of PERL shows a good efficiency of 21.3% with open-circuit voltage (Voc) of 662 mV, short-circuit current density (Jsc) of 39.6 mA cm−2, and a high FF of 81.1%.@en