Space-Air-Ground Integrated Mobile Crowdsensing for Partially Observable Data Collection by Multi-Scale Convolutional Graph Reinforcement Learning

Mobile crowdsensing (MCS) is attracting considerable attention in the past few years as a new paradigm for large-scale information sensing. Unmanned aerial vehicles (UAVs) have played a significant role in MCS tasks and served as crucial nodes in the newly-proposed space-air-ground integrated network (SAGIN). In this paper, we incorporate SAGIN into MCS task and present a Space-Air-Ground integrated Mobile CrowdSensing (SAG-MCS) problem. Based on multi-source observations from embedded sensors and satellites, an aerial UAV swarm is required to carry out energy-efficient data collection and recharging tasks. Up to date, few studies have explored such multi-task MCS problem with the cooperation of UAV swarm and satellites. To address this multi-agent problem, we propose a novel deep reinforcement learning (DRL) based method called Multi-Scale Soft Deep Recurrent Graph Network (ms-SDRGN). Our ms-SDRGN approach incorporates a multi-scale convolutional encoder to process multi-source raw observations for better feature exploitation. We also use a graph attention mechanism to model inter-UAV communications and aggregate extra neighboring information, and utilize a gated recurrent unit for long-term performance. In addition, a stochastic policy can be learned through a maximum-entropy method with an adjustable temperature parameter. Specifically, we design a heuristic reward function to encourage the agents to achieve global cooperation under partial observability. We train the model to convergence and conduct a series of case studies. Evaluation results show statistical significance and that ms-SDRGN outperforms three state-of-the-art DRL baselines in SAG-MCS. Compared with the best-performing baseline, ms-SDRGN improves 29.0% reward and 3.8% CFE score. We also investigate the scalability and robustness of ms-SDRGN towards DRL environments with diverse observation scales or demanding communication conditions.     

振动噪声等高图的电驱动桥噪声预测及优化*

在项目开发前期通过优化电驱动桥扭矩特性的设计,可以规避噪声大问题。根据电驱动桥台架在对应扭矩下的振动噪声特性,提出了一套稳定高效的测试流程和分析方法。首先设计了一套完整的试验流程,制定了精准的数据分析方法。然后绘制出能够全面反映电驱动桥振动噪声特性的等高图。最后利用电驱动桥台架的振动噪声等高图,准确评估电驱动桥加速工况下振动噪声风险,为主机厂和电驱动桥零部件企业提供电驱动桥扭矩特性设计前期指导。利用该方法成功识别到某电驱动桥匀速及加速工况下电机及齿轮的啸叫问题。通过优化电驱动桥扭矩特性设计,电机48阶噪声峰值降低了8.5dB(A),确认该方法准确可靠,具备推广应用价值。     

An Algorithm Framework for Drug-Induced Liver Injury Prediction Based on Genetic Algorithm and Ensemble Learning

In the process of drug discovery, drug-induced liver injury (DILI) is still an active research field and is one of the most common and important issues in toxicity evaluation research. It directly leads to the high wear attrition of the drug. At present, there are a variety of computer algorithms based on molecular representations to predict DILI. It is found that a single molecular representation method is insufficient to complete the task of toxicity prediction, and multiple molecular fingerprint fusion methods have been used as model input. In order to solve the problem of high dimensional and unbalanced DILI prediction data, this paper integrates existing datasets and designs a new algorithm framework, Rotation-Ensemble-GA (R-E-GA). The main idea is to find a feature subset with better predictive performance after rotating the fusion vector of high-dimensional molecular representation in the feature space. Then, an Adaboost-type ensemble learning method is integrated into R-E-GA to improve the prediction accuracy. The experimental results show that the performance of R-E-GA is better than other state-of-art algorithms including ensemble learning-based and graph neural network-based methods. Through five-fold cross-validation, the R-E-GA obtains an ACC of 0.77, an F1 score of 0.769, and an AUC of 0.842.     

Development of a ~3He Gas Filling Station at the China Spallation Neutron Source

At the China Spallation Neutron Source(CSNS), we have developed a custom gas-filling station, a glassblowing workshop, and a spin-exchange optical pumping(SEOP) system for producing high-quality ~3He-based neutron spin filter(NSF) cells. The gas-filling station is capable of routinely filling ~3He cells made from GE180 glass of various dimensions, to be used as neutron polarizers and analyzers on beamlines at the CSNS. Performance tests on cells fabricated at our gas-filling station are conducted via neutron transmission and nuclear-magneticresonance measurements, revealing nominal filling pressures, and a saturated ~3He polarization in the region of 80%, with a lifetime of approximately 240 hours. These results demonstrate our ability to produce competitive NSF cells to meet the ever-increasing research needs of the polarized neutron research community.     

PTC: Full and Drift Particle Orbit Tracing Code for α Particles in Tokamak Plasmas

Fusion born α particle confinement is one of the most important issues in burning plasmas,such as ITER and CFETR.However,it is extremely complex due to the nonequilibrium characteristics,and multiple temporal and spatial scales coupling with background plasma.A numerical code using particle orbit tracing method(PTC)has been developed to study energetic particle confinement in tokamak plasmas.Both full orbit and drift orbit solvers are implemented to analyze the Larmor radius effects on α particle confinement.The elastic collisions between alpha particles and thermal plasma are calculated by a Monte Carlo method.A triangle mesh in poloidal section is generated for electromagnetic fields expression.Benchmark between PTC and ORBIT has been accomplished for verification.For CFETR burning plasmas,PTC code is used for α particle source and slowing down process calculation in 2D equilibrium.In future work,3D field like toroidal field ripples,Alfven and magnetohydrodynamics instabilities perturbation inducing α particle transport will be analyzed.     

Characterization of Roasting Time on Sensory Quality,Color, Taste,and Nonvolatile Compounds of Yuan An Yellow Tea

Roasting is crucial for producing Yuan An yellow tea (YAYT) as it substantially affects sensory quality. However, the effect of roasting time on YAYT flavor quality is not clear. To investigate the effect of roasting time on the sensory qualities, chemical components, odor profiles, and metabolic profile of YAYTs produced with 13 min roasting, 16 min roasting, 19 min roasting, 22 min roasting, and 25 min roasting were determined. The YAYTs roasted for 22 min got higher sensory scores and better chemical qualities, such as the content of gallocatechin (GC), gallocatechin gallate (GCG), free amino acids, solutable sugar, meanwhile the lightness decreased, the hue of tea brew color (b) increased, which meant the tea brew got darker and yellower. YAYTs roasted for 22 min also increased the contents of key odorants, such as benzaldehyde, nonanal, β-cyclocitral, linalool, nerol, α-cedrol, β-ionone, limonene, 2-methylfuran, indole, and longiborneol. Moreover, non-targeted metabolomics identified up to 14 differentially expressed metabolites through pair-wise comparisons, such as flavonoids, phenolic acids, sucrose, and critical metabolites, which were the main components corresponding to YAYT roasted for 22 min. In summary, the current results provide scientific guidance for the production of high quality YAYT.     

A MALDI-MS sensing chip prepared by non-covalent assembly for quantitation of acid phosphatase

A novel sensing chip was designed for MALDI-MS quantitation of acid phosphatase(ACP).The ACP sensing chip was constructed through non-covalent interaction of streptavidin and biotin for the assembly of biotinylated peptide substrate on biotinylated polyethylene-glycol(PEG)modified indium-tin oxide(ITO)slide.In the presence of ACP,the peptide substrate was dephosphorylated under acidic condition to generate a new mass signal.The quantitative assay of ACP was achieved with the mass signal ratio of product to the sum of product and left peptide substrate.Under optimal detection conditions,the ratio was linearly correlated with the concentration of ACP in the range of 0.05–12 g/L with a detection limit(LOD)of 0.04 g/L.The designed ACP sensing chip has been used to analyze ACP in complex clinical samples,which exhibited high selectivity,good repeatability,and admirably anti-interference ability.This work further demonstrates the concept of MS sensing and the application of MALDI-MS in quantitative analysis,and provides a convenient method for the quantitation of proteases in clinical diagnosis.     

Luminescence interference-free lifetime nanothermometry pinpoints in vivo temperature

Luminescence nanothermometry makes non-invasive and real-time temperature readings possible in living animals. However,the spectral fluctuation in tissues and fluids, as well as the interaction between fluorophores and environment hinders accuracy of the thermometry. Here, we report a luminescence lifetime-based nanothermometry which specifically addresses this problem. A temporal based calibration(lifetime sensing) in the NIR range, an endogenous thermal response as well as a polymer encapsulation evading environmental factors, altogether help to pinpoint temperature in vivo. Thanks to the highly condensed NdYb ions in a well-protected tiny core-shell nanocrystal(overall 11 nm), a temperature sensitivity about 2.07% K~(-1)(with 5% Yb~(3+) doped nanoparticles) and an accuracy of 0.27 K(with 25% Yb~(3+) doped nanoparticles) in biological fluids are achieved.Hopefully, combining thermally activated energy transfer nanothermometer with anti-interference lifetime thermometry would provide a more accurate temperature measurement for biological and preclinical studies.     

Mechanism Study on Nanoparticle Negative Surface Charge Modification by Ascorbyl Palmitate and Its Improvement of Tumor Targeting Ability

Surface charge polarity and density influence the immune clearance and cellular uptake of intravenously administered lipid nanoparticles (LNPs), thus determining the efficiency of their delivery to the target. Here, we modified the surface charge with ascorbyl palmitate (AsP) used as a negatively charged lipid. AsP-PC-LNPs were prepared by dispersion and ultrasonication of AsP and phosphatidylcholine (PC) composite films at various ratios. AsP inserted into the PC film with its polar head outward. The pK_a for AsP was 4.34, and its ion form conferred the LNPs with negative surface charge. Zeta potentials were correlated with the amount and distribution of AsP on the LNPs surface. DSC, Raman and FTIR spectra, and molecular dynamics simulations disclosed that AsP distributed homogeneously in PC at 1–8% (w/w), and there were strong hydrogen bonds between the polar heads of AsP and PC (PO²⁻), which favored LNPs’ stability. But at AsP:PC > 8% (w/w), the excessive AsP changed the interaction modes between AsP and PC. The AsP–PC composite films became inhomogeneous, and their phase transition behaviors and Raman and FTIR spectra were altered. Our results clarified the mechanism of surface charge modification by AsP and provided a rational use of AsP as a charged lipid to modify LNP surface properties in targeted drug delivery systems. Furthermore, AsP–PC composites were used as phospholipid-based biological membranes to prepare paclitaxel-loaded LNPs, which had stable surface negative charge, better tumor targeting and tumor inhibitory effects.