A NIR-II AIEgen-Based Supramolecular Nanodot for Peroxynitrite-Potentiated Mild-Temperature Photothermal Therapy of Hepatocellular Carcinoma

Compared to conventional photothermal therapy (PTT) which requires hyperthermia higher than 50 °C, mild-temperature PTT is a more promising antitumor strategy with much lower phototoxicity to neighboring normal tissues. However, the therapeutic efficacy of mild-temperature PTT is always restricted by the thermoresistance of cancer cells. To address this issue, a supramolecular drug nanocarrier is fabricated to co-deliver nitric oxide (NO) and photothermal agent DCTBT with NIR-II aggregation-induced emission (AIE) characteristic for mild-temperature PTT. NO can be effectively released from the nanocarriers in intracellular reductive environment and DCTBT is capable of simultaneously producing reactive oxygen species (ROS) and hyperthermia upon 808 nm laser irradiation. The generated ROS can further react with NO to produce peroxynitrite (ONOOˉ) bearing strong oxidization and nitration capability. ONOOˉ can inhibit the expression of heat shock proteins (HSP) to reduce the thermoresistance of cancer cells, which is necessary to achieve excellent therapeutic efficacy of DCTBT-based PTT at mild temperature (<50 °C). The antitumor performance of ONOOˉ-potentiated mild-temperature PTT is validated on subcutaneous and orthotopic hepatocellular carcinoma (HCC) models. This research puts forward an innovative strategy to overcome thermoresistance for mild-temperature PTT, which provides new inspirations to explore ONOOˉ-sensitized tumor therapy strategies.     

Vapor Phase Dealloying Derived Nanoporous Co@CoO/RuO2 Composites for Efficient and Durable Oxygen Evolution Reaction (Adv. Funct. Mater. 17/2023)

The development of low-cost and effective oxygen evolution reaction (OER) electrocatalysts to expedite the slow kinetics of water splitting is crucial for increasing the efficiency of energy conversion from electricity to hydrogen fuel. Herein, 3D bicontinuous nanoporous Co@CoO/RuO₂ composites with tunable sizes and chemical compositions are fabricated by introducing vapor phase dealloying of cobalt-based alloys. The influence of physical parameters on the formation of nanoporous Co substrates with various feature ligament sizes is systematically investigated. The CoO/RuO₂ shell is constructed by integrating a thin layer of RuO₂ on the inner surface of nanoporous Co, where the CoO interlayer is formed by annealing oxidization. The composite catalyst delivers an ultralow overpotential of 198 mV at 10 mA cm⁻², Tafel slope of 57.1 mV dec⁻¹, and long-term stability of 50 h. The superior OER activity and fast reaction kinetics are attributed to charge transfer through the coupling of Co O Ru bonds at the interface and the excellent nanopore connectivity, while the durability originates from the highly stable CoO/RuO₂ interface.     

LRSS-Net:轻量级遥感地物分割网络

针对目前基于深度学习的高分辨率遥感图像分割模型由于参数量大、计算复杂而导致高延迟、低响应的问题,提出了一种轻量级遥感地物分割方法,较好的平衡了速度和精度.该方法使用MobileNetV2进行特征粗提取,通过构建空间信息嵌入分支实现不同尺度的特征细提取,不同层次之间引入密集连接以获取密集的上下文信息.解码端设计特征融合优化策略逐层融合不同尺度的特征增加对细粒度特征的感知,同时以反卷积与双线性插值交替的上采样方式减少图像边缘信息丢失.最后采用交叉熵损失与Dice损失结合的方式加快网络收敛速度.为了验证所提方法的有效性,与几种常用的语义分割方法进行了对比实验.实验结果表明,所提算法的分割准确率为93.7%,MIoU为88.01%,可以实现地物的有效分割.     

融合MAML和CBAM的安卓恶意应用家族分类模型

为满足对新兴安卓恶意应用家族的快速检测需求,提出一种融合MAML(model-agnostic meta-learning)和CBAM(convolutional block attention module)的安卓恶意应用家族分类模型MAML-CAS。将安卓恶意应用样本集中的DEX文件可视化为灰度图,并构建任务集;融合混合域注意力机制CBAM,设计两个具有同等结构的卷积神经网络,分别作为基学习器和元学习器,这两个学习器在自动提取任务集中样本特征的同时,可从通道和空间两个维度来增强关键特征表达;利用元学习方法 MAML对两个学习器进行训练,其中基学习器完成特定恶意家族分类任务的属性学习,元学习器则学习不同任务的共性;在两个学习器训练完成后,MAML-CAS将获得初始化参数,在面对新的安卓恶意应用家族分类任务时,不需要重新训练,只需要少量样本就可以快速迭代;利用训练完成的基学习器提取安卓恶意应用家族特征,并利用SVM进行恶意家族分类。实验结果表明,MAML-CAS模型对新兴小样本安卓恶意应用家族具有良好的检测效果,检测速度较快,并具有较好的稳定性。     

Disturbance observer-based tracking control with prescribed performance specifications for a class of nonlinear systems subject to mismatched disturbances

This work investigates simultaneous prescribed performance tracking control and mismatched disturbance rejection problems for a class of strict-feedback nonlinear systems. A novel control scheme combining prescribed performance control, disturbance observer technique, and backstepping method is proposed. The disturbance estimations are introduced into the design of virtual control law design in each step to compensate the mismatched disturbances. To further improve the control performance, a prescribed performance function characterizing the error convergence rate, maximum overshoot, and steady-state error is used to construct the composite controller. The proposed controller guarantees transient and steady-state performance specifications of tracking error and provides much better disturbance attenuation ability simultaneously. Rigorous stability analysis for the closed-loop system is established by direct Lyapunov function method. It is shown that all the states in the resulting closed-loop system are stable, and the tracking error evolves within the prescribed performance boundaries and asymptotically converges to zero even in the presence of mismatched external disturbances. Finally, theoretical results are illustrated and demonstrated by two simulation examples.     

基于区块链的物联网任务协作信任管理方案

在物联网智能设备任务协作场景中,为解决设备交互不受信以及存在恶意设备破坏协作的问题,提出一种基于区块链的信任管理方案。半分布式的架构克服了集中式和分布式架构的不足,任务协作过程中兼顾了双方的信任,并设计了审查机制确保任务评价的真实可靠。信任的衡量中使用狄利克雷分布模拟任务评价的多样性,构建信任为数个状态,并利用马尔可夫链评估信任状态和全局信任。仿真结果表明,提出的方案可以客观衡量设备的信任,同时在检测恶意设备、提高协作任务成功率及避免误判方面具有优势,因此可以有效确保任务交互的可信和协作环境的安全。     

面向配电网弹性提升的需求响应优化配置策略

社会和经济的发展对电力系统供电质量提出更高的要求,尤其是与用户相连的配电网易受扰动影响,需加强其供能弹性。随着智能电网技术的发展,通过优化资源配置、灵活调度需求响应资源,可以提高电网供能弹性。采用基于力学映射模型的多阶段弹性评估模型量化需求响应的优化配置对配电网弹性的提升效果,并通过遗传算法求解该非线性优化问题。最后,以IEEE 33节点系统为例验证该模型在需求响应配置问题中的合理性和有效性,并分析优化配置方案对配电网弹性的提升效果。     

柔性直流输电系统接入交流电网的高频谐振风险运行方式辨识

柔性直流输电系统接入交流电网后存在高频谐振的风险,研究表明交流电网运行方式变化与高频谐振的产生存在内在关联。针对交直流系统中的高频谐振问题,首先构建了柔性直流输电系统直流侧和交流侧计及串补站的等效阻抗模型,分析了柔性直流输电系统接入交流电网的高频谐振机理。对柔性直流输电系统交流侧运行方式变化(特别是线路发生N-1、N-2故障)下等效阻抗的变化规律进行了理论分析和总结。此外,提出了一种柔性直流输电系统接入交流电网的高频谐振风险运行方式辨识方法,并以某地区电网局部交直流系统实际数据为例进行仿真,验证了所提辨识方法的有效性。     

IGBT芯片静态输出曲线连续测量方法

绝缘栅双极型晶体管(IGBT)芯片的静态输出曲线是考核其能量损耗及指导多芯片并联设计的重要指标之一。现有测量IGBT静态输出曲线的方法多采用商用化的功率器件分析仪,然而商业化功率器件分析仪存在价格昂贵、夹具单一的问题。亟需开发一种简单、快速、有效的静态输出曲线测量方法。面向高压IGBT芯片,提出一种新的静态输出曲线连续测量方法及测试电路,有效减小了IGBT芯片的电导调制效应和温升效应对静态输出曲线的影响。通过实时测量动态过程中的电压及电流,可以快速得到IGBT芯片静态输出曲线。通过对比本文连续法与功率器件分析仪的测量结果,证明了所提方法的有效性。