Suppressing the spiking of a synchronized array of Izhikevich neurons

Nonlinear Dynamics - The aim of this research work is to investigate the synchronization and desynchronization of an array of non-identical Izhikevich neurons in a star-like configuration. Both...     

Solvability of the power flow problem in DC overhead wire circuit modeling

Applications of Mathematics - Proper traffic simulation of electric vehicles, which draw energy from overhead wires, requires adequate modeling of traction infrastructure. Such vehicles include...     

超声波燃气表矩形流道的雷诺修正系数仿真研究

在超声波流量计测量技术中, 雷诺修正系数相关的研究对于提高计量精度有重要作用. 为研究矩形流道的雷诺修正系数与雷诺数的关系, 对矩形流道在常温常压流量较小情况下进行仿真, 结果发现: 矩形流道层流状态下的雷诺修正系数与雷诺数呈线性相关. 保持压强、体积流量不变, 在不同温度下进行仿真及拟合, 结果表明: 在不同温度下雷诺修正系数与雷诺数的线性关系依然满足. 在上述实验基础上, 对矩形流道湍流状态下的雷诺修正系数与雷诺数关系进行研究, 通过改变温度、压强和体积流量进行仿真及拟合发现, 矩形流道湍流状态下雷诺修正系数与雷诺数呈非线性相关.     

污泥基吸附剂对亚甲基蓝和环丙沙星的吸附性能研究

亚甲基蓝和环丙沙星是水体中2种污染物, 对生态环境有潜在危害. 本文以市政剩余活性污泥为原料, 氯化锌为活化剂热解制备污泥基吸附剂, 研究盐酸酸洗浓度、氯化锌浓度、热解温度、热解时间等对污泥基吸附剂吸附水中亚甲基蓝和环丙沙星性能的影响. 结果表明 (1)污泥基吸附剂对亚甲基蓝的吸附性能随盐酸酸洗浓度的增大而增加, 对环丙沙星的吸附性能则随盐酸酸洗浓度的增大呈先降后增趋势, 两者均在1.500mol·L-1盐酸浓度下取得最优值. (2)污泥基吸附剂对亚甲基蓝和环丙沙星的吸附性能随氯化锌浓度和热解温度的增加呈先升后降趋势, 在氯化锌浓度为4.0mol·L-1、热解温度为500℃时有最优值; 随着热解时间的延长, 污泥基吸附剂对亚甲基蓝和环丙沙星的吸附性能分别在500℃热解70min和80min时有最优值. (3)污泥基吸附剂的最佳制备条件为 氯化锌4.0mol·L-1活化2h、500℃热解70min和80min、1.500mol·L-1盐酸酸洗; 以此制得的污泥基吸附剂对亚甲基蓝和环丙沙星的去除率分别为97.7%和96.4%, 平衡吸附量分别为97.9mg·g-1和3.9mg·g-1, 且污泥基吸附剂对亚甲基蓝和环丙沙星的吸附过程均符合准二级动力学方程.     

Dynamics of laser-induced cavitation bubble during expansion over sharp-edge geometry submerged in liquid – an inside view by diffuse illumination

Laser ablation in liquids is growing in popularity for various applications including nanoparticle production, breakdown spectroscopy, and surface functionalization. When laser pulse ablates the solid target submerged in liquid, a cavitation bubble develops. In case of “finite” geometries of ablated solids, liquid dynamical phenomena can occur inside the bubble when the bubble overflows the surface edge. To observe this dynamics, we use diffuse illumination of a flashlamp in combination with a high-speed videography by exposure times down to 250 ns. The developed theoretical modelling and its comparison with the experimental observations clearly prove that this approach widens the observable area inside the bubble. We thereby use it to study the dynamics of laser-induced cavitation bubble during its expansion over a sharp-edge (“cliff-like” 90°) geometry submerged in water, ethanol, and polyethylene glycol 300. The samples are 17 mm wide stainless steel plates with thickness in the range of 0.025–2 mm. Bubbles are induced on the samples by 1064-nm laser pulses with pulse durations of 7–60 ns and pulse energies of 10–55 mJ. We observe formation of a fixed-type secondary cavity behind the edge where low-pressure area develops due to bubble-driven flow of the liquid. This occurs when the velocity of liquid overflow exceeds ~20 m s⁻¹. A re-entrant liquid injection with up to ~40 m s⁻¹ velocity may occur inside the bubble when the bubble overflows the edge of the sample. Formation and characteristics of the jet evidently depend on the relation between the breakdown-edge offset and the bubble energy, as well as the properties of the surrounding liquid. Higher viscosity of the liquid prevents the generation of the jet.     

Transfer learning in deep neural network based under-sampled MR image reconstruction

In Magnetic Resonance Imaging (MRI), the success of deep learning-based under-sampled MR image reconstruction depends on: (i) size of the training dataset, (ii) generalization capabilities of the trained neural network. Whenever there is a mismatch between the training and testing data, there is a need to retrain the neural network from scratch with thousands of MR images obtained using the same protocol. This may not be possible in MRI as it is costly and time consuming to acquire data. In this research, a transfer learning approach i.e. end-to-end fine tuning is proposed for U-Net to address the data scarcity and generalization problems of deep learning-based MR image reconstruction. First the generalization capabilities of a pre-trained U-Net (initially trained on the human brain images of 1.5 T scanner) are assessed for: (a) MR images acquired from MRI scanners of different magnetic field strengths, (b) MR images of different anatomies and (c) MR images under-sampled by different acceleration factors. Later, end-to-end fine tuning of the pre-trained U-Net is proposed for the reconstruction of the above-mentioned MR images (i.e. (a), (b) and (c)). The results show successful reconstructions obtained from the proposed method as reflected by the Structural SIMilarity index, Root Mean Square Error, Peak Signal-to-Noise Ratio and central line profile of the reconstructed images.