Optimal design of Raman fibre amplifier based on terminal value optimization strategy and shuffled frog leaping algorithm

This paper introduces an evolutionary algorithm, Shuffled Frog Leaping Algorithm (SFLA), to solve the optimization problem in designing the multi-pumped Raman Fibre Amplifier (RFA). SFLA is a powerful optimizer tool because of its efficient mathematical expressions and global search capability. We utilize SFLA to determine the optimal pump wavelengths and pump powers by minimizing the gain ripple of RFA. To accelerate calculations, a terminal value optimization strategy (TVOS) is incorporated into the evolution of SFLA. This proposed strategy takes the terminal power values of pumps as the decision variables in optimization. Then, the optimal original power values of the pumps are obtained by solving the Power Coupled Equations once, without using the traditional method of repetitive guesses.The combination of SFLA and TVOS enhances the efficiency of optimization and accelerates calculation, while satisfying the design requirements of RFA.The simulation results show that nearly 65% of computational time has been saved compared with the traditional average power analysis. The 4-pumped C+L band of backward multi-pumped RFA with the average net gain of 0 dB, 1 dB and 2 dB are designed individually, where the gain ripple is less than 0.64 dB. The combination of SFLA and TVOS enhance the optimization efficiency and improve the performance of RFA with good gain profile.     

利用调度优化模型研究原油调合方案的可行性

针对某沿海炼油企业拟投用原油调合系统以稳定混炼原油的性质,设计方案要求将部分码头罐的操作模式由混储改为单储的情况,采用约束规划与数学规划复合建模,实现了码头至厂区大容积长输线模拟,建立了该企业的原油调度优化模型并研究了设计方案的可行性。针对240h调度周期、10种可加工原油、3艘到港油轮的复杂工况进行了优化计算。结果表明:该设计方案具备可行性,且油轮仍可实现到港即卸油,蒸馏装置混炼油种也可长时间保持稳定;模型的优化计算时间较短(小于15min),且对长输线收付及存油情况模拟准确,复合建模方法满足工业应用要求。     

R2D2: A scalable deep learning toolkit for medical imaging segmentation

Deep learning has gained a significant popularity in recent years thanks to its tremendous success across a wide range of relevant fields of applications, including medical image analysis domain in particular. Although convolutional neural networks (CNNs) based medical applications have been providing powerful solutions and revolutionizing medicine, efficiently training of CNNs models is a tedious and challenging task. It is a computationally intensive process taking long time and rare system resources, which represents a significant hindrance to scientific research progress. In order to address this challenge, we propose in this article, R2D2, a scalable intuitive deep learning toolkit for medical imaging semantic segmentation. To the best of our knowledge, the present work is the first that aims to tackle this issue by offering a novel distributed versions of two well-known and widely used CNN segmentation architectures [ie, fully convolutional network (FCN) and U-Net]. We introduce the design and the core building blocks of R2D2. We further present and analyze its experimental evaluation results on two different concrete medical imaging segmentation use cases. R2D2 achieves up to 17.5× and 10.4× speedup than single-node based training of U-Net and FCN, respectively, with a negligible, though still unexpected segmentation accuracy loss. R2D2 offers not only an empirical evidence and investigates in-depth the latest published works but also it facilitates and significantly reduces the effort required by researchers to quickly prototype and easily discover cutting-edge CNN configurations and architectures.     

Void region restriction for additive manufacturing via a diffusion physics approach

A longstanding challenge in additive manufacturing (AM), the presence of void regions in additively manufactured components, causes two main issues: the enclosing of build material powder in powder bed fusion techniques and limiting tool access in critical post-processing operations to remove sacrificial support structures. As topology optimization has embraced and overcome many of the obstacles of incorporating AM constraints into the underlying numerical optimization statement, there exist few solutions that directly address this fundamental void region issue. By developing computationally efficient and effective solutions to this problem, the integration of these two advanced technologies can be fully realized. Drawing on inspiration from the principles of diffusion physics, a particle diffusion void restriction (PDVR) method is presented in this work that is capable of encouraging the optimization scheme to generate final designs that are fully accessible. Additionally, this method empowers the user to choose the type of post-processing method to clear support material (eg, three-axis or five-axis milling operations, number and orientation of part set-ups) and, therefore, quantify the level of costs associated with the post-processing operation. The PDVR optimization framework is demonstrated on multiple two- and three-dimensional test problems, with physically manufactured examples depicting the real-world benefits this method admits.     

An intelligent EGWO-SCA-CS algorithm for PSS parameter tuning under system uncertainties

This paper proposes a novel hybrid technique called enhanced grey wolf optimization-sine cosine algorithm-cuckoo search (EGWO-SCA-CS) algorithm to improve the electrical power system stability. The proposed method comprises of a popular grey wolf optimization (GWO) in an enhanced and hybrid form. It embraces the well-balanced exploration and exploitation using the cuckoo search (CS) algorithm and enhanced search capability through the sine cosine algorithm (SCA) to elude the stuck to the local optima. The proposed technique is validated with the 23 benchmark functions and compared with state-of-the-art methods. The benchmark functions consist of unimodal, multimodal function from which the best suitability of the proposed technique can be identified. The robustness analysis also presented with the proposed method through boxplot, and a detailed statistical analysis is performed for a set of 30 individual runs. From the inferences gathered from the benchmark functions, the proposed technique is applied to the stability problem of a power system, which is heavily stressed with the nonlinear variation of the load and thereby operating conditions. The dynamics of power system components have been considered for the mathematical model of a multimachine system, and multiobjective function has been framed in tuning the optimal controller parameters. The effectiveness of the proposed algorithm has been assessed by considering two case studies, namely, (i) the optimal controller parameter tuning, and (ii) the coordination of oscillation damping devices in the power system stability enhancement. In the first case study, the power system stabilizer (PSS) is considered as a controller, and a self-clearing three-phase fault is considered as the system uncertainty. In contrast, static synchronous compensator (STATCOM) and PSS are considered as controllers to be coordinated, and perturbation in the system states as uncertainty in the second case study.     

烟气CO2捕集热能梯级利用节能工艺耦合优化

醇胺法捕集CO₂技术是一种较成熟的CO₂捕集技术，具有吸收速度快、脱除效果好等显著优点，但其操作费用高、解吸能耗大。本文以降低醇胺法捕集烟气中CO₂系统再生能耗为出发点，对常规醇胺法捕集CO₂工艺统进行了节能优化研究。在常规工艺流程基础上引入压缩式热泵节能技术，并利用Aspen Plus软件建立了基于压缩式热泵技术的CO₂捕集工艺流程模型。研究了压缩式热泵与机械蒸汽压缩回收（MVR）热泵、分流解吸、分布式换热、级间冷却4种节能工艺耦合，通过模拟计算与优化，结果说明了最佳节能工艺组合为“解吸塔压缩式热泵+贫液MVR热泵+分流解吸+级间冷却”耦合的CO₂捕集工艺流程，当解吸塔顶气体分流比为0.25∶0.75、冷富液分流比为0.05∶0.95、级间冷却器位于吸收塔17块塔板位置、吸收塔输入冷量为-3.0GJ/h时，系统再生能耗最低，为2.533 GJ/tCO₂，相比常规有机胺工艺（再生能耗4.204GJ/tCO₂）节能率39.748%。     

大方坯结晶器电磁搅拌器结构优化

为改善连铸结晶器内钢液的流动状况，采用有限元法和有限容积法研究了大方坯结晶器电磁搅拌器结构对连铸结晶器内电磁场、流场和温度场分布的影响。结果表明，新型结晶器电磁搅拌器在400 A、3 Hz时产生的电磁力比传统结晶器电磁搅拌器在500 A、3 Hz时产生的电磁力更加均匀，对结晶器内钢液的扰动更强，可提高结晶器内的传热效率，有助于初始凝固坯壳和等轴晶的形成，改善方坯质量。     

楼宇型分布式能源系统设备容量和运行策略优化研究

设备容量优化和运行策略优化是分布式能源系统设计，运行的关键问题。为实现分布式能源系统的经济效益，能效水平和环境效益最大化，针对楼宇型分布式能源系统建立了相对普适化的物理模型和数学模型，以粒子群优化算法和线性规划相结合，采用两阶段优化方法计算系统的最优容量配置，并给出运行策略。以某写字楼的分布式能源系统为例，得到最优的系统设备容量和全年逐时运行策略，并采用遍历法验证计算结果的准确性。优化的分布式能源系统与传统供能系统相比，费用年值降低7.79%，年总能耗降低24.18%，污染物排放量减少了62.77 %。     

γ辐照加工用新型工作模体设计研究

本文研究设计了一种用于获取γ辐照装置辐射场和货物剂量场真实分布的新型工作模体。基于真实γ辐照装置参数构建了有效可用的模拟辐射场，提出了新型工作模体的结构和填料设计方法。采用蒙特卡罗方法和随机填充方法（RCS）模拟计算剂量计套管材料与壁厚、填料小球尺寸、空心填料小球尺寸与壁厚、小球填充方式等影响模体剂量学性能的主要因素的分布规律。在满足计算置信度的前提下，参数优化取值范围为：套管采用壁厚为3~5 mm的铝管，对辐射场干扰不超过4.372 23%；填料选择外径为1~4 cm、壁厚为1.1~11.5 mm、材料等效性好、货物密度模拟范围为0.1~0.5 g/cm³的聚丙烯小球，对辐射场干扰不超过9.998 44%；均匀填充和随机填充模式对辐射场干扰基本持平，且均不超过10%。结果表明，当前设计可行有效，投入低、兼顾参数多且量程更宽，适合推广应用。     

基于梯度分布调节策略的Xgboost算法优化

为了解决集成学习模型Xgboost在二分类问题中少数类检出率低的问题，提出了基于梯度分布调节策略的改进的Xgboost算法——LCGHA-Xgboost。首先，通过定义损失贡献（LC）来模拟Xgboost算法中样本个体的损失量；而后，通过定义损失贡献密度（LCD）来衡量Xgboost算法中样本被正确分类的难易程度；最后，提出了梯度分布调节算法LCGHA，依据LCD动态调整样本个体的一阶梯度分布，间接地增大难分样本（主要存在于少数类中）的损失量，减小易分样本（主要存在于多数类中）的损失量，使Xgboost算法偏向对难分样本的学习。实验结果表明,与Xgboost、GBDT、随机森林（Random_Forest）这三大集成学习算法相比，LCGHA-Xgboost算法在多个UCI数据集上的召回率（Recall）值有5.4%~16.7%的提高，AUC值有0.94%~7.41%的提高；在垃圾网页数据集WebSpam-UK2007和DC2010数据集上所提算法的Recall值更是有44.4%~383.3%的提高，AUC值有5.8%~35.6%的提高。LCGHA-Xgboost算法可以有效提高对少数类的分类检出能力，减小少数类的分类错误率。