基于多通道的ICU脑血管疾病死亡风险预测模型

死亡风险预测指根据病人临床体征监测数据来预测未来一段时间的死亡风险。对于ICU病患，通过死亡风险预测可以有针对性地对病人做出临床诊断，以及合理安排有限的医疗资源。基于临床使用的MEWS和Glasgow昏迷评分量表，针对ICU病人临床监测的17项生理参数，提出一种基于多通道的ICU脑血管疾病死亡风险预测模型。引入多通道概念应用于BiLSTM模型，用于突出每个生理参数对死亡风险预测的作用。采用Attention机制用于提高模型预测精度。实验数据来自MIMIC [Ⅲ]数据库，从中提取3?080位脑血管疾病患者的16?260条记录用于此次研究，除了六组超参数实验之外，将所提模型与LSTM、Multichannel-BiLSTM、逻辑回归（logistic regression）和支持向量机（support vector machine, SVM）四种模型进行了对比分析，准确率Accuracy、灵敏度Sensitive、特异性Specificity、AUC-ROC和AUC-PRC作为评价指标，实验结果表明，所提模型性能优于其他模型，AUC值达到94.3%。     

Automatic design for shop scheduling strategies based on hyper-heuristics: A systematic review

Against the background of smart manufacturing and Industry 4.0, how to achieve real-time scheduling has become a problem to be solved. In this regard, automatic design for shop scheduling based on hyper-heuristics has been widely studied, and a number of reviews and scheduling algorithms have been presented. Few studies, however, have specifically discussed the technical points involved in algorithm development. This study, therefore, constructs a general framework for automatic design for shop scheduling strategies based on hyper-heuristics, and various state-of-the-art technical points in the development process are summarized. First, we summarize the existing types of shop scheduling strategies and classify them using a new classification method. Second, we summarize an automatic design algorithm for shop scheduling. Then, we investigate surrogate-assisted methods that are popular in the current algorithm field. Finally, current problems and challenges are discussed, and potential directions for future research are proposed.     

An improved feedrate scheduling method for NURBS interpolation in five-axis machining

Five-axis machining plays an important role in manufacturing by dint of its high efficiency and accuracy. While two rotation axes benefit the flexibility of machining, it also brings limitations and challenges. In order to further balance machining precision and efficiency, an improved feedrate scheduling method is presented considering geometric error and kinematic constraints for the Non Uniform Rational B-Spline (NURBS) interpolation in five-axis machining. A simplification method is proposed to calculate the geometric error which describes the deviation between the ideal tool path and the real tool path induced by the non-linear movement. A linear relation between geometric error and feedrate is built to limit the feedrate. The constraints determined by single axis kinematic performance and tangential kinematic performance are also considered. Under these constraints, a constrained feedrate profile is determined. Aiming to get more constant feedrate in the difficult-to-machine areas, this work proposes a scheduling method which combines morphological filtering and S-shape acceleration/deceleration (acc/dec) mode. Simulations and experiments are performed to compare the proposed feedrate scheduling method with two previous feedrate scheduling method and the results prove that the proposed feedrate scheduling method is reliable and effective.     

改进Elman神经网络在短期热负荷预测中的应用

热电厂的短期热负荷预测在城市集中供暖中起着至关重要的作用,直接影响热电厂的经济效益和热能利用率。电厂的短期热负荷一般采用神经网络预测模型进行预测,而BP神经网络应用最为广泛。Elman神经网络算法在BP神经网络基础上加入了承接层,作为一步延时算子,实现记忆能力,使系统具备适应时变能力,增强系统全局稳定性。但Elman神经网络算法模型的构造依然需要大量样本的支撑,而且输入层的变量多,导致预测时间依然很长,收敛速度慢。该文在Elman神经网络预测前,进行了相关系数预处理和对样本中异常值的平均化预处理,通过数据归一化运算,使Elman神经网络输入层变量大幅减少。仿真实验表明,改进的Elman神经网络算法使预测模型快速寻优,减少预测时间的同时明显提高预测精度。     

Hierarchical multi-reservoir optimization modeling for real-world complexity with application to the Three Gorges system

High dimensionality in real-world multi-reservoir systems greatly hinders the application and popularity of evolutionary algorithms, especially for systems with heterogeneous units. An efficient hierarchical optimization framework is presented for search space reduction, determining the best water distributions, not only between cascade reservoirs, but also among different types of hydropower units. The framework is applied to the Three Gorges Project (TGP) system and the results demonstrate that the difficulties of multi-reservoir optimization caused by high dimensionality can be effectively solved by the proposed hierarchical method. For the day studied, power output could be increased by 6.79 GWh using an optimal decision with the same amount of water actually used; while the same amount of power could be generated with 2.59 × 10⁷ m³ less water compared to the historical policy. The methodology proposed is general in that it can be used for other reservoir systems and other types of heterogeneous unit generators.     

巨型水电站孤岛试验期间AVC电压波动的解决

本文首先分析国内典型的AVC的一般配置、控制策略和控制原理,针对巨型水电站孤岛试验期间的AVC控制的复杂工况进行阐述,对出现的AVC电压波动的情况进行分析,并对采用的解决方案进行介绍,为今后类似电站的孤岛运行期间的AVC电压控制提供一些参考建议。     

Lagrangian relaxation hybrid with evolutionary algorithm for short-term generation scheduling

Short-term generation scheduling is an important function in daily operational planning of power systems. It is defined as optimal scheduling of power generators over a scheduling period while respecting various generator constraints and system constraints. Objective of the problem includes costs associated with energy production, start-up cost and shut-down cost along with profits. The resulting problem is a large scale nonlinear mixed-integer optimization problem for which there is no exact solution technique available. The solution to the problem can be obtained only by complete enumeration, often at the cost of a prohibitively computation time requirement for realistic power systems. This paper presents a hybrid algorithm which combines Lagrangian Relaxation (LR) together with Evolutionary Algorithm (EA) to solve the problem in cooperative and competitive energy environments. Simulation studies were carried out on different systems containing various numbers of units. The outcomes from different algorithms are compared with that from the proposed hybrid algorithm and the advantages of the proposed algorithm are briefly discussed.     

基于组合双向拍卖的云资源调度方法

针对跨数据中心的资源调度问题，提出了一种基于组合双向拍卖（PCDA）的资源调度方案。首先，将云资源拍卖分为三个部分：云用户代理报价、云资源提供商要价、拍卖代理组织拍卖；其次，在定义用户的优先级及任务紧迫度的基础上，在拍卖过程中估算每一个工作发生的服务等级协议（SLA）违规并以此计算云提供商的收益，同时每轮竞拍允许成交多项交易；最终达到根据用户等级合理分配云资源调度的效果。仿真实验结果表明该算法保证了竞拍成功率，与传统一次拍卖成交一项的组合双向拍卖方案相比，PCDA在竞拍时间段产生的能耗降低了35.00%，拍卖云提供商的利润提高了约38.84%。     

Supply chain planning and scheduling integration using Lagrangian decomposition in a knowledge management environment

The integration of planning and scheduling decisions in rigorous mathematical models usually results in large scale problems. In order to tackle the problem complexity, decomposition techniques based on duality and information flows between a master and a set of subproblems are widely applied. In this sense, ontologies improve information sharing and communication in enterprises and can even represent holistic mathematical models facilitating the use of analytic tools and providing higher flexibility for model building. In this work, we exploit this ontologies’ capability to address the optimal integration of planning and scheduling using a Lagrangian decomposition approach. Scheduling/planning sub-problems are created for each facility/supply chain entity and their dual solution information is shared by means of the ontological framework. Two case studies based on a STN representation of supply chain planning and scheduling models are presented to emphasize the advantages and limitations of the proposed approach.     

Multi-resource scheduling for FPGA systems

In modern cloud data centers, reconfigurable devices (FPGAs) are used as an alternative to Graphics Processing Units to accelerate data-intensive computations (e.g., machine learning, image and signal processing). Currently, FPGAs are configured to execute fixed workloads, repeatedly over long periods of time. This conflicts with the needs, proper to cloud computing, to flexibly allocate different workloads and to offer the use of physical devices to multiple users. This raises the need for novel, efficient FPGA scheduling algorithms that can decide execution orders close to the optimum in a short time. In this context, we propose a novel scheduling heuristic where groups of tasks that execute together are interposed by hardware reconfigurations. Our contribution is based on gathering tasks around a high-latency task that hides the latency of tasks, within the same group, that run in parallel and have shorter latencies. We evaluated our solution on a benchmark of 37500 random workloads, synthesized from realistic designs (i.e., topology, resource occupancy). For this testbench, on average, our heuristic produces optimum makespan solutions in 47.4% of the cases. It produces acceptable solutions for moderately constrained systems (i.e., the deadline falls within 10% of the optimum makespan) in 90.1% of the cases.