基于全寿命周期成本的高速铁路连续梁桥减隔震方案研究

高烈度地震区大跨连续梁桥一般采用减隔震设计以减小结构地震反应及由此带来的经济损失,而不同的减隔震方案亦决定着结构的抗震性能和初始成本。为了从抗震性能和经济性两方面综合评价连续梁桥减隔震设计的合理性,以一座高烈度区高速铁路连续梁桥为背景,设计多种减隔震布置方案开展桥梁地震易损性分析,并基于全寿命周期成本最小准则探讨了桥梁在全寿命期内各方案的经济性。结果表明：全桥采用双曲面球型支座和黏滞阻尼器配合使用方案在全寿命周期的地震损失最小,但初始成本过高致其经济性较差;全桥布置双曲面球型支座、边墩配置黏滞阻尼器方案满足抗震设防要求且全寿命周期成本最小,为本文减隔震设计中的最优方案。本文提出的桥梁减隔震方案研究方法实现了结构抗震性能与其经济效益的双目标评价,为利益相关者对桥梁减隔震方案的合理决策提供了新思路。     

直喷柴油机喷油系统对燃烧和排放影响的三维模拟研究

直喷柴油机的喷油过程影响燃烧过程、影响发动机性能和污染物排放。为了研究喷油特性(如不同的喷油压力、喷油定时和喷孔直径等)对燃烧性能的影响,利用三维数值模拟软件K IVA-3代码,对喷油过程、燃烧过程和发动机性能进行了数值模拟。计算结果显示了不同喷油特性下的NO和碳烟排放的变化趋势。     

基于特征选择和时间卷积网络的工业控制系统入侵检测

针对工业控制系统流量数据存在特征冗余及深度学习模型对较小规模数据集检测能力较差的问题,提出了一种基于特征选择和时间卷积网络的工业控制系统入侵检测模型。首先,对源域数据集的异常特征和样本不平衡数据进行处理,提高源域数据集质量。其次,针对流量数据的特征冗余,利用信息增益率和主成分分析法构建IGR-PCA特征选择算法,筛选出最优特征子集实现数据降维。然后,根据工业控制系统流量数据的时间序列特性,在较大规模的源域数据集上,利用时间卷积网络（temporal convolution network,TCN）对时间序列数据优异的处理能力,构建源域时间卷积网络预训练模型。最后,在较小规模的目标域数据集上,结合迁移学习（transfer learning,TL）微调策略,获取源域样本数据的流量特征,构建目标域TCN-TL模型。利用公开的工业控制系统数据集进行实验测试,实验结果表明：流量数据经本文特征算法处理后,相较于其他方法,在降低数据维度减少计算量的同时仍具有良好的检测效果;在较大规模的源域数据集和较小规模的目标域数据集上,本文模型均取得了良好的检测效果,在目标域中利用迁移学习微调策略能够学习到源域中的知识,模型检测准确率为99.06%,在训练时间对比中,本文模型训练时间消耗更少,具有更好的泛化能力,能够更好地保护工业控制系统安全。     

Approach and structure of stabilizing control system preventing an extension of synchronism loss

Recently, large-capacity power stations have been built far from load centers with long-distance transmission lines. Therefore, if a fault occurs at the line, generators in the large-capacity power station may lose synchronism from the remaining generators [1, 3]. To prevent an extension of loss of synchronism to other generators, fast shedding of some generators is an effective measure. The following methods have been proposed so far: A method [3] based on energy function approach; and a method [2] which estimates and predicts the relative swing among generators using data measured on-line at each generator and proposes shedding some generators for stabilization. This method requires a large communication network for an exchange of data between power stations. This paper presents a new stabilizing method and a newly developed stabilizing control system which does not require a large communication network because it utilizes generator output, voltage and current measured on-line in the vicinity of the large-capacity power station. Using these measured data, the system estimates and predicts the relative swing between generators in the large-capacity power station and the remaining generators. Finally, the number of generators to be shed for stabilization is decided and shedding is performed. Configuration and characteristics of the developed stabilizing control system are shown. The validity of this method is confirmed by simulation and testing using an artificial power system.     

火电厂灰水回收处理远程监控系统设计

火电厂灰水回收处理远程监控系统由PLC控制器、现场操作站、上位监控计算机及工业以太网构成。系统采用光纤通信和基于工业以太网的信息传送,将基于PLC的工业自动控制和视频监控及防火防盗监控有机的结合起来,实现了灰水回收处理设备的远程自动控制及工业处理过程综合监视。     

Influence of concrete composition on the characterization of fracture surface

An experimental method based on laser triangulation was used to measure the 3-D profile of fracture surface of concrete. The projective-covering method was established to determine the fractal characteristics of fracture surface. Based on the experiment, the influence of composition of concrete on the fractal dimension, including water–binder ratio, maximum aggregate size and aggregate type, was investigated. The variation of fractal dimension with the material composition was also analysed to make further understanding of fracture mechanism of concrete.     

On probabilistic fatigue models for composite materials

The general purpose of this article is to review the main ideas in fatigue analysis of composites in the context of the application of probabilistic methods, both theoretical and computational. That is why most deterministic concepts of composite materials fatigue are summarized together with stochastic approaches. The application of the perturbation based Stochastic Finite Element Method (SFEM) to fatigue analysis of homogeneous and heterogeneous media is shown. Further, homogenization method in its effective modules approach is proposed below for application in fatigue processes modeling of linear elastic periodic random composites. Considering stochastic character of the analysis, the reliability tools appropriate to multicomponent materials are presented together with the specially adopted brittle and ductile fracture criteria.     

Shear testing of thick adhesive layers using the ENF-specimen

An existing experimental method to determine cohesive laws for adhesive layers loaded in shear is further developed. The method is based on differentiation of the energy release rate (ERR) with respect to the adhesive shear deformation at the crack tip. The test geometry used is an ENF-specimen for which the adherends are assumed to deform linearly elastic. The original method is expanded to account for situations where the thickness of the adhesive layer is not negligible as compared to the adherend thickness. To this end, a novel mathematical expression for the energy release rate (ERR) is derived. No assumptions on the form of the cohesive law are made; it is implicitly included in the derivation. The expression for the ERR contains the applied load and the shear deformation of the adhesive layer at the initial position of the crack tip, in addition to geometrical properties and the elastic modulus of the adherend material. Numerical simulations are performed to verify the accuracy of the mathematical expression for the ERR. Preliminary results from experiments performed on an epoxy adhesive are presented. The cohesive law of the adhesive layer is extracted by using a blunted crack tip. Verifying simulations confirm that the local pre-fracture behavior is accurately captured.     

Agent-based monitoring service for management of disruptive events in supply chains

Schedules of supply chains are generated with buffers to absorb the effect of disruptive events that could occur during their execution. Schedules can be systematically repaired through specific modifications within buffers by using appropriate decision models that consider the distributed nature of a supply chain. To this aim, information of disruptive events at occurrence or in advance allows decision models to make better decisions. To detect and predict disruptive events along a schedule execution, a service-oriented monitoring subsystem that uses a reference model for defining monitoring models was proposed. This subsystem offers services for collecting execution data of a schedule and environment data, and assessing them to detect/anticipate disruptive events. Because of the distributed nature and the complexity of these services functionalities, this paper presents an agent-based approach for their implementation. This technology allows dealing with supply chain monitoring by structuring monitoring subsystem functionalities as a set of autonomous entities. These entities are able to perform tailored plans created at execution time to concurrently monitor different schedules. A case study is described to try out the implemented prototype system.     

Robot Cloud: Bridging the power of robotics and cloud computing

Cloud computing is shaping the cyber world and evolves as a key computing and service platform for sharing resources including platforms, software applications and everything in the form of services. This is known “X as a Service”. Although it brings our age unparalleled computing ability and economic benefits, the application of cloud computing is still limited currently in the cyberspace due to the cloud services can only reside in cloud instead of our daily life environment. In fact, there are still a plethora of physical position based on-site service demands that cloud computing could help little due to the “cyber limitation”. In this paper, we aim to integrate the cyber world and the physical world by bringing up the idea of “Robot Cloud” to bridge the power of robotics and cloud computing. To make it possible, we design a novel Robot Cloud stack to support our idea and adopt the service-oriented architecture (SOA) to make the functional modules in the Robot Cloud more flexible, extensible and reusable. Then we develop a prototype of Robot Cloud using the popular Google App Engine to demonstrate our design method. Finally, we conduct the simulation experiments with a “robot show” application scenario to evaluate our scheduling policy and identify the effect of different request distributions and robot center solutions.