Integrated evolutionary neural network approach with distributed computing for congestion management

Electric supply industry is facing deregulation all over the world. Under deregulated power supply scenario, power transmission congestion has become more intensified and recurrent, as compared to conventional regulated power system. Congestion may lead to violation of voltage or transmission capacity limits, thus threatens the power system security and reliability. Also the growing congestion may lead to unanticipated divergent electricity pricing. Owing to these facts congestion management has become a crucial issue in the deregulated power system scenario.Fast and precise prediction of nodal congestion prices in real time deregulated/spot power market may enable market participants and system operators to keep pace with the congestion by taking preventive measures like transaction rescheduling, bids (both for supplying and consuming electricity) modification, regulated dispatch of electric power, etc. This paper proposes an integrated evolutionary neural network (ENN) approach to predict nodal congestion prices (NCPs) for congestion management in spot power market. Distributed computing is employed to tackle the heterogeneity of the data in the prediction of NCP values. Developed ENNs have been trained and tested under distributed computing environment, using a message passing paradigm. Proposed hybrid approach for NCP prediction is demonstrated on a 6-bus test power system with and without distributed computing. The proposed approach not only demonstrated the computing efficiency of the developed ENN model over the conventional optimal power flow method but also shows the time saving aspect of distributed computing.     

并行稀疏矩阵与向量乘的负载平衡和通信优化

本文考虑了在工作站机群上实现大型稀疏矩阵和向量乘的负载平衡。提出了一个快速负载平衡和有效的消息传递技术相结合的方法,来缓解计算和节点间通信。该方法的性能在工作站机群上进行测试,可获得良好结果;并且,通过I/O延迟隐藏和整体负载平衡使I/O开销能有效地分摊。     

AN ARGUMENT-BASED APPROACH TO NONMONOTONIC REASONING*

We define an argument system to be a pair consisting of a set of inference rules and a set of completeness conditions. Inference rules are used to build arguments. Completeness conditions are used to define argument structures, which are sets of arguments supporting belief sets. We reformulate Reiter's default logic as special argument systems. This enables us, among other things, to apply the negation-as-failure rule to general default theories. We also speculate on some other potential uses of our argument systems.     

PVM与网络并行计算

松散耦合的异构型并行处理系统近年来又有较大的发展，这种只能支持粗粒度并行的计算环境，由于通讯硬件的更新和软件环境的开发，其性能已经能够与某些ＭＰＰ机相媲美，因而形成并行处理领域一个强有力的分支－网络计算，本文围绕ＰＶＭ，讨论网络计算的软件环境，概述其特色和发展，并与其它基于消息传递的软件环境相比较，最后预测ＰＶＭ未来可能面临的问题和发展方向。     

一类IMBq方程的第二初边值问题

本文应用Galerkin方法和Leray-Schauder不动点原理,建立了一类IMAq方程的第二初边值问题解的存在唯一性。     

一种优化的LTPB网络带宽分配方法及实时性能分析

研究LTPB（线性令牌传递总线）通信网络中带宽分配方法对网络实时性能的影响。提出了一种优化的带宽分配方法和基于“最差情形下可达负载率”的网络实时性能分析方法，证明了该带宽分配方法优于所有传统的分配方法，最后举例说明了结论的正确性。     

Compositional priority specification in real-time distributed systems

In this paper, we develop a compositional denotational semantics for prioritized real-time distributed programming languages. One of the interesting features is that it extends the existing compositional theory proposed by Koymanset al (1988) for prioritized real-time languages preserving the compositionality of the semantics. The language permits users to define situations in which an action has priority over another action without the requirement of preassigning priorities to actions for partially ordering the alphabet of actions. These features are part of the languages such as Ada designed specifically keeping in view the needs of real-time embedded systems. Further, the approach does not have the restriction of other approaches such as prioritized internal moves can pre-empt unprioritized actions etc. Our notion of priority in the environment is based on the intuition that a low priority action can proceed only if the high priority action cannot proceed due to lack of the handshaking partner at that point of execution. In other words, if some action is possible corresponding to that environment at some point of execution then the action takes place without unnecessary waiting. The proposed semantic theory provides a clear distinction between the semantic model and the execution model — this has enabled us to fully ensure that there is no unnecessary waiting.     

Load-Balanced Sparse Matrix–Vector Multiplication on Parallel Computers

We considered the load-balanced multiplication of a large sparse matrix with a large sequence of vectors on parallel computers. We propose a method that combines fast load-balancing with efficient message-passing techniques to alleviate computational and inter-node communications challenges. The performance of the proposed method was evaluated on benchmark as well as on synthetically generated matrices and compared with the current work. It is shown that, by using our approach, a tangible improvement over prior work can be obtained, particularly for very sparse and skewed matrices. Moreover, it is also shown that I/O overhead for this problem can be efficiently amortized through I/O latency hiding and overall load-balancing.     

大采高工作面过断层围岩控制分析

以大采高工作面过断层围岩控制为对象开展探究.结合具体工程实际,在利用数值模拟手段分析围岩破坏特征的基础上,对围岩控制方式做出探究,并对其应用效果进行分析.结果表明,所设计方案实现了对围岩的有效控制,提升了工作面过断层的安全性.     

自适应SW-ADI方法解反应扩散方程的并行实现

根据解反应扩散方程的自适应样条小波-交替方向(SW-ADI)方法,使用MPI、OpenMP两种并行编程模式,对串行程序进行了直接并行化,并在上海大学的高性能计算机自强2000上分别用MPI和OpenMP实现了对方程的求解。对运算结果进行了分析并给出了与串行程序相比较的并行加速比。