A parallel computation of power system equations

Streaming SIMD Extensions (SSE) is a unique feature embedded in the Pentium III and P4 classes of microprocessors. By fully exploiting SSE, parallel algorithms can be implemented on a standard personal computer and a theoretical speedup of four can be achieved. In this paper, we demonstrate the implementation of a parallel LU matrix decomposition algorithm for solving power systems network equations with SSE and discuss advantages and disadvantages of this approach.

SSE based parallel solution for power systems network equations

Streaming SIMD Extensions (SSE) is a unique feature embedded in the Pentium III class of microprocessors. By fully exploiting SSE, parallel algorithms can be implemented on a standard personal computer and a theoretical speedup of four can be achieved. In this paper, we demonstrate the implementation of a parallel LU matrix decomposition algorithm for solving power systems network equations with SSE and discuss advantages and disadvantages of this approach.

A ∑GIi/D/1/∞ queue with heterogeneous input/output slot times

In this paper, we present a ∑GIi/D/1/∞ queue with heterogeneous input/output slot times. This queueing model can be regarded as an extension of the ordinary GI/D/1/∞ model. For this ∑GIi/D/1/∞ queue, we assume that several input streams arrive at the system according to different slot times. In other words, there are different slot times for different input/output processes in the queueing model. The queueing model can therefore be used for an ATM multiplexer with heterogeneous input/output link capacities. Several cases of the queueing model are discussed to reflect different relationships among the input/output link capacities of an ATM multiplexer. In the queueing analysis, two approaches: the Markov model and the probability generating function technique, are adopted to develop the queue length distributions observed at different epochs. This model is particularly useful in the performance analysis of ATM multiplexers with heterogeneous input/output link capacities.

A study of the generalised max–min fair rate allocation for ABR control in ATM

In the rate-based flow control for ATM Available Bit Rate service, fairness is an important requirement, i.e. each flow should be allocated a fair share of the available bandwidth in the network. Max–min fairness, which is widely adopted in ATM, is appropriate only when the minimum cell rates (MCRs) of the flows are zero or neglected. Generalised max–min (GMM) fairness extends the principle of the max–min fairness to accommodate MCR. In this paper, we will discuss the formulation of the GMM fair rate allocation, propose a centralised algorithm, analyse its bottleneck structure and develop an efficient distributed explicit rate allocation algorithm to achieve the GMM fairness in an ATM network. The study in this paper addresses certain theoretical and practical issues of the GMM fair rate allocation.

Weighted fairness with MCR guarantee and a general-purpose explicit rate allocation algorithm for ATM ABR service

In this paper, weighted fair rate allocation for ATM available bit rate (ABR) service is discussed with the concern of the minimum cell rate (MCR). Weighted fairness with MCR guarantee has been discussed recently in the literature. In those studies, each ABR virtual connection (VC) is first allocated its MCR, then the remaining available bandwidth is further shared among ABR VCs according to their weights. For the weighted fairness defined in this paper, the bandwidth is first allocated according to each VC's weight; if a VC's weighted share is less than its MCR, it should be allocated its MCR instead of the weighted share. This weighted fairness with MCR guarantee is referred to as extended weighted (EXW) fairness. Certain theoretical issues related to EXW, such as its global solution and bottleneck structure, are first discussed in the paper. A distributed explicit rate allocation algorithm is then proposed to achieve EXW fairness in ATM networks. The algorithm is a general-purpose explicit rate algorithm in the sense that it can realise almost all the fairness principles proposed for ABR so far whilst only minor modifications may be needed.

Fair flow control of ABR service by per-VC virtual queuing

In this paper, a rate-based flow control scheme based upon per-VC virtual queuing is proposed for the Available Bit Rate (ABR) service in ATM. In this scheme, each VC in a shared buffer is assigned a virtual queue, which is a counter. To achieve a specific kind of fairness, an appropriate scheduler is applied to the virtual queues. Each VC's bottleneck rate (fair share) is derived from its virtual cell departure rate. This approach of deriving a VC's fair share is simple and accurate. By controlling each VC with respect to its virtual queue and queue build-up in the shared buffer, network congestion is avoided. The principle of the control scheme is first illustrated by max–min flow control, which is realised by scheduling the virtual queues in round-robin. Further application of the control scheme is demonstrated with the achievement of weighted fairness through weighted round robin scheduling. Simulation results show that with a simple computation, the proposed scheme achieves the desired fairness exactly and controls network congestion effectively.

An enhanced explicit rate algorithm for ABR traffic control in ATM networks

A high performance, low computational complexity rate-based flow control algorithm which can avoid congestion and achieve fairness is important to ATM available bit rate service. The explicit rate allocation algorithm proposed by Kalampoukas et al. is designed to achieve max–min fairness in ATM networks. It has several attractive features, such as a fixed computational complexity of O(1) and the guaranteed convergence to max–min fairness. In this paper, certain drawbacks of the algorithm, such as the severe overload of an outgoing link during transient period and the non-conforming use of the current cell rate field in a resource management cell, have been identified and analysed; a new algorithm which overcomes these drawbacks is proposed. The proposed algorithm simplifies the rate computation as well. Compared with Kalampoukas's algorithm, it has better performance in terms of congestion avoidance and smoothness of rate allocation.

A parallel solution to linear systems

Streaming SIMD Extensions (SSE) is a unique feature embedded in the Pentium III and IV classes of microprocessors. By fully exploiting SSE, parallel algorithms can be implemented on a standard personal computer and a theoretical speedup of four can be achieved. In this paper, we demonstrate the implementation of a parallel LU matrix decomposition algorithm for solving linear systems with SSE and discuss advantages and disadvantages of this approach based on our experimental study.

Performance optimization for parallel processing on a multiple-CPU server

Symmetric multi-processor (SMP) systems, or multiple-CPU servers, are suitable for implementing parallel algorithms because they employ dedicated communication devices to enhance the inter-processor communication bandwidth, so that a better performance can be obtained. However, the cost for a multiple-CPU server is high and therefore, the server is usually shared among many users. The work-load due to other users will certainly affect the performance of the parallel programs so it is desirable to derive a method to optimize parallel programs under different loading conditions. In this paper, we present a simple method, which can be applied in SPMD type parallel programs, to improve the speedup by controlling the number of threads within the programs.

Coasting control of train operation by genetic algorithm

Balancing between the provision of high quality of service and running within a tight budget is one of the biggest challenges for most metro railway operators around the world. Conventionally, one possible approach for the operator to adjust the time schedule is to alter the stop time at stations, if other system constraints, such as traction equipment characteristic, are not taken into account. Yet it is not an effective, flexible and economical method because the run-time of a train simply cannot be extended without limitation, and a balance between run-time and energy consumption has to be maintained. Modification or installation of a new signalling system not only increases the capital cost, but also affects the normal train service. Therefore, in order to procure a more effective, flexible and economical means to improve the quality of service, optimisation of train performance by coasting point identification has become more attractive and popular. However, identifying the necessary starting points for coasting under the constraints of current service conditions is no simple task because train movement is attributed by a large number of factors, most of which are non-linear and inter-dependent. This paper presents an application of genetic algorithms (GA) to search for the appropriate coasting points and investigates the possible improvement on computation time and fitness of genes.

Restarting trains under moving block signalling : an expert system approach

A high peak power demand at substations will result under Moving Block Signalling (MBS) when a dense queue of trains begins to start from a complete stop at the same time in an electrified railway system. This may cause the power supply interruption and in turn affect the train service substantially. In a recent study, measures of Starting Time Delay (STD) and Acceleration Rate Limit (ARL) are the possible approaches to reduce the peak power demand on the supply system under MBS. Nevertheless, there is no well-defined relationship between the two measures and peak power demand reduction (PDR). In order to attain a lower peak demand at substations on different traffic conditions and system requirements, an expert system is one of the possible approaches to procure the appropriate use of peak demand reduction measures. The main objective of this paper is to study the effect of the train re-starting strategies on the power demand at substations and the time delay suffered by the trains with the aid of computer simulation. An expert system is a useful tool to select various adoptions of STD and ARL under different operational conditions and system requirements.

Generation evolution in GA application on railway traffic control

This paper presents a Genetic Algorithms (GA) approach to resolve traffic conflicts at a railway junction. The formulation of the problem for the suitable application of GA will be discussed and three neighborhoods have been proposed for generation evolution. The performance of the GA is evaluated by computer simulation. This study paves the way for more applications of artificial intelligence techniques on a rather conservative industry.

Railway junction conflict resolution by local search method

This study investigates the application of local search methods on the railway junction traffic conflict-resolution problem, with the objective of attaining a quick and reasonable solution. A procedure based on local search relies on finding a better solution than the current one by a search in the neighbourhood of the current one. The structure of neighbourhood is therefore very important to an efficient local search procedure. In this paper, the formulation of the structure of the solution, which is the right-of-way sequence assignment, is first described. Two new neighbourhood definitions are then proposed and the performance of the corresponding local search procedures is evaluated by simulation. It has been shown that they provide similar results but they can be used to handle different traffic conditions and system requirements.

A general-purpose simulator for train operations

Computer simulation has been widely accepted as an essential tool for the analysis of many engineering systems. It is nowadays perceived to be the most readily available and feasible means of evaluating operations in real railway systems. Based on practical experience and theoretical models developed in various applications, this paper describes the design of a general-purpose simulation system for train operations. Its prime objective is to provide a single comprehensive computer-aided engineering tool for most studies on railway operations so that various aspects of the railway systems with different operation characteristics can be investigated and analysed in depth. This system consists of three levels of simulation. The first is a single-train simulator calculating the running time of a train between specific points under different track geometry and traction conditions. The second is a dual-train simulator which is to find the minimum headway between two trains under different movement constraints, such as signalling systems. The third is a whole-system multi-train simulator which carries out process simulation of the real operation of a railway system according to a practical or planned train schedule or headway; and produces an overall evaluation of system performance.

Design and evaluation of a prototype tracking system

In this paper, we present the design and construction of a prototype target tracking system. The experimental set up consists of three main modules for moving the object, detecting the motion of the object and its tracking. The mechanism for moving the object includes an object and two stepper motors and their driving and control circuitry. The detection of the object’s motion is realized by photo switch array. The tracking mechanism consists of a laser beam and two DC servomotors and their associated circuitry. The control algorithm is a standard fuzzy logic controller. The system is designed to operate in two modes in such a way that the role of target and tracker can be interchanged. Experimental results indicate that the fuzzy controller is capable of controlling the system in both modes.