Scheduling thermal generation units plays an important role in power system economic operations. Each day power generating units have to be selected to realize a reliable production of electric energy with the fewest fuel costs. This paper presents a new branch-and-bound algorithm for the unit scheduling problem. An efficient branching method based on the 'heap' data structure and a simple intuitive bounding rule are proposed.Computational results indicate that the presented approach locates the optimum schedule in less time than many existing techniques. INTRODUCT 1ONTo economically commit available thermal generating units under constraints is still the subject under intensive research [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. Costs of unit commitment are incurred both from starting up generating units and from dispatching them. The overall cost over the study period is to be minimized and the variables to be determined are the hourly statuses of generating units, namely ON(committed) or OFF(unco"itted).Constraints in the unit commitment problem can naturally be divided into two categories : the "coupling" constraints and the "local" constraints. The "coupling" constraints reflect the sum of the power generated by all units. This whole generation must meet the system demand, including network transmission losses and spinning reserves. The "local" constraints deal with each thermal unit individually. The first is the upper and lower limits on generated power, if the unit is ON. And there are technical constraints, together with the requirement of limiting equipment fatigue, leading to the imposition of minimum up times and nunimum down times.Optimal unit commitment is the cheapest production policy that se!ects the most economical start-up and shut-down times for each unit such that all constraints are satisfied for the study period. Many attempts have been made to solve this problem. References [l-41 propose solution methods based on unit priority lists and heuristic rules to improve a given feasible solution. In references [5, 92 SM 483-8 EC A paper recommended and approved by the IEEE Energy Development and Power Generation Committee of the IEEE Power Engineering society for presentation at the IEEE/Shun-Chung Wang 61 a solution methodology based on duality, Lagrangian relaxation, and nondifferentiable optimization is described. Merlin and Sandrin [7] propose another Lagrangian relaxation method: this is a decomposition method using Lagrange multipliers which provides a new solution to the conventional problem of thermal unit commitment. Dillion et al. [8] and Pang et al. [9]formulate the unit commitment problem as a linear mixed-integer programming problem and then use standard integer programming algorithms to solve for the commitment schedule. Dynamic programming has long been used by many authors to solve the unit commitment problem [lo-151. In order to keep the problem tractable, these methods assume some partial ordering which specifies which set of units can be used for dispatch at a given ...

This article describes the successful experiences of National Taiwan University Hospital (NTUH) in moving from IBM Mainframe to connected networking computer systems. We use multi-tier architecture and HL7 standard to implement our new outpatient Hospital Information System (HIS). The NTUH HIS is a complex environment with several operating systems, databases, and information systems. We adopt ServiceOriented Architecture (SOA) to reduce the complex relations between systems and solve data consistency problems among databases. We also show that the distributed architecture can provide us stable and reasonable system performances. Our main contribution is proving that the distributed environment with HL7 standard and SOA can sustain in a highly demanding environment.

No abstract

No abstract