The gap between processor speed and memory access performance increases continuingly. In order to improve performance degradation due to memory accesses, we can reduce memory block accesses by improving data locality. In this research, we focus on optimizing data locality in loops. We adopt an iterative greedy algorithm. It iteratively perform loop fission and loop fusion by moving strongly-connected components among loops subject to data dependence constraints. It evaluates data locality change of each move and decide if accept such move. Subsequently loops can be improved gradually forward optimal result.

The protection system plays an important role in the power system to detect the fault, isolate the faulted zone from the unaulted zone and restore the power supply to the healthy part. In order to validate the developed protection system, hardware in the loop (HIL) tests have been performed on several power system test cases using commercial overcurrent and distance relays from two vendors and a Real Time Digital Simulator (RTDS). Two-bus and eight-bus terrestrial power systems (TPS) as well as a 4-bus shipboard power system (SPS) are developed in RSCAD and used as power system test cases. The simulation results of the HIL tests provides the background to design the software overcurrent relay model, which should have the same functionality as that of the hardware commercial overcurrent relay. The test case power system with the developed software overcurrent relay model is simulated on RTDS for the same fault conditions in the software in the loop (SIL) test. The comparison of the HIL and SIL simulation results was used to validate the software overcurrent relay model. This research work is further extended to implement more complex operations like multiple relay operation and relay coordination using the developed overcurrent relay model. The first step is to conduct the multiple relay operation and relay coordination tests on the SPS and TPS using multiple commercial hardware overcurrent relays. The second step is to modify the RSCAD software overcurrent relay model by adding more functionality to produce the matching results as obtained in the first step for the same test cases under similar fault scenarios. In addition to modeling overcurrent relay, a differential relay was also modeled using LabVIEW. The relay model was tested for different types of fault conditions.

No

No description available.

Optimisation of a fed-batch fermentation process typically uses the calculus of variations or Pontryagin's maximum principle to determine an optimal feed rate profile. This often results in a singular control problem and an open loop control structure. The singular feed rate is the optimal feed rate during the singular control period and is used to control the substrate concentration in the fermenter at an optimal level. This approach is supported by biological knowledge that biochemical reaction rates are controlled by the environmental conditions in the fermenter; in this case, the substrate concentration. Since an accurate neural net-based on-line estimation of the substrate concentration has recently become available and is currently employed in industry, we are therefore able to propose a method which makes use of this estimation. The proposed method divides the optimisation problem into two parts. First, an optimal substrate concentration profile which governs the biochemical reactions in the fermentation process is determined. Then a controller is designed to track the obtained optimal profile. Since the proposed method determines the optimal substrate concentration profile, the singular control problem is therefore avoided because the substrate concentration appears nonlinearly in the system equations. Also, the process is then operated in closed loop control of the substrate concentration. The proposed method is then called "closed loop optimal control". The proposed closed loop optimal control method is then compared with the open loop optimal feed rate profile method. The comparison simulations from both primary and secondary metabolite production processes show that both methods give similar performance in a case of perfect model while the closed loop optimal control provides better performance than the open loop method in a case of plant/model mismatch. The better performance of the closed loop optimal control is due to an ability to compensate for the modelling errors using feedback.

Dissertação (mestrado) - Universidade Federal de Santa Catarina, Centro Tecnológico. Programa de Pós-Graduação em Ciência da Computação. / Made available in DSpace on 2012-10-20T09:40:08Z (GMT). No. of bitstreams: 1 188157.pdf: 431174 bytes, checksum: 32f4521af19e9fdd365fc47b62c259a6 (MD5) / Este trabalho tem como objetivo resolver um problema clássico da Síntese de Alto Nível através de uma abordagem orientada à exploração de soluções alternativas. O problema consiste no escalonamento de operações de um dado algoritmo sob restrição de recursos físicos de forma que cada operação é executada respeitando a ordem de precedência imposta pelo algoritmo. Para abordar o problema acima, utilizou-se as técnicas de Loop Pipelining e Loop Unrolling, onde operações de diferentes iterações podem ser executadas em um mesmo estado. Estas técnicas, por exporem mais paralelismo, permitem uma melhor utilização dos recursos. Este trabalho descreve a abordagem proposta, a modelagem que a ampara e a implementação de ferramentas que a suportam (escalonador e paralelizador). São apresentados resultados experimentais obtidos a partir de exemplos clássicos da literatura.

No description available.

No description available.

No description available.

No description available.