Aluminum Conductor Steel Reinforced (ACSR) conductors have been applied in electric power transmission and distribution for over 80 years. Research about ACSR includes its possible properties in electrical, mechanical, and thermal areas. We postulate that these properties predict certain behaviours in power transmission and distribution lines. Four models have been established by various authors for determining conductor behaviour. They are the electromagnetic, mechanical, radial conduction, and steady-state thermal models. These models were developed independently,. Although they can be used in their fields individually, there are no experimental studies verifying a combined model. Also, using them separately does not yield the required information for determining conductor performance. The unified model connects these models probabilistically by considering power system loads and meteorological factors. Based on the unified model and its modules, it is possible to use mathematical tools to optimize the ACSR design and analyze conductor characteristics when conductor parameters are changed,. Evolutionary Computation is an optimization process simulating natural evolution on the computer. Instances based on evolutionary principles are Evolutionary Algorithms that historically include Genetic Algorithms, Evolution Strategies, and Evolutionary Programming. Genetic Algorithms are used in the optimization of multi-dimensional problems in this work. Evolutionary Algorithms are empirically robust in finding near-optimal solutions to complex problems through parallel searches of solution space. Evolution Computations imitates natural evolution and genetic variation, and lays the mathematical foundation for problems in which many inputs are variable. Especially, Genetic Algorithms are extensively applied in engineering to solve problems without satisfying gradient descent, deterministic hill climbing, or purely random search. This project introduces the Evolutionary Algorithms and applies the Genetic Algorithms to the unified models. The problem solved by applying Genetic Algorithms to optimize the unified model is to select optimum multi-dimensional input parameters for the model. This provides an effective way to find conductor size for optimizing conductor design. The results give the variation of electrical, thermal, and mechanical characteristics according to conductor loss changes and predict the variation ranges of electric and magnetic fields of three-layer conductors within ASTM standards. The procedure to apply Genetic Algorithms to optimize ACSR design is unique to the problem. Objective functions are found according to the characteristics of each model. The results are put into the unified model. Comparing results gives rules to change geometrical parameters of ACSR to reach minimal Joule loss. / Thesis / Master of Engineering (ME)
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/23214 |
| Date | 01 1900 |
| Creators | Liu, Hongyan |
| Contributors | Findlay, Raymond, Electrical and Computer Engineering |
| Source Sets | McMaster University |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0018 seconds