The Study on the Optimization of Heat-Sink Parameters for HB-LED with ANN
| Autor: | Chien Ting Hsu, 許建霆 |
|---|---|
| Rok vydání: | 2011 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 100 In this study, the performance of the heat sink that dissipates heat from a surface mount package with four 1 W HB-LEDs (high brightness light emitting diodes) is investigated. Since the surface mount structure of the LEDs minimized the thermal resistance from the chip to the heat sink that we only need to focus on the optimization of the heat sink performance. An experimentally verified numerical model is established for the simulations of the junction temperature, Tj, of HB-LED lightings. The length of the fins on the heat sink should be kept at a minimum value of 15 mm. Then the simulation results of the heat sinks with various thicknesses and numbers of fin are used to construct a feedforward type ANN (artificial neural network). The best structure of the ANN has only one hidden layer with 6 neurons in the layer. The constructed ANN is then used to find the relationship between the minimum junction temperature, Tj,min, and the depth of the heat sink, H, and between the critical number of fins, Nc, and H. One can use these relationships as a tool to design an optimal heat sink if Tj,min is specified. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|