The Acrtifucial Neural Networks Applied to TemperatureControl for A Variable Refrigerant Flow Air-conditioning System

• Main Authors: Tien-Huai Ma, 馬天懷
• Other Authors: Ming-Huei Chu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/64035720125391800138

碩士 === 東南科技大學 === 機械工程研究所 === 100 === The variable refrigerant flow (VRF) air-conditioning systems are developed for energy saving applications. A variable speed compressor droved by an electrical inverter has been used to VRF air-conditioning systems for energy saving, but the AC driver will generate high heat dissipation and induce high operating temperature especially compressor in low speed operation conditions.This will decrease the energy saving efficiency. Recent years, the air conditioner compressors droved by a Brushless DC motor (BLDC motor) will improve the high heat generation problem in wide operation speed. The DC drivers have better energy saving efficiency than the conventional AC drivers. This study proposes the direct neural control with specified reference model applied to control the room temperature. The conventional PID control is applied to control the speed of compressor, and the back propagation error (BPE) is approximated by a linear combination of error and error’s differential, so that it is not necessary to build the neural emulator to estimate the Jacobin of plant, and the convergent speed can be increased by proposed methods.The dynamic simulations will analyze the time response of room temperature and energy saving efficiency.The simulation resuits show the dynamic response of room temperatnre will follow the response of the reference model with the energy-saving performance.