The electricity savings by using probabilistic neural network for room air-conditioners

In this paper, an effective tool is presents to perform the electrical energy management (EEM) of room air-conditioners. A practical air-conditioner is installed to measure the on-line operating information, which includes temperature, humidity, and power consumption in a room. Based on the theory of enthalpy, the training data for probabilistic neural network (PNN) is derived to decide the status of the electromagnetic valves and the operating frequency of compressor. The PNN can be fast learning and recalling process, no iteration for weight regulations in learning process, and adaptability for architecture changes. Testing results show that it provides a good tool to make better control strategies for achieving the EEM of room air-conditioners.