Optimizing for Large Time Delay Systems by BP Neural Network and Evolutionary Algorithm Improving

BP artificial neural network is a non-feedback network. This paper utilizes the initial weights of neural network to choose controller performance. Simultaneously according to the characteristics that process of central air-conditioning energy saving control is the system of multi-parameter and nonlinear time-varying complexity, we analysis and study its algorithm and system architecture. The experimental results   demonstrate   that new control system gets better results and energy saving.