基于神经网络预测节能中央空调控制策略

由于传统中央空调具有大滞后、大惯性、非线性特性，造成常规控制方法下系统供给的能量与负载所需能量不匹配，使得中央空调与使用环境能量供求不平衡，浪费了大量的电能.针对中央空调的控制特性，提出了一种基于神经网络技术的预测控制方法，将Elman神经网络预测器和神经网络控制器有机结合，通过预测未来能量需求，实时调节控制策略，使系统所需能量和空调输出能量达到匹配.采用Elman神经网络预测器和神经网络控制器有机结合的控制方法，使系统具有良好的动态性能和稳态性能，节能效果显著.采用神经网络预测型节能中央空调，可有效控制中央空调与使用环境能量供求的关系，为降低智能建筑能耗提供了可靠的保障.

Control strategy for predicting energy-saving central air-conditioning system based on neural network

Traditional central air conditioning system has large hysteretic, inertial and non linear characteristics, 〖JP〗which may cause the energy mismatch between system power supply and load energy demand in regular control method and result in the unbalance of energy supply and demand between central air conditioning system and application environment. And thus, a great amount of electric energy is wasted. Aiming at the control performance of central air conditioning system, a neural network based prediction control method to combine Elman neural network predictor with neural network controller was proposed. The control strategy can be adjusted real time in order to match the system energy demand and air conditioning energy output through predicting future energy demand. With adopting the control method of combining Elman neural network predictor and neural network controller, the system exhibits good performances of dynamics and stability as well as obvious energy saving effect. The predictive energy saving central air conditioning system using neural network can effectively control the relationship between the central air conditioning system and application environment energy supply, and provide a reliable guarantee for reducing energy consumption of intelligent buildings.〖