论文部分内容阅读
冷库温度预测优化控制在果蔬冷藏方面的应用尚有许多不足之处。主要问题之一是不能通过简练有效的计算完成制冷系统的在线优化控制计算。 Ｒ Ｂ Ｆ神经网络有极强的非线性映照能力和良好的插补性能， 且训练速度快。该文提出使用二级 Ｒ Ｂ Ｆ神经网络， 并合理地综合利用状态量以往的测量值和预测的未来值来实现库温的在线预测优化控制。将该方法用于某冷库库温控制系统， 取得了满意的结果。
Optimization of cold storage temperature control in the application of fruit and vegetable refrigeration there are still many deficiencies. One of the main problems is that the online optimization control calculation of the refrigeration system can not be completed by concise and effective calculation. RBF neural network has strong nonlinear mapping ability and good interpolation performance, and training speed. In this paper, a two-level RBF neural network is proposed to optimize the on-line forecasting of the reservoir temperature by comprehensively using the previous measurements of the state quantities and the predicted future values. The method is applied to a cold storage temperature control system, and achieved satisfactory results.