制冷涡旋压缩机热力性能模型适用性研究

摘要

对4种模型的内插、外推性能进行了对比分析。结果表明：遗传算法优化的BP神经网络模型对压缩机热力性能有较好的预测能力，而且有较强的泛化能力。%Four models' interpolation performance and extrapolation performance were compared and analyzed. The results showed that BP neural network model based on genetic algorithms to optimize has a good ability to predict the thermal performance of the com-pressor, but there is strong generalization ability.