摘要
离心压缩机性能的预测在其研究和应用领域有着十分重要的作用。传统的性能预测方法是根据有限的实验或者某一模型机的已有数据进行换算，大部分预测结果仅在数据采样点附近难确，远离这些点的预测结果不很准确。本文应用神经网络技术进行离心压缩机性能预测。该方法虽基于有限的实验数据，但可以在大范围内有效预测离心压缩机性能，方法经过对两个典型性能预测，数值模拟显示了预测性能的有效性和可行性。
The Performance predictions of centrifugal compressors take a very important part in the field of turbomachinery research and application. Traditional methods for the Performance Predictions are based on computations from limited experiment measurements of the machine or limited data obtained from its models. The computed results are incorrect. The paper proposed an alternative method, neural network technique. The method has been applied to some typical performance prediction tasks. The numerical results are quite satisfactory.
关键词
神经网络
离心压缩机
性能预测
neural network
centrifugal compressor
performance prediction
