基于人工神经网络的混响时间预测方法

A method of forecast of the reverberation time based on artificial neural network

在分析混响时间的各种影响因素的基础上，运用人工神经网络的ＢＰ算法，建立了厅堂混响时间的预涮模型，该方法具有速度快、使用方便、精度高、可考虑的因素多等特点，具有一定的工程实用价值，有待实践的考验并逐步提高完善．

Parameters influencing the reverberation time in rooms and halls are an- alyzed, and a forecast model of reverberation time is established based on ANN. The model is shown to be speedy, convenient, sufficiently accurate and practical in engineer- ing design. Tests in practices will help to check and improve its worth.