子波神经网络在火灾探测中的应用

The Application of a Wavelet Neural Network on Fire Detection

论述了用于信号识别的子波神经网络的结构和算法，并根据火灾传感器信号处理的特点，提出了将其用于火灾探测的方法，在子波神经网络中采用了子波函数和共轭梯度优化方法．实验表明，子波神经网络对火灾信号具有很好的学习和探测能力，与ＢＰ神经网络火灾探测方法相比，所提出的方法能够更快和更准确地探测各种标准实验火．

The structure and algorithm of a wavelet neural network are discussed. The wavelet neural network is used in fire detection depending on the features of fire signals. Wavelet function and conjugate gradient algorithm are applied in the wavelet network. The experimental results indicate that the wavelet neural network has a good capability to learn and detect fire signals. It is faster and more accurate to detect fires.