论文部分内容阅读
利用MATLAB工具箱函数分别建立了起重机钢丝绳断丝数目检测的BP神经网络和RBF神经网络模型,并对2种模型的结构、预测精度和训练过程进行了对比研究。结果表明,在一定的样本集和训练条件下,BP和RBF神经网络均能对钢丝绳的断丝数目进行很好预测,可以解决峰值、阀值波宽、小波能量和波形下面积对钢丝绳断丝数目的非线性映射关系,能够满足工程预测的需要。但RBF神经网络较BP神经网络在迭代次数、收敛速度和网络结构方面更具优势,因此其预测能力和泛化能力都优于BP神经网络。
The BP neural network and the RBF neural network model of detecting the broken wire number of the crane wire rope were established respectively by using the MATLAB toolbox function. The structure, prediction accuracy and training process of the two models were compared. The results show that both BP and RBF neural networks can predict the number of broken wire ropes under certain sample sets and training conditions, which can solve the problems of peak value, threshold wave width, wavelet energy and the area under the waveform of broken wire rope The number of non-linear mapping, to meet the needs of project prediction. However, the RBF neural network is more advantageous than the BP neural network in terms of the number of iterations, the convergence speed and the network structure, and therefore its prediction capability and generalization ability are superior to the BP neural network.