【摘要】 针对非线性预测问题,提出了小波神经网络算法,建立了小波神经网络的趋势预测模型,通过对钢丝绳磨损度的时间序列预测,实现了故障预报。实践表明:小波网络具有更快的收敛速度和更高的预报精度,仿真结果与实测数据相比最大相对误差为4.23%,预报精度满足要求。更多还原

【Abstract】 Aiming at nonlinear forecast problem, the paper offered wavelet neural network algorithm, established trend forecasting model for wavelet neural network. By forecasting time series of abrasion de- gree of steel rope, malfunction forecast was realized. Practice showed that wavelet network possessed faster convergence rate and higher forecasting accuracy, max. relative error between simulation result and testing data was 4.23%, the forecasting accuracy met demand.更多还原