基于遗传对角回归神经网络的农机总动力预测

Forecasting of Agricultural Machine Power Based on Genetic Diagonal Recurrent Neural Network

利用遗传算法强全局随机搜索特点,结合DRNN神经网络对非线性数据具有鲁棒性和自学习能力的优点,通过将历年农机总动力数据作为时间序列进行分析,建立DRNN神经网络预测模型对农机总量进行预测。本文采用遗传算法对DRNN神经网络进行训练,可克服基于梯度算法的神经训练算法的缺点,收敛速度快,可达到全局最优。通过与校验用数据的比较证明本文建立的预测模型具有较高的精度。

Agricultural machine power is forecast by diagonal recurrent neural network （DRNN） forecasting model ,which is combined the characteristic of global random search of genetic algorithm with the virtue of robust and self-study for nonlineardata of DRNN and agricultural machine power data in history is analyzed by time series. The DRNN is trained by genetic algorithm in this paper, which result indicates the method has rapid speed and reach global best of all, and can avoid the defect of neural network training algorithm based on grads algorithm. The forecasting results is compared with verify data, that prove the forecasting model in this paper has higher precision.