论文部分内容阅读
建立了一个自适应性神经网络模型，它在Ｂ－Ｐ网络模型基础上，对网络的自身结构及学习规则进行了动态优化。网络能自组织和自学习自己的结构，即在学习过程中，网络可根据具体问题自动调整本身的结构，从而使结构达到最优。学习速度具有动态调节功能，根据每次学习时得到的误差不同，网络不断调整学习速率，从而在不引起系统振荡的情况下加速了收敛过程。在此基础上，对我国农机总动力需求进行了预测，预测结果和实际结果有很好的一致性。
An adaptive neural network model is established, which dynamically optimizes the network structure and learning rules based on the B-P network model. The network can self-organize and self-learn its own structure, that is, in the learning process, the network can automatically adjust its structure according to specific problems so that the structure can be optimized. The learning speed has the function of dynamic adjustment. According to the error obtained in each learning, the network adjusts the learning rate continuously so as to accelerate the convergence process without causing system oscillation. On this basis, the total power demand of agricultural machinery in our country is predicted, and the predicted results are in good agreement with the actual results.