论文部分内容阅读
研究不同水分处理所得的作物实际产量，建立前馈神经网络模型，得到不同土壤含水量时的旱作水稻预测经济产量，模型通过学习，使误差减小到0.1。
The actual crop yield of different moisture treatments was studied, the feedforward neural network model was established, and the predicted economic yield of dry-crop rice with different soil water contents was obtained. The model reduced the error to 0.1 by studying.