In order to reduce the influences of sensor's random errors on three point support small agricultural machinery centroid position measuring system,a new test data processing method was studied with fewer times and small-angle test operations.A construction method of gray BP neural network model was proposed,which integrated the gray GM(1,1) model and the BP neural network model.And the prediction model was verified with an actual small agricultural machinery centroid measuring platform.The results show that the construction method and forecasting model are credible,and the relative error is as 0.759% which is better than original error.

为减小传感器的随机误差对三点支撑小型农业机质心测试系统质心高度数据的影响,同时提高质心高度的测试效率和安全性,对该类测试系统在有限次、小角度条件下测试获得质心高度的数据处理方法进行了研究。提出灰色GM(1,1)模型与BP神经网络相结合的灰色BP神经网络模型的构建方法,

采用该方法构建的预测模型,对质心高度测试获得的数据进行处理可以获得较高精度的质心高度数据,并搭建了小型试验台对预测模型进行验证。结果表明,采用该方法构建的质心高度预测模型得到的质心高度数据相对误差为0.759%。

数据处理, 预测模型, 灰色GM(1,1)模型, BP神经网络

 ,