highly significantP > 0.1
Excluding insignificant items, the quadratic regression model of the error between the stratification effect of the actual fertilizer application device and the stratification effect of the target fertilizer application device is obtained as 9 According to Table 5 , it can be seen that the quadratic regression model P < 0.01 was highly significant, and the loss of fit term P > 0.1 was not significant, which indicates that the fitted model can correctly reflect the relationship between each factor and the error and can better predict the test results; A, B, C, AB, BC, and B 2 are significant, and the rest of the terms are not significant, and according to the size of the model regression coefficient, it can be seen that the influence of each factor on the error in order from the largest to smallest is C, B, and A.