The temperatures, electrical conductivities, capacitances and somatic cell counts of cow fresh milk were measured accurately, and a four-layer BP neural networks regression model was established. The temperatures, electrical conductivities and capacitances were used as the model input data, and somatic cell was counted as output data. The model results were compared with those of the model without the capacitances parameters. It showed that the detection accuracy had been significantly improved with .the capacitances parameters, the correct cow mastitis detection rate for validation sample set was 100% .%通过准确测量奶牛新鲜乳汁的温度、电导率、电容和乳汁中的体细胞含量,建立了温度、电导率、电容为输入,体细胞数为输出的四层BP神经网络模型.并将结果与没有电容参数的网络模型结果进行比较.结果表明:电容参数的加入使检测精度有显著提高,验证集奶牛乳腺炎等级的正确检出率达到100%.
展开▼