BP神经网络在液浆浓度检测中的应用

针对液浆浓度检测系统中,浓度与折射率、温度之间的非线性关系问题,以实验数据为基础,采用了多元线性回归建模和BP神经网络建模2种方法,分别建立了检测系统的数学模型。利用MATLAB进行仿真,研究了采用不同的数据处理算法的差异,通过误差对比分析,得出了应采用BP神经网络技术建立智能化液浆浓度测量模型,提高测量精度,克服温度变化的非线性影响的结论。

Application of BP Neural Network in Luquid Concentration Detection

To the non-linear problem of the concentration measurement system,two modeling methods including multivariate linear regression model and BP neural networks model were used,and each model was set up based on test data.We researched the difference of each algorithm to solve this measurement system by MATLAB.Through the simulation and error comparison in two models,the results showed that BP neural network model should be applied in liquid concentration measurement system to improve measurement accuracy and decrease the non-linear effect from temperature.