Double Artificial Neural Networks Research Articles

Noninvasive Blood Glucose Measurement Based on NIR Spectrums and Double ANN Analysis

This paper presents a new noninvasive blood glucose monitoring method based on four near infrared spectrums and double artificial neural network analysis. We choose four near infrared wavelengths, 820 nm, 875 nm, 945 nm, 1050 nm, as transmission spectrums, and capture four fingers transmission PPG signals simultaneously. The wavelet transform algorithm is used to remove baseline drift, smooth signals and extract eight eigenvalues of each PPG signal. The eigenvalues are the input parameters of double artificial neural network analysis model. Double artificial neural network regression combines the classification recognition algorithm with prediction algorithm to improve the accuracy of measurement. Experiments show that the root mean square error of the prediction is between 0.97 mg/dL - 6.69 mg/dL, the average of root mean square error is 3.80 mg/dL.

Study of Wind Farm Power Output Predicting Model Based on Nonlinear Time Series

To solve the problem of the variancy of the wind power when wind farm connect with the power grid, a wind power predicting model of wind farm based on double ANNs is proposed in the paper. Wind velocity and wind direction on wind farm are the key of wind power predicting, and other circumstance conditions such as temperature, humidity, atmospheric pressure, are also great influence on it. The observed values of these five circumstance conditions can be treated as a nonlinear time series and be analyzed by the nonlinear time series ANNs model. The wind power predicting model consists of double artificial neural networks. The first is consisted of five artificial neural networks which is used to prediction the circumstance conditions time series, the second is employed to prediction the power of wind farm use predicting value of the five conditions. A series of simulation show that the results of the predicting model is acceptable in engineering application.

Analysis of ultraviolet absorption spectrum of Chinese herbal medicine–Cortex Fraxini by double ANN

A fast, accurate and convenient method for the simultaneous determination of multi-component in the Chinese herbal medicine was proposed by using ultraviolet absorption spectrum. In this method, dummy components were added to training sample, and a double artificial neural network (DANN) that has the function of high self-revision and self-simulation was used. Effect of other interference components could be eliminated by adjusting concentration of dummy components. Therefore, the accuracy of concentration prediction for multi-component in the complicated Chinese herbal medicine was improved. It has been realized that two effective components of Cortex Fraxini, aesculin and aesculetin, were simultaneously determined, without any separation. The predicted accuracy was 92% within the permitted relative errors. The measurement precisions of the aesculin and aesculetin were 0.37% and 1.5%, respectively.