Application of artificial neural networks to intelligent weighing systems
暂无分享,去创建一个
The authors present a new method for dynamic weighing, using a feature extractor and two-layer artificial neural network capable of predicting the final value of the sensor response in a noisy environment while it is still in oscillation. The method permits arbitrary input and initial conditions and requires no restriction on the order of the sensor. Introducing a pre-processor as a feature extraction block before the neural network dramatically reduces the required number of neurones. This, in turn, reduces the complexity of computation and offers the possibility of real-time procedures for dynamic force measurements. The proposed method is established by theoretical analysis and justified by means of both simulation and real data measurements.
[1] Bernard Widrow,et al. 30 years of adaptive neural networks: perceptron, Madaline, and backpropagation , 1990, Proc. IEEE.
[2] Neil M. White,et al. Adaptive filters in load cell response correction , 1993 .
[3] Wei-Qun Shu. Dynamic weighing under nonzero initial conditions , 1993 .