Sensor sensitivity is usually effected by crossed factors, thus its output characteristic is not only changed with object parameter, but also easily interfered with measurement circumstance such as temperature, humidity and supply voltage fluctuations etc. The method of monitoring these parameters synchronously with different sensors and fusing these data with wavelet neural network is proposed. Pressure sensor is chosen as a simulation example, and the upper method is used to improve its output performance. The simulation results show that the method can effectively eliminate the infection of circumstance. The rapid convergence rate and compensation accuracy are better than traditional methods and neural network. The infection of non-object parameters is eliminated, and measurement accuracy is improved. The algorithm can easily be extended to other kinds of intelligent sensors and has important practical application value.
[1]
Qinghua Zhang,et al.
Wavelet networks
,
1992,
IEEE Trans. Neural Networks.
[2]
Jianjuan Liu.
Application of Fuzzy Neural Networks Based on Genetic Algorithms in Integrated Navigation System
,
2009,
2009 Second International Conference on Intelligent Computation Technology and Automation.
[3]
Hou Zhi.
Data fusion of pressure sensor based on neural network
,
2002
.
[4]
Mohsen Razzaghi,et al.
The Legendre wavelets operational matrix of integration
,
2001,
Int. J. Syst. Sci..
[5]
Xu Xiao-su.
Application of wavelet neural networks based on genetic algorithms in integrated navigation system
,
2006
.