Aiming at the data processing of waste gas volume fraction

multi-sensor is used to measure the environmental temperature

humidity

one waste gas volume fraction.For the data distortion duing to the sensor faults

the data fusion menthod is used based on RBF neural networks which colligates manifold factors to the measurement of one waste gas

to estimate the volume fractortion of this waste gas

consequently achieving the resumption of the data distortion.Take the data processing of the volume fraction of NH3 for example

the Matiab emulation result shows that the estimate error is less than 6.7 %

the efficiency of data fusion menthod based on RBF is proved.