Non-linearity Compensation for Syrup Concentration Sensor Based on BP ANN

The sensor signals would affected by the surroundings in the syrup concentration measuring, so its performances are unstable and low accuracy. It is necessary to calibrate the non-linearity for the sensors. Pointing out that the BP ANN is the best method in sensor no-linearity calibrating under comparing the advantages and disadvantages of the methods that applied in calibrating. As for the non-linearity on sensor, the BP artificial neural network model was built to compensate it. The sensor data were processed by BP ANN function in the MATLAB ANN toolbox, it has solved the problem that sensor affected by temperature. The BP ANN simulation show that the sensor output stability reached 0.25% after the temperature compensation; this enhances the sensor´s precision and anti-jamming ability.