Data fusion based on first optimization and its comparison with the traditional algorithms

Piezoresistive pressure sensors are widely used in industrial measurement and control systems and can greatly affect the performance of these systems. However, cross-sensitivity exists in most pressure sensors, whose static characteristics are not only influenced by the variety of target parameters but also subjected to non-target parameters. In this paper, we have proposed a new method based on 1stOpt (First Optimization) to reduce cross-sensitivity and improve the stability and measuring accuracy of pressure sensors. It can be applied for the fusion of two data sets generated by pressure sensors. To demonstrate the usefulness of this method, a practical case study is investigated. Compared with two widely used methods, SVR (support vector regression) and BP neural network (back propagation neural network), data fusion based on 1stOpt proves to be of higher accuracy, better robustness and wider application range.