Study on Data Fusion Algorithms of Hot-Pressing Measurement Process of Wood-Based Panels

Measurement error often exists during hot pressing of the wood-based panels due to the complex mechanism of the process, technological conditions, environment noises and performance of the sensors. Multi-sensor technology is used for mat temperature and pressure data processing to improve the precision and reliability of data. By determining the information weight coefficient with the corresponding measurement variance of each sensor the weighted least square algorithm is applied to achieve high accuracy of overall estimation. The precision of overall estimation value is preferable than the partial estimation of each sensor. A target recognition frame of control strategy is set, and D-S evidence combination principles and reasoning theory are introduced to select the object of greatest credibility for decision making. The results of experiment demonstrate the effectiveness and reliability of the algorithm.