In response to the accuracy issue in multi-source data fusion for smart factory monitoring environments,a two-tier fusion-based multi-sensor data fusion method was proposed to enhance the accuracy and reliability of multi-source data fusion.The method consists of a primary data fusion stage and a secondary decision fusion stage.Firstly,a combination of Kalman filte-ring and adaptive weighted averaging is employed to denoise and fuse data from sensors of the same type.Subsequently,the artifi-cial rabbit optimization(ARO)algorithm was utilized to optimize the extreme learning machine(ELM)neural network for deci-sion fusion.Experimental results demonstrate that the ARO-ELM-based multi-sensor data fusion algorithm outperforms other ad-vanced algorithms in terms of fusion accuracy.After verification,the proposed two-tier fusion-based multi-sensor data fusion scheme has superior fusion performance,effectively enhancing the reliability and robustness of the perception system,thereby ena-bling more accurate and reliable monitoring and prediction.
关键词
多传感器数据融合/卡尔曼滤波/自适应加权平均/人工兔优化算法/ELM神经网络
Key words
multi-sensor data fusion/Kalman filter/adaptive weighted averaging/artificial rabbits optimization/ELM neural net-work
维普
万方数据