The Research of Data Fusion Method for Sample Mean Random Weighting Estimation

First a new method of random weighting estimation is applied to multi-sensor data fusion, a random weighting data fusion method of multi-source sensor is proposed. According to maximum likelihood theory, random samples coming from different sensors (same mean different variance) are fused in an effective way, and a new estimation of mean is gained, thus interference from low-precision detecting results is eliminated and accuracy of the detection results is enhanced. Secondly, an optimal fusion algorithm of multidimensional position based on the random weighting estimation is proposed. Research result shows that the new algorithm has a better performance than traditional information fusion method