基于多次约束匹配的室内定位算法

传统增强动态K加权算法(enhanced weighted K nearest neighbors,EWKNN)算法相比K加权邻近算法(weighted K nearest neighbors,WKNN)算法,虽然能有效提高定位精度,但是仍然存在定位结果在空间中跳动跨度大的问题。针对传统EWKNN算法的不足,在EWKNN算法的基础上提出了基于多次约束匹配的室内定位算法。考虑了行人前后位置的关联性,用上一步预测的位置对当前步的Wi Fi匹配进行多次约束,剔除匹配到的较远的参考点。实验表明,相比于传统的EWKNN算法,研究结果具有较高的定位精度。

Indoor Location Algorithm Based on Multiple Constraint Matching

Compared with the WKNN algorithm, the traditional EWKNN algorithm can effectively improve the positioning accuracy, but there is still a problem that the localization result has a long jump span in space. Aiming at the shortcomings of the traditional EWKNN algorithm, this paper proposes an indoor localization algorithm based on multiple constraint matching based on the EWKNN algorithm. The algorithm takes the correlation between the front and rear position of the pedestrian into account, and uses the predicted position of the previous step to constrain the WiFi matching of the current step many times, and remove the distant reference points from the matching. Experiments show that compared with the traditional EWKNN algorithm, this algorithm has a higher positioning accuracy.