Integrating WSN and Laser SLAM for Mobile Robot Indoor Localization

Localization plays a vital role in the mobile robot navigation system and is a fundamental capability for the following path planning task.In an indoor environment where the global positioning system signal fails or becomes weak,the wireless sensor network(WSN)or simultaneous localization and mapping(SLAM)scheme gradually becomes a research hot spot.WSN method uses received signal strength indicator(RSSI)values to determine the position of the target signal node,however,the orientation of the target node is not clear.Besides,the distance error is large when the indoor signal receives interference.The laser SLAM-based method usually uses a 2D laser Lidar to build an occupancy grid map,then locates the robot according to the known grid map.Unfortunately,this scheme only works effectively in those areas with salient geometrical features.The traditional particle filter always fails for areas with similar structures,such as a long corridor.To solve their shortcomings,this paper proposes a novel coarse-to-fine paradigm that uses WSN to assist mobile robot localization in a geometrically similar environment.Firstly,the fingerprints database is built in the offline stage to get reference distance information.The distance data is determined by the statistical mean value of multiple RSSI values.Secondly,a hybrid map with grid cells and RSSI values is constructed when the mobile robot moves from a starting point to the ending place.Thirdly,the RSSI values are thought of as a basic reference to get a coarse localization.Finally,an improved particle filteringmethod is presented to achieve fine localization.Experimental results demonstrate that our approach is effective and robust for global localization.The localization success rate reaches 97.0%and the average moving distance is only 0.74 meters,while the traditional method always fails.In addition,the method also works well when the mobile robot is kidnapped to another position in the environment.

A graph autoencoder network to measure the geometric similarity of drainage networks in scaling transformation

Similarity measurement has been a prevailing research topic geographic information science.Geometric similarity measurement inin scaling transformation(GSM_ST)is critical to ensure spatial data quality while balancing detailed information with distinctive features.However,GSM_ST is an uncertain problem due to subjective spatial cognition,global and local concerns,and geometric complexity.Traditional rule-based methods considering multiple consistent conditions require subjective adjustments to characteristics and weights,leading to poor robustness in addressing GSM_ST.This study proposes an unsupervised representation learning framework for automated GSM_ST,using a Graph Autoencoder Network(GAE)and drainage networks as an example.The framework involves constructing a drainage graph,designing the GAE architecture for GSM_ST,and using Cosine similarity to measure similarity based on the GAE-derived drainage embeddings in different scales.We perform extensive experiments and compare methods across 71 drainage networks duringfive scaling transformations.The results show that the proposed GAE method outperforms other methods with a satisfaction ratio of around 88%and has strong robustness.Moreover,our proposed method also can be applied to other scenarios,such as measuring similarity between geographical entities at different times and data from different datasets.