Research On WiFi-based Indoor Robot Localization Technology

With the development of robotics,more and more robots are used in various fields.For indoor mobile robots,the positioning and navigation of the robot itself is a core issue.In recent years,many robots have adopted visual SLAM(simultaneous localization and mapping)technology for positioning and navigation.However,this technology is sensitive to indoor lighting,and the image processing is complicated and the calculation is large,which affects the real-time positioning of the technology to a certain extent.Compared with the visual SLAM technology,WiFi-based indoor positioning technology has the advantages of not being interfered by light,simple positioning algorithm and good positioning real-time performance.At present,WiFi equipment has a high indoor penetration rate and relatively low cost.Therefore,this paper studies the indoor positioning of robots based on WiFi.First of all,in view of the traditional WiFi location fingerprint positioning algorithm(hereinafter referred to as the traditional location fingerprint algorithm in this article),the positioning accuracy is not high,and it is difficult to meet the needs of robot indoor positioning.The reasons for the low positioning accuracy of the traditional location fingerprint algorithm are studied.Through experiments,it is found that the main factors that affect the positioning accuracy of the traditional location fingerprint algorithm are: the size of the reference point spacing,the number of WiFi devices participating in the positioning,the large fluctuations of WiFi signal strength,and the number of WiFi acquisitions.Through the above experimental conclusions,a series of improvements are made to the traditional location fingerprint algorithm,and the improved algorithm is applied to the robot’s indoor positioning.The improved algorithm is divided into the phase of offline establishment of location fingerprint database and the phase of real-time positioning.During the offline establishment of the location fingerprint database,the distance between the reference points of the traditional location fingerprint algorithm is reduced,and then the WiFi signal is collected at each reference point by the time-based collection method.Aiming at the problem that the location fingerprint of the traditional location fingerprint algorithm is not very specific,the standardized processing method is used to standardize the WiFi signal strength data to improve the specificity of the location fingerprint.In the real-time positioning stage,the Mahalanobis distance is used as the similarity reference,and the coordinates corresponding to the fingerprint of the position with the smallest Mahalanobis distance are used as the estimated positioning coordinates.It has been verified by many experiments that the positioning accuracy of the improved algorithm is higher than that of the traditional position fingerprint algorithm,but there is a problem of large fluctuations in the positioning error.In order to solve this problem,an improved adaptive K-value WKNN algorithm is fused on the basis of the improved algorithm.The purpose of fusing this algorithm is to introduce position fingerprints that are effective for improving positioning accuracy,and to eliminate position fingerprints that reduce positioning accuracy.This effectively improves the positioning performance of the system and reduces the fluctuation range of the positioning error.The experimental results show that after fusing the improved adaptive K-value WKNN algorithm,the fluctuation range of the positioning error is significantly reduced,the variance of the positioning error is reduced from the original 0.31 to 0.21,and the positioning accuracy is also improved.Secondly,according to the characteristics of the improved algorithm,a grid-based navigation algorithm is proposed for indoor robot navigation.Compared with the existing navigation algorithms,this algorithm is simpler to implement,but requires stricter speed control of the robot’s motor,and the PID algorithm is used to control the motor’s speed.Obstacles are unavoidable in the navigation process,and a 3-way infrared obstacle avoidance module is used to avoid obstacles.After many experiments,it is found that the navigation accuracy of the proposed grid navigation algorithm meets the requirements of robot indoor navigation accuracy.Finally,a web-based robot control platform is designed according to the positioning algorithm and robot navigation requirements.On the offline data collection page,the robot can be controlled to collect WiFi signals and establish a location fingerprint database.The robot can be positioned on the positioning page.By setting the coordinates of the end point,the user realizes the robot’s autonomous navigation and can view the coordinates of the robot’s indoor position in real time.