Research On Service Object Search Based On Home Service Robot System

With the development of urbanization and the aging degree of society to deepencontinuously, the need of service robots into the home to provide daily services for theelderly is more and more urgent. When the service robot is execute the service order,firstly it often needs to search the service object in home environment, and service objectsearch includes service object identification and location. How to complete the serviceobject identification and location accurately in the home environment is an important partof the home service robot in study. In order to overcome the shortcoming of existingmethods of services object identification, based on existing technologies fully, this paper isto propose a home service robot service object search method in aid of smart space-basedradio frequency identification (RFID) technology. This paper will focus on the followingaspects.Firstly, this paper proposes a RFID-based service object identification method. Withpasting electronic tags on service objects in the home environment, the method convertsthe service object identification to the electronic tag identification, to reduce thecomplexity of the service object identification, and improve service object identificationaccuracy and efficiency.Secondly, this paper proposes a method that locates service object coarsely, based onthe k-nearest neighbor algorithms. While locating the object coarsely, conventionalmethods directly convert the received signal strength from tags by readers to distanceinformation, which cause large errors. Therefore this paper makes the similarity ofreceived energy from object and reference tags by RFID reader as the weight of thesimilarity of their position, and then uses the k-nearest neighbor algorithm for coarselylocating the service object.Again, this paper proposes a accurate service object location method, based onBayesian probability search. After gaining a rough location of service object by using ofk-nearest neighbor algorithm, we propose a probability search algorithm based onBayesian, which uses the rough location information as a priori information to obtain precise position of the service object.Finally, we design experiments in this paper targets to validate the feasibility andefficiency of the home service robot service object search algorithm proposed in this paper,and collate and analyze experimental data.