| Due to the large area of public places in the city,it is impossible to obtain the status and usage of large-scale garbage cans by manual work in time.In order to improve the management level of garbage cans,intelligent management of large-scale garbage cans by using Internet of Things technology has attracted widespread attention.In order to manage garbage cans timely and conveniently,an intelligent garbage can system based on Internet of Things technology was designed and implemented,and garbage can positioning algorithm was explored.Aiming at the problems of single function,high cost and lack of indoor public environment application scenarios of the existing intelligent garbage can system,an indoor intelligent garbage can system was designed.The system was controlled by STM32 with high performance and low cost,combined with six functional modules:temperature and humidity detection module,garbage residual detection module,gas state detection module,tilt state detection module,switch control module and Wi Fi information collection module.By using Message Queuing Telemetry Transport(MQTT)to communicate with host computer,the remote status monitoring,control switch and indoor positioning functions were realized.Firstly,the system status detection and control scheme was designed to complete the collection of system status data.Secondly,the ESP8266 Wi Fi module was bound to Smart Config to improve the efficiency of network distribution by listening on wireless Access Point(AP)channel.Thirdly,through analyzing the characteristics and the communication process of the MQTT protocol,designed suitable message topic and content,implemented message transmission access authentication and SSL security optimization,and improved the stability and reliability of mqtt communication in complex networks through a series of improvement measures such as adaptive heartbeat keeping alive mechanism,disconnection reconnection mechanism,connection state feedback and repeated loss message processing.Finally,mobile phone software was developed to realize the display of system status data and remote control equipment.The experiment showed that the test results of each functional module of the system are accurate and reliable,and the communication is relatively stable,which verifies the feasibility of the system design scheme.In order to improve the positioning accuracy and efficiency of indoor trash cans,an indoor trash can positioning scheme based on Wi Fi location fingerprint was designed.Firstly,Wi Fi information was collected offline to build the location fingerprint library.By analyzing Wi Fi signal characteristics,the fingerprint library optimization scheme based on Kalman filter and multi-coefficient AP selection was adopted to reduce the dimension of the fingerprint library and improve the robustness.Secondly,the initial clustering center was determined based on the similarity of AP set,cosine similarity was introduced as the measure of clustering distance,and Thompson was used to screen the abnormal reference points of the fingerprint database for clustering partitioning,so as to improve the accuracy of partitioning location.Thirdly,the Wi Fi information collected by ESP8266 was uploaded to the MQTT server as the fingerprint of the test points.In the online matching stage,the test point was located in the partition and the edge points of the partition are processed.Finally,Kernel Principal Component Analysis(KPCA)was used to extract fingerprint features.The improved Self-adaption-k Weighted K Nearest Neighbor(SWKNN)algorithm was used to locate the test points accurately.Experimental results showed that the improved partition and KPCA-SWKNN fusion localization algorithm is better than the traditional localization algorithm and other improved algorithms in terms of positioning accuracy and efficiency,and the online positioning accuracy of garbage cans is about 1.64 m. |
PDF Full Text Request