| Intelligent agricultural equipment is facing the development of full unmanned operation.In the past,a single machine was used for operation.The demand for multi-machine collaborative work gradually increased.A single unmanned platform could not meet the complex and diverse tasks of the task,so the coordination between multiple machines Work will be a potential direction of development.This subject research uses elastic cloud server,Mongo DB database cluster and Node.js technology to achieve collaborative communication between multiple robots.This research starts from the cloud platform of the air-ground collaborative system,considers the communication method between multi-machine coordination,and aims at the mutual communication requirements of the ground mobile robot and the UAV in the case of cooperative operation,and studies a Node.The communication cloud platform of js technology realizes the mutual sharing of data between multiple computers.The main research contents are as follows:1.Analyzed the development process of multi-robot communication,air-ground robot collaboration system,Node.js and Express.js applications,Mongo DB-based data storage applications,and cloud computing and cloud robot platform development in different fields.The deficiencies of the research were summarized through the comparative article analysis,and determining the method and technical route of the communication cloud platform research.2.Through the analysis of the overall system architecture,determine the key technologies and research platforms used by the cloud platform.And carry out demand analysis and architecture design of the system.On this basis,the design of the air-ground robot platform and the classification of sensor data are carried out.Finally,the database is selected according to the requirements of the communication cloud platform sensor data compared with the traditional relational database,and also the real-time communication requirement and massive data generated from the cloud platform.Last,the database cluster architecture were designed and implemented.3.According to the functional requirements of the system,the implementation method of developing a cloud communication platform between multiple robots based on Node.js and Express.js is determined.Analyzed the format requirements of GNSS,numerical,and image sensors,designed the sensor’s data communication operation interface based on RESTful design principles;at the same time,designed the Mongoose database storage mode(Schema)based on the Mongodb database;finally implemented it based on Python The robot end accesses the cloud service platform program,and the browser end realizes the real-time view of the cloud platform status.4.According to the requirements of cloud platform communication,the corresponding prototype system was implemented.Based on binocular vision in the orchard,the deep neural network Yolov4 was used to achieve obstacle information acquisition and positioning,and shared to the cloud platform according to the requirements of the cloud platform data interface format.The cloud platform realizes real-time data display of data and completes information sharing with other robots.In order to further test the data communication capability of the cloud platform and the local embedded computer,the network cloud platform network delay,file upload and download test,the test data shows that the maximum network delay is 36.122 ms,the minimum is 23.347 ms;in the file upload test,the minimum The value is 1.2 MB/s and the maximum value is 2.9 MB/s;the minimum value of the file download test is 580.9 KB/s,and the maximum value is 680.9KB/s.Finally,the Apache Bench tool was used to conduct a concurrent test on the communication cloud platform,and finally it was concluded that the web server can achieve a relatively stable response to high concurrent processing capabilities under a concurrent load of 500 requests/sec. |
PDF Full Text Request