| One of the important links of science and engineering teaching is experiment,a large number of experimental operations can deepen the understanding of the theoretical knowledge learned,therefore,practical teaching has attracted attention with the teaching reform.However,the existing distance experimental teaching system mainly integrates multimedia network teaching resources into the database,and students log in to the system for online learning or interact with teachers,although it overcomes the problems of uneven distribution of teaching resources and time and space,it does not take into account the practical teaching of science and engineering courses.In view of the existing problems,this dissertation puts forward the functional design of the combination of local experiment and remote online experiment,the combination of virtual and real,the monitoring of experimental process and the automatic collection of experimental data,and the prediction of experimental results,for the mechanical remote experimental teaching system,this dissertation constructs a modular design method,and develops relevant software and hardware systems,the main research contents include:(1)A comprehensive experimental platform is designed.The experimental platform includes two parts: mechanical and electric control,the mechanical arm can be upside down installed on the beam reinforcement of the experimental workbench or directly installed on the profile desktop;as a mechanical experiment case,this dissertation designs a small belt drive experiment module,which is mainly composed of motor part,load part,speed detection part,torque detection part and weight part.(2)The PLC control and data acquisition system is designed.In this dissertation,PLCRS422 communication interface is connected with the port of the computer,and FXPLC communication protocol is used to communicate with the host computer to realize the sensing acquisition and control of physical signals.(3)A remote browser client is developed.In B/S mode,through the combination of Django framework,Python language,HTML language,CSS language and Java Script language,the web client is developed,and the front-end function module and background management system are designed,the front-end function module includes basic functions,course functions and user operation management,and the background management system includes experimental course management,authentication and authorization management and online experiment management;the E-R diagram is drawn,and the corresponding database is designed according to the E-R diagram.(4)The local server software program is developed.According to the requirements of modularization of experimental platform and openness of experimental content,a new obstacle calibration algorithm is proposed,and the obstacle avoidance function software module is developed;Integrate Windows platform and ROS platform technology,communicate with ROS host through rosbridge protocol,realize diversified data information collection and ensure the reliability of system operation;finally,this dissertation develops the neural network performance prediction program,uses the multi-layer perceptron MLP model,trains the data set through the super parameter optimization setting,obtains the regression model,predicts the characteristics of students’ performance distribution according to the regression model,analyzes the teaching quality,and improves or maintains it according to the performance distribution.Taking the belt drive experiment as an example,according to the operation requirements of the belt drive experiment,the comprehensive test is carried out in the laboratory environment with Anno robot,the experimental test results show that the browser client developed by Django framework and the local server software program developed based on rosbridge run stably and can realize diversified data information collection,the new obstacle calibration algorithm proposed in this dissertation can also adapt to the changes of obstacles,the content of this dissertation is verified. |
