Performance Improvement And Evaluation Of Industrial Real-time Communication Network Based On Artificial Intelligence

Industrial communication network is the Core Technique of enterprise automation system,is very important in industrial automation.With the rapid development of the Internet technology and network business,the network performance becomes the focus of people’s concern.Therefore,we need to analyse and evaluate the network performance of industrial communication network,in order to improve network performance and can be used in various industrial applications.This thesis regard CAN (Controller Area Network) as research object,the Real-time Communication performance is researched.Firstly,the simulation model of CAN communication control protocol was designed using the tool of STATEFLOW in MATLAB/SIMULINK,from the results of the model,the network performance of CAN bus such as throughput,average delay of message,collision probability,network utilization,network efficiency,and the message finish rate were acquired.After the simulation and analysis,the results is accords with CAN bus protocol,It approves that modeling and simulation of CAN bus protocol with Stateflow is feasible,and Stateflow is a new approach to analyse and research CAN bus protocol.However,the static priority algorithm of CAN still exists some unsolved issues in the communication scheduling.Secondly,the evaluation index system is established,and the BP neural network is used to evaluate network performance of CAN bus.After learning and testing,the model can be used in evaluating network performance of CAN bus,showed the effectiveness and correctness of the methods.Finally,the dynamic priority scheduling algorithm for real-time systems is researched.The EDF scheduling algorithm Is proposed to improve network performance of CAN bus.The simulation model of CAN communication control protocol was designed using the tool of STATEFLOW in MATLAB/SIMULINK based on the EDF scheduling algorithm,from the results of the model,the success rate of EDF scheduling algorithm were acquired.To verify this method,comparing the result with the success rate of static scheduling algorithm,the communication performance is improved.