Performance Evaluation On SharkBus Characteristic Indexes And Optimization Deployment Research

Buses are the cornerstone of the next generation of intelligent manufacturing and the Internet of Things.At present,domestic high-end bus technology is under embargo.On the other hand,the development of information technology has put forward higher requirements for buses in the industrial field.Shark Bus independently developed by the Shanxi Engineering Research Center for New Industrial Bus(North University of China)emerged under this background.Shark Bus is a masterless network that uses a storage and forwarding mechanism.It has functions such as high dynamic network automatic reconstruction,multipath mesh redundancy,automatic time synchronization,time triggering,traffic control,etc.It is a kind of high reliability network suitable for military and industrial fields.Shark Bus has many excellent features such as automatic reconstruction,high speed,high reliability,and high real-time.Among them,high reliability and high real-time are the application characteristics that users are most concerned about.However,in the specific application,the performance of these indexes can only be obtained through actual measurement and cannot be quantitatively compared with other buses.Therefore,it is of great significance to systematically evaluate the performance of Shark Bus indexes and conduct a quantitative comparative analysis with other buses through theoretical research and simulation.This can provide a quantitative basis for Shark Bus relevant performance.This dissertation systematically investigates the connectivity reliability,the optimized deployment scheme,and the time synchronization accuracy of Shark Bus.It compares and analyzes the topological reliability and the time synchronization accuracy of Shark Bus.The main research contents are as follows:(1)In order to compare and evaluate the connectivity reliability of Shark Bus with other several mainstream buses,a virtual node method is proposed.This algorithm extends the application range of the connectivity reliability algorithm to the bus network and solves the problem that existing algorithms are not suitable for evaluating the bus network.A comparative evaluation of the invulnerability and the survivability of topology architectures of seven types of buses is performed,including Shark Bus,AFDX,the bus network without redundancy,the bus network with redundancy,the star network,the ring network,and the tree network.In addition,the impact of the number of terminal nodes,the link reliability,and the operating time on the connectivity reliability of the network is studied.The results show that compared with other buses,Shark Bus has higher structural robustness,and the network is more robust and stable.No matter in the case of a large number of terminal nodes,or in the case of low link reliability,or in the case of long runtime,Shark Bus has better performance.(2)In order to guide the application deployment of Shark Bus network,we study the impact of different deployment schemes and different numbers of switches and terminal devices on the connectivity reliability of the network,and propose an optimized deployment scheme for improving the connectivity reliability of Shark Bus.The results show that the fewer the number of network nodes,the higher the connectivity reliability of the network;under the deployment scheme of multi-layer cascading of switches,the fewer the number of switches in the last layer,the higher the connectivity reliability of the network;the closer the terminal devices is to the bottom layer,the higher the connectivity reliability of the network;any one of different deployment schemes of switches,different numbers of switches,and different deployment schemes of terminal devices can ensure the stability of the connectivity reliability.However,different numbers of terminal devices in the network have a greater impact on the connectivity reliability.The content of this research provides important guiding principles for the application deployment of Shark Bus network.(3)The impact of crystal oscillator frequency,crystal oscillator accuracy,and asymmetry of the communication path on the time synchronization accuracy is studied,respectively.In order to achieve higher time synchronization accuracy between master and slave nodes,a single-cycle proportional compensation algorithm is proposed.The time synchronization accuracy of Shark Bus is evaluated from both simulation and hardware testing aspects.The results show that a time synchronization accuracy of ±1 clock period can be achieved between the master node and the slave node.Compared with the methods proposed in the existing literature,the time synchronization method of Shark Bus can achieve higher time synchronization performance with lower overhead.The research works of this dissertation have important significance for the application and promotion of Shark Bus and provide technical support for the next generation of the intelligent manufacturing industry.