The Research On Real-time Dynamic Scheduling Based On CAN In Industry Control Systems

With the development of computer hardware, software techniques, and integrated circuit technology, the industrial control system has become one of the most active branches in computer technology applications, which has made a huge progress. The research on Fieldbus and the applications have become very hot in the domain of industrial data bus. As fieldbus standards are yet to be reunified in order to make the fieldbus users and applications to be able to work without any restraint of standards. Controller Area Network has arisen to meet the requirements.As the basic CAN protocol does not answer the clear limitation of the real time data transmission. That is why at high load under these circumstances can lead system performance to decline to some extent, or even collapse. It can even affect the stability of links in the CAN bus among the system applications.In this thesis, we have carried out an in-depth study on CAN bus in the dynamic scheduling problem and proposed a real time dynamic scheduling algorithm for CAN bus. One aspect is that we have introduced the concept of time intervals into CAN networks to make the bandwidth distribution more reasonable, in order to speed up the real-time data transmission, to minimize the bandwidth consumption, and to reduce the communication latency. The other aspect is that we have considered the data retransmission mechanism. When data transmission is failed, we will re-transmit the data using the non-real-time interval without allocating additional network resources. This Retransmission mechanism can reduce the latency of real-time data.This thesis summarizes both interval selection mechanism and re-transmission mechanism, given the stability of data to meet all conditions:Through rigorous simulations, we have demonstrated that our proposed algorithm out-performs some existing algorithms in the literature.This thesis has described the improvement and development in the two aspects of CAN networks. We also introduce the experimental settings for the simulations to examine our proposed algorithm.