| In recent years,with the rapid development of industrial Internet,the manufacturing industry has gradually changed from large-scale mass production to customized flexible production,which is characterized by small-scale,diversified and customized.In order to meet the needs of traditional manufacturing factories evolving to flexible intelligent customization factories,this paper proposes to use cloud PLC instead of traditional PLC to deploy in the industrial edge computing environment,and realize the rapid processing of industrial data nearby.In order to ensure the rapid control and real-time response of key tasks,cloud PLC must solve the high timeliness of data update.High data timeliness is mainly achieved by minimizing the age of data information(AoI).On the one hand,the transmission delay caused by the transmission of equipment update data to cloud PLC is an important factor affecting AoI.This problem can be solved by optimizing the deployment of cloud PLC and planning the routing path between equipment and cloud PLC.On the other hand,increasing the update frequency of data can reduce the peak AoI and improve the precision of cloud PLC's perception of equipment state.However,the high data update frequency may increase the amount of network data,and thus increase the queuing delay of data in the gateway device.To solve these problems,this paper studies the AoI adaptive optimization mechanism of coordinated cloud PLC optimal deployment and queue scheduling policy scheduling,and achieves the goal of minimizing AoI of cloud PLC data peak.Firstly,this paper proposes a cloud PLC optimization deployment algorithm based on bottleneck resource dependence to solve the problem that some key resources are exhausted in the process of data transmission,which leads to greater delay in data detour transmission.The algorithm analyzes the resource distribution of cloud PLC application and establishes the bottleneck level of cloud PLC virtual resources.Ignore some cloud PLC application requests that abuse bottleneck resources to ensure that the total amount of network resources can meet the needs of all cloud PLC applications.On this basis,the deployment position and data routing optimization of cloud PLC are completed,and the total transmission delay of the system is reduced by about 30%;compared with existing algorithm,the reduction ratio of rejected application requests caused by insufficient resource can reach up to 66.7%.In order to achieve high time-effectiveness of data update,this paper proposes an AoI adaptive optimization mechanism based on cloud-based PLC behavior prediction to minimize the average peak AoI of cloud-based PLC applications.In this paper,the sequential flow pattern of the data exchange process between the cloud PLC application and its equipment is modeled as a Markov process system model.The time series of the flow is regarded as the observation series,and it is assumed that it is controlled by the underlying hidden state of markov process.And build the flow prediction algorithm to realize the cloud PLC behavior pattern prediction.The cloud PLC traffic scheduling strategy of the gateway is optimized according to the prediction law,and the optimal traffic scheduling strategy is adaptively formulated according to the time distribution law of the gateway traffic,so as to reduce the network transmission delay and the queue delay of the gateway,and finally reduce the average peak AoI of 40%-50%. |
PDF Full Text Request