Service Routing Method Of Smart Grid Communication Network Based On Deep Reinforcement Learning

With the rapid development of the information society,big data the Internet of Things and other technologies have gradually entered our lives.The role of power communication networks in people’s lives has become increasingly important.Smart grid is the trend of development and direction of the next generation grid.Compared with the traditional grid,it can fully meet user demand for electricity and optimize resource allocation,which reflects the high degree of integration of power flow,information flow,and service flow.In the backbone network of the power communication network,the number of services and the scale of traffic are getting larger and larger.The traffic model is eventual and sudden.The network needs to have the ability to adjust environment flexibly and respond to service changes quickly.This brings tremendous pressure on existing networks.In this context,it is particularly important to distribute the power service routing reasonably.According to the characteristics of different power services in the current backbone network,this paper proposes a smart grid communication network service routing method based on deep reinforcement learning,which is in the smart grid communication network scenario.This is of great significance for the safe and reliable operation of the smart grid communication network.Reasonable distribution of service routing,fast and reliable transmission of services,improvement of communication resource utilization,and reduction of network risks are the focus and difficulty of power communication networks.This topic studies two different service scenarios.According to the characteristics of their service scenarios,this paper proposes different efficient and stable service routing methods,and uses corresponding deep reinforcement learning algorithms to solve them,which greatly improves the efficiency of problem solving.For power control-type services,they have higher requirements for channel performance and delays within a smaller range.Most control-type services use fiber optic cables for transmission.Therefore,in the context of the SDH optical transmission network,this paper proposes a method for power control-type service route distribution for link balancing.From the perspective of link load balancing,this paper proposes a routing distribution model for power control-type services carried on the transmission network,and uses the deep reinforcement learning DQN algorithm to solve it,which can not only guarantee the overall QoS of the service,but also take into account the link balancing of the network.Besides,it can make up for inefficiency of traditional methods.This is of great significance for the safe and stable operation of control-type services in the smart grid communication network.In the face of rapidly emerging and explosion IP services,IP+optical communication network architecture will become an important mode of communication for smart grid communication network.Under the control of SDN,management and maintenance of IP+optical networks can be realized effectively.In order to improve the collaborative ability and resource utilization of IP+optical networks,this paper proposes a joint routing algorithm for IP+optical network service routing based on deep reinforcement learning A3 C.Firstly,this paper combines the characteristics of IP data services,and designs a service risk balance indicator based on the bearing characteristics of IP and optical networks.Then,based on the network delay,bandwidth,site level difference,and similarity of primary and alternate routes,a reasonable distribution model of primary and alternate route is designed.Finally,the A3 C algorithm is used to solve the model.Simulation results and comparative analysis show that this method can make full use of the resources of the IP+optical network,while ensuring the average delay of services and reducing network risks.Also it can effectively improve the convergence speed.This provides a reference for the routing of power data services in the smart grid communication network.This topic analyzes the current problems and services of the smart grid.Based on the existing routing distribution methods and solving methods,a smart grid communication network service routing method based on deep reinforcement learning is proposed.This method not only guarantees the safe transmission of the services,but also improves the efficiency of routing distribution in complex power grids.It can provide guidance for future large-scale service concurrent routing distribution schemes,and also provide a demonstration and theoretical basis for the construction of smart grid communication networks.