Research On Computing Load Modelling And Cloud-Edge Collaborative Strategy For Distribution Power Internet Of Things Application

Under the background of Power Internet of Things,the distribution applications have presented a complex and diversified trend of massive information access,concurrent user interaction,real-time closed-loop control and so on,which puts forward higher requirements for the real-time,economical efficiency,security and reliability of the decision-making of the Power Distribution Internet of things(PD-Io T)applications.The PD-Io T integrates cloud computing and edge computing technology to realize cloud-edge collaboration and resource scheduling of distribution application,so as to give full play to its advantages of process software definition and elastic resource allocation.Therefore,the research on computing load modelling and cloud-edge collaboration scheme of distribution application is an important and basic work.Starting from the demand of the PD-Io T application,this paper analyzes its computing load characteristics and cloud-edge collaboration mechanism,and puts forward a computing load modelling method of distribution application considering uncertainty.Further,based on the uncertain computing load model of distribution application,the cloud-edge collaboration scheme of PD-Io T is studied from the two dimensions of cloud-edge collaborative application offloading and computing resource allocation of edge computing terminals.The main research work of this paper is as follows:(1)The overall architecture,key components and distribution application requirements of the power distribution Internet of things are analyzed.Combined with the application characteristics and the cloud-edge collaborative resource provision mode,the cloud-edge collaborative resource optimization way is analyzed from two dimensions: cloud-edge collaborative application offloading and cloud-edge computing resource allocation.(2)Analyze the nature and characteristics of deterministic and uncertain computing load,and map the uncertainty of allocation computing load into the number of application requests.An adaptive kernel density estimation probability model optimization method based on pseudo variable method is proposed.The probability density curve and probability distribution curve of the number of distribution application requests are generated,and the distribution application computing load model considering uncertainty is obtained.(3)The cloud-edge collaborative offloading problem of distribution application is studied.Taking the typical periodic energy management application as an example,a cloud-edge collaborative offloading method of energy management application considering offloading cost is proposed.Based on the two different offloading paths of cloud-edge offloading and edge-edge offloading,the computing cost,communication cost and queuing cost models are established respectively.Taking the minimization of the overall offloading cost of the system as the objective function,the optimal offloading path of the applications is solved and determined,and the low-cost and low delay offloading of the energy management application of the power distribution Internet of things is realized.(4)Considering the uncertainty of application computing load,the problem of cloud-edge collaborative application offloading and the computing resource allocation of edge computing terminals are studied.For the orderly charging application of electric vehicles with strong random fluctuation in the number of applications,firstly,based on the computing load model,the application coverage level is proposed to measure the uncertainty of computing load,then the computing resource allocation model of edge computing terminal and the cloud-edge collaborative application offloading model with minimizing application delay as the objective function are established,It realizes the reasonable allocation of computing resources of edge computing terminals under different application coverage levels and the low-delay cloud-edge collaborative offloading of electric vehicle orderly charging application.