Design And Implementation Of A Task Allocation System Based On Edge Computing In Power Internet Of Things

With the organic integration of the power system and the Internet of Things(IoT)system,the concept of power IoT is gradually formed in the construction process of the smart grid.In the power IoT,there are many different types of power tasks,such as the power Wireless Sensor Networks(WSNs)clustering routing scheduling task.In this context,an important research direction of intelligent power IoT is to design more efficient task allocation technology to solve the problem of deep interaction between grid workers and power tasks.Therefore,this thesis designs a task allocation simulation system with a desirable graphical interface for the power IoT.In general,the main work of this thesis is divided into the following three parts.(1)Aiming at the task of power WSNs clustering routing scheduling,this thesis establishes an efficient clustering model and proposes a new clustering method.Firstly,this thesis improved the traditional cluster-head election method in building the clustering model and designed an evaluation function aiming at maximizing the value of the function by considering factors such as the residual energy,intra-cluster distance,BS distance,and data transmission delay.Then,HPCP-QCWOA,a clustering routing method based on the improved whale optimization algorithm,is proposed,and a new quantum rotation angle table as well as a new hierarchical cloning strategy are provided.Finally,a test environment of power WSNs is built using MATLAB,and the proposed clustering method is compared with some state-of-the-art clustering protocols,namely O-LEACH,LDIWPSO,and ARSH-FATI-CHS.The test results show that the average transmission delay obtained by HPCP-QCWOA is 16.95%,15.55%,and 20.29% lower than the above clustering protocols,respectively.(2)To accelerate the execution speed of power tasks such as power WSNs clustering routing schedule,a power task edge offloading model is designed based on the concept of edge computing,and a new power task edge offloading method is proposed.Firstly,the offloading delay of power tasks is divided into three different parts,namely the task upload delay,execution delay,and result return delay.Some delay coefficients are introduced to obtain an evaluation function aiming at minimizing the function value.Then,an edge offloading method of power tasks based on an improved genetic algorithm is proposed,a new quantum operator is designed without searching the quantum rotation angle table,and an elite pool is constructed for storing historical optimal solutions.Finally,Cloud Sim is used to build an edge offloading environment for power tasks,and the proposed offloading method is compared with a variety of existing offloading methods,namely,GWO,PSO,sequential offloading,and random offloading.The test results show that the task completion delay of the proposed power task offloading method is reduced by 12.29%,8.80%,28.22%,and 40.92%,respectively.(3)Based on the mathematical models and algorithms related to task allocation in the power IoT,this thesis applies the MATLAB App Designer platform to build a task allocation simulation system with a desirable graphical interface.The system mainly consists of four modules,which are the system login module,system menu module,power WSNs clustering routing scheduling module,and power task edge offloading scheduling module.In addition,the system also provides some auxiliary functions,such as cleaning up the operating environment and viewing historical assignment results and user manuals,and grid staff can also save the appropriate task assignment results for subsequent operations.Finally,to enhance the reliability of the system,a comprehensive system function test is carried out.The test results show that the designed functions of the system have reached the expected goals.