Research On Key Problems Of Power Internet Of Things System Optimization

Power system is a very complex nonlinear artificial system.Its intelligent development is the key to guarantee security and reliability.Nowadays,the development direction of enterprises is to improve the service level of the industry in the Power Internet of Things which constructed by basic integration and control device that both utilize the information network.The thesis focuses on the three key problems of the perceptual layer,the network layer and the edge computing layer in the system structure.Through reasonable and effective methods,each layer is optimized to achieve the goal of improving the system performance.In the perception layer of the Power Internet of Things system,in order to increase the available communication modes of the measuring devices and improve the reliability of the data transmitted by the devices,the thesis designs the perceptual communication architecture which integrates power line carrier wired transmission and wireless micro power transmission.Moreover,for the purpose of improving the reliability of devices communication and transmission,the thesis studies the process of devices access and maintenance of fusion network,evaluating the channel index and constructing the selection table of transmission mode.Due to the large number of devices and the very complex topology in the network layer of the Power Internet of Things system,the path of transmitting broadcast frames and transmission delay are relatively long.In order to shorten them,the thesis determines the best route for transmitting data between devices.Firstly,based on the iterative search of traditional ant colony algorithm,the updating model of pheromone matrix is optimized in the thesis.Secondly,after the global optimal solution is obtained in every iteration,the checking mechanism of local optimal solution is added to overcome the shortcoming that the traditional ant colony algorithm is easily trapped in the local optimal value in the iterative search process.Meanwhile,and the iterative search performance of the algorithm is improved.In the edge computing layer of the Power Internet of Things system,the processing delay of devices is long because of the insufficient storage capacity and computing processing resources.In order to optimize it,the computational offloading algorithm is introduced in the thesis.Considering the dependencies among the sub-tasks,the optimization model is established.The centralized scheduling algorithm is studied to solve the problem.For solving the offloading strategy faster and overcome the shortcomings of the central scheduling algorithm,the distributed offloading algorithm is optimized according to the solving ideas of the classical distributed offloading algorithm.The optimized distributed offloading algorithm in the thesis not only takes into account the sub-task dependencies and reduces the task processing time,but also utilizes the advantages of the classical distributed offloading algorithm to ensure that all edge devices stop iteration and reach equilibrium steady state.