Research On Transmission Line Path Planning Method Combining Artificial Intelligence Technology

With the rapid economic development,China’s power grid construction has entered a new era of ultra-high voltage from the era of ultra-high voltage,and the application of power electronic equipment has become more and more extensive,the status quo like this all puts higher and higher requirements on the grid design.Faced with the increasingly complex power grid construction environment and external Changes in the environment,the needs of industry development and changes in design concepts,it is imminent to adopt a new method of intelligent route selection for transmission line design.At the same time,artificial intelligence and big data technology have entered a new era of widespread application,and the transmission line path is a professional field suitable for artificial intelligence and other technologies.At this time,it is the right time to introduce artificial intelligence into the grid transmission line path planning.Therefore,how to combine the successful experience of existing transmission line path planning,and use advanced technologies such as artificial intelligence to greatly improve the efficiency and design quality of transmission line planning and selection is a subject of great value and worth studying.In this context,this paper takes the research of the transmission line path planning method combined with artificial intelligence as the research direction.Among them,the applied artificial intelligence technology is mainly manifested in BP neural network algorithm,intelligent quadtree image segmentation technology and path planning intelligent algorithm.The research work in this paper mainly includes the evaluation model of route selection scheme,geographic information grid cost model and path planning algorithm technology.In terms of the evaluation process of the line selection scheme,this paper proposes an evaluation model of transmission line path planning method combining AHP and BP neural network technology.The specific implementation process is: using Delphi method and AHP to obtain the original of each alternative The indicator data and the ideal output score divide the obtained data samples and results into a training set and a verification set,which are used for training and verification of the BP neural network model,respectively,and finally obtain the BP neural network model whose accuracy meets the evaluation requirements.The BP neural network model can intelligently evaluate other alternative route planning schemes with similar planning requirements,so as to achieve the goal of intelligently and autonomously completing the scoring and ranking.Finally,the proposed evaluation method is analyzed to verify the feasibility of the program.In terms of the modeling process of geographic information grid cost model,this paper first proposes to rasterize the original high-definition remote sensing image based on quadtree segmentation technology,and focuses on the analysis and design of the realization method of the intelligent adaptive quadtree image segmentation algorithm and the termination judgment condition of the quadtree segmentation process,so as to establish an adaptive variable resolution geographic information grid cost model.It is conducive to the formation of accurate and streamlined optimal rasterized fitting results of remote sensing maps,and then the detailed analysis and design of the neighborhood structure of the geographic information cost model and the calculation rules of raster cost data.Finally,by comparing this model with the unified resolution grid geographic information cost model,the feasibility and advantages of the proposed method are verified.In terms of path planning algorithm research,this chapter proposes a progressive multilevel path planning algorithm,and detailed design and discussion of the algorithm implementation process.The algorithm uses an intelligent adaptive resolution geographic information grid cost model to divide the route design process into three progressive levels.The planning process strictly follows these three levels of order,by progressively reducing the appropriate channel range,using intelligent ant colony algorithm for fine path planning within the selected channel range,and finally supplementing the line selection results with line distortion correction.