Design of Power Grid Flexible Planning Algorithm Based on Deep Learning

Conventional methods usually only consider one possible future environment, and cannot deal with the influence of various uncertain factors, so the obtained flexible planning scheme of power grid has poor adaptability to external environmental changes. This paper mainly discusses the flexible planning algorithm of power grid affected by uncertain factors. In this paper, the global optimization ability of GA(genetic algorithm) and the nonlinear approximation ability of BPNN(BP neural network) are combined for analysis and improvement, and an algorithm model of power grid flexible planning based on deep learning is established. It can not only reduce the search space of GA and improve the search efficiency, but also easily converge to the optimal solution, which provides a new strategy for solving the flexible power grid planning problem. It is found that the cost of the control group without considering the flexible balance of supply and demand is much higher than the total cost of this model, which shows the economic advantages of this model. The flexible cost of wind power and the flexible cost of load respectively represent that the lower the cost of waste air volume and load shedding, the better the wind power absorption capacity and the stronger the power supply reliability. The obtained planning scheme is more flexible and adaptable in the uncertain environment in the future.