Research on power industry engineering optimization algorithm based on deep learning

Power system plays an important role in the development of national economy, and maintaining the balance between supply and demand of power grid is the key condition to ensure the stable operation of power system. At present, the super-capacity electric energy storage technology has not been broken, so it is necessary to implement the corresponding generation scheduling plan to ensure the stable operation of the power system. Effective short-term forecasting of power load is the basis for making the dispatching plan. Accurate short-term load forecasting can make the dispatching plan more accurate, and can reduce the power loss caused by inaccurate forecasting. As one of the hottest technologies in recent years, deep learning has achieved a lot of breakthrough development results so far. The technology is gradually maturing and the theory is gradually enriched, which has been widely used in many fields at present. However, most of the deep neural networks have a huge network architecture, which is deep and wide, with a large number of parameters and a large amount of calculation, so it requires higher computer hardware, and generally requires a high-profile graphics card to run. Aiming at the problem that the smart grid can’t effectively coordinate the production, transportation and distribution of electric energy due to the inaccurate short-term electric load forecast in the power system, in order to reduce the waste of resources caused by overload or low load and unnecessary carbon dioxide emissions, this paper proposes a new deep learning method to solve the problem of reliable short-term electric load forecast in this kind of power grid.