Abstract: High-quality development planning for power grid enterprises is not only a key path to realize high-quality development of the energy industry, but also an important support to promote the realization of the ”double carbon” goal and the construction of a new energy system, through the latent dirichlet allocation(LDA) topic identification model, indicators are selected from six aspects, such as safety and reliability, scientific and technological innovation, service capability, production efficiency, business performance, green development, etc., and the evaluation system of high-quality development planning for power grid enterprises is established by combining the hierarchical analysis method and BP neural network, and empirical analysis is carried out. It is found that the effectiveness of high-quality development planning of power grid enterprises is on the rise, but the annual growth rate fluctuates greatly, and compared with safety, reliability and service capacity, the degree of importance attached to scientific and technological innovation and green development is relatively insufficient. The key factors influencing the high-quality development planning of power grid enterprises should be systematically optimized, and the planning and construction of scientific and technological innovation and green development should be strengthened.

Key words: power grid enterprise, high-quality development planning evaluation, LDA model, AHP-BP neural network