Research on power grid investment capacity prediction model based on multiple method combination

With the increasingly complex investment environment of China's power grid, accurately predicting the investment capacity of power grid enterprises has become an important prerequisite. Firstly, based on the internal and external environment of the enterprise, factors and indicators that affect the investment capacity of the power grid are selected from three dimensions: internal finance, economy and society, and the power industry. Principal component analysis is conducted to obtain the principal component factors after dimensionality reduction. Then, multiple regression models, backpropagation neural network models, and autoregressive integrated moving average models are used to predict the investment capacity of the power grid. Finally, a combination prediction model is constructed based on the three single prediction models mentioned above, and the weight coefficient of the single prediction model is calculated using nonlinear programming method. The results indicate that compared with a single prediction model, the combined prediction model can achieve higher prediction accuracy in predicting the investment capacity of the power grid, providing a reference for power grid investment decision-making and management.