Power standard cost prediction method based on LSTM optimizer

Abstract

The invention relates to an electric power standard cost prediction technology, and aims to provide an electric power standard cost prediction method based on an LSTM optimizer. The method comprises the steps of collecting historical power cost data and daily cash flow data of daily electricity selling income of a power company, performing standardized preprocessing, and preparing a data set for training; constructing an LSTM prediction network and an LSTM optimizer, and training and alternately updating weight parameters of the LSTM prediction network and the LSTM optimizer; and performing power standard cost prediction by using the trained LSTM prediction network to obtain a prediction result. According to the method, the electricity selling income daily cash flow data set with similar fluctuation ratios and the power cost data are adopted to participate in training of the LSTM prediction network and the LSTM optimizer together, and the problems of over-fitting, unstable algorithm, low precision and slow convergence caused by insufficient data quantity can be solved. The power standard cost can be automatically predicted, and the defect that in the prior art, the accuracy difference of different optimizers is large is overcome.

Description

technical field [0001] The present invention relates to the electricity standard cost prediction technology, in particular to a standard cost prediction method based on LSTM optimizer. Background technique [0002] Smart grid has become a new global energy strategy in the 21st century. The social responsibility of my country's power grid enterprises is to provide high-quality and low-cost power services. Its core business mainly includes power grid construction, power grid operation and maintenance, procurement and sales, while pursuing benefits maximize. This requires power grid companies to improve their management level and constantly strengthen cost management. The corresponding costs are mainly the cost of power grid construction; the second is the cost of power grid maintenance and operation; the third is the cost of management and operation. In the budget management of power grid enterprises, take the budget as the leader, establish a relatively complete cost accounti...

AI Technical Summary

Problems solved by technology

But from the current point of view, there are still three problems: first, the level of refinement needs to be improved; second, the degree of intensification is not high

Financial personnel focus on back-end accounting, and implement supervision based on the total cost, unable to accurately grasp the economy of project arrangements and the rationality of cost expenditures

Business personnel are proficient in project cost and business processing, but lack the concept and macroscopic grasp of total cost expenditure

Images