Method for forecasting electric power system short-term load based on method for improving uttermost learning machine

Abstract

The invention discloses a power system short-term load forecasting method based on improving extreme learning machine (IELM) methods, which adopts an extreme learning machine (ELM) as the basic structure of a forecasting model and is an iteration-analysis learning algorithm which puts forward BFGS (Broyden, Fletcher, Goldfarb, Shanno) simulated Newton's method to give optimization, training and adjustment to network left metrics and analyze to get right metric parameters. The invention establishes the short-term load forecasting model based on improving extreme learning machine (IELM) method, puts forward extreme learning machine network reference hidden node concepts, trains the equidimensional extreme learning machine networks which have the same numbers of hidden nodes and samples, orderly clusters module values of equidimensional network right metrics vector, finds out a plurality of corresponding module value break points, and regards the break points as the reference hidden nodes of predicting networks. The method of the invention is also greatly improved in the aspect of prediction precision and speed.

Description

technical field [0001] The invention belongs to a method for short-term load forecasting of a power system, in particular to a short-term load forecasting model of a power system based on an improved extreme learning machine theory. Background technique [0002] Scholars at home and abroad have done a lot of research on short-term load forecasting models of power systems. The main models are: traditional forecasting models based on time series and regression analysis; modern forecasting based on artificial neural network, wavelet analysis, expert system and other artificial intelligence theories Model. [0003] The short-term load forecasting model based on the time series method treats the load change as a time-varying sequence, finds out the change law in the historical load data sequence, and then extrapolates it for load forecasting. Commonly used models include autoregressive model, moving average model, autoregressive moving average model and cumulative autoregressive...

AI Technical Summary

Problems solved by technology

[0012] 1) BP algorithm learning rate η is difficult to choose

If η is too large, the training process will be unstable or difficult to converge; if η is too small, the training time may be greatly increased;

[0013] 2) When η is too small, the iterative process may fall into the "small pit" of the local extremum and fail to reach the global optimum, which not only wastes training time, but also has poor generalization ability;

[0014] 3) The number of hidden nodes in the network structure directly affects the learning ability and generalization ability of the network prediction model. However, at present, no effective method has been proposed for neural network training to give a reasonable number of hidden nodes.

[0015] The existing artificial neural network prediction model, on the one hand, because the neural network’s own learning and training needs to iteratively adjust all value parameters, the network optimization process always stays in a high-dimensional space, the complexity of optimization calculation increases, and the training time is too long , and it is easy to fall into the local optimum; on the other hand, the hidden node parameters of the network model have not been proposed a reasonable selection method, which limits the improvement of prediction accuracy

Images