A photovoltaic power prediction model optimization method based on star oscillation optimization algorithm

By introducing the Stellar Oscillation Optimization (BSOO) algorithm to globally optimize the neural network parameters, the problem of local optima in photovoltaic power prediction is solved, achieving high-precision and stable photovoltaic power prediction and improving the model's predictive ability.

CN122315618APending Publication Date: 2026-06-30HUANENG CLEAN ENERGY RES INST +2
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HUANENG CLEAN ENERGY RES INST
Filing Date
2026-03-27
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing photovoltaic power prediction methods are prone to getting stuck in local optima during the neural network parameter optimization process, resulting in limited prediction accuracy and unstable model performance. In particular, in high-dimensional, non-convex search spaces, traditional optimization algorithms struggle to balance global exploration and local exploitation.

Method used

The Stellar Oscillation Optimization (BSOO) algorithm is used to globally optimize the parameters of the neural network. The parameter positions are updated iteratively through the oscillation mechanism and the Levy flight strategy. Combined with parameter standardization and destandardization mechanisms, the connection weights of the neural network model are optimized.

Benefits of technology

It improves the accuracy and stability of photovoltaic power prediction, enhances the model's generalization ability, and significantly improves the accuracy and consistency of prediction results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122315618A_ABST
    Figure CN122315618A_ABST
Patent Text Reader

Abstract

This invention proposes a photovoltaic power prediction model optimization method based on a stellar oscillation optimization algorithm. The method includes: acquiring historical electrical operation data and corresponding meteorological environmental data of a photovoltaic array; constructing training and testing datasets and performing normalization preprocessing; constructing a neural network prediction model containing an input layer, hidden layer, and output layer based on the preprocessed datasets; setting the connection weights and activation functions for each layer; using the stellar oscillation optimization algorithm to globally optimize the model's connection weights, iteratively updating the parameters until convergence through an oscillation mechanism and flight strategy; inputting the preprocessed test dataset into the optimized model, and outputting the photovoltaic power prediction value. This invention improves the efficiency and accuracy of parameter optimization for photovoltaic power prediction models, effectively enhancing the accuracy and reliability of photovoltaic power prediction results, and providing data support for the operation and regulation of photovoltaic systems.
Need to check novelty before this filing date? Find Prior Art