A multi-strategy improved dung beetle optimisation algorithm and its application

To address the shortcomings of the dung beetle optimizer, such as low convergence precision and a tendency to fall into local optima, a multi-strategy improved dung beetle optimizer (IDBO) is proposed. Firstly, a Cubic chaos mapping strategy is introduced to enhance the diversity of the initial population; secondly, a global exploration strategy from the Osprey optimization algorithm is incorporated, endowing the dung beetle algorithm with the ability to identify the best areas and escape from local optima, which preliminarily improves the convergence speed and optimization precision of the algorithm; finally, an adaptive t-distribution perturbation strategy is adopted to disturb the foraging behavior of the dung beetles, allowing the algorithm to further accelerate the convergence speed while enhancing global exploitation and local exploration capabilities. The effectiveness of the three improvement strategies is verified through testing and analysis with the CEC2021 and CEC2017 test functions, and a convergence analysis of the improved algorithm’s optimization results compared to other algorithms is conducted. The Wilcoxon rank-sum test demonstrates that the IDBO algorithm has good convergence speed and optimization precision. Moreover, the IDBO algorithm is used to optimize the parameters of the HKELM prediction model and applied to short-term photovoltaic power generation prediction simulation comparison experiments. The experimental results show that compared to the DBO-HKELM prediction model, the error metrics MAE and RMSE of the IDBO-HKELM are reduced by 43.95% and 50.79% respectively, further verifying the feasibility and effectiveness of the IDBO algorithm in solving practical application problems. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.