Efficiency prediction of foam drainage measures using enhanced dung beetle optimization

Abstract

The dung beetle algorithm, when applied to complex multimodal function problems, tends to fall into local optima and may converge prematurely to suboptimal solutions, thereby restricting its application in high-dimensional and complex optimization issues. To address these limitations, a new dynamic weight adjustment mechanism and escape strategy have been introduced. Through the solution and comparison of 12 benchmark test functions, the improved dung beetle optimization algorithm (DGDBO) has demonstrated enhanced global search capabilities and resistance to premature convergence. This is evidenced by comparison results with other popular optimization algorithms. The improved optimization algorithm was applied in predicting the effectiveness of foam drainage gas recovery measures, the LightGBM prediction model and BiLSTM time series model optimized by DGDBO are established. Experimental results show that the DGDBO model can quickly and accurately identify the optimal parameters of the model, and significantly improve the accuracy and efficiency of the model prediction. This verifies the effectiveness of the DGDBO algorithm and shows its potential in practical engineering applications.