Fruit fly optimization algorithm based on membrane computing deals with grain temperature prediction
Jing Yu, Xia Li · 2024
To address the issues of local optima trapping and slow convergence in the Fruit Fly Optimization Algorithm (FOA), an enhanced approach called Membrane Computing-based Fruit Fly Optimization Algorithm (MCFOA) is proposed to optimize models. The MCFOA algorithm merges the parallelism of membrane computing with the efficient search capabilities of the fruit fly optimization algorithm, thereby enhancing both search accuracy and convergence speed. First, the improved fruit fly optimization algorithm was built by the cell-like P system consisting of three cells to find the optimal initial weights and thresholds for the BP neural network, to improve the accuracy of grain temperature prediction. In addition, the three cells used different fruit fly optimization algorithms as their evolutionary mechanisms. Under the control of evolutionary mechanisms and communication mechanisms, the model was applied to improve the convergence performance of the algorithm, overcome the defect of falling into local optimal, and continuously optimize the weights and thresholds of the BP neural network to achieve the aiming of improving the prediction accuracy. The results showed that the prediction curve of the MCFOA-BP model is closer to the actual grain temperature curve, and the prediction performance of the MCFOA-BP model is better than other models.