Meta Improved Differential Evolution Algorithm for Asymmetric Hammerstein System Model Parameter Identification

Traditional Differential Evolution (DE) algorithm has been widely used in black-box optimization problems due to its simplicity and efficiency. However, its performance is heavily dependent on the choice of hyper-parameters and mutation strategies used that has proven difficult to configure. To address these limitations, Meta Improved Differential Evolution (Meta IDE) algorithm auto-adjusts its hyper-parameters and strategies to avoid stagnations and adapt search behavior based on the search landscape (exploration and exploitation), and integrates the GPU accelerated parallel computing strategy to accelerate computation time thus improving optimization performance. For feasibility, this paper uses benchmark functions to verify the optimization performance of the proposed algorithm, and then implements it on smart piezoelectric aero-engine blade system model for parameters identification. Asymmetric Hammerstein model is used to characterize the smart aero-engine fan blade system. To validate the robustness and effectiveness of the proposed method, simulations and experiments were conducted and the results analysis establishes that the proposed Meta IDE algorithm has better optimization performance as compared to the traditional DE algorithm in terms of convergence speed and identification accuracy.