Multi-model Adaptive Control Based on Heuristic Algorithms

In order to reduce the waste of computational resources in traditional weighted multiple-model control methods, this paper proposes a new adaptive control method for controlling multi-model systems subject to Laplace noise interference. Utilizing Bayesian methods to estimate the unknown parameters of the noise, it then performs online modeling of the multi-model system. By employing the Bald Eagle Search algorithm, it searches for the globally optimal points and partially local optimal points corresponding to the controllers of the multi-model system. Only the selected controllers are weighted, reducing the number of models that need to be computed. The results show that compared to traditional weighted multiple-model control systems, this control system achieves an average reduction of 91% in computational resource redundancy while ensuring the stability and robustness of the control system.