Journal
INTERNATIONAL JOURNAL OF CONTROL AUTOMATION AND SYSTEMS
Volume 19, Issue 10, Pages 3297-3308Publisher
INST CONTROL ROBOTICS & SYSTEMS, KOREAN INST ELECTRICAL ENGINEERS
DOI: 10.1007/s12555-020-0474-x
Keywords
Neural network; randomly occurring uncertainties; receding horizon stabilization; time delay
Categories
Funding
- National Natural Science Foundation of China [61403278, 61503280]
Ask authors/readers for more resources
This paper proposes an improved algorithm for delayed neural networks with randomly occurring uncertainties, optimizing stability in a receding horizon. By transforming the ROU problem into a linear matrix inequality and generating new matrix variables with more information, the method reduces conservatism and increases delay upper limit. Illustration of method superiority is provided through simulation numerical examples.
For delayed neural networks with randomly occurring uncertainties (ROU), this paper uses an improved integral inequality to optimize the stability of the receding horizon. The ROU follows some uncorrelated Bernoulli distribution white noise sequence, which it can enter the neural network in a free and random manner. By using a suitable lemma, the ROU problem added in this paper is transformed into a linear matrix inequality. Based on the auxiliary function-based integral inequality method, the new cross terms matrix of linear matrix inequality in the improved Lyapunov-Krasovskii functional is processed. Therefore, some new matrix variables containing more information are generated, so that the results have more degrees of freedom. This paper has obtained the new condition of the end-weighting matrix of the receding horizon cost function, thereby reducing its conservativeness and increasing its upper limit of delay. Finally, the superiority of the method has be illustrated by giving some simulation numerical examples.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available