$\mathcal {H}_{\infty }$ State Estimation for PDT-Switched Coupled Neural Networks Under Round-Robin Protocol: A Cooperation-Competition-Based Mechanism

作     者：Hao Shen Yinsheng Song Jing Wang Ju H. Park 

作者机构：AnHui Province Key Laboratory of Special Heavy Load Robot Anhui University of Technology Ma'anshan China Key Laboratory of Advanced Control and Optimization for Chemical Processes Ministry of Education East China University of Science and Technology Shanghai China School of Electrical and Information Engineering Anhui University of Technology Ma'anshan China School of Information Science and Engineering East China University of Science and Technology Shanghai China Department of Electrical Engineering Yeungnam University Kyongsan Republic of Korea 

出 版 物：《IEEE Transactions on Network Science and Engineering》 

年 卷 期：2022年第10卷第2期

页      面：911-921页

学科分类：0808[工学-电气工程] 08[工学] 

基　　金：National Research Foundation of Korea Ministry of Science and Information and Communications Technology National Natural Science Foundation of China Natural Science Foundation for Excellent Young Scholars of Anhui Province Natural Science Foundation of Higher Education Institutions of Anhui Province Major Technologies Research and Development Special Program of Anhui Province Key research and development projects of Anhui Province Open Project of China International Science and Technology Cooperation Base on Intelligent Equipment Manufacturing in Special Service Environment Fundamental Research Funds for the Central Universities 

主　　题：Switches Topology Switching circuits Protocols Neurons State estimation Quantization (signal) 

摘      要：This paper is concentrated on the $\mathcal {H_{\infty }}$ state estimation problem for switched coupled neural networks based on a Takagi-Sugeno fuzzy model. Notably, the time-variant network topology with alternate fast switchings and slow ones is described suitably by a persistent dwell-time rule, and the interactive dynamics with both cooperative properties and antagonistic ones among nodes are featured comprehensively by the switching signed graph. In view of the communication pressures brought by network-induced problems and the requirements in digital control, the round-robin protocol and logarithmic quantization are flexibly integrated for more transmission efficiency and fewer data collisions. Thereafter, by utilizing a relaxed multiple Lyapunov function method and some novel matrix process techniques, sufficient criteria guaranteeing the exponential stability in a globally uniform sense with a prescribed $\mathcal {H_{\infty }}$ performance level of the estimation error system are established. Finally, the synthesized analysis of the proposed method is presented with an illustrative example.