Scholarlink

<p>To solve the chattering problem in the synchronization process of competitive neural networks (CNNs) and predict the working time of the controller, this paper studies the predefined-time (PDT) synchronization and energy consumption prediction of stochastic memristive competitive neural networks (SMCNNs). First, CNNs with multiple key dynamic factors such as memristive weights, time-scale parameters, stochastic disturbances, and mixed time-delays are considered. Based on the obtained PDT stability lemma, the PDT synchronization criterion is obtained. Second, a continuous PDT controller is designed to overcome the high-frequency chattering and control instability problems caused by traditional sign function control methods, which guarantees that the system synchronizes smoothly within the PDT. Third, the upper bound of energy consumption under the synchronization process of SMCNNs is derived, which helps to assess the working time of the controller ahead of practical application. At last, the effectiveness of the obtained method is addressed through a simulation example.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predefined-Time Synchronization and Energy Consumption Prediction of Stochastic Memristive Competitive Neural Networks With Multiple Time-Scales and Mixed Time-Delays

  • 期刊论文

  • 发表日期：

    2026年3月6日

查看全文
  • Henglei Liu,
  • Qiaoyu Chen,
  • Dongbing Tong,
  • Wuneng Zhou

摘要

To solve the chattering problem in the synchronization process of competitive neural networks (CNNs) and predict the working time of the controller, this paper studies the predefined-time (PDT) synchronization and energy consumption prediction of stochastic memristive competitive neural networks (SMCNNs). First, CNNs with multiple key dynamic factors such as memristive weights, time-scale parameters, stochastic disturbances, and mixed time-delays are considered. Based on the obtained PDT stability lemma, the PDT synchronization criterion is obtained. Second, a continuous PDT controller is designed to overcome the high-frequency chattering and control instability problems caused by traditional sign function control methods, which guarantees that the system synchronizes smoothly within the PDT. Third, the upper bound of energy consumption under the synchronization process of SMCNNs is derived, which helps to assess the working time of the controller ahead of practical application. At last, the effectiveness of the obtained method is addressed through a simulation example.