This article focuses on the fixed-time (FxT) synchronization of discontinuous fuzzy competitive neural networks (DFCNNs) with leakage-delays and state-dependent switching. Discontinuous activations, fuzzy terms, leakage delays, and state-dependent switching are considered into the competitive neural networks to increase the generality in practical applications. Before studying the synchronization, as an important theoretical basis, a new FxT stability lemma is established based on the method of nonscaling integral denominator, which is low conservative. Five different settling-time estimates are explicitly given by discussing the range of the exponents. Previous results are included and greatly improved. In order to achieve the FxT synchronization of the addressed DFCNNs, a continuously bounded control scheme is proposed by introducing a saturation function rather than a sign function, which cannot only solve the chattering phenomenon, but also ensure the boundedness of the control input. Some new leakage-delay-dependent synchronization criteria are derived for the DFCNNs. Finally, the validity of the main results is verified by numerical simulations.

中文翻译：

---

具有状态相关切换的不连续模糊竞争网络的基于有界和饱和控制的固定时间同步


本文重点讨论具有泄漏延迟和状态相关切换的不连续模糊竞争神经网络 (DFCNN) 的固定时间 (FxT) 同步。竞争性神经网络考虑了不连续激活、模糊项、泄漏延迟和状态相关切换，以增加实际应用中的通用性。在研究同步之前，作为重要的理论基础，基于非标度积分分母的方法建立了新的FxT稳定性引理，该引理具有低保守性。通过讨论指数范围明确给出了五种不同的稳定时间估计。以前的结果被包括在内并得到了很大的改进。为了实现所讨论的DFCNN的FxT同步，通过引入饱和函数而不是符号函数，提出了连续有界控制方案，不仅解决了抖振现象，而且保证了控制输入的有界性。为 DFCNN 导出了一些新的与泄漏延迟相关的同步标准。最后通过数值模拟验证了主要结果的有效性。