国产a片 学术报告[2026]101号

（高水平大学建设系列报告1360号）

报告题目：On the optimal design of stochastic resonance system based on Bayesian optimization

报告人：李志保 副教授 （中南大学）

报告时间：2026年10月26日11:00 - 12:00

报告地点：腾讯会议 465628393

报告摘要：A stochastic resonance (SR) system is typically formulated using the Langevin equation to model a bistable or multistable nonlinear system. The equilibria of the system are determined by its parameters, which play a crucial role in overall performance, particularly in high-dimensional SR systems. In this paper, we focus on the optimal design of a four-dimensional SR system derived from a modified Lorenz force chaotic system. This modification introduces additional adjustable parameters, increasing the total to eight, which enhances detection performance, improves signal processing flexibility, and enables adaptation to diverse application scenarios. To determine the optimal parameter configuration, we employ the signal-to-noise ratio improvement (SNRI) to evaluate signal detection performance and guide the development of a Bayesian optimization-based parameter selection algorithm. By maximizing SNRI, we identify the optimal system configuration. Additionally, we introduce specific metrics to assess the reliability of SR systems when detecting weak signals subject to characteristic frequency disturbances. Experimental results indicate that the proposed algorithm exhibits superior performance across diverse challenging scenarios, thereby establishing a robust basis for subsequent enhancements to SR models.

报告人简介：李志保，中南大学数学与统计学院副教授，硕士生导师。研究方向包括最优化理论与方法、传感器阵列信号处理、随机共振声等。在 Inverse Problems，Applied Mathematical Modelling，Journal of Computational and Applied Mathematics，Engineering Applications of Artificial Intelligence，IEEE/ACM Transactions on Audio, Speech, and Language Processing，Digital Signal Processing 等期刊发表学术论文30余篇。（曾）主持国家自然科学基金面上项目、青年项目、湖南省自然科学基金面上项目、青年项目与长沙市科技计划项目等。

邀请人：涂凯

国产a片

2026年10月23日