arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.04134math.FA
收藏

通过加权算子平均法获得数值半径的更精细上界

Refined upper bounds for the numerical radius via weighted operator means

University of Central Florida(中佛罗里达大学)

Shankhadeep Mondal, Ram Narayan Mohapatra, Kasun Tharuka Dewage

AI总结:

本文通过加权算子平均法改进数值半径上界,并揭示非正常算子的刚性现象。

摘要:

我们通过引入算子模与其共轭算子的加权几何平均,建立了有界线性算子在复Hilbert空间上数值半径的新上界。该方法产生了一组不等式,扩展并严格改进了Kittaneh和Bhunia--Paul等人已知的几个上界,除了在正常或退化情况下除外。进一步通过插值数值半径估计与谱半径界,得到一组混合不等式,为非正常算子提供更精确的控制。给出了$2\times2$算子矩阵的应用,并完全 characterize 等号情况,揭示了强烈的刚性现象。包含明确的例子以说明新界的严格性。

英文摘要:

We establish a parameterized family of upper bounds for the numerical radius of bounded linear operators on a complex Hilbert space, based on weighted expressions involving the modulus of an operator and its adjoint. The proposed estimates encompass a known numerical-radius inequality as a particular symmetric case, while optimization over the weight parameter provides additional flexibility and can lead to sharper bounds. Under suitable structural assumptions on the associated auxiliary operators, we further derive corresponding spectral-radius estimates. The weighted approach is also extended to $2\times2$ off-diagonal operator matrices. Finally, we examine sharpness and equality cases and provide explicit finite-dimensional examples illustrating the obtained estimates.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑