摘要
为提高多输入多输出(Multiple input multiple output,MIMO)线性随机振动试验系统中具有相关特性的多路驱动信号生成精度,基于驱动谱非负定Hermite特性及矩阵分解理论提出了一种CD(Cholesky decomposition)白噪声滤波信号生成方法。构造具有特定"频响特性"的传递系统,将一系列独立白噪声通过该系统得到具有预期相干特性、相位差以及自谱的平稳随机信号。建立悬臂梁两输入两输出线性振动系统仿真模型,将传统傅里叶逆变换时频转换方法与CD滤波法生成的驱动信号进行对比分析。结果表明新方法精度与参考值误差不足1dB。两轴振动试验验证结果表明,应用CD法既满足工程实际标准需求,又为振动环境试验提供理论依据。
To improve the generating precision of driving signal in multiple-input multiple-out (MIMO) linear random vibration test system, a new method called Cholesky decomposition(CD) filter based on white noise is proposed according to the matrix decomposition theory in view of the nonnegative definite Hermite characteristic of driving powr spectral density(PSD). The lower triangular matrix with particular frequency response transfer characteristics is constructed. A series of independent white noises are passed through the system, and the multiple stationary random signals with articipative coherence performance, phases and self-PSD are obtained. A two-input two-output simulation system based on a cantilever beam is established to verify the effectiveness of the proposed method. The mixed reference spectrum including board and narrow random signals is set and the results between the traditional time-frequency transformation method based on Fourier transform and the new method are compared. The results show that the new method has high accuracy in the self-spectrum and cross-spectrum, and can meet the technical requirements of engineering application, thus improving the effect of random vibration environment test control.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2017年第6期839-844,共6页
Journal of Nanjing University of Aeronautics & Astronautics
基金
国家自然科学基金(51705205)资助项目
江苏省高校自然科学研究面上(16KJD460001)资助项目
江苏大学高级人才科研启动基金(15JDG166)资助项目
关键词
驱动谱
白噪声
随机振动
CHOLESKY分解
频响模型
driving spectrum
white noise
random vibration
Cholesky decomposition
frequency response model