基于特征分解谱估计的蓄电池内阻检测方法

Battery Resistance Detection based on Eigendecomposition Spectrum Estimation

摘要:

针对目前在蓄电池内阻检测中被广泛采用的锁相放大法存在硬件电路复杂、成本较高、操作复杂等问题,提出采用特征分解谱估计的方法来替代蓄电池内阻检测中的锁相放大环节,以软件计算的方法替代硬件电路,从而降低硬件成本及操作难度；通过对特征分解谱估计的原理分析以及对运用特征分解谱估计方法进行蓄电池内阻测量时的关键步骤,即电压信号的提取与幅值检测进行仿真,仿真表明特征分解谱估计方法在高功率噪声背景下仍有较高的频率分辨力,幅值测量结果在50 dB高斯白噪声背景下测量误差约为2%,而在20 dB非高斯白噪声背景下测量误差约为3%；分析表明基于特征分解谱估计的蓄电池内阻检测方法在较高功率非高斯噪声背景下具有良好的检测效果,可以作为微弱信号检测的软件实现方法以替代锁相放大环节。

Abstract:

Method based on eigendecomposition spectrum estimation is used to replace lock-in amplifier which is widely used in battery resistance detection. The new method can reduce hardware costs and operational difficulties due to use software way instead of hardware circuit. Series analyses about the theory of eigendecomposition spectrum estimation are taken. Simulations about the extraction and measuring for voltage signal,which is the key step in the method, are done in the paper. Results of simulations shows good frequency resolution. The measurement error of the method can be around 2% under 50dB white Gaussian noise, while around 3% under 20dB non-Gaussian noise background. Analysis and simulations shows the method based on eigendecomposition spectrum estimation is efective in battery resistance detection under higher power non-Gaussian noise background and can be software method of Weak Signal Detection to replace lock-in amplifier.