【摘要】 基于非线性动力学理论,研究800 kW离心压缩机出口动态压力的最大Lyapunov指数特征,并用于离心压缩机喘振识别。首先,控制防喘振阀改变离心压缩机的工作状态,使其从稳定运行经过渡过程进入到喘振状态;通过分析出口动态压力的波形和频谱,研究压缩机运行状态与出口动态压力变化之间的关系;然后,分析出口动态压力在不同工作状态下的最大Lyapunov指数特征,以识别离心压缩机的喘振特征;最后,为验证所提出的分析方法的有效性及抗干扰性,引入随机噪声,进一步分析含有干扰信号的离心压缩机出口动态压力的最大Lyapunov指数特征。研究结果显示,在稳定状态下工作时,离心压缩机的出口动态压力的波动幅值较小,波形表现出随机特征。此时,最大Lyapunov指数较大并且波动较为剧烈,说明系统具有混沌特征;当系统由过渡过程进入喘振状态时,其出口动态压力的波动幅值增加并达到稳定,同时表现出周期特性;而在此状态,最大Lyapunov指数很小,接近于零,表明系统的周期性变强;而在引入干扰信号后的研究发现,动态压力的最大Lyapunov指数的抗干扰能力较强。研究结果将为实现离心压缩机的初始喘振状态识别和预测及防喘振控制等方面带来新的思路。更多还原

【Abstract】 Based on the nonlinear dynamics theory, the characteristics of the largest Lyapunov exponent of the flow pressure at the outlet of an 800 kW centrifugal compressor was studied, which may be used for the surge identification of the centrifugal compressor. First, the centrifugal compressor entered the surge state from a stable state through transition process gradually by controlling two anti-surge valves. After analyzing the waveform and frequency spectrum of the flow pressure, the relationship between compressor working state and change rule of flow pressure was studied. Then, the largest Lyapunov exponent characteristics of flow pressure under different working conditions were applied to identify the surge characteristics of the centrifugal compressor. Finally, in order to verify the effectiveness and anti-interference of the proposed methods, random noise was added to the original flow pressure and then the largest Lyapunov exponent was analyzed. The results show that when the centrifugal compressor works in a stable state, the fluctuation amplitude of the flow pressure of the centrifugal compressor is small, and the waveform is random. In this stage, the corresponding largest Lyapunov exponent is large and fluctuates violently, indicating that the system has certain chaotic characteristics. When the system enters the surge state, the fluctuation amplitude of the flow pressure increases and reaches the stability, showing a certain periodic characteristic. In the surge state, the largest Lyapunov exponent is very small and close to zero, indicating that the periodicity of the system becomes strong. Further, after the introduction of interference signals, it is found that the largest Lyapunov exponent of dynamic pressure has strong anti-interference ability. This work may bring new ideas for realizing the surge state identification and prediction and anti-surge control of centrifugal compressor.更多还原