NF-6风洞轴流压缩机喘振信号检测、分析与应用
|
摘要
NF-6风洞是我国第一座具有国际先进水平的大型连续式高雷诺数跨声速翼型风洞，由于其轴流压缩机工作在高压低温环境中，受风洞结构限制入口气流脉动很大，很容易导致压缩机进入喘振区，严重危及运行人员的人身安全和设备的正常工作。当压缩机发生喘振时，其流量、压力、振动和温度等喘振表征信号会出现常规条件下难以想像的现象，目前在国际上能借鉴的资料不多，在国内也是首次研究，因此对有关信号的检测、分析与预测对于预防喘振现象的发生具有非常重要的科学意义和社会价值。然而，由于对风洞压缩机喘振现象的研究属于交叉学科，内容涉及空气动力学、信号检测与处理、旋转机械和控制等多个领域，需要考虑的因素很多。本论文利用多学科的知识，特别从信号检测与处理的角度，主要通过理论分析与实验相结合的方法，在考虑风洞和压缩机相互耦合、相互影响的基础上，将风洞和压缩机作为一个整体系统来研究，主要创新点和结果如下：
1．经过分析，得到了NF-6风洞和轴流压缩机联合运行时的稳定工作条件，为喘振信号的检测提供了理论依据。
提出用信号与系统的观点研究连续式高速风洞这个复杂的气体网络系统。由于风洞与压缩机二者形成闭环回路，整个系统的工作是一个相互耦合、相互作用的演化过程，这是风洞中压缩机与一般工业压缩机最大的区别，因此讨论风洞中压缩机的稳定运行与不稳定工况时必须考虑这个耦合过程。通过理论分析结果可以看出，即使压缩机本身工作在稳定区，由于风洞的耦合作用，工作点也会处于不稳定的工作状态，有向压缩机不稳定工况点漂移的倾向，因此压缩机的防喘振系统是NF-6风洞安全运行的必备条件。
2．通过对流量信号检测方式的分析，提出了采用等圆环面积法可以准确测量通过压缩机的流量，但必须同时考虑风洞总压和总温对流量测量准确度的影响。
流量信号是NF-6风洞轴流压缩机防喘振控制系统正常工作的关键信号之一，论文通过计算给出了NF-6风洞轴流压缩机工作的流量范围包线。由于NF-6风洞轴流压缩机入口气流是典型的非充分发展管流，目前工业上已经成熟应用的流量测量方法无法直接应用于NF-6风洞，因此论文提出采用经过标定的风速管在风洞第一拐角后测量该截面的平均流速，同时测量风洞稳定段的总压和总温，进而准确得到进入轴流压缩机的质量流量。论文还对传感器位置的选择、传感器形式、传感器量程和精度对测量结果的影响进行了详细的计算和讨论，这些结果指导了NF-6风洞在压缩机流量测量方面设备的建设，并更改了原设计中不合理的部分，降低了设备调试和运行中的风险。
3．通过对压力信号检测的模拟实验和分析，验证了滤波器截止频率的工程估算公式，采用小波变换分析了压缩机发生喘振时的压力突变信号，发现了稳态和动态压力信号检测中的部分规律。
轴流压缩机出口总压和进口总压之比称为压比信号，是防喘振控制系统正常工作的另一个关键信号。由于压缩机发生喘振的最根本的内在原因还是气流分离造成的，因此压缩机入口截面气流的不均匀性会加快或加剧喘振的发生。而NF-6风洞轴流压缩机由于受到风洞结构的限制，其进口截面气流存在脉动和偏转，单纯采用测量总压而得到的压比信号作为压缩机防喘振控制参数，一方面会造成压缩机的运行范围变窄，压缩机效率下降；另一方面也会使风洞在同样的工况条件下能耗增大，造成不必要的浪费，使风洞实验的成本增加。因此，论文在国内首次提出在压缩机入口和出口安装动态压力传感器，得到这两个位置上的动态总压信号。为了得到相关规律，通过一个进气道实验模拟了动态压力测量的有关问题。同时应用小波变换分析了压缩机发生喘振后压力信号的奇异性特征，提出将其应用于NF-6风洞轴流压缩机防喘振系统中，可以提高控制系统的快速性和准确性，有效地拓宽NF-6风洞的运行范围。
4．分析了NF-6风洞轴流压缩机轴系的振动情况以及进气温度对压缩机防喘振曲线的影响，其结果指导了振动信号和温度信号检测系统的建设。
分析结果表明，NF-6风洞压缩机轴系的结构一阶固有振动频率落在相应于压缩机正常工作转速的激励范围之内。进一步的分析则显示，源于电机、压缩机和齿轮的一般激振，都不会引起较大的扭振应力，因此可以认为整个系统能在设计的转速范围内安全运行。但当高速齿轮存在缺陷时，两阶振型系数都较大，若压缩机转速对应的频率接近一阶固有振动频率时，可能会引起整个轴系的振动，从而造成整个结构的破坏。因此，为了保证NF-6风洞的安全运行，在此计算结果的指导下，我们建立了NF-6风洞压缩机振动监测系统。同时，由于压缩机进气温度变化范围很宽，还讨论了温度测量精度对整个系统的影响，据此设计了专用的温度检测和保护系统。这些措施在NF-6风洞建设过程中都已经得到了应用，为整个系统的成功调试打下了坚实的基础。
5．初步建立了系统的数学模型，将多变量耦合系统进行了解耦，给出了连续式风洞控制的一般规律。
通过分析指出，NF-6风洞控制系统是典型的多变量耦合系统，从输入和输出关系上是典型的多输入多输出(MIMO)模型，因此单纯采用PID控制无法满足系统控制的要求。但由于工程的特殊要求，在风洞调试和运转的初期，我们可以将MIMO系统进行解耦，分为四个单独的回路，采用传统的PID调节对风洞进行控制，在保证风洞安全调试和运行的前提下，通过大量的实验获取风洞运行过程中的有关参数，对其进行分析和处理后，逐步采用智能控制等更为高级的控制方式，提高风洞运转的效率，扩大运行范围。
以上理论分析结果和实际计算结果已经成功用于指导NF-6风洞轴流压缩机防喘振控制系统的建设，取得了较为满意的结果。作者同时希望本论文的研究结果能够为以后我国建设类似的风洞提供一定的参考和帮助。
The NF-6 wind tunnel is the first internationally advanced large-scale continuously-running high Reynolds number transonic airfoil wind tunnel in China, because the axial-flow compressor installed in the wind tunnel loop works under a condition of high pressure and low temperature, the flow at its entrance fluctuates greatly and would easily cause the compressor to go into surging zone, severely endangering the safety of the people working on site and make the equipment unable to work normally. When surge occurs in the compressor, the signals such as flow rate, pressure, temperature, vibration, etc., which symbolize and characterize the surge, will exhibit phenomena that are incredible and impossible under normal conventional conditions. The research on these phenomena in this dissertation is the first time in China and the references are seldom found internationally. Therefore detecting, analyzing, and predicting the signals are very important to preventing surge in scientific sense and in social value. However, the research on axial-flow surge phenomena as a cross discipline, involving aerodynamics, detecting and processing of signals, rotating machinery, control, etc., has to take many factors into account. This dissertation took the wind tunnel and the compressor as an integrated unity, exploiting the knowledge of many disciplines especially from the view of signal detecting and processing, mainly through the combination of theoretical analysis and experiments and based on the consideration of the coupling and interacting between the wind tunnel and the compressor, and produced following main creative contributions and results:
1. Through analysis the conditions under which the NF-6 wind tunnel and its axial-flow compressor can jointly run stably were obtained, laying a theoretical basis for the detection of the surge signal.
It was raised and put forward in the dissertation that such a complicated network of fluid flow system as the continuously-running high-speed wind tunnel be studied based on the concept of signal and system. Since the wind tunnel along with the compressor forms a closed loop, the working process of the whole system is a evolutionary process of mutually coupling and mutually interacting, which is the most significant difference between a compressor in the wind tunnel loop and that in used conventional industries. Therefore, the mutually coupling and interacting process must be taken into account when discussing the stable operating conditions and unstable working situations of the compressor in a wind tunnel. It was found from theoretical analysis that even if the compressor itself works in stable zones, the operating point of the system of the wind tunnel and the compressor might be in an unstable state and have a trend to shift to an unstable operating point of the compressor because of the coupling effect from the wind tunnel and thus, as a whole, the surge-preventing system of the compressor is a necessary precondition for the safe running of NF-6 wind tunnel.
2. By analyzing the manner to detect flow rate signal a constant-annular-area method was proposed to accurately measure the flow rate through the compressor while at the same time the effect of the wind tunnel total pressure and total temperature on the measurement precision must be counted in.
In view of that the flow rate signal is one of the key signals for the anti-surging control system of the axial-flow compressor to work normally in the NF-6 wind tunnel, the dissertation calculated the envelope of the range of flow rate values which the axial-flow compressor should obey when working in the NF-6 wind tunnel. At the same time, since the entrance flow entering the axial-flow compressor of the NF-6 wind tunnel is typical non-fully-developed pipe flow, the matured methods widely used in common industries for measuring flow rate are unable to be directly applied to the NF-6 wind tunnel, thus the dissertation proposed to obtain the accurate mass flow rate entering the axial-flow compressor by both measuring the average flow velocity at the cross section immediately behind the wind tunnel's first comer with a calibrated anemometer and, at the same time, measuring the total pressure and total temperature of the settling chamber of the wind tunnel. Additionally, in this dissertation, the effects of location, form, range and precision of transducers on the measured results were calculated and discussed in detail. These results gave guidance to the construction of the compressor flow rate measuring equipment in the NF-6 wind tunnel and were used to modify and improve some part of the predecessor design and contributed to the reduction of the risk of debugging, testing, and running of the equipment.
3. The dissertation validated the empirical formulae for estimating the cutoff frequency of the filter through the simulation test of pressure signal detection and the analysis of the test, analyzed the signal of abrupt pressure variation when surge occurs in the compressor using wavelet transform, and found some rules in detecting stable and dynamic pressure signals.
The pressure ratio signal defined as the ratio of the outlet total pressure to the inlet total pressure of the axial-flow compressor is another key signal on which the anti-surging control system of the axial-flow compressor relies to work normally in NF-6 wind tunnel. Since the most leading and fundamental inherent cause of surging in compressor is flow separation, thus the flow non-uniformity at the compressor entrance section would accelerate or aggravate the occurring of surging. Because of the structural limitations of the wind tunnel there exist flow parameter fluctuations and flow deflections in the entrance section of the axial-flow compressor of the NF-6 wind tunnel. Therefore adopting the pressure ratio signal obtained by only measuring total pressure as the surging-preventing control parameter of the compressor will, on the one hand, lead to narrower compressor operating zone and lower compressor efficiency, while on the other hand, increase the energy consumption and the cost of the wind tunnel tests under the same operating conditions, hence cause unnecessary waste. In this dissertation a new concept was put forward for the first time nationwide that dynamic pressure sensors be installed both in the entrance and exit of the compressor to obtain the dynamic total pressure signals at the two positions. To acquire the corresponding rules, the issues associated with measuring the dynamic pressure were simulated by an inlet test. Furthermore the singularity characteristics of the pressure signals when were analyzed using wavelet transformation when surge occurs in the compressor, which was proposed to be applied in the design of the surging-preventing system of the axial-flow compressor in the NF-6 wind tunnel to improve the rapidity and accuracy of the control system and to widen the operating envelope of the wind tunnel.
4. This dissertation discussed the vibration of the axial-flow compressor's shafting of the NF-6 wind tunnel and the influence of the entering flow temperature of NF-6 wind tunnel on the surging-preventing curve, the results from the discussion gave a good guidance to the construction of the detecting and measuring system of vibration signals and temperature signals.
The analysis indicated that the first-order natural vibration frequency of the structure of the axial-flow compressor's shafting in NF-6 wind tunnel fell within the excitation region corresponding to the normal working rotating speed of the compressor. Further analysis showed that the general excitation originating from motors, compressors and gears would not cause large stress of twisting vibration. Therefore the whole system could be considered to be able to run safely in the designed range of rotating speed. But if the high-speed gears have flaws the coefficients of the vibration mode of both orders are large. Meanwhile, if the frequency corresponding to the rotating speed of the compressor is approaching the first-order natural vibration frequency the whole compressor's shafting system may vibrate and hence cause the destruction of the whole structure. Therefore, in order to ensure that the whole NF-6 wind tunnel run safely a monitoring system of compressor vibration was set up guided by the above computed results. Additionally, considering the influence the entering temperature of the compressor on the surging-preventing curve, the effect of the measuring precision of temperature on the whole system was discussed and a temperature detecting, checking and protecting system was designed based on the discussions. All these measures have been utilized in the construction of NF-6 wind tunnel and laid a solid foundation for the successful running of the whole system.
5. The dissertation set up an elementary mathematic model for the system, decoupled the multi-argument coupling system, and gave a generally applicable rule for the control of continuously-running wind tunnel.
Analysis shows that the control system of NF-6 wind tunnel was a typical multi-argument coupling system and a typical model of multi-input and multi-output (MIMO) as viewed from the input and output. Therefore the adopting purely PID control did not meet the requirements of the NF-6 wind tunnel system control. However, from the point of view of engineering in the early period of debugging and operating of the wind tunnel, the MIMO system can be decoupled into four single separate loops and then the wind tunnel can be controlled with traditional PID technique. With the safe debugging and running of the wind tunnel guaranteed, the parameters associated with the wind tunnel running process were acquired through a massive number of wind tunnel tests. After accumulating, analyzing and processing these parameters, more advanced control mode such as artificial intelligence control could be put into use gradually to improve the running efficiency and broaden the running envelop of the wind tunnel.
The above-mentioned results theoretical analysis and practical computations have been successfully used to guide the construction of the anti-surge system of the axial-flow compressor in the NF-6 wind tunnel and led to very satisfactory effects. The author hopes that the results in this dissertation will be helpful valuable to the construction of the similar wind tunnels in our country in the future.
1．经过分析，得到了NF-6风洞和轴流压缩机联合运行时的稳定工作条件，为喘振信号的检测提供了理论依据。
提出用信号与系统的观点研究连续式高速风洞这个复杂的气体网络系统。由于风洞与压缩机二者形成闭环回路，整个系统的工作是一个相互耦合、相互作用的演化过程，这是风洞中压缩机与一般工业压缩机最大的区别，因此讨论风洞中压缩机的稳定运行与不稳定工况时必须考虑这个耦合过程。通过理论分析结果可以看出，即使压缩机本身工作在稳定区，由于风洞的耦合作用，工作点也会处于不稳定的工作状态，有向压缩机不稳定工况点漂移的倾向，因此压缩机的防喘振系统是NF-6风洞安全运行的必备条件。
2．通过对流量信号检测方式的分析，提出了采用等圆环面积法可以准确测量通过压缩机的流量，但必须同时考虑风洞总压和总温对流量测量准确度的影响。
流量信号是NF-6风洞轴流压缩机防喘振控制系统正常工作的关键信号之一，论文通过计算给出了NF-6风洞轴流压缩机工作的流量范围包线。由于NF-6风洞轴流压缩机入口气流是典型的非充分发展管流，目前工业上已经成熟应用的流量测量方法无法直接应用于NF-6风洞，因此论文提出采用经过标定的风速管在风洞第一拐角后测量该截面的平均流速，同时测量风洞稳定段的总压和总温，进而准确得到进入轴流压缩机的质量流量。论文还对传感器位置的选择、传感器形式、传感器量程和精度对测量结果的影响进行了详细的计算和讨论，这些结果指导了NF-6风洞在压缩机流量测量方面设备的建设，并更改了原设计中不合理的部分，降低了设备调试和运行中的风险。
3．通过对压力信号检测的模拟实验和分析，验证了滤波器截止频率的工程估算公式，采用小波变换分析了压缩机发生喘振时的压力突变信号，发现了稳态和动态压力信号检测中的部分规律。
轴流压缩机出口总压和进口总压之比称为压比信号，是防喘振控制系统正常工作的另一个关键信号。由于压缩机发生喘振的最根本的内在原因还是气流分离造成的，因此压缩机入口截面气流的不均匀性会加快或加剧喘振的发生。而NF-6风洞轴流压缩机由于受到风洞结构的限制，其进口截面气流存在脉动和偏转，单纯采用测量总压而得到的压比信号作为压缩机防喘振控制参数，一方面会造成压缩机的运行范围变窄，压缩机效率下降；另一方面也会使风洞在同样的工况条件下能耗增大，造成不必要的浪费，使风洞实验的成本增加。因此，论文在国内首次提出在压缩机入口和出口安装动态压力传感器，得到这两个位置上的动态总压信号。为了得到相关规律，通过一个进气道实验模拟了动态压力测量的有关问题。同时应用小波变换分析了压缩机发生喘振后压力信号的奇异性特征，提出将其应用于NF-6风洞轴流压缩机防喘振系统中，可以提高控制系统的快速性和准确性，有效地拓宽NF-6风洞的运行范围。
4．分析了NF-6风洞轴流压缩机轴系的振动情况以及进气温度对压缩机防喘振曲线的影响，其结果指导了振动信号和温度信号检测系统的建设。
分析结果表明，NF-6风洞压缩机轴系的结构一阶固有振动频率落在相应于压缩机正常工作转速的激励范围之内。进一步的分析则显示，源于电机、压缩机和齿轮的一般激振，都不会引起较大的扭振应力，因此可以认为整个系统能在设计的转速范围内安全运行。但当高速齿轮存在缺陷时，两阶振型系数都较大，若压缩机转速对应的频率接近一阶固有振动频率时，可能会引起整个轴系的振动，从而造成整个结构的破坏。因此，为了保证NF-6风洞的安全运行，在此计算结果的指导下，我们建立了NF-6风洞压缩机振动监测系统。同时，由于压缩机进气温度变化范围很宽，还讨论了温度测量精度对整个系统的影响，据此设计了专用的温度检测和保护系统。这些措施在NF-6风洞建设过程中都已经得到了应用，为整个系统的成功调试打下了坚实的基础。
5．初步建立了系统的数学模型，将多变量耦合系统进行了解耦，给出了连续式风洞控制的一般规律。
通过分析指出，NF-6风洞控制系统是典型的多变量耦合系统，从输入和输出关系上是典型的多输入多输出(MIMO)模型，因此单纯采用PID控制无法满足系统控制的要求。但由于工程的特殊要求，在风洞调试和运转的初期，我们可以将MIMO系统进行解耦，分为四个单独的回路，采用传统的PID调节对风洞进行控制，在保证风洞安全调试和运行的前提下，通过大量的实验获取风洞运行过程中的有关参数，对其进行分析和处理后，逐步采用智能控制等更为高级的控制方式，提高风洞运转的效率，扩大运行范围。
以上理论分析结果和实际计算结果已经成功用于指导NF-6风洞轴流压缩机防喘振控制系统的建设，取得了较为满意的结果。作者同时希望本论文的研究结果能够为以后我国建设类似的风洞提供一定的参考和帮助。
The NF-6 wind tunnel is the first internationally advanced large-scale continuously-running high Reynolds number transonic airfoil wind tunnel in China, because the axial-flow compressor installed in the wind tunnel loop works under a condition of high pressure and low temperature, the flow at its entrance fluctuates greatly and would easily cause the compressor to go into surging zone, severely endangering the safety of the people working on site and make the equipment unable to work normally. When surge occurs in the compressor, the signals such as flow rate, pressure, temperature, vibration, etc., which symbolize and characterize the surge, will exhibit phenomena that are incredible and impossible under normal conventional conditions. The research on these phenomena in this dissertation is the first time in China and the references are seldom found internationally. Therefore detecting, analyzing, and predicting the signals are very important to preventing surge in scientific sense and in social value. However, the research on axial-flow surge phenomena as a cross discipline, involving aerodynamics, detecting and processing of signals, rotating machinery, control, etc., has to take many factors into account. This dissertation took the wind tunnel and the compressor as an integrated unity, exploiting the knowledge of many disciplines especially from the view of signal detecting and processing, mainly through the combination of theoretical analysis and experiments and based on the consideration of the coupling and interacting between the wind tunnel and the compressor, and produced following main creative contributions and results:
1. Through analysis the conditions under which the NF-6 wind tunnel and its axial-flow compressor can jointly run stably were obtained, laying a theoretical basis for the detection of the surge signal.
It was raised and put forward in the dissertation that such a complicated network of fluid flow system as the continuously-running high-speed wind tunnel be studied based on the concept of signal and system. Since the wind tunnel along with the compressor forms a closed loop, the working process of the whole system is a evolutionary process of mutually coupling and mutually interacting, which is the most significant difference between a compressor in the wind tunnel loop and that in used conventional industries. Therefore, the mutually coupling and interacting process must be taken into account when discussing the stable operating conditions and unstable working situations of the compressor in a wind tunnel. It was found from theoretical analysis that even if the compressor itself works in stable zones, the operating point of the system of the wind tunnel and the compressor might be in an unstable state and have a trend to shift to an unstable operating point of the compressor because of the coupling effect from the wind tunnel and thus, as a whole, the surge-preventing system of the compressor is a necessary precondition for the safe running of NF-6 wind tunnel.
2. By analyzing the manner to detect flow rate signal a constant-annular-area method was proposed to accurately measure the flow rate through the compressor while at the same time the effect of the wind tunnel total pressure and total temperature on the measurement precision must be counted in.
In view of that the flow rate signal is one of the key signals for the anti-surging control system of the axial-flow compressor to work normally in the NF-6 wind tunnel, the dissertation calculated the envelope of the range of flow rate values which the axial-flow compressor should obey when working in the NF-6 wind tunnel. At the same time, since the entrance flow entering the axial-flow compressor of the NF-6 wind tunnel is typical non-fully-developed pipe flow, the matured methods widely used in common industries for measuring flow rate are unable to be directly applied to the NF-6 wind tunnel, thus the dissertation proposed to obtain the accurate mass flow rate entering the axial-flow compressor by both measuring the average flow velocity at the cross section immediately behind the wind tunnel's first comer with a calibrated anemometer and, at the same time, measuring the total pressure and total temperature of the settling chamber of the wind tunnel. Additionally, in this dissertation, the effects of location, form, range and precision of transducers on the measured results were calculated and discussed in detail. These results gave guidance to the construction of the compressor flow rate measuring equipment in the NF-6 wind tunnel and were used to modify and improve some part of the predecessor design and contributed to the reduction of the risk of debugging, testing, and running of the equipment.
3. The dissertation validated the empirical formulae for estimating the cutoff frequency of the filter through the simulation test of pressure signal detection and the analysis of the test, analyzed the signal of abrupt pressure variation when surge occurs in the compressor using wavelet transform, and found some rules in detecting stable and dynamic pressure signals.
The pressure ratio signal defined as the ratio of the outlet total pressure to the inlet total pressure of the axial-flow compressor is another key signal on which the anti-surging control system of the axial-flow compressor relies to work normally in NF-6 wind tunnel. Since the most leading and fundamental inherent cause of surging in compressor is flow separation, thus the flow non-uniformity at the compressor entrance section would accelerate or aggravate the occurring of surging. Because of the structural limitations of the wind tunnel there exist flow parameter fluctuations and flow deflections in the entrance section of the axial-flow compressor of the NF-6 wind tunnel. Therefore adopting the pressure ratio signal obtained by only measuring total pressure as the surging-preventing control parameter of the compressor will, on the one hand, lead to narrower compressor operating zone and lower compressor efficiency, while on the other hand, increase the energy consumption and the cost of the wind tunnel tests under the same operating conditions, hence cause unnecessary waste. In this dissertation a new concept was put forward for the first time nationwide that dynamic pressure sensors be installed both in the entrance and exit of the compressor to obtain the dynamic total pressure signals at the two positions. To acquire the corresponding rules, the issues associated with measuring the dynamic pressure were simulated by an inlet test. Furthermore the singularity characteristics of the pressure signals when were analyzed using wavelet transformation when surge occurs in the compressor, which was proposed to be applied in the design of the surging-preventing system of the axial-flow compressor in the NF-6 wind tunnel to improve the rapidity and accuracy of the control system and to widen the operating envelope of the wind tunnel.
4. This dissertation discussed the vibration of the axial-flow compressor's shafting of the NF-6 wind tunnel and the influence of the entering flow temperature of NF-6 wind tunnel on the surging-preventing curve, the results from the discussion gave a good guidance to the construction of the detecting and measuring system of vibration signals and temperature signals.
The analysis indicated that the first-order natural vibration frequency of the structure of the axial-flow compressor's shafting in NF-6 wind tunnel fell within the excitation region corresponding to the normal working rotating speed of the compressor. Further analysis showed that the general excitation originating from motors, compressors and gears would not cause large stress of twisting vibration. Therefore the whole system could be considered to be able to run safely in the designed range of rotating speed. But if the high-speed gears have flaws the coefficients of the vibration mode of both orders are large. Meanwhile, if the frequency corresponding to the rotating speed of the compressor is approaching the first-order natural vibration frequency the whole compressor's shafting system may vibrate and hence cause the destruction of the whole structure. Therefore, in order to ensure that the whole NF-6 wind tunnel run safely a monitoring system of compressor vibration was set up guided by the above computed results. Additionally, considering the influence the entering temperature of the compressor on the surging-preventing curve, the effect of the measuring precision of temperature on the whole system was discussed and a temperature detecting, checking and protecting system was designed based on the discussions. All these measures have been utilized in the construction of NF-6 wind tunnel and laid a solid foundation for the successful running of the whole system.
5. The dissertation set up an elementary mathematic model for the system, decoupled the multi-argument coupling system, and gave a generally applicable rule for the control of continuously-running wind tunnel.
Analysis shows that the control system of NF-6 wind tunnel was a typical multi-argument coupling system and a typical model of multi-input and multi-output (MIMO) as viewed from the input and output. Therefore the adopting purely PID control did not meet the requirements of the NF-6 wind tunnel system control. However, from the point of view of engineering in the early period of debugging and operating of the wind tunnel, the MIMO system can be decoupled into four single separate loops and then the wind tunnel can be controlled with traditional PID technique. With the safe debugging and running of the wind tunnel guaranteed, the parameters associated with the wind tunnel running process were acquired through a massive number of wind tunnel tests. After accumulating, analyzing and processing these parameters, more advanced control mode such as artificial intelligence control could be put into use gradually to improve the running efficiency and broaden the running envelop of the wind tunnel.
The above-mentioned results theoretical analysis and practical computations have been successfully used to guide the construction of the anti-surge system of the axial-flow compressor in the NF-6 wind tunnel and led to very satisfactory effects. The author hopes that the results in this dissertation will be helpful valuable to the construction of the similar wind tunnels in our country in the future.
引文
[1] 黄钟岳，王晓放．透平式压缩机．北京，化学工业出版社．2004．8
[2] 周渭，于建国，刘海霞．测试与计量技术基础．西安，西安电子科技大学出版社．2004．4
[3] 张贤达，保铮．非平稳信号分析与处理．北京，国防工业出版社．2001．7
[4] 王志清．透平压缩机的调节运行与振动．北京，机械工业出版社，1996．1
[5] 赵庆海主编．测试技术与工程应用．北京，化学工业出版社．2005．5
[6] 杜水友主编．压力测量技术及仪表．北京，机械工业出版社．2005．7
[7] 姜忠良，陈秀云．温度的测量与控制．北京，清华大学出版社．2005．9
[8] 李士勇．模糊控制、神经控制和智能控制论．哈尔滨，哈尔滨工业大学出版社．2002．6
[9] 华绍曾等．实用流体阻力手册．北京，国防工业出版社．1985
[10] 超声速扩压器．1961年俄文翻译版．
[11] 航空技术研究所2m×2m跨声速风洞的设计和构造．NAR-TR-25．1962
[12] 范洁川等．世界风洞．北京，航空工业出版社．1992
[13] 梁国伟，蔡武昌．流量测量技术及仪表．北京，机械工业出版社．2002．6
[14] 许第昌．压力计量测试．北京，中国计量出版社．1988．10
[15] 叶长青，张琪．轴流压缩机与离心压缩机综合比较．流体机械，1998年12月，第26卷第12期：21～24
[16] 刘波，王掩刚，赵旭民．轴流压缩机出口级附近流道形状变化对静子叶片气动性能的影响．流体机械．2000年第28卷第4期：20～24
[17] 竹朝霞，惠增宏，尹迪义．NF-6风洞压缩机监控系统的研制．流体力学实验与测量．2003年第17卷：112～115
[18] 高永卫，肖春生，金承信．在NF-6增压连续式回流高速风洞中轴流压缩机启动问题的探讨．2003第17卷：11～14
[19] 唐汇虎，余庭辉．AV50-11轴流压缩机的防喘振控制．压缩机技术，2001年第4期(总第168期)：11～14
[20] 魏宗宪，卢世忠，李白荣．大型透平式压缩机防喘振控制及应用．化工机械，2002年第29卷第5期：301～305
[21] 姜国栋，张三元，史晓云．大型轴流压缩机国产化设计的新发展．风机 技术．2001年第4期：3～5
[22] 李江平．相似原理在压缩机防喘振控制中的应用．大氮肥．1997年第20卷第5期：294～296
[23] 阎斌．压缩机喘振控制方法．应用科技．1998年第6期：4
[24] 侯平．压缩机的喘振与3C防喘振控制器的基本控制．石油化工自动化．2001年第4期：28～31
[25] 张春梅，安卫平．压缩机防喘振自控系统研制．自动化仪表．2000年第21卷第1期：41～42
[26] R. Kurz, R. C. White, Surge Avoidance in Gas Compression System, Journal of Turbomachinery, October 2004, Vol. 126/501.
[27] Roland A, Keith M, Kelly Cohen, Wind Tunnel Compressor Optimization and Efficiency Roll-Off Prediction Using Fuzzy Logic, AIAA2003-0089.
[28] Roland A, Frank W, Neural Network Application for Optimizing Multi-Stage Wind Tunnel Compressor Efficiency, AIAA2002-0308.
[29] Norikazu Sudani, Mamoru Sato, etc. Assessment of Sidewall Interference in a Two-Dimensional Transonic Wind Tunnel, AIAA2001-2423.
[30] Dale L, Kevin M, Dustin S, A Review of Standards Model Testing in the Boeing Transonic Wind Tunnel (Invited), AIAA2002-2789.
[31] Renato Paciorri, Filippo Sabetta, Wave Reflection on Porous Walls: Numerical Modeling and Application to Transonic Wind Tunnels, AIAA2002-1060.
[32] Max. A, Sreedhara V, Flow Quality Measurements in the NASA-Ames Upgraded 11-by 11-ft Transonic Wind Tunnel (Invited Paper), AIAA2000-2681.
[33] Thombi Layukallo, Yoshiaki Nakamura, The Efficiency of a Second Throat in Stabilizing the Flow in a Transonic Wind Tunnel, AIAA99-3413.
[34] T. Wolf, State of the Art in Circuit Loss Analysis of Transonic Wind Tunnel. 1998.
[35] Tsuguji Nakano, Hidekazu Kodama, Numeric and Experimental Investigation of Instability Inception in a High-speed Axial-Flow Compressor Using Chaos Theory, AIAA2002-4089.
[36] Hideo Sawada, Takashi Kohno, Tetsuya Kunimasu, Pressure Fluctuation Measurements in the NAL 0.2-m Supersonic Wind Tunnel, AIAA98-0636.
[37] Wing-fai Ng, Joel C, A Fast-Response Aspirating Probe for Measurements of Total Temperature and Pressure in Transonic Cryogenic Wind Tunnel, AIAA-86-0126.
[38] R. N. Gamache, E. M. Greitzer, Reverse Flow in Multistage Axial Compressors, AIAA-86-1747.
[39] J. Dugundji, A. H. Epstein, V. Garnier, E. M. Greitzer, G. Guenette, A Progress Report on Active Control of Flow Instabilities: Rotating Stall Stabilization in Axial Compressors, AIAA-89-1008.
[40] H. Ishii, Y. Kashiwabara, Surge and Rotating Stall in Axial Compressors, AIAA 89-2683.
[41] H. Ishii, Y. Kashiwabara, A Numeric Study of Surge and Rotating Stall in Axial Compressors, AIAA 91-1896.
[42] H. Ishii, Y. Kashiwabara, Numeric Investigation of Surge and Rotating Stall in Multistage Axial Compressors, AIAA 92-3193.
[43] James A., Martindale, D. W. Wanger, Performance Evaluation of a Transonic Wind Tunnel Compressor, AIAA 92-3927.
[44] Nhan Nguyen, Doug Muzzio, Roy Guist, Experimental Investigation of the Rotor Blade Vibration in the Three-Stage Compressor of the 11-By 11-Foot Transonic Wind Tunnel, AIAA 95-3139.
[45] W. Vikatos, O. C. Jadayel, H. Ekerol, Three Dimensional Flow Measurements Down-Stream of an Axial Compressor Rotor Using a Four-hole Pressure Probe, AIAA 98-0972.
[46] Craig A. Matthew A. Sanford Fleeter. The Design of a Control Law for Rotating Stall in an Axial Compressor, AIAA-98-3311.
[47] Daniel J. Domey, Om P. Sharma. The Effects of Thermal Gradients On the Performance of Axial Compressors, AIAA-98-3412.
[48] Assaad Krichene, J. V. R. Prasad, Synthesis of Adaptive Neural Network / Fuzzy Logic Based Rotating Stall Controllers for Axial Compressor, AIAA 99-2126.
[49] Charles A., Harold F., Richard G.., NASA Low-Speed Axial Compressor for Fundamental Research, NASA TM 4635.
[50] Guoxiang Gu, Andrew Sparks, Siva Banda, Bifurcation Based Nonlinear Feedback Control for Rotating Stall in Axial Compressor, DRAFT SF 298.
[51] Teri Kaiser, Reed S. Hansen, Experimental/Analytical Approach to Understanding Mistuning in a Transonic Wind Tunnel Compressor, NASA TM 108833.
[52] Pierce L. Lawing, Jerry B. Adcock, Charles L. Ladson, A Fan Pressure Ratio Correlation in Terms of Mach Number and Reynolds Number for the Langley 0.3-Meter Transonic Cryogenic Tunnel, NASA TP 1752.
[53] Blair B. Gloss, Donna Nystrom, Estimation of Fan Pressure ratio Requirements and Operation Performance for the National Transonic Facility, NASA TM 81802.
[54] Shahnaz M. Puniani, Predication of an Axial Compressor Stage Characteristic From a One-Point Measurement, AFIT/GAE/ENY/95D-21.
[55] 刘原，李鹏飞，王晓娟．应用可编程型单回路调节器防止轴流压缩机喘振．西北纺织工学院学报．2000年第14卷第2期 (总第54期)：180～184
[56] 杨忠学，孙大生．轴流式压缩机喘振特性的研究．流体机械．1997年第25卷第4期：3～6
[57] 徐晓秋，梅筱琴，陈永健．轴流式压缩机防喘振控制系统的研制．四川大学学报 (工程科学版)．2002年第34卷第1期：85～88
[58] 王群．轴流压缩机TURBOLOG防喘振控制系统分析研究．流体机械．2002年第30卷第10期：20～22
[59] 李怀庆．轴流压缩机喘振初期现象的特性分析．流体机械．2001年第29卷第11期：31～34
[60] 王双群，程力．轴流压缩机的保护控制．能源研究与利用．2000年第6期：13～15
[61] 李怀庆．轴流压缩机喉部压差／常数C的确定及应用．风机技术．2002年第2期：24～25
[62] 惠增宏，何明一．NF-6风洞马赫数控制系统研制．流体力学实验与测量． 2005．Vol．19 No．1：61～65
[63] 惠增宏，何明一．NF-6风洞压缩机及驱动系统研制．流体力学实验与测量，2005．Vol．19 No．4：31～35
[64] 惠增宏，竹朝霞，张理．进气道实验中动态数据采集及处理方法的研究．实验力学．2002．Vol．17 No．1：48～54
[65] 刘大响，叶培梁．俄罗斯的发动机进口流场畸变评定指南．燃气涡轮实验与研究，1994年第3期．
[66] 鲍国华主编．风洞特种实验．西安：西北工业大学出版社．1990年5月：1～9
[67] 范立钦，周鼎义合编．飞机空气动力学．西安：西北工业大学出版社．1989年9月：225～P230
[68] 胡广书．数字信号处理—理论、算法与实现．北京：清华大学出版社．1999年9月：55～84
[69] 成礼智，王红霞，罗永．小波的理论与应用．北京：科学出版社，2004．
[70] 李弼程，罗建书．小波分析及其应用．北京：电子工业出版社，2003．
[71] 彭玉华．小波变换与工程应用．北京：科学出版社，1999．
[72] Ingrid Daubechies著，李建平，杨万年译．小波十讲．北京：国防工业出版社，2004．
[73] Rafael C．Gonzalez著，阮秋琦，阮宇智译，数字图像处理(第二版)．北京：电子工业出版社，2003．
[74] 程云鹏．矩阵论(第二版)．西安：西北工业大学出版社，2001．
[75] 程正兴，杨守志．多小波理论的发展与研究．工程数学学报，Vol．18 No．5Dec．2001：1～16
[76] 程正兴，张玲玲．多小波分析与应用．工程数学学报，Vol．18 No．1Feb 2001．99～107
[77] Strela V., Multiwavelets: Theory and Application, Ph. D. Thesis, MIT, 1996.
[2] 周渭，于建国，刘海霞．测试与计量技术基础．西安，西安电子科技大学出版社．2004．4
[3] 张贤达，保铮．非平稳信号分析与处理．北京，国防工业出版社．2001．7
[4] 王志清．透平压缩机的调节运行与振动．北京，机械工业出版社，1996．1
[5] 赵庆海主编．测试技术与工程应用．北京，化学工业出版社．2005．5
[6] 杜水友主编．压力测量技术及仪表．北京，机械工业出版社．2005．7
[7] 姜忠良，陈秀云．温度的测量与控制．北京，清华大学出版社．2005．9
[8] 李士勇．模糊控制、神经控制和智能控制论．哈尔滨，哈尔滨工业大学出版社．2002．6
[9] 华绍曾等．实用流体阻力手册．北京，国防工业出版社．1985
[10] 超声速扩压器．1961年俄文翻译版．
[11] 航空技术研究所2m×2m跨声速风洞的设计和构造．NAR-TR-25．1962
[12] 范洁川等．世界风洞．北京，航空工业出版社．1992
[13] 梁国伟，蔡武昌．流量测量技术及仪表．北京，机械工业出版社．2002．6
[14] 许第昌．压力计量测试．北京，中国计量出版社．1988．10
[15] 叶长青，张琪．轴流压缩机与离心压缩机综合比较．流体机械，1998年12月，第26卷第12期：21～24
[16] 刘波，王掩刚，赵旭民．轴流压缩机出口级附近流道形状变化对静子叶片气动性能的影响．流体机械．2000年第28卷第4期：20～24
[17] 竹朝霞，惠增宏，尹迪义．NF-6风洞压缩机监控系统的研制．流体力学实验与测量．2003年第17卷：112～115
[18] 高永卫，肖春生，金承信．在NF-6增压连续式回流高速风洞中轴流压缩机启动问题的探讨．2003第17卷：11～14
[19] 唐汇虎，余庭辉．AV50-11轴流压缩机的防喘振控制．压缩机技术，2001年第4期(总第168期)：11～14
[20] 魏宗宪，卢世忠，李白荣．大型透平式压缩机防喘振控制及应用．化工机械，2002年第29卷第5期：301～305
[21] 姜国栋，张三元，史晓云．大型轴流压缩机国产化设计的新发展．风机 技术．2001年第4期：3～5
[22] 李江平．相似原理在压缩机防喘振控制中的应用．大氮肥．1997年第20卷第5期：294～296
[23] 阎斌．压缩机喘振控制方法．应用科技．1998年第6期：4
[24] 侯平．压缩机的喘振与3C防喘振控制器的基本控制．石油化工自动化．2001年第4期：28～31
[25] 张春梅，安卫平．压缩机防喘振自控系统研制．自动化仪表．2000年第21卷第1期：41～42
[26] R. Kurz, R. C. White, Surge Avoidance in Gas Compression System, Journal of Turbomachinery, October 2004, Vol. 126/501.
[27] Roland A, Keith M, Kelly Cohen, Wind Tunnel Compressor Optimization and Efficiency Roll-Off Prediction Using Fuzzy Logic, AIAA2003-0089.
[28] Roland A, Frank W, Neural Network Application for Optimizing Multi-Stage Wind Tunnel Compressor Efficiency, AIAA2002-0308.
[29] Norikazu Sudani, Mamoru Sato, etc. Assessment of Sidewall Interference in a Two-Dimensional Transonic Wind Tunnel, AIAA2001-2423.
[30] Dale L, Kevin M, Dustin S, A Review of Standards Model Testing in the Boeing Transonic Wind Tunnel (Invited), AIAA2002-2789.
[31] Renato Paciorri, Filippo Sabetta, Wave Reflection on Porous Walls: Numerical Modeling and Application to Transonic Wind Tunnels, AIAA2002-1060.
[32] Max. A, Sreedhara V, Flow Quality Measurements in the NASA-Ames Upgraded 11-by 11-ft Transonic Wind Tunnel (Invited Paper), AIAA2000-2681.
[33] Thombi Layukallo, Yoshiaki Nakamura, The Efficiency of a Second Throat in Stabilizing the Flow in a Transonic Wind Tunnel, AIAA99-3413.
[34] T. Wolf, State of the Art in Circuit Loss Analysis of Transonic Wind Tunnel. 1998.
[35] Tsuguji Nakano, Hidekazu Kodama, Numeric and Experimental Investigation of Instability Inception in a High-speed Axial-Flow Compressor Using Chaos Theory, AIAA2002-4089.
[36] Hideo Sawada, Takashi Kohno, Tetsuya Kunimasu, Pressure Fluctuation Measurements in the NAL 0.2-m Supersonic Wind Tunnel, AIAA98-0636.
[37] Wing-fai Ng, Joel C, A Fast-Response Aspirating Probe for Measurements of Total Temperature and Pressure in Transonic Cryogenic Wind Tunnel, AIAA-86-0126.
[38] R. N. Gamache, E. M. Greitzer, Reverse Flow in Multistage Axial Compressors, AIAA-86-1747.
[39] J. Dugundji, A. H. Epstein, V. Garnier, E. M. Greitzer, G. Guenette, A Progress Report on Active Control of Flow Instabilities: Rotating Stall Stabilization in Axial Compressors, AIAA-89-1008.
[40] H. Ishii, Y. Kashiwabara, Surge and Rotating Stall in Axial Compressors, AIAA 89-2683.
[41] H. Ishii, Y. Kashiwabara, A Numeric Study of Surge and Rotating Stall in Axial Compressors, AIAA 91-1896.
[42] H. Ishii, Y. Kashiwabara, Numeric Investigation of Surge and Rotating Stall in Multistage Axial Compressors, AIAA 92-3193.
[43] James A., Martindale, D. W. Wanger, Performance Evaluation of a Transonic Wind Tunnel Compressor, AIAA 92-3927.
[44] Nhan Nguyen, Doug Muzzio, Roy Guist, Experimental Investigation of the Rotor Blade Vibration in the Three-Stage Compressor of the 11-By 11-Foot Transonic Wind Tunnel, AIAA 95-3139.
[45] W. Vikatos, O. C. Jadayel, H. Ekerol, Three Dimensional Flow Measurements Down-Stream of an Axial Compressor Rotor Using a Four-hole Pressure Probe, AIAA 98-0972.
[46] Craig A. Matthew A. Sanford Fleeter. The Design of a Control Law for Rotating Stall in an Axial Compressor, AIAA-98-3311.
[47] Daniel J. Domey, Om P. Sharma. The Effects of Thermal Gradients On the Performance of Axial Compressors, AIAA-98-3412.
[48] Assaad Krichene, J. V. R. Prasad, Synthesis of Adaptive Neural Network / Fuzzy Logic Based Rotating Stall Controllers for Axial Compressor, AIAA 99-2126.
[49] Charles A., Harold F., Richard G.., NASA Low-Speed Axial Compressor for Fundamental Research, NASA TM 4635.
[50] Guoxiang Gu, Andrew Sparks, Siva Banda, Bifurcation Based Nonlinear Feedback Control for Rotating Stall in Axial Compressor, DRAFT SF 298.
[51] Teri Kaiser, Reed S. Hansen, Experimental/Analytical Approach to Understanding Mistuning in a Transonic Wind Tunnel Compressor, NASA TM 108833.
[52] Pierce L. Lawing, Jerry B. Adcock, Charles L. Ladson, A Fan Pressure Ratio Correlation in Terms of Mach Number and Reynolds Number for the Langley 0.3-Meter Transonic Cryogenic Tunnel, NASA TP 1752.
[53] Blair B. Gloss, Donna Nystrom, Estimation of Fan Pressure ratio Requirements and Operation Performance for the National Transonic Facility, NASA TM 81802.
[54] Shahnaz M. Puniani, Predication of an Axial Compressor Stage Characteristic From a One-Point Measurement, AFIT/GAE/ENY/95D-21.
[55] 刘原，李鹏飞，王晓娟．应用可编程型单回路调节器防止轴流压缩机喘振．西北纺织工学院学报．2000年第14卷第2期 (总第54期)：180～184
[56] 杨忠学，孙大生．轴流式压缩机喘振特性的研究．流体机械．1997年第25卷第4期：3～6
[57] 徐晓秋，梅筱琴，陈永健．轴流式压缩机防喘振控制系统的研制．四川大学学报 (工程科学版)．2002年第34卷第1期：85～88
[58] 王群．轴流压缩机TURBOLOG防喘振控制系统分析研究．流体机械．2002年第30卷第10期：20～22
[59] 李怀庆．轴流压缩机喘振初期现象的特性分析．流体机械．2001年第29卷第11期：31～34
[60] 王双群，程力．轴流压缩机的保护控制．能源研究与利用．2000年第6期：13～15
[61] 李怀庆．轴流压缩机喉部压差／常数C的确定及应用．风机技术．2002年第2期：24～25
[62] 惠增宏，何明一．NF-6风洞马赫数控制系统研制．流体力学实验与测量． 2005．Vol．19 No．1：61～65
[63] 惠增宏，何明一．NF-6风洞压缩机及驱动系统研制．流体力学实验与测量，2005．Vol．19 No．4：31～35
[64] 惠增宏，竹朝霞，张理．进气道实验中动态数据采集及处理方法的研究．实验力学．2002．Vol．17 No．1：48～54
[65] 刘大响，叶培梁．俄罗斯的发动机进口流场畸变评定指南．燃气涡轮实验与研究，1994年第3期．
[66] 鲍国华主编．风洞特种实验．西安：西北工业大学出版社．1990年5月：1～9
[67] 范立钦，周鼎义合编．飞机空气动力学．西安：西北工业大学出版社．1989年9月：225～P230
[68] 胡广书．数字信号处理—理论、算法与实现．北京：清华大学出版社．1999年9月：55～84
[69] 成礼智，王红霞，罗永．小波的理论与应用．北京：科学出版社，2004．
[70] 李弼程，罗建书．小波分析及其应用．北京：电子工业出版社，2003．
[71] 彭玉华．小波变换与工程应用．北京：科学出版社，1999．
[72] Ingrid Daubechies著，李建平，杨万年译．小波十讲．北京：国防工业出版社，2004．
[73] Rafael C．Gonzalez著，阮秋琦，阮宇智译，数字图像处理(第二版)．北京：电子工业出版社，2003．
[74] 程云鹏．矩阵论(第二版)．西安：西北工业大学出版社，2001．
[75] 程正兴，杨守志．多小波理论的发展与研究．工程数学学报，Vol．18 No．5Dec．2001：1～16
[76] 程正兴，张玲玲．多小波分析与应用．工程数学学报，Vol．18 No．1Feb 2001．99～107
[77] Strela V., Multiwavelets: Theory and Application, Ph. D. Thesis, MIT, 1996.