Compressor rotor assembly quality online detection method and system

Abstract

The invention discloses a compressor rotor assembly quality on-line detection method and system, and relates to the technical field of compressor component assembly detection.The compressor rotor assembly quality on-line detection method comprises the steps that microgravity operation data of a compressor are collected, artificial disturbance simulation is conducted, and assembly microdefect response signals are induced; in combination with real-time axial micro-vibration feature extraction and quantitative calculation of an oil film formation delay factor Ym, stability and hysteresis of oil film response are accurately described, a modal frequency aggregation trend is further identified through an aggregation trend factor Jq, a structure coupling precursor behavior is captured in advance, and finally, a multi-dimensional feature fusion comprehensive scoring mode is used to identify a modal frequency aggregation trend. And intelligent grade evaluation and early warning output are carried out on the rotor assembly quality, so that the online detection of the compressor rotor assembly quality has the beneficial effects of high early diagnosis capability, high tiny defect identification precision, clear structural response analysis and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of compressor component assembly detection, and in particular to an online detection method and system for compressor rotor assembly quality. Background Art

[0002] Compressors, as core actuators in power and refrigeration systems, are widely deployed in energy, rail transit, petrochemical, and tunneling scenarios. During operation, they are subjected to high-frequency thermal cycles, starting load disturbances, and superposition of structural stresses. Assembly accuracy directly affects operational stability and lifespan reliability. This is especially true for high-speed, heavy-duty compressors. If the rotor structure exhibits problems such as assembly eccentricity, uneven support, and abnormal initial preload during assembly, these issues can easily induce hysteresis in the oil film pressure-bearing process, vibration mode aliasing, and even axial resonance, seriously interfering with the stable operation of the control system. Currently, online assessment of rotor assembly status still primarily relies on terminal vibration threshold alarm mechanisms, which are difficult to meet the needs of early identification of structural anomalies. Therefore, an integrated assembly quality inspection technology system is needed that integrates disturbance simulation, vibration signal feature extraction, and resonance trend assessment.

[0003] In a Chinese invention patent application with publication number CN108932712A, a rotor winding quality inspection system is described. The system includes: an image preprocessing module that segments the rotor winding image using image template positioning to obtain a set of images of the parts to be inspected, labels each image of the parts to be inspected as qualified or unqualified, and uses the labeled set of images of the parts to be inspected as a training set; a feature extraction module that uses the LBP method to extract the texture of each image of the parts to be inspected; a neural network training module that uses a convolutional neural network to further extract target features from the texture features of the rotor winding images extracted by the LBP operator to distinguish the rotor winding quality and learn network parameters; and an image inspection module that uses the trained neural network to perform quality inspection on the rotor winding images to be inspected. By learning the rotor winding texture information and continuously optimizing the weight parameters of the network system, the detection and recognition rate of the rotor winding qualification is greatly improved, especially for components with poor reflection.

[0004] The above method greatly improves the detection and recognition rate of rotor winding qualifications by learning the rotor winding texture information and continuously optimizing the weight parameters of the network system. However, in addition to this, in the existing online detection method of compressor rotor assembly quality, it is usually evaluated through manual measurement of fitting clearance, vibration amplitude monitoring under no-load test, or motor starting current analysis.

[0005] However, this method mostly relies on experience or indirect judgment of macro signals, and it is difficult to accurately identify the weak response in the early stage of oil film establishment or structural deviations caused by assembly. There are problems such as response lag, low recognition accuracy and insufficient early warning capabilities. In actual applications, compressors are deployed in plateau areas or areas with unstable environmental parameters, such as areas with drastic altitude changes or low air pressure. The microgravity conditions in such environments will cause the inertial response characteristics of the compressor system to change, manifested as premature oil film formation, excessive torque output, excessive disturbance loading excitation, etc., which in turn masks the weak micro-vibration signals caused by assembly deviations of the rotor structure, making it difficult to accurately identify assembly defects such as loose bearing assembly, end face contact eccentricity and initial imbalance, resulting in axial resonance of the compressor rotor, oil film collapse or component fatigue failure, which in turn induces unstable operation of the entire machine, surge in vibration, reduced energy efficiency and even catastrophic damage.

[0006] To this end, the present invention provides a method and system for online detection of compressor rotor assembly quality. Summary of the Invention

[0007] In view of the deficiencies in the prior art, the present invention provides a method and system for online detection of compressor rotor assembly quality, which solves the problems in the above-mentioned background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: a compressor rotor assembly quality online detection system, including a disturbance simulation loading module, a micro-vibration monitoring and analysis module, an oil film delay analysis module, an aggregation trend analysis module and a rotor assembly stability early warning module; The disturbance simulation loading module is used to collect the microgravity operation data of the compressor, perform microgravity operation simulation, and construct a disturbance sampling window, which is then sent to the microvibration monitoring and analysis module for data collection and analysis. The micro-vibration monitoring and analysis module is used to collect the axial acceleration response signal of the compressor rotor within the disturbance sampling window in real time. , and perform feature extraction to obtain a feature data set; The oil film delay analysis module is used to calculate the oil film formation delay factor Ym based on the characteristic data set, evaluate the oil film delay risk level, and trigger the aggregation trend analysis module; The aggregation trend analysis module is used to analyze the frequency domain aggregation trend and calculate the aggregation trend factor Jq; The rotor assembly stability warning module is used to calculate the rotor assembly comprehensive score Rs based on the oil film formation delay factor Ym and the aggregation tendency factor Jq, evaluate the compressor rotor assembly quality, and issue an assembly quality warning.

[0009] Preferably, the disturbance simulation loading module is used to collect microgravity operation data at the location of the compressor, including the altitude of the location of the compressor. and atmospheric pressure , and calculate the impact of microgravity operation data on the working state of the compressor, and obtain the equivalent gravity correction coefficient , where the equivalent gravity correction coefficient is The specific way to obtain it is: ; Where, Indicates the altitude where the compressor is located. Indicates the standard reference altitude, Indicates the atmospheric pressure at the location of the compressor, represents standard atmospheric pressure, represents the power transformation factor; Based on the equivalent gravity correction factor , calculate and obtain the compressor control parameter set, wherein the compressor control parameter set includes the initial maximum acceleration value , load limit torque upper limit and ramp-up time constant τ; Initial maximum acceleration value The specific acquisition method is as follows: ; Where, represents the standard acceleration due to gravity; Load limit torque upper limit The specific acquisition method is as follows: ; Where, Indicates the minimum torque output ratio of the compressor, represents the environmental disturbance sensitivity amplification factor, Indicates the maximum output torque value of the compressor under rated voltage, rated frequency, stable load, standard ambient temperature and normal cooling conditions; The specific method of obtaining the ramp-up time constant τ is as follows: ; Where, Indicates the standard ramp-up time. represents the disturbance time sensitivity amplification factor; Based on the generated compressor control parameter set, the compressor control parameter set is transmitted to the PLC industrial control system through the standard industrial communication protocol, and the PLC industrial control system is used to map the compressor control parameters into an internal execution control curve function, and automatically output the compressor dynamic control instructions. According to the compressor dynamic control instructions, the compressor is driven to perform disturbance simulation, and the slow rise time constant τ is used as the upper limit to set the disturbance sampling window [ τ].

[0010] Preferably, the micro-vibration monitoring and analysis module includes a spectrum analysis unit and a feature extraction unit; The spectrum analysis unit is used to deploy the axial acceleration sensor on the compressor and detect the axial acceleration in the disturbance sampling window [ τ], the axial acceleration sensor is used to collect the axial acceleration response signal of the compressor rotor in real time , and respond to the collected axial acceleration signal Perform preprocessing, including normalization and wavelet denoising; According to the pre-processed axial acceleration response signal , perform fast Fourier transform FFT to obtain the frequency spectrum of the compressor rotor, wherein the specific process of fast Fourier transform FFT is as follows: ; In the formula, A( ) indicates the compressor rotor at the frequency point The frequency domain energy density at represents the fast Fourier transform, represents the axial acceleration response signal of the compressor rotor at time point t, t∈[ τ], represents the i-th discrete frequency point; And based on the frequency spectrum of the compressor rotor, a frequency data set S is constructed, wherein the frequency data set S includes the frequency domain energy density of each frequency point.

[0011] Preferably, the feature extraction unit is used to Axial acceleration response signal , perform integral summation to obtain the disturbance sampling window [ The disturbance energy response factor NL of the energy intensity of the compressor rotor structure response in τ] is obtained as follows: ; Where, represents the axial acceleration response signal of the compressor rotor at time point t, represents the start time of disturbance loading, Indicates the ramp-up time constant; Based on the frequency data set S, identify the main frequency points in the frequency data set S , and the main frequency point Axial acceleration response signal Use bandpass filtering to obtain the main frequency response component , and the main frequency response component Perform Hilbert transform to obtain the main frequency response component The instantaneous phase φ , and calculate the main frequency response components The instantaneous phase change of is obtained to obtain the phase change factor Xw, wherein the phase change factor Xw is obtained in the following manner: ; And based on the obtained disturbance energy response factor NL and phase change factor Xw, a feature data set is constructed.

[0012] Preferably, the oil film delay analysis module includes an oil film formation delay factor calculation unit and an early warning unit; The oil film formation delay factor calculation unit is used to calculate the oil film formation delay factor in the disturbance sampling window [ τ], from the starting interference time point At the beginning, take a length of Time sliding window to obtain several lengths of Time sliding window; For each time sliding window, the main frequency point is identified, and the main frequency response component and instantaneous phase of each time sliding window are obtained to obtain the phase change factor Xw of each time sliding window, and the phase change factor Xw of all time sliding windows is integrated to obtain the standard deviation of the phase change factor ; Set the phase change stability threshold Xwyz. For each time sliding window, if the phase change factor standard deviation is less than the phase change stability threshold Xwyz, then the marking time sliding window is a stable window. If the phase change factor standard deviation If it is greater than or equal to the phase change stability threshold Xwyz, no processing is required; Set the continuous stability window threshold , for each time sliding window obtained, if the continuous If all the time sliding windows are stable windows, it is determined that the compressor has entered the main frequency response stable state, and the continuous The starting time point of the initial time sliding window in the time sliding window is the time point when the main frequency response is stable ; Based on the characteristic data set and the main frequency response stable time point , quantify the lag degree of oil film establishment of the compressor rotor after the disturbance loading process to obtain the oil film formation delay factor Ym, where the oil film formation delay factor Ym is specifically obtained as follows: ; Where, Indicates the time point when the main frequency response stabilizes, Indicates the standard response time point, represents the disturbance energy response factor, Indicates the time point when the main frequency response stabilizes The phase change factor of the compressor is represents the adjustment factor, ln represents the natural logarithm function, represents the coupled inertial damping constant.

[0013] Preferably, the early warning unit is used to preset a first delay risk threshold and the second delay risk threshold and the oil film formation delay factor Ym is combined with the first delay risk threshold and the second delay risk threshold Conduct comparative analysis and assess the oil film delay risk level. The specific assessment contents are as follows: If the oil film formation delay factor Ym is less than or equal to the first delay risk threshold , then the oil film is judged to be in a normal formation state. At this time, the oil film delay risk level is judged to be the first risk level and no treatment is required; If the oil film formation delay factor Ym is greater than the first delay risk threshold , and is less than the second delay risk threshold , then the oil film is judged to be in a state where delayed formation is allowed. At this time, the oil film delay risk level is judged to be the second risk level, and the rotor operation status needs to be continuously monitored; If the oil film formation delay factor Ym is greater than or equal to the second delay risk threshold , the oil film is judged to be in an abnormally delayed state. At this time, the oil film delay risk level is judged to be the third risk level. The compressor rotor assembly torque is abnormal, there is a stress concentration point, and the lubrication system is partially blocked and the oil supply pressure is insufficient, which automatically triggers the aggregation trend analysis module.

[0014] Preferably, the aggregation trend analysis module includes a frequency domain energy analysis unit and an aggregation trend index calculation unit; The frequency domain energy analysis unit is used to calculate the energy of each frequency point based on the frequency data set S. and main frequency point The absolute distance between , and form a frequency difference vector; And based on the frequency difference vector, perform global statistical analysis and calculate the frequency mean and frequency fluctuation variance .

[0015] Preferably, the clustering trend indicator calculation unit is used to calculate the frequency difference vector and the frequency fluctuation variance. , perform summary calculation to obtain the aggregation trend factor Jq, where the specific method of obtaining the aggregation trend factor Jq is: ; Where, Indicates the main frequency point The frequency domain energy density, Indicates frequency point and main frequency point The absolute distance between , represents the frequency fluctuation variance, represents the i-th frequency point, Indicates frequency point The frequency domain energy density, n represents the total number of frequency points, i represents the frequency point index, i={1, 2, 3, ..., n}, Represents the correction constant.

[0016] Preferably, the rotor assembly stability warning module is used to perform summary calculation based on the aggregation tendency factor Jq and the oil film formation delay factor Ym to obtain the rotor assembly comprehensive score Rs, wherein the rotor assembly comprehensive score Rs is specifically obtained as follows: ; Where, and Jq represent the oil film formation delay factor and aggregation tendency factor, respectively. represents the sensitivity adjustment factor, It represents the maximum oil film formation delay factor allowed by the compressor, and e represents the base of the natural logarithm; A comprehensive scoring threshold YZ is preset, and the rotor assembly comprehensive score Rs and the comprehensive scoring threshold YZ are compared and analyzed to evaluate the compressor rotor assembly quality. The specific evaluation contents are as follows: If the rotor assembly comprehensive score Rs is less than or equal to the comprehensive score threshold YZ, it is determined that the oil film response is timely and the compressor rotor assembly quality is normal. At this time, the "assembly structure qualified" label is output, the aggregation tendency factor Jq and the oil film formation delay factor Ym are recorded, and an instruction is sent to control the normal operation of the compressor system; If the rotor assembly comprehensive score Rs is greater than the comprehensive score threshold YZ, it is determined that the oil film formation delay is abnormal, the frequency aggregation trend is obvious, the compressor has a hidden danger of inducing structural resonance, and the compressor rotor assembly quality is at risk. At this time, the "assembly structure abnormality risk" label is output, the equipment is suspended, and the quality inspector is notified to perform a load test. At the same time, recommended inspection items are generated, including "bearing installation eccentricity inspection", "initial rotor balance state inspection", "end face contact deviation inspection" and "lubrication chamber residual impurity inspection", and the characteristic data is recorded and stored as a "fault sample" in the data analysis library.

[0017] Preferably, a method for online detection of compressor rotor assembly quality comprises the following steps: Step 1: Collect the compressor microgravity operation data, perform microgravity operation simulation, and construct a disturbance sampling window, which is then sent to the micro-vibration monitoring and analysis module for data collection and analysis. Step 2: Real-time collection of the axial acceleration response signal of the compressor rotor within the disturbance sampling window , and perform feature extraction to obtain a feature data set; Step 3: Based on the characteristic data set, calculate and obtain the oil film formation delay factor Ym, evaluate the oil film delay risk level, and trigger the aggregation trend analysis module to perform frequency domain aggregation trend analysis; Step 4: Analyze the frequency domain aggregation trend and calculate the aggregation trend factor Jq; Step 5: Based on the oil film formation delay factor Ym and the aggregation tendency factor Jq, calculate and obtain the rotor assembly comprehensive score Rs, evaluate the compressor rotor assembly quality, and issue an assembly quality warning.

[0018] The present invention provides a method and system for online detection of compressor rotor assembly quality, which has the following beneficial effects: (1) By constructing a disturbance simulation loading module, a microgravity operation data simulation mechanism is introduced in the online assembly state of the compressor rotor, and combined with the disturbance start boundary control strategy, the dynamic reconstruction and equivalent excitation control of the compressor operating state in a special plateau or low-pressure environment are realized. This method can suppress the masking effect of the oil film inertia drive, significantly enhance the observability of the structural response characteristics caused by small assembly deviations, and provide a highly sensitive disturbance triggering basis for the online detection system, solving the current assembly state identification blind spot of "micro defects are difficult to expose and the problem of excessive excitation masking is serious", and has good field adaptability and signal controllability.

[0019] (2) By constructing the micro-vibration monitoring and analysis module and the oil film delay analysis module, feature extraction is performed to obtain NL and phase change factor Xw, and further obtain the oil film formation delay factor Ym. By combining the axial acceleration signal and phase fluctuation characteristics during the disturbance for calculation, the oil film delay behavior caused by pressure deviation, oil supply fluctuation or insufficient film thickness can be actively identified during the assembly stage. The system actively triggers the subsequent focus trend analysis process by determining the oil film status level, thereby realizing the structured evaluation and early warning response of "abnormal oil film response", solving the problems of delayed identification means lag, fuzzy status division, and no risk early warning in the existing solutions.

[0020] (3) The aggregation trend factor Jq is constructed by calculating the frequency difference vector and energy density to identify whether the multi-modal frequencies have abnormal convergence in the main frequency area. This method effectively captures the structural frequency fusion trend and solves the fuzzy perception problem of "resonance precursors" in existing methods. It can provide early warning of the risk of systemic structural instability caused by modal coupling during the rotor assembly stage. At the same time, the rotor assembly comprehensive score Rs obtained by the oil film delay factor Ym and the aggregation trend factor Jq is used to quantitatively evaluate the rotor assembly quality. It has the advantages of clear risk response, accurate abnormal correlation, and traceable evaluation results, which makes up for the evaluation gap of the lack of structure-lubrication collaborative risk indicators in traditional systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a block diagram of an online detection system for compressor rotor assembly quality according to the present invention; Figure 2 This is a flow chart of an online detection method for compressor rotor assembly quality according to the present invention; Figure 3 This is a block diagram of the disturbance simulation loading module of the present invention; Figure 4 Schematic diagram of the rotor assembly quality evaluation results of the present invention. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention. Example 1

[0023] See also Figure 1 , the present invention provides a compressor rotor assembly quality online detection system, including a disturbance simulation loading module, a micro-vibration monitoring and analysis module, an oil film delay analysis module, an aggregation trend analysis module and a rotor assembly stability early warning module; The disturbance simulation loading module is used to collect the microgravity operation data of the compressor, perform microgravity operation simulation, and construct a disturbance sampling window, which is then sent to the microvibration monitoring and analysis module for data collection and analysis. The micro-vibration monitoring and analysis module is used to collect the axial acceleration response signal of the compressor rotor within the disturbance sampling window in real time. , and perform feature extraction to obtain a feature data set; The oil film delay analysis module is used to calculate the oil film formation delay factor Ym based on the characteristic data set, evaluate the oil film delay risk level, and trigger the aggregation trend analysis module to perform frequency domain aggregation trend analysis; The aggregation trend analysis module is used to analyze the frequency domain aggregation trend and calculate the aggregation trend factor Jq; The rotor assembly stability warning module is used to calculate the rotor assembly comprehensive score Rs based on the oil film formation delay factor Ym and the aggregation tendency factor Jq, evaluate the compressor rotor assembly quality, and issue an assembly quality warning.

[0024] In the embodiment, by constructing a disturbance simulation loading module, a micro-vibration monitoring and analysis module, an oil film delay analysis module, an aggregation trend analysis module and a rotor assembly stability warning module, a full-link structural diagnosis of the compressor operation response under the rotor assembly state is realized. The system induces assembly micro-defect response signals through microgravity environment simulation, and combines real-time axial micro-vibration feature extraction with quantitative calculation of the oil film formation delay factor to accurately characterize the stability and hysteresis of the oil film response, and further identifies the modal frequency aggregation trend through the aggregation trend factor to capture the structural coupling precursor behavior in advance. Finally, the rotor assembly quality is intelligently evaluated and warned in a comprehensive scoring method based on multi-dimensional feature fusion, which has the beneficial effects of strong early diagnosis capability, high accuracy in identifying minor defects and clear analysis of structural responses. Example 2

[0025] Please refer to Figure 1 and Figure 3 Specifically: The disturbance simulation loading module is used to collect microgravity operation data at the location of the compressor, including the altitude of the compressor location. and atmospheric pressure , and calculate the impact of microgravity operation data on the working state of the compressor, and obtain the equivalent gravity correction coefficient , where the equivalent gravity correction coefficient is The specific way to obtain it is: ; Where, Indicates the altitude where the compressor is located. Indicates the standard reference altitude, Indicates the atmospheric pressure at the location of the compressor, represents standard atmospheric pressure, represents the power transformation factor; It represents the altitude ratio, which is the ratio of the sea wave height in the compressor operating environment to the standard altitude. Under normal circumstances, the higher the altitude, the lower the air density, the lower the oil film damping capacity will be, and the disturbance will be more easily amplified. It is used to measure the spatial characteristics of the impact of air rarefaction on the compressor control system. It indicates the ratio of atmospheric pressure. The lower the air pressure, the thinner the molecular density per unit volume, the lower the heat transfer and fluidity. The air pressure will affect the air load characteristics, oil film formation speed and disturbance attenuation ability of the compressor. , which can reflect the impact of air thinness on the compressor control system. When it decreases, it indicates the air coefficient, the system damping capacity decreases, and disturbances are more easily excited, thereby amplifying the phenomena of oil film delay, energy accumulation and structural resonance; Altitude ratio Ratio to atmospheric pressure Multiplying them together, a dimensionless environmental correction benchmark factor is constructed, which can reflect the comprehensive degree of deviation of the compressor operating environment from the standard operating state. When both deviate significantly from the standard value at the same time, the product will significantly decrease or increase, reflecting the nonlinear attenuation effect. When the compressor is operated in a plateau, high altitude or low pressure environment, the system's dynamic characteristics and the lubricating oil film formation process will be affected by the coupling of two factors, including "reduced atmospheric pressure → reduced air density → faster evaporation of lubricating oil, making it more difficult to form an oil film" and "increased altitude → increased gravitational potential energy and further decreased air density". Therefore, the disturbance loading effect and assembly response quality have significant differences in different environments, requiring a comprehensive correction factor, namely the equivalent gravity correction factor. To make adjustments; Among them, the altitude of the compressor operating environment Obtained through electronic map database, standard reference altitude Obtained from the compressor system equipment table, the atmospheric pressure of the compressor operating environment Obtained through pressure sensor; Power conversion factor A nonlinear control factor used to adjust the degree of equivalent gravity influence, reflecting the nonlinear attenuation trend of the compressor's disturbance response under non-standard operating conditions, such as plateau conditions, and is set by experts based on actual conditions; In plateau or abnormal air pressure environment, the control characteristics of the compressor system, such as rotational inertia, oil film establishment speed and torque output curve, will be different from those in the standard environment. These changes will interfere with the accuracy of the compressor's disturbance loading behavior. Therefore, the equivalent gravity correction coefficient is used. , which can realize the dynamic correction of the physical boundaries of the compressor control parameters; In a specific example, suppose that compressors of the same model work under different environments. Under standard conditions, the compressor control system starts according to the standard starting curve. The rotor oil film is established synergistically between the change in rotor inertia and the system response. The oil film delay window is reasonable, which can expose response characteristics such as micro-vibration and phase anomaly caused by assembly eccentricity, end face pressure deviation, etc., which is convenient for identifying assembly defects. However, in plateau environments, due to the lower atmospheric pressure and thin density, the system damping is reduced and the rotor inertia is enhanced, which makes the oil film respond earlier and the interference loading window shortened, causing the oil film to form prematurely, suppressing structural micro-vibration, and making the detection system judge that the rotor is "assembled normally", but in fact there are "unexposed potential defects".

[0026] Based on the equivalent gravity correction factor , calculate and obtain the compressor control parameter set, wherein the compressor control parameter set includes the initial maximum acceleration value , load limit torque upper limit and ramp-up time constant τ; Initial maximum acceleration value The specific acquisition method is as follows: ; Where, represents the standard acceleration due to gravity; Initial maximum acceleration value It is the maximum acceleration of the motor speed increase when the compressor control system starts. The rotor oil film will be quickly established under the dominance of inertia, covering up the rotor assembly deviation. This is a phenomenon in which strong inertia drives the system weakness, which will make it difficult to identify the weak vibration characteristics caused by assembly eccentricity, unequal bearing clearance, end face pressure deviation, etc. The monitoring system will find it difficult to observe the micro-vibration characteristic signals caused by assembly defects, and thus judge the compressor rotor assembly as qualified, and introduce the initial maximum acceleration value. , it can dynamically adjust the maximum acceleration boundary at the initial stage of disturbance loading, thereby limiting the excessive increase in motor speed and ensuring that the disturbance loading process remains in a state of moderate excitation and observable response, which is conducive to the true exposure of structural micro-vibration characteristics and the accurate identification of assembly quality.

[0027] Load limit torque upper limit The specific acquisition method is as follows: ; Where, Indicates the minimum torque output ratio of the compressor, represents the environmental disturbance sensitivity amplification factor, Indicates the maximum output torque value of the compressor under rated voltage, rated frequency, stable load, standard ambient temperature and normal cooling conditions; Among them, the upper limit of load torque Refers to the maximum output torque allowed by the compressor under the current environment, which exceeds the upper limit of the load torque The load limit and alarm protection will be triggered. Indicates the basic driving capacity that the compressor must ensure under extremely low output conditions. Indicates the maximum continuous output capacity of the compressor under rated voltage, rated frequency, stable load, standard ambient temperature and normal cooling conditions. It represents the disturbance-environment joint correction factor, that is, the response adjustment amplitude of the load limit control capability to external disturbances under different operating environments. When the external environment deteriorates, such as increasing altitude, decreasing air pressure and thinning air, the equivalent gravity correction coefficient will increase, making the compressor system more difficult to operate, where "1" is the baseline, " is an adjustable item; Load limit torque upper limit The calculation formula adopts a multiplication structure design to achieve a dynamic coupling relationship between the disturbance correction factor and the system rated capacity on the basis of ensuring the consistency of the physical quantity dimensions. Provides a theoretical upper bound, It reflects the amplifying and regulating effect of environmental changes on load limits, and It serves as the minimum protection coefficient to ensure that the system is not completely unloaded under any circumstances, thereby building a composite load limit control mechanism with nonlinear adjustment capabilities, disturbance sensitivity mapping capabilities and safety bottom line protection capabilities. Among them, unloading refers to the behavior of manually or automatically reducing the output load, torque or power of the drive system during the operation of the compressor, which is usually used to protect equipment, slow down the spread of faults or terminate operation. Incomplete unloading means that when a disturbance risk or abnormal operation is detected, the output capacity is not reduced to zero, but is controlled to operate within a certain minimum output range; Among them, the minimum torque output ratio of the compressor drive motor is It is used to prevent the disturbance amplitude from being too small to stimulate a response, and is obtained by performing a rated power matching test during the compressor control manufacturing stage; Environmental disturbance sensitivity amplification factor It is used to measure the "sensitivity amplification" effect of the compressor system structure to the disturbance torque in a microgravity environment, that is, the amplification factor of the disturbance force required for the micro-response of the structure under the perturbation state. It is obtained by fitting the result response model. The specific acquisition method is: ; Where, Indicates the minimum torque required for the disturbance loading to reach the "main frequency response steady state" at the current altitude. It indicates the minimum torque required for disturbance loading to reach the "main frequency response steady state" at standard altitude. The main frequency response steady state refers to the state in which the main frequency response component of the compressor rotor tends to be relatively stable and has no significant fluctuations within a certain period of time during the disturbance loading process.

[0028] Load limit torque upper limit It indicates the maximum allowable torque boundary of the pure output of the compressor drive system during the disturbance simulation loading period. It is used to control the "mechanical amplitude" of the disturbance imposed by the system on the rotor. In the rotor assembly state, if the disturbance torque is too strong, it will forcibly suppress the oil film or mislead the structural response, covering up the assembly defect characteristics, and introduce the load-limiting torque upper limit value. , which can limit the maximum torque output of the compressor during the disturbance simulation loading process, prevent the premature compaction of the oil film and the abnormal nonlinear enhancement of the structural response due to excessive excitation. In order to adapt to the changes in the mechanical response ability of the compressor system in plateau or low-pressure environments, the upper limit of the load torque is set. Combined with the equivalent gravity correction factor Dynamic calculations are performed to ensure that the disturbance loading process has stable, weak, and comparable control behavior, thereby ensuring that the structural micro-response characteristics can be accurately stimulated and identified, providing a reliable data basis for subsequent assembly deviation analysis.

[0029] The specific method of obtaining the ramp-up time constant τ is as follows: ; Where, Indicates the standard ramp-up time. represents the disturbance time sensitivity amplification factor; Among them, the standard ramp-up time Refers to the time constant required for the system to slowly rise from the zero disturbance state to the set disturbance amplitude when the compressor is subjected to disturbance loading under ideal working conditions with standard working environment, no external interference, stable oil film structure and balanced load. The specific values ​​are obtained from the compressor standard operating parameter table; Disturbance time sensitivity amplification factor It refers to the sensitivity amplification factor of the compressor's structure or lubrication response to the "disturbance loading time rhythm" when the compressor is subjected to disturbance loading. It is a parameter used to characterize the response sensitivity of the compressor structure system to the change of the disturbance loading time rhythm. Its specific value is set by experts according to actual conditions. The disturbance time sensitivity amplification factor is used. and equivalent gravity correction factor Combined, it is used to determine whether the compressor needs to slow down the disturbance loading speed to prevent the system from causing structural resonance, lubrication instability or abnormal oil film adhesion due to excessive loading speed; Indicates the disturbance rhythm adjustment factor, which is used to determine whether the disturbance loading speed needs to be slowed down to protect the structural safety and the stability of the lubrication system; The ramp-up time constant τ refers to the duration of the disturbance loading process, which is based on the standard ramp-up time As a benchmark, combined with the sensitivity of the current equipment structure to the disturbance loading rhythm Risk intensity of deviations from standard operating conditions Dynamic calculation acquisition; The ramp-up time constant τ is used to realize the dynamic medium condition of the compressor disturbance loading process, avoiding the problem of premature oil film coverage and masking of abnormal structural response due to too fast starting speed. By introducing the ramp-up time constant τ, the compressor system can maintain an equivalent disturbance rhythm under different environmental conditions, thereby enhancing the stability of rotor assembly deviation identification.

[0030] Based on the generated compressor control parameter set, the compressor control parameter set is transmitted to the PLC industrial control system through the standard industrial communication protocol, and the PLC industrial control system is used to map the compressor control parameters into an internal execution control curve function, and automatically output the compressor dynamic control instructions. According to the compressor dynamic control instructions, the compressor is driven to perform disturbance simulation, and the slow rise time constant τ is used as the upper limit to set the disturbance sampling window [ τ].

[0031] The PLC industrial control system is an industrial automation control system with a programmable logic controller as its core. It has functions such as parameter reception, curve mapping, control logic execution and command output. It can communicate with the host system through the industrial bus interface and map the disturbance control parameters into execution control curves to achieve dynamic loading disturbance control of the compressor, ensuring that the disturbance process is controllable and the sampling window is clearly set. It is the core execution module for realizing full closed-loop disturbance control simulation.

[0032] In the embodiment, by introducing the disturbance simulation loading module, the equivalent gravity correction coefficient is introduced to solve the problem that the disturbance loading process is uncontrollable due to the change of rotational inertia and abnormal oil film formation rate of the compressor system, and the micro-vibration characteristics of assembly defects are difficult to excite. , dynamically adjust the initial maximum acceleration , load limit torque upper limit and key boundary control parameters such as the ramp-up time constant τ, to achieve fine-grained control of the disturbance loading intensity and rhythm. By limiting the excitation speed and mechanical amplitude of the motor during the startup phase, it effectively avoids premature stabilization of the oil film or nonlinear distortion of the response signal, allowing the compressor system to truly expose the weak vibration characteristics caused by rotor assembly deviation. This module provides a disturbance excitation mechanism that can be adapted to multiple environmental conditions, improving the sensitivity, stability and environmental adaptability of the compressor rotor assembly online detection system in early defect identification, and providing an accurate and reliable response data basis for subsequent diagnostic analysis. Example 3

[0033] Please refer to Figure 1 ,Specifically: the micro-vibration monitoring and analysis module includes a spectrum ,analysis unit and a feature extraction unit; The spectrum analysis unit is used to deploy the axial acceleration sensor on the compressor and detect the axial acceleration in the disturbance sampling window [ τ], the axial acceleration sensor is used to collect the axial acceleration response signal of the compressor rotor in real time , and respond to the collected axial acceleration signal Perform preprocessing, including normalization and wavelet denoising; According to the pre-processed axial acceleration response signal , perform fast Fourier transform FFT to obtain the frequency spectrum of the compressor rotor, wherein the specific process of fast Fourier transform FFT is as follows: ; In the formula, A( ) indicates the compressor rotor at the frequency point The frequency domain energy density at represents the fast Fourier transform, represents the axial acceleration response signal of the compressor rotor at time point t, t∈[ τ], represents the i-th discrete frequency point; And based on the frequency spectrum of the compressor rotor, a frequency data set S is constructed, wherein the frequency data set S includes the frequency domain energy density of each frequency point.

[0034] The feature extraction unit is used to extract the axial acceleration response signal , perform integral summation to obtain the disturbance sampling window [ The disturbance energy response factor NL of the energy intensity of the compressor rotor structure response in τ] is obtained as follows: ; Where, represents the axial acceleration response signal of the compressor rotor at time point t, represents the start time of disturbance loading, represents the ramp-up time constant, i.e. the duration of the disturbance loading; The disturbance energy response factor NL is used to represent the axial structural energy coupling intensity of the compressor rotor during the disturbance loading process, and can be used as a basis for dynamic evaluation of oil film stability and rotor assembly status. It can quantify the axial vibration energy intensity of the compressor rotor during the simulated disturbance loading process. It calculates the axial acceleration response signal during the disturbance loading period in the disturbance sampling window [ In the case of defects such as rotor eccentricity, improper bearing installation, and insufficient oil film coverage, the rotor structure's ability to absorb external disturbances will be abnormally low. At this time, the disturbance energy response factor NL will deviate from the normal range. It can be combined with the phase change factor Xw to obtain the oil film formation delay factor Ym, and perform a preliminary assessment of the rotor assembly quality.

[0035] Based on the frequency data set S, identify the main frequency points in the frequency data set S , and the main frequency point Axial acceleration response signal Use bandpass filtering to obtain the main frequency response component , and the main frequency response component Perform Hilbert transform to obtain the main frequency response component The instantaneous phase φ , and calculate the main frequency response components The instantaneous phase change of is obtained to obtain the phase change factor Xw, wherein the phase change factor Xw is obtained in the following manner: ; Among them, bandpass filtering refers to extracting the main frequency point The main frequency response component , for the main frequency point , only keep the main frequency point Nearby response signals are filtered to remove frequency noise and non-target components; Main frequency response components Refers to the compressor axial acceleration response signal In the main frequency point The corresponding frequency response component represents the compressor axial acceleration response signal , the actual response behavior in the direction of the main frequency excitation; Instantaneous phase refers to the phase rotation angle of the disturbance signal and is often used to identify problems such as "assembly instability", "relative delay of oil film" and "disturbance signal offset". The Hilbert transform is a mathematical tool commonly used in signal analysis that can construct a complex envelope signal from a real signal and then obtain the instantaneous phase of the signal. And based on the obtained disturbance energy response factor NL and phase change factor Xw, a feature data set is constructed.

[0036] The phase change factor Xw is used to represent the main frequency response component of the compressor system The phase fluctuation rate is an indicator to measure whether the rotor structure stiffness and oil film stability fluctuate under the perturbation state, and can identify "local loose parts and oil film delay problems."

[0037] In the embodiment, by integrating spectrum analysis and feature extraction mechanisms, the axial acceleration response signal of the compressor rotor is collected and processed in real time within the disturbance sampling window, and signal preprocessing, spectrum analysis and energy feature extraction are completed successively. By constructing a frequency data set, the system can identify the main frequency point and obtain its main frequency response component, and further use bandpass filtering and Hilbert transform to extract the instantaneous phase trajectory, calculate the phase change factor Xw, and combine the disturbance energy response factor NL and the phase change factor Xw to form a feature data set, thereby realizing accurate identification of compressor structural rigidity fluctuations, oil film establishment delays and local loosening anomalies. This method effectively breaks through the problem of difficulty in identifying weak vibration features in traditional assembly inspection, and improves the system's perception sensitivity and risk warning capabilities for slight structural anomalies and unstable oil film behavior. Example 4

[0038] Please refer to Figure 1 ,Specifically: the oil film delay analysis module includes an oil film formation delay factor ,calculation unit and an early warning unit; The oil film formation delay factor calculation unit is used to calculate the oil film formation delay factor in the disturbance sampling window [ τ], from the starting interference time point At the beginning, take a length of Time sliding window to obtain several lengths of Time sliding window; For each time sliding window, the main frequency point is identified, and the main frequency response component and instantaneous phase of each time sliding window are obtained to obtain the phase change factor Xw of each time sliding window, and the phase change factor Xw of all time sliding windows is integrated to obtain the standard deviation of the phase change factor ; Set the phase change stability threshold Xwyz. For each time sliding window, if the phase change factor standard deviation is less than the phase change stability threshold Xwyz, then the marking time sliding window is a stable window. If the phase change factor standard deviation If it is greater than or equal to the phase change stability threshold Xwyz, no processing is required; Set the continuous stability window threshold , for each time sliding window obtained, if the continuous If all the time sliding windows are stable windows, it is determined that the compressor has entered the main frequency response stable state, and the continuous The starting time point of the initial time sliding window in the time sliding window is the time point when the main frequency response is stable ; Set the continuous stability window threshold The purpose is to enhance the anti-interference ability of the stable identification of disturbance response, prevent the occurrence of misjudgment that the rotor oil film has been stably formed due to short-term fluctuations or transient errors, and thus improve the credibility and stability of the oil film delay factor calculation; The time sliding window mechanism can effectively cope with the non-stationary characteristics of the disturbance signal. By extracting the main frequency response and instantaneous phase in segments, it captures the dynamic evolution characteristics of the assembly process. Further, the standard deviation of the phase change factors of adjacent windows is used for fluctuation analysis to identify continuous stable windows. This mechanism can not only accurately define the steady-state range of the oil film formation process, but also serve as an important benchmark segment for subsequent fault warning, structural adhesion stability assessment, and risk model building, significantly improving the system's recognition sensitivity and assessment accuracy for assembly anomalies.

[0039] Based on the characteristic data set and the main frequency response stable time point , quantify the lag degree of oil film establishment of the compressor rotor after the disturbance loading process to obtain the oil film formation delay factor Ym, where the oil film formation delay factor Ym is specifically obtained as follows: ; Where, Indicates the time point when the main frequency response stabilizes, Indicates the standard response time point, represents the disturbance energy response factor, Indicates the time point when the main frequency response stabilizes The phase change factor of the compressor is represents the adjustment factor, ln represents the natural logarithm function, represents the coupled inertial damping constant, where .

[0040] in, It represents the time delay normalization term, which is used to indicate the relative delay degree of the rotor oil film response, and indicates the time extension from startup to main frequency stabilization state after disturbance loading. The larger the value, the more delayed the oil film response; Represents the disturbance intensity adjustment factor, where the adjustment factor The results were obtained by regression fitting experiments using the least squares method; It represents the oil film structure stability correction term, which is used to measure whether the main frequency phase is still "stable" or "violently fluctuating" after the disturbance. Among them, the coupled inertial damping constant It is used to reflect the coupling resistance between the compressor rotor stiffness structure and the oil film. The specific value is obtained from the compressor database. Standard response time points The coupled inertial damping constant is obtained from the compressor rotor structure database. The inertia of compressor rotor structures is obtained from the compressor rotor structure database and the rotor structure and lubrication coupling modeling simulation. The inertia of compressor structures of different models varies greatly. The “disturbance bearing capacity” of the rotor structure can then be normalized to make the algorithm universal.

[0041] The early warning unit is used to preset the first delay risk threshold and the second delay risk threshold and the oil film formation delay factor Ym is combined with the first delay risk threshold and the second delay risk threshold Conduct comparative analysis and assess the oil film delay risk level. The specific assessment contents are as follows: If the oil film formation delay factor Ym is less than or equal to the first delay risk threshold , then the oil film is judged to be in a normal formation state. At this time, the oil film delay risk level is judged to be the first risk level and no treatment is required; If the oil film formation delay factor Ym is greater than the first delay risk threshold , and is less than the second delay risk threshold , then the oil film is judged to be in a state where delayed formation is allowed. At this time, the oil film delay risk level is judged to be the second risk level, and the rotor operation status needs to be continuously monitored; If the oil film formation delay factor Ym is greater than or equal to the second delay risk threshold , the oil film is judged to be in an abnormally delayed state. At this time, the oil film delay risk level is judged to be the third risk level. The compressor rotor assembly torque is abnormal, there is a stress concentration point, and the lubrication system is partially blocked and the oil supply pressure is insufficient, which automatically triggers the aggregation trend analysis module.

[0042] In the embodiment, by introducing the disturbance response sliding window analysis mechanism and combining the main frequency response phase fluctuation characteristics, the oil film formation delay factor Ym is constructed, which realizes the quantitative expression of the hysteresis of the initial film formation process of the compressor oil film. Compared with the existing technology that is difficult to clearly judge the stabilization time point of the main frequency response, the body response is fuzzy, and it is difficult to quantify the degree of film formation lag, the present invention can not only automatically identify the oil film establishment delay time, but also realize the graded judgment and early warning triggering of the delay state based on the preset multi-level risk threshold, solve the technical shortcomings that problems such as abnormal assembly stress and lubrication resistance are difficult to manifest in the early stage, and significantly enhance the system's recognition ability and response foresight to problems caused by poor lubrication and structural deviation. Example 5

[0043] Please refer to Figure 1 ,Specifically: the aggregation trend analysis module includes a frequency domain energy ,analysis unit and an aggregation trend index calculation unit; The frequency domain energy analysis unit is used to calculate the energy of each frequency point based on the frequency data set S. and main frequency point The absolute distance between , and form a frequency difference vector; And based on the frequency difference vector, perform global statistical analysis and calculate the frequency mean and frequency fluctuation variance .

[0044] The calculation unit of the cluster trend indicator is used to calculate the frequency difference vector and the frequency fluctuation variance , perform summary calculation to obtain the aggregation trend factor Jq, where the specific method of obtaining the aggregation trend factor Jq is: ; Where, Indicates the main frequency point The frequency domain energy density, Indicates frequency point and main frequency point The absolute distance between , represents the frequency fluctuation variance, represents the i-th frequency point, Indicates frequency point The frequency domain energy density, n represents the total number of frequency points, i represents the frequency point index, i={1, 2, 3, ..., n}, represents the correction constant; Refers to the square value of the spectrum energy density at the main frequency point, Represents the sum of the weighted spectral energy density of all frequency points, Represents a penalty factor used to suppress the interference of frequency components that deviate from the main frequency point on the frequency aggregation score. Multiplying by the spectrum energy density can reduce the sum of the weighted spectrum energy density of each frequency point with "high spectrum energy density but far from the main frequency point" The farther the frequency point is from the main frequency point, the heavier the penalty and the larger the weighted value. It indicates the degree of deviation of the frequency point in the entire spectrum distribution; The essence of the aggregation trend factor Jq is the ratio of the "spectral energy density of the main frequency point" to the "sum of the weighted spectral energy density of all frequency points". Other frequency points except the main frequency point will be penalized for deviating from the main frequency. The aggregation trend factor Jq is used to identify whether there is an abnormal aggregation trend in the rotor structure frequency response during the compressor disturbance loading process. During the rotor startup phase, the lubricating oil film will experience critical thinning under different operating conditions, including low load and frequent start-stop conditions. The compressor control system will experience micro-vibration and the bearing support force will fluctuate, which will trigger axial resonance coupling and ultimately destroy the steady-state startup conditions.

[0045] In an embodiment, an aggregation trend analysis module is set up, and a frequency domain energy analysis unit is used to extract the absolute offset between each frequency point in the frequency data set and the main frequency band, to construct a frequency difference vector, and to calculate the frequency fluctuation variance by a statistical method. Further, an aggregation trend index calculation unit obtains an aggregation trend factor Jq based on the relationship between the frequency difference and the energy distribution, quantifies the modal focusing degree of the frequency response structure, and accurately identifies whether the multi-modal frequency tends to abnormally aggregate toward the main frequency after disturbance loading, thereby capturing the precursor characteristics of axial resonance. Compared with the traditional frequency domain peak recognition method, this indicator can reveal in advance the frequency fusion trend caused by oil film instability, uneven bearing clearance, etc., and has the advantages of sensitive structural risk response and accurate focusing behavior judgment, thereby improving the recognition ability and systematic warning level of potential resonance hazards of the rotor assembly structure. Example 6

[0046] Please refer to Figure 1 and Figure 4 Specifically: The rotor assembly stability warning module is used to perform summary calculations based on the aggregation tendency factor Jq and the oil film formation delay factor Ym to obtain the rotor assembly comprehensive score Rs. The rotor assembly comprehensive score Rs is specifically obtained as follows: ; Where, and Jq represent the oil film formation delay factor and aggregation tendency factor, respectively. represents the sensitivity adjustment factor, Indicates the maximum oil film formation delay factor allowed by the compressor, e represents the base of the natural logarithm, where the sensitivity adjustment factor By setting multiple experimental samples and performing data regression fitting, the maximum oil film formation delay factor allowed by the compressor is obtained. The oil film formation delay factors of multiple "known qualified assembly" compressors are recorded under standard assembly conditions, including various rotor bearing fit accuracy and lubrication system loads, and statistical distribution analysis is performed to obtain the results. It represents the oil film delay degree × frequency aggregation degree, reflecting the fault intensity of the rotor assembly. Is an exponential nonlinear adjustment coefficient, which is a sigmoid function The deformation structure of the oil film delay factor Ym is constructed in an exponential form to dynamically adjust the rotor assembly comprehensive score Rs. The maximum oil film formation delay factor allowed by the compressor The greater the difference, the smaller the delay of the oil film. ≪ , the larger the denominator, the lower the rotor assembly comprehensive score Rs, and the better the rotor assembly quality. ≫ , the closer the denominator is to 1, the higher the rotor assembly comprehensive score Rs is, and the worse the compressor rotor assembly quality is; The rotor assembly comprehensive score Rs is constructed by coupling the oil film formation delay factor Ym with the aggregation tendency factor Jq to construct the assembly score item, and the maximum oil film formation delay factor allowed by the compressor is introduced. The exponential nonlinear adjustment coefficient centered on is used to achieve interval response control of the score; A comprehensive scoring threshold YZ is preset, and the rotor assembly comprehensive score Rs and the comprehensive scoring threshold YZ are compared and analyzed to evaluate the compressor rotor assembly quality. The specific evaluation contents are as follows: If the rotor assembly comprehensive score Rs is less than or equal to the comprehensive score threshold YZ, it is determined that the oil film response is timely and the compressor rotor assembly quality is normal. At this time, the "assembly structure qualified" label is output, the aggregation tendency factor Jq and the oil film formation delay factor Ym are recorded, and an instruction is sent to control the normal operation of the compressor system; If the rotor assembly comprehensive score Rs is greater than the comprehensive score threshold YZ, it is determined that the oil film formation delay is abnormal, the frequency aggregation trend is obvious, the compressor has a hidden danger of inducing structural resonance, and the compressor rotor assembly quality is at risk. At this time, the "assembly structure abnormality risk" label is output, the equipment is suspended, and the quality inspector is notified to perform a load test. At the same time, recommended inspection items are generated, including "bearing installation eccentricity inspection", "initial rotor balance state inspection", "end face contact deviation inspection" and "lubrication chamber residual impurity inspection", and the characteristic data is recorded and stored as a "fault sample" in the data analysis library.

[0047] In the embodiment, a fusion judgment mechanism is constructed through the rotor assembly stability early warning module, and a nonlinear comprehensive score is performed based on the oil film formation delay factor Ym and the frequency aggregation trend factor Jq. The assembly comprehensive score Rs is calculated and compared with the preset threshold value YZ, thereby achieving accurate judgment and classification response to the rotor assembly status. It can effectively identify potential resonance precursors caused by the hysteresis of the compressor's lubrication system and abnormal aggregation of structural modes, and improve the system's sensitivity and response speed to assembly defects. When Rs is abnormal, the system can automatically suspend the equipment operation and generate diagnostic suggestions, effectively ensuring the safety of equipment operation and supporting the construction of an offline fault sample library, with significant engineering practical value and field deployability. Example 7

[0048] Please refer to Figure 2 Specifically: A compressor rotor assembly quality online detection method, including: Step 1: Collect the compressor microgravity operation data, perform microgravity operation simulation, and construct a disturbance sampling window, which is then sent to the micro-vibration monitoring and analysis module for data collection and analysis. Step 2: Real-time collection of the axial acceleration response signal of the compressor rotor within the disturbance sampling window , and perform feature extraction to obtain a feature data set; Step 3: Based on the characteristic data set, calculate and obtain the oil film formation delay factor Ym, evaluate the oil film delay risk level, and trigger the aggregation trend analysis module to perform frequency domain aggregation trend analysis; Step 4: Analyze the frequency domain aggregation trend and calculate the aggregation trend factor Jq; Step 5: Based on the oil film formation delay factor Ym and the aggregation tendency factor Jq, calculate and obtain the rotor assembly comprehensive score Rs, evaluate the compressor rotor assembly quality, and issue an assembly quality warning.

[0049] In the embodiment, by constructing a microgravity disturbance simulation environment, the problem of insufficient accuracy of compressor rotor assembly quality detection under special working conditions such as plateaus is effectively solved. The whole process extracts axial acceleration characteristics from the disturbance response signal, quantifies the delay degree of the oil film establishment process, obtains the oil film formation delay factor Ym, and extracts the aggregation trend factor Jq in combination with the frequency domain energy analysis results. Finally, the rotor assembly comprehensive score Rs is used to achieve intelligent early warning of the compressor rotor assembly quality. This method has the ability to adapt to dynamic environments, can timely identify structural abnormalities caused by assembly eccentricity, poor lubrication, etc., realize early quality defect identification, and improve rotor assembly reliability and operation safety.

[0050] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A compressor rotor assembly quality online detection system, characterized by: It includes disturbance simulation loading module, micro-vibration monitoring and analysis module, oil film delay analysis module, aggregation trend analysis module and rotor assembly stability early warning module; The disturbance simulation loading module is used to collect the microgravity operation data of the compressor, perform microgravity operation simulation, and construct a disturbance sampling window, which is then sent to the microvibration monitoring and analysis module for data collection and analysis. The micro-vibration monitoring and analysis module is used to collect the axial acceleration response signal of the compressor rotor within the disturbance sampling window in real time. , and perform feature extraction to obtain a feature data set; The oil film delay analysis module is used to calculate the oil film formation delay factor Ym based on the characteristic data set, evaluate the oil film delay risk level, and trigger the aggregation trend analysis module to perform frequency domain aggregation trend analysis; The aggregation trend analysis module is used to analyze the frequency domain aggregation trend and calculate the aggregation trend factor Jq; The rotor assembly stability warning module is used to calculate the rotor assembly comprehensive score Rs based on the oil film formation delay factor Ym and the aggregation tendency factor Jq, evaluate the compressor rotor assembly quality, and issue an assembly quality warning.

2. The compressor rotor assembly quality online detection system according to claim 1, characterized in that: The disturbance simulation loading module is used to collect microgravity operation data at the compressor location, including the altitude of the compressor location. and atmospheric pressure , and calculate the impact of microgravity operation data on the working state of the compressor, and obtain the equivalent gravity correction coefficient , where the equivalent gravity correction coefficient is The specific way to obtain it is: ； Where, Indicates the altitude where the compressor is located. Indicates the standard reference altitude, Indicates the atmospheric pressure at the location of the compressor, represents standard atmospheric pressure, represents the power transformation factor; Based on the equivalent gravity correction factor , calculate and obtain the compressor control parameter set, wherein the compressor control parameter set includes the initial maximum acceleration value , load limit torque upper limit and ramp-up time constant τ; Initial maximum acceleration value The specific acquisition method is as follows: ； Where, represents the standard acceleration due to gravity; Load limit torque upper limit The specific acquisition method is as follows: ； Where, Indicates the minimum torque output ratio of the compressor, represents the environmental disturbance sensitivity amplification factor, Indicates the maximum output torque value of the compressor under rated voltage, rated frequency, stable load, standard ambient temperature and normal cooling conditions; The specific method of obtaining the ramp-up time constant τ is as follows: ； Where, Indicates the standard ramp-up time. represents the disturbance time sensitivity amplification factor; Based on the generated compressor control parameter set, the compressor control parameter set is transmitted to the PLC industrial control system through the standard industrial communication protocol, and the PLC industrial control system is used to map the compressor control parameters into an internal execution control curve function, and automatically output the compressor dynamic control instructions. According to the compressor dynamic control instructions, the compressor is driven to perform disturbance simulation, and the slow rise time constant τ is used as the upper limit to set the disturbance sampling window [ τ].

3. The compressor rotor assembly quality online detection system according to claim 2, characterized in that: The micro-vibration monitoring and analysis module includes a spectrum analysis unit and a feature extraction unit; The spectrum analysis unit is used to deploy the axial acceleration sensor on the compressor and detect the axial acceleration in the disturbance sampling window [ τ], the axial acceleration sensor is used to collect the axial acceleration response signal of the compressor rotor in real time , and respond to the collected axial acceleration signal Perform preprocessing, including normalization and wavelet denoising; According to the pre-processed axial acceleration response signal , perform fast Fourier transform FFT to obtain the frequency spectrum of the compressor rotor, wherein the specific process of fast Fourier transform FFT is as follows: ； In the formula, A( ) indicates the compressor rotor at the frequency point The frequency domain energy density at represents the fast Fourier transform, represents the axial acceleration response signal of the compressor rotor at time point t, t∈[ τ], represents the i-th discrete frequency point; And based on the frequency spectrum of the compressor rotor, a frequency data set S is constructed, wherein the frequency data set S includes the frequency domain energy density of each frequency point.

4. The compressor rotor assembly quality online detection system according to claim 3, characterized in that: The feature extraction unit is used to extract the axial acceleration response signal , perform integral summation to obtain the disturbance sampling window [ The disturbance energy response factor NL of the energy intensity of the compressor rotor structure response in τ] is obtained as follows: ； Where, represents the axial acceleration response signal of the compressor rotor at time point t, represents the start time of disturbance loading, Indicates the ramp-up time constant; Based on the frequency data set S, identify the main frequency points in the frequency data set S , and the main frequency point Axial acceleration response signal Use bandpass filtering to obtain the main frequency response component , and the main frequency response component Perform Hilbert transform to obtain the main frequency response component The instantaneous phase φ , and calculate the main frequency response components The instantaneous phase change of is obtained to obtain the phase change factor Xw, wherein the phase change factor Xw is obtained in the following manner: ； And based on the obtained disturbance energy response factor NL and phase change factor Xw, a feature data set is constructed.

5. The compressor rotor assembly quality online detection system according to claim 4, characterized in that: The oil film delay analysis module includes an oil film formation delay factor calculation unit and an early warning unit; The oil film formation delay factor calculation unit is used to calculate the oil film formation delay factor in the disturbance sampling window [ τ], from the starting interference time point At the beginning, take a length of Time sliding window to obtain several lengths of Time sliding window; For each time sliding window, the main frequency point is identified, and the main frequency response component and instantaneous phase of each time sliding window are obtained to obtain the phase change factor Xw of each time sliding window, and the phase change factor Xw of all time sliding windows is integrated to obtain the standard deviation of the phase change factor ; Set the phase change stability threshold Xwyz. For each time sliding window, if the phase change factor standard deviation is less than the phase change stability threshold Xwyz, then the marking time sliding window is a stable window. If the phase change factor standard deviation If it is greater than or equal to the phase change stability threshold Xwyz, no processing is required; Set the continuous stability window threshold , for each time sliding window obtained, if the continuous If all the time sliding windows are stable windows, it is determined that the compressor has entered the main frequency response stable state, and the continuous The starting time point of the initial time sliding window in the time sliding window is the time point when the main frequency response is stable ; Based on the characteristic data set and the main frequency response stable time point , quantify the lag degree of oil film establishment of the compressor rotor after the disturbance loading process to obtain the oil film formation delay factor Ym, where the oil film formation delay factor Ym is specifically obtained as follows: ； Where, Indicates the time point when the main frequency response stabilizes, Indicates the standard response time point, represents the disturbance energy response factor, Indicates the time point when the main frequency response stabilizes The phase change factor of the compressor is represents the adjustment factor, ln represents the natural logarithm function, represents the coupled inertial damping constant.

6. The compressor rotor assembly quality online detection system according to claim 5, characterized in that: The early warning unit is used to preset the first delay risk threshold and the second delay risk threshold and the oil film formation delay factor Ym is combined with the first delay risk threshold and the second delay risk threshold Conduct comparative analysis and assess the oil film delay risk level. The specific assessment contents are as follows: If the oil film formation delay factor Ym is less than or equal to the first delay risk threshold , then the oil film is judged to be in a normal formation state. At this time, the oil film delay risk level is judged to be the first risk level and no treatment is required; If the oil film formation delay factor Ym is greater than the first delay risk threshold , and is less than the second delay risk threshold , then the oil film is judged to be in a state where delayed formation is allowed. At this time, the oil film delay risk level is judged to be the second risk level, and the rotor operation status needs to be continuously monitored; If the oil film formation delay factor Ym is greater than or equal to the second delay risk threshold , the oil film is judged to be in an abnormally delayed state. At this time, the oil film delay risk level is judged to be the third risk level. The compressor rotor assembly torque is abnormal, there is a stress concentration point, and the lubrication system is partially blocked and the oil supply pressure is insufficient, which automatically triggers the aggregation trend analysis module.

7. The compressor rotor assembly quality online detection system according to claim 6, characterized in that: The aggregation trend analysis module includes a frequency domain energy analysis unit and an aggregation trend index calculation unit; The frequency domain energy analysis unit is used to calculate the energy of each frequency point based on the frequency data set S. and main frequency point The absolute distance between , and form a frequency difference vector; And based on the frequency difference vector, perform global statistical analysis and calculate the frequency mean and frequency fluctuation variance .

8. The compressor rotor assembly quality online detection system according to claim 7, characterized in that: The calculation unit of the cluster trend indicator is used to calculate the frequency difference vector and the frequency fluctuation variance , perform summary calculation to obtain the aggregation trend factor Jq, where the specific method of obtaining the aggregation trend factor Jq is: ； Where, Indicates the main frequency point The frequency domain energy density, Indicates frequency point and main frequency point The absolute distance between , represents the frequency fluctuation variance, represents the i-th frequency point, Indicates frequency point The frequency domain energy density, n represents the total number of frequency points, i represents the frequency point index, i={1, 2, 3, ..., n}, Represents the correction constant.

9. The compressor rotor assembly quality online detection system according to claim 8, characterized in that: The rotor assembly stability warning module is used to calculate the rotor assembly comprehensive score Rs based on the aggregation tendency factor Jq and the oil film formation delay factor Ym. The rotor assembly comprehensive score Rs is obtained as follows: ； Where, and Jq represent the oil film formation delay factor and aggregation tendency factor, respectively. represents the sensitivity adjustment factor, It represents the maximum oil film formation delay factor allowed by the compressor, and e represents the base of the natural logarithm; A comprehensive scoring threshold YZ is preset, and the rotor assembly comprehensive score Rs and the comprehensive scoring threshold YZ are compared and analyzed to evaluate the compressor rotor assembly quality. The specific evaluation contents are as follows: If the rotor assembly comprehensive score Rs is less than or equal to the comprehensive score threshold YZ, the oil film response is timely and the compressor rotor assembly quality is normal. In this case, the "assembly structure qualified" label is output, the aggregation tendency factor Jq and the oil film formation delay factor Ym are recorded, and a command is sent to control the normal operation of the compressor system. If the rotor assembly comprehensive score Rs is greater than the comprehensive score threshold YZ, it is determined that the oil film formation delay is abnormal, the frequency aggregation trend is obvious, the compressor has a hidden danger of inducing structural resonance, and the compressor rotor assembly quality is at risk. At this time, the "assembly structure abnormality risk" label is output, the equipment is suspended, and the quality inspector is notified to perform a load test. At the same time, recommended inspection items are generated, including "bearing installation eccentricity check", "initial rotor balance state check", "end face contact deviation check" and "lubrication chamber residual impurity check", and the characteristic data is recorded and stored as "fault samples" in the data analysis library.

10. A method for online detection of compressor rotor assembly quality, for implementing a system for online detection of compressor rotor assembly quality according to any one of claims 1 to 9, comprising the following steps: Step 1: Collect the compressor microgravity operation data, perform microgravity operation simulation, and construct a disturbance sampling window, which is then sent to the micro-vibration monitoring and analysis module for data collection and analysis. Step 2: Real-time collection of the axial acceleration response signal of the compressor rotor within the disturbance sampling window , and perform feature extraction to obtain a feature data set; Step 3: Based on the characteristic data set, calculate and obtain the oil film formation delay factor Ym, evaluate the oil film delay risk level, and trigger the aggregation trend analysis module to perform frequency domain aggregation trend analysis; Step 4: Analyze the frequency domain aggregation trend and calculate the aggregation trend factor Jq; Step 5: Based on the oil film formation delay factor Ym and the aggregation tendency factor Jq, calculate and obtain the rotor assembly comprehensive score Rs, evaluate the compressor rotor assembly quality, and issue an assembly quality warning.

Images