Over-critical feature-based engine complete machine vibration fault diagnosis method and system

Abstract

The invention belongs to the technical field of engine fault diagnosis, and provides an engine complete machine vibration fault diagnosis method and system based on an over-critical characteristic, and the method comprises the steps: carrying out the real-time monitoring of an engine in an operation state through a laser vibration meter in an operation monitoring period, obtaining the operation amplitude data, and carrying out the calculation of the fault of the engine according to the operation amplitude data, the method comprises the following steps: performing difference analysis on phases of a plurality of adjacent monitoring points in an engine, and combining operation amplitude data to obtain a shafting fault diagnosis value, so as to perform fault diagnosis on the engine according to the difference degree between the corresponding phases of the adjacent monitoring points and the difference degree between the amplitudes, thereby identifying whether the engine has a fault or not; and the engine fault type is further identified, so that the diagnosis accuracy of the engine vibration fault is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of engine fault diagnosis, and specifically relates to an engine overall vibration fault diagnosis method and system based on over-critical characteristics. Background Art

[0002] Overall vibration fault is one of the common problems of engines. Traditional fault diagnosis methods often rely on single-parameter analysis. For example, only judging faults by vibration amplitude, the diagnostic accuracy is relatively low, and it is difficult to accurately identify the fault type and locate the fault position.

[0003] In the prior art, most rely on single parameters or simple analysis means. For example, only judging faults by the magnitude of vibration amplitude, the diagnostic accuracy is poor, different types of faults cannot be effectively distinguished, and it is also difficult to accurately locate the position where the fault occurs. Therefore, this application solves the problem of how to accurately identify the type of shaft misalignment fault in the engine shafting by analyzing the relationship between the shaft rotation frequency and the phase difference frequency and the periodicity of the phase difference when the engine shafting fails, improves the diagnostic accuracy of engine vibration faults, and further identifies the engine fault type, provides data support for subsequent maintenance personnel to carry out targeted maintenance, speeds up the efficiency of engine fault maintenance, and further determines the key over-critical characteristics.

[0004] Therefore, the present invention provides an engine overall vibration fault diagnosis method and system based on over-critical characteristics. Summary of the Invention

[0005] In order to make up for the deficiencies of the prior art and solve at least one of the technical problems proposed in the background art.

[0006] The technical solution adopted by the present invention to solve its technical problems is as follows: In a first aspect, an engine overall vibration fault diagnosis method based on over-critical characteristics includes the following steps: During the operation monitoring period, a laser vibrometer is used to monitor the engine in the running state in real time to obtain running amplitude data; Based on the running amplitude data, the phases of multiple monitoring points in the engine are obtained, the phase difference analysis between adjacent monitoring points is carried out, and combined with the running amplitude data, it is evaluated whether the engine shafting fails; If the engine shafting fails, the shaft rotation frequency is obtained through a speed sensor, and the correlation with the phase difference frequency is analyzed to determine whether the phase difference changes periodically, and the shafting fault type is identified; Record the diagnostic results of multiple shafting fault types and compare them with the actual diagnostic results actually detected by the corresponding maintenance personnel to evaluate whether the diagnostic results of the shafting fault type are accurate, and obtain the key over-critical characteristics.

[0007] A further solution of the present invention is as follows: The process of obtaining the running amplitude data is as follows: Evenly divide the running monitoring period into several monitoring nodes; Evenly divide the engine surface into several monitoring points, obtain the running amplitude of each monitoring point at each monitoring node through a laser vibrometer, sort them according to the time series, and integrate them into running amplitude data.

[0008] A further solution of the present invention is as follows: The process of obtaining the phase of the monitoring point is as follows: Taking the X-axis as time and the Y-axis as amplitude, construct a running amplitude change curve based on the running amplitude data, and extract all the wave peaks within the running amplitude change curve; Respectively obtain the time difference between adjacent wave peaks corresponding to the monitoring nodes, and obtain the monitoring point phase through the phase acquisition formula.

[0009] A further solution of the present invention is as follows: The process of analyzing the phase difference between adjacent monitoring points is as follows: Based on the same time series, process the monitoring point phases corresponding to adjacent monitoring points through the Manhattan distance formula, and output the phase difference value.

[0010] A further solution of the present invention is as follows: Combining the running amplitude data, evaluate whether the engine shafting fails, and the process is as follows: Extract all the running amplitudes in the running amplitude data corresponding to the monitoring points, combine the running amplitudes corresponding to different monitoring points within the same monitoring node to obtain multiple amplitude analysis groups, and process them through the Manhattan distance formula, and output the amplitude difference value; Input the phase difference values and amplitude difference values corresponding to adjacent monitoring points into the geometric product model, and perform mean calculation, output the shafting fault diagnosis value. If the shafting fault diagnosis value is greater than the shafting fault diagnosis threshold, generate a shafting fault signal.

[0011] A further solution of the present invention is as follows: The rotational speed sensor obtains the shaft rotation frequency, and the process is as follows: According to the electromagnetic induction law and the alternating signal frequency output by the magnetoelectric rotational speed sensor, calculate the shaft rotation frequency through the shaft rotation frequency formula within the running monitoring period.

[0012] A further solution of the present invention is as follows: Analyze the correlation between the shaft rotation frequency and the phase difference frequency, and the process is as follows: Statistically analyze multiple phase differences corresponding to adjacent monitoring points, take the X-axis as time and the Y-axis as the phase difference, and construct a phase difference change curve; On the phase difference change curve, respectively extract the phase difference wave peaks; Sort all phase difference frequencies and shaft system rotation frequencies according to the time series respectively to obtain a phase difference frequency sequence and a shaft system rotation frequency sequence; Input all the phase difference frequencies in the phase difference frequency sequence and all the shaft system rotation frequencies in the shaft system rotation frequency sequence into the improved Pearson correlation coefficient model respectively, and output the frequency correlation coefficient; If the frequency correlation coefficient is less than or equal to the frequency correlation standard coefficient, the correlation degree between the shaft rotation frequency and the phase difference frequency is relatively tight.

[0013] As a further solution of the present invention: Determine whether the phase difference changes periodically and identify the shaft system fault type. The process is as follows: If the correlation degree between the shaft rotation frequency and the phase difference frequency is relatively tight, after averaging all the shaft rotation frequencies, take the reciprocal to obtain the shaft rotation period; Based on the duration corresponding to the shaft rotation period, divide the phase difference change curve to obtain several periodic analysis curves. Respectively obtain the starting coordinates and ending coordinates corresponding to the periodic analysis curves, and perform fitting by the least square method to obtain the periodic analysis fitting line. Use the slope calculation formula for the starting coordinates and the terminal coordinates to obtain the slope of the periodic analysis fitting line; Arbitrarily combine the slopes of adjacent periodic analysis fitting lines and input them into the Manhattan distance formula, and output the periodic trend analysis value; Input the starting Y coordinates and ending Y coordinates corresponding to all the periodic analysis curves into the Manhattan distance formula for calculation to obtain the periodic deviation analysis value; Calculate the periodic trend analysis value and the periodic deviation analysis by the geometric product method to obtain the periodic determination value; If the periodic determination value is less than or equal to the periodic determination threshold, it is determined that the shaft misalignment fault type is angular misalignment.

[0014] As a further solution of the present invention: Determine whether the phase difference is a key critical passing feature. The process is as follows: Sort the diagnosis results of multiple shaft system fault types and the corresponding actual diagnosis results according to the diagnosis time series, and respectively construct a diagnosis result sequence and an actual diagnosis result sequence; Arbitrarily extract a diagnosis result from the diagnosis result sequence, and correspondingly extract an actual diagnosis result from the actual diagnosis result sequence for combination to obtain multiple diagnosis result comparison groups. Obtain the proportion of the number of diagnosis result coincidence groups in the total number of diagnosis result groups to obtain the diagnosis result coincidence proportion. If the diagnosis result coincidence proportion is greater than the diagnosis result coincidence proportion threshold, record the phase difference as a key critical passing feature.

[0015] In a second aspect, an engine overall vibration fault diagnosis system based on critical passing features includes: Amplitude data acquisition module: During the operation monitoring period, the engine under operation is monitored in real time by a laser vibrometer to obtain the running amplitude data; Fault diagnosis and analysis module: Based on the running amplitude data, the phases of multiple monitoring points in the engine are obtained, the phase difference between adjacent monitoring points is analyzed, and combined with the running amplitude data, it is evaluated whether there is a fault in the engine shafting; Shafting fault identification module: If there is a fault in the engine shafting, the shaft rotation frequency is obtained through a speed sensor, and the correlation with the phase difference frequency is analyzed to determine whether the phase difference changes periodically, and the type of shafting fault is identified; Key feature screening module: Records the diagnostic results of multiple shafting fault types, and compares them with the actual diagnostic results detected by the corresponding maintenance personnel to evaluate whether the diagnostic results of the shafting fault types are accurate, and obtains the key critical passing features.

[0016] The beneficial effects of the present invention are as follows: 1. During the operation monitoring period of the present invention, the engine under operation is monitored in real time by a laser vibrometer to obtain the running amplitude data. Based on the running amplitude data, the phase difference between multiple adjacent monitoring points in the engine is analyzed, and combined with the running amplitude data, the shafting fault diagnosis value is obtained. Thus, the engine is fault diagnosed respectively from the difference degree between the corresponding phases of adjacent monitoring points and the difference degree between amplitudes, not only identifying whether the engine has a fault, but also further identifying the type of engine fault, thereby improving the diagnostic accuracy of the engine vibration fault; 2. The present invention uses a speed sensor such as a magnetoelectric type to calculate the shaft rotation frequency, simultaneously counts the phase difference between adjacent monitoring points to construct a curve and calculates the phase difference frequency, and then uses the Pearson correlation coefficient improved model to analyze the correlation between the two. If the correlation is close, then according to the shaft rotation period, the phase difference curve is divided, and the period determination value is obtained by calculating the slope of the fitting line, the trend analysis value, and the deviation analysis value, thereby identifying whether the shafting fault is angular misalignment, solving the problem of how to accurately identify the type of angular misalignment fault of the shafting by analyzing the relationship between the shaft rotation frequency and the phase difference frequency and the periodicity of the phase difference when there is a fault in the engine shafting, improving the diagnostic accuracy of the engine vibration fault, providing data support for the subsequent targeted maintenance of maintenance personnel, and accelerating the efficiency of engine fault maintenance; 3. The diagnostic results of multiple shafting fault types in the present invention are compared with the actual diagnostic results detected by the corresponding maintenance personnel to obtain a diagnostic result coincidence group and a diagnostic result non - coincidence group. If the diagnostic result coincidence group is smaller than the result non - coincidence group, it indicates that the accuracy of the diagnostic results for diagnosing the engine shafting fault types by whether there is periodicity in the phase difference change is relatively low. If the diagnostic result coincidence group is larger than the result non - coincidence group, it indicates that the accuracy of the diagnostic results for diagnosing the engine shafting fault types by whether there is periodicity in the phase difference change is relatively high. The phase difference is the key critical - passing feature. Thus, it can not only identify whether there is periodicity in the phase difference change and the accuracy level of the results for diagnosing the engine shafting fault types, but also provide a key diagnostic direction for subsequent diagnosis, improving the efficiency of engine diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below with reference to the drawings.

[0018] Figure 1 is a flowchart of the steps of a method for diagnosing the vibration fault of an entire engine based on the critical - passing feature of the present invention; Figure 2 is a schematic diagram of a system for diagnosing the vibration fault of an entire engine based on the critical - passing feature of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0019] In order to make the technical means, creative features, achieved purposes and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0020] Embodiment 1:

[0021] Please refer to Figure 1 As shown, a method for diagnosing the vibration fault of an entire engine based on the critical - passing feature in the embodiment of the present invention, wherein the critical - passing feature includes amplitude, vibration phase difference, vibration frequency, rotational speed, etc. Since different types of faults will cause the engine vibration to show different characteristics in amplitude and phase, therefore, combining the amplitude and the phase difference for diagnosing the vibration fault of the entire engine is beneficial to identifying the vibration fault type while identifying the vibration fault of the entire engine. It not only improves the diagnostic accuracy of the vibration fault of the entire engine, but also improves the subsequent maintenance efficiency of the vibration fault on the basis of identifying the vibration fault type, and includes the following steps: Step 1: During the operation monitoring period, use a laser vibrometer to monitor the engine in the running state in real - time to obtain the running amplitude data; In a preferred embodiment, the engine surface is evenly divided into several monitoring points; It should be noted that the distance between adjacent monitoring points is equal; The engine in the operating state is monitored in real time by a laser vibrometer, and the process is as follows: The laser vibrometer is used to emit laser to any monitoring point on the surface of the engine in the operating state. Due to the principle of laser interference, the frequency of the reflected laser will change due to the vibration of the engine in the operating state. Therefore, the operating amplitude is obtained; It should be noted that when using the laser vibrometer to emit laser to any monitoring point on the surface of the engine in the operating state, the purpose is as follows: since the surface of the engine is evenly divided into several monitoring points, the accuracy of engine fault diagnosis can be improved by comparing the amplitude changes or phase difference changes of different monitoring points during the operating monitoring period, providing data support for diagnosing engine faults; The operating monitoring period is evenly divided into several monitoring nodes; It should be noted that the time intervals between adjacent monitoring nodes are equal; The monitoring node corresponding to each operating amplitude is obtained and sorted according to the time series, and integrated into operating amplitude data; Step two: Based on the operating amplitude data, obtain the phases of multiple monitoring points in the engine, perform adjacent monitoring point phase difference analysis, and combine the operating amplitude data to evaluate whether the engine shafting has failed; In a preferred embodiment, with the X-axis as time and the Y-axis as amplitude, based on the operating amplitude data, an operating amplitude change curve is constructed and analyzed by the peak method. The process is as follows: All the wave peaks or wave valleys in the operating amplitude change curve are extracted; Exemplarily, taking the wave peak as an example, the time difference between adjacent wave peak corresponding monitoring nodes is obtained respectively. Through the phase acquisition formula: , the monitoring point phase is calculated , where represents the time difference between adjacent wave peak corresponding monitoring nodes, represents the operating monitoring period; Based on the same time series, the monitoring point phases corresponding to adjacent monitoring points are processed by the Manhattan distance formula, and the phase difference value is output ; It should be noted that based on the same time series, that is, the monitoring nodes corresponding to adjacent monitoring points are nodes A, B, and C respectively; The monitoring point phases of adjacent monitoring points at the AB monitoring nodes and the monitoring point phases at the BC monitoring nodes are respectively processed by the Manhattan distance formula to obtain the phase difference value; Specifically, the Manhattan distance formula: , where represents the pth monitoring point phase corresponding to one of the adjacent monitoring points, The phase of the p-th monitoring point, which is represented as the teammate of another monitoring point among adjacent monitoring points Represents the total number of adjacent monitoring nodes; Any monitoring point; Extract all the running amplitudes within the running amplitude data corresponding to the monitoring points, combine the running amplitudes corresponding to different monitoring points within the same monitoring node, obtain multiple amplitude analysis groups, and process them through the Manhattan distance formula to output the amplitude difference value ; Specifically, the amplitude difference formula: , where Represents the total number of amplitude analysis groups, Represents the One of the running amplitudes within the i-th amplitude analysis group, Is the Another running amplitude within the i-th amplitude analysis group; It should be noted that the function of using the Manhattan distance formula is as follows: Function 1: From the perspective of amplitude difference analysis, processing the amplitudes of the same monitoring node corresponding to adjacent monitoring points can overall quantify the difference degree between the amplitudes corresponding to all monitoring points on the engine surface, and then determine whether the engine has a fault; Function 2: From the perspective of phase difference analysis, processing the phases corresponding to adjacent monitoring points can overall quantify the difference degree between the phases corresponding to all monitoring points on the engine surface, and then, on the basis of amplitude difference judgment, further determine whether there is a fault in the engine shafting, improving the accuracy of identifying the type of engine fault and the accuracy of identifying the fault location; Input the phase difference value and amplitude difference value corresponding to adjacent monitoring points into the geometric product model, and perform mean calculation to output the shafting fault diagnosis value; It can be understood that the meaning represented by the shafting fault diagnosis value is: obtained by inputting the phase difference value and adjacent amplitude difference value corresponding to adjacent monitoring points into the geometric product model and performing mean calculation. Specifically, the phase difference value quantifies the difference degree between the phases corresponding to adjacent monitoring points, while the adjacent amplitude difference value quantifies the difference degree between the amplitudes corresponding to different monitoring points within the same monitoring node. Now, based on the two, it can not only identify whether the engine has a fault, but also further identify the type of engine fault; If the shafting fault diagnosis value is greater than the shafting fault diagnosis threshold, it indicates that the amplitude and phase differences between adjacent monitoring points are relatively obvious, and a shafting fault signal is generated; If the shafting fault diagnosis value is less than or equal to the shafting fault diagnosis threshold, it indicates that the amplitude and phase differences between adjacent monitoring points are relatively less obvious, and a shafting normal signal is generated; The technical solution of the embodiment is as follows: During the operation monitoring period, the engine in the operating state is monitored in real time by a laser vibrometer to obtain the operating amplitude data. Based on the operating amplitude data, the phase differences between multiple adjacent monitoring points in the engine are analyzed, and combined with the operating amplitude data, the shafting fault diagnosis value is obtained. Thus, the engine is diagnosed for faults from the difference degree between the corresponding phases of adjacent monitoring points and the difference degree between the amplitudes, not only identifying whether the engine has a fault, but also further identifying the type of engine fault, thereby improving the diagnostic accuracy of the engine vibration fault.

[0022] Embodiment 2:

[0023] Please refer to Figure 1 As shown, a method for diagnosing the overall vibration fault of an engine based on the critical speed characteristics according to an embodiment of the present invention includes the following steps: Step Three: If there is a shafting fault in the engine, obtain the shaft rotation frequency through a speed sensor, analyze the correlation with the phase difference frequency, determine whether the phase difference changes periodically, and diagnose the type of shafting fault; In a preferred embodiment, the shaft rotation frequency is obtained through a speed sensor, and the process is as follows: It should be noted that the speed sensor includes an electromagnetic speed sensor, a Hall sensor, or an optoelectronic sensor; Exemplarily, taking the electromagnetic speed sensor as an example, according to the electromagnetic induction law and the alternating signal frequency after the output of the electromagnetic speed sensor, within the operation monitoring period, through the shaft rotation frequency formula: , the shaft rotation frequency is calculated , where represents the alternating signal frequency, represents the total number of teeth on the induction gear in the electromagnetic speed sensor; Statistically analyze the multiple phase differences corresponding to adjacent monitoring points. Taking the X-axis as time and the Y-axis as the phase difference, construct a phase difference change curve; On the phase difference change curve, extract the phase difference wave peaks or phase difference wave valleys respectively; Exemplarily, taking the phase difference wave peak as an example, extract the interval time corresponding to adjacent wave peaks, perform averaging processing and then take the reciprocal to output the phase difference frequency; Sort all the phase difference frequencies in time series to obtain a phase difference frequency sequence; Similarly, sort all the shafting rotation frequencies in time series to obtain a shafting rotation frequency sequence; Analyze the correlation degree between the phase difference frequencies in the phase difference frequency sequence and the shafting rotation frequencies in the shafting rotation frequency sequence through the Pearson correlation coefficient improvement model. The process is as follows: A1, average all the phase difference frequencies in the phase difference frequency sequence, and output the average phase difference frequency; Average all the shaft rotation frequencies in the shaft rotation frequency sequence, and output the average rotation frequency; A2, input the average phase difference frequency and the average rotation frequency into the improved Pearson correlation coefficient model respectively, and output the frequency correlation coefficient ; Specifically, the improved Pearson correlation coefficient model: , where represents the average phase difference frequency, represents the e-th phase difference frequency in the phase difference frequency sequence, represents the average rotation frequency, represents the e-th shaft rotation frequency in the shaft rotation frequency sequence; It should be noted that since the Pearson correlation coefficient is within , if the Pearson correlation coefficient is close to 0, it means that the degree of correlation is not close. If the Pearson correlation coefficient is close to 1 or -1, it means that the degree of correlation is close. Therefore, the absolute value of the Pearson correlation coefficient is processed. If the Pearson correlation coefficient is close to 1 or -1, the frequency correlation coefficient obtained according to the improved Pearson correlation coefficient model is smaller, indicating that the correlation degree between the shaft rotation frequency and the phase difference frequency is relatively close. If the Pearson correlation coefficient is close to 0, the frequency correlation coefficient obtained according to the improved Pearson correlation coefficient model is larger, indicating that the correlation degree between the shaft rotation frequency and the phase difference frequency is less close; If the frequency correlation coefficient is greater than the frequency correlation standard coefficient, it means that the correlation degree between the shaft rotation frequency and the phase difference frequency is less close; If the frequency correlation coefficient is less than or equal to the frequency correlation standard coefficient, it means that the correlation degree between the shaft rotation frequency and the phase difference frequency is relatively close. Then, after averaging all the shaft rotation frequencies, take the reciprocal to obtain the shaft rotation period; Based on the duration corresponding to the shaft rotation period, divide the phase difference change curve to obtain several periodic analysis curves; Arbitrarily select a periodic analysis curve, respectively obtain the starting coordinate and the ending coordinate corresponding to the periodic analysis curve, and perform fitting by the least squares method to obtain the periodic analysis fitting line. Use the slope calculation formula for the starting coordinate and the terminal coordinate to obtain the slope of the periodic analysis fitting line; Arbitrarily combine the slopes of adjacent periodic analysis fitting lines and input them into the Manhattan distance formula, and output the periodic trend analysis value; After inputting the starting Y coordinate and the ending Y coordinate corresponding to all the periodic analysis curves into the Manhattan distance formula for calculation, a periodic deviation analysis value is obtained; The periodic trend analysis value and the periodic deviation analysis are calculated by the geometric product method to obtain a periodic determination value; It can be understood that the meaning represented by the periodic determination value is: a value obtained by comprehensively considering the slope change of the periodic analysis curve (reflected by the periodic trend analysis value) and the change amplitude of the phase difference within one period (reflected by the periodic deviation analysis value), which is used to comprehensively judge the periodic characteristics of the phase difference change curve; Compare the periodic determination value with the periodic determination threshold, and the process is as follows: If the periodic determination value is greater than the periodic determination threshold, it indicates that there is no periodicity in the phase difference change within the shaft rotation period, and a non-periodic phase difference signal is generated; If the periodic determination value is less than or equal to the periodic determination threshold, it indicates that there is periodicity in the phase difference change within the shaft rotation period, a periodic phase difference signal is generated, and it is determined that the misalignment fault type of the shafting is angular misalignment; The specific implementation scheme of this embodiment is: use a magnetic-electric equal-speed sensor to calculate the shaft rotation frequency, and at the same time count the phase difference between adjacent monitoring points to construct a curve and calculate the phase difference frequency. Then use the Pearson correlation coefficient improved model to analyze the correlation between the two. If the correlation is close, then according to the shaft rotation period, divide the phase difference curve accordingly. By calculating the slope of the fitting line, the trend analysis value, and the deviation analysis value, a periodic determination value is obtained, thereby identifying whether the shafting fault is angular misalignment. This solves the problem of how to accurately identify the angular misalignment fault type of the shafting by analyzing the relationship between the shaft rotation frequency and the phase difference frequency and the periodicity of the phase difference during the engine shafting fault, improves the accuracy of the engine vibration fault diagnosis, provides data support for the subsequent maintenance personnel to carry out targeted maintenance, and speeds up the efficiency of the engine fault repair.

[0024] Embodiment 3:

[0025] A method for diagnosing engine overall vibration faults based on critical speed characteristics according to an embodiment of the present invention further includes the following steps: Step Four: Record the diagnosis results of multiple shafting fault types and compare them with the actual diagnosis results actually detected by the corresponding maintenance personnel to determine whether the phase difference is a key critical speed characteristic; Among them, the diagnosis results include angular misalignment or parallel misalignment; Exemplarily, sort the diagnosis results of multiple shafting fault types according to the time series of each diagnosis to obtain a diagnosis result sequence; For example, the diagnosis result sequence can be diagnosing angular misalignment, diagnosing angular misalignment, diagnosing parallel misalignment, diagnosing angular misalignment, diagnosing parallel misalignment; Similarly, the actual diagnosis results of multiple corresponding maintenance personnel's actual detections are sorted according to the time series of each actual detection to obtain an actual diagnosis result sequence; For example, the actual diagnosis result sequence can be actual diagnosis angular misalignment, actual diagnosis parallel misalignment, actual diagnosis parallel misalignment, actual diagnosis angular misalignment, actual diagnosis angular misalignment; Arbitrarily extract a diagnosis result from the diagnosis result sequence, and correspondingly extract an actual diagnosis result from the actual diagnosis result sequence, and combine them to obtain multiple diagnosis result comparison groups; It should be noted that the method of extracting and combining into diagnosis result comparison groups is: extracting and combining according to the sorting correspondence of the results in the diagnosis result sequence and the actual diagnosis result sequence, and the number of results in the diagnosis result sequence and the actual diagnosis result sequence is the same, and the number of results is odd; For example: diagnosis angular misalignment and actual diagnosis angular misalignment are a group of diagnosis result comparison groups, diagnosis angular misalignment and actual diagnosis parallel misalignment are a group of diagnosis result comparison groups, diagnosis parallel misalignment and actual diagnosis angular misalignment are a group of diagnosis result comparison groups, diagnosis parallel misalignment and actual diagnosis angular misalignment are a group of diagnosis result comparison groups; Based on multiple diagnosis result comparison groups, if the diagnosis result in the diagnosis result comparison group is consistent with the actual diagnosis result, then record it as a diagnosis result coincidence group; If the diagnosis result in the diagnosis result comparison group is inconsistent with the actual diagnosis result, then record it as a diagnosis result non - coincidence group; Count the number of diagnosis result coincidence groups, and calculate the ratio with the total number of diagnosis result groups, and output the diagnosis result coincidence ratio; If the diagnosis result coincidence ratio is greater than the diagnosis result coincidence ratio threshold, it indicates that the diagnosis result accuracy of the engine shafting fault type by whether there is periodicity in the phase difference change is relatively high, and the phase difference is the key critical - passing feature; If the diagnosis result coincidence ratio is less than or equal to the diagnosis result coincidence ratio threshold, it indicates that the diagnosis result accuracy of the engine shafting fault type by whether there is periodicity in the phase difference change is relatively low, and the phase difference is a non - key critical - passing feature; The technical solution of this embodiment is as follows: Compare the diagnostic results of multiple shafting fault types with the actual diagnostic results detected by the corresponding maintenance personnel to obtain a diagnostic result coincidence group and a diagnostic result non-coincidence group. If the coincidence ratio of the diagnostic results is less than or equal to the diagnostic result coincidence ratio threshold, it indicates that the accuracy of the diagnostic results for diagnosing the engine shafting fault types by whether there is periodicity in the phase difference change is relatively low. If the coincidence ratio of the diagnostic results is greater than the diagnostic result coincidence ratio threshold, it indicates that the accuracy of the diagnostic results for diagnosing the engine shafting fault types by whether there is periodicity in the phase difference change is relatively high. The phase difference is a key critical passing feature, so that not only can it identify whether there is periodicity in the phase difference change and the accuracy of the results for diagnosing the engine shafting fault types, but also it can provide a key diagnostic direction for subsequent diagnosis and improve the efficiency of engine diagnosis.

[0026] Embodiment 4:

[0027] As Figure 2 shown, a whole-engine vibration fault diagnosis system for an engine based on critical passing features according to an embodiment of the present invention includes the following modules: Amplitude data acquisition module: During the operation monitoring period, use a laser vibrometer to monitor the engine in the running state in real time to obtain running amplitude data; Fault diagnosis and analysis module: Based on the running amplitude data, obtain the phases of multiple monitoring points in the engine, perform adjacent monitoring point phase difference analysis, and combine the running amplitude data to evaluate whether the engine shafting has a fault; Shafting fault identification module: If there is a fault in the engine shafting, obtain the shaft rotation frequency through a speed sensor, analyze the correlation with the phase difference frequency, and determine whether the phase difference changes periodically to identify the shafting fault type; Key feature determination module: Record the diagnostic results of multiple shafting fault types and compare them with the actual diagnostic results detected by the corresponding maintenance personnel to determine whether the phase difference is a key critical passing feature.

[0028] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. An engine overall vibration fault diagnosis method based on supercritical characteristics, characterized in that: Including: During the operation monitoring period, the engine in the operating state is monitored in real time by a laser vibrometer to obtain the operating amplitude data; Based on the operating amplitude data, the phases of multiple monitoring points in the engine are obtained, the phase difference analysis between adjacent monitoring points is carried out, and combined with the operating amplitude data, it is evaluated whether the engine shafting fails; If the engine shafting fails, the shaft rotation frequency is obtained through a speed sensor, and the correlation with the phase difference frequency is analyzed to determine whether the phase difference changes periodically, and the type of shafting failure is identified; Record the diagnostic results of multiple shafting failure types, and compare them with the actual diagnostic results detected by the corresponding maintenance personnel to evaluate whether the diagnostic results of the shafting failure types are accurate, and obtain the key critical passing characteristics.

2. The engine overall vibration fault diagnosis method based on overcritical characteristics according to claim 1, characterized in that: The process of obtaining the operating amplitude data is as follows: The operation monitoring period is evenly divided into several monitoring nodes; The engine surface is evenly divided into several monitoring points, the operating amplitudes of the monitoring points at the monitoring nodes are obtained, sorted according to the time series, and integrated into the operating amplitude data.

3. A method for diagnosing engine overall vibration faults based on supercritical characteristics according to claim 1, characterized in that: The process of obtaining the phase of the monitoring point is as follows: Based on the operating amplitude data, an operating amplitude change curve is constructed, all wave peaks are extracted, the time difference between adjacent wave peak corresponding monitoring nodes is obtained respectively, and the phase of the monitoring point is obtained through the phase acquisition formula.

4. A method for diagnosing engine vibration faults based on supercritical characteristics according to claim 3, characterized in that: The process of analyzing the phase difference between adjacent monitoring points is as follows: Based on the same time series, the monitoring point phases corresponding to adjacent monitoring points are input into the Manhattan distance formula, and the phase difference value is output.

5. A method for diagnosing engine overall vibration faults based on supercritical characteristics according to claim 1, characterized in that: The process of evaluating whether the engine shafting fails by combining the operating amplitude data is as follows: Extract all the operating amplitudes in the operating amplitude data corresponding to the monitoring points, extract the operating amplitudes of different monitoring points in the same monitoring node period for combination, obtain multiple amplitude analysis groups, and process them through the Manhattan distance formula, and the amplitude difference value is output; Input the phase difference values and amplitude difference values corresponding to adjacent monitoring points into the geometric product model, and perform mean calculation, and the shafting failure diagnosis value is output. If it is greater than the shafting failure diagnosis threshold, a shafting failure signal is generated.

6. The engine overall vibration fault diagnosis method based on supercritical characteristics according to claim 1, characterized in that: The process of the speed sensor obtaining the shaft rotation frequency is as follows: According to the electromagnetic induction law and the alternating signal frequency after the output of the magnetoelectric speed sensor, the shaft rotation frequency is calculated through the shaft rotation frequency formula during the operation monitoring period.

7. A method for diagnosing engine overall vibration faults based on supercritical characteristics according to claim 1, characterized in that: The process of analyzing the correlation between the shaft rotation frequency and the phase difference frequency is as follows: Statistically analyze the multiple phase differences corresponding to adjacent monitoring points, construct a phase difference change curve, extract the interval time between all adjacent phase difference wave peaks respectively, perform averaging processing and then take the reciprocal to obtain the phase difference frequency; Sort all the phase difference frequencies and the shafting rotation frequencies according to the time series respectively to obtain the phase difference frequency sequence and the shafting rotation frequency sequence; Input all the phase difference frequencies in the phase difference frequency sequence and all the shafting rotation frequencies in the shafting rotation frequency sequence into the improved Pearson correlation coefficient model respectively, and the frequency correlation coefficient is output; If the frequency correlation coefficient is less than or equal to the frequency correlation standard coefficient, the correlation degree between the shaft rotation frequency and the phase difference frequency is relatively close.

8. A method for diagnosing the vibration fault of an entire engine based on the supercritical characteristics according to claim 7, characterized in that: Determine whether the phase difference changes periodically and identify the type of shafting fault. The process is as follows: If the correlation between the shaft rotation frequency and the phase difference frequency is relatively close, after averaging all the shaft rotation frequencies, take the reciprocal to obtain the shaft rotation period; Based on the duration corresponding to the shaft rotation period, divide the phase difference change curve and perform fitting by the least squares method to obtain multiple periodic analysis fitting lines and the slopes of the periodic analysis fitting lines; Input the slopes of all adjacent periodic analysis fitting lines into the Manhattan distance formula, and output to obtain the periodic trend analysis value; Input the starting Y coordinates and ending Y coordinates of all periodic analysis curves into the Manhattan distance formula for calculation to obtain the periodic deviation analysis value; Calculate the geometric product of the periodic trend analysis value and the periodic deviation analysis to obtain the periodic determination value; If the periodic determination value is less than or equal to the periodic determination threshold, determine that the type of shaft misalignment fault is angular misalignment.

9. A method for diagnosing the vibration fault of an entire engine based on supercritical characteristics according to claim 1, characterized in that: Judge whether the phase difference is a key critical passing feature. The process is as follows: Sort the diagnostic results of multiple shafting fault types and the corresponding actual diagnostic results according to the diagnostic time series, and construct a diagnostic result sequence and an actual diagnostic result sequence respectively; Arbitrarily extract a diagnostic result from the diagnostic result sequence, and correspondingly extract an actual diagnostic result from the actual diagnostic result sequence for combination to obtain multiple diagnostic result comparison groups. Obtain the proportion of the number of diagnostic result coincidence groups in the total number of diagnostic result groups to get the diagnostic result coincidence proportion. If the diagnostic result coincidence proportion is greater than the diagnostic result coincidence proportion threshold, record the phase difference as a key critical passing feature.

10. An engine overall vibration fault diagnosis system based on supercritical characteristics, characterized in that: It includes the following modules: Amplitude data acquisition module: During the operation monitoring period, use a laser vibrometer to monitor the engine in operation in real time to obtain the operation amplitude data; Fault diagnosis and analysis module: Based on the operation amplitude data, obtain the phases of multiple monitoring points in the engine, perform adjacent monitoring point phase difference analysis, and combine with the operation amplitude data to evaluate whether the engine shafting has a fault; Shafting fault identification module: If there is an engine shafting fault, obtain the shaft rotation frequency through a speed sensor, analyze the correlation with the phase difference frequency, determine whether the phase difference changes periodically, and identify the type of shafting fault; Key feature screening module: Record the diagnostic results of multiple shafting fault types and compare them with the actual diagnostic results detected by the corresponding maintenance personnel to judge whether the phase difference is a key critical passing feature.

Images