A vibration-based multi-dimensional aircraft accessory casing health assessment method

Abstract

The present invention provides a vibration-based multi-dimensional aircraft accessory casing health assessment method, which belongs to the technical field of fixed-wing aircraft health management. The multi-dimensional aircraft accessory casing health assessment method first determines the vibration monitoring state index, determines the flight state, and determines the working state parameters that affect the health state of the flight accessory casing; secondly, establishes a health assessment model, and optimizes the model parameters by applying data. Compared with the traditional aircraft accessory casing health monitoring function, the vibration-based multi-dimensional aircraft accessory casing health assessment model proposed by the present invention adds vibration data that can directly reflect the health state of the flight accessory casing, thereby increasing the effectiveness of health monitoring; the vibration data is analyzed and processed according to different flight states and working states, and the lubricating oil monitoring results are referred to to increase the accuracy of health monitoring; the health index integrates all factors that can reflect the aircraft accessory casing, and can perform trend analysis and fault diagnosis, thereby improving the practicality of aircraft accessory casing health monitoring.

Description

Technical Field

[0001] The present invention belongs to the technical field of fixed-wing aircraft health management, and relates to a vibration-based multi-dimensional aircraft accessory casing (hereinafter referred to as: flight accessory casing) health assessment method, which is applied to a fixed-wing aircraft flight accessory casing health monitoring system. Background Art

[0002] The auxiliary casing connects the engine and starter of the fixed-wing aircraft. During the starting process, the auxiliary power unit first drives the starter to operate, and the starter drives the engine turbine through the auxiliary casing until the engine is fully started; when the engine starts and can work independently, the starter is disengaged, and the engine drives the generator, oil pump and other accessories through the auxiliary casing to provide electricity and oil for the normal operation of the engine. The auxiliary casing plays a vital role in the engine starting process and the stable operation after starting.

[0003] In order to ensure the normal operation of the flight attachment casing, it is necessary to monitor the health performance of the flight attachment casing. The existing conventional solution is to monitor the working state parameters such as speed, torque, lubricating oil temperature, lubricating oil pressure, lubricating oil metal chips, etc., and set a normal working range or threshold for each parameter to monitor whether this series of working state parameters are within the normal range to determine whether the flight attachment casing is working normally. However, this conventional method still has the following shortcomings:

[0004] (1) Lack of vibration monitoring: The main function of the flight attachment casing is to transmit mechanical rotational torque through the mechanical transmission mechanism (including gears, shafts and bearings). Vibration is an important indicator reflecting the health status of the mechanical transmission mechanism. However, there is a lack of vibration monitoring of the flight attachment casing for current fixed-wing aircraft.

[0005] (2) For the monitored working status parameters, manual evaluation is performed separately, which relies too much on the experience of experts and engineers and cannot effectively extract the coupling relationship between the status parameters.

[0006] (3) Without a digital model, it is impossible to conduct a comprehensive analysis of health status. Summary of the invention

[0007] In view of the problems existing in the prior art, the present invention provides a vibration-based multi-dimensional flight attachment casing health assessment method. The present invention establishes a multi-dimensional health assessment model including vibration state indicators, flight state, and working state parameters by collecting vibration signals, flight state information, and working state parameters of the flight attachment casing. The model parameters are optimized and the health threshold is determined through the application data accumulated during the flight and the relevant laboratory test data. Through the health assessment model, quantifiable health assessment and trend analysis of the flight attachment casing can be performed.

[0008] In order to achieve the above object, the technical solution adopted by the present invention is:

[0009] A vibration-based multi-dimensional flight attachment casing health assessment method first determines the vibration monitoring status indicators, vibration monitoring channels, and flight status; secondly, a health assessment model is established, and finally, the model parameters are optimized by applying data to conduct subsequent health assessment. It includes the following steps (see the attached Figure 1 ):

[0010] (I) Determine the vibration monitoring status indicators

[0011] The internal components of the flight attachment casing include gears, shafts and bearings. Various faults will occur during their normal operation due to the alternating loads. The monitoring faults of gears are mainly gear imbalance, misalignment, wear and cracks. For shafts, the faults are mainly rotor imbalance and misalignment. The bearing faults are mostly concentrated in the inner ring, outer ring, rolling element and cage. The above faults can be analyzed and confirmed by analyzing high-frequency and low-frequency vibration signals. By monitoring and analyzing the high-frequency vibration signals of the flight attachment casing, status indicators that can reflect the health status of internal components can be extracted. At the same time, the flight attachment casing as a whole is affected by the aircraft's aerodynamic loads, landing impact, etc., and low-frequency vibration signals will also be generated. It is also necessary to monitor the low-frequency vibration signals of the flight attachment casing to ensure that the extracted health status indicators are more accurate.

[0012] In the present invention, according to the structural characteristics of the flying attachment casing, after theoretical analysis of the main failure modes and failure mechanisms of the internal components (gears, shafts and bearings) of the flying attachment casing, the following vibration monitoring status indicators are selected to perform health assessment of the flying attachment casing:

[0013] (1) Gear vibration monitoring status indicators and their corresponding failure modes

[0014] ●Root mean square (RMS): monitors the overall energy of the signal;

[0015] ●1st order rotation frequency amplitude (gear OM1): monitors gear imbalance fault;

[0016] ●2nd order rotation frequency amplitude (gear OM2): monitors gear misalignment faults;

[0017] ●1st order meshing frequency amplitude (OMX): monitors wear distributed faults;

[0018] ●2nd order meshing frequency amplitude (OM2X): monitors wear distributed faults;

[0019] ● Band Energy Index (MODx): monitors gear hub cracks and misalignment faults.

[0020] (2) Shaft vibration monitoring status indicators and their corresponding failure modes

[0021] ●1st order rotation frequency amplitude (axis OM1): monitors the unbalance fault of the shaft;

[0022] ●2nd order rotation frequency amplitude (axis OM2): monitors shaft misalignment, bending and other faults.

[0023] (3) Bearing monitoring indicators and their corresponding failure modes

[0024] ● Outer ring fault characteristic frequency amplitude (FOP): monitor bearing outer ring fault;

[0025] ● Inner ring fault characteristic frequency amplitude (FIP): monitor bearing inner ring fault;

[0026] ● Rolling element fault characteristic frequency amplitude (FBC): monitor bearing rolling element fault;

[0027] ●Frequency amplitude of cage fault characteristics (FOC): monitor bearing cage fault.

[0028] (II) Determine the vibration monitoring channel

[0029] According to the structural characteristics of the flight attachment casing and the health monitoring requirements, the vibration sensors and the number of their channels used for health monitoring are determined.

[0030] In order to monitor the high-frequency vibration signals generated by the internal components (gears, shafts and bearings) of the flight attachment casing, a single-axis high-frequency monitoring sensor is installed under the flight attachment casing (occupying one vibration monitoring channel); in order to monitor the low-frequency signals generated by the overall vibration of the flight attachment casing, a three-axis low-frequency monitoring sensor is installed behind the flight attachment casing (occupying three vibration monitoring channels).

[0031] (III) Determine flight status

[0032] In different flight states, the speed and power of the flight attachment casing and its internal components are different, the stress loads they are subjected to are also different, and the status indicators reflecting their health status are also quite different. Therefore, the health assessment of the flight attachment casing using working status parameters needs to be combined with the actual flight status. According to the working principle and mission requirements of fixed-wing aircraft, the engine high-pressure rotor speed S is selected. n2 , Ground speed S g The wheel load signal W is used as the dividing standard to divide the three flight states: starting, taxiing, and flying. The vibration data in these three states are relatively stable and have obvious characteristics. According to the design requirements of the aircraft and the analysis of application data, the specific definitions of the three main flight states are as follows:

[0033] ●Engine start: W = ground, S n2 >5%, S g <10km / ;

[0034] ●Ground sliding: W = ground, S n2 >5%, S g ≥10km / ;

[0035] ●Flying: W=ground.

[0036] 4. Establishing a health assessment model

[0037] There are many factors that affect the health status of the flight attachment casing, which are mainly reflected by the status indicators of each component and each channel under different flight conditions. The present invention integrates various monitoring parameters to define a comprehensive health indicator (HI) of the flight attachment casing to characterize the health status of the flight attachment casing.

[0038] The comprehensive health status index is obtained by weighted average of normalized data of each status index of each monitoring channel under different flight conditions. The relationship between HI and each input parameter is specifically expressed as follows:

[0039]

[0040] Where X = [X RMS ,X OM1 ,X FOC ] T Represents each state index, K=[k RMS ,k OM1 ,…,k FOC ] T Represents the coefficient of the indicator, and N represents the total number of state indicators. R represents different flight states, R=1 represents the engine start state, R=2 represents the ground taxiing state, and R=3 represents the flight state. C represents the number of channels for monitoring vibration signals, C=1 is the high-frequency channel, and C=2,3,4 are low-frequency channels.

[0041] 5. Determination and optimization of health assessment model parameters

[0042] The parameters of the health assessment model are determined based on the degree of influence of the vibration-based state indicators corresponding to each channel under a certain flight state on the overall health state of the flight attachment casing. After the initial value is calculated and fitted, it is initially corrected based on the experience of experts. As the actual flight data of each aircraft is accumulated, the newly collected working state parameters are added to optimize and update the coefficients.

[0043] In the actual application process, the specific measured data X of each channel and different status indicators under different flight states need to be manually scored according to the actual health status of the flight attachment casing. The scoring range is (0,1), corresponding to the full state of the flight attachment casing from failure to health. Based on the continuous collection of new actual data, the least squares method is used to solve the overdetermined equations and optimize the coefficient K of the health assessment model. * . See the attached for the specific process. Figure 2 .

[0044] The beneficial effects of the present invention are:

[0045] (1) Compared with the traditional flight attachment casing health monitoring function, the present invention proposes a vibration-based multi-dimensional flight attachment casing health assessment model, which makes the flight attachment casing health detection more complete.

[0046] (2) The technical solution proposed in the present invention classifies, analyzes and processes the vibration data according to different flight states, which is more representative of the health status of the flight attachment casing and effectively improves the accuracy of health monitoring.

[0047] (3) The health index (HI) proposed in the present invention integrates all factors that can reflect the health status of the flight attachment casing to form a highly integrated comprehensive index, which can perform trend analysis and fault diagnosis simply and intuitively, thereby improving the practicability of flight attachment casing health monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Design a methodological process for a vibration-based multi-dimensional flight attachment casing health assessment model;

[0049] Figure 2 Flowchart for model parameter determination and optimization. DETAILED DESCRIPTION

[0050] The present invention is further described below in conjunction with specific embodiments.

[0051] The method described in the present invention is applied in a certain type of aircraft flight casing health assessment system as follows:

[0052] The first step is to determine the vibration monitoring status indicators:

[0053] The flight attachment casing includes 10 gears, 8 shafts, and 16 bearings. According to the first step above, a total of 140 status indicators are defined, including:

[0054] ●Gear status indicators: RMS, OM1, OM2, OMX, OM2X, MODx of 10 gears, a total of 60 status indicators;

[0055] ●Axis status indicators: OM1 and OM2 of 8 axes record a total of 16 status indicators;

[0056] ●Bearing status indicators: FOP, FIP, FBC, FOC of 16 bearings, a total of 64 status indicators.

[0057] The second step is to determine the vibration monitoring channel:

[0058] According to the specific installation situation of the vibration sensor, determine that 1 channel is for high-frequency monitoring and 3 channels (X, Y, Z) are for low-frequency monitoring.

[0059] The third step is to determine the flight status of the aircraft attachment based on the three selected indication signals, including three states: ground, taxiing and flying.

[0060] The fourth step is to establish the health assessment model of the flight attachment casing based on the above parameters as follows:

[0061]

[0062] The fifth step is to determine and optimize the model coefficient K based on actual flight data. The method for determining and optimizing the weighted coefficient K is as follows:

[0063] (1) The vibration data collected in various states and channels are calculated to obtain a normalized state index X. The comprehensive health index HI of the flight attachment casing is scored according to the actual health state of the flight attachment casing corresponding to the data.

[0064] (2) Collect flight data from 20 flights And the corresponding comprehensive health indicators for each flight Substituting into the above model formula, we get the overdetermined equations

[0065] (3) The least squares method is used to solve the constructed overdetermined equations of health indicators and obtain the weighted coefficients of the model.

[0066] (4) After every 20 flights, the model coefficient K is calculated again based on the flight data of the first 20 flights, totaling 40 flights. * , to achieve continuous iterative optimization and update of the model coefficients until the coefficients are stable and no longer change.

[0067] The above-described embodiments merely express the implementation methods of the present invention, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A vibration-based multi-dimensional aircraft accessory casing health assessment method, which can perform quantifiable health assessment and trend analysis on the aircraft accessory casing through a health assessment model, characterized in that: The following steps are involved: (I) Determine the vibration monitoring status indicators According to the structural characteristics of the flight attachment casing, after theoretical analysis of the failure modes and failure mechanisms of the internal components of the flight attachment casing, the following vibration monitoring status indicators are selected for health assessment of the flight attachment casing: (1) Gear vibration monitoring status indicators and their corresponding failure modes Root mean square (RMS): monitors the overall energy of the signal; 1st order rotation frequency amplitude (gear OM1): monitors gear imbalance fault; ·2nd order rotation frequency amplitude (gear OM2): monitors gear misalignment faults; 1st order meshing frequency amplitude (OMX): monitors wear distributed faults; ·2nd order meshing frequency amplitude (OM2X): monitors wear distributed faults; Band Energy Index (MODx): monitors gear hub cracks and misalignment faults; (2) Shaft vibration monitoring status indicators and their corresponding failure modes 1st order rotation frequency amplitude (axis OM1): monitors the unbalance fault of the shaft; ·2nd order rotation frequency amplitude (axis OM2): monitors shaft misalignment and bending faults; (3) Bearing monitoring indicators and their corresponding failure modes Outer ring fault characteristic frequency amplitude (FOP): monitor bearing outer ring fault; Inner ring fault characteristic frequency amplitude (FIP): monitor bearing inner ring fault; Rolling element fault characteristic frequency amplitude (FBC): monitor bearing rolling element fault; Cage fault characteristic frequency amplitude (FOC): monitor bearing cage fault; (II) Determine the vibration monitoring channel According to the structural characteristics and health monitoring requirements of the flight attachment casing, determine the vibration sensors used for health monitoring and the number of their channels; In order to monitor the high-frequency vibration signals generated by the internal components of the flight attachment casing, including gears, shafts and bearings, a single-axis high-frequency monitoring sensor is installed under the flight attachment casing, which occupies one vibration monitoring channel; in order to monitor the low-frequency signals generated by the overall vibration of the flight attachment casing, a three-axis low-frequency monitoring sensor is installed behind the flight attachment casing, which occupies three vibration monitoring channels; (III) Determine flight status Combined with the actual flight status, the working status parameters are used to evaluate the health of the flight attachment casing; according to the working principle and mission requirements of the fixed-wing aircraft, the engine high-pressure rotor speed S is selected. n2 , ground speed S g The wheel load signal W is used as the dividing standard to divide the three flight states: starting, taxiing, and flying. According to the design requirements of the aircraft and the analysis of application data, the specific definitions of the three flight states are as follows: Engine start: W = ground, S n2 >5%, S g <10km / h Ground sliding: W = ground, S n2 >5%, S g ≥10km / h Flight: W = ground; 4. Establishing a health assessment model Based on various monitoring parameters, a comprehensive health index HI of the flight attachment casing is defined to represent the health status of the flight attachment casing; The comprehensive health status index is obtained by weighted average of normalized data of each status index of each monitoring channel under different flight states; the relationship between HI and each input parameter is specifically expressed as follows: Where X = [X RMS ,X OM1 ,X FOC ] T Represents each state index, K=[k RMS ,k OM1 ,…,k FOC ] T Represents the coefficient of the index, N represents the total number of state indicators; R represents different flight states, R = 1 represents the engine start state, R = 2 represents the ground taxiing state, and R = 3 represents the flight state; C represents the number of channels for monitoring vibration signals, C = 1 is the high-frequency channel, and C = 2, 3, 4 are low-frequency channels; 5. Determination and optimization of health assessment model parameters The influence of the vibration-based status indicators corresponding to each channel under a certain flight state condition on the overall health state of the flight attachment casing is determined; the initial value is calculated and fitted, and then a preliminary correction is performed. With the accumulation of actual flight data of each aircraft, the newly collected working state parameters are added to optimize and update the coefficients.

Images