A method for monitoring the status and diagnosing faults of moving parts of an unmanned helicopter

Abstract

The present invention provides a method for monitoring the status and diagnosing faults of the moving components of an unmanned helicopter, comprising the following steps: Step 1, selecting measurement points: sensors are installed at key locations on the unmanned helicopter fuselage. The sensors pick up signals from rotating components and electrical equipment in the transmission system, power unit, rotor system, and tail rotor system, and transmit analog signals to an onboard data acquisition device via dedicated cables; Step 2, data acquisition; Step 3, data processing; Step 4, data transmission; Step 5, developing a diagnostic plan; Step 6, outputting diagnostic results; and Step 7, comprehensive data analysis. This method is suitable for monitoring the status and diagnosing faults of the moving components of unmanned helicopters, with clear fault mechanisms and a simple and reliable implementation.

Description

Technical Field

[0001] The invention relates to a method for monitoring the state of moving parts of an unmanned helicopter and diagnosing faults. Background Art

[0002] Unmanned helicopters are capable of long-term hovering, vertical takeoff and landing, low-speed flight, maneuverability, and the ability to perform missions in complex environments with zero casualties in the event of a crash. They have low requirements for landing and takeoff sites and can operate on unpaved surfaces or rooftops of urban buildings. Their flexibility makes helicopters irreplaceable in the aviation industry. This ability to perform missions in complex environments also leads to a high incidence of unmanned helicopter-related failures, necessitating a state detection and fault diagnosis method tailored to these needs.

[0003] The HUMS (Health and Usage Monitoring System), which is currently widely used on manned helicopters, also collects information such as rotation speed, temperature, and vibration, but does not have the ability to transmit data back to the ground station in real time. It can only display the results in the cockpit and does not have an early warning of the corresponding component failure type. It can only judge subsystem failures based on the overall characteristics of the signal.

[0004] Therefore, it is of great practical significance to design a method for unmanned helicopter state monitoring and mechanical fault diagnosis that is suitable for unmanned helicopters, based on fault mode analysis, and performs real-time tracking of major moving parts. Summary of the Invention

[0005] Purpose of the invention: The technical problem to be solved by the present invention is to address the deficiencies of the existing technology and provide a method for monitoring the status and diagnosing faults of the moving parts of an unmanned helicopter, including: measurement point selection, data acquisition, data processing, data transmission, formulation of a diagnostic plan, output of diagnostic results, and comprehensive data analysis.

[0006] The present invention comprises the following steps:

[0007] Step 1, measurement point selection: Install sensors at key locations on the unmanned helicopter fuselage. The sensors pick up signals from rotating components in the transmission system, power unit, rotor system, tail rotor system, and electrical equipment, and transmit analog signals to the onboard data acquisition equipment via dedicated cables.

[0008] Step 2, data collection;

[0009] Step 3, data processing;

[0010] Step 4, data transmission;

[0011] Step 5: Develop a diagnostic plan;

[0012] Step 6: Output the diagnosis result;

[0013] Step 7: Comprehensive data analysis.

[0014] In step 1, the power transmission mode of the unmanned helicopter system is divided into two paths: one is the engine, belt, main reducer, and rotor, and the other is the engine, belt, tail drive shaft, tail reducer-tail rotor. The objects to be monitored and the types of faults expected to occur in the monitored objects are determined. The fault types include rotor failure, tail rotor failure, gear failure, bearing failure, belt failure, drive shaft failure, and engine failure.

[0015] According to the signal transmission route, the measurement point locations were selected based on the principle of shortest path and maximum stiffness. The top output bearing seat of the main gearbox, the bottom support bearing seat of the main gearbox, the power input bearing seat of the main gearbox, the output bearing seat of the tail gearbox, and the output bearing seat of the engine output gearbox were selected as the key locations of the unmanned helicopter.

[0016] Depending on the fault identification method (see step 5 for the identification method), speed sensors are installed at the support positions of the main gearbox output shaft, tail drive shaft, and tail rotor main shaft.

[0017] For example, to determine rotor dynamic balance failure, the two parameters of vibration amplitude and vibration phase must be used, and a vibration acceleration sensor and a speed sensor must be installed at the output shaft of the gearbox supporting the rotor shaft; to determine gear and bearing failures, an acceleration sensor with a large enough dynamic range must be selected, covering the range from 2Hz to 6 times the gear meshing frequency.

[0018] Step 2 includes: conducting a mechanical failure mechanism analysis based on the types of failures that may occur in each system on the power transmission path of the two unmanned helicopter systems, and based on the analysis results, determining the frequency range that needs to be analyzed at the rated operating speed, and then calculating the sampling frequency required for each key position, specifically including: the six channels of the output bearing seat and the power input bearing seat at the top of the main gearbox need to monitor the main gear meshing frequency information, the three channels of the support bearing seat measuring point at the bottom of the main gearbox only need to monitor the information of the support bearing, the three channels of the tail gearbox output bearing seat need to monitor the tail gear meshing frequency information, and the frequency range that electrical equipment and mission equipment need to pay attention to is only up to four times the frequency of the engine crankshaft speed.

[0019] Step 3 includes: calculating the time domain characteristic values ​​of each channel, including the effective value, peak factor, skewness index, and kurtosis index. The specific algorithm is shown in Chapter 4 "Engineering Signal Analysis Technology" of "Fault Mechanism and Diagnosis Technology of Rotating Machinery";

[0020] Process the corresponding fault information according to the location of the measuring point, including:

[0021] Processing of rotor dynamic balance state data: Select the vibration acceleration data of the horizontal left and right channels of the output bearing seat at the top of the main gearbox, perform frequency domain digital filtering based on the speed information of the main gearbox output shaft, and only retain the signals of the 1 / 4 speed range on both sides of the speed frequency. Then calculate the span of the speed peak interval, determine the vibration acceleration peak position, and then calculate the dynamic balance phase based on the relationship between the peak position and the speed pulse. The current peak value is recorded as the amplitude data, and the mean and variance DIA of the phase PHA and amplitude AMP of the current group of data are recorded.

[0022] In step 3, the vibration data of the horizontal left and right channels of the output bearing seat at the top of the main gearbox are set as the X sequence x1, x2, x3, ..., xn, where xn represents n vibration data, and n is a natural number. According to the sampling frequency and the requirements of 20kHz and frequency domain resolution, the number of calculation points is 2 16 , that is, every 2 16 The data is collected as a group, and the current group of data is recorded as the current group of data. The calculation method includes the following steps:

[0023] Step 3-1, calculate the spectrum data using the radix 2-FFT algorithm (for the algorithm, refer to the Fast Fourier Transform (FFT) in Section 3 of Chapter 3 "Digital Signal Processing" of "Fault Mechanism and Diagnosis Technology of Rotating Machinery");

[0024] Step 3-2: Based on the speed sensor information, select the frequency domain data of the 1 / 4 speed range on both sides of the speed frequency, set the other frequency domain data to zero, and perform inverse FFT analysis (also using the base-2 algorithm with a rotation factor of -1) to obtain the vibration data Y_ at the rotation frequency;

[0025] Step 3-3: Starting from the first rising edge of the speed sensor pulse signal, search for the relative time position of the vibration peak. Convert the two rising edges of the pulse signal to 360 degrees and calculate the relative position of the corresponding peak value for each circle of data. This relative position is also given in degrees (°). The relative position is called the phase and the peak value is called the amplitude. Record the peak value and phase as AMP1 and PHA1 respectively.

[0026] Step 3-4: Calculate the amplitude and phase of all cycles of the current group vibration data according to the method of step 3-3 to obtain the data sequence AMP1, APM2, AMP3, ..., AMP n and PHA1, PHA2, PHA3, ... PHA n ; (The subscript of the symbol indicates the number of cycles, and n indicates the maximum number of complete cycles contained in the current set of vibration data);

[0027] Step 3-5: Calculate the amplitude mean MEN_AMP, phase mean MEN_PHA, amplitude variance DIA_AMP, and phase variance DIA_PHA of the two data sequences according to the formula for calculating variance in the university textbook "Probability Theory and Mathematical Statistics";

[0028] Step 3 also includes: processing the dynamic balance data of the tail drive shaft and tail rotor: calculating the dynamic balance of the tail drive shaft by selecting the vertical channel acceleration data of the main gearbox power input bearing seat and the tail gearbox output bearing seat; calculating the dynamic balance of the tail rotor by selecting the vertical channel acceleration data of the tail gearbox output bearing seat. For the specific algorithm, see steps 3-1 to 3-5;

[0029] Shaft bending state data processing: Select the axial channel acceleration data from the sensors at both ends of the shaft, combine it with the shaft speed information, perform frequency domain digital filtering, retain only the signal near the speed frequency, calculate the span of the speed peak range, determine the vibration acceleration peak position, calculate the vibration phase and amplitude data based on the relationship between the peak position and the speed pulse, calculate the mean and variance of the vibration phase and amplitude of the current group of data, and perform real-time analysis of the amplitude AMP and phase difference PHA at both ends of the shaft. For the specific algorithm, see steps 3-1 to 3-5;

[0030] Belt wear status data processing: Select measurement points at both ends of the belt, and select two channels parallel to the belt. For example, if the belt is arranged vertically, select the upper and lower directions of the measurement points at both ends. Use the base 2-FFT calculation to obtain the amplitude of a series of harmonics of the belt frequency, and use the amplitude of the 2nd harmonic as the parameter indicating belt wear;

[0031] Gear status data processing includes the following steps:

[0032] In step a1, time-domain synchronous averaging, the number of sampling points must be set based on the number of teeth on each gear. The digital resampling frequency Fs is set to a base-2 number greater than six times the number of teeth per revolution. For example, if the gear fixed to the main shaft has 17 teeth and the operating frequency of the gear shaft is Ω, and Ω is set to 6 Hz, then the digital resampling frequency Fs is:

[0033] 6×17×Ω=612

[0034] Due to 2 9 =512,2 10 =1024, then Fs=1024.

[0035] Step a2: Select the 8-week time domain average data at the latest moment and perform radix 2-FFT calculation;

[0036] Step a3: Perform a refined spectrum analysis. The center frequency is selected as the 1st, 2nd, and 3rd harmonics of the gear meshing frequency. The refinement multiple is reasonably designed based on the analysis bandwidth and the rotational frequency of each shaft: the bandwidth is required to be greater than 12 times the rotational frequency of the shaft where the gear is located, and the frequency resolution is less than 0.1 Hz.

[0037] For example, a transmission shaft in a gearbox rotates at a frequency of 10 Hz, and the number of teeth on the fixed gear is 23. The meshing frequency of this pair of teeth is 230 Hz. In this case, 230 Hz, 460 Hz, and 690 Hz are selected as center frequencies, and the bandwidth is greater than 12 × 10 = 120 Hz. For the specific algorithm, see "Research on the Algorithm of Complex Modulation Refinement Spectrum Analysis Based on Complex Analysis Bandpass Filter".

[0038] Step a4: For each sideband of the 1st, 2nd, and 3rd harmonic frequency refined spectra of the gear meshing, extract the amplitude of each sideband respectively, and convert the amplitude to dimensionless according to the peak value of each harmonic frequency of the gear meshing, and record the dimensionless data of each gear meshing frequency and sideband:

[0039] Bearing status data processing includes the following steps:

[0040] Step b1, calculate the fault frequency according to the formulas given in various reference books, including the inner ring fault frequency, outer ring fault frequency, cage fault frequency and rolling element fault frequency;

[0041] Step b2: Perform FFT calculation on the original points with the same number of points as the time domain synchronous averaging points, subtract the obtained data column from the FFT data of the time domain synchronous averaging of the same measuring point in the aforementioned gear fault analysis, perform inverse FFT transformation, and then calculate the time domain eigenvalue;

[0042] Step b3, recording the peak values ​​of each fault frequency, the dimensionless amplitudes of the shaft frequency sidebands on both sides of the inner and outer ring fault frequencies, and the dimensionless amplitudes of the cage sidebands on both sides of the 1st and 2nd times of the rolling element fault frequency;

[0043] Step b4, recording the amplitudes of the 1st, 2nd, and 3rd times of the shaft frequency in the FFT spectrum of the bearing axial vibration signal;

[0044] Step b5, recording the amplitudes of the 1st, 2nd, ..., 10th times of the shaft frequency in the FFT spectrum of the radial vibration signal of the measuring point near the bearing;

[0045] Step b6: remove the peak information of the frequency corresponding to each exciting force in the entire spectrum, and calculate the energy of the remaining spectrum, which is recorded as the background noise energy.

[0046] Step 4 includes: transmitting the characteristic values ​​obtained in step 3 to the ground station for reception.

[0047] Step 5 includes: the diagnostic scheme is based on the parameters obtained from the failure mechanism analysis of the rotor, tail rotor, tail drive shaft, gears and bearings, and is used for the actual failure analysis of the unmanned helicopter;

[0048] The primary source of rotor system imbalance is the rotor, with a frequency equal to the rotor's operating speed. Under stable operating conditions, the vibration source is also stable, resulting in stable amplitude and phase. According to national standards, the amplitude can be directly determined, while phase deviations are determined through on-site testing analysis.

[0049] The same imbalance mechanism applies to the tail rotor and tail drive shaft.

[0050] Gear failure is also based on the analysis of the failure mechanism. The different types of gear failure are tested on the test bench with the same transmission power, and the rules of various parameters are summarized to guide the actual fault diagnosis.

[0051] Bearing fault diagnosis is based on the bearing failure frequency mechanism, combined with the actual use of the unmanned helicopter, to analyze the manifestation of each fault, and to determine the diagnosis plan by implanting the fault on the test bench.

[0052] Step 6 includes: the ground station receives data from the link, the ground station display and control software determines the characteristic value data of each field according to the transmission protocol, and then matches the characteristic values ​​of each fault mode according to the diagnosis plan to give a diagnosis result. If the domain characteristic value and each fault information do not exceed the threshold, the status is good. If it exceeds the threshold, a fault code is given according to the diagnosis plan.

[0053] Step 7 includes: after the unmanned helicopter completes its flight and returns to the ground, the flight data of the unmanned helicopter is downloaded to the data comprehensive analysis system via a dedicated data cable, data analysis is performed on the flight conditions, trend tracking of each characteristic value over the flight time is established, and fault prediction is given based on the life curve determined by the test.

[0054] The beneficial effects of the present invention are as follows: This method selects measurement points on the fuselage of the unmanned helicopter according to the possible faults of each system of the unmanned helicopter, and then collects signals such as vibration and rotation speed according to a specific data acquisition method, discretizes them, and obtains a series of characteristic values ​​according to the data processing method, which are transmitted to the ground station through the limited bandwidth of the link. The ground station display and control software matches each fault mode (or combination). If each characteristic value (combination) reaches the warning condition of each fault mode, the display and control software will give an alarm and display a specific fault code, thereby realizing real-time monitoring and tracking of various faults. After the unmanned helicopter flight is completed, the original data is downloaded to the data comprehensive analysis system for multi-condition analysis, and trend tracking is given to perform life prediction and fault prediction. This method is suitable for status monitoring and fault diagnosis of moving parts of unmanned helicopters. The fault mechanism is clear and the implementation method is simple and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more apparent.

[0056] Figure 1 Flowchart of the present invention.

[0057] Figure 2 This is a system diagram that needs to be considered in the measurement point selection method of the present invention.

[0058] Figure 3 Schematic diagram of the data collection method of the present invention.

[0059] Figure 4 This is a data processing flow chart of the present invention. DETAILED DESCRIPTION

[0060] The present invention provides a method for monitoring the status of moving parts of an unmanned helicopter and diagnosing faults, including: selecting measuring points, collecting data, processing data, transmitting data, formulating a diagnosis plan, outputting diagnosis results, and conducting comprehensive data analysis. Figure 1 This is the principle and data transmission path of this method.

[0061] For selection of the measuring points, see Figure 2 , including the following steps:

[0062] Based on the unmanned helicopter system's power transmission method, the objects that need to be monitored and the types of faults expected to occur in these objects are determined, including rotor failure, tail rotor failure, gear failure, bearing failure, belt failure, drive shaft failure, and engine failure. Based on the signal transmission route, measurement point locations are selected based on the principles of shortest path and maximum stiffness. The top output bearing seat of the main gearbox, the bottom support bearing seat of the main gearbox, the power input bearing seat of the main gearbox, the tail gearbox output bearing seat, and the engine output gearbox output bearing seat are selected as key measurement points for the unmanned helicopter. Additionally, as needed, the installation locations of electrical equipment and mission equipment are also selected as measurement points.

[0063] Depending on the fault identification method, speed sensors are installed at the supporting positions of the main gearbox output shaft, tail drive shaft, and tail rotor main shaft.

[0064] The data collection method is described in Figure 3 ,According to the fault types that may occur in each system, the mechanical fault mechanism is analyzed, and based on the analysis results, the frequency range that needs to be analyzed at the rated working speed is determined, and then the sampling frequency required for each measuring point is calculated.

[0065] For example, the six channels of the output bearing seat and the power input bearing seat at the top of the main gearbox need to monitor the main gear meshing frequency information, the three channels of the support bearing seat measuring point at the bottom of the main gearbox only need to monitor the information of the support bearing, and the three channels of the output bearing seat of the tail gearbox need to monitor the tail gear meshing frequency information. The frequency range that electrical equipment and mission equipment need to pay attention to is only up to four times the frequency of the engine crankshaft speed, and the sampling frequency should be set accordingly.

[0066] The data processing method, data transmission see Figure 4 , including the following:

[0067] First, calculate the time domain characteristic values ​​of each channel, including: effective value, peak coefficient, skewness index, kurtosis index

[0068] Then, the corresponding fault information is processed according to the location of the measuring point, for example:

[0069] The rotor dynamic balance data processing method selects the vibration acceleration data from the horizontal left and right channels of the main gearbox's top output bearing seat. Based on the main gearbox output shaft speed information, frequency-domain digital filtering is performed, retaining only the signals within the 1 / 4 speed range on both sides of the speed frequency. The span of the speed peak interval is then calculated to determine the peak position of the vibration acceleration. The dynamic balance phase is then calculated based on the relationship between the peak position and the speed pulse, and the current peak value is recorded as the amplitude data. The mean and variance DIA of the phase PHA and amplitude AMP of the current set of data are recorded.

[0070] The calculation of the dynamic balance state of the tail drive shaft and tail rotor is similar to that of the rotor. The difference is that the vertical channel acceleration data of the main gearbox power input bearing seat and the tail gearbox output bearing seat are selected to calculate the dynamic balance of the tail drive shaft, and the vertical channel acceleration data of the tail gearbox output bearing seat are selected to calculate the dynamic balance of the tail rotor.

[0071] The method for processing shaft bending state data involves selecting axial channel acceleration data from sensors at both ends of the shaft. This data is then combined with the shaft's speed information and subjected to frequency-domain digital filtering, retaining only signals near the speed frequency. The span of the speed peak interval is then calculated to determine the peak location of the vibration acceleration. The vibration phase and amplitude data are then calculated based on the relationship between the peak location and the speed pulse. The mean and variance of the phase and amplitude values ​​for the current set of data are then calculated. The amplitude AMP and phase difference PHA at both ends of the shaft are then analyzed in real time.

[0072] Belt wear status data processing method: Select measurement points at both ends of the belt and two channels parallel to the belt. Calculate the amplitude of a series of belt frequency harmonics through FFT calculations. Use the amplitude of the 2nd harmonic as the parameter indicating belt wear.

[0073] Gear status data processing method:

[0074] Step a1, time-domain synchronous averaging, uses an algorithm as described in the literature "Vibration Signal Processing of Helicopter Transmission Systems Based on TSA." However, the number of sampling points must be considered in the number of teeth on each shaft, and the digital resampling frequency must be set to a base-2 number greater than six times the number of teeth per cycle.

[0075] Step a2: Select the 8-week time domain average data at the latest moment for FFT analysis;

[0076] Step a3: Perform a refined spectrum analysis. The center frequency is selected as the 1st, 2nd, and 3rd harmonics of the gear meshing frequency. The refinement multiple is reasonably designed based on the analysis bandwidth and the rotational frequency of each shaft. The bandwidth is required to be greater than 12 times the rotational frequency of the shaft where the gear is located, and the frequency resolution is less than 0.1 Hz.

[0077] Step a4: For each sideband of the 1st, 2nd, and 3rd harmonic spectrum of the gear meshing, extract the amplitude of each sideband respectively, and convert the amplitude to dimensionless value for each harmonic peak of the gear meshing, and record the dimensionless data of each gear meshing frequency and sideband:

[0078] Bearing status data processing method:

[0079] Step b1: first, calculate the fault frequency according to the public bearing fault frequency calculation formula, including the inner ring fault frequency, outer ring fault frequency, cage fault frequency and rolling element fault frequency;

[0080] Step b2: Perform FFT analysis on the original points with the same number of points as the time-domain synchronous averaging points, take the modulus, and subtract the data column from the FFT data of the time-domain synchronous averaging. Perform inverse FFT transform, and then calculate the time-domain eigenvalues.

[0081] Step b3, record the peak values ​​of each fault frequency and the dimensionless amplitudes of the shaft frequency sidebands on both sides of the inner and outer ring fault frequencies, as well as the dimensionless amplitudes of the cage sidebands on both sides of the 1st and 2nd times of the rolling element fault frequency.

[0082] Step b4, recording the amplitudes of the 1st, 2nd and 3rd times of the shaft frequency in the FFT spectrum of the bearing axial vibration signal.

[0083] Step b5: record the amplitudes of the 1st, 2nd, ..., 10th times of the axis frequency in the FFT spectrum of the radial vibration signal of the measuring point near the bearing.

[0084] Step b6: remove the peak information of the frequency corresponding to each exciting force in the entire spectrum, and calculate the energy of the remaining spectrum, which is recorded as the background noise energy.

[0085] The data transmission method described above cannot transmit the original data in real time due to the limited downlink bandwidth of the link. Therefore, the method is designed to transmit the processed eigenvalues ​​of the data, including time domain eigenvalues, frequency domain eigenvalues, etc., to be received by the ground station.

[0086] The diagnosis scheme is formulated based on the failure mode of each component, as shown in Table 1:

[0087] Table 1

[0088]

[0089]

[0090]

[0091]

[0092] The output of the diagnosis result is specifically:

[0093] The ground station receives data from the link. The ground station display and control software determines the characteristic value data of each field according to the transmission protocol, and then matches the characteristic values ​​of each fault mode according to the diagnosis plan to give a diagnosis result. When the domain characteristic value and each fault information do not exceed the threshold, the status is good. If the characteristic value (or combination) exceeds the threshold, a fault code is given according to the diagnosis plan.

[0094] The comprehensive analysis of the data is as follows:

[0095] After the unmanned helicopter completes its flight and returns to the ground, the flight data is downloaded to the data comprehensive analysis system via a dedicated data cable. Data analysis of the flight conditions is performed, and trend tracking of each characteristic value over flight time is established. Based on the life curve determined by the test, a fault prediction is given.

[0096] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium is capable of storing a computer program that, when executed by the data processing unit, executes the invention content of the method for monitoring the status and fault diagnosis of the dynamic components of an unmanned helicopter provided by the present invention and some or all of the steps in each embodiment. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0097] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, in essence or in other words, the part that contributes to the prior art, can be embodied in the form of a computer program, i.e., a software product. The computer program software product can be stored in a storage medium and includes several instructions for enabling a device including a data processing unit (which can be a personal computer, a server, a single-chip microcomputer, a MUU, or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0098] The present invention provides a method for monitoring the status and diagnosing faults of the moving components of an unmanned helicopter. While many methods and approaches exist for implementing this technical solution, the above-described preferred embodiments of the present invention are merely exemplary. It should be noted that those skilled in the art may make improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Any components not specified in this embodiment may be implemented using existing technologies.

Claims

1. A method for monitoring the status and diagnosing faults of moving parts of an unmanned helicopter, characterized in that: The following steps are involved: Step 1, measurement point selection: Install sensors at key locations on the unmanned helicopter fuselage. The sensors pick up signals from rotating components in the transmission system, power unit, rotor system, tail rotor system, and electrical equipment, and transmit analog signals to the onboard data acquisition equipment via dedicated cables. Step 2, data collection; Step 3, data processing; Step 4, data transmission; Step 5: Develop a diagnostic plan; Step 6: Output the diagnosis result; Step 7, comprehensive data analysis; In step 1, the power transmission mode of the unmanned helicopter system is divided into two paths: one path is the engine, belt, main reducer, and rotor, and the other path is the engine, belt, tail drive shaft, tail reducer, and tail rotor. The objects to be monitored and the types of faults expected to occur in the monitored objects are determined. The fault types include rotor failure, tail rotor failure, gear failure, bearing failure, belt failure, drive shaft failure, and engine failure. According to the signal transmission route, the measurement point locations were selected based on the principle of shortest path and maximum stiffness. The top output bearing seat of the main gearbox, the bottom support bearing seat of the main gearbox, the power input bearing seat of the main gearbox, the output bearing seat of the tail gearbox, and the output bearing seat of the engine output gearbox were selected as the key locations of the unmanned helicopter. Depending on the fault identification method, speed sensors are installed at the support positions of the main gearbox output shaft, tail drive shaft, and tail rotor main shaft; Step 2 includes: conducting a mechanical failure mechanism analysis based on the types of failures that may occur in each system on the power transmission path of the two unmanned helicopter systems, and based on the analysis results, determining the frequency range that needs to be analyzed at the rated operating speed, and then calculating the sampling frequency required for each key position, specifically including: the six channels of the output bearing seat and the power input bearing seat at the top of the main gearbox need to monitor the main gear meshing frequency information, the three channels of the support bearing seat measuring point at the bottom of the main gearbox only need to monitor the information of the support bearing, the three channels of the tail gearbox output bearing seat need to monitor the tail gear meshing frequency information, and the frequency range that electrical equipment and mission equipment need to pay attention to is only up to four times the frequency of the engine crankshaft speed.

2. The method according to claim 1, characterized in that Step 3 includes: calculating the time domain characteristic values ​​of each channel, including effective value, peak coefficient, skewness index, and kurtosis index; Process the corresponding fault information according to the location of the measuring point, including: Processing of rotor dynamic balance state data: Select the vibration acceleration data of the horizontal left and right channels of the output bearing seat at the top of the main gearbox, perform frequency domain digital filtering according to the speed information of the main gearbox output shaft, and only retain the signals of the 1 / 4 speed range on both sides of the speed frequency. Then calculate the span of the speed peak interval, determine the vibration acceleration peak position, and then calculate the dynamic balance phase based on the relationship between the peak position and the speed pulse, and record the current peak value as the amplitude data, and record the mean and variance DIA of the phase PHA and amplitude AMP of the current group of data.

3. The method according to claim 2, characterized in that In step 3, the vibration data of the horizontal left and right channels of the output bearing seat at the top of the main gearbox are set as the X sequence x1, x2, x3, ..., xn, where xn represents n vibration data, and n is a natural number. According to the sampling frequency and the requirements of 20kHz and frequency domain resolution, the number of calculation points is 2 16 , that is, every 2 16 The data is collected as a group, and the current group of data is recorded as the current group of data. The calculation method includes the following steps: Step 3-1, calculating spectrum data according to the radix 2-FFT algorithm; Step 3-2: Based on the speed sensor information, select the frequency domain data of the 1 / 4 speed range on both sides of the speed frequency, set the other frequency domain data to zero, and perform inverse FFT analysis to obtain the vibration data Y_ at the rotation frequency; Step 3-3: Starting from the first rising edge of the speed sensor pulse signal, search for the relative time position of the vibration peak. Convert the two rising edges of the pulse signal to 360° and calculate the relative position of the corresponding peak value for each circle of data. The relative position is also given in degrees. The relative position is called the phase and the peak value is called the amplitude. The peak value and phase are recorded as AMP1 and PHA1 respectively. Step 3-4: Calculate the amplitude and phase of all cycles of the current group vibration data according to the method of step 3-3 to obtain the data sequence AMP1, APM2, AMP3, ..., AMP n and PHA1, PHA2, PHA3, ... PHA n ; Step 3-5: Calculate the amplitude mean MEN_AMP, phase mean MEN_PHA, amplitude variance DIA_AMP, and phase variance DIA_PHA of the two data sequences.

4. The method according to claim 3, characterized in that Step 3 also includes: processing the dynamic balance data of the tail drive shaft and the tail rotor: calculating the dynamic balance of the tail drive shaft by selecting the vertical channel acceleration data of the main gearbox power input bearing seat and the tail gearbox output bearing seat; calculating the dynamic balance of the tail rotor by selecting the vertical channel acceleration data of the tail gearbox output bearing seat; Shaft bending state data processing: The axial channel acceleration data of the sensors at both ends of the shaft are selected and combined with the shaft speed information to perform frequency domain digital filtering, retaining only the signal near the speed frequency. The span of the speed peak interval is calculated to determine the peak position of the vibration acceleration. The vibration phase and amplitude data are calculated based on the relationship between the peak position and the speed pulse. The mean and variance of the vibration phase and amplitude of the current group of data are calculated, and the amplitude AMP and phase PHA at both ends of the shaft are analyzed in real time. Belt wear status data processing: Select measuring points at both ends of the belt, and select the two channels parallel to the belt. If the belt is arranged in a vertical direction, select the upper and lower directions of the measuring points at both ends. Use the 2-base FFT calculation to obtain the amplitude of a series of harmonics of the belt frequency, and use the amplitude of the 2nd harmonic as the parameter indicating belt wear. Gear status data processing includes the following steps: Step a1, time domain synchronous averaging: the number of sampling points must be set considering the number of teeth on each shaft gear, and the digital resampling frequency Fs is set to a base 2 number greater than 6 times the number of teeth per cycle; Step a2: Select the 8-week time domain average data at the latest moment and perform radix 2-FFT calculation; Step a3: Perform a refined spectrum analysis: The center frequency is selected as the 1st, 2nd, and 3rd harmonics of the gear meshing frequency. The refinement multiple is reasonably designed based on the analysis bandwidth and the rotational frequency of each shaft: the bandwidth is required to be greater than 12 times the rotational frequency of the shaft where the gear is located, and the frequency resolution is less than 0.1 Hz. Step a4: For each sideband of the 1st, 2nd, and 3rd harmonic frequency refined spectra of the gear meshing, extract the amplitude of each sideband respectively, and convert the amplitude to dimensionless according to the peak value of each harmonic frequency of the gear meshing, and record the dimensionless data of each gear meshing frequency and sideband: Bearing status data processing includes the following steps: Step b1, calculate the fault frequency according to the formulas given in various reference books, including the inner ring fault frequency, outer ring fault frequency, cage fault frequency and rolling element fault frequency; Step b2: Perform FFT calculation on the original points with the same number of points as the time domain synchronous averaging points, subtract the obtained data column from the time domain synchronous averaging FFT data of the same measuring point in the aforementioned gear fault analysis, perform inverse FFT transformation, and then calculate the time domain eigenvalue; Step b3, recording the peak values ​​of each fault frequency, the dimensionless amplitudes of the shaft frequency sidebands on both sides of the inner and outer ring fault frequencies, and the dimensionless amplitudes of the cage sidebands on both sides of the 1st and 2nd times of the rolling element fault frequency; Step b4, recording the amplitudes of the 1st, 2nd, and 3rd times of the shaft frequency in the FFT spectrum of the bearing axial vibration signal; Step b5, recording the amplitudes of the 1st, 2nd, ..., 10th times of the shaft frequency in the FFT spectrum of the radial vibration signal of the measuring point near the bearing; Step b6: remove the peak information of the frequency corresponding to each exciting force in the entire spectrum, and calculate the energy of the remaining spectrum, which is recorded as the background noise energy.

5. The method according to claim 4, characterized in that Step 4 includes: transmitting the characteristic values ​​obtained in step 3 to the ground station for reception.

6. The method according to claim 5, characterized in that Step 5 includes: the diagnostic scheme is based on parameters obtained from the failure mechanism analysis of the rotor, tail rotor, tail drive shaft, gears and bearings, and is used for failure analysis of an actual unmanned helicopter.

7. The method according to claim 6, characterized in that Step 6 includes: the ground station receives data from the link, the ground station display and control software determines the characteristic value data of each field according to the transmission protocol, and then matches the characteristic values ​​of each fault mode according to the diagnosis plan to give a diagnosis result. If the domain characteristic value and each fault information do not exceed the threshold, the status is good. If it exceeds the threshold, a fault code is given according to the diagnosis plan.

8. The method according to claim 7, characterized in that Step 7 includes: after the unmanned helicopter completes its flight and returns to the ground, the flight data of the unmanned helicopter is downloaded to the data comprehensive analysis system via a dedicated data cable, data analysis is performed on the flight conditions, trend tracking of each characteristic value over the flight time is established, and fault prediction is given based on the life curve determined by the test.

Images