Hydraulic cylinder internal leakage detection method and system based on transient pressure wave

Abstract

This invention belongs to the field of fluid machinery fault diagnosis technology, and relates to a method and system for detecting internal leakage in hydraulic cylinders based on transient pressure waves. The invention generates a global hardware trigger signal by acquiring the control signal of an electromagnetic directional valve, simultaneously activating a high-frequency dynamic pressure sensor and a displacement sensor to acquire the original water hammer pressure pulse signal and dynamic physical cavity length at the hydraulic cylinder inlet, respectively. The invention combines the physical velocity of sound of the hydraulic oil at the current temperature to calculate the theoretical echo arrival time and sets a hardware-gated time window to accurately extract the target echo signal. By comparing the phase difference and amplitude attenuation characteristics between the target echo and the original signal, the system can accurately identify the acoustic impedance abrupt change mechanism, distinguish between real leakage and false attenuation, and effectively reduce the false alarm rate. This invention achieves real-time online detection without requiring shutdown or disassembly of the equipment, improving detection efficiency and equipment availability, and providing reliable technical support for hydraulic system maintenance.

Description

Technical Field

[0001] This invention belongs to the field of fluid machinery fault diagnosis technology, and relates to a method and system for detecting internal leakage in hydraulic cylinders based on transient pressure waves. Background Technology

[0002] Hydraulic cylinders, as the core actuators of hydraulic systems, are widely used in engineering machinery, metallurgical equipment, and aerospace. Their main function is to convert hydraulic energy into mechanical energy to drive loads in reciprocating linear motion. The sealing performance of hydraulic cylinders is fundamental to their normal operation. Wear or damage to the piston seals can lead to internal leakage between the high-pressure and low-pressure chambers. This fault can cause a drop in system pressure, slow and weak actuator movement, or even failure, and in severe cases, may even lead to safety accidents.

[0003] Currently, the detection methods for internal leaks in hydraulic cylinders mainly rely on offline or indirect measurements. Common offline methods include pressure testing, which involves pressurizing a single chamber of the hydraulic cylinder while the equipment is stopped and measuring its pressure-holding performance. While direct, this method is cumbersome and cannot reflect leaks under dynamic operating conditions. Online detection methods include monitoring changes in hydraulic oil temperature or analyzing the movement speed of the actuator. For example, infrared thermal imaging technology can be used to capture the localized high temperatures generated by leaks, or displacement sensors can be used to monitor speed decay under load. In addition, some technologies use vibration or acoustic emission sensors to pick up the high-frequency signals generated when fluid flows through a gap during a leak.

[0004] However, the aforementioned existing technical solutions have several drawbacks in practical applications. Offline detection methods interrupt the production process, resulting in low equipment utilization and long detection times. Indirect measurement methods based on temperature or speed changes have low sensitivity, making it difficult to detect early, minute leaks, and are highly susceptible to interference from external factors such as load changes, ambient temperature, and oil viscosity fluctuations, leading to unstable detection results. Meanwhile, detection technologies based on vibration or acoustic emission suffer from extremely low signal-to-noise ratios in the high-noise and high-vibration background environments of heavy equipment, where leak characteristic signals are often submerged in environmental noise, making effective extraction difficult and resulting in frequent false positives and false negatives. Summary of the Invention

[0005] In view of this, in order to solve the problems mentioned in the background technology, a method and system for detecting internal leakage in hydraulic cylinders based on transient pressure waves are proposed.

[0006] The objective of this invention can be achieved through the following technical solution: The first aspect of this invention provides a method for detecting internal leakage in a hydraulic cylinder based on transient pressure waves, comprising:

[0007] S1. Collect the control signal of the electromagnetic directional valve in the hydraulic circuit, perform edge detection, extract the falling edge level, and generate a global hardware trigger signal.

[0008] S2. Responding to the global hardware trigger signal, the high-frequency dynamic pressure sensor is activated to collect fluid pressure data at the hydraulic cylinder inlet pipe. By identifying the high-pressure jump edge in the pressure sequence that exceeds the preset differential steepness threshold, the absolute zero point time of the waveform emission is established, and the original water hammer pressure pulse signal is generated.

[0009] S3. At the instant the absolute zero point is established, the displacement sensor data attached to the hydraulic cylinder piston rod is triggered and read. The absolute physical coordinates are obtained and combined with the fixed total length of the cylinder for algebraic calculation to calculate the dynamic physical cavity length.

[0010] S4. Obtain the physical sound velocity of the hydraulic oil at the current temperature, and use the physical sound velocity and dynamic physical cavity length to perform kinematic calculations to generate the theoretical echo arrival time.

[0011] S5. Generate a hardware-gated time window based on the theoretical echo arrival time, and use the hardware-gated time window to dynamically capture the original water hammer pressure pulse signal to extract the target echo signal.

[0012] S6. The target echo signal and the original water hammer pressure pulse signal are compared and analyzed by inputting the hardware zero-crossing detector to extract phase difference features and amplitude attenuation features.

[0013] S7. Combine phase difference characteristics and amplitude attenuation characteristics to identify the acoustic impedance mutation mechanism and generate internal leakage diagnosis results.

[0014] The second aspect of the present invention provides a hydraulic cylinder internal leakage detection system based on transient pressure waves, comprising: a global hardware trigger signal generation module, which collects the control signal of the electromagnetic directional valve in the hydraulic circuit for edge detection and extracts the falling edge level to generate a global hardware trigger signal.

[0015] The native water hammer pressure pulse signal generation module responds to the global hardware trigger signal to start the high-frequency dynamic pressure sensor to collect fluid pressure data at the hydraulic cylinder inlet pipeline. By identifying the high pressure jump edge in the pressure sequence that exceeds the preset differential steepness threshold, the absolute zero point time of the waveform transmission is established, and the native water hammer pressure pulse signal is generated.

[0016] The dynamic physical cavity length calculation module triggers and reads the displacement sensor data attached to the hydraulic cylinder piston rod at the instant the absolute zero point is established. It obtains the absolute physical coordinates and performs algebraic calculations in combination with the fixed total length of the cylinder to calculate the dynamic physical cavity length.

[0017] The theoretical echo arrival time generation module obtains the physical sound velocity of hydraulic oil at the current temperature, and uses the physical sound velocity and dynamic physical cavity length to perform kinematic calculations to generate the theoretical echo arrival time.

[0018] The target echo signal extraction module generates a hardware-gated time window based on the theoretical echo arrival time, and uses the hardware-gated time window to dynamically capture the original water hammer pressure pulse signal to extract the target echo signal.

[0019] The feature extraction module compares and analyzes the target echo signal with the original water hammer pressure pulse signal by inputting it into a hardware zero-crossing detector, and extracts phase difference features and amplitude attenuation features.

[0020] The internal leakage diagnosis result generation module combines phase difference characteristics and amplitude attenuation characteristics to identify the acoustic impedance change mechanism and generate internal leakage diagnosis results.

[0021] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) The present invention realizes real-time online detection of hydraulic cylinders in actual working process through precise synchronous hardware triggering and dynamic cavity length measurement. This method does not require stopping or disassembling the equipment, and directly uses the pressure fluctuation generated by the action of the solenoid valve during normal system operation as a diagnostic signal, which can capture the internal leakage status of the hydraulic cylinder in any piston position and movement state, thereby improving detection efficiency and equipment availability.

[0022] (2) The detection method proposed in this invention has good anti-interference ability and diagnostic accuracy. By setting a hardware-gated time window that follows the piston position, the target echo signal can be accurately captured, effectively eliminating the interference of complex pressure fluctuations and mechanical vibration noise inherent in the hydraulic system. This high signal-to-noise ratio signal extraction method lays the data foundation for subsequent feature analysis and ensures the reliability of the diagnostic results.

[0023] (3) This invention establishes an intelligent identification logic based on the acoustic impedance mutation mechanism by jointly analyzing the phase difference characteristics and amplitude attenuation characteristics of the echo signal. This method can not only determine whether there is energy attenuation, but also accurately distinguish between physical leakage caused by real sealing gaps and false attenuation caused by air bubbles in hydraulic oil by the physical phenomenon of whether the phase is reversed. This reduces the defect of traditional detection methods that are easily affected by oil quality and produce false alarms, and effectively reduces the false alarm rate.

[0024] (4) This invention ensures the synchronization accuracy of the trigger signal at the microsecond level by accurately matching the de-energization action of the electromagnetic coil through voltage sampling and differential operation; the weighted moving average filter can filter out high-frequency electromagnetic interference while maintaining the physical characteristics of the rising edge of the water hammer pulse without distortion; and the introduction of multi-parameter mapping compensation of physical sound speed, temperature and pressure overcomes the time calculation error caused by changes in the oil environment. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0027] Figure 2 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Please see Figure 1 The first aspect of the present invention provides a method for detecting internal leakage in a hydraulic cylinder based on transient pressure waves, comprising: S1, acquiring the control signal of the electromagnetic directional valve in the hydraulic circuit for edge detection, and extracting the falling edge level to generate a global hardware trigger signal.

[0030] In a specific embodiment of the present invention, the control signal of the electromagnetic directional valve in the hydraulic circuit is collected for edge detection, and the falling edge level is extracted to generate a global hardware trigger signal, including: receiving the control signal of the electromagnetic directional valve in the hydraulic circuit for voltage level sampling, and generating a real-time voltage sequence.

[0031] Differential operations are performed on the real-time voltage sequence to extract voltage abrupt change points.

[0032] The voltage abrupt change point is matched with the de-energization action of the electromagnetic coil, and the falling edge level is extracted to generate a global hardware trigger signal.

[0033] Specifically, this step aims to precisely capture the instantaneous state switching of the solenoid directional valve in the hydraulic system and generate a reference time signal for synchronizing all subsequent measurement operations. The entire process begins with continuous monitoring of the control signal of the solenoid directional valve in the hydraulic circuit. This control signal is a voltage level signal; its voltage fluctuations directly instruct the opening and closing of the solenoid valve, thereby controlling the flow of hydraulic oil. By sampling this signal at high frequency, it can be transformed from a continuous analog form into a discrete digital sequence, i.e., generating a real-time voltage sequence. The real-time voltage sequence forms the basis for subsequent digital analysis; it consists of a series of voltage values ​​measured at specific points in time.

[0034] To identify abrupt voltage changes from a stable real-time voltage sequence, differential operations are introduced. In this context, differential operations calculate the difference between the voltage values ​​at two adjacent sampling time points; this difference directly reflects the rate of voltage change within that small time interval. This allows for the quantification and differentiation between stable and abrupt voltage states, thereby extracting voltage abrupt change points. A voltage abrupt change point is defined as the moment when the absolute value of the differential operation result exceeds a preset threshold, indicating a significant change in the control signal state of the electromagnetic directional valve.

[0035] In one specific embodiment of the present invention, the typical value of the preset threshold for the voltage mutation point is set to 5. Its physical setting is based on the following: through spectrum monitoring of the electromagnetic directional valve control circuit under multiple operating conditions, the measured maximum background noise fluctuation voltage is 0.5. Set the difference threshold to 5 This ensures that at 100 Sufficient anti-interference redundancy is obtained under high-frequency sampling (0.01ms interval), which effectively filters out background glitches caused by contactor bounce or electromagnetic coupling, and can also prevent the voltage from dropping to 24 when the electromagnetic coil is de-energized. Precise locking is achieved immediately at approximately 80% of the range, thereby establishing a time reference with improved signal-to-noise ratio.

[0036] A voltage abrupt change, especially a sharp drop in voltage, is usually closely associated with the de-energization of the solenoid coil. The de-energization of the solenoid coil is the physical process of the solenoid directional valve switching from an energized to a de-energized state. This process causes its control signal voltage to drop rapidly from a high level to a low level, forming a falling edge. By matching the detected negative voltage abrupt change with this physical action, the instant the solenoid coil is de-energized can be precisely pinpointed. Once the system confirms the capture of this voltage abrupt change representing the falling edge level, it immediately generates a standard 5V transistor-to-transistor logic (TTL) level pulse signal, which serves as the global hardware trigger signal. This global hardware trigger signal acts as a high-precision synchronization beacon, distributed to other sensors and data acquisition units in the system, ensuring that all subsequent operations are activated under a unified time reference.

[0037] During implementation, differential operations are used to identify voltage abrupt changes, and their relationship can be expressed by the following formula:

[0038]

[0039] in, Represents the current sampling time point. This represents the previous sampling time point. At a certain point in time The collected real-time voltage values, At a certain point in time The collected real-time voltage value. At a certain point in time The calculated differential voltage is expressed in volts (V). ), and voltage The dimensions are consistent.

[0040] A voltage mutation point The conditions for being identified are:

[0041]

[0042] Here, It refers to the specific time point when the voltage surge occurs. It is a positive voltage threshold value used to distinguish between the true falling edge of the signal and background noise fluctuations, set to 5. . The setting is based on statistical analysis of the noise levels of electromagnetic directional valve control signals under 100 different operating conditions, selecting the minimum value greater than the maximum noise fluctuation amplitude to ensure the reliability of the detection. When the differential voltage... For a value less than a negative threshold When the number is negative, the system determines that... A falling edge occurred at a certain moment, triggering the generation of a global hardware trigger signal.

[0043] For example, in a hydraulic testing platform, the control signal for the solenoid directional valve is 24. DC signal, high level is 24 Low level is 0 The data acquisition system uses 100 The voltage level of the signal is sampled at a frequency of [frequency missing].

[0044] At a certain moment, the real-time voltage sequence generated by the system is at time point [time value missing]. to The following is a fragment with a time interval of 0.01 milliseconds: In At that moment, the voltage value was 24.01. ;exist At that moment, the voltage value was 23.99. ;exist At that moment, the voltage value was 24.00. ;exist At that moment, the operator issued a command to de-energize the electromagnetic coil, with the voltage value at 16.50. ;exist At that moment, the voltage value was 5.20. ;exist At that moment, the voltage value was 0.05. ;exist At that moment, the voltage value was 0.04. .

[0045] The system performs a difference operation on the real-time voltage sequence. According to the formula... calculate: exist time, .

[0046] exist time, .

[0047] exist time, .

[0048] exist time, .

[0049] exist time, .

[0050] Set voltage threshold 5 This setting is based on a maximum noise voltage of 0.5 on the control signal line for this model of solenoid directional valve when there is no operation. The data was determined based on actual measurements, with sufficient safety margins.

[0051] The system will calculate the differential voltage and the negative threshold. That is, compare with -5V.

[0052] exist time, This value is less than -5V, therefore the system is... It can identify a voltage change point at any time.

[0053] This voltage jump point matches the de-energization action of the electromagnetic coil, and the system confirms that this is a valid falling edge level.

[0054] Immediately, the system... Generate a standard 5 at any time The TTL signal serves as a global hardware trigger signal, which is distributed to subsequent pressure and displacement acquisition modules to achieve precise time synchronization throughout the diagnostic process.

[0055] S2. Responding to the global hardware trigger signal, the high-frequency dynamic pressure sensor is activated to collect fluid pressure data at the hydraulic cylinder inlet pipe. By identifying the high-pressure jump edge in the pressure sequence that exceeds the preset differential steepness threshold, the absolute zero point time of the waveform emission is established, and the original water hammer pressure pulse signal is generated.

[0056] In a specific embodiment of the present invention, a high-frequency dynamic pressure sensor is activated in response to a global hardware trigger signal to collect fluid pressure data at the hydraulic cylinder inlet pipe. The absolute zero point time of waveform transmission is established by identifying a high-pressure transition edge in the pressure sequence that exceeds a preset differential steepness threshold, and a native water hammer pressure pulse signal is generated. This includes receiving a global hardware trigger signal for clock synchronization and establishing the absolute zero point time of transient pressure wave transmission.

[0057] The waveform acquisition array of the high-frequency dynamic pressure sensor is activated at absolute zero time to capture fluid pressure data.

[0058] After high-frequency noise filtering of the fluid pressure data, steepness detection is performed. The pressure rise edge that exceeds the preset differential steepness threshold is established as the absolute zero point time of transient pressure wave emission. The original water hammer pressure pulse signal is generated by segmenting based on this absolute zero point time.

[0059] Specifically, this step, which closely follows the precise capture of the solenoid directional valve's de-energization, is primarily tasked with responding to the global hardware trigger signal generated in step S1 and utilizing a high-frequency dynamic pressure sensor to capture pressure fluctuations caused by the sudden fluid cutoff. When the global hardware trigger signal arrives, it provides a wake-up signal for the entire measurement system. Due to the millisecond-level mechanical hysteresis between the de-energization of the solenoid coil and the complete closure of the valve core, this moment cannot be used as the starting point for acoustic time-of-flight calculations. Transient pressure waves, commonly known as water hammer, are pressure pulses formed in the pipeline due to a rapid change in fluid momentum in the hydraulic circuit; their starting point is strongly correlated with the solenoid valve's closing action in time. Establishing the absolute zero point time is equivalent to calibrating the zero point for all subsequent time-related measurements, ensuring the accuracy of data analysis.

[0060] Upon receiving the wake-up signal, the system immediately activates the waveform acquisition array of the high-frequency dynamic pressure sensor located at the hydraulic cylinder inlet pipe. The high-frequency dynamic pressure sensor is a measuring element capable of responding to extremely rapid pressure changes; its waveform acquisition array is an electronic component within the sensor used to convert analog pressure signals into digital sequences. Activating this array means the sensor begins continuously capturing fluid pressure data at the hydraulic cylinder inlet pipe at an extremely high sampling rate, such as millions of times per second. This raw fluid pressure data is a mixed signal containing real pressure fluctuations and various interference noises. After filtering the acquired fluid pressure data, the system performs differential steepness calculations or sets a high-pressure jump threshold on the continuous data. The moment a rapid jump in pressure is detected is established by the system as the absolute zero point time of the transient pressure wave emission. Establishing the absolute zero point time is equivalent to calibrating the true physical zero point for all subsequent measurements related to the flight time of microsecond-level sound waves, overcoming the systematic errors caused by electromagnetic-mechanical response delays, and ensuring the accuracy of data analysis.

[0061] In one specific embodiment of the present invention, the typical value of the differential steepness threshold is set as follows: The corresponding high-voltage jump threshold is set as follows: in a single sampling period, such as... Internally, pressure suddenly increases The physical basis for setting this threshold is that, under normal operating conditions, the background hydraulic noise such as flow pulsation or mechanical vibration of the hydraulic main pump has a relatively smooth pressure change and a fluctuation rate that is usually lower than [the threshold value is missing here]. When the solenoid valve core instantly and completely blocks the oil flow, triggering a "water hammer effect," the fluid momentum undergoes a rapid conversion, and the pressure wave front exhibits significant shock wave characteristics. In the measured data, the pressure is at... Internal leap .

[0062] To extract clear pressure pulse characteristics, high-frequency noise filtering is required for the acquired fluid pressure data. High-frequency noise mainly originates from electromagnetic interference in sensor circuits or high-frequency vibrations in mechanical systems, which can obscure the details of the true pressure signal. High-frequency noise filtering is typically implemented using digital filters, which allow the relatively low-frequency main signal characterizing water hammer to pass through while suppressing noise components with frequencies higher than a set cutoff frequency. After this processing, the signal-to-noise ratio (SNR) is improved, ultimately generating a native water hammer pressure pulse signal with a steep rising edge. The native water hammer pressure pulse signal is a high SNR digital sequence characterizing the initial waveform of the water hammer phenomenon. Its steep rising edge represents the rapid process of pressure jumping from normal operating pressure to peak pressure; this clear edge is crucial for subsequent time measurements.

[0063] When performing high-frequency noise filtering, a weighted moving average filter can be used, and its calculation process is as follows:

[0064]

[0065] in, At a certain point in time The calculated filtered pressure value constitutes part of the original water hammer pressure pulse signal. At a certain point in time The raw fluid pressure data collected. These are the weighting coefficients of the filter, which determine the contribution of neighboring data points to the current filtering result. The sum of all coefficients is 1. This is to ensure that the DC component of the signal remains unchanged. It is the width of the filtering window. and The setting is based on a spectral analysis of 200 sets of sensor noise signals collected under standard operating conditions, selecting the parameter combination that can filter out the main noise frequency bands and has the least impact on the rising edge shape of the original water hammer pressure pulse signal. In this formula, all pressure terms... and The units are all Pascals (Pa) ), weighting coefficient Since it is a dimensionless pure number, the consistency of dimensions between the two ends of the formula is guaranteed.

[0066] For example, upon receiving the data from step S1... After the global hardware trigger signal is generated at any time, the system immediately executes step S2.

[0067] The system will The time is defined as the absolute zero point time of the transient pressure wave emission, denoted as . .

[0068] exist At a certain moment, the waveform acquisition array of the high-frequency dynamic pressure sensor installed at the hydraulic cylinder inlet pipe is activated, at 5... The sampling rate begins to capture fluid pressure data. Assuming that... Over the next few microseconds, a sequence of fluid pressure data (in megapascals) was acquired. The following section contains significant noise fluctuations: At that time, the pressure was 10.1. ;exist At that time, the pressure was 10.3. ;exist At that time, the pressure was 25.5. (Water hammer wave arrives); in At that time, the pressure was 35.1. ;exist At that time, the pressure was 34.8. ;exist At that time, the pressure was 35.2. .

[0069] The system then performs high-frequency noise filtering on these fluid pressure data. A window width of 3 (i.e., ...) is used. The weighted moving average filter with weights of 1) has weighting coefficients of 1. , , .

[0070] right The pressure value at any given time is filtered and calculated. ; .

[0071] By performing this filtering operation on the entire data sequence, a new pressure data sequence is obtained, which is the original water hammer pressure pulse signal. For example, the filtered sequence might look like this: At that time, the pressure was 10.15. ;exist At that time, the pressure was 15.35. ;exist At that time, the pressure was 27.8. ;exist At that time, the pressure was 32.625. ;exist At that time, the pressure was 34.975. .

[0072] Compared to the original data, this new sequence has smoother fluctuations. arrive The pressure pulse pattern exhibits a clear shape with a steep rising edge, which constitutes the original water hammer pressure pulse signal for subsequent analysis.

[0073] S3. At the instant the absolute zero point is established, the displacement sensor data attached to the hydraulic cylinder piston rod is triggered and read. The absolute physical coordinates are obtained and combined with the fixed total length of the cylinder for algebraic calculation to calculate the dynamic physical cavity length.

[0074] In a specific embodiment of the present invention, at the instant of absolute zero time, the displacement sensor data attached to the piston rod of the hydraulic cylinder is triggered and read, the absolute physical coordinates are obtained, and algebraic calculations are performed in combination with the fixed total length of the cylinder to calculate the dynamic physical cavity length, including: at the instant of establishing absolute zero time, triggering the data reading command of the displacement sensor to obtain the absolute physical coordinates at the current instant.

[0075] Algebraic calculations are performed by subtracting the absolute physical coordinates from the fixed total length of the hydraulic cylinder.

[0076] By subtracting the absolute physical coordinates from the fixed total length of the hydraulic cylinder and performing algebraic calculations, a dynamic physical cavity length representing the true physical length of the detection cavity at the current instant is generated.

[0077] Specifically, this step is executed in parallel with step S2, and its purpose is to accurately measure the internal geometry of the hydraulic cylinder at the instant the transient pressure wave is generated. The entire process is also triggered by the global hardware trigger signal generated in step S1, but in order to eliminate the displacement error caused by the piston continuing to move during the solenoid valve closing hysteresis, the precise measurement command trigger is strictly synchronized with the absolute zero point time of the pressure wave emission. When the system precisely pinpoints the moment of actual physical water hammer activation in step S2. Immediately afterward, the system sends a data reading command to the displacement sensor attached to the hydraulic cylinder piston rod. The displacement sensor is a measuring device capable of monitoring the extension or retraction position of the piston rod in real time. This command causes the displacement sensor to lock onto and output its current measurement value the instant it receives the signal; this value represents the absolute physical coordinates at the precise moment the water hammer event occurred. The absolute physical coordinates are a scalar quantity representing the displacement of the piston rod relative to a preset zero point at the exact moment the water hammer event occurred.

[0078] To calculate the actual distance the pressure wave needs to propagate from absolute physical coordinates, the system requires a crucial reference parameter: the fixed total length of the hydraulic cylinder. The fixed total length of the hydraulic cylinder is an inherent physical property of the cylinder; it represents the distance from the end face of the pressure sensor mounting position to the piston end face when the piston is fully retracted. This value is constant and is typically obtained from the equipment's design drawings or through calibration measurements.

[0079] After obtaining the instantaneously changing absolute physical coordinates and the constant total length of the hydraulic cylinder, the system determines the dynamic physical cavity length through a precise algebraic calculation. The dynamic physical cavity length is the final output of this step; it characterizes the actual physical length of the fluid cavity from the pressure sensor location to the piston's working surface at the instant the global hardware trigger signal is activated. This length is dynamically changing because it directly depends on the piston's real-time position. The dynamic physical cavity length is obtained by subtracting the absolute physical coordinates from the total fixed length of the hydraulic cylinder.

[0080] In this embodiment, the rod chamber is used as the oil inlet detection chamber and the origin of the displacement sensor coordinates is set at the fully retracted piston rod position. The calculation process can be described by the following formula:

[0081]

[0082] in, Represents the dynamic physical cavity length, with the unit being meters (m). ). This represents the fixed total length of the hydraulic cylinder, also in meters (m). This value is preset as a system constant, and its setting is based on the hydraulic cylinder design manual or by calibration measurement of the physical prototype using a laser interferometer. This represents the absolute physical coordinates read by the displacement sensor at the moment of triggering, in meters (m). In this formula, all variables are of the dimension of length, and the dimensions of both sides of the equation are consistent, so no additional correction is required.

[0083] For example, in the hydraulic system of a working excavator, its boom hydraulic cylinder is performing a retraction action. At point in time... The global hardware trigger signal generated in step S1 is sent to the high-frequency dynamic pressure sensor and displacement sensor controller to wake up the system.

[0084] Upon receiving a global hardware trigger signal, the displacement sensor controller immediately triggers a data reading command from the displacement sensor. This displacement sensor is a magnetostrictive type, with its zero point set at the fully retracted piston rod position. The displacement sensor controller... At that moment, due to command response locking, the captured and output absolute physical coordinates were: This indicates that the piston rod has now extended 0.352 meters from its fully retracted position.

[0085] Simultaneously, the system retrieves the fixed total length of the hydraulic cylinder for this model from its configuration parameter library. This parameter is preset according to the hydraulic cylinder's design drawings, and its value is... rice.

[0086] The system then performs precise algebraic calculations by subtracting the absolute physical coordinates from the fixed total length of the hydraulic cylinder. According to the formula... Substitute the specific values ​​into the calculation:

[0087] Calculation results show that at the precise instant the solenoid valve closes and the water hammer effect occurs, the actual physical length of the fluid cavity on the hydraulic cylinder inlet side, i.e., from the pressure sensor to the piston end face, is 0.648 meters. The system outputs this calculated result of 0.648 meters as the dynamic physical cavity length for subsequent step S4 to calculate the theoretical propagation time of the pressure wave.

[0088] S4. Obtain the physical sound velocity of the hydraulic oil at the current temperature, and use the physical sound velocity and dynamic physical cavity length to perform kinematic calculations to generate the theoretical echo arrival time.

[0089] In a specific embodiment of the present invention, the physical sound velocity of hydraulic oil at the current temperature is obtained, and kinematic calculations are performed using the physical sound velocity and the dynamic physical cavity length to generate the theoretical echo arrival time. This includes: obtaining real-time temperature data and static system pressure data inside the hydraulic cylinder, querying the fluid sound velocity mapping table based on the real-time temperature data and static system pressure data, and obtaining the physical sound velocity of hydraulic oil at the current temperature and pressure.

[0090] The round-trip propagation distance is obtained by multiplying the dynamic physical cavity length by a constant coefficient.

[0091] By dividing the round-trip propagation distance by the physical speed of sound, kinematic calculations are performed to generate the theoretical echo arrival time of the transient pressure wave's round-trip dynamic physical cavity length.

[0092] Specifically, the core task of this step is to predict the arrival time of the pressure wave echo based on known physical parameters through kinematic calculations. This prediction is crucial for the subsequent accurate capture of the echo signal. The entire process begins with acquiring a key physical characteristic of the hydraulic oil: the physical velocity of sound. The physical velocity of sound refers to the speed at which pressure disturbances (i.e., sound waves) propagate in a fluid medium, and it is affected by the temperature, pressure, and composition of the medium. In this embodiment, the system first acquires real-time temperature data and static system pressure data inside the hydraulic cylinder using a temperature sensor installed inside the cylinder or close to its outer wall. Since the physical velocity of sound in hydraulic oil varies with temperature, the system does not use a fixed value but instead queries a pre-calibrated fluid velocity mapping table containing both temperature and pressure parameters. This fluid velocity mapping table is a database storing the corresponding physical velocities of a specific type of hydraulic oil at different temperatures and pressures, and its data is established based on experimental measurements of the sound velocity of that type of hydraulic oil at different temperatures and pressures. By using real-time temperature data as an index, the system can accurately retrieve the physical velocity of sound under the current operating conditions from the table.

[0093] After determining the propagation speed of the pressure wave, it is also necessary to determine the length of its propagation path. The transient pressure wave is generated at the pressure sensor located at the oil inlet, propagates to the piston end face, and then reflects back to the pressure sensor. Therefore, the total path of the waveform propagation is twice the length of the dynamic physical cavity. The system multiplies the dynamic physical cavity length calculated in step S3 by a constant coefficient of 2 to obtain the round-trip propagation distance.

[0094] Finally, the system divides the calculated round-trip propagation distance by the physical speed of sound obtained from the fluid sound speed mapping table, performing a basic kinematic calculation. The result of this calculation is the theoretical flight time required for the transient pressure wave to travel from emission to its echo being received by the sensor again, i.e., the theoretical echo arrival time. This time is a relative time value relative to the absolute zero point time established in step S2.

[0095] The above kinematic calculation process can be precisely expressed by the following formula:

[0096]

[0097] in, Represents the theoretical echo arrival time, and its unit is seconds ( ). The dynamic physical cavity length is calculated from step S3, and the unit is meters (m). ). It is the physical speed of sound of hydraulic oil at the current temperature, obtained by looking up a table based on real-time temperature data, in meters per second (m / s). The constant coefficient 2 is dimensionless and represents one round trip of the pressure wave. In this formula, the dimensionless value on the right side is... The time dimension is consistent with that on the left, ensuring the correctness of the physical meaning.

[0098] For example, continuing with the aforementioned example of the excavator boom hydraulic cylinder, the system has already calculated the dynamic physical cavity length in step S3 at the instant the global hardware trigger signal is triggered. It is 0.648 meters.

[0099] Now, the system executes step S4. The temperature sensor installed in the hydraulic cylinder's oil circuit collects and uploads real-time temperature data inside the hydraulic cylinder, which is 45 degrees Celsius. Upon receiving this temperature value, the system immediately queries the preset fluid sound velocity mapping table for the L-HM46 anti-wear hydraulic oil used in this excavator. The table lookup result shows that at 45 degrees Celsius, the physical sound velocity of this hydraulic oil is 1380 m / s. Therefore, the system obtains the physical sound velocity of the hydraulic oil at the current temperature. It is 1380 m / s.

[0100] Next, the system multiplies the dynamic physical cavity length by a constant coefficient of 2 to obtain the round-trip propagation distance: Round-trip propagation distance .

[0101] Finally, the system divides the round-trip propagation distance by the physical speed of sound to perform kinematic calculations, generating the theoretical echo arrival time. According to the formula... Perform the calculation:

[0102] To facilitate subsequent processing, the system converts this result to microseconds, specifically 939.13 microseconds. This value represents the theoretical time required for the transient pressure wave to travel from its generation to its echo returning to the pressure sensor. The system ultimately generates and outputs a theoretical echo arrival time of 939.13 microseconds, which will serve as the core basis for setting the capture window in the next step.

[0103] S5. Generate a hardware-gated time window based on the theoretical echo arrival time, and use the hardware-gated time window to dynamically capture the original water hammer pressure pulse signal to extract the target echo signal.

[0104] In a specific embodiment of the present invention, a hardware-gated time window is generated based on the theoretical echo arrival time. The original water hammer pressure pulse signal is dynamically intercepted using the hardware-gated time window to extract the target echo signal. This includes: extending a preset time margin forward and backward with the theoretical echo arrival time as the center node to generate the hardware-gated time window.

[0105] The hardware-gated time window is applied to the acquisition time axis of the original water hammer pressure pulse signal to perform dynamic follow-up interception operation.

[0106] Pressure waveform data within the hardware-gated time window is extracted to generate a target echo signal that excludes piston displacement interference.

[0107] Specifically, this step aims to precisely separate the echo reflected from the piston end face from the continuously acquired pressure signal. The entire process utilizes the theoretical echo arrival time calculated in step S4 as a positioning beacon, capturing the target signal by setting a dynamic time window. First, the system generates a hardware-gated time window with the theoretical echo arrival time as the central reference. The hardware-gated time window is a precisely defined time interval, its function being to allow only data within that interval to pass through. To generate this window, the system extends a preset time margin in both directions before and after the theoretical echo arrival time, based on the central node. This preset time margin is to compensate for minor errors that may exist during the calculation process, such as slight fluctuations in the physical speed of sound or the limit accuracy error of the dynamic physical cavity length measurement, thereby ensuring that the actual echo signal can fall completely within the window.

[0108] After setting the start and end times of the hardware gating time window, the system performs a dynamic follow-up interception operation. This operation applies the hardware gating time window to the acquisition time axis of the original water hammer pressure pulse signal generated and continuously acquired in step S2. The "follow-up" characteristic here is crucial because it means that in each diagnostic cycle, the position of the hardware gating time window is dynamically adjusted according to the latest calculated theoretical echo arrival time, thereby adapting to changes in the echo arrival time of the piston in different positions or even in different motion states.

[0109] During the interception operation, the system compares the timestamp of each data point in the original water hammer pressure pulse signal with the boundary of the hardware-gated time window. Only data points whose timestamps are after the start time and before the end time of the window are selected and extracted. In this way, the system accurately intercepts a small segment of pressure waveform data containing echo characteristics from the long-term pressure signal. This extracted data sequence constitutes the target echo signal. Since the position of the hardware-gated time window is precisely calculated based on the instantaneous position of the piston, this method can effectively eliminate signal aliasing or misjudgment caused by piston movement displacement, ensuring that the extracted target echo signal is interference-filtered and directly related to the initial pulse.

[0110] Start time of hardware-gated time window and end time It can be defined by the following formula:

[0111]

[0112]

[0113] in, The theoretical echo arrival time is calculated from step S4, in seconds. ). It is the preset time margin, in seconds. The setting is based on experiments conducted on 500 hydraulic cylinders under different working conditions, statistical analysis of the deviation distribution between theoretical and actual echo arrival times, and selection of time values ​​that can cover 99.7% of the deviation range to ensure the integrity of the capture. and The opening and closing times of the hardware gating time window are defined separately, both in seconds. ), which is consistent with the dimensions of other terms in the formula.

[0114] For example, following the steps described above, the system has already calculated the theoretical echo arrival time in step S4. It is 939.13 microseconds.

[0115] The system now generates a hardware-gated time window based on this theoretical echo arrival time. The system sets an extremely narrow time margin. The value is set at 10 microseconds. This value is based on 500 tests conducted on the hydraulic cylinder of this model of excavator boom under various load and temperature conditions. The statistical results show that the maximum absolute value of the difference between the theoretical and measured echo time is 8.5 microseconds. Therefore, a margin of 10 microseconds is sufficient to cover all expected errors.

[0116] The system calculates the boundaries of the hardware-gated time window based on the formula:

[0117] Start time .

[0118] End time .

[0119] This means that the hardware gating time window is set relative to absolute zero. The interval between 929.13 microseconds and 949.13 microseconds.

[0120] Next, the system applies this hardware-gated time window to the acquisition time axis of the continuously acquired native water hammer pressure pulse signal, performing a dynamic retrieval operation. The data acquisition system uses 5... The sampling rate (i.e., one data point every 0.2 microseconds) records the pressure. The system will filter out all timestamps. satisfy Pressure data points.

[0121] Assume that during this time period, some data of the original water hammer pressure pulse signal were collected as follows: At that time, the pressure was 10.2. ;exist At that time, the pressure was 10.3. ;exist At that time, the pressure was 11.5. ;exist At that time, the pressure was 18.2. (Echo arrives); in At that time, the pressure was 24.5. ;exist At that time, the pressure was 10.5. ;exist At that time, the pressure was 10.4. .

[0122] The system will extract from the timestamp arrive All pressure data points between these points are combined to form a new data sequence of 19.8 microseconds in length. This precisely extracted data sequence is the target echo signal, excluding interference from piston movement displacement. It clearly contains the pressure echo characteristics arriving around 939.0 microseconds, which can be used for precise feature analysis in subsequent steps.

[0123] S6. The target echo signal and the original water hammer pressure pulse signal are compared and analyzed by inputting the hardware zero-crossing detector to extract phase difference features and amplitude attenuation features.

[0124] In a specific embodiment of the present invention, the target echo signal and the original water hammer pressure pulse signal are compared and analyzed by inputting the hardware zero-crossing detector to extract phase difference features and amplitude attenuation features, including: inputting the target echo signal and the original water hammer pressure pulse signal into the hardware zero-crossing detector to extract the zero-crossing time and generate the zero-crossing time difference.

[0125] Phase difference features are extracted by converting the electrical angle based on the zero-crossing time difference and the signal period.

[0126] By comparing the peak amplitudes of the target echo signal and the original water hammer pressure pulse signal, the amplitude attenuation characteristics are extracted.

[0127] Specifically, the core of this step is to perform a precise quantization comparison between the original pressure pulse and its echo, aiming to extract the changes that occur during the propagation and reflection of the pressure wave within the hydraulic cylinder. These changes directly reflect the acoustic characteristics of the reflecting interface. The entire process begins by simultaneously inputting two key digital signals—the original water hammer pressure pulse signal generated in step S2 and the target echo signal extracted in step S5—into a dedicated hardware circuit, namely a hardware zero-crossing detector. The hardware zero-crossing detector is a high-speed comparison circuit whose function is to accurately capture the moment when the waveform of the input signal crosses a preset reference level, typically the signal's average value or zero level, and output a corresponding timestamp.

[0128] To extract phase information, the hardware zero-crossing detector first extracts the zero-crossing time of the two signals. It identifies the first zero-crossing point after the main pulse in both the original water hammer pressure pulse signal and the target echo signal. By recording the timestamps of these two zero-crossing points and calculating the time difference between them, the system generates the zero-crossing time difference. This time difference directly reflects the morphological time delay or lead of the echo signal relative to the original signal.

[0129] However, a simple time difference is insufficient to fully characterize phase changes because it is related to the frequency of the signal itself. Therefore, the system needs to perform electrical angle conversion based on the zero-crossing time difference and the signal period. The signal period is the duration of the main oscillation portion of the original water hammer pressure pulse signal, which is determined by the inherent physical characteristics of the hydraulic system and can be predetermined through spectral analysis of the signal or based on a system model. By dividing the zero-crossing time difference by the signal period and then multiplying by 360 degrees, the system converts the time difference into an angular difference, thereby extracting the phase difference characteristic.

[0130] Meanwhile, to extract energy loss information, the system compares the peak amplitudes of the target echo signal and the original water hammer pressure pulse signal. The system iterates through the data sequences of both signals to find their respective maximum values, i.e., peak pressures. By calculating the ratio of the peak amplitude of the target echo signal to the peak amplitude of the original water hammer pressure pulse signal, the system extracts the amplitude attenuation characteristics. This dimensionless ratio intuitively represents the degree of attenuation of pressure wave energy during its round-trip propagation.

[0131] The formula for calculating the phase difference characteristic is as follows:

[0132]

[0133] in, This is the final extracted phase difference feature, in degrees ( ). It is the zero-crossing timestamp of the target echo signal measured by the hardware zero-crossing detector. It is the zero-crossing timestamp of the original water hammer pressure pulse signal, and both are in seconds. ). It is the signal period, in seconds. The formula is based on a Fast Fourier Transform analysis of 1000 original water hammer pressure pulse signals under normal operating conditions, using the reciprocal of the dominant frequency as the characteristic period. (Right side of the formula) It is dimensionless; after multiplying by the angle, the dimension becomes degrees, consistent with the left side.

[0134] The formula for calculating the amplitude attenuation characteristic is as follows:

[0135]

[0136] in, It is the extracted amplitude decay feature, which is a dimensionless pure number. It is the peak amplitude of the target echo signal. This refers to the peak amplitude of the original water hammer pressure pulse signal; both are measured in Pascals (Pa). (or multiples thereof) units are eliminated when calculating ratios, ensuring the dimensionless nature of the results.

[0137] For example, let's continue with the diagnostic scenario for the hydraulic cylinder of the excavator boom.

[0138] The system inputs the native water hammer pressure pulse signal generated in step S2 and the target echo signal extracted in step S5 into the hardware zero-crossing detector.

[0139] Data from the original water hammer pressure pulse signal shows that its peak amplitude It is 35.0 The first zero-crossing timestamp after its main pulse The time measured by the hardware zero-crossing detector is relative to the absolute zero point. The next 5.2 microseconds.

[0140] Data from the target echo signal shows that its peak amplitude It is 28.0 The first zero-crossing timestamp after its main pulse The measurement is relative to the absolute zero time. The next 944.5 microseconds.

[0141] To perform phase comparison, the two signals need to be aligned. The system subtracts the theoretical echo arrival time from the time axis of the target echo signal. (939.13 microseconds) to observe the time changes in its internal morphology. After alignment, the zero-crossing time of the target echo signal is... .

[0142] Therefore, the time difference at the zero crossing is .

[0143] Based on the analysis of the hydraulic system model, the system presets the signal cycle. It is 20.0 microseconds.

[0144] The system calculates the phase difference characteristics according to the formula: .

[0145] The phase difference feature extracted by the system is 3.06 degrees.

[0146] Next, the system compares the peak amplitudes of the two signals to extract amplitude attenuation features.

[0147] Calculate according to the formula: .

[0148] The amplitude attenuation feature extracted by the system is 0.8.

[0149] Finally, step S6 outputs two key quantization metrics: a phase difference of 3.06 degrees and an amplitude attenuation of 0.8. These two features will be used in step S7 for final fault diagnosis.

[0150] S7. Combine phase difference characteristics and amplitude attenuation characteristics to identify the acoustic impedance mutation mechanism and generate internal leakage diagnosis results.

[0151] In a specific embodiment of the present invention, the acoustic impedance mutation mechanism is identified by combining phase difference characteristics and amplitude attenuation characteristics, and an internal leakage diagnosis result is generated, including: if the attenuation judgment flag is false, it is determined that the system has not experienced significant energy dissipation, and an internal leakage diagnosis result confirming that the internal sealing state of the hydraulic cylinder is healthy is generated.

[0152] When the attenuation determination flag is true, the phase difference characteristics are analyzed to determine whether a physical phase reversal has occurred, and a phase reversal flag is generated.

[0153] By combining attenuation judgment flags and phase reversal flags, the acoustic impedance mutation mechanism is identified, and internal leakage diagnosis results are generated.

[0154] In a specific embodiment of the present invention, the acoustic impedance mutation mechanism is identified by combining the attenuation determination flag and the phase reversal flag to generate an internal leakage diagnosis result, including: extracting the attenuation determination flag and the phase reversal flag and logically combining them to generate a feature combination vector.

[0155] When the feature combination vector indicates that there is amplitude attenuation and physical phase reversal, it is determined that there is a local acoustic impedance mutation, and an internal leakage diagnosis result is generated, confirming that there is a physical leakage in the hydraulic cylinder piston due to a real sealing gap.

[0156] When the characteristic combination vector indicates that there is amplitude attenuation and no physical phase reversal occurs, it is determined that there is false attenuation of two-phase flow, and an internal leakage diagnosis result is generated that automatically filters out bubble interference and does not trigger an alarm.

[0157] Specifically, this final step is the decision-making stage of the entire diagnostic process. Its core is based on acoustic physics principles, comprehensively interpreting the phase difference and amplitude attenuation characteristics extracted from step S6 to identify the physical properties of the pressure wave reflection interface, thereby generating a clear internal leakage diagnosis result. This process is called acoustic impedance mutation mechanism identification, which involves analyzing the differences in reflection characteristics of pressure waves when encountering different media interfaces to determine the root cause of the fault.

[0158] The diagnostic logic first analyzes the amplitude attenuation characteristics to determine whether there is significant energy dissipation during the propagation and reflection of the pressure wave. The system compares the amplitude attenuation characteristics calculated in step S6 with a preset attenuation threshold. If the amplitude attenuation characteristics are less than the threshold, it indicates that the energy of the echo signal is much lower than that of the original signal, indicating energy dissipation, and the system generates an attenuation judgment flag with a value of "true". Conversely, if the amplitude attenuation characteristics are greater than or equal to the threshold, it is considered that the energy attenuation is within the normal range, and the attenuation judgment flag is "false".

[0159] In one specific embodiment of the present invention, a preset attenuation threshold for the comparison amplitude is used. The typical value is set at 0.95. The physical and statistical basis for this threshold setting is as follows: Through benchmark tests on water hammer wave propagation of 300 hydraulic cylinders of different models under healthy sealing conditions across the entire operating range, the actual measurements showed that even under fault-free and seamless conditions, due to the inherent viscous resistance of hydraulic oil and friction along the pipe wall, the pressure waveform still produces a slight natural energy dissipation, and the healthy echo amplitude retention rate typically fluctuates between 0.97 and 0.99. Based on this massive amount of measured data, the system extracted a natural attenuation distribution model and selected a value slightly below its 99% confidence interval lower limit (i.e., 0.95) as the judgment threshold.

[0160] Assuming the attenuation determination flag is "true," the system further analyzes the phase difference characteristics to determine if the reflection is caused by an interface with reduced acoustic impedance. The system checks if the phase difference characteristics are close to 180 degrees, indicating whether a physical phase reversal has occurred. Phase reversal occurs when a pressure wave propagates from a high acoustic impedance medium (hydraulic oil) to an interface with a low acoustic impedance medium. The system confirms this by determining if the phase difference characteristics fall within a small range centered at 180 degrees. If the condition is met, the system generates a "true" phase reversal flag; otherwise, the flag is "false."

[0161] Subsequently, the system logically combines the attenuation judgment flag and the phase reversal flag to generate a feature combination vector, and uses this vector to perform the final acoustic impedance abrupt change mechanism identification. When the feature combination vector indicates the presence of amplitude attenuation and simultaneous physical phase reversal, this perfectly matches the physical behavior of a pressure wave encountering a local acoustic impedance abrupt change in a leakage pore. Therefore, the system determines that the hydraulic cylinder piston has a physical leak caused by a real sealing pore and generates the corresponding internal leakage diagnosis result. The physical mechanism is as follows: when a pressure wave in the hydraulic oil encounters a tiny pore (leading to the low-pressure chamber) caused by piston seal failure, this pore is equivalent to the "open end" model in acoustics. The local acoustic impedance of the reflecting surface instantly drops to a level far lower than the impedance of the hydraulic oil system, causing the reflected pressure wave to undergo a 180-degree phase jump relative to the incident pressure wave. This process can produce a phase difference under ideal acoustic conditions, while in practical applications, the complex topological geometry and viscous boundary constraints are overcome by using a tolerance threshold.

[0162] When the characteristic combination vector indicates amplitude attenuation but no physical phase reversal, this situation is usually caused by air bubbles mixed in the hydraulic oil, resulting in pseudo-attenuation of the two-phase flow. Air bubbles scatter and absorb sound wave energy, leading to attenuation, but their reflection characteristics do not cause a 180-degree phase reversal. Therefore, the system determines this to be air bubble interference and generates an internal leak diagnosis result that automatically filters out this interference and does not trigger an alarm. Through quantitative comparison of physical characteristics, it overcomes the shortcomings of traditional methods that are easily affected by air bubble interference, improving the scientific accuracy of the detection. The physical mechanism is that the tiny air bubbles in the oil are diffusely distributed, mainly causing energy dissipation (amplitude attenuation) through scattering effects and multiple reflections. However, the macroscopic reflection interface (piston end face) in a statistical sense is still a solid high-impedance boundary, which does not meet the impedance transition condition for phase reversal, thus achieving automatic filtering of non-leakage interference.

[0163] Attenuation judgment flag The generation logic is as follows:

[0164]

[0165] in, It is the amplitude attenuation feature extracted from step S6. It is the attenuation threshold, a dimensionless pure number less than 1. It is set based on testing 300 sets of healthy hydraulic cylinders of different models in the full range of working conditions, statistically analyzing the distribution of their amplitude attenuation characteristics, and selecting a value lower than the lower limit of the 99% confidence interval to ensure that normal attenuation can be distinguished from abnormal energy dissipation.

[0166] Phase Reversal Flag The generation logic is as follows, and this logic only applies to... Execute if true:

[0167]

[0168] in, It is the phase difference feature extracted in step S6, in degrees ( ). It is the tolerance threshold for phase reversal, in degrees ( This is used to define the range of "approaching" 180 degrees. Its setting is based on analyzing the fluctuation range of phase difference through finite element acoustic simulation of hydraulic cylinders containing leakage orifices of different sizes, and selecting the maximum tolerance that can stably identify phase reversal phenomena.

[0169] In a specific embodiment of the present invention, the tolerance threshold for phase reversal is... The typical value is set as When the measured phase difference falls within to Within a certain range, it is considered 'close' and a physical phase reversal occurs. The physical and engineering basis for this threshold setting is that in the ideal acoustic classical absolute 'open end' model, impedance abrupt changes will theoretically cause a 180-degree phase reversal; however, inside industrial hydraulic cylinders, the real tiny leakage pores formed by piston seal damage often have extremely complex topological geometry and viscous boundary constraints. This non-ideal 'slot-type low impedance' boundary, accompanied by weak acoustic blind reactance, coupled with the discrete sampling truncation error of the high-frequency data acquisition system, will inevitably cause a slight forward or backward phase distortion of the reflected signal. In the early research and development, through three-dimensional finite element acoustic simulation and experimental benchmarking of hydraulic cylinders containing a large number of pores of different sizes and wear morphologies, it was found that the extreme value of the phase reversal fluctuation caused by pore reflection is usually within a certain range. Within the range.

[0170] For example, following all the aforementioned steps, the system has already extracted the phase difference features in step S6. and amplitude attenuation characteristics .

[0171] System preset attenuation threshold The tolerance threshold for phase reversal is 0.95. It is 20 degrees.

[0172] The system first analyzes the amplitude attenuation characteristics. and Comparison. Because The condition for energy dissipation is met, therefore the system generates a decay judgment flag. It is "true".

[0173] When the amplitude attenuation characteristic is greater than or equal to the preset attenuation threshold (e.g., 0.95), the system generates an attenuation judgment flag with a value of "false," determining that the internal sealing state of the hydraulic cylinder is healthy. No further phase analysis logic is needed, and a normal result is directly output. This branch logic completes the system's automatic identification capability for healthy operating conditions.

[0174] Since the attenuation determination flag is "true," the system continues to analyze the phase difference characteristics. Calculation Compare this result with the tolerance threshold. Comparison. Because The conditions for a physical phase reversal are not met, so the system generates a phase reversal flag. It is "fake".

[0175] Next, the system extracts the attenuation determination flag and the phase reversal flag and performs a logical combination to generate a feature combination vector. .

[0176] The system identifies the acoustic impedance abrupt change mechanism based on this characteristic combination vector. This vector indicates the presence of amplitude attenuation without physical phase reversal, which matches the physical mechanism of pseudo-attenuation in two-phase flow. Therefore, the system ultimately generates an internal leakage diagnosis result that automatically filters out bubble interference and does not trigger an alarm.

[0177] In another test, it is assumed that the feature extracted in step S6 is the amplitude attenuation feature. Phase difference characteristics .

[0178] System analysis of amplitude attenuation characteristics: Therefore, the attenuation judgment flag It is "true".

[0179] System analysis of phase difference characteristics: .because The conditions for a physical phase reversal are met, therefore the phase reversal indicator... It is "true".

[0180] The feature combination vector generated by the system is .

[0181] This vector indicates amplitude attenuation and a physical phase reversal, indicating a local acoustic impedance abrupt change. The system ultimately generates a diagnostic result confirming an internal leak in the hydraulic cylinder piston due to a genuine physical leak in the sealing pores, and sends an alarm to the monitoring system.

[0182] Reference Figure 2 The second aspect of the present invention provides a hydraulic cylinder internal leakage detection system based on transient pressure waves, including: a global hardware trigger signal generation module, a native water hammer pressure pulse signal generation module, a dynamic physical cavity length calculation module, a theoretical echo arrival time generation module, a target echo signal extraction module, a feature extraction module, and an internal leakage diagnosis result generation module.

[0183] The global hardware trigger signal generation module is connected to the native water hammer pressure pulse signal generation module. The global hardware trigger signal generation module is connected to the dynamic physical cavity length calculation module. The dynamic physical cavity length calculation module is connected to the theoretical echo arrival time generation module. Both the native water hammer pressure pulse signal generation module and the theoretical echo arrival time generation module are connected to the target echo signal extraction module. Both the native water hammer pressure pulse signal generation module and the target echo signal extraction module are connected to the feature extraction module. The feature extraction module is connected to the internal leakage diagnosis result generation module.

[0184] The global hardware trigger signal generation module collects the control signal of the electromagnetic directional valve in the hydraulic circuit, performs edge detection, and extracts the falling edge level to generate a global hardware trigger signal.

[0185] The native water hammer pressure pulse signal generation module responds to the global hardware trigger signal to start the high-frequency dynamic pressure sensor to collect fluid pressure data at the hydraulic cylinder inlet pipeline. By identifying the high pressure jump edge in the pressure sequence that exceeds the preset differential steepness threshold, the absolute zero point time of the waveform transmission is established, and the native water hammer pressure pulse signal is generated.

[0186] The dynamic physical cavity length calculation module triggers and reads the displacement sensor data attached to the hydraulic cylinder piston rod at the instant the absolute zero point is established. It obtains the absolute physical coordinates and performs algebraic calculations in combination with the fixed total length of the cylinder to calculate the dynamic physical cavity length.

[0187] The theoretical echo arrival time generation module obtains the physical sound velocity of hydraulic oil at the current temperature, and uses the physical sound velocity and dynamic physical cavity length to perform kinematic calculations to generate the theoretical echo arrival time.

[0188] The target echo signal extraction module generates a hardware-gated time window based on the theoretical echo arrival time, and uses the hardware-gated time window to dynamically capture the original water hammer pressure pulse signal to extract the target echo signal.

[0189] The feature extraction module compares and analyzes the target echo signal with the original water hammer pressure pulse signal by inputting it into a hardware zero-crossing detector, and extracts phase difference features and amplitude attenuation features.

[0190] The internal leakage diagnosis result generation module combines phase difference characteristics and amplitude attenuation characteristics to identify the acoustic impedance change mechanism and generate internal leakage diagnosis results.

[0191] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for detecting internal leak in a hydraulic cylinder based on transient pressure wave, characterized in that, include: S1. Collect the control signal of the electromagnetic directional valve in the hydraulic circuit, perform edge detection, extract the falling edge level, and generate a global hardware trigger signal; S2. Responding to the global hardware trigger signal, the high-frequency dynamic pressure sensor is activated to collect fluid pressure data at the hydraulic cylinder inlet pipeline. By identifying the high-pressure jump edge in the pressure sequence that exceeds the preset differential steepness threshold, the absolute zero point time of the waveform transmission is established, and the original water hammer pressure pulse signal is generated. S3. At the instant the absolute zero point is established, the displacement sensor data attached to the hydraulic cylinder piston rod is triggered and read. The absolute physical coordinates are obtained and combined with the fixed total length of the cylinder for algebraic calculation to calculate the dynamic physical cavity length. S4. Obtain the physical sound velocity of the hydraulic oil at the current temperature, and use the physical sound velocity and dynamic physical cavity length to perform kinematic calculations to generate the theoretical echo arrival time. S5. Generate a hardware-gated time window based on the theoretical echo arrival time, and use the hardware-gated time window to dynamically capture the original water hammer pressure pulse signal and extract the target echo signal. S6. Input the target echo signal and the original water hammer pressure pulse signal into the hardware zero-crossing detector for comparison and analysis, and extract the phase difference characteristics and amplitude attenuation characteristics. S7. Combine phase difference characteristics and amplitude attenuation characteristics to identify the acoustic impedance mutation mechanism and generate internal leakage diagnosis results.

2. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, The control signal of the electromagnetic directional valve in the hydraulic circuit is acquired, and edge detection is performed to extract the falling edge level and generate a global hardware trigger signal, including: The system receives control signals from the electromagnetic directional valve in the hydraulic circuit, samples the voltage level, and generates a real-time voltage sequence. Perform differential operations on the real-time voltage sequence to extract voltage abrupt changes; The voltage abrupt change point is matched with the de-energization action of the electromagnetic coil, and the falling edge level is extracted to generate a global hardware trigger signal.

3. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, The response to the global hardware trigger signal activates the high-frequency dynamic pressure sensor to collect fluid pressure data at the hydraulic cylinder inlet pipe. By identifying the absolute zero point time of the waveform emission at the high-pressure jump edge exceeding the preset differential steepness threshold in the pressure sequence, a native water hammer pressure pulse signal is generated, including: Receive global hardware trigger signals for clock synchronization and establish the absolute zero point time for transient pressure wave transmission; The waveform acquisition array of the high-frequency dynamic pressure sensor is activated at absolute zero time to capture fluid pressure data; After high-frequency noise filtering of the fluid pressure data, steepness detection is performed. The pressure rise edge that exceeds the preset differential steepness threshold is established as the absolute zero point time of transient pressure wave emission. The original water hammer pressure pulse signal is generated by segmenting based on this absolute zero point time.

4. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, At the instant the absolute zero point is established, the displacement sensor data attached to the hydraulic cylinder piston rod is triggered and read to obtain the absolute physical coordinates. Combined with the fixed total length of the cylinder, algebraic calculations are performed to calculate the dynamic physical cavity length, including: At the instant the absolute zero point is established, the data reading command of the displacement sensor is triggered to obtain the absolute physical coordinates at the current instant; Algebraic calculations were performed by subtracting the absolute physical coordinates from the fixed total length of the hydraulic cylinder. By subtracting the absolute physical coordinates from the fixed total length of the hydraulic cylinder and performing algebraic calculations, a dynamic physical cavity length representing the true physical length of the detection cavity at the current instant is generated.

5. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, The process of obtaining the physical velocity of sound of hydraulic oil at the current temperature, performing kinematic calculations using the physical velocity of sound and the dynamic physical cavity length, and generating the theoretical echo arrival time includes: Obtain real-time temperature data and static system pressure data inside the hydraulic cylinder, and query the fluid sound velocity mapping table based on the real-time temperature data and static system pressure data to obtain the physical sound velocity of the hydraulic oil at the current temperature and pressure. Multiplying the dynamic physical cavity length by a constant coefficient yields the round-trip propagation distance; By dividing the round-trip propagation distance by the physical speed of sound, kinematic calculations are performed to generate the theoretical echo arrival time of the transient pressure wave's round-trip dynamic physical cavity length.

6. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, The process involves generating a hardware-gated time window based on the theoretical echo arrival time, and using this window to dynamically capture the original water hammer pressure pulse signal to extract the target echo signal, including: A hardware-gated time window is generated by extending a preset time margin forward and backward with the theoretical echo arrival time as the center node. Apply the hardware-gated time window to the acquisition time axis of the original water hammer pressure pulse signal to perform dynamic follow-up interception operation; Pressure waveform data within the hardware-gated time window is extracted to generate a target echo signal that excludes piston displacement interference.

7. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, The process of inputting the target echo signal and the original water hammer pressure pulse signal into a hardware zero-crossing detector for comparison and analysis, and extracting phase difference features and amplitude attenuation features, includes: The target echo signal and the original water hammer pressure pulse signal are input into a hardware zero-crossing detector to extract the zero-crossing time and generate the zero-crossing time difference. The phase difference feature is extracted by converting the electrical angle based on the zero-crossing time difference and the signal period; By comparing the peak amplitudes of the target echo signal and the original water hammer pressure pulse signal, amplitude attenuation characteristics are extracted.

8. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 1, wherein, The method of identifying acoustic impedance abrupt change mechanisms by combining phase difference characteristics and amplitude attenuation characteristics, and generating internal leakage diagnostic results, includes: If the attenuation judgment flag is false, it is determined that no abnormal energy dissipation has occurred in the system, and an internal leakage diagnosis result is generated confirming that the internal sealing of the hydraulic cylinder is healthy. When the attenuation determination flag is true, the phase difference characteristics are analyzed to determine whether a physical phase reversal has occurred, and a phase reversal flag is generated. By combining attenuation judgment flags and phase reversal flags, the acoustic impedance mutation mechanism is identified, and internal leakage diagnosis results are generated.

9. The transient pressure wave based hydraulic cylinder internal leak detection method of claim 8, wherein, The method of identifying acoustic impedance abrupt change mechanisms by combining attenuation determination flags and phase reversal flags to generate internal leakage diagnostic results includes: Extract the attenuation determination flag and the phase reversal flag, and logically combine them to generate a feature combination vector; When the feature combination vector indicates that there is amplitude attenuation and physical phase reversal, it is determined that there is a local acoustic impedance mutation, and an internal leakage diagnosis result is generated, confirming that there is a physical leakage in the actual sealing orifice of the hydraulic cylinder piston. When the characteristic combination vector indicates that there is amplitude attenuation and no physical phase reversal occurs, it is determined that there is false attenuation of two-phase flow, and an internal leakage diagnosis result is generated that automatically filters out bubble interference and does not trigger an alarm.

10. A hydraulic cylinder internal leak detection system based on transient pressure waves, characterized in that, include: The global hardware trigger signal generation module collects the control signal of the electromagnetic directional valve in the hydraulic circuit, performs edge detection, extracts the falling edge level, and generates a global hardware trigger signal. The native water hammer pressure pulse signal generation module responds to the global hardware trigger signal to start the high-frequency dynamic pressure sensor to collect fluid pressure data at the hydraulic cylinder inlet pipeline. By identifying the high pressure jump edge in the pressure sequence that exceeds the preset differential steepness threshold, the absolute zero point time of the waveform transmission is established, and the native water hammer pressure pulse signal is generated. The dynamic physical cavity length calculation module triggers and reads the displacement sensor data attached to the piston rod of the hydraulic cylinder at the instant the absolute zero point is established. It obtains the absolute physical coordinates and performs algebraic calculations in combination with the fixed total length of the oil cylinder to calculate the dynamic physical cavity length. The theoretical echo arrival time generation module obtains the physical sound velocity of hydraulic oil at the current temperature, and uses the physical sound velocity and dynamic physical cavity length to perform kinematic calculations to generate the theoretical echo arrival time. The target echo signal extraction module generates a hardware-gated time window based on the theoretical echo arrival time, and uses the hardware-gated time window to dynamically capture the original water hammer pressure pulse signal to extract the target echo signal. The feature extraction module compares and analyzes the target echo signal with the original water hammer pressure pulse signal by inputting them into a hardware zero-crossing detector to extract phase difference features and amplitude attenuation features. The internal leakage diagnosis result generation module combines phase difference characteristics and amplitude attenuation characteristics to identify the acoustic impedance change mechanism and generate internal leakage diagnosis results.

Images