Data processing method and system for vibration state of a rotary engine

Abstract

The application provides a rotor engine vibration state data processing method and system, and relates to the technical field of data processing.The method comprises the following steps: collecting real-time rotating speed data sets of a single-cylinder rotor aero-engine under multiple different load conditions and instantaneous torque measurement value data sets synchronously measured with each rotating speed data point; based on the rotating speed data sets and the torque measurement value data sets, performing three-dimensional space mapping processing on rotating speed dimension, torque dimension and time dimension data to generate a three-dimensional torque data space distribution model representing torque distribution with rotating speed and time; based on the three-dimensional torque data space distribution model, extracting a polygon vertex sequence composed of discrete measurement data points on an engine external characteristic curve, and obtaining a continuous and derivable torque-rotating speed relationship smooth curve mathematical model by performing smooth fitting on the polygon vertex sequence.The application realizes precise monitoring of engine vibration state and torque performance.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data processing method and system for the vibration state of a rotary engine. Background Technology

[0002] Rotary engines, with their advantages of compact structure and high power density, are widely used in the field of light aviation equipment. Their vibration status is directly related to operational safety and reliability. Accurate data processing is the core prerequisite for vibration monitoring and fault early warning, but there are still significant deficiencies in the relevant data processing technology.

[0003] For example, during a high-altitude test flight of a certain light aircraft single-cylinder rotary engine, the traditional data processing method only collected vibration signals under a single load condition, without synchronously linking parameters such as speed, torque, and real-time atmospheric pressure. When the engine switched between idle and maximum power conditions, the vibration amplitude fluctuated abnormally. Because the data processing did not achieve multi-dimensional data fusion and environmental compensation, the staff misjudged it as a rotor wear fault. After shutdown and inspection, it was found that there was no mechanical damage, and the vibration abnormality was caused by environmental factors and the switching of operating conditions. This has the defects of incomplete operating condition coverage, asynchronous multi-dimensional data, and lack of environmental parameter correction mechanism, which can easily lead to misjudgment of vibration status and delayed fault warning. Summary of the Invention

[0004] This invention provides a data processing method and system for the vibration state of a rotary engine, enabling accurate monitoring of the engine's vibration state and torque performance.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a data processing method for the vibration state of a rotary engine, the method comprising: Step 1: Collect real-time speed data sets of a single-cylinder rotary aero-engine under multiple different load conditions and instantaneous torque measurement data sets measured synchronously with each speed data point; Step 2: Based on the speed dataset and torque measurement dataset, perform three-dimensional spatial mapping processing on the speed dimension, torque dimension and time dimension data to generate a three-dimensional torque data spatial distribution model that represents the torque distribution with speed and time. Step 3: Based on the three-dimensional torque data spatial distribution model, extract the polygon vertex sequence composed of discrete measurement data points on the engine external characteristic curve. By smoothing the polygon vertex sequence, a continuous and differentiable torque-speed relationship smooth curve mathematical model is obtained. Step 4: Based on the mathematical model of the torque-speed relationship smooth curve, input the real-time monitored speed parameters of the single-cylinder rotor aero-engine, and obtain the standard torque reference value data of the real-time speed through model interpolation calculation. Step 5: Compensate the standard torque reference value data with environmental parameters by introducing real-time atmospheric pressure and intake air temperature environmental monitoring data, and correct the standard torque reference value by using a preset environmental influence coefficient to obtain the torque test result data after environmental correction. Step 6: Based on the torque detection results data after environmental correction, and combined with the engine's fuel consumption rate benchmark data under this operating condition, perform cross-validation and error calibration of the torque value to obtain the final verified and calibrated torque detection value data.

[0006] Furthermore, real-time speed datasets of a single-cylinder rotary aero-engine under multiple different load conditions and instantaneous torque measurement datasets measured synchronously with each speed data point are collected, including: Multiple load operating points are set, covering the engine speed range from idle to maximum power, and including at least the idle point, the maximum torque speed point, and the maximum power speed point. At each load operating point, speed and torque signals are collected in real time at a preset sampling frequency by a speed sensor and a torque sensor installed on the engine output shaft, respectively. A synchronization trigger is used to ensure that each speed data point and torque data point have the same timestamp, thereby obtaining a synchronized real-time speed dataset and an instantaneous torque measurement dataset.

[0007] Furthermore, based on the speed dataset and torque measurement dataset, the speed, torque, and time dimensions of the data are mapped in three dimensions to generate a three-dimensional torque data spatial distribution model representing the torque distribution with speed and time, including: Using the rotational speed, torque, and time values ​​as the coordinate axes of a three-dimensional coordinate system, each synchronously measured data point is mapped to a discrete point in three-dimensional space, thus forming a three-dimensional scattered dataset. Based on a 3D scattered dataset, a spherical neighborhood radius is preset, and a spherical neighborhood is constructed with any point to be interpolated in 3D space as the center. All discrete points falling into the spherical neighborhood are searched. Based on the discrete points falling within the spherical neighborhood and their distance from the center point, the torque representative value of the interpolation point is calculated using a distance-weighted average algorithm, where discrete points that are closer to the center point are given higher weights. Based on the distance-weighted average algorithm, the torque representative value is calculated sequentially for each grid node in the three-dimensional space after being divided into grids according to a preset step size, thereby generating a three-dimensional gridded data field composed of grid nodes and torque representative values. The three-dimensional gridded data field is used as a three-dimensional torque data spatial distribution model that characterizes the continuous distribution of torque with rotational speed and time.

[0008] Furthermore, based on the three-dimensional torque data spatial distribution model, a sequence of polygonal vertices composed of discrete measurement data points is extracted from the engine external characteristic curve. By smoothing the polygonal vertex sequence, a continuously differentiable mathematical model of the torque-speed relationship smooth curve is obtained, including: Based on the three-dimensional torque data spatial distribution model, the data is sliced ​​at equal intervals along the speed axis with a preset speed step size to obtain multiple two-dimensional torque-time slices perpendicular to the speed axis. For each two-dimensional torque-time slice, the torque representative values ​​of all grid nodes in the slice are traversed, and the maximum value is extracted as the external characteristic torque representative value corresponding to the speed point, thereby obtaining a series of discrete data points composed of speed values ​​and corresponding external characteristic torque representative values. The obtained series of discrete data points are connected sequentially in order of increasing rotational speed to form a polygon vertex sequence that represents the preliminary broken line shape of the engine's external characteristics. Based on the polygon vertex sequence, by setting the order of the curve and adjusting the position of the control points, the generated Bézier curve approximates the polygon vertex sequence under the least squares error criterion, while ensuring that the curve has a continuous first derivative at each vertex, thus obtaining a smooth and continuously differentiable torque-speed relationship curve. The torque-speed relationship curve is stored in the form of a mathematical expression, serving as a mathematical model for the smoothing curve of the torque-speed relationship.

[0009] Furthermore, based on the mathematical model of the torque-speed relationship smooth curve, the real-time monitored speed parameters of the single-cylinder rotary aero engine are input, and the standard torque reference value data of the real-time speed is obtained through model interpolation calculation, including: Based on the mathematical model of the smooth curve of torque-speed relationship, the position range of the real-time monitored speed parameter on the torque-speed relationship curve is determined, and two known speed nodes adjacent to the real-time monitored speed parameter are identified. Based on two known speed nodes and their corresponding representative torque values, linear interpolation is used to calculate the torque value corresponding to the real-time monitored speed parameters; the calculated torque value is used as the standard torque reference value data under the real-time monitored speed.

[0010] Furthermore, environmental parameter compensation is applied to the standard torque reference value data. Real-time atmospheric pressure and intake air temperature monitoring data are introduced, and the standard torque reference value is corrected using a preset environmental impact coefficient to obtain environmentally corrected torque test results, including: Real-time atmospheric pressure monitoring data and real-time intake air temperature monitoring data of a single-cylinder rotary aero-engine operating site are acquired, and the real-time atmospheric pressure monitoring data and the real-time intake air temperature monitoring data are used as input parameters for environmental compensation. Based on the real-time atmospheric pressure monitoring data and the preset standard atmospheric pressure value, the atmospheric pressure correction ratio is calculated, and the intake temperature correction ratio is calculated based on the real-time intake temperature monitoring data and the preset standard intake temperature value. The atmospheric pressure correction coefficient is obtained by performing a power exponent calculation based on the atmospheric pressure correction ratio and the preset atmospheric pressure influence index, and the intake temperature correction coefficient is obtained by performing a power exponent calculation based on the intake temperature correction ratio and the preset intake temperature influence index. The atmospheric pressure correction factor and the intake air temperature correction factor are multiplied together to generate the comprehensive environmental correction factor. Based on the standard torque reference value data, the standard torque reference value data is multiplied together with the comprehensive environmental correction factor to obtain the environmentally corrected torque test result data.

[0011] Furthermore, based on the environmentally corrected torque detection results and combined with the engine's fuel consumption rate benchmark data under operating conditions, cross-validation and error calibration of the torque values ​​are performed to obtain the final validated and calibrated torque detection values, including: The real-time fuel consumption rate monitoring data of a single-cylinder rotary aero engine under the current operating conditions is obtained, and the real-time fuel consumption rate monitoring data is used as the input parameter for cross-validation; based on the environmentally corrected torque detection results and the current speed parameters, the effective power value of the engine under the current operating conditions is calculated. Based on the effective power value and real-time fuel consumption rate monitoring data, the torque back-calculated value based on fuel consumption is obtained according to the energy conversion relationship. The torque detection results after environmental correction are compared with the torque back-calculated value based on fuel consumption to determine whether the deviation exceeds the preset error allowable threshold. If the deviation does not exceed the preset error allowable threshold, the torque detection result data after environmental correction will be directly used as the final verification calibration torque detection value data output. If the deviation exceeds the preset error allowable threshold, the torque detection result data after environmental correction and the torque back-calculated value based on fuel consumption will be fused and calibrated according to the magnitude and direction of the deviation. The fused and calibrated value will be used as the final torque detection value data for verification calibration.

[0012] Secondly, the data processing system for the vibration state of the rotary engine includes: The acquisition module is used to collect real-time speed data sets of a single-cylinder rotary aero-engine under multiple different load conditions, as well as instantaneous torque measurement data sets measured synchronously with each speed data point. The mapping module is used to perform three-dimensional spatial mapping processing on the speed, torque and time dimensions of the speed dataset and the torque measurement dataset to generate a three-dimensional torque data spatial distribution model that represents the torque distribution with speed and time. The fitting module is used to extract the sequence of polygon vertices composed of discrete measurement data points on the engine external characteristic curve based on the three-dimensional torque data spatial distribution model. By smoothly fitting the sequence of polygon vertices, a continuous and differentiable mathematical model of the torque-speed relationship smooth curve is obtained. The calculation module is used to input the real-time monitored speed parameters of a single-cylinder rotor aero engine based on the mathematical model of the torque-speed relationship smooth curve, and obtain the standard torque reference value data of the real-time speed through model interpolation calculation. The correction module is used to compensate for environmental parameters in the standard torque reference value data. It introduces real-time atmospheric pressure and intake air temperature environmental monitoring data, corrects the standard torque reference value through a preset environmental influence coefficient, and obtains the torque detection result data after environmental correction. The processing module is used to perform cross-validation and error calibration of torque values ​​based on the environmentally corrected torque detection results data and the engine's fuel consumption rate benchmark data under this operating condition, so as to obtain the final verified and calibrated torque detection value data.

[0013] Thirdly, a computing device, comprising: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0014] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0015] The above-described solution of the present invention has at least the following beneficial effects: This invention employs techniques such as synchronously acquiring speed and torque data under multiple load conditions, performing three-dimensional spatial mapping modeling of speed, torque, and time-dimensional data, smoothing and fitting the discrete vertex sequence of engine external characteristic curves, introducing atmospheric pressure and intake air temperature for environmental parameter compensation, and combining fuel consumption rate for torque value cross-validation and error calibration. Therefore, it effectively overcomes the technical problems existing in current rotary engine vibration state data processing, such as incomplete operating condition coverage, asynchronous multi-dimensional data, lack of environmental parameter correction mechanisms, vibration state misjudgment due to lack of cross-validation, delayed fault warning, and low torque detection accuracy. This achieves the technical effect of accurately monitoring engine vibration state and torque performance, improving the reliability and accuracy of rotary engine operating state assessment, avoiding misjudgments of vibration anomalies, reducing ineffective maintenance, and meeting the technical requirements for safe operation and state monitoring of rotary engines used in light aviation equipment under complex operating conditions. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of a data processing method for the vibration state of a rotary engine provided in an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram of a data processing system for the vibration state of a rotary engine provided in an embodiment of the present invention. Detailed Implementation

[0018] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art.

[0019] like Figure 1 As shown, an embodiment of the present invention proposes a data processing method for the vibration state of a rotary engine, the method comprising the following steps: Step 1: Collect real-time speed data sets of a single-cylinder rotary aero-engine under multiple different load conditions and instantaneous torque measurement data sets measured synchronously with each speed data point; Step 2: Based on the speed dataset and torque measurement dataset, perform three-dimensional spatial mapping processing on the speed dimension, torque dimension and time dimension data to generate a three-dimensional torque data spatial distribution model that represents the torque distribution with speed and time. Step 3: Based on the three-dimensional torque data spatial distribution model, extract the polygon vertex sequence composed of discrete measurement data points on the engine external characteristic curve. By smoothing the polygon vertex sequence, a continuous and differentiable torque-speed relationship smooth curve mathematical model is obtained. Step 4: Based on the mathematical model of the torque-speed relationship smooth curve, input the real-time monitored speed parameters of the single-cylinder rotor aero-engine, and obtain the standard torque reference value data of the real-time speed through model interpolation calculation. Step 5: Compensate the standard torque reference value data with environmental parameters by introducing real-time atmospheric pressure and intake air temperature environmental monitoring data, and correct the standard torque reference value by using a preset environmental influence coefficient to obtain the torque test result data after environmental correction. Step 6: Based on the torque detection results data after environmental correction, and combined with the engine's fuel consumption rate benchmark data under this operating condition, perform cross-validation and error calibration of the torque value to obtain the final verified and calibrated torque detection value data.

[0020] In this embodiment of the invention, the following technical means are employed: synchronously acquiring real-time speed data and instantaneous torque measurement data of a single-cylinder rotary aero-engine under multiple different load conditions; mapping the speed, torque, and time dimension data into a three-dimensional space to generate a three-dimensional torque data spatial distribution model; smoothly fitting the discrete polygon vertex sequence of the engine's external characteristic curve to obtain a continuously differentiable torque-speed relationship smooth curve mathematical model; calculating the standard torque reference value corresponding to the real-time speed through model interpolation; introducing real-time atmospheric pressure and intake air temperature for environmental parameter compensation and correction; and combining fuel consumption rate benchmark data for torque value cross-validation and error calibration. Therefore, this approach overcomes the technical problems in existing rotary engine vibration state data processing, such as incomplete operating condition coverage, asynchronous multi-dimensional data, lack of scientific three-dimensional data modeling and curve fitting methods, absence of environmental parameter correction mechanisms and cross-validation steps, leading to vibration state misjudgment, low torque detection accuracy, and inaccurate fault warnings. This achieves the technical effect of accurately monitoring the vibration state and torque performance of a single-cylinder rotary aero-engine, improving the stability and reliability of data processing, providing accurate data support for engine vibration state assessment and fault warning, and ensuring the safe and stable operation of the engine under complex operating conditions.

[0021] In a preferred embodiment of the present invention, step 1 above may include: Step 1.1: Set multiple load operating points, covering the engine's speed range from idle to maximum power, including at least the idle point, the maximum torque speed point, and the maximum power speed point. At each load operating point, speed and torque signals are collected in real time at a preset sampling frequency using speed and torque sensors mounted on the engine output shaft. A synchronization trigger ensures that each speed data point and torque data point has the same timestamp, thereby obtaining a synchronized real-time speed dataset and instantaneous torque measurement dataset. Specifically, this includes: firstly, comprehensively planning the operating conditions of the single-cylinder rotary aero engine, setting multiple load operating points, all of which cover the engine's complete speed range from idle to maximum power. The engine's idle speed is limited to a lower limit of 600 rpm and its maximum power speed is limited to an upper limit of 5500 rpm, thus fully covering the speed range of 600 rpm to 5500 rpm. The load operating points set in this study must include the idle speed point, the maximum torque speed point, and the maximum power speed point. Intermediate transition operating points are also added to ensure data comprehensiveness. The specific operating point values ​​are: idle speed point 600 rpm, transition operating point 1500 rpm, transition operating point 2500 rpm, maximum torque speed point 3500 rpm, transition operating point 4500 rpm, and maximum power speed point 5500 rpm, for a total of 6 load operating points, fully covering the speed range from low to high.

[0022] A speed sensor and a torque sensor are fixedly installed coaxially on the engine output shaft, ensuring that the detection centers of both sensors coincide with the axis of the engine output shaft to avoid signal errors caused by installation misalignment. The real-time sampling frequency of both the speed sensor and the torque sensor is set to 1000Hz, meaning 1000 sets of speed and torque signals are collected synchronously per second. The continuous signal acquisition duration for a single load condition point is set to 30 seconds, and the total number of data points collected for a single load condition point is calculated using the following formula: In the formula: This refers to the number of data points collected under a single operating condition. This is the sensor sampling frequency, with a value of 1000Hz. The acquisition time for a single working condition is set to 30 seconds. Substituting the value into the formula, we can calculate: N = 1000 × 30 = 30000. That is, 30,000 sets of speed signals and corresponding torque signals can be stably acquired at each load working condition point.

[0023] Synchronous triggers are integrated into the acquisition circuits of the speed and torque sensors. The synchronization accuracy of the timestamps of the synchronous triggers is set to ≤1ms to ensure the accuracy of timing calibration. The acquisition timing is uniformly controlled by the synchronous triggers. After each speed data point is acquired, the synchronous trigger immediately sends a trigger signal to control the torque sensor to acquire the instantaneous torque data at the same moment, ensuring that each speed data point and torque data point has the exact same timestamp and no time offset error. All speed signals acquired under 6 load conditions are integrated to form a complete real-time speed dataset. All torque signals that correspond exactly to the speed data timestamps are integrated to form an instantaneous torque measurement dataset. Finally, two sets of basic datasets with full operating condition coverage and complete timing synchronization are obtained.

[0024] In this embodiment of the invention, multiple load condition points are set to cover the engine's speed range from idle to maximum power, including at least the idle point, the maximum torque speed point, and the maximum power speed point. Simultaneously, speed and torque signals are collected in real time at a preset sampling frequency using speed and torque sensors mounted on the engine output shaft. A synchronous trigger is used to ensure that each speed data point and torque data point has the same timestamp to obtain a synchronized dataset. This overcomes the technical problems in existing technologies, such as incomplete load condition coverage and asynchronous speed and torque data acquisition, which lead to a lack of comprehensive and accurate basic data for data processing and are prone to data deviation. Therefore, it provides comprehensive, synchronous, and accurate basic data for steps such as three-dimensional torque data space modeling and torque-speed curve fitting, ensuring the accuracy and reliability of the entire rotor engine vibration state data processing flow.

[0025] In a preferred embodiment of the present invention, step 2 above may include: Step 2.1: Using the speed, torque, and time values ​​as the coordinate axes of a three-dimensional coordinate system, each synchronously measured data point is mapped to a discrete point in three-dimensional space, thus forming a three-dimensional scattered dataset. Specifically, after completing multi-condition synchronous data acquisition, based on the acquired synchronous speed dataset and instantaneous torque measurement dataset, and matching the acquisition time information corresponding to each data set, a dedicated three-dimensional Cartesian coordinate system is built to achieve spatial fusion and visualization mapping of multi-dimensional data. The speed value is defined as the X-axis of the three-dimensional Cartesian coordinate system. Combined with the actual engine operating range, the X-axis speed value range is set to 600 r / min to 550 r / min. 0 r / min; The torque value is defined as the Y-axis of the three-dimensional Cartesian coordinate system. Combined with the rated output performance of this type of single-cylinder rotor aero-engine, the torque value range of the Y-axis is set to 0 N·m to 80 N·m; The acquisition time value is defined as the Z-axis of the three-dimensional Cartesian coordinate system. Combined with the acquisition duration of 30 seconds for a single working condition, the time value range of the Z-axis is set to 0 s to 30 s. The speed data, torque data and corresponding acquisition time of each group of synchronously acquired data are combined into an independent three-dimensional spatial coordinate point. The coordinates are mapped one by one to the constructed three-dimensional coordinate system according to the coordinate correspondence. After all the mapped three-dimensional spatial coordinate points are summarized, a complete three-dimensional scattered dataset is formed.

[0026] Step 2.2: Based on the 3D scattered data set, a spherical neighborhood radius is preset, and a spherical neighborhood is constructed centered on any point to be interpolated in 3D space. All discrete points falling within the spherical neighborhood are searched. Specifically, this includes: Based on the formed 3D scattered data set, to achieve the conversion of discrete data to continuous data, relevant parameters of the spherical neighborhood are preset. In this embodiment, the radius of the spherical neighborhood is fixed at 50 3D coordinate units. This radius value has been verified through multiple experiments to ensure that the spherical neighborhood of each point to be interpolated contains a sufficient number of valid scattered points, avoiding calculation deviations due to insufficient samples, and preventing data interference caused by introducing distant invalid scattered points due to an excessively large selection range. Any point to be interpolated in 3D space is selected as the geometric center point of the spherical neighborhood. Using the 3D space Euclidean distance calculation formula, the spatial distance between the center point and all scattered points in the 3D scattered data set is calculated point by point. The distance value is used to determine whether the scattered points fall within the spherical neighborhood. The formula for calculating the spatial distance between the center point and scattered points in 3D space is: In the formula: This represents the three-dimensional spatial distance between the center point and the scattered points. , , These are the rotational speed, torque, and time coordinates of the center point to be interpolated; , , The rotational speed, torque, and time coordinates of the three-dimensional scattered points to be judged are respectively. After the distance calculation of all scattered points is completed, the distance values ​​are judged. When the calculated spatial distance d is less than or equal to the preset spherical neighborhood radius of 50, the scattered point is judged to fall into the spherical neighborhood of the current interpolation point and is included in the valid calculation sample. Finally, the screening of all discrete points in the spherical neighborhood is completed.

[0027] Step 2.3: Based on the discrete points falling within the spherical neighborhood and their distances to the center point, calculate the representative torque value of the interpolation point using a distance-weighted average algorithm. The closer the discrete points are, the higher their weights are assigned. Specifically, for the selected discrete points within the spherical neighborhood, calculate the representative torque value of the center point using a distance-weighted average algorithm. This algorithm follows the principle that closer discrete points have a greater influence on the center point's value, assigning higher weight coefficients to effectively reduce interference from distant scattered points and improve the stability and accuracy of the interpolation results. First, calculate the weight value corresponding to each discrete point within the spherical neighborhood. The weight calculation formula is: In the formula: For the spherical neighborhood, the first Weight values ​​for each discrete point; For the first The three-dimensional spatial distance between each discrete point and the center point to be interpolated is calculated. After calculating the weight of each discrete point, the representative torque value of the center point to be interpolated is obtained through a weighted average operation. The specific calculation formula is as follows: In the formula: The final representative value of the torque at the center point to be interpolated; This represents the total number of discrete points obtained by filtering within the spherical neighborhood. For the spherical neighborhood, the first By using the weighted calculation process described above, the actual torque measurement values ​​at each discrete point can effectively eliminate random fluctuations and noise interference at a single discrete measurement point, resulting in smooth and stable torque values ​​at the interpolation point.

[0028] Step 2.4: Based on the distance-weighted average algorithm, the torque representative value is calculated sequentially for each grid node in the three-dimensional space after being gridded according to a preset step size. This generates a three-dimensional gridded data field composed of grid nodes and torque representative values. The three-dimensional gridded data field is used as a three-dimensional torque data space distribution model representing the continuous distribution of torque with speed and time. Specifically, to construct a continuously distributed three-dimensional data model, the constructed three-dimensional space of speed, torque, and time is regularly gridded. In this embodiment, a uniform grid division step size is preset, where the grid step size for the speed axis is set to 100 r / min, the grid step size for the torque axis is set to 1 N·m, and the grid step size for the time axis is set to 1 s. This set of step size parameters can ensure the model's accuracy. While reducing redundant computation and balancing data processing efficiency with modeling effectiveness, the entire three-dimensional space is divided into a uniform and regular three-dimensional grid structure according to the set step size parameters. The intersection of all grids is the center point to be interpolated. Each grid node is traversed sequentially from top to bottom and from left to right, and the spherical neighborhood discrete point screening and distance-weighted average calculation are repeatedly performed to obtain the accurate torque representative value corresponding to each grid node. The three-dimensional coordinate information of all grid nodes and the corresponding torque representative value are integrated to form a three-dimensional gridded data field composed of grid node coordinates and torque representative values. This three-dimensional gridded data field is a three-dimensional torque data spatial distribution model that fully represents the continuous distribution law of torque with speed and time.

[0029] The core of constructing a three-dimensional torque data spatial distribution model is to transform the obtained discrete three-dimensional scattered data into a three-dimensional gridded data field that can completely represent the continuous distribution law of torque with speed and time. The specific process is as follows: Based on a three-dimensional Cartesian coordinate system with speed X-axis (600 r / min ~ 5500 r / min), torque Y-axis (0 N·m ~ 80 N·m), and time Z-axis (0 s ~ 30 s), the core parameters for grid division are first determined. The step size for the speed axis is set to 100 r / min, the step size for the torque axis is 1 N·m, and the step size for the time axis is 1 s. The number of grid nodes for each axis is calculated to be 50, 81, and 31 respectively, totaling 125,550 grid nodes. Then, the grid is further divided using this step size... The three-dimensional space is divided into uniform three-dimensional grid cells, and the intersection of adjacent grid cells is used as the interpolation center point. Following a hierarchical traversal rule of fixing the time axis, then the rotational speed axis, and finally the torque axis, all grid nodes are traversed without omission. For each traversed grid node, effective scattered points within the neighborhood are selected using a spherical neighborhood radius of 50 three-dimensional coordinate units. The torque representative value of each node is calculated using a distance-weighted average algorithm. Finally, the three-dimensional coordinate information of all grid nodes is associated with their corresponding torque representative values ​​and stored uniformly as a structured dataset, forming a three-dimensional gridded data field, i.e., a three-dimensional torque data spatial distribution model, realizing the transformation of discrete data into a continuous distribution representation.

[0030] In this embodiment of the invention, the synchronously measured data points are mapped to three-dimensional discrete points using rotational speed, torque, and time values ​​as coordinate axes of a three-dimensional coordinate system, forming a three-dimensional scattered dataset. A spherical neighborhood radius is preset, and a spherical neighborhood is constructed with the point to be interpolated as the center. Discrete points within the neighborhood are searched, and the representative torque value of the point to be interpolated is calculated using a distance-weighted average algorithm. Then, the representative torque value is calculated sequentially for each grid node in the gridded three-dimensional space, generating a three-dimensional gridded data field as a three-dimensional torque data spatial distribution model. This overcomes the technical problem in the prior art that it is impossible to effectively integrate multi-dimensional data of rotational speed, torque, and time, and that discrete measurement data is difficult to characterize the continuous distribution law of torque with rotational speed and time, resulting in a lack of accurate and continuous model support for data processing and easy data deviation. Thus, it achieves the accurate construction of a three-dimensional model characterizing the continuous distribution of torque with rotational speed and time, effectively eliminating the volatility of discrete data.

[0031] In a preferred embodiment of the present invention, step 3 above may include: Step 3.1: Based on the three-dimensional torque data spatial distribution model, perform equidistant slicing along the speed coordinate axis with a preset speed step size to obtain multiple two-dimensional torque-time slices perpendicular to the speed axis. Specifically, based on the generated three-dimensional torque data spatial distribution model, to extract the torque variation law at different engine speeds, perform equidistant slicing along the speed coordinate axis X-axis. The slicing direction is perpendicular to the speed axis, ultimately obtaining multiple independent two-dimensional torque-time slices. Each slice corresponds to a fixed speed point, achieving hierarchical extraction of the speed dimension. The speed step size of the slice is consistent with the grid step size of the speed axis, set to 100 r / min, ensuring that the speed points after slicing completely correspond to the speeds of the grid nodes in the three-dimensional model, avoiding data disconnection. Combining the value range of the speed axis in the three-dimensional model (600 r / min ~ 5500 r / min), the total number of slices can be calculated using the following formula: : This represents the total number of slices. This represents the maximum rotational speed of the shaft, set at 5500 r / min. This is the minimum rotational speed of the shaft, taken as 600 r / min; The rotational speed slice step size is set to 100 r / min. Substituting the values ​​into the calculation, 50 two-dimensional torque-time slices perpendicular to the rotational speed axis are obtained. Each slice corresponds to a fixed rotational speed, namely 600 r / min, 700 r / min, 800 r / min, ..., 5500 r / min. Each slice contains all the grid nodes and corresponding torque representative values ​​for the time from 0 s to 30 s and the torque from 0 N·m to 80 N·m at that rotational speed.

[0032] Step 3.2: For each two-dimensional torque-time slice, iterate through the representative torque values ​​of all grid nodes within the slice, extracting the maximum value as the representative external characteristic torque value corresponding to that speed point. This yields a series of discrete data points composed of speed values ​​and their corresponding representative external characteristic torque values. Specifically, for each obtained two-dimensional torque-time slice, iterate through the grid nodes one by one, extracting the maximum value among the representative torque values ​​of all grid nodes within each slice. This maximum value is the representative external characteristic torque value corresponding to the corresponding speed point. The representative external characteristic torque value reflects the maximum torque that the engine can output at that speed and is the core data for constructing the engine's external characteristic curve. The number of grid nodes in each two-dimensional torque-time slice is the product of the number of time axis nodes and the number of torque axis nodes, calculated using the following formula: In the formula: The total number of grid nodes in a single slice; This represents the number of grid nodes on the time axis, with a value of 31. The value is 81, representing the number of mesh nodes for the torque axis. Substituting the numerical values, each slice contains 2511 mesh nodes. The torque representative values ​​of these 2511 mesh nodes within each slice are iterated and compared one by one, and the maximum value is selected as the external characteristic torque representative value for that speed. For example, in the 600 r / min idle speed slice, after iterating through all 2511 nodes, the maximum value is 8.2 N·m, which is the external characteristic torque representative value corresponding to 600 r / min. In the 3500 r / min maximum torque speed slice, the maximum torque representative value is 45.5 N·m, which is the external characteristic torque representative value corresponding to that speed. In the 5500 r / min maximum power speed slice, the maximum torque representative value is 38.3 N·m, which is the external characteristic torque representative value corresponding to that speed. The speed values ​​corresponding to all 50 speed points are mapped one-to-one with the representative values ​​of external characteristic torque to form a series of discrete data points. The format of each discrete data point is speed value followed by representative value of external characteristic torque, resulting in a total of 50 discrete data points that fully cover the external characteristic torque data of the engine across the entire speed range.

[0033] Step 3.3 involves connecting the obtained discrete data points sequentially according to their speed values ​​from smallest to largest, forming a polygonal vertex sequence representing the initial broken-line shape of the engine's external characteristics. Specifically, this includes sorting the 50 discrete data points in ascending order of their speed values, ranging from a minimum speed of 600 r / min to a maximum speed of 5500 r / min, ensuring the data points are arranged according to the increasing engine speed. After sorting, adjacent discrete data points are connected sequentially with straight lines to form a broken line. This broken line represents the polygon representing the initial shape of the engine's external characteristics. Each turning point of the broken line is a discrete data point, and these discrete data points together constitute the polygonal vertex sequence. Some specific nodes in the polygonal vertex sequence are as follows: (600 r / min, 8.2) The following parameters are listed: (700 r / min, 10.5 N·m), (1500 r / min, 22.3 N·m), (2500 r / min, 34.1 N·m), (3500 r / min, 45.5 N·m), (4500 r / min, 41.2 N·m), (5500 r / min, 38.3 N·m). The remaining 43 nodes are arranged in ascending order of speed, forming a complete polygonal vertex sequence. This polygonal vertex sequence initially reflects the trend of engine torque with speed, that is, from idle speed to maximum torque speed, torque gradually increases; from maximum torque speed to maximum power speed, torque gradually decreases. However, since this sequence is formed by connecting discrete points, the broken line has obvious inflection points and cannot reflect the continuous change law of torque with speed.

[0034] Step 3.4: Based on the polygon vertex sequence, by setting the order of the curve and adjusting the position of the control points, the generated Bézier curve approximates the polygon vertex sequence under the least squares error criterion, while ensuring that the curve has a continuous first derivative at each vertex, thus obtaining a smooth and continuously differentiable torque-speed relationship curve. Specifically, this includes: based on the polygon vertex sequence constructed in step 3.3, using a Bézier curve for smooth fitting, by setting the curve order and adjusting the control point position, ensuring that the fitted curve approximates the polygon vertex sequence under the least squares error criterion, while ensuring that the curve has a continuous first derivative at each vertex, achieving smoothness and continuous differentiability of the torque-speed relationship curve. In this embodiment, considering the number of polygon vertex sequences (50), the order of the Bézier curve is set to 5. A 5th-order Bézier curve has sufficient flexibility to accurately approximate the discrete vertex sequence, while avoiding curve distortion caused by excessively high orders. The fitting of the Bézier curve follows the least squares error criterion, and the formula for calculating the least squares error is: In the formula: This represents the fitting error; This represents the number of nodes in the polygon vertex sequence, with a value of 50. This represents the actual external characteristic torque value of the i-th vertex; For the Bézier curve in the th Each vertex corresponds to a fitted torque value at a given speed. During the fitting process, the positions of the control points on the Bézier curve are adjusted. A 5th-order Bézier curve has six control points, which helps to reduce the fitting error. To achieve the minimum value, a fitting error threshold of 0.1 N·m is set to ensure that the deviation between the fitted curve and the actual vertex sequence is within the allowable range. Simultaneously, to ensure the curve has a continuous first derivative at each vertex, the positions of adjacent control points are adjusted to make the tangent slope of the curve continuous at the vertex, avoiding inflection points and abrupt changes. This ensures the fitted curve is smooth, continuous, and differentiable, truly reflecting the continuous change of torque with speed. For example, at the vertex of the maximum torque speed of 3500 r / min, by adjusting the control point positions, the tangent slope of the curve to the left of this vertex is 0.008 N·m / (r / min), and the tangent slope to the right is also 0.008 N·m / (r / min), achieving continuous first derivative and ensuring a smooth transition of the curve at this vertex without abrupt changes.

[0035] Step 3.5: Store the torque-speed relationship curve in the form of a mathematical expression, serving as the mathematical model for the smooth torque-speed relationship curve. Specifically, this includes converting the fitted, smooth, and continuously differentiable torque-speed relationship curve into a mathematical expression and storing it in a structured format. This mathematical expression is the mathematical model for the smooth torque-speed relationship curve. The mathematical expression corresponding to the fitted 5th-order Bézier curve is in polynomial form, as detailed below: In the formula: Torque values ​​are measured in N·m. This represents the rotational speed (unit: r / min). , , , , , The fitting coefficients are calculated and their specific values ​​are as follows: =1.2×10 -12 , =-8.5×10 -9 , =2.3×10 -5 , =-0.028, =18.6, =-4200, store the mathematical expression and the corresponding fitting coefficients in a unified manner to form a complete mathematical model of the smooth curve of torque-speed relationship. This model can directly receive the engine's real-time monitoring speed parameters and calculate the corresponding torque value by substituting them in.

[0036] In this embodiment of the invention, the external characteristic polygon vertex sequence is constructed by slicing the three-dimensional torque data spatial distribution model at equal intervals along the rotational speed axis, extracting the maximum torque value of each slice, and then fitting the vertex sequence with a Bézier curve with continuous first derivative based on the least squares error criterion, ultimately forming a continuously differentiable torque. The technique of using a mathematical model to smooth the speed-relationship curve overcomes the shortcomings of traditional discrete data points in engine external characteristic curves, which are unsmooth, discontinuous, and non-differentiable, and cannot accurately reflect torque. The technical problem of accurately constructing a smooth, continuously differentiable torque curve stems from understanding the true variation of rotational speed. A mathematical model for rotational speed improves the accuracy of external characteristic curve representation.

[0037] In a preferred embodiment of the present invention, step 4 above may include: Step 4.1: Based on the mathematical model of the torque-speed relationship smooth curve, determine the position range of the real-time monitored speed parameter on the torque-speed relationship curve, and identify two known speed nodes adjacent to the real-time monitored speed parameter. Specifically, this includes: based on the mathematical model of the torque-speed relationship smooth curve, determining the position range of the real-time monitored speed in the model curve, and accurately identifying two known speed nodes adjacent to the real-time speed. The known speed nodes are the 50 discrete speed points extracted in Step 3.2: 600 r / min, 700 r / min, 800 r / min, ..., 5500 r / min. The nodes are evenly distributed across the entire speed range of 600 r / min to 5500 r / min, with intervals of 100 r. / min, each known speed node corresponds to a specific external characteristic torque value, forming a complete speed-torque reference table; the specific process of interval judgment is to compare the real-time monitored speed with all known speed nodes one by one, find two known speed nodes, one node whose speed value is less than the real-time monitored speed, and is the closest to the real-time speed among all nodes less than the real-time speed, and is recorded as the lower limit node; the other node whose speed value is greater than the real-time monitored speed, and is the closest to the real-time speed among all nodes greater than the real-time speed, is recorded as the upper limit node. The interval between the two nodes is the position interval of the real-time monitored speed. Combining the three selected real-time monitored speed examples, the specific interval judgment and adjacent node determination process is as follows: When the real-time monitored speed is 1850 r / min, the 50 known speed nodes are compared one by one. The closest lower limit node is found to be less than 1850 r / min, which is 1800 r / min, and the closest upper limit node is found to be greater than 1850 r / min, which is 1900 r / min. Therefore, the position range of the real-time speed is 1800 r / min to 1900 r / min, and the adjacent known speed nodes are 1800 r / min and 1900 r / min. When the real-time monitored speed is 3200 r / min, the lower limit node is determined to be 3200 r / min. The closest lower limit node is 3100 r / min, the upper limit node is 3300 r / min, the position range is 3100 r / min to 3300 r / min, and the adjacent known speed nodes are 3100 r / min and 3300 r / min. When the real-time monitored speed is 5200 r / min, the lower limit node is determined to be 5200 r / min. The closest lower limit node corresponding to 5200 r / min is 5200 r / min, the upper limit node is 5300 r / min, the position range is 5200 r / min to 5300 r / min, and the adjacent known speed nodes are 5200 r / min and 5300 r / min.

[0038] Meanwhile, the representative values ​​of the external characteristic torque corresponding to each adjacent known speed node are extracted, and the specific correspondences are as follows: 1800 r / min corresponds to a torque of 25.4 N·m, 1900 r / min corresponds to a torque of 27.6 N·m; 3100 r / min corresponds to a torque of 41.5 N·m, 3300 r / min corresponds to a torque of 44.8 N·m; 5200 r / min corresponds to a torque of 38.8 N·m, 5300 r / min corresponds to a torque of 38.5 N·m.

[0039] Step 4.2: Based on two known speed nodes and their corresponding representative torque values, linear interpolation is used to calculate the torque value corresponding to the real-time monitored speed parameters. The calculated torque value is used as the standard torque reference value data under the real-time monitored speed. Specifically, based on two determined adjacent known speed nodes and their corresponding representative torque values, a linear interpolation algorithm is used to accurately calculate the torque value corresponding to the real-time monitored speed. This torque value is the standard torque reference value under the real-time monitored speed, solving the defects of large deviations and lack of accurate reference benchmarks in traditional technologies. The core logic of the linear interpolation algorithm is to assume that the torque changes linearly with the speed between two adjacent known speed nodes. A linear relationship is established through the speed and torque data of the two nodes, and then the torque value corresponding to the real-time speed is calculated. The linear interpolation calculation formula is: In the formula: The standard torque reference value for real-time monitoring of rotational speed is in N·m. The unit for real-time monitoring of rotational speed parameters is r / min; The speed value of the node with the lower limit known speed is in r / min; The torque value corresponding to the known speed node at the lower limit is represented by N·m. The speed value of the node with a known upper limit speed is in r / min; The torque value corresponding to the node with a known upper limit speed is represented by N·m. When real-time speed monitoring =1850 r / min, corresponding lower limit node =1800 r / min =25.4 N·m, upper limit node =1900r / min =27.6 N·m, substituting into the formula, the standard torque reference value corresponding to a real-time speed of 1850 r / min is 26.5 N·m; When real-time speed monitoring When the speed is 3200 r / min, the corresponding lower limit node is... =3100r / min =41.5 N·m, upper limit node =3300r / min =44.8 N·m, substituting into the formula, the standard torque reference value corresponding to the real-time speed of 3200 r / min is 43.15 N·m; When real-time speed monitoring At 5200 r / min, the corresponding lower limit node =5200r / min =38.8 N·m, upper limit node =5300r / min =38.5N・m, substituting into the formula, the standard torque reference value corresponding to the real-time speed of 5200r / min is 38.8N・m.

[0040] After the calculation is completed, the standard torque reference value corresponding to each real-time monitored speed is stored to form a real-time torque reference dataset.

[0041] In this embodiment of the invention, the input torque of the rotational speed will be monitored in real time. The mathematical model of the speed-speed relationship smooth curve determines the adjacent known speed nodes and calculates the corresponding torque value through linear interpolation to obtain the standard torque reference value data. Therefore, it overcomes the technical problems of lack of continuous smooth mathematical model support, low accuracy of torque calculation corresponding to real-time speed, and unreliable interpolation results, and thus achieves the goal of quickly and accurately obtaining the standard torque reference value at real-time speed.

[0042] In a preferred embodiment of the present invention, step 5 above may include: Step 5.1: Acquire real-time atmospheric pressure monitoring data and real-time intake air temperature monitoring data at the operating site of the single-cylinder rotary aero-engine. Use these two data as input parameters for environmental compensation. Specifically, this includes: acquiring real-time environmental parameters at the operating site of the single-cylinder rotary aero-engine; clarifying the input data for environmental compensation; and considering the influence of environmental factors such as atmospheric pressure in high-altitude flight test scenarios, selecting real-time atmospheric pressure and real-time intake air temperature as core environmental compensation parameters. These two parameters are key environmental factors affecting engine torque output, and their changes directly lead to deviations between the actual engine torque and the standard torque reference value. Dedicated environmental sensors are used to collect relevant data. The atmospheric pressure sensor is installed near the engine intake duct to monitor the atmospheric pressure at the operating site in real time, and the intake air temperature sensor is integrated... Located within the intake manifold, sensors are used to monitor engine intake air temperature in real time. Both sensors are set to a sampling frequency of 1000Hz, consistent with the speed sampling frequency in steps 1 and 4, ensuring time synchronization between environmental data and speed / torque data, and avoiding compensation deviations caused by inconsistent sampling frequencies. The environmental signals collected by the sensors are filtered and noise-reduced, then converted into directly calculable digital parameters. Simultaneously, preset standard environmental parameter values ​​are established to match the engine's rated operating environment. After multiple tests, the preset standard atmospheric pressure is determined to be 101.325 kPa, and the preset standard intake air temperature is 20℃ (293.15 K). These standard values ​​are the rated environmental parameters used in the engine design and serve as a benchmark for quantifying environmental deviations. Three sets of typical real-time environmental parameters are selected as examples, corresponding one-to-one with the three real-time speed examples in step 4, as follows: The corresponding real-time speed is 1850 r / min, the real-time atmospheric pressure is 90.0 kPa, and the real-time intake air temperature is 15℃; The corresponding real-time speed is 3200 r / min, the real-time atmospheric pressure is 85.0 kPa, and the real-time intake air temperature is 12℃; The corresponding real-time speed is 5200 r / min, the real-time atmospheric pressure is 80.0 kPa, and the real-time intake air temperature is 10℃.

[0043] The above three sets of real-time atmospheric pressure monitoring data and real-time intake air temperature monitoring data are used as input parameters for environmental compensation at the corresponding real-time speeds.

[0044] Step 5.2: Based on the real-time atmospheric pressure monitoring data and the preset standard atmospheric pressure value, calculate the atmospheric pressure correction ratio. Simultaneously, based on the real-time intake air temperature monitoring data and the preset standard intake air temperature value, calculate the intake air temperature correction ratio. Specifically, this includes dividing the real-time environmental parameter value by the corresponding preset standard environmental parameter value. The resulting ratio is used to characterize the degree of deviation of the real-time environment from the standard environment. A ratio greater than 1 indicates that the real-time environment is better than the standard environment, a ratio less than 1 indicates that the real-time environment is worse than the standard environment, and a ratio equal to 1 indicates that the real-time environment is consistent with the standard environment and no correction is required.

[0045] Formula for calculating atmospheric pressure correction ratio: In the formula: This is the atmospheric pressure correction ratio; The unit for real-time atmospheric pressure monitoring data is kPa; The preset standard atmospheric pressure value is in kPa, and the value is 101.325 kPa.

[0046] Intake air temperature correction ratio calculation formula: In the formula: This is the intake air temperature correction ratio; The real-time intake air temperature monitoring data is in °C. When calculating, it needs to be converted to thermodynamic temperature in K. The conversion method is: thermodynamic temperature = temperature in Celsius + 273.15. The preset standard intake air temperature value is in K, and the value is 293.15K, which is 20℃.

[0047] Step 5.3: Based on the atmospheric pressure correction ratio and the preset atmospheric pressure influence index, a power exponent calculation is performed to obtain the atmospheric pressure correction coefficient. Similarly, based on the intake air temperature correction ratio and the preset intake air temperature influence index, a power exponent calculation is performed to obtain the intake air temperature correction coefficient. Specifically, this includes: based on the calculated correction ratio and combined with the preset environmental influence index, the atmospheric pressure correction coefficient and intake air temperature correction coefficient are obtained through power exponent calculation. The quantified value of the environmental deviation is converted into a correction coefficient that can be directly used for torque compensation. The preset environmental influence index has been verified through multiple experiments and determined in conjunction with the performance parameters of this type of single-cylinder rotary aero-engine. The specific values ​​are: atmospheric pressure influence index 0.7, intake air temperature influence index 0.3. This index characterizes the degree of influence of the corresponding environmental parameters on the engine torque output. The larger the index, the more significant the influence, which conforms to the operating law of aero-engines. The influence of atmospheric pressure on torque is greater than that of intake air temperature. The specific calculation formula is as follows: Formula for calculating atmospheric pressure correction factor: In the formula: This is the atmospheric pressure correction factor; This is the atmospheric pressure correction ratio; The atmospheric pressure influence index is set to 0.7.

[0048] Intake air temperature correction factor calculation formula: In the formula: This is the intake air temperature correction factor; This is the intake air temperature correction ratio; The value is 0.3, which represents the influence of intake air temperature.

[0049] Combining the three sets of correction ratios, the correction coefficients obtained after exponential calculation are as follows: Corresponding real-time rotational speed 1850 r / min: Atmospheric pressure correction ratio The atmospheric pressure correction factor is approximately 0.888. ≈0.920; Intake air temperature correction ratio The calculation yields an intake air temperature correction factor of approximately 0.983. ≈0.995.

[0050] Corresponding real-time rotational speed 3200 r / min: Atmospheric pressure correction ratio The atmospheric pressure correction factor is approximately 0.839. ≈0.884; Intake air temperature correction ratio The calculation yields an intake air temperature correction factor of approximately 0.973. ≈0.992.

[0051] Corresponding real-time speed 5200 r / min: Atmospheric pressure correction ratio The atmospheric pressure correction factor is approximately 0.790. ≈0.848; Intake air temperature correction ratio The calculation yields an intake air temperature correction factor of approximately 0.966. ≈0.990.

[0052] The above correction coefficients are all less than 1, which is consistent with the conclusion that the real-time environment is inferior to the standard environment. This indicates that the standard torque reference value needs to be reduced by the correction coefficient to obtain torque detection results that fit the actual environment.

[0053] Step 5.4: Multiply the atmospheric pressure correction factor and the intake air temperature correction factor to generate the comprehensive environmental correction factor; based on the standard torque reference value data, multiply the standard torque reference value data with the comprehensive environmental correction factor to obtain the environmentally corrected torque detection result data. Specifically, this includes multiplying the atmospheric pressure correction factor and the intake air temperature correction factor to obtain the comprehensive environmental correction factor. This factor integrates the combined influence of the two environmental parameters on torque, ensuring that the compensation calculation fully eliminates environmental interference. The calculation formula is as follows: In the formula: This is a comprehensive environmental correction factor; This is the atmospheric pressure correction factor; This is the intake air temperature correction factor.

[0054] Multiplying the obtained standard torque reference value by the comprehensive environmental correction factor yields the torque test result that conforms to the actual operating environment. Calculation formula: In the formula: The torque test results are after environmental correction; This is the standard torque reference value; This is the comprehensive environmental correction factor.

[0055] Combining the standard torque reference value and the correction factor, the calculation results for the three sets of examples are as follows: Corresponding real-time speed of 1850 r / min: Standard torque reference value is 26.5 N·m; Comprehensive environmental correction factor. =0.920×0.995≈0.915; Torque detection result after environmental correction =26.5×0.915≈24.25N·m; Corresponding real-time speed of 3200 r / min: Standard torque reference value is 43.15 N·m; Comprehensive environmental correction factor. =0.884×0.992≈0.877; Torque detection result after environmental correction =43.15×0.877≈37.84N·m; Corresponding real-time speed of 5200 r / min: Standard torque reference value is 38.8 N·m; Comprehensive environmental correction factor. =0.848×0.990≈0.839; Torque detection result after environmental correction =38.8×0.839≈32.55N·m; After the calculation is completed, the environmentally corrected torque detection results corresponding to each group of real-time speeds are stored in a unified manner to form an environmentally corrected torque detection dataset.

[0056] In this embodiment of the invention, real-time atmospheric pressure and intake air temperature are used as environmental compensation parameters. The pressure and temperature correction coefficients are obtained by calculating the correction ratio and combining it with a preset influence index to perform power exponentiation. This generates a comprehensive environmental correction coefficient to correct the standard torque reference value. Therefore, this method overcomes the technical problem in the prior art that the influence of environmental factors on torque detection is not considered, which leads to the detection results being easily affected by atmospheric pressure and intake air temperature, resulting in large errors. This effectively eliminates the detection deviation caused by environmental factors and improves the accuracy and applicability of torque detection results under complex aviation conditions.

[0057] In a preferred embodiment of the present invention, step 6 above may include: Step 6.1: Obtain real-time fuel consumption rate monitoring data of the single-cylinder rotary aero engine under the current operating conditions, and use the real-time fuel consumption rate monitoring data as the input parameter for cross-validation; based on the environmentally corrected torque detection results and the current speed parameters, calculate the effective power value of the engine under the current operating conditions. Specifically, this includes: using a fuel consumption rate sensor to collect real-time data. This sensor is installed in the engine fuel supply line to monitor the fuel consumption of the engine per unit time in real time. The sampling frequency is set to 1000Hz, which is consistent with the sampling frequency of the speed and environmental parameters mentioned above, to ensure that the fuel consumption rate data is synchronized with the torque, speed, and environmental data in time sequence, and to avoid verification errors caused by time sequence deviations.

[0058] The fuel consumption rate signal collected by the sensor is filtered and noise-reduced before being converted into directly calculable digital parameters. The unit of real-time fuel consumption rate is g / (kW·h). Based on three sets of real-time speed examples, the corresponding real-time fuel consumption rates are as follows: The corresponding real-time speed is 1850 r / min, the environmentally corrected torque detection result is 24.25 N·m, and the real-time fuel consumption rate is 320 g / (kW·h). The corresponding real-time speed is 3200 r / min, the environmentally corrected torque measurement result is 37.84 N·m, and the real-time fuel consumption rate is 305 g / (kW·h). The corresponding real-time speed is 5200 r / min, the environmentally corrected torque is 32.55 N·m, and the real-time fuel consumption rate is 330 g / (kW·h).

[0059] The real-time fuel consumption rate monitoring data mentioned above is used as the input parameter for cross-validation. At the same time, based on the obtained environmentally corrected torque detection results and the current real-time speed parameters, the engine effective power value under the current operating conditions is calculated according to the inherent relationship between engine effective power, torque, and speed.

[0060] The formula for calculating the effective power of an engine is as follows: In the formula: This refers to the engine's effective power. The torque detection result after environmental correction obtained in step 5; The current real-time speed is 9550; 9550 is a fixed conversion factor used to convert the units of torque and speed into the units of effective power.

[0061] Based on the three examples, the effective power is calculated respectively. The specific process is as follows: Corresponding to a real-time speed of 1850 r / min: The torque detection result after environmental correction is 24.25 N·m. With a real-time speed of 1850 r / min, substituting into the formula for calculation: First calculate the numerator: 24.25 × 1850 = 44862.5; then divide the numerator by 9550: 44862.5 ÷ 9550 ≈ 4.70 kW; that is, the effective power of the engine under the current operating conditions is approximately 4.70 kW.

[0062] Corresponding to a real-time speed of 3200 r / min: The torque detection result after environmental correction is 37.84 N·m. With a real-time speed of 3200 r / min, substituting into the formula for calculation: First calculate the numerator: 37.84 × 3200 = 121088; then divide the numerator by 9550: 121088 ÷ 9550 ≈ 12.68 kW; that is, the effective power of the engine under the current operating conditions is approximately 12.68 kW.

[0063] Corresponding to a real-time speed of 5200 r / min: The torque detection result after environmental correction is 32.55 N·m. With a real-time speed of 5200 r / min, substituting into the formula for calculation: First calculate the numerator: 32.55 × 5200 = 169260; then divide the numerator by 9550: 169260 ÷ 9550 ≈ 17.72 kW; that is, the effective power of the engine under the current operating conditions is approximately 17.72 kW.

[0064] The calculated effective power value accurately reflects the actual output power of the engine under the current operating conditions.

[0065] Step 6.2: Based on the effective power value and real-time fuel consumption rate monitoring data, the torque back-calculation value based on fuel consumption is calculated according to the energy conversion relationship. Specifically, the core of Step 6.2 is to calculate the torque back-calculation value based on fuel consumption by combining the effective power value calculated in Step 6.1 with the real-time fuel consumption rate monitoring data and back-calculating according to the energy conversion relationship. This back-calculated value serves as the benchmark value for cross-validation, used to compare the accuracy of the torque detection results after environmental correction, and to ensure the reliability of the torque data.

[0066] In the energy conversion relationship of an engine, effective power is fixedly related to fuel consumption rate, torque, and speed. The core logic of calculating torque based on fuel consumption is: first, verify the rationality of the power through effective power and fuel consumption rate, and then, based on the relationship between power, torque, and speed, calculate the corresponding torque value in reverse. This calculated value is not affected by environmental factors and is only related to fuel consumption and power output, and can be used as an independent verification benchmark.

[0067] The formula for calculating torque based on fuel consumption is as follows: In the formula: This is the torque value calculated based on fuel consumption. The calculated effective engine power; The current real-time rotational speed is 9550; 9550 is a fixed conversion factor, consistent with the conversion factor in step 6.1, to ensure consistent calculation logic.

[0068] It should be noted that the real-time fuel consumption rate is used to verify the rationality of the effective power. Based on the performance parameters of this type of single-cylinder rotary aero engine, the matching range between effective power and fuel consumption rate is 280 to 350 g / (kW·h). The fuel consumption rates of the three sets of examples are all within this range, indicating that the effective power calculation is reasonable and can be used for torque back-calculation.

[0069] Combining the three sets of effective power and real-time speed, the torque back-calculation value based on fuel consumption is calculated respectively. The specific process is as follows: Corresponding to a real-time speed of 1850 r / min: the effective power is approximately 4.70 kW. Substituting into the formula, we first calculate the numerator: 4.70 × 9550 = 44885; then divide the numerator by 1850: 44885 ÷ 1850 ≈ 24.26 N·m; that is, the torque value calculated based on fuel consumption is approximately 24.26 N·m.

[0070] Corresponding to a real-time speed of 3200 r / min: the effective power is approximately 12.68 kW. Substituting into the formula at a real-time speed of 3200 r / min: first calculate the numerator: 12.68 × 9550 = 121094; then divide the numerator by 3200: 121094 ÷ 3200 ≈ 37.84 N·m; that is, the torque value calculated based on fuel consumption is approximately 37.84 N·m.

[0071] Corresponding to a real-time speed of 5200 r / min: the effective power is approximately 17.72 kW. Substituting into the formula at a real-time speed of 5200 r / min: first calculate the numerator: 17.72 × 9550 = 169226; then divide the numerator by 5200: 169226 ÷ 5200 ≈ 32.54 N·m; that is, the torque value calculated based on fuel consumption is approximately 32.54 N·m.

[0072] The calculated torque back-calculation value is very close to the obtained torque detection result after environmental correction, which initially indicates that the torque data after environmental correction has a certain degree of accuracy.

[0073] Step 6.3 involves comparing the environmentally corrected torque detection result with the torque back-calculated value based on fuel consumption to determine if the deviation exceeds a preset error tolerance threshold. Specifically, this includes comparing the environmentally corrected torque detection result with the obtained torque back-calculated value based on fuel consumption, calculating the deviation between the two, and then comparing it with the preset error tolerance threshold to determine if the current environmentally corrected torque detection result meets the accuracy requirements. This provides a basis for determining whether subsequent fusion calibration is needed. In this embodiment, the preset error tolerance threshold was determined through multiple experiments and in conjunction with the torque measurement accuracy requirements of this type of single-cylinder rotary aero-engine. The specific value is 0.5 Nm. This threshold represents the maximum permissible deviation of the torque detection result. If the deviation is within this threshold range, it indicates that the accuracy of the environmentally corrected torque detection result meets the standard and can be used directly; if it exceeds this threshold, it indicates that the torque data has a deviation and further fusion calibration is required. The specific logic of the deviation comparison is as follows: calculate the absolute difference between the torque detection result after environmental correction and the torque back-calculated value based on fuel consumption, that is, take the non-negative result after subtracting the two values, compare the absolute difference with the error allowable threshold, and determine whether it exceeds the threshold.

[0074] Based on the three sets of examples, the deviations were calculated and threshold judgments were made. The specific process is as follows: At a real-time speed of 1850 rpm, the torque detection result after environmental correction is 24.25 Nm, and the torque calculated based on fuel consumption is 24.26 Nm. The difference between the two values ​​is calculated and the non-negative result is taken, resulting in a deviation of 0.01 Nm. This deviation is compared with the preset error allowable threshold of 0.5 Nm. Since the deviation is less than the threshold, it is determined that the deviation does not exceed the preset error allowable threshold, and the accuracy of the torque detection result after environmental correction meets the standard. At a real-time speed of 3200 rpm, the torque detection result after environmental correction is 37.84 Nm. The torque calculated based on fuel consumption is also 37.84 Nm. The difference between the two values ​​is calculated and the non-negative result is taken, resulting in a deviation of 0.00 Nm. This deviation is compared with the preset error allowable threshold of 0.5 Nm. Since the deviation is less than the threshold, it is determined that the deviation does not exceed the preset error allowable threshold, and the accuracy of the torque detection result after environmental correction meets the standard. At a real-time speed of 5200 rpm, the torque detection result after environmental correction is 32.55 Nm, and the torque calculated based on fuel consumption is 32.54 Nm. The difference between the two values ​​is calculated and the non-negative result is taken, resulting in a deviation of 0.01 Nm. This deviation is compared with the preset error allowable threshold of 0.5 Nm. Since the deviation is less than the threshold, it is determined that the deviation does not exceed the preset error allowable threshold, and the accuracy of the torque detection result after environmental correction meets the standard.

[0075] Step 6.4: If the deviation does not exceed the preset error allowable threshold, the environmentally corrected torque detection result data is directly used as the final verification calibration torque detection value data output. Specifically, when the deviation does not exceed the preset error allowable threshold, it indicates that the environmentally corrected torque detection result is highly consistent with the torque back-calculated value based on fuel consumption. The data accuracy meets the requirements of vibration state analysis and fault early warning for this type of single-cylinder rotor aero-engine. Moreover, the environmentally corrected torque detection result has eliminated the interference of environmental factors and can truly reflect the torque output state under the actual operating conditions of the engine. Therefore, it can be directly used as the final torque detection value output.

[0076] Based on the three sets of examples, the deviations did not exceed the threshold. The specific output process is as follows: The corresponding real-time speed is 1850 r / min. The torque detection result after environmental correction is 24.25 N·m, with a deviation of 0.01 N·m, which does not exceed the threshold. Therefore, 24.25 N·m is directly output as the final torque detection value for verification and calibration. The corresponding real-time speed is 3200 r / min. The torque detection result after environmental correction is 37.84 N·m, with a deviation of 0.00 N·m, which does not exceed the threshold. Therefore, 37.84 N·m is directly output as the final torque detection value for verification and calibration. The corresponding real-time speed is 5200 r / min. The torque detection result after environmental correction is 32.55 N·m, with a deviation of 0.01 N·m, which does not exceed the threshold. Therefore, 32.55 N·m is directly output as the final torque detection value for verification and calibration.

[0077] The final torque detection values ​​are stored uniformly to form a final torque detection dataset, which has the required accuracy and closely matches actual working conditions.

[0078] Step 6.5: If the deviation exceeds the preset error allowable threshold, then based on the magnitude and direction of the deviation, the torque detection result data after environmental correction and the torque back-calculated value based on fuel consumption are fused and calibrated. The fused and calibrated value is used as the final verification calibration torque detection value data. Specifically, this includes: combining the reliability of the two torque data and assigning weights. The torque detection result after environmental correction has eliminated environmental interference and fits the actual operating environment, so the weight is assigned to 0.6; the torque back-calculated value based on fuel consumption is used as an independent cross-validation benchmark, with high data stability, so the weight is assigned to 0.4. The total weight is 1. This weight allocation has been verified through multiple experiments to ensure that the calibrated torque detection value not only fits the actual environment but also has high stability.

[0079] The calculation formula for fusion calibration is as follows: In the formula: The torque detection value is the final verification calibration value after fusion calibration; The obtained torque detection results after environmental correction; The weight of the environmentally corrected torque detection result is set to 0.6. This is the torque value calculated based on fuel consumption. The weight for the torque inverse calculation based on fuel consumption is set to 0.4.

[0080] In conjunction with an additional example where the increased deviation exceeds the threshold, at a real-time rotational speed of 2500 r / min, the specific fusion calibration process is as follows: The relevant data is as follows: the torque detection result after environmental correction is 33.8 N·m, the torque back-calculated value based on fuel consumption is 33.2 N·m, the deviation is 0.6 N·m, which exceeds the preset threshold of 0.5 N·m, and fusion calibration is required; Substituting into the fusion calibration formula: First, calculate the product of the environmentally corrected torque detection result and the corresponding weight: 33.8 × 0.6 = 20.28 N·m; then calculate the product of the torque back-calculated value based on fuel consumption and the corresponding weight: 33.2 × 0.4 = 13.28 N·m; finally, add the two products to obtain the final torque detection value after fusion calibration: 20.28 + 13.28 = 33.56 N·m; Deviation verification: The final torque detection value after fusion calibration is 33.56 N·m, which deviates from the torque detection result after environmental correction by 0.24 N·m and from the torque back-calculated value based on fuel consumption by 0.36 N·m. Both are less than the preset error allowable threshold of 0.5 N·m, indicating that the accuracy of the calibrated torque detection value meets the standard. Output: 33.56 N·m is output as the final verified and calibrated torque detection value at a real-time speed of 2500 r / min, and stored in the final torque detection dataset.

[0081] In this embodiment of the invention, by introducing real-time fuel consumption rate monitoring data for cross-validation, calculating the torque back-calculation value based on the relationship between effective power and energy conversion, comparing the difference with the environmentally corrected torque detection result, judging the error threshold, and performing fusion calibration of the two types of torque data when the deviation exceeds the limit, the technical means overcome the technical problems of low reliability of single data source detection, lack of cross-validation mechanism, and easy occurrence of data anomalies and misjudgments. Thus, it achieves the technical effect of improving the accuracy and reliability of the final torque detection value, providing reliable data support for accurate assessment of the vibration state of the rotary engine, and avoiding misjudgments of anomalies and faults.

[0082] like Figure 2 As shown, embodiments of the present invention also provide a data processing system for the vibration state of a rotary engine, comprising: The acquisition module is used to collect real-time speed data sets of a single-cylinder rotary aero-engine under multiple different load conditions, as well as instantaneous torque measurement data sets measured synchronously with each speed data point. The mapping module is used to perform three-dimensional spatial mapping processing on the speed, torque and time dimensions of the speed dataset and the torque measurement dataset to generate a three-dimensional torque data spatial distribution model that represents the torque distribution with speed and time. The fitting module is used to extract the sequence of polygon vertices composed of discrete measurement data points on the engine external characteristic curve based on the three-dimensional torque data spatial distribution model. By smoothly fitting the sequence of polygon vertices, a continuous and differentiable mathematical model of the torque-speed relationship smooth curve is obtained. The calculation module is used to input the real-time monitored speed parameters of a single-cylinder rotor aero engine based on the mathematical model of the torque-speed relationship smooth curve, and obtain the standard torque reference value data of the real-time speed through model interpolation calculation. The correction module is used to compensate for environmental parameters in the standard torque reference value data. It introduces real-time atmospheric pressure and intake air temperature environmental monitoring data, corrects the standard torque reference value through a preset environmental influence coefficient, and obtains the torque detection result data after environmental correction. The processing module is used to perform cross-validation and error calibration of torque values ​​based on the environmentally corrected torque detection results data and the engine's fuel consumption rate benchmark data under this operating condition, so as to obtain the final verified and calibrated torque detection value data.

[0083] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0084] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0085] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0086] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A data processing method for the vibration state of a rotary engine, characterized in that, The method includes: Collect real-time speed data sets of a single-cylinder rotary aero-engine under multiple different load conditions, as well as instantaneous torque measurement data sets measured synchronously with each speed data point; Based on the speed dataset and torque measurement dataset, the speed dimension, torque dimension and time dimension data are processed by three-dimensional spatial mapping to generate a three-dimensional torque data spatial distribution model that represents the torque distribution with speed and time. Based on the three-dimensional torque data spatial distribution model, a sequence of polygon vertices composed of discrete measurement data points is extracted from the engine external characteristic curve. By smoothing the sequence of polygon vertices, a mathematical model of a smooth torque-speed relationship curve with continuous and differentiable characteristics is obtained. Based on the mathematical model of the torque-speed relationship smooth curve, the real-time monitored speed parameters of the single-cylinder rotor aero-engine are input, and the standard torque reference value data of the real-time speed is obtained through model interpolation calculation. Environmental parameter compensation is performed on the standard torque reference value data. Real-time atmospheric pressure and intake air temperature environmental monitoring data are introduced. The standard torque reference value is corrected by a preset environmental influence coefficient to obtain the torque test result data after environmental correction. Based on the torque detection results after environmental correction, and combined with the engine's fuel consumption rate benchmark data under this operating condition, the torque value is cross-validated and the error is calibrated to obtain the final verified and calibrated torque detection value data.

2. The data processing method for the vibration state of a rotary engine according to claim 1, characterized in that, The system collects real-time speed datasets of a single-cylinder rotary aero-engine under multiple different load conditions, as well as instantaneous torque measurement datasets measured synchronously with each speed data point, including: Multiple load operating points are set, covering the engine speed range from idle to maximum power, and including at least the idle point, the maximum torque speed point, and the maximum power speed point. At each load operating point, speed and torque signals are collected in real time at a preset sampling frequency by a speed sensor and a torque sensor installed on the engine output shaft, respectively. A synchronization trigger is used to ensure that each speed data point and torque data point have the same timestamp, thereby obtaining a synchronized real-time speed dataset and an instantaneous torque measurement dataset.

3. The data processing method for the vibration state of a rotary engine according to claim 2, characterized in that, Based on the speed dataset and torque measurement dataset, the speed, torque, and time dimensions of the data are mapped in three dimensions to generate a three-dimensional torque data spatial distribution model representing the torque distribution with speed and time, including: Using the rotational speed, torque, and time values ​​as the coordinate axes of a three-dimensional coordinate system, each synchronously measured data point is mapped to a discrete point in three-dimensional space, thus forming a three-dimensional scattered dataset. Based on a 3D scattered dataset, a spherical neighborhood radius is preset, and a spherical neighborhood is constructed with any point to be interpolated in 3D space as the center. All discrete points falling into the spherical neighborhood are searched. Based on the discrete points falling within the spherical neighborhood and their distance from the center point, the torque representative value of the interpolation point is calculated using a distance-weighted average algorithm, where discrete points that are closer to each other are given higher weights. Based on the distance-weighted average algorithm, the torque representative value is calculated sequentially for each grid node in the three-dimensional space after being divided into grids according to a preset step size, thereby generating a three-dimensional gridded data field composed of grid nodes and torque representative values. The three-dimensional gridded data field is used as a three-dimensional torque data spatial distribution model that characterizes the continuous distribution of torque with rotational speed and time.

4. The data processing method for the vibration state of a rotary engine according to claim 3, characterized in that, Based on a three-dimensional torque data spatial distribution model, a sequence of polygonal vertices composed of discrete measurement data points is extracted from the engine's external characteristic curve. By smoothly fitting this polygonal vertex sequence, a continuously differentiable mathematical model of the torque-speed relationship smooth curve is obtained, including: Based on the three-dimensional torque data spatial distribution model, the data is sliced ​​at equal intervals along the speed axis with a preset speed step size to obtain multiple two-dimensional torque-time slices perpendicular to the speed axis. For each two-dimensional torque-time slice, the torque representative values ​​of all grid nodes in the slice are traversed, and the maximum value is extracted as the external characteristic torque representative value corresponding to the speed point, thereby obtaining a series of discrete data points composed of speed values ​​and corresponding external characteristic torque representative values. The obtained discrete data points are connected sequentially in order of increasing rotational speed to form a polygonal vertex sequence that represents the preliminary broken line shape of the engine's external characteristics. Based on the polygon vertex sequence, by setting the order of the curve and adjusting the position of the control points, the generated Bézier curve approximates the polygon vertex sequence under the least squares error criterion, while ensuring that the curve has a continuous first derivative at each vertex, thus obtaining a smooth and continuously differentiable torque-speed relationship curve. The torque-speed relationship curve is stored in the form of a mathematical expression, serving as a mathematical model for the smoothing curve of the torque-speed relationship.

5. The data processing method for the vibration state of a rotary engine according to claim 4, characterized in that, Based on the mathematical model of the torque-speed relationship smooth curve, the real-time monitored speed parameters of a single-cylinder rotary aero engine are input, and the standard torque reference value data for the real-time speed is obtained through model interpolation calculation, including: Based on the mathematical model of the smooth curve of torque-speed relationship, the position range of the real-time monitored speed parameter on the torque-speed relationship curve is determined, and two known speed nodes adjacent to the real-time monitored speed parameter are identified. Based on two known speed nodes and their corresponding representative torque values, linear interpolation is used to calculate the torque value corresponding to the real-time monitored speed parameters; the calculated torque value is used as the standard torque reference value data under the real-time monitored speed.

6. The data processing method for the vibration state of a rotary engine according to claim 5, characterized in that, Environmental parameter compensation is applied to the standard torque reference value data. Real-time atmospheric pressure and intake air temperature monitoring data are introduced, and the standard torque reference value is corrected using a preset environmental impact coefficient to obtain the environmentally corrected torque test results, including: Real-time atmospheric pressure monitoring data and real-time intake air temperature monitoring data of a single-cylinder rotary aero-engine operating site are acquired, and the real-time atmospheric pressure monitoring data and the real-time intake air temperature monitoring data are used as input parameters for environmental compensation. Based on the real-time atmospheric pressure monitoring data and the preset standard atmospheric pressure value, the atmospheric pressure correction ratio is calculated, and the intake temperature correction ratio is calculated based on the real-time intake temperature monitoring data and the preset standard intake temperature value. The atmospheric pressure correction coefficient is obtained by performing a power exponent calculation based on the atmospheric pressure correction ratio and the preset atmospheric pressure influence index, and the intake temperature correction coefficient is obtained by performing a power exponent calculation based on the intake temperature correction ratio and the preset intake temperature influence index. The atmospheric pressure correction factor and the intake air temperature correction factor are multiplied together to generate the comprehensive environmental correction factor. Based on the standard torque reference value data, the standard torque reference value data is multiplied together with the comprehensive environmental correction factor to obtain the environmentally corrected torque test result data.

7. The data processing method for the vibration state of a rotary engine according to claim 6, characterized in that, Based on the environmentally corrected torque detection data, combined with the engine's fuel consumption rate benchmark data under operating conditions, cross-validation and error calibration of the torque values ​​are performed to obtain the final validated and calibrated torque detection data, including: The real-time fuel consumption rate monitoring data of a single-cylinder rotary aero engine under the current operating conditions is obtained, and the real-time fuel consumption rate monitoring data is used as the input parameter for cross-validation; based on the environmentally corrected torque detection results and the current speed parameters, the effective power value of the engine under the current operating conditions is calculated. Based on the effective power value and real-time fuel consumption rate monitoring data, the torque back-calculated value based on fuel consumption is obtained according to the energy conversion relationship. The torque detection results after environmental correction are compared with the torque back-calculated value based on fuel consumption to determine whether the deviation exceeds the preset error allowable threshold. If the deviation does not exceed the preset error allowable threshold, the torque detection result data after environmental correction will be directly used as the final verification calibration torque detection value data output. If the deviation exceeds the preset error allowable threshold, the torque detection result data after environmental correction and the torque back-calculated value based on fuel consumption will be fused and calibrated according to the magnitude and direction of the deviation. The fused and calibrated value will be used as the final torque detection value data for verification calibration.

8. A data processing system for the vibration state of a rotary engine, the system implementing the method as described in any one of claims 1 to 7, characterized in that, include: The acquisition module is used to collect real-time speed data sets of a single-cylinder rotary aero-engine under multiple different load conditions, as well as instantaneous torque measurement data sets measured synchronously with each speed data point. The mapping module is used to perform three-dimensional spatial mapping processing on the speed, torque and time dimensions of the speed dataset and the torque measurement dataset to generate a three-dimensional torque data spatial distribution model that represents the torque distribution with speed and time. The fitting module is used to extract the sequence of polygon vertices composed of discrete measurement data points on the engine external characteristic curve based on the three-dimensional torque data spatial distribution model. By smoothly fitting the sequence of polygon vertices, a continuous and differentiable mathematical model of the torque-speed relationship smooth curve is obtained. The calculation module is used to input the real-time monitored speed parameters of a single-cylinder rotor aero engine based on the mathematical model of the torque-speed relationship smooth curve, and obtain the standard torque reference value data of the real-time speed through model interpolation calculation. The correction module is used to compensate for environmental parameters in the standard torque reference value data. It introduces real-time atmospheric pressure and intake air temperature environmental monitoring data, corrects the standard torque reference value through a preset environmental influence coefficient, and obtains the torque detection result data after environmental correction. The processing module is used to perform cross-validation and error calibration of torque values ​​based on the environmentally corrected torque detection results data and the engine's fuel consumption rate benchmark data under this operating condition, so as to obtain the final verified and calibrated torque detection value data.

9. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images