A gear test bench and a gear evaluation method for simulating a space environment

Abstract

The application provides a kind of simulation space environment gear test bench and gear evaluation method, including support rack, high-low temperature vacuum environment cabin is installed inside the support rack, drive motor is installed on the top side of the support rack, the bottom of the drive motor is connected with one test gear in high-low temperature vacuum environment cabin through transmission shaft after passing through high-low temperature vacuum environment cabin, magnetic powder brake is installed on the other side of the top of the support rack, the bottom of the magnetic powder brake is connected with another test gear in high-low temperature vacuum environment cabin through transmission shaft after passing through high-low temperature vacuum environment cabin, and the two test gears are engaged transmission.The application has the beneficial effects: high-precision simulation of the influence of space environment on gear box under ground conditions, realize the non-interference dynamic loading of gear engagement under microgravity environment, and synchronously monitor the multi-dimensional performance parameters of gear box under extreme working conditions.

Description

Technical Field

[0001] This invention belongs to the field of ground simulation test technology for spacecraft mechanical transmission systems, and in particular relates to a gear test bench for simulating space environment and a gear evaluation method. Background Technology

[0002] Currently, gearbox test benches are mainly designed for ground-based environments and cannot reproduce the unique spatial characteristics of test chambers with dimensions less than 10. -6 microgravity of g, less than 10 -6 High vacuum (Pa), extreme temperatures ranging from -180°C to +200°C, and radiation, such as gamma / X-rays, with a dose rate greater than or equal to 5 kGy / h, are all present. Furthermore, traditional test benches rely on mechanical contact loading methods, such as springs and hydraulic systems, which are subject to gravitational friction interference, making it difficult to accurately simulate the weightless meshing characteristics of space gearboxes. Existing equipment lacks the comprehensive monitoring capability for the dynamic performance of gearboxes under long-term space environments, such as vibration, stress, and lubrication. Summary of the Invention

[0003] In view of this, the present invention aims to propose a gear test bench for simulating a space environment and a gear evaluation method to solve the problem of environmental adaptability testing of space gearboxes in the design verification stage.

[0004] To achieve the above objectives, the technical solution of the present invention is implemented as follows: To achieve the above objectives, the technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides a gear test bench for simulating a space environment, comprising a support frame, a high and low temperature vacuum environment chamber installed inside the support frame, a drive motor installed on one side above the support frame, the bottom of the drive motor passing through the high and low temperature vacuum environment chamber via a transmission shaft and connected to a test gear inside the high and low temperature vacuum environment chamber, and a magnetic powder brake installed on the other side above the support frame, the bottom of the magnetic powder brake passing through the high and low temperature vacuum environment chamber via a transmission shaft and connected to another test gear inside the high and low temperature vacuum environment chamber, and the two test gears meshing and transmitting power.

[0005] Furthermore, a zero-gravity unit is installed at the bottom of the high and low temperature vacuum environment chamber.

[0006] Furthermore, a triaxial accelerometer is installed on one side of the high and low temperature vacuum environment chamber.

[0007] Furthermore, both the drive motor and the magnetic powder brake are equipped with constant speed and torque sensors.

[0008] Furthermore, fiber Bragg grating strain gauges were installed at the connection between the two test gears.

[0009] Secondly, based on the same concept, the present invention also provides a gear evaluation method for a gear test bench simulating a space environment, including an evaluation algorithm for evaluating test gears. The evaluation algorithm includes the following steps: S1. Preparation stage: Space environment factor calibration and coupling factor calculation, vibration signal acquisition and preprocessing, WPD optimization parameter definition, and algorithm hyperparameter setting are carried out respectively. S2, ECAPSO algorithm initialization phase: Perform multidimensional particle encoding, particle position and velocity initialization, and individual optimal and global optimal solution initialization respectively; S3, ECAPSO multi-objective iterative optimization WPD parameter stage: performing cyclic iterations of particle multi-objective fitness value calculation, adaptive parameter update, particle position and velocity update, population diversity maintenance, and individual optimal and global optimal solution update; S4. Customized signal analysis stage after optimization of WPD: WPD constructs a set of orthogonal wavelet basis functions by performing layer-by-layer filtering and sampling operations on the signal, realizing multi-scale decomposition of the signal in the whole frequency band, and obtaining the sub-band coefficients and reconstructed signal of the optimized WPD decomposition. S5. Feature Extraction and Output Stage: Based on the sub-band coefficients and reconstructed signals of the optimized WPD decomposition, a spatial gear feature set is extracted. The spatial gear feature set includes 4 basic statistical features, 3 time-frequency decomposition features, 3 environmental coupling features, and 2 dynamic features. The 12 features are standardized and used as the core input data for the subsequent evaluation of the spatial gear transmission performance.

[0010] Furthermore, the calibration of space environment factors and the calculation of coupling factors include: For the service environment of space gears, space environment influence factors are calibrated respectively. The space environment influence factors include microgravity influence factor λ1, high and low temperature alternation factor λ2, performance degradation influence factor λ3, and boundary support influence factor λ4.

[0011] The spatial environment coupling factor λ is calculated based on the environmental impact weight, where... The weights α, β, γ, and ζ are determined by the similarity between the ground simulation environment and the actual space environment, and satisfy α, β, γ, and ζ = 1. Vibration signal acquisition and preprocessing, including: Signal acquisition: Vibration signals of the space gearbox under the same working conditions are acquired on the ground simulation platform. The sampling frequency is not less than 10 times the gear meshing frequency. At the same time, corresponding simulated environmental parameters are acquired: microgravity simulation value, radiation dose, and temperature change range.

[0012] Basic preprocessing: The original vibration signal is subjected to DC component removal, trend term elimination, and outlier removal operations to obtain a clean signal sequence to be analyzed; Define the WPD parameters to be optimized: Identify the core WPD parameters that ECAPSO needs to optimize, and set the value range of each parameter according to the characteristics of the space gear vibration signal.

[0013] Furthermore, multidimensional particle encoding includes: The four parameters to be optimized in WPD—decomposition layer number, wavelet basis function, subband threshold, and reconstructed subband combination—are used as the four dimensions of the particle. Each dimension is standardized and numerically encoded, and each particle is ultimately represented as a four-dimensional vector, corresponding to a set of WPD parameter combinations. Particle position and velocity are randomly initialized, including: Within a preset parameter solution space, the initial position and initial velocity of each particle are randomly and uniformly generated; the velocity range is set according to the position solution space. Initialization of individual optimal and global optimal solutions includes: Individual optimal solution: The initial position of each particle is taken as its individual historical optimal position; Global optimal solution: Calculate the fitness value of the initial position of all particles, and take the position of the particle with the best fitness value as the global optimal position; Population diversity initialization: Initialize the particle crowding factor, with an initial value of 0.

[0014] Furthermore, the multi-target fitness value of the particles is calculated, including: A three-objective fitness function is constructed to comprehensively evaluate the performance of each WPD parameter combination; the smaller the fitness value, the better the parameter combination. The three objectives are transformed into a single-objective fitness function through a linear weighting method. The three-objective fitness function includes: Signal reconstruction error: The mean square error between the optimized WPD reconstructed signal and the original preprocessed signal, reflecting the signal reconstruction accuracy; Signal reconstruction error formula: ; in, Number of sampling points; : The value of the pure vibration signal after preprocessing at the i-th sampling point; The value of the signal reconstructed by WPD at the i-th sampling point; Feature discriminability: The feature separation degree between the high-frequency impact subband and the low-frequency drift subband after WPD decomposition; Feature discriminability formula: ; Similarity between ground and space features: the cosine distance between the decomposed features of the ground-based analog signal and the features of the actual space signal; the distance formula is: ; in, : Represents the dimension of the feature vector; The i-th feature value extracted from the ground-based analog signal; The i-th feature value extracted from the actual signal in space; The calculation of the single-objective fitness function includes: Calculate the fitness value of all particles in the population using the formula above. The smaller the calculated fitness value, the better the combination. The fitness value formula is as follows: ; in: : The value of the pure vibration signal after preprocessing at the i-th sampling point; The value of the signal reconstructed by WPD at the i-th sampling point; High-frequency impact sub-band coefficient; Low-frequency impact subband coefficient; Characteristics of ground-based analog signal decomposition; : Actual signal characteristics in space; Adaptive parameter updates, including: Based on the environmental coupling factor and the current iteration number, the environmental coupling inertial weight and dual-drive acceleration coefficient of the particle are dynamically updated. The environmental coupling inertial weight update incorporates the environmental coupling factor into the calculation, and its formula is as follows: ; in: : Upper limit of inertia weight; Lower bound of inertia weight; : Current iteration number; η: Maximum number of iterations; η: Environmental coupling coefficient; Among them, the dual-drive acceleration coefficient is divided into individual learning factor C1 and global learning factor C2. Updates to individual optimal and global optimal solutions include: Individual optimal solution update: Compare the fitness value of the particle's new position with its own historical best fitness value. If the calculated result is smaller than the original, update; otherwise, keep. Global optimal solution update: Compare the new individual optimal fitness value of all particles with the current global optimal fitness value; Population diversity maintenance includes: Calculate the particle crowding factor C of the current population: ; in, : The position vector of the i-th particle in the solution space; : The position vector of the j-th particle; : Indicates the total number of particles; If the preset threshold is exceeded The local optimal particle is updated by random perturbation; ; That is, adding a tiny random number to one dimension of its position vector, where: Local optimal particle position; : Amplitude of small disturbances; Unit vector, representing a randomly selected dimension; Particle position and velocity updates include: Each particle represents a candidate solution, with its position xi = (xi1, xi2, ..., xid) corresponding to a point in the solution space, and its velocity vi = (vi1, vi2, ..., vid) determining its direction and step size. The particle updates its velocity and position based on its historical best position (pbesti) and global best position (gbest). ; ; Where: w: inertia weight, controlling the particle's motion inertia; C1, C2: acceleration coefficients; r1, r2: random numbers in the interval [0,1]; The individual historical best position of particle i; : Global optimal position; : The velocity of particle i in generation t; The position of particle i in the kth generation; Iterative convergence criteria include: Determine if the iteration termination condition is met: if the termination condition is met, proceed to the output stage; if the termination condition is not met, return to the evaluation stage and continue iterative optimization.

[0015] Furthermore, in the optimized WPD customized signal analysis stage, WPD constructs a set of orthogonal wavelet basis functions by performing layer-by-layer filtering and sampling operations on the signal, thereby achieving multi-scale decomposition of the signal across the entire frequency band; its basic iterative formula is: ; In the formula, Indicates the first The basis functions of the layer and These are low-pass and high-pass filters, respectively. , satisfies orthogonality; with Construct orthogonal wavelet packets for the initial scaling function, assuming the signal ,but It can be represented as: ; In the formula, Indicates the number of decomposition layers. Indicates frequency band index For the first Layer Wavelet coefficients. These are the basis functions of wavelet decomposition; given the decomposition coefficients of a certain level... At that time, the first The coefficients of the layer can be obtained through the filter coefficients. , The results were obtained from the sampling operation: ; Therefore, the wavelet packet reconstruction is represented as: ; Using the optimal WPD parameter combination obtained through ECAPSO iterative optimization, customized wavelet packet decomposition, hierarchical adaptive denoising, and optimal sub-band reconstruction are performed on the preprocessed gear vibration signal to fully adapt to the signal characteristics under spatial environment coupling. The steps are as follows: Multi-scale wavelet packet decomposition: Based on the optimal number of decomposition layers and the optimal wavelet basis function, the clean signal is decomposed into layers of wavelet packet to obtain the wavelet packet coefficients of each sub-band. Hierarchical adaptive threshold denoising: Based on the optimal sub-band threshold, differentiated denoising strategies are adopted for different frequency sub-bands, which can be determined by ECAPSO optimization: soft thresholding is used for high-frequency sub-bands, and hard thresholding is used for low-frequency sub-bands; the wavelet packet coefficients after denoising are corrected to obtain the clean coefficients; Optimal subband reconstruction: Based on the optimal subband combination, effective feature subbands are selected from the subbands, mainly by eliminating noise-dominant subbands. Inverse wavelet packet transform is performed on the corrected wavelet packet coefficients to obtain the denoised reconstructed signal. This signal retains the core features of spatial gear vibration and the noise is effectively filtered out.

[0016] Compared with existing technologies, the gear test bench and gear evaluation method for simulating a space environment described in this invention have the following advantages: The gear test bench and gear evaluation method for simulating a space environment described in this invention can demonstrate environmental reproducibility: covering extreme space conditions with an error of less than 5%; demonstrate loading accuracy: magnetic levitation loading eliminates mechanical friction interference, with torque control accuracy of ±0.5%; demonstrate data comprehensiveness: simultaneous monitoring of multiple physical fields supports fault mode analysis; and demonstrate scalability: adaptable to the testing needs of other space transmission components. Attached Figure Description

[0017] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the principle of the simulated space environment gear test bench according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the PSO-WPD optimization process described in an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0020] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0021] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] like Figures 1 to 2 As shown, a simulated space environment gear test bench includes a support frame. A high and low temperature vacuum environment chamber is installed inside the support frame. A drive motor is installed on one side above the support frame. The bottom of the drive motor passes through the high and low temperature vacuum environment chamber via a transmission shaft and is connected to a test gear inside the high and low temperature vacuum environment chamber. A magnetic powder brake is installed on the other side above the support frame. The bottom of the magnetic powder brake passes through the high and low temperature vacuum environment chamber via a transmission shaft and is connected to another test gear inside the high and low temperature vacuum environment chamber. The two test gears mesh and drive each other.

[0023] This simulated space environment gear test bench offers the following advantages: Environmental reproducibility: Covers extreme spatial conditions such as vacuum, temperature change, and radiation, with an error of less than 5%; Loading accuracy: Magnetic levitation loading eliminates mechanical friction interference, with torque control accuracy of ±0.5%; Data comprehensiveness: Simultaneous monitoring of multiple physical fields supports fault mode analysis, such as tooth surface spalling and lubrication failure; Expandability: Adaptable to the testing needs of other spatial transmission components such as bearings and couplings.

[0024] In a preferred embodiment of the invention, a zero-gravity unit is installed at the bottom of the high-low temperature vacuum environment chamber, and a triaxial accelerometer is installed on one side of the chamber. In this embodiment, the zero-gravity unit is a device for isolating vibrations, achieving vibration attenuation through a special stiffness design. In the high-low temperature vacuum environment chamber, it is used to reduce the interference of external vibrations on the experiment. The triaxial accelerometer is a sensor that measures the vibration acceleration of the equipment in three orthogonal directions (X, Y, and Z axes), used to monitor the vibration characteristics inside the simulated chamber in real time, providing dynamic data support for gear testing.

[0025] In a preferred embodiment of the invention, both the drive motor and the magnetic powder brake are equipped with constant speed and torque sensors. In this embodiment, the constant speed and torque sensor is a sensor used to monitor the speed and torque parameters of the equipment in real time.

[0026] In a preferred embodiment of the invention, a fiber Bragg grating strain gauge is also installed at the connection between the two test gears. In this embodiment, the fiber Bragg grating strain gauge is a sensor based on the fiber grating principle, used to measure strain changes on the surface of a structure.

[0027] In this embodiment, the core structure specifically includes: High and low temperature vacuum chamber: Double-layer structure, consisting of an inner stainless steel layer and an outer thermal insulation material layer, integrated with a vacuum pump, capable of maintaining a vacuum level of 10°C. -6 Pa, liquid nitrogen / electric heating temperature control system, temperature range -180°C to +200°C, temperature change rate ±30°C / min, radiation source, γ / X-ray, dose rate adjustable.

[0028] Gravity compensation system: The test bench axis is parallel to the gravity direction, and it is used in conjunction with a quasi-zero stiffness vibration isolation platform to simulate microgravity meshing conditions.

[0029] Dynamic loading and transmission system: magnetic powder brake and torque sensor. The magnetic powder brake adopts dynamic torque loading of 0-500Nm, and the torque sensor can monitor the input / output shaft speed and torque in real time.

[0030] Multimodal monitoring system: vibration sensor for collecting shell vibration, fiber optic strain gauge for collecting tooth surface stress, temperature sensor for collecting lubricant and housing temperature, and speed and torque sensor.

[0031] A working process for a gear test bench simulating a space environment: Secure the gearbox inside the environmental tank, aligning its axis with the direction of gravity. Start the vacuum pump, temperature control system, and radiation source to establish the space environment; The drive motor drives the gearbox through a reducer, and the magnetic powder brake applies a dynamic load. Multiple sensors simultaneously collect data such as vibration, temperature, stress, and torque.

[0032] This invention also proposes a gear evaluation method for a gear test bench simulating a space environment, including an evaluation algorithm. The evaluation algorithm evaluates the test gears and includes the following steps: S1. Preparation stage: Space environment factor calibration and coupling factor calculation, vibration signal acquisition and preprocessing, WPD optimization parameter definition, and algorithm hyperparameter setting are carried out respectively. S2, ECAPSO algorithm initialization phase: Perform multidimensional particle encoding, particle position and velocity initialization, and individual optimal and global optimal solution initialization respectively; S3, ECAPSO multi-objective iterative optimization of WPD parameters stage: S4. Optimized WPD Customized Signal Analysis Stage: S5, Feature Extraction and Output Stage.

[0033] In a preferred embodiment of the present invention, the Environment Coupled Adaptive Particle Swarm Optimization-Wavelet Packet Decomposition (ECAPSO-WPD) algorithm is a customized upgrade of the traditional PSO-WPD algorithm, designed for the nonlinear, non-stationary, and multi-interference characteristics of vibration signals in the microgravity, high radiation, and high-low temperature alternating coupling environment of a space gearbox. The core of the algorithm achieves environment-aware dynamic optimization of key parameters of wavelet packet decomposition (WPD) through Environment Coupled Adaptive Particle Swarm Optimization (ECAPSO), and then uses the optimized WPD to complete high-precision feature extraction of the space gear vibration signal, providing support for the consistent dynamic evaluation of space gear transmission performance. The algorithm is divided into five parts: a preparation stage, an ECAPSO algorithm initialization stage, an ECAPSO multi-objective iterative optimization stage for WPD parameters, a customized signal analysis stage for the optimized WPD, and a feature extraction and output stage.

[0034] In a preferred embodiment of the present invention, the preparation stage includes spatial environment factor calibration and coupling factor calculation, vibration signal acquisition and preprocessing, WPD optimization parameter definition, and algorithm hyperparameter setting. This quantifies the coupling influence of the spatial environment on the gear vibration signal, laying the foundation for subsequent algorithm operation. This is a key pre-step that distinguishes ECAPSO from traditional PSO.

[0035] In a preferred embodiment of the present invention, space environment factor calibration and coupling factor calculation are performed as follows: (1) For the service environment of space gears, environmental impact factors are calibrated respectively (the values ​​are all [0,1], and the larger the value, the stronger the environmental impact): Microgravity influence factor λ1: Characterizes the frictional vibration coefficient caused by uneven lubrication, quantified by gear meshing friction test data; High and low temperature alternation factor λ2: Characterizes the change coefficient of meshing clearance caused by thermal expansion and contraction, which is quantified by fitting the amount of clearance change and vibration frequency drift in high and low temperature tests; Performance degradation factor λ3: Characterizes the stiffness change coefficient caused by material wear and aging; Boundary support influence factor λ4: Characterizes the boundary support factor of the ground simulation system. Ground gravity unloading is difficult to completely simulate the microgravity state in space, and the boundary conditions need to be modified.

[0036] (2) Calculate the spatial environment coupling factor λ based on environmental impact weights. The weights α, β, γ, and ζ are determined by the similarity between the ground simulation environment and the actual space environment, and satisfy α, β, γ, and ζ = 1.

[0037] (3) The environmental coupling factor λ provides the core basis for the adaptive adjustment of the inertia weight and acceleration coefficient of the subsequent ECAPSO algorithm.

[0038] In a preferred embodiment of the present invention, vibration signal acquisition and preprocessing are performed as follows: Signal acquisition: Vibration signals of the space gearbox under the same working conditions are acquired on the ground simulation platform. The sampling frequency is not less than 10 times the gear meshing frequency to ensure that high-frequency impact signals are not lost. At the same time, corresponding simulated environmental parameters are acquired: microgravity simulation value, radiation dose, and temperature change range.

[0039] Basic preprocessing: The original vibration signal is processed by removing the DC component, eliminating the trend term, and removing outliers to obtain a clean signal sequence to be analyzed.

[0040] In a preferred embodiment of the present invention, the WPD parameters to be optimized and the solution space are defined as follows: The core WPD parameters that ECAPSO needs to optimize are identified, and the value range (solution space) of each parameter is set according to the characteristics of the spatial gear vibration signal, providing a basis for particle coding.

[0041] In a preferred embodiment of the present invention, the basic hyperparameters of the algorithm are set as follows: The basic hyperparameters of the ECAPSO algorithm are determined as follows, and can be fine-tuned according to the actual signal analysis requirements: particle population size; maximum number of iterations; upper and lower limits of inertial weight; upper and lower limits of acceleration coefficient.

[0042] In a preferred embodiment of the present invention, the ECAPSO algorithm initialization stage completes the multidimensional encoding of particles, the initialization of particle position and velocity, and the initialization of individual optimal and global optimal solutions. This stage transforms the WPD parameters to be optimized into particle features that can be processed by the ECAPSO algorithm and is the starting point for iterative optimization.

[0043] In a preferred embodiment of the present invention, particle multidimensional encoding is performed as follows: The four parameters to be optimized in WPD—decomposition level, wavelet basis function, subband threshold, and reconstructed subband combination—are used as the four dimensions of the particle. Each dimension is standardized and numerically encoded, and each particle is ultimately represented as a four-dimensional vector, corresponding to a set of WPD parameter combinations.

[0044] In a preferred embodiment of the present invention, the particle position and velocity are randomly initialized: Within the preset parameter solution space, the initial position and initial velocity of each particle are randomly and uniformly generated; the velocity value range is set according to the position solution space, usually ±0.5 times the difference between the upper and lower limits of the position, to ensure that the initial distribution of particles is uniform and to avoid the algorithm getting trapped in local optima.

[0045] In a preferred embodiment of the present invention, the individual optimal and global optimal solutions are initialized as follows: Individual optimal solution: The initial position of each particle is taken as its individual historical optimal position; Global optimal solution: Calculate the fitness value of the initial position of all particles, and take the position of the particle with the best fitness value as the global optimal position; Population diversity initialization: Initialize the particle crowding factor for subsequent population diversity maintenance, with an initial value of 0.

[0046] In a preferred embodiment of the present invention, the ECAPSO multi-objective iterative optimization stage for WPD parameters achieves global optimal search for WPD parameters through iterative processes including particle multi-objective fitness value calculation, adaptive parameter update, particle position and velocity update, population diversity maintenance, and updates of individual optimal and global optimal solutions. The core difference between this stage and traditional PSO is the integration of adaptive adjustment of the environmental coupling factor λ and the construction of a multi-objective fitness function specific to the space gear signal. The iteration termination condition is: reaching the maximum number of iterations, or the global optimal fitness value showing no significant change for 20 consecutive generations (convergence accuracy ≤ 10). -4 ).

[0047] In a preferred embodiment of the present invention, the multi-target fitness value of the particle is calculated: A three-objective fitness function is constructed to comprehensively evaluate the performance of each WPD parameter combination. The smaller the fitness value, the better the parameter combination. The three objectives are transformed into a single-objective fitness function by a linear weighting method. The weights are set according to engineering requirements (such as signal reconstruction accuracy 0.4, feature recognition 0.3, and sky-ground similarity 0.3), and the weight sum is 1.

[0048] Definition of three objective functions Signal reconstruction error: The mean square error between the optimized WPD reconstructed signal and the original preprocessed signal, reflecting the signal reconstruction accuracy. Signal reconstruction error formula: ; in, Number of sampling points; : The value of the pure vibration signal after preprocessing at the i-th sampling point; The value of the signal reconstructed by WPD at the i-th sampling point; Feature discriminability: The feature separation degree between the high-frequency impact subband and the low-frequency drift subband after WPD decomposition. Feature discriminability formula: ; Similarity between ground and space features: The cosine distance between the decomposed features of the ground-based analog signal and the features of the actual space signal. The distance formula is: ; in, : Represents the dimension of the feature vector; The i-th feature value extracted from the ground-based analog signal; The i-th feature value extracted from the actual signal in space; Single-objective fitness function calculation Calculate the fitness value of all particles in the population using the formula above. The smaller the calculated fitness value, the better the combination. The fitness value formula is: ; in: : The value of the pure vibration signal after preprocessing at the i-th sampling point; The value of the signal reconstructed by WPD at the i-th sampling point; High-frequency impact sub-band coefficient; Low-frequency impact subband coefficient; Characteristics of ground-based analog signal decomposition; : Actual signal characteristics in space.

[0049] In a preferred embodiment of the present invention, ECAPSO adaptive parameter updates are performed as follows: Based on the environmental coupling factor and the current iteration number, the inertial weight and acceleration coefficient of the particles are dynamically updated to achieve an environmentally adaptive balance between the algorithm's global exploration capability and local development capability.

[0050] Environment-coupled inertial weight update: This method abandons traditional linear inertial weights and incorporates environmental coupling factors into the calculation, achieving stronger global exploration capabilities with stronger environmental disturbances. The calculation formula is as follows: ; in: : Upper limit of inertia weight; Lower bound of inertia weight; : Current iteration number; : Maximum number of iterations; η: Environmental coupling coefficient.

[0051] Dual-drive acceleration coefficient update: The acceleration coefficient is divided into individual learning factor C1 (driven by the similarity of local signal features) and global learning factor C2 (driven by environmental coupling factors), realizing a dynamic balance between individual exploration and global collaboration.

[0052] In a preferred embodiment of the present invention, the individual optimal and global optimal solutions are updated as follows: Individual optimal solution update: Compare the fitness value of the particle's new position with its own historical best fitness value. If the calculated result is smaller than the original, update; otherwise, keep. Global optimal solution update: Compare the new individual optimal fitness value of all particles with the current global optimal fitness value.

[0053] In a preferred embodiment of the present invention, population diversity is maintained: Calculate the particle crowding factor C of the current population (characterizing the degree of particle aggregation in the solution space; the larger the factor, the more severe the aggregation). ; in, : The position vector of the i-th particle in the solution space; : The position vector of the j-th particle; : Indicates the total number of particles; If the preset threshold is exceeded (Usually taken as 0.8), the locally optimal particles (the top 20% of particles in terms of fitness value) are updated by random perturbation; ; This involves adding a tiny random number to one dimension of the particle's position vector to break the particle aggregation state and prevent the algorithm from getting trapped in local optima. Specifically: Local optimal particle position; : Amplitude of small disturbances; Unit vector, representing a randomly selected dimension. In a preferred embodiment of the present invention, particle position and velocity are updated as follows: Each particle represents a candidate solution, with its position xi = (xi1, xi2, ..., xid) corresponding to a point in the solution space, and its velocity vi = (vi1, vi2, ..., vid) determining its direction and step size. The particle updates its velocity and position based on its historical best position (pbesti) and global best position (gbest). ; ; Where: w: inertia weight, controlling the particle's motion inertia; C1, C2: acceleration coefficients; r1, r2: random numbers in the interval [0,1]; The individual historical best position of particle i; : Global optimal position; : The velocity of particle i in generation t; : The position of particle i in the kth generation.

[0054] In a preferred embodiment of the present invention, the iterative convergence determination is as follows: Determine if the iteration termination condition is met: if the termination condition is met, proceed to the output stage; if the termination condition is not met, return to the evaluation stage and continue iterative optimization.

[0055] In a preferred embodiment of the present invention, during the optimized WPD customized signal analysis stage, WPD constructs a set of orthogonal wavelet basis functions by performing layer-by-layer filtering and sampling operations on the signal, thereby achieving multi-scale decomposition of the signal across the entire frequency band. Its basic iterative formula is: ; In the formula, Indicates the first The basis functions of the layer and These are low-pass and high-pass filters, respectively. They satisfy orthogonality. Construct orthogonal wavelet packets for the initial scaling function, assuming the signal ,but It can be represented as: ; In the formula, Indicates the number of decomposition layers. Indicates frequency band index For the first Layer Wavelet coefficients. These are the basis functions of wavelet decomposition. Given the decomposition coefficients of a certain level... At that time, the first The coefficients of the layer can be obtained through the filter coefficients. , The results were obtained from the sampling operation: ; Therefore, wavelet packet reconstruction can be represented as: ; Using the optimal WPD parameter combination obtained through ECAPSO iterative optimization, customized wavelet packet decomposition, hierarchical adaptive denoising, and optimal sub-band reconstruction are performed on the preprocessed gear vibration signal to fully adapt to the signal characteristics (high-frequency impact, low-frequency drift, broadband noise) under spatial environment coupling. The steps are as follows: Multi-scale wavelet packet decomposition: Based on the optimal number of decomposition layers and the optimal wavelet basis function, layer wavelet packet decomposition is performed on the clean signal to obtain the wavelet packet coefficients of each sub-band.

[0056] Hierarchical adaptive threshold denoising: Based on the optimal sub-band threshold, differentiated denoising strategies are adopted for different frequency sub-bands, which can be determined by ECAPSO optimization: soft thresholding is used for high-frequency sub-bands, and hard thresholding is used for low-frequency sub-bands; the wavelet packet coefficients after denoising are corrected to obtain the clean coefficients.

[0057] Optimal subband reconstruction: Based on the optimal subband combination, effective feature subbands are selected from the subbands, mainly by eliminating noise-dominant subbands. Inverse wavelet packet transform is performed on the corrected wavelet packet coefficients to obtain the denoised reconstructed signal. This signal retains the core features of spatial gear vibration and the noise is effectively filtered out.

[0058] In a preferred embodiment of the present invention, during the feature extraction and output stage, a feature set of the space gear is extracted based on the sub-band coefficients and reconstructed signal of the optimized WPD decomposition. The feature set combines the coupling characteristics of the space environment and covers four major categories of features, comprehensively reflecting the vibration characteristics and dynamic state of the space gear. This provides accurate feature data support for the subsequent space-ground dynamic feature mapping model and the space gear transmission performance space-ground consistent evaluation system. The specific feature classifications and indicators are as follows: Basic statistical characteristics (4 items): kurtosis, skewness, root mean square, and peak factor, reflecting the nonlinearity, impulsiveness, and energy distribution characteristics of the signal.

[0059] Time-frequency decomposition characteristics (3 items): energy entropy of each sub-band, energy proportion of high-frequency impact sub-band, frequency offset of low-frequency drift sub-band, reflecting the time-frequency distribution and characteristic change trend of the signal.

[0060] Environmental coupling characteristics (3 items): radiation-induced stiffness variation characteristics, temperature-induced clearance variation characteristics, and microgravity-induced uneven lubrication friction characteristics, which are obtained by fitting the sub-band coefficient with environmental factors and reflect the coupling effect of the space environment on gears.

[0061] Dynamic characteristics (2 items): meshing frequency modulation coefficient, proportion of nonlinear vibration components, and high-frequency impact concentration, reflecting the dynamic characteristics and transmission smoothness of gear meshing.

[0062] The above 12-dimensional features are standardized, that is, the output is after eliminating the influence of dimensions, and serves as the core input data for the subsequent evaluation of the consistency between space and ground in the performance of spatial gear transmission.

[0063] The essence of the ECAPSO-WPD algorithm is a closed-loop logic of environmental perception, parameter optimization, signal analysis, and feature extraction. Its core advantages are reflected in: In terms of environmental adaptation, the space environment coupling factor λi is integrated throughout the entire ECAPSO parameter update process, realizing the upgrade of WPD parameters from manual setting / traditional random optimization to environment-aware dynamic optimization; In terms of customized analysis, WPD's decomposition, denoising, and reconstruction all use ECAPSO's optimized proprietary parameters to accurately adapt to the coupling characteristics of high-frequency impact, low-frequency drift, and broadband noise in space gear vibration signals. In terms of high-precision feature extraction, the extracted 12-dimensional exclusive feature set, combined with the coupling characteristics of the spatial environment, is more in line with the actual transmission state of the space gear, providing a reliable data foundation for the consistent evaluation between space and ground.

[0064] Example 1 A gear test bench simulating a space environment: Environmental simulation tank commissioning: Evacuate to 10 -6Pa, set the temperature cycle (-180°C→+200°C, rate ±30°C / min); turn on the gamma-ray source and adjust the dose rate to 5 kGy / h.

[0065] Gearbox installation and alignment: Adjust the shaft system to be parallel to the direction of gravity, and calibrate the vibration isolation platform to a near-zero stiffness state.

[0066] Dynamic loading and data acquisition: The drive motor operates at a set speed, the magnetic powder brake applies a set torque, and data such as vibration spectrum, temperature gradient, and tooth surface strain are recorded simultaneously.

[0067] This invention relates to the field of ground simulation testing technology for spacecraft mechanical transmission systems, specifically to a gearbox performance testing device capable of simulating extreme space environments (microgravity, high vacuum, extreme temperature, radiation), used for reliability verification and life assessment of spacecraft gear transmission systems.

[0068] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A gear test bench simulating a space environment, characterized in that: The device includes a support frame, inside which a high and low temperature vacuum environment chamber is installed. A drive motor is installed on one side of the support frame. The bottom of the drive motor passes through the high and low temperature vacuum environment chamber via a transmission shaft and is connected to a test gear inside the chamber. A magnetic powder brake is installed on the other side of the support frame. The bottom of the magnetic powder brake passes through the high and low temperature vacuum environment chamber via a transmission shaft and is connected to another test gear inside the chamber. The two test gears mesh and drive each other.

2. The gear test bench for simulating a space environment according to claim 1, characterized in that: The bottom of the high and low temperature vacuum environment chamber is equipped with a zero gravity unit.

3. The gear test bench for simulating a space environment according to claim 1, characterized in that: A triaxial accelerometer is installed on one side of the high and low temperature vacuum environment chamber.

4. The gear test bench for simulating a space environment according to claim 1, characterized in that: Both the drive motor and the magnetic powder brake are equipped with constant speed and torque sensors.

5. The gear test bench for simulating a space environment according to claim 1, characterized in that: Fiber Bragg grating strain gauges were also installed at the connection between the two test gears.

6. A gear evaluation method for a gear test bench simulating a space environment, applied to the gear test bench simulating a space environment as described in any one of claims 1 to 5, characterized in that: This includes an evaluation algorithm for evaluating test gears. The evaluation algorithm includes the following steps: S1. Preparation stage: Space environment factor calibration and coupling factor calculation, vibration signal acquisition and preprocessing, WPD optimization parameter definition, and algorithm hyperparameter setting are carried out respectively. S2, ECAPSO algorithm initialization phase: Perform multidimensional particle encoding, particle position and velocity initialization, and individual optimal and global optimal solution initialization respectively; S3, ECAPSO multi-objective iterative optimization WPD parameter stage: performing cyclic iterations of particle multi-objective fitness value calculation, adaptive parameter update, particle position and velocity update, population diversity maintenance, and individual optimal and global optimal solution update; S4. Customized signal analysis stage after optimization of WPD: WPD constructs a set of orthogonal wavelet basis functions by performing layer-by-layer filtering and sampling operations on the signal, realizing multi-scale decomposition of the signal in the whole frequency band, and obtaining the sub-band coefficients and reconstructed signal of the optimized WPD decomposition. S5. Feature Extraction and Output Stage: Based on the sub-band coefficients and reconstructed signals of the optimized WPD decomposition, a spatial gear feature set is extracted. The spatial gear feature set includes 4 basic statistical features, 3 time-frequency decomposition features, 3 environmental coupling features, and 2 dynamic features. The 12 features are standardized and used as the core input data for the subsequent evaluation of the spatial gear transmission performance.

7. The gear evaluation method for a gear test bench simulating a space environment according to claim 6, characterized in that: Space environment factor calibration and coupling factor calculation include: For the service environment of space gears, space environment influence factors are calibrated respectively. The space environment influence factors include microgravity influence factor λ1, high and low temperature alternation factor λ2, performance degradation influence factor λ3, and boundary support influence factor λ4. The spatial environment coupling factor λ is calculated based on the environmental impact weight, where... The weights α, β, γ, and ζ are determined by the similarity between the ground simulation environment and the actual space environment, and satisfy α, β, γ, and ζ = 1. Vibration signal acquisition and preprocessing, including: Signal acquisition: Vibration signals of the space gearbox under the same working conditions are acquired on the ground simulation platform. The sampling frequency is not less than 10 times the gear meshing frequency. At the same time, corresponding simulated environmental parameters are acquired: microgravity simulation value, radiation dose, and temperature change range. Basic preprocessing: The original vibration signal is subjected to DC component removal, trend term elimination, and outlier removal operations to obtain a clean signal sequence to be analyzed; Define the WPD parameters to be optimized: Identify the core WPD parameters that ECAPSO needs to optimize, and set the value range of each parameter according to the characteristics of the space gear vibration signal.

8. The gear evaluation method for a gear test bench simulating a space environment according to claim 6, characterized in that: Particle multidimensional encoding, including: The four parameters to be optimized in WPD—decomposition layer number, wavelet basis function, subband threshold, and reconstructed subband combination—are used as the four dimensions of the particle. Each dimension is standardized and numerically encoded, and each particle is ultimately represented as a four-dimensional vector, corresponding to a set of WPD parameter combinations. Particle position and velocity are randomly initialized, including: Within a preset parameter solution space, the initial position and initial velocity of each particle are randomly and uniformly generated; the velocity range is set according to the position solution space. Initialization of individual optimal and global optimal solutions includes: Individual optimal solution: The initial position of each particle is taken as its individual historical optimal position; Global optimal solution: Calculate the fitness value of the initial position of all particles, and take the position of the particle with the best fitness value as the global optimal position; Population diversity initialization: Initialize the particle crowding factor, with an initial value of 0.

9. The gear evaluation method for a gear test bench simulating a space environment according to claim 6, characterized in that: Calculate the multi-target fitness value of particles, including: A three-objective fitness function is constructed to comprehensively evaluate the performance of each WPD parameter combination; the smaller the fitness value, the better the parameter combination. The three objectives are transformed into a single-objective fitness function through a linear weighting method. The three-objective fitness function includes: Signal reconstruction error: The mean square error between the optimized WPD reconstructed signal and the original preprocessed signal, reflecting the signal reconstruction accuracy; Signal reconstruction error formula: ； in, Number of sampling points; : The value of the pure vibration signal after preprocessing at the i-th sampling point; The value of the signal reconstructed by WPD at the i-th sampling point; Feature discriminability: The feature separation degree between the high-frequency impact subband and the low-frequency drift subband after WPD decomposition; Feature discriminability formula: ； Similarity between ground and space features: the cosine distance between the decomposed features of the ground-based analog signal and the features of the actual space signal; the distance formula is: ； in, : Represents the dimension of the feature vector; The i-th feature value extracted from the ground-based analog signal; The i-th feature value extracted from the actual signal in space; The calculation of the single-objective fitness function includes: Calculate the fitness value of all particles in the population using the formula above. The smaller the calculated fitness value, the better the combination. The fitness value formula is as follows: ； in: ; : The value of the pure vibration signal after preprocessing at the i-th sampling point; : The value of the signal reconstructed by WPD at the i-th sampling point; : High-frequency impact sub-band coefficient; Low-frequency impact subband coefficient; : Characteristics of ground-based analog signal decomposition; : Actual signal characteristics in space; Adaptive parameter updates, including: Based on the environmental coupling factor and the current iteration number, the environmental coupling inertial weight and dual-drive acceleration coefficient of the particle are dynamically updated. The environmental coupling inertial weight update incorporates the environmental coupling factor into the calculation, and its formula is as follows: ； in: : Upper limit of inertia weight; Lower bound of inertia weight; : Current iteration number; η: Maximum number of iterations; η: Environmental coupling coefficient; Among them, the dual-drive acceleration coefficient is divided into individual learning factor C1 and global learning factor C2. Updates to individual optimal and global optimal solutions include: Individual optimal solution update: Compare the fitness value of the particle's new position with its own historical best fitness value. If the calculated result is smaller than the original, update; otherwise, keep. Global optimal solution update: Compare the new individual optimal fitness value of all particles with the current global optimal fitness value; Population diversity maintenance includes: Calculate the particle crowding factor C of the current population: ； in, : The position vector of the i-th particle in the solution space; : The position vector of the j-th particle; : Indicates the total number of particles; If the preset threshold is exceeded The local optimal particle is updated by random perturbation; ； That is, adding a tiny random number to one dimension of its position vector, where: Local optimal particle position; : Amplitude of small perturbations; Unit vector, representing a randomly selected dimension; Particle position and velocity updates include: Each particle represents a candidate solution, with its position xi = (xi1, xi2, ..., xid) corresponding to a point in the solution space, and its velocity vi = (vi1, vi2, ..., vid) determining the particle's direction and step size. The particle updates its velocity and position based on its historical best position and the global best position. ； ； Where: w: inertia weight, controlling the particle's motion inertia; C1, C2: acceleration coefficients; r1, r2: random numbers in the interval [0,1]; The individual best position in the history of particle i; : Global optimal position; : The velocity of particle i in generation t; The position of particle i in the kth generation; Iterative convergence criteria include: Determine if the iteration termination condition is met: if the termination condition is met, proceed to the output stage; if the termination condition is not met, return to the evaluation stage and continue iterative optimization.

10. The gear evaluation method for a gear test bench simulating a space environment according to claim 9, characterized in that: In the optimized WPD customized signal analysis stage, WPD constructs a set of orthogonal wavelet basis functions by performing layer-by-layer filtering and sampling operations on the signal, thereby realizing the multi-scale decomposition of the signal across the entire frequency band. Its basic iterative formula is: ； In the formula, Indicates the first The basis functions of the layer and These are low-pass and high-pass filters, respectively. , satisfying orthogonality; with Construct orthogonal wavelet packets for the initial scaling function, assuming the signal ,but It can be represented as: ； In the formula, Indicates the number of decomposition layers. Indicates frequency band index For the first Layer Wavelet coefficients; These are the basis functions of wavelet decomposition; given the decomposition coefficients of a certain level... At that time, the first The coefficients of the layer can be obtained through the filter coefficients. , The results were obtained from the sampling operation: ； Therefore, the wavelet packet reconstruction is represented as: ； Using the optimal WPD parameter combination obtained through ECAPSO iterative optimization, customized wavelet packet decomposition, hierarchical adaptive denoising, and optimal sub-band reconstruction are performed on the preprocessed gear vibration signal to fully adapt to the signal characteristics under spatial environment coupling. The steps are as follows: Multi-scale wavelet packet decomposition: Based on the optimal number of decomposition layers and the optimal wavelet basis function, the clean signal is decomposed into layers of wavelet packet to obtain the wavelet packet coefficients of each sub-band. Hierarchical adaptive threshold denoising: Based on the optimal sub-band threshold, differentiated denoising strategies are adopted for different frequency sub-bands, which can be determined by ECAPSO optimization: soft thresholding is used for high-frequency sub-bands, and hard thresholding is used for low-frequency sub-bands; the wavelet packet coefficients after denoising are corrected to obtain the clean coefficients; Optimal subband reconstruction: Based on the optimal subband combination, effective feature subbands are selected from the subbands, mainly by eliminating noise-dominant subbands. Inverse wavelet packet transform is performed on the corrected wavelet packet coefficients to obtain the denoised reconstructed signal. This signal retains the core features of spatial gear vibration and the noise is effectively filtered out.

Images