A fatigue bench test method and system based on working device space posture

Abstract

The present application relates to the technical field of fatigue test, provide a kind of fatigue test bench test method and system based on the spatial posture of working device, comprising: collecting target load data under each working condition operation;Working condition at least includes straight push, left inclined push and right inclined push;Obtain cylinder stroke, after obtaining the spatial configuration angle of all components in working device by the three-dimensional space vector equation and constraint equation of working device, the coordinate conversion matrix of the relationship of working device spatial configuration and test bench configuration is constructed, the target load data under each working condition operation is converted into test bench target load, and each working condition proportion is obtained, the test bench target load under each working condition operation is combined, the test bench target load of full working condition is obtained, the excitation signal of virtual actuator is calculated, excitation signal is applied to virtual vibration test bench system, simulate working device full working condition load, carry out fatigue test analysis of each component of working device.Accuracy and reliability of fatigue simulation analysis result are guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of fatigue testing technology, and in particular relates to a fatigue bench testing method and system based on the spatial posture of the working device. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Tracked / wheeled construction machinery is highly susceptible to fatigue failure due to the intense continuous loading cycles and harsh environmental conditions it endures. Therefore, vibration tests on key components are necessary to verify their durability throughout their lifespan. The key to conducting vibration tests is inputting an excitation signal with an equivalent actual working condition load to the actuator. This equivalent load must include the bulldozer's working device's direct and oblique pushing operating conditions. Although fatigue vibration tests for passenger cars have been maturely applied for a long time, their use in construction machinery, such as bulldozers, is extremely limited, mainly due to the following issues: (1) Compared with passenger cars and commercial vehicles, the working device of bulldozers has obvious uncertainties in the position, direction and magnitude of the load on the blade due to the different physical properties such as the accumulation form, volume and density of the working material in front of the blade during actual operation. Current analysis usually simplifies the load on the blade as a constant load that does not change with the working conditions, and cannot reflect the real load information of the actual operation process of the construction machinery. (2) Compared with passenger cars and commercial vehicles, the configuration of the working device of construction machinery changes with the working conditions. That is, the relative angle between each component changes with the working conditions, which leads to the load distribution ratio and transmission path of each component changing with the working conditions. However, the working device is fixed in position and configuration on the fatigue vibration test bench. The working device configuration corresponding to the collected working condition load is inconsistent with the working device configuration on the fatigue test bench, which makes the working condition load collected with the motion working conditions unable to be directly used on the fatigue vibration test bench. (3) Compared with the working devices of loaders and excavators, the working device of bulldozers moves in three-dimensional space. Under the oblique pushing condition, the projection of the bottom edge of the blade on the ground is no longer perpendicular to the forward direction. The bottom edge of the blade forms a spatial configuration with the ground and the normal plane of the forward direction. The working device not only has to overcome the soil resistance in the forward direction and vertical direction, but also has to deal with the lateral force and additional torque generated by the oblique pushing, which significantly increases the stress complexity of the working device. (4) In addition, the oblique pushing condition causes the soil piled in front of the blade to have an asymmetrical distribution compared to the center symmetry plane of the bulldozer body. The lateral component force causes the bulldozer to shift laterally and tilt, generating additional torque, which causes the bulldozer body to deflect and deviate from the predetermined driving line. Therefore, the oblique pushing operation of the bulldozer is difficult to control, and there is a lack of vehicle stability control strategy and operation control strategy for oblique pushing operation simulation. (5) The data measured from the sensors of the engineering machinery components or systems in the working condition load acquisition test are response signals output at a specific position by the external load excitation of the engineering machinery, and cannot be used as the excitation signal of the fatigue vibration test actuator; (6) The spatial structure of the rods of the bulldozer working device is complex, making it difficult to directly collect the spatial configuration angles of each working rod in the working condition test; (7) The working device of the engineering machinery and the working material are mostly in surface contact and the contact situation is complex. The contact area between the vibrating table actuator and the working device is small, which makes it difficult to determine the actuator layout scheme, load spectrum loading method and model boundary condition setting. (8) The fatigue life analysis of the bulldozer working device needs to consider the load of the straight push and inclined push conditions throughout the entire life cycle of the bulldozer. The straight push, left tilt push and right tilt push result in different spatial configurations of the working device. Therefore, it is necessary to calculate the impact of all working conditions of the straight push, left tilt push and right tilt push on the fatigue life of the device.

[0004] In summary, the key to bench analysis lies in how to achieve simulated vehicle stability and unmanned operation through control logic; how to obtain the spatial configuration position of the working device through easily acquired cylinder stroke data; how to convert experimentally acquired variable configuration target data into vibration bench data with a fixed configuration; how to construct a bulldozer working device vibration bench system, including the specific position, quantity, and type of actuators; and how to improve the accuracy of excitation signals. Summary of the Invention

[0005] To address the technical problems mentioned above, this invention provides a fatigue bench test method and system based on the spatial attitude of the working device. Under direct push, left tilt push, and right tilt push conditions, coordinate transformation matrices expressing the relationship between the working device's working condition configuration and the bench configuration are constructed respectively. This enables accurate conversion and calculation of the collected working condition response load and the bench load. Furthermore, by considering the proportion of working conditions, a full working condition load of the working device that integrates the loads of direct push, left tilt push, and right tilt push conditions is proposed. Using the full working condition load as the target object, iterative calculation of the fatigue vibration bench actuator signal is carried out, ensuring the accuracy and reliability of the fatigue simulation analysis results.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The first aspect of the present invention provides a fatigue bench test method based on the spatial posture of a working device, comprising: Collect target load data under various working conditions; the working conditions include at least direct thrust, left tilt thrust, and right tilt thrust. The cylinder stroke is obtained, and the three-dimensional spatial vector equation and constraint equation of the working device are solved nonlinearly to obtain the spatial configuration angles of all components in the working device. Then, a coordinate transformation matrix is ​​constructed to determine the relationship between the spatial configuration of the working device and the configuration of the test bench. The target load data under each working condition is converted into the test bench target load, and the proportion of each working condition is obtained. The test bench target loads under each working condition are combined to obtain the test bench target load for all working conditions. Based on the geometry and hinge relationships of each component of the working device, a virtual vibration test system is constructed. Based on the target load of the test system under full working conditions, the excitation signal of the virtual actuator is calculated. The excitation signal is applied to the virtual vibration test system to simulate the full working load of the working device and to conduct fatigue test analysis of each component of the working device.

[0007] Furthermore, the target load data includes the longitudinal soil load and the vertical soil load on the working device; the axial thrust of the push rod is selected to characterize the longitudinal soil load on the working device, and the driving force of the lifting cylinder is selected to characterize the vertical soil load on the working device.

[0008] Furthermore, the three-dimensional spatial vector equation includes:

[0009] in, l 1. l 2 and l 3 represents the strokes of the left lifting cylinder, right lifting cylinder, and tilting cylinder, respectively. , , , , and These are the rotation matrices of the configuration coordinates of the left push rod, right push rod, left lifting cylinder, right lifting cylinder, tilting cylinder, and blade relative to the global coordinate system; A0, A1, A2, A3, A4, and A5 represent the spatial position coordinates of the rear hinge point of the right push rod, the front hinge point of the right push rod, the rear hinge point of the tilting cylinder, the front hinge point of the tilting cylinder, the lower hinge point of the right lifting cylinder, and the upper hinge point of the right lifting cylinder in the global coordinate system, respectively; B0, B1, B2, B3, B4, and B5 represent the spatial position coordinates of the rear hinge point of the left push rod, the front hinge point of the left push rod, the rear hinge point of the inclined support, the front hinge point of the inclined support, the lower hinge point of the left lifting cylinder, and the upper hinge point of the left lifting cylinder in the global coordinate system, respectively.

[0010] Furthermore, the constraint equations include: ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; in, , and Both represent the spatial distance between the two hinge points; the X-axis of the global coordinate system is represented as... X G The Y-axis is represented as Y G The Z-axis is represented as Z G ; α for B 0 B 5 in X G Z G Plane projection and A 0 B Angle of 0; β for B 0 B 5 and X G Z G Angle between planes; γ For the shovel B 4 B 1 and B 1 A The angle between 1 and 2; ω For the shovel A 3 B 4 andB 4 A The angle between 4; τ For the putter A 2 A 0 and A 0 A The angle between 1 and 2; υ For the putter A 2 A 1 and A 1 A The angle between 0 and 0; φ for B 5 A 0 in X G Z G Plane projection and A 0 B Angle of 0; σ for B 5 A 0 and X G Z G Angle between planes; θ 1. θ 2. θ 3 represents the rotation angle of the three coordinate axes of the left push rod coordinate system; θ 4. θ 5. θ 6 represents the rotation angle of the three coordinate axes of the right push rod coordinate system; θ 7. θ 8 θ 9 represents the rotation angle of the three coordinate axes of the shovel coordinate system; ε 1. ε 2 ε 3 represents the rotation angle of the three coordinate axes of the left lifting cylinder coordinate system; ε 4. ε 5. ε 6 represents the rotation angle of the three coordinate axes of the right lifting cylinder coordinate system; ε 7. ε 8 ε 9 represents the rotation angle of the three coordinate axes of the tilting cylinder coordinate system; l 1 represents the stroke of the left lifting cylinder; l 2 represents the stroke of the right lifting cylinder; l 3 represents the tilting cylinder stroke.

[0011] Furthermore, the conversion of target load data under various working conditions into bench target load specifically includes: acquiring the axial strain of both push rods, calculating the axial load of both push rods, and then, based on the relationship between the push rod coordinate system and the blade coordinate system, calculating the three components of the axial load of both push rods applied to the blade acting at the push rod-blade hinge point in the blade coordinate system; acquiring the hydraulic pressure of the two lifting cylinders, calculating the cylinder axial load, and then, based on the relationship between the lifting cylinder coordinate system and the blade coordinate system, calculating the three components of the axial load of both lifting cylinders applied to the blade acting at the cylinder-blade hinge point in the blade coordinate system.

[0012] Furthermore, the specific method for combining the target loads of the test bench under various operating conditions is as follows: based on the target loads of the test bench under various operating conditions, draw histograms of the mean and amplitude frequency distributions and probability density curves to quantify the load characteristics of each operating condition, and use a weighted fusion method to perform a weighted average of the target loads of the test bench under various operating conditions.

[0013] Furthermore, the construction steps of the virtual vibration table system include: The working device in the joint simulation model is retained, and actuators and sensors are placed on the shovel. The modal method is used to divide the push rod and inclined support components in the working device into flexible bodies, and a force distribution unit is used to realize the connection between the rigid body components and the flexible body. Create virtual actuators and set actuators on each actuator. The actuators are used to apply excitation signals of joint entities or force entity components to the virtual actuators.

[0014] Furthermore, the calculation steps for the excitation signal of the virtual actuator include: Extract the frequency response function that characterizes the virtual vibration test system; Based on the target load of the test bench under all working conditions, the excitation signal is calibrated using the Newton-Raphael algorithm and the inverse function of the frequency response function; The target load and excitation signal of the test bench under all working conditions are compared in the time domain and frequency domain, and the error rate is calculated to determine whether the excitation signal meets the requirements.

[0015] Furthermore, the application of the excitation signal to the virtual vibration table system specifically involves: smoothing the excitation signal using a high-order spline interpolation algorithm to obtain a spline curve showing the change of the excitation load over time, and then applying the smoothed spline curve to the virtual vibration table system.

[0016] A second aspect of the present invention provides a fatigue bench testing system based on the spatial posture of a working device, comprising: The signal determination module is configured to: collect target load data under various operating conditions; the operating conditions include at least direct thrust, left tilt thrust, and right tilt thrust. The load data conversion module is configured to: acquire the cylinder stroke, perform nonlinear solution through the three-dimensional spatial vector equation and constraint equation of the working device to obtain the spatial configuration angle of all components in the working device, construct a coordinate transformation matrix of the relationship between the spatial configuration of the working device and the frame configuration, convert the target load data under each working condition into the frame target load, acquire the proportion of each working condition, combine the frame target loads under each working condition, and obtain the frame target load for all working conditions. The fatigue testing module is configured to: construct a virtual vibration test bench system based on the geometry and hinge relationships of each component of the working device; calculate the excitation signal of the virtual actuator based on the target load of the bench under full working conditions; apply the excitation signal to the virtual vibration test bench system to simulate the full working condition load of the working device; and perform fatigue test analysis on each component of the working device.

[0017] Compared with the prior art, the beneficial effects of the present invention are: This invention constructs coordinate transformation matrices to express the relationship between the working device configuration and the test bench configuration under direct push, left tilt push, and right tilt push conditions, respectively. This enables accurate conversion and calculation of the collected working condition response load and the test bench load. Furthermore, by proposing the full working condition load of the working device by integrating the working condition loads of direct push, left tilt push, and right tilt push conditions through the working condition ratio, and using the full working condition load as the target object, iterative calculation of the fatigue vibration test bench actuator signal is carried out, ensuring the accuracy and reliability of the fatigue simulation analysis results.

[0018] This invention utilizes a virtual vibration table to perform fatigue durability analysis on the working device, which is cheaper than testing on actual equipment, saving testing costs and reducing the risk of damage to actual equipment.

[0019] The virtual vibration test bench of this invention can use computer simulation and modeling techniques to analyze fatigue damage and failure conditions. These techniques can process large amounts of data and simulate various conditions to enable rapid necessary improvements and optimizations, thereby improving the efficiency of fatigue testing and enabling rapid fatigue testing of components.

[0020] The iterative learning algorithm used in this invention is highly efficient, greatly reducing the experimental iteration cycle and resulting in a small error rate in the measured durability performance data of the working device. This improves the accuracy and reliability of fatigue simulation analysis results while enhancing the precision of fatigue testing.

[0021] Once the test conditions are determined, the method of the present invention will be carried out in the same way each time and will not be affected by the environment, thus having good repeatability and improving the reliability of the test results.

[0022] The method of the present invention can be used to obtain excitation signals from vibration test benches of different types and models of engineering machinery. It is applicable to various working conditions and is not affected by the structural details of the test prototype. The types of excitation signals include position, velocity, acceleration, force, stress, etc., and fully considers the feasibility and effectiveness of bench fatigue testing. Attached Figure Description

[0023] 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 improper limitation of the invention.

[0024] Figure 1 A flowchart of a fatigue bench test method based on the spatial posture of a working device provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of a bulldozer operating under multiple conditions provided in Embodiment 1 of the present invention; Figure 3 This is a control logic diagram of the tracked walking system provided in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of the vector equation provided in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the location of the shovel hinge point provided in Embodiment 1 of the present invention; Figure 6 This is a side view of the relationship between the external load and the hinge point load of the working device provided in Embodiment 1 of the present invention; Figure 7 This is a front view of the relationship between the external load and the hinge point load of the working device provided in Embodiment 1 of the present invention; Figure 8 This is a hydraulic cylinder stroke diagram provided in Embodiment 1 of the present invention; Figure 9 This is a schematic diagram of the rotation angle of the rear hinge points of the push rods on both sides in three-dimensional space, provided in Embodiment 1 of the present invention. Figure 10 This is a schematic diagram of the rotation angle of the front hinge point of the lifting cylinder on both sides in three-dimensional space, provided in Embodiment 1 of the present invention. Figure 11 This is a schematic diagram of the virtual vibration table system provided in Embodiment 1 of the present invention; Figure 12 This is a comparison diagram of the virtual sensor response signal obtained from the final iteration provided in Embodiment 1 of the present invention and its corresponding time-domain target signal; Figure 13 This is a comparison diagram of the virtual sensor response signal obtained from the final iteration provided in Embodiment 1 of the present invention and its corresponding frequency domain signal; Figure 14 The following is a trend of change in the error rate of the response signal provided in Embodiment 1 of the present invention; Figure 15This is a schematic diagram of the virtual actuator excitation signal provided in Embodiment 1 of the present invention; Figure 16 The life cloud diagram and damage distribution diagram of the left push rod provided in Embodiment 1 of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0026] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0027] Example 1 This embodiment provides a fatigue bench test method based on the spatial posture of the working device.

[0028] In this embodiment, a bulldozer is used as an example for explanation.

[0029] This embodiment provides a fatigue bench test method based on the spatial posture of the working device, such as... Figure 1 As shown, it includes the following steps: Step 1: Build a multi-condition unmanned control system for bulldozers (configured with a multi-condition operation motion control strategy), and collect target load data (referred to as target signal) under multi-condition operation by means of multi-body multi-field joint simulation (also known as multi-body multi-field integrated simulation).

[0030] Taking the bulldozer's inclined pushing of clay as an example, this working condition includes five stages: lifting and tilting the blade, starting forward, collecting and transporting soil at a fixed depth, lifting and unloading the soil, straightening the blade, and reversing and lowering the blade.

[0031] Among them, multi-condition operation includes at least direct push, left tilt push and right tilt push.

[0032] The target load data includes the longitudinal soil load and the vertical soil load on the working device. Based on the relationship between the material load and the constraint force at the hinge point of the working device during operation, and the load ratio, the axial thrust of the push rod is selected to characterize the longitudinal soil load on the working device, and the driving force of the lifting cylinder is selected to characterize the vertical soil load on the working device. The specific data sources mainly include the following three methods: (1) The target signal is directly collected in a single physical experiment.

[0033] The strain signals of the left push rod and the right push rod, as well as the oil pressure of the left and right lifting cylinders, were collected through physical experiments. The axial thrust of the push rod was calculated from the surface strain, material properties, and cross-sectional area of ​​the push rod, and the driving force of the lifting cylinder was calculated from the oil pressure and cylinder size.

[0034] Among them, the strain signal is a signal collected by attaching strain gauges to a preset area under the condition of inclined pushing clay, which reflects the change of relative deformation of a local area of ​​the target pusher structure over time.

[0035] (2) Obtain the target signal through a single joint simulation.

[0036] Based on the established unmanned control logic for multi-condition operation of bulldozers, the axial thrust of the push rod and the driving force of the hydraulic cylinder are directly output through co-simulation, such as... Figure 2 As shown, it can simulate direct push, left tilt push, and right tilt push conditions.

[0037] Among them, the unmanned control logic realizes the adaptive adjustment of the extension and retraction motion of each cylinder by monitoring two key parameters, track slip rate and blade depth in real time; at the same time, the vehicle body attitude angle is introduced as a control variable to perform differential steering correction on the track.

[0038] In actual operation, the driver primarily relies on visual feedback and pre-planned work paths, adjusting the bulldozer's working devices (including lifting and lowering the shovel) by manipulating control handles. The control strategy essentially utilizes the differential control of the two track drive speeds to generate steering torque, correcting the steering and maintaining a straight working trajectory. Similar to the control strategy in actual operation, the simulation model lacks visual perception, directly generating steering torque through the differential control of the two track drive speeds. It controls the lowering and lifting of the shovel through shovel depth, achieving a control effect equivalent to actual machine operation through the coupled simulation of the hydraulic system and the dynamic model. Figure 3 The diagram shows the track drive force distribution control strategy, based on multi-parameter real-time feedback. The real-time feedback parameters include the left and right variable pump control currents IL and IR, tilt cylinder stroke K, vehicle attitude angle E, slip ratio S, simulation time T, blade depth D, and lifting cylinder stroke L at each time step. The specific track drive force distribution control strategy is as follows: Blade lifting and tilting phase: The bulldozer first falls onto the pellet bed under gravity within 0.5 seconds, aligning the blade tip with the preset working starting point. At this time, all hydraulic cylinders remain locked to ensure parameter stability. Then, it enters the active attitude adjustment phase, where the lifting cylinder retracts at a constant speed to lift the blade. When the blade height above the ground reaches the preset value, it immediately holds the position. Next, the tilting cylinder extends at the same speed until it reaches its maximum stroke of 67mm, driving the blade to the preset right tilt limit position. After tilting adjustment is completed, the lifting cylinder extends and retracts to begin lowering the blade. When the blade depth sensor detects that the center of the lower edge of the blade is above the ground at the preset value, it enters the holding state and prepares for startup. Initial and forward phases: The bulldozer remains stationary for the first second, during which the variable displacement pump maintains a displacement of 0 ml / r (the current controlling the pump displacement is 220 mA). From the first second (i.e., time T > 1, record the current time T as t1), the bulldozer gradually accelerates, and the pump displacement current increases by 220 + 600. The speed change of (T-t1) causes the variable pump displacement to gradually increase from 0 ml / r to 45 ml / r (at which point the current controlling the pump displacement is 345 mA) and remain stable. During the fixed-depth soil collection and transportation phase: the control system dynamically adjusts the extension and retraction of the lifting cylinder by real-time monitoring of the blade depth and track slip ratio to ensure that the blade depth remains within the target range during operation, while strictly controlling the slip ratio below the preset value; when a deviation of less than -1 degree in the heading angle is detected (i.e., E < -1, at which point the bulldozer yaws to the right), and T < 32, the current time T is recorded as t2, and I1 = IL is set for differential speed adjustment. The current of the left pump is adjusted according to I1 - 600. With the speed change of (T-t2), the current of the right pump remains constant. This means the control system automatically adjusts the speed difference between the left and right tracks to correct the trajectory (at this time, the current of the left pump is reduced to 260mA and remains constant, while the current of the right pump remains constant at 345mA; under this current, the bulldozer begins to turn left). When a deviation in heading angle greater than 1 degree is detected (i.e., E>1, at which point the bulldozer yaws to the left), the current time T is recorded as t3. Let I1=IL, and differential speed adjustment is performed. The current of the left pump is adjusted according to I1+600... With the speed change of (T-t3), the current of the right pump displacement remains unchanged. That is, the control system automatically adjusts the speed difference between the left and right tracks to avoid excessive correction of the travel direction (at this time, the current controlling the left pump displacement increases from 260mA to 345mA and then remains unchanged, while the current controlling the right pump displacement remains unchanged at 345mA). After the differential speed adjustment is completed, the bulldozer maintains a constant speed. The two differential speed adjustment processes ensure that the heading angle of the bulldozer fluctuates between -1 degree and 1 degree during the fixed-depth soil collection and transportation stages, maintaining straight-line travel. During the lifting and unloading of soil and the return of the blade to center phase: When the bulldozer has been running for 28 seconds (i.e., T > 28, record the current time T as t4), set I1 = IL and I2 = IR, and begin to decelerate. The rate of change of the current of the left pump displacement is I1 - 600. (T-t4), the rate of change of current in the pump displacement on the right side I2-600 (T-t4), until it reaches zero (at this point, it returns to the initial state, and the current is reduced to 220mA); when the blade depth is less than -0.4 meters (D<=-0.4), it begins to retreat at a constant speed (the current controlling the pump displacement decreases from 220mA to -160mA and then remains constant, and the current time T is recorded as t5); then at the 29th second, the lifting cylinder quickly retracts at a preset speed to perform the unloading operation; at the 31.5th second, the tilting cylinder retracts at a set speed to complete the blade posture reset; Reverse blade lowering phase: After the blade returns to center and is raised to the preset ground clearance height, the bulldozer immediately reverses, i.e., reverses at a constant speed. At this time, the variable displacement pump control current is set to -160mA to achieve reverse drive. After the vehicle speed stabilizes, it begins to decelerate from the 33rd second (i.e., T>33, record the current time T as t6) until it comes to a complete stop (at this time, the control current is 220mA). The pump displacement current changes at a rate of 220+600. (T-t6); When the stop command is triggered, the lifting cylinder immediately extends at a preset speed, so that the blade falls back smoothly to the ground, completing the termination operation of the entire work cycle.

[0039] (3) Combine physical experiments and joint simulations to obtain multiple sets of load data that conform to actual working conditions.

[0040] The two acquisition methods described above can complement each other. Physical test data can be used to verify the co-simulation model. Afterwards, more test data can be obtained through co-simulation by changing the control logic of multiple materials and multiple operating conditions.

[0041] In this embodiment, the target signal duration is 35 seconds. A low-pass filter is used to filter out the portion above 50 Hz, and the signal is saved as a CSV file as an evaluation metric for iterative learning.

[0042] In this implementation, the obtained raw target signal needs to undergo low-pass filtering, de-straining, and trend removal to reduce the influence of interference signals and improve the authenticity of the load spectrum data.

[0043] Step 2: Establish the relationship between the cylinder stroke and the spatial angles of each component of the working device (establish the relationship between the lifting cylinder stroke, the tilting cylinder stroke and the spatial angles of each link of the working device).

[0044] Specifically, using the global coordinate system as a reference, the configuration coordinates of the push rod, blade, and cylinder are established; through the hinge structure and kinematic analysis of the working device, the three-dimensional spatial vector equations and constraint equations of the working device are established; using the known cylinder stroke (such as...) Figure 8 The cylinder stroke diagram shown is used to obtain the spatial configuration angles of all components in the working device by nonlinearly solving the established configuration coordinates using the Newton-Raphson method (NS).

[0045] The established configuration coordinates include: 6 rotations of the push rods on both sides relative to the global coordinates, 3 rotations of the blade relative to the global coordinates, and 9 rotations of the three hydraulic cylinders, totaling 18 coordinate vectors.

[0046] The configuration motion and hinge structure of the inclined thrust working device are as follows: Figure 4 , Figure 5 , Figure 6 and Figure 7 As shown, the blade rotates from the leftmost tilt position to the rightmost tilt position (or moves in the opposite direction), forming a specific included-angle spatial configuration, thereby enabling the working device to move in three-dimensional space. Here, A0, A1, A2, A3, A4, and A5 represent the spatial coordinates of the right push rod rear hinge point, right push rod front hinge point, tilt cylinder rear hinge point, tilt cylinder front hinge point, right lifting cylinder lower hinge point, and right lifting cylinder upper hinge point in the global coordinate system, respectively; B0, B1, B2, B3, B4, and B5 represent the spatial coordinates of the left push rod rear hinge point, left push rod front hinge point, inclined support rear hinge point, inclined support front hinge point, left lifting cylinder lower hinge point, and left lifting cylinder upper hinge point in the global coordinate system, respectively. B 0 、A 0 、 A 5 、B 5. It is fixed on the frame, and its position coordinates in the global coordinate system are known.

[0047] The working device has six three-dimensional spatial vector equations: Under the three-dimensional spatial vectors, the spatial vector from the rear hinge point of the right push rod to the upper hinge point of the right lifting cylinder is constant, so equation (1) is established; the spatial vector from the front hinge point of the right push rod to the upper and lower hinge points of the tilting cylinder is constant, so equation (2) is established; the spatial vector from the rear hinge point of the right push rod to the upper hinge point of the left lifting cylinder is constant, so equation (3) is established; the spatial vector from the upper hinge point of the right lifting cylinder to the upper hinge point of the left lifting cylinder is constant, so equation (4) is established; the spatial vector from the rear hinge point of the left push rod to the upper hinge point of the left lifting cylinder is constant, so equation (5) is established; the spatial vector from the upper hinge point of the right lifting cylinder to the upper hinge point of the left lifting cylinder is parallel to the spatial vector from the front hinge point of the right push rod to the front hinge point of the left push rod, so equation (6) is established. The details are as follows:

[0048] in, l 1.l 2. l 3 represents the strokes of the left lifting cylinder, right lifting cylinder, and tilting cylinder, respectively. These strokes can be measured by stroke sensors and are known quantities. Figure 8 As shown, the lengths of each member are known; , , , , , These are the configuration coordinates of the left push rod, right push rod, left lifting cylinder, right lifting cylinder, tilting cylinder, and blade, respectively, and the rotation matrix of the global coordinate system.

[0049] Each vector equation can be used to establish 3 constraint equations, for a total of 18 constraint equations, as shown in equations (7) to (24): (7); (8); (9); (10); (11); (12); (13); (14); (15); (16); (17); (18); (19); (20); (twenty one); (twenty two); (twenty three); (twenty four); in, , , Both represent the spatial distance between the two hinge points; like Figure 7As shown, the origin of the global coordinate system is chosen at the bulldozer's center of mass. The X-axis points horizontally to the left, the Y-axis points vertically upwards, and the Z-axis points longitudinally along the bulldozer, i.e., the bulldozer's forward direction, with forward being the positive direction. The X-axis of the global coordinate system is represented as... X G The Y-axis is represented as Y G The Z-axis is represented as Z G ; α for B 0 B 5 in X G Z G Plane projection and A 0 B Angle of 0; β for B 0 B 5 and X G Z G Angle between planes; γ For the shovel B 4 B 1 and B 1 A The angle between 1 and 2; ω For the shovel A 3 B 4 and B 4 A The angle between 4; τ For the putter A 2 A 0 and A 0 A The angle between 1 and 2; υ For the putter A 2 A 1 and A 1 A The angle between 0 and 0; φ for B 5 A 0 in X G Z G Plane projection and A 0 B Angle of 0; σ for B 5 A 0 and X G Z G Angle between planes; The origin of the left pusher coordinate system is chosen at the rear hinge point of the left pusher. The X-axis points horizontally to the left, the Z-axis is along the axial direction of the left pusher, and the Y-axis is perpendicular to both the X and Z axes and points upwards. The X-axis of the left pusher coordinate system is represented as... x BO The Y-axis is represented as y BO The Z-axis is represented as z BO ; θ 1. θ 2. θ 3 represents the coordinate system of the left push rod. x BO 、y BO 、z BO Shaft rotation angle; The origin of the right pusher coordinate system is chosen at the rear hinge point of the right pusher. The X-axis points horizontally to the left, the Z-axis is along the axial direction of the right pusher, and the Y-axis is perpendicular to both the X and Z axes and points upwards. The X-axis of the right pusher coordinate system is represented as... x AO The Y-axis is represented as y AO The Z-axis is represented as z AO ; θ 4. θ 5. θ 6 represents the coordinate system of the right push rod. x AO 、y AO 、z AO Shaft rotation angle; The blade coordinate system is set at the blade's center of mass. The X-axis points horizontally to the left, the Z-axis is perpendicular to the blade plane with forward as the positive direction, and the Y-axis is perpendicular to both the X and Z axes and points upward. The X-axis of the blade coordinate system is represented as... x C The Y-axis is represented as y C The Z-axis is represented as z C ; θ 7. θ 8 θ 9 represent the shovel coordinate system. x C 、y C 、z C Shaft rotation angle; The origin of the coordinate system for the left lifting cylinder is chosen at the lower hinge point of the left lifting cylinder. The X-axis points horizontally to the left, the Z-axis is along the axial direction of the left lifting cylinder, and the Y-axis is perpendicular to both the X and Z axes and points forward. The X-axis of the coordinate system for the left lifting cylinder is represented as... xB4 The Y-axis is represented as y B4 The Z-axis is represented as z B4 ; ε 1. ε 2 ε 3 represents the coordinate system of the left lifting cylinder. x B4 、y B4 、z B4 Shaft rotation angle; The origin of the coordinate system for the right lifting cylinder is chosen at the lower hinge point of the right lifting cylinder. The X-axis points horizontally to the left, the Z-axis is along the axial direction of the right lifting cylinder, and the Y-axis is perpendicular to both the X and Z axes and points forward. The X-axis of the right lifting cylinder coordinate system is represented as... x A4 The Y-axis is represented as y A4 The Z-axis is represented as z A4 ; ε 4. ε 5. ε 6 represents the coordinate system of the right lifting cylinder. x A4 、y A4 、z A4 Shaft rotation angle; The origin of the tilting cylinder coordinate system is chosen at the rear hinge point of the tilting cylinder. The X-axis points horizontally to the left, the Z-axis is along the axial direction of the tilting cylinder, and the Y-axis is perpendicular to the X and Z axes, pointing in the forward direction of the bulldozer. The X-axis of the tilting cylinder coordinate system is represented as... x A2 The Y-axis is represented as y A2 The Z-axis is represented as z A2 ; ε 7. ε 8 ε 9 represent the coordinate system of the tilting cylinder. x A2 、y A2 、z A2 Shaft rotation angle; l 1 represents the stroke of the left lifting cylinder; l 2 represents the stroke of the right lifting cylinder; l 3 represents the tilting cylinder stroke.

[0050] Based on the above 18 coordinate vectors and 18 constraint equations, the Newton-Raphson method is used for nonlinear solution to establish the relationship between the lifting cylinder stroke, the tilting cylinder stroke and the configuration coordinates of each link of the working device.

[0051] like Figure 9 and Figure 10 The figures show the calculated angle relationship between the left and right push rods and the lifting cylinder. Figure 9 Let be the rotation angles of the rear hinge points of the push rods on both sides in three-dimensional space, where (a) is the rotation angle of the rear hinge point of the left push rod and (b) is the rotation angle of the rear hinge point of the right push rod. Figure 10 Let be the rotation angle of the front hinge point of the lifting cylinders on both sides in three-dimensional space, where (a) is the rotation angle of the left lifting cylinder and (b) is the rotation angle of the right lifting cylinder.

[0052] Step 3: Construct a coordinate transformation matrix to determine the relationship between the spatial configuration (operating condition configuration) of the working device and the platform configuration, and convert the collected target load data under multiple operating conditions into the platform target load.

[0053] Specifically, based on the axial loads of the push rods on both sides obtained from experimental strain calculations or co-simulation, and based on the relationship between the push rod coordinate system and the blade coordinate system in the global coordinate system, three components of the axial load of the push rod applied to the blade acting at the push rod-blade hinge point in the blade coordinate system are obtained; based on the hydraulic pressure of the lifting cylinder collected from the experiment, or the axial load of the hydraulic cylinder obtained from co-simulation, and based on the relationship between the lifting cylinder coordinate system and the blade coordinate system in the global coordinate system, three components of the axial load of the lifting cylinder applied to the blade acting at the cylinder-blade hinge point in the blade coordinate system are obtained.

[0054] The relationship between the local coordinate systems of the two components in the global coordinate system can be represented by a rotation matrix. R This is illustrated by taking the rear hinge point of the left push rod as an example, rotating it around the three coordinate axes respectively. θ 1. θ 2. θ 3 degrees. Using equations (26) and (27), the original data in the local coordinate system of the left push rod is equivalent to the target load data in the global coordinate system applicable to the fixed configuration vibration table.

[0055] (26); (27); in, F i Force at the rear hinge point of the left push rod in the actual working condition configuration under the local coordinate system; F j This represents the force at the rear hinge point of the left pusher in the global coordinate system. The configuration transformation method at other hinge points is the same as that at the rear hinge point of the pusher, and will not be repeated here. The final calculated target data is as follows: Figure 9 As shown.

[0056] Step 4: Based on the proportion of each working condition (direct push, left tilt push, right tilt push working condition), combine the target load of the working device on the test bench for each working condition to obtain the target load (target signal) of the test bench for all working conditions.

[0057] Specifically, based on the dynamic load survey data of different working conditions under long-term monitoring, mean, amplitude frequency distribution histograms and probability density curves are plotted to quantify the load characteristics of each working condition. The weighted fusion method is used to perform a weighted average on the collected load data. The bulldozer working device load under all working conditions is obtained by fusing the loads of the straight push, left tilt push and right tilt push working conditions.

[0058] Step 5: Construct a virtual vibration table system based on the geometry and hinge relationships of each component of the working device.

[0059] The dynamic model of the working device in the co-simulation model is retained, and virtual actuators and virtual sensors are set on the dynamic model.

[0060] Among them, a multibody dynamics model of the working device is established based on the geometry and hinge relationship of each component of the engineering machinery working device.

[0061] The working device consists of a blade, lifting cylinder, push rod, diagonal support, horizontal support, and tilting cylinder. To ensure accurate transfer of soil load from the blade to each hinge point, this embodiment utilizes the modal method to divide the blade into a flexible body and mesh it, thereby improving analysis speed and ensuring calculation accuracy by increasing the element density. The minimum size of the mesh element is 2mm, and the maximum size is 6mm, resulting in a total of 104,534 nodes. Eight force distribution elements, similar to multi-point constraint equations, are used to connect the rigid body component and the flexible blade.

[0062] The force distribution unit can rigidly connect the master node and all slave nodes, with no relative displacement between the master and slave nodes, thereby accurately transferring the load acting on the master node to all slave nodes.

[0063] Virtual actuators are key components for implementing active vibration control. During construction, the blade is mainly subjected to tangential digging resistance in the Y direction, normal digging resistance in the Z direction, and lateral force in the X direction. These three digging resistances are distributed. In this embodiment, six virtual actuators are established based on the current blade plane coordinate system, with one actuator for each. The actuators apply excitation signals from joint or force components to the corresponding virtual actuators, achieving coordinated correlation of multi-dimensional input signals. This ensures that there are loading axes in all six spatial directions, guaranteeing that the applied excitation signals achieve full spatial freedom of motion. This satisfies the loads on the working device in the X, Y, and Z directions, as well as the torques around the X, Y, and Z axes. The torque is generated by the combination of two loads in the same direction applied by different virtual actuators, thus reproducing the spatial motion of the construction machinery's working device. To facilitate loading during research and fatigue testing, the component forces in each direction are concentrated at the center of the blade, as shown in the specific layout below. Figure 11 As shown, the first virtual actuator 1 and the second virtual actuator 2 provide X-axis excitation, the third virtual actuator 3 and the fourth virtual actuator 4 provide Y-axis excitation, and the fifth virtual actuator 5 and the sixth virtual actuator 6 provide Z-axis excitation.

[0064] Virtual sensors are used to acquire iterative response signals and compare them with target signals to detect the convergence of the two signals. This embodiment establishes four virtual sensors. The layout of the virtual sensors, including their position, type, and number, is determined based on the relationship between the material load and the constraint force at the hinge point of the working device, obtained from load identification, and the load ratio. Their positions correspond to the acquisition positions of the target signals. Each sensor can acquire loads in three directions at the hinge point, resulting in a total of 12 load signals. The specific layout is as follows... Figure 11 As shown, the first virtual sensor 7 and the second virtual sensor 8 are defined at the rear hinge points of the left and right push rods, respectively, and the third virtual sensor 9 and the fourth virtual sensor 10 are defined at the left and right lifting cylinders to collect the lifting cylinder drive data.

[0065] Step 6: Based on the full-condition test bench target load data, obtain the excitation signal of the virtual actuator. Specifically, based on the converted target load data and the inverse function of the test bench frequency response function, use iterative learning error control to obtain the actuator excitation signal.

[0066] The 12 loads at the blade-push rod hinge point and the blade-lifting cylinder hinge point are used as target signals, and the actuator excitation signal is obtained by iterative learning control calculation.

[0067] Iterative learning error control is a method that uses the known response of the entire system to inversely calculate the input signal, i.e., the external driving load, that causes the response. It uses white noise signal and response to calculate the frequency response function and its inverse function of the virtual vibration test system; then, it uses the inverse function of the frequency response function and the target signal to inversely calculate the excitation signal of the actuator.

[0068] To ensure accurate reproduction of the load input signal and avoid the influence of nonlinear systems on the target signal, the Newton-Raphael algorithm is used to calibrate the excitation signal multiple times during the calculation process. (28); In the formula, u ( i () represents the excitation signal from the previous step; u ( i+ 1) This is the updated excitation signal; u re ( i The input correction signal can be calculated using equation (29): (29); (30); In the formula, R -1 Let be the inverse frequency response function of the vibration test bench; e ( i ) is the target signal y tar Difference between the response and the response; G(i) A diagonal matrix is ​​used to control convergence. Since the movements of the bulldozer's actuators are not independent, they affect the load signal of each sensor. G (i) To account for the interaction effect, a 6×12 matrix is ​​used. The algorithm terminates when the simulation results converge with the target signal, outputting the final input signal applied to the virtual actuator. u ( i +1).

[0069] After the iteration is completed, the response signal obtained by the iteration is further judged to see if it is consistent with the target signal. The comparison is mainly made from two aspects: time domain and frequency domain comparison and error rate change trend. The simulation iteration response is checked to see if it meets the requirements in order to obtain a high-accuracy excitation signal.

[0070] like Figure 12 The image shows the final virtual sensor time-domain response signal and its comparison with the corresponding time-domain target signal, specifically the comparison results of the axial force (Z-direction) of the left and right push rods and the driving force (Z-direction) of the left and right lifting cylinders. Figure 12 (a) in the image represents the comparison of signals from the first virtual sensor. Figure 12 (b) in the image represents the comparison of signals from the second virtual sensor. Figure 12 (c) in the figure represents the signal comparison of the third virtual sensor. Figure 12 (d) in the figure represents the comparison of the signals from the fourth virtual sensor. It can be seen that the amplitude and trend of the time-domain comparison between the two are basically the same, and the difference between the two tends to be 0.

[0071] like Figure 13 The image shows the final iteration of the virtual sensor's frequency domain response signal and its comparison with the corresponding frequency domain target signal, specifically the comparison results of the axial force (Z-direction) of the left and right push rods and the driving force (Z-direction) of the left and right lifting cylinders. Figure 13 (a) in the image represents the comparison of signals from the first virtual sensor. Figure 13 (b) in the image represents the comparison of signals from the second virtual sensor. Figure 13 (c) in the figure represents the signal comparison of the third virtual sensor. Figure 13 (d) in the diagram represents the comparison of the signals from the fourth virtual sensor. It can be seen that the amplitudes and trends of the two signals in the frequency domain are basically the same, and the differences between them are close.

[0072] The change in error rate δ is calculated using equation (31), and the curve is shown below. Figure 14 As shown, the error rate gradually decreases during the iteration process, eventually approaching 10%, which meets the fatigue error requirement.

[0073] (31).

[0074] The virtual actuator excitation signal obtained by the above virtual iteration method is as follows: Figure 15 As shown, these are the excitation signals of the six actuators when a bulldozer is pushing clay at an angle. Figure 15 In the diagram, (a) and (b) represent the loads to which the first and second virtual actuators apply X-axis excitation signals. Figure 15 In the diagram, (c) and (d) represent the loads for applying Y-axis excitation signals to the third and fourth virtual actuators. Figure 15 In the diagram, (e) and (f) represent the loads for applying Z-axis excitation signals to the fifth and sixth virtual actuators.

[0075] Step 7: Smooth the obtained actuator excitation signal to obtain a spline curve of the excitation load changing over time, specifically including: The acquired actuator excitation signal has a low frequency and contains unreasonable amplitude anomalies, which can affect the fatigue simulation results when applied to the actuator. Therefore, a high-order spline interpolation algorithm is used to connect the data points. By adaptively adjusting the interpolation density, over-computation is avoided while ensuring the accuracy of curve fitting, eliminating the influence of abnormal disturbances on the fatigue simulation results, and ensuring the physical rationality of the excitation load curve.

[0076] In this embodiment, a smoothed load spline curve is applied to the virtual vibration table system to simulate the full-condition load of the bulldozer's working device, and fatigue testing and analysis are performed on each component of the working device. Figure 16 The diagram shows the life cloud map and damage distribution of the left push rod.

[0077] It is understandable that fatigue life analysis of various components of engineering machinery can be performed using commercial software, and no specific limitations are made here.

[0078] Example 2 This embodiment provides a fatigue bench testing system based on the spatial posture of the working device, which specifically includes: The signal determination module is configured to: collect target load data under various operating conditions; the operating conditions include at least direct thrust, left tilt thrust, and right tilt thrust. The load data conversion module is configured to: acquire the cylinder stroke, perform nonlinear solution through the three-dimensional spatial vector equation and constraint equation of the working device to obtain the spatial configuration angle of all components in the working device, construct a coordinate transformation matrix of the relationship between the spatial configuration of the working device and the frame configuration, convert the target load data under each working condition into the frame target load, acquire the proportion of each working condition, combine the frame target loads under each working condition, and obtain the frame target load for all working conditions. The fatigue testing module is configured to: construct a virtual vibration test bench system based on the geometry and hinge relationships of each component of the working device; calculate the excitation signal of the virtual actuator based on the target load of the bench under full working conditions; apply the excitation signal to the virtual vibration test bench system to simulate the full working condition load of the working device; and perform fatigue test analysis on each component of the working device.

[0079] It should be noted that each module in this embodiment corresponds one-to-one with each step in Embodiment 1, and their specific implementation processes are the same, so they will not be repeated here.

[0080] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A fatigue bench test method based on the spatial posture of a working device, characterized in that, include: Establish a multi-condition unmanned control system for bulldozers and collect target load data under various operating conditions. The working conditions include at least direct push, left tilt push, and right tilt push; the target load data includes the longitudinal soil load and the vertical soil load on the working device; the axial thrust of the push rod is selected to characterize the longitudinal soil load on the working device, and the driving force of the lifting cylinder is selected to characterize the vertical soil load on the working device; The relationship between the cylinder stroke and the spatial angles of each component of the working device is established. The cylinder stroke is obtained, and nonlinear solutions are performed using the three-dimensional spatial vector equations and constraint equations of the working device to obtain the spatial configuration angles of all components. A coordinate transformation matrix is ​​then constructed to determine the relationship between the spatial configuration of the working device and the platform configuration. The target load data under various operating conditions are converted into target loads on the platform. Specifically, this includes: obtaining the axial strain of the push rods on both sides, calculating the axial loads of the push rods on both sides, and then, based on the relationship between the push rod coordinate system and the blade coordinate system, calculating the axial loads of the push rods on both sides applied to the blade at the push rod-blade hinge point in the blade coordinate system. The system calculates the hydraulic pressure of the two lifting cylinders, calculates the axial load of the cylinders, and then calculates the three components of the axial load of the two lifting cylinders acting on the cylinder-blade hinge point in the blade coordinate system, based on the relationship between the lifting cylinder coordinate system and the blade coordinate system. It also obtains the proportion of each working condition and combines the target load of the test bench under each working condition. Specifically, based on the target load of the test bench under each working condition, it plots the mean, amplitude frequency distribution histograms and probability density curves to quantify the load characteristics of each working condition. Using a weighted fusion method, it performs a weighted average of the target load of the test bench under each working condition to obtain the target load of the test bench under all working conditions. Based on the geometry and hinge relationships of each component of the working device, a virtual vibration test bench system is constructed. The construction steps include: retaining the working device in the co-simulation model and placing actuators and sensors on the blade; using the modal method, dividing the push rod and inclined support components in the working device into flexible bodies, and using force distribution units to connect the rigid body components and flexible bodies; establishing virtual actuators, setting actuators on each actuator to apply excitation signals from joint entities or force entities to the virtual actuators; and calculating the excitation signals of the virtual actuators based on the target load of the test bench under full working conditions. The calculation steps include: extracting and characterizing the virtual... The frequency response function of the vibration test system is determined. Based on the target load of the test bench under full working conditions, the excitation signal is calibrated using the Newton-Raphael algorithm and the inverse function of the frequency response function. The target load of the test bench under full working conditions and the excitation signal are compared in the time and frequency domains, and the error rate is calculated to determine whether the excitation signal meets the requirements. The excitation signal is applied to the virtual vibration test system, specifically by using a high-order spline interpolation algorithm to smooth the excitation signal and obtain the spline curve of the excitation load changing with time. The smoothed spline curve is then applied to the virtual vibration test system. The full working condition load of the working device is simulated, and fatigue test analysis of each component of the working device is performed.

2. A fatigue bench testing system based on the spatial attitude of a working device, employing the fatigue bench testing method based on the spatial attitude of a working device as described in claim 1, characterized in that... include: The signal determination module is configured to collect target load data under various operating conditions. The operating conditions include at least direct push, left tilt push, and right tilt push; The load data conversion module is configured to: acquire the cylinder stroke, perform nonlinear solution through the three-dimensional spatial vector equation and constraint equation of the working device to obtain the spatial configuration angle of all components in the working device, construct a coordinate transformation matrix of the relationship between the spatial configuration of the working device and the frame configuration, convert the target load data under each working condition into the frame target load, acquire the proportion of each working condition, combine the frame target loads under each working condition, and obtain the frame target load for all working conditions. The fatigue testing module is configured to: construct a virtual vibration test bench system based on the geometry and hinge relationships of each component of the working device; calculate the excitation signal of the virtual actuator based on the target load of the bench under full working conditions; apply the excitation signal to the virtual vibration test bench system to simulate the full working condition load of the working device; and perform fatigue test analysis on each component of the working device.

Images