Hydraulic motor plunger-roller friction pair hydrostatic oil groove and self-evolution design method thereof

Abstract

This invention discloses a hydrostatic oil groove for a hydraulic motor plunger-roller friction pair and its self-evolutionary design method. Aiming at the globally optimal configuration of the hydrostatic oil groove through self-evolutionary design, a hydrostatic oil groove with high load-bearing capacity and superior lubrication is designed on the interface of the hydraulic motor plunger-roller friction pair, exhibiting a globally optimal configuration. A threshold-truncated Gaussian kernel function is used to describe any oil groove profile, achieving a mathematical representation of the oil groove morphology. Subsequently, an AI-assisted generalized pattern search algorithm is used to iteratively update the oil groove configuration based on contact force data fed back from the lubrication model, achieving self-evolutionary optimization of load-bearing capacity, and ultimately obtaining the globally optimal hydrostatic oil groove. This invention, through a self-evolutionary design method, designs hydrostatic oil grooves with different parameters on the interface of the hydraulic motor plunger-roller friction pair, achieving a globally optimal self-evolutionary design of the hydrostatic oil groove, reducing frictional torque, increasing load-bearing capacity, and significantly improving plunger life.

Description

Technical Field

[0001] This invention relates to the field of hydraulic motors, and specifically to a hydrostatic oil groove for a piston-roller friction pair in a hydraulic motor and its self-evolutionary design method. Background Technology

[0002] The hydrostatic grooves in a hydraulic motor plunger-roller friction pair are key structures affecting the load-bearing, lubrication, and wear performance of the friction pair. The grooves provide necessary hydrostatic support to the friction pair, reducing the contact pressure of local micro-protrusions at the interface. Furthermore, they promote lubricating oil film formation and heat dissipation, thereby reducing friction and improving wear characteristics. The groove configuration has a significant impact on its load-bearing capacity.

[0003] Current research on the design and optimization of hydrostatic oil groove configurations is mostly based on empirical configuration design, biomimetic configuration design, and parameter optimization methods. These methods rely heavily on the designer's experience and inspiration, and theoretically, it is difficult to obtain a globally optimal oil groove configuration. In order to achieve a globally optimal self-evolutionary design of the hydrostatic oil groove, reduce frictional torque, improve load-bearing capacity, and significantly extend plunger life, this invention provides a self-evolutionary design method for the hydrostatic oil groove of the hydraulic motor plunger-roller friction pair. Summary of the Invention

[0004] The technical problem to be solved by this invention is to address the shortcomings of the prior art by proposing a self-evolution design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair, comprising the following steps: S1: Anchor points are set at equal intervals on the rectangular design domain and assigned values ​​to form a control surface. The description function between the anchor points and the oil groove profile is established by using the threshold truncation method of the control surface to obtain the mathematical representation of the oil groove shape. Then, the optimization conditions are set with minimizing the contact force of the roller-piston pair as the optimization objective. S2: Based on the mathematical representation and optimization conditions of the oil groove morphology in S1, the initial oil groove configuration and its contact force are obtained, and the generalized pattern search algorithm is used to iteratively solve the oil groove configuration and its contact force. S3: Based on the oil groove configuration and its contact force data in S2, when the number of recorded oil groove configurations exceeds the threshold MG, the online deep learning AI model adds perturbation to the input through the perturbation layer, and evolves different new oil groove configurations with the current optimal solution as a template, and predicts the contact force of the new oil groove configuration. S4: The generalized pattern search algorithm is used to poll and optimize several new oil tanks evolved from the online deep learning AI model in S3. When the number of oil tanks polled by the generalized pattern search algorithm increases by a preset value, the current self-perturbed online deep learning AI model is updated once. S5: Online deep learning AI model iterative updates determine the final shape of the oil tank.

[0005] Furthermore, in step S1, a small number of anchor points are evenly distributed at equal intervals throughout the rectangular design domain, and the minimum distance between adjacent anchor points is selected to be less than the longest side of the design domain.

[0006] Furthermore, in step S1, each anchor point is assigned an anchor point value, and all anchor point values ​​are connected to form a control surface; a threshold plane is established, and the control surface intersects with the threshold plane to form an intersection line, which is the outline of the oil groove. Changing the anchor point assignment value yields a clear outline of any oil groove.

[0007] Furthermore, in step S1, the optimization conditions are set based on the machinability and sealing performance of the left and right boundaries of the oil tank and the narrowest distance between adjacent oil tanks.

[0008] Furthermore, in step S2, the initial anchor point value is converted into the initial oil tank configuration using the oil tank description function, and the initial contact force is obtained by solving the initial oil tank configuration. The initial oil tank configuration and contact force are used as the initial base point for the iteration of the generalized pattern search algorithm.

[0009] Furthermore, in step S2, in the j-th iteration before the online deep learning AI model intervenes, the generalized pattern search algorithm will use the anchor value of the (j-1)-th iteration. As a base point, at the base point Within the N-dimensional neighborhood, 2N test points are polled in turn. The contact force corresponding to each oil tank test configuration is obtained by solving the oil tank configuration. The oil tank configuration with the smallest contact force is selected as the optimal solution for the j-th iteration.

[0010] Furthermore, in step S3, the optimal solution of the j-th iteration in the online deep learning AI model is used as the model input in the form of an image. Perturbations are actively added to the input to generate a new image, which evolves into a new oil tank configuration.

[0011] Furthermore, in step S3, a dynamic neighborhood particle swarm optimization algorithm is designed in the online deep learning AI model to iteratively adjust the perturbation mixing in the input based on the predicted contact force, so that the contact force of the evolved oil tank configuration is as small as possible.

[0012] Furthermore, in step S4, the online deep learning AI model can evolve an optimal oil tank configuration in each run, and the optimal oil tank configuration obtained in each run is submitted to the generalized pattern search algorithm for analysis.

[0013] Furthermore, a hydrostatic oil groove for a hydraulic motor plunger-roller friction pair is designed using the aforementioned self-evolutionary design method.

[0014] Compared with the prior art, the present invention achieves the following technical effects: 1. A novel oil groove was designed using a new self-evolutionary design method for the hydrostatic oil groove of the hydraulic motor plunger-roller friction pair. Compared with the traditional oil groove designed based on experience and inspiration, this design greatly improves the load-bearing capacity and extends the plunger life.

[0015] 2. A novel oil groove was designed using a new self-evolutionary design method for the hydrostatic oil groove of the hydraulic motor plunger-roller friction pair. Compared with the traditional oil groove designed based on experience and inspiration, it provides better lubrication, reduces friction torque by up to 88%, and extends plunger life.

[0016] 3. By using a new self-evolutionary design method for the hydrostatic oil groove of the hydraulic motor plunger-roller friction pair, a novel oil groove is designed. Compared with the traditional oil groove designed based on experience and inspiration, the design process is standardized and reasonable, and the designed oil groove has a globally optimal solution. Attached Figure Description

[0017] For ease of explanation, the present invention will be described in detail below with reference to specific embodiments and accompanying drawings.

[0018] Figure 1 This application provides a design flow diagram of a hydrostatic oil groove for a hydraulic motor plunger-roller friction pair. Figure 2 The evolution process of the contact force Fc objective function and the corresponding oil groove configuration; Figure 3 This application provides a structural diagram of a hydrostatic oil groove for a novel hydraulic motor plunger-roller friction pair. Figure 4 (a)-(b) are two structural diagrams of a traditional hydraulic motor plunger-roller friction hydrostatic oil groove; Figure 5 A comparison diagram of the simulation results of a novel hydraulic motor plunger-roller friction pair hydrostatic oil groove and a traditional hydrostatic oil groove provided in this application. Detailed Implementation

[0019] The following are specific embodiments of the present invention, described in conjunction with the accompanying drawings, to further illustrate the technical solutions of the present invention. However, the present invention is not limited to these embodiments. Specific details, such as particular configurations, are provided in the following description merely to aid in a comprehensive understanding of the embodiments of the present invention. Therefore, those skilled in the art should understand that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention.

[0020] The following description, with reference to the accompanying drawings, further illustrates the present invention. The present invention provides a hydrostatic oil groove for a piston-roller friction pair in a hydraulic motor and its self-evolutionary design method. Taking the optimized design of the hydrostatic oil groove for the piston-roller friction pair in a CRM-HA50 cam hydraulic motor as an example, as follows... Figure 1 As shown, it includes the following steps: S1: The inner curve sets the anchor points at equal intervals on the rectangular design domain and assigns values ​​to them to form a control surface. The mapping relationship between the anchor points and the oil groove profile is established by using the threshold truncation method of the control surface, so as to realize the mathematical representation of the oil groove shape. Then, the corresponding optimization conditions are set with minimizing the contact force of the roller-piston pair as the optimization objective.

[0021] Furthermore, in step S1, a small number of anchor points are evenly distributed at equal intervals throughout the rectangular design domain, and the minimum distance between adjacent anchor points is selected to be at least 12.5% ​​of the longest side of the design domain. While ensuring optimization accuracy, considering the time and difficulty of optimization solution, the minimum interval between adjacent anchor points can be selected as xs2 / 10, and the anchor point set Ps contains 11×6 anchor points.

[0022] Furthermore, in step S1, a Gaussian kernel function is used to assign an anchor value φ to each anchor point. si By connecting all anchor point values, a control surface can be formed. A threshold plane is established such that the control surface intersects with the threshold plane to form an intersection line, which is the outline of the oil groove. Changing the anchor point values ​​can yield a clear outline of any oil groove.

[0023] Furthermore, in step S1, corresponding optimization conditions should be established to ensure the actual machinability of the oil groove. To ensure that its left and right boundaries are machinable, the designed oil groove must only appear within the effective area, and the effective area limits the left and right span of the oil groove to no more than 130°; to ensure sufficient sealing, the minimum distance between the oil groove and the friction pair must also be limited, and the minimum distance between the oil groove and the friction pair is limited to 2.5 mm; to ensure the narrowest distance of the oil groove, the minimum grid size of the oil groove configuration also needs to be considered, and the minimum grid size of the oil groove configuration is set to 1 mm.

[0024] S2: Based on the mathematical representation and optimization conditions of the oil tank in S1, the initial oil tank configuration and its contact force are first obtained. Then, the generalized pattern search algorithm (GPS) is used to iteratively solve the oil tank configuration and its contact force. In each iteration, all oil tank configurations and their contact forces queried by the generalized pattern search algorithm are recorded in the database.

[0025] Furthermore, in step S2, the initial anchor point value is determined using the oil tank description function. The initial oil groove configuration is converted to G0. The initial contact force is obtained by solving the initial oil groove configuration. The initial oil groove configuration and the contact force are used as the initial base point for the iteration of the generalized pattern search algorithm.

[0026] Furthermore, in step S2, in the j-th iteration before AI intervention, the generalized pattern search algorithm will use the anchor value of the (j-1)-th iteration. As a base point, 2N test points are generated in the neighborhood of this base point, i.e., in the N-dimensional hypersphere space. The contact force Fc corresponding to each oil tank test configuration is obtained by solving the oil tank configuration. The oil tank configuration with the minimum contact force is selected as the optimal solution for the j-th iteration, and its corresponding anchor point value is used as the base point for the next iteration. Then, the generalized pattern search algorithm begins its (j+1)th iteration. In each iteration, each oil tank configuration and its contact force that the generalized pattern search algorithm has polled is stored in the database.

[0027] S3: Based on the oil groove configuration and its contact force data in S2, when the number of recorded oil groove configurations exceeds the threshold MG, the online deep learning AI model assistance function is activated, and a perturbation layer is set at the input. Add perturbations to the current optimal solution. Different new oil tank configurations are derived from the template, and then the contact force Fc of these new oil tank configurations is predicted using an online deep learning AI model.

[0028] Furthermore, in step S3, the deep learning AI model specifically employs a convolutional neural network G-Net designed based on the characteristics of the oil tank configuration. This network includes one perturbation layer, four convolutional layers, one pooling layer, and three fully connected layers. The perturbation layer is specifically designed to achieve abrupt changes in the oil tank configuration. Additionally, the last fully connected layer of G-Net directly outputs the contact force Fc corresponding to the oil tank configuration. The convolutional layers use large-size convolutional kernels of 11×11 and 7×7, which have a larger effective receptive field, i.e., a wider field of view. The current optimal solution... Using an image as the model input, and actively adding perturbations (image noise) to the input, a new image is generated, which evolves into a new oil tank configuration.

[0029] Furthermore, in step S3, a swarm intelligence algorithm is designed in the online deep learning AI model to minimize the contact force Fc of the evolved oil tank configuration. The contact force Fc output by G-Net is input into the swarm intelligence algorithm, specifically the Dynamic Neighborhood Particle Swarm Optimization (DNPSO) algorithm. Each perturbation is considered an individual and corresponds to a mutant oil tank configuration, and all perturbations form a population. If the contact force of the oil tank configuration generated by the k-th perturbation is less than the previous minimum contact force, then the k-th perturbation is the minimum contact force discovered so far, representing the current optimal perturbation. The perturbation mixture in the input is iteratively adjusted according to the predicted contact force Fc.

[0030] S4: Based on the multiple new oil tanks evolved by AI in S3, these new oil tanks are fed back to the generalized pattern search algorithm for polling and optimization. Whenever the number of oil tanks polled by the generalized pattern search algorithm increases by a preset value Mn, the current self-perturbed online deep learning AI model is updated.

[0031] Furthermore, in step S4, considering that the neural network prediction accuracy cannot reach 100% and the DNPSO algorithm may get stuck in local optima, the online deep learning AI model can evolve an optimal oil tank configuration in each run, and submit the optimal oil tank configuration obtained in each run to the generalized pattern search algorithm for analysis.

[0032] S5: As Figure 2 As shown, the oil tank configuration is optimized with the number of iterations. When the contact force fluctuation is less than a set threshold after several consecutive iterations and there is no significant room for optimization of the oil tank configuration, the optimization is considered to have converged, and the final shape of the oil tank is determined. Figure 3 The novel hydraulic motor plunger-roller friction pair hydrostatic oil groove structure is shown.

[0033] The final oil tank designed in this embodiment is as follows: Figure 3 The novel hydraulic motor plunger-roller friction pair hydrostatic oil groove structure shown is similar to... Figure 4 Simulation comparisons were performed on the traditional oil tanks shown in (a) and (b) (experience-designed oil tanks provided by manufacturers S and Z). Figure 5 As shown, the results indicate that the hydrostatic oil groove designed in this invention can reduce frictional torque by up to 88%, improve load-bearing capacity, and extend plunger life.

[0034] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other.

[0035] Those skilled in the art to which this application pertains may make various modifications or additions to the specific embodiments described, or adopt similar methods to replace them, without departing from the inventive concept of this application or exceeding the scope defined by the appended claims.

Claims

1. A self-evolutionary design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair, characterized in that, Includes the following steps: S1: Anchor points are set at equal intervals on the rectangular design domain and assigned values ​​to form a control surface. The description function between the anchor points and the oil groove profile is established by using the threshold truncation method of the control surface to obtain the mathematical representation of the oil groove shape. Then, the optimization conditions are set with minimizing the contact force of the roller-plunger pair as the optimization objective. S2: Based on the mathematical representation and optimization conditions of the oil groove morphology in S1, the initial oil groove configuration and its contact force are obtained, and the generalized pattern search algorithm is used to iteratively solve the oil groove configuration and its contact force. S3: Based on the oil groove configuration and its contact force data in S2, when the number of recorded oil groove configurations exceeds the threshold, the online deep learning AI model adds perturbation to the input through a perturbation layer, and evolves different new oil groove configurations using the current optimal solution as a template, predicting the contact force of the new oil groove configuration. S4: The new oil tank configurations evolved from the online deep learning AI model in S3 are fed to the generalized pattern search algorithm for polling and optimization. When the number of oil tanks polled by the generalized pattern search algorithm increases by a preset value, the current self-perturbed online deep learning AI model is updated once. S5: Online deep learning AI model iterative updates determine the final shape of the oil tank; In step S1, a small number of anchor points are evenly distributed at equal intervals throughout the rectangular design domain, and the minimum distance between adjacent anchor points is selected to be less than the longest side of the design domain. In step S2, the initial anchor point value is converted into the initial oil tank configuration using the oil tank description function, and the initial contact force is obtained by solving the initial oil tank configuration. The initial oil tank configuration and contact force are used as the initial base point for the iteration of the generalized pattern search algorithm. In step S2, during the j-th iteration before the online deep learning AI model intervenes, the generalized pattern search algorithm sets the anchor value of the (j-1)-th iteration. As a base point, at the base point Within the N-dimensional neighborhood, 2N test points are polled in turn. The contact force corresponding to each oil tank test configuration is obtained by solving the oil tank configuration. The oil tank configuration with the smallest contact force is selected as the optimal solution for the j-th iteration.

2. The self-evolution design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair according to claim 1, characterized in that, In step S1, each anchor point is assigned an anchor point value, and all anchor point values ​​are connected to form a control surface; a threshold plane is established, and the control surface intersects with the threshold plane to form an intersection line, which is the outline of the oil groove. Changing the anchor point assignment value yields a clear outline of the oil groove.

3. The self-evolution design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair according to claim 1, characterized in that, In step S1, the optimization conditions are set based on the machinability and sealing performance of the left and right boundaries of the oil tank and the narrowest distance between adjacent oil tanks.

4. The self-evolution design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair according to claim 1, characterized in that, In step S3, the optimal solution of the j-th iteration is used as an image input in the online deep learning AI model. Perturbations are actively added to the input to generate a new image, which evolves into a new oil tank configuration.

5. The self-evolution design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair according to claim 1, characterized in that, In step S3, a dynamic neighborhood particle swarm optimization algorithm is designed in the online deep learning AI model to iteratively adjust the perturbation mixing in the input based on the predicted contact force, so that the contact force of the evolved oil tank configuration is as small as possible.

6. The self-evolution design method for the hydrostatic oil groove of a hydraulic motor plunger-roller friction pair according to claim 1, characterized in that, In step S4, the online deep learning AI model can evolve an optimal oil tank configuration in each run, and the optimal oil tank configuration obtained in each run is submitted to the generalized pattern search algorithm for analysis.

7. A hydrostatic oil groove for a hydraulic motor plunger-roller friction pair, characterized in that, It is designed using the self-evolutionary design method as described in any one of claims 1-6.

Images