Multi-parameter collaborative optimization method for deterministic polishing of confined space robot

Abstract

The application discloses a multi-parameter collaborative optimization method for deterministic polishing of a robot in a confined space, and the method is characterized in that: through a multi-step collaborative optimization framework, a material removal amount prediction model, a polishing track, a robot posture and dynamic error compensation are integrated into a unified optimization process, the nonlinear coupling influence of robot posture adjustment and feed speed fluctuation on the removal amount distribution is reduced, the collaborative problems of removal amount out-of-tolerance, coverage uniformity and interference avoidance caused by multi-parameter coupling in a narrow flow channel are solved, and finally, the high-dimensional nonlinear optimization problem is decomposed through an alternating direction multiplier method, the convergence efficiency is significantly improved, the processing surface can meet the preset tolerance requirement, and the deterministic polishing with high precision in the confined space is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of complex integral component polishing technology, and in particular to a multi-parameter collaborative optimization method for deterministic polishing of a robot in a restricted space. BACKGROUND

[0002] Limitations of traditional polishing process: The polishing of complex integral components (such as turbine blades of an aero-engine and turbine blades of a nuclear reactor cooling pump) has long relied on manual operation, which has problems such as low efficiency, poor consistency, and uncontrollable surface quality, and is difficult to meet the needs of high-precision and batch production.

[0003] Industrial robots are gradually applied to the polishing of complex curved surfaces due to their high flexibility and flexibility, but they still face the following problems due to the narrow flow channel, large curvature change, and other working conditions: (1) Inaccurate material removal model: the traditional Hertz contact theory fails to work under the nonlinear working condition of side blade contact, resulting in large prediction error of removal depth. (2) Insufficient trajectory planning capability: existing trajectory planning methods (such as line cutting and cycloid) do not coordinate the tool path, robot pose, and process parameters, resulting in fluctuations in dwell time and removal amount.

[0004] Deterministic polishing is a method that uses a high-precision material removal amount prediction model to optimize the polishing trajectory and dwell time, etc., to control the material removal amount. However, in complex integral components such as integral turbine blades and turbine disks, the motion space of the robot for polishing is severely limited, making it difficult to accurately control multiple parameters. SUMMARY

[0005] To solve the above problems, the present application provides a multi-parameter collaborative optimization method for deterministic polishing of a robot in a restricted space, which optimizes the robot polishing trajectory, pose, and dwell time under the driving of material removal simulation, and is expected to improve the polishing removal accuracy and achieve higher quality polishing.

[0006] To solve the above technical problems, the technical solution of the present application is as follows:

[0007] A multi-parameter collaborative optimization method for deterministic polishing of a robot in a restricted space, comprising the following steps:

[0008] Step 1: For the target narrow flow channel, a material removal amount prediction model based on a modified Preston equation and a side blade contact mechanics model is established;

[0009] Step 2: The target workpiece model is planned using a line cutting or ring cutting trajectory to generate a polishing trajectory of the center point of the side blade tool;

[0010] Step 3, discretize the polishing trajectory to obtain discrete machining points, and then establish a robot non-interference configuration space corresponding to each discrete machining point under the constraint of non-collision machining, and perform smooth non-collision pose optimization along the polishing trajectory in the robot non-interference configuration space to obtain a robot pose change sequence;

[0011] Step 4, calculate the material removal amount according to the polishing trajectory and the robot pose change sequence, and in combination with the material removal amount prediction model.

[0012] Step 5, optimize and adjust the residence time of each discrete machining point in combination with the material removal amount requirement required by the task, and calculate the polishing speed of the robot according to the residence time;

[0013] Step 6, input the optimized robot pose change sequence, residence time and polishing speed into the material removal amount prediction model to predict the material removal amount of the target workpiece model, generate a topography simulation result, and repeatedly perform steps 2-5 using the alternating direction multiplier method based on the topography simulation result until a preset material removal amount tolerance is met.

[0014] In some embodiments, in step 2, a computational conformal geometry method is introduced, and different polishing trajectories are optimized using an adaptive algorithm.

[0015] In some embodiments, in step 2, the polishing trajectory planning strategy includes:

[0016] The main path of the polishing trajectory adopts a line-cut mode to suppress periodic ripples, locally superimposes a cycloid trajectory to compensate for random removal fluctuations, and then evaluates the removal amount accumulation effect based on Monte Carlo simulation, constructs a joint function with uniformity of residual height and path smoothness as the optimization objective, and dynamically corrects the polishing trajectory and overlap rate parameters according to the joint function.

[0017] In some embodiments, in step 2, the parameters of the polishing trajectory include at least trajectory spacing and tool inclination angle.

[0018] In some embodiments, in step 3, the constraint conditions set in the process of non-collision pose optimization of the robot pose change sequence include at least polishing pose geometric smoothness of the robot, joint angular velocity and angular acceleration.

[0019] In some embodiments, in step 3, the method for obtaining the robot pose change sequence includes:

[0020] The robot non-interference configuration space is converted into a weighted directed graph, a mapping relationship between the weighted directed graph and a mapping model is established, and the shortest distance of the weighted directed graph is calculated, the shortest distance being generated by a swarm intelligence optimization algorithm to generate the robot posture change sequence.

[0021] In some embodiments, in step 5, the optimization method of the residence time includes: inputting the material removal amount into an inversion equation to solve the residence time of the tool, the inversion equation including:

[0022]

[0023] In the formula, F -1 (·) is an inversion function, h target is a material removal amount, and Δt comp is a compensation residence time, used to correct deviations caused by factors such as process parameters.

[0024] In some embodiments, in step 6, the topography simulation result is a set of the material removal amounts corresponding to all the discrete machining points.

[0025] The beneficial effects of the present application are: through a multi-step collaborative optimization framework, the material removal amount prediction model, the grinding and polishing trajectory, the robot posture, and the dynamic error compensation are integrated into a unified optimization process, the nonlinear coupling influence of the robot posture adjustment and the feed speed fluctuation on the removal amount distribution is reduced, the collaborative problems of removal amount out-of-tolerance, coverage uniformity, and interference avoidance caused by multi-parameter coupling in narrow flow channels are solved, finally, the high-dimensional nonlinear optimization problem is decomposed by the alternating direction multiplier method, the convergence efficiency is significantly improved, the machining surface meets the preset tolerance requirement, and high-precision deterministic grinding and polishing in a restricted space are realized. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 The flowchart of the multi-parameter collaborative optimization method for deterministic grinding and polishing of a robot in a restricted space disclosed in the embodiments of the present application. DETAILED DESCRIPTION

[0027] To make the purpose, technical solutions and advantages of the present application clearer and more explicit, the content of the present application will be further described in detail below in combination with the drawings and specific embodiments. It can be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings, but not all the contents.

[0028] The embodiment proposes a multi-parameter collaborative optimization method for deterministic polishing of a robot in a confined space. Through a multi-step collaborative optimization framework (model establishment→ trajectory planning→ posture optimization→ closed-loop feedback→ global iteration), the material removal amount prediction model, polishing trajectory, robot posture, and dynamic error compensation are integrated into a unified optimization process, reducing the nonlinear coupling influence of robot posture adjustment and feed speed fluctuation on the removal amount distribution, solving the collaborative problems of removal amount out-of-tolerance, coverage uniformity, and interference avoidance caused by multi-parameter coupling in narrow flow channels, and finally decomposing high-dimensional nonlinear optimization problems through the alternating direction multiplier method, significantly improving the convergence efficiency, ensuring that the processed surface meets the preset tolerance requirements, and realizing deterministic polishing with high precision in a confined space. Figure 1 as shown, comprising the following steps:

[0029] A multi-parameter collaborative optimization method for deterministic polishing of a robot in a confined space, comprising the following steps:

[0030] Step 1: For the target narrow flow channel, a material removal amount prediction model based on the modified Preston equation and the side blade contact mechanics model is established. In this step, by establishing a material removal amount prediction model based on the modified Preston equation and the side blade contact mechanics model, the limitations of traditional Hertz contact theory in nonlinear working conditions of the side blade are broken through, and the prediction accuracy of material removal depth in the narrow flow channel is significantly improved, providing a theoretical basis for subsequent trajectory planning and parameter optimization.

[0031] Step 2: The target workpiece model is planned using a line-cut or ring-cut trajectory to generate a polishing trajectory of the center point of the side blade cutter, improving the path coverage uniformity and smoothness of the polishing trajectory, and solving the contradiction between cutter accessibility and removal amount stability in the narrow flow channel.

[0032] Optionally, it also includes introducing a computational conformal geometry method and using an adaptive algorithm to optimize different polishing trajectories. In this optional scheme, by introducing a computational conformal geometry method and an adaptive algorithm to optimize the polishing trajectory, the local geometric features of complex surfaces are dynamically matched, the residual height fluctuation caused by the axial mutation of the cutter is reduced, and the collaborative accuracy of trajectory planning and material removal model is enhanced, thereby improving the uniformity of the allowance removal in the narrow flow channel.

[0033] In an example, the planning strategy of the above polishing trajectory includes:

[0034] The main path of the grinding and polishing trajectory adopts a line-cut mode to suppress periodic ripples, and a local trochoidal trajectory is superimposed to compensate for random removal fluctuations. Then, based on Monte Carlo simulation, the cumulative effect of the removal amount is evaluated, a joint function is constructed with the uniformity of the residual height and the smoothness of the path as the optimization objectives, and the grinding and polishing trajectory and the overlap rate parameters are dynamically corrected according to the joint function. In this example, the line-cut main path is used to suppress periodic ripples, the local trochoidal trajectory is superimposed to compensate for random removal fluctuations, and the dynamic parameter optimization of the Monte Carlo simulation is combined to effectively balance the coverage uniformity and the path smoothness, avoid the defects of fixed wave peaks and valleys or random overshoot in a single trajectory mode, and improve the overall grinding and polishing quality of complex surfaces.

[0035] In this scheme, the parameters of the grinding and polishing trajectory at least include the trajectory spacing and the tool inclination angle. By cooperatively adjusting the trajectory spacing (controlling the coverage density) and the tool inclination angle (optimizing the contact pressure distribution), the prediction error problem of the traditional model under the nonlinear working condition of the side edge contact is solved, the controllability of the material removal behavior in the narrow flow channel is enhanced, and high-consistency surface processing is realized.

[0036] Step 3: Discretize the grinding and polishing trajectory to obtain discrete machining points, and then establish a robot non-interference configuration space corresponding to each discrete machining point under the constraint condition of collision-free machining. The robot posture change sequence is obtained by optimizing the collision-free posture of the grinding and polishing trajectory in the robot non-interference configuration space. In this step, the robot smooth posture sequence is generated in the strong constraint feasible region (collision-free, high stiffness, and non-singularity) by optimizing the collision-free posture of the discrete machining points, avoiding the dynamic error accumulation caused by the sudden change of the robot posture in the traditional method, and ensuring the motion stability of the grinding and polishing process in the narrow space.

[0037] In the process of robot posture change sequence optimization, the constraint conditions set at least include the grinding and polishing posture geometric smoothness of the robot, the joint angular velocity and angular acceleration, so as to suppress the posture mutation and dynamic error transmission, ensure the stability of the robot motion in the narrow space, and reduce the risk of surface quality degradation caused by vibration or interference.

[0038] In an example, the method for obtaining the robot posture change sequence comprises:

[0039] The robot non-interference configuration space is converted into a weighted directed graph, and a mapping relationship between the weighted directed graph and the mapping model is established. The shortest distance of the weighted directed graph is calculated, and the robot posture change sequence is generated by a group intelligence optimization algorithm. In this example, the non-interference configuration space is mapped into a weighted directed graph, and the group intelligence algorithm is combined to quickly screen the global optimal posture sequence, significantly reducing the high computational complexity of the traditional enumeration method, and realizing efficient planning of the high-performance non-collision posture of the robot in the narrow flow channel.

[0040] Step 4, calculate the material removal amount according to the polishing trajectory and the robot pose change sequence, and in combination with the material removal amount prediction model. In this step, based on the material removal amount prediction model of step 1 and the robot pose change sequence calculated in step 3, the material removal amount distribution is calculated in real time, the dynamic coupling analysis of trajectory-pose-process parameters is realized, the data basis for residence time optimization is provided, and the problem of disconnection between removal amount prediction and actual processing in traditional open-loop planning is solved.

[0041] Step 5, optimize and adjust the residence time of each discrete machining point according to the material removal amount combined with the material removal amount requirement of the task, and calculate the polishing speed of the robot according to the residence time.

[0042] In step 5, the optimization method of residence time includes: inputting the material removal amount into the inverse equation to solve the residence time of the tool, and the inverse equation includes:

[0043]

[0044] In the formula, F -1 (·) is the inverse function, h target is the material removal amount, Δt comp is the compensation residence time, which is used to correct the deviation caused by process parameters and other factors. In step 5, the theoretical residence time is solved by the inverse equation, and the compensation term is introduced to correct the dynamic errors such as speed fluctuation and tool wear, generate the polishing speed curve that meets the kinematic constraints of the robot, realize the collaborative adaptation of process parameters and robot dynamic performance, and improve the real-time performance and robustness of removal amount control.

[0045] Step 6, input the optimized robot pose change sequence, residence time and polishing speed into the material removal amount prediction model to predict the material removal amount of the target workpiece model, generate the topography simulation result, and based on the topography simulation result, repeat steps 2-5 using the alternating direction multiplier method until the preset material removal amount tolerance is met. In this step, the alternating direction multiplier method (ADMM) is used to iteratively optimize the trajectory, pose and residence time globally, decompose the high-dimensional nonlinear coupling problem into sub-modules that can be solved in parallel, significantly improve the convergence efficiency, ensure that the material removal amount error finally converges to the preset tolerance range, and achieve the deterministic polishing target in the limited space.

[0046] Wherein, the topography simulation result is a set of material removal amounts corresponding to all discrete machining points, which aims to illustrate that the iterative optimization process in step 6 needs to be specific to the material removal amount corresponding to each discrete machining point.

[0047] The above examples are only for illustrating the technical concept and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the present application and to implement it, and cannot limit the protection scope of the present application. Any equivalent changes or modifications made according to the essence of the present application should be covered within the protection scope of the present application.

Claims

1. A multi-parameter collaborative optimization method for deterministic grinding and polishing of robots in confined spaces, characterized in that, Includes the following steps: Step 1: For the target narrow flow channel, establish a material removal prediction model based on the modified Preston equation and the side blade contact mechanics model; Step 2: The target workpiece model is planned using line cutting or circumferential cutting trajectories to generate the grinding and polishing trajectory of the center point of the side cutting tool; Step 3: Discretize the polishing trajectory to obtain discrete processing points. Then, with collision-free processing as a constraint, establish a robot non-interference configuration space for each discrete processing point. Perform smooth, collision-free posture optimization along the polishing trajectory within the robot non-interference configuration space to obtain a robot posture change sequence. Step 4: Calculate the amount of material removed based on the polishing trajectory, the robot posture change sequence, and the material removal prediction model. Step 5: Based on the material removal amount required by the task, optimize and adjust the dwell time of each discrete processing point, and calculate the robot's polishing speed based on the dwell time; Step 6: Input the optimized robot posture change sequence, dwell time and polishing speed into the material removal amount prediction model, predict the material removal amount of the target workpiece model, generate morphology simulation results, and repeat steps 2-5 based on the morphology simulation results using the alternating direction multiplier method until the preset material removal amount tolerance is met.

2. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, Step 2 also includes introducing a computational conformal geometry method and using an adaptive algorithm to optimize different polishing trajectories.

3. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, In step 2, the planning strategy for the polishing trajectory includes: The main path of the polishing trajectory uses a line-cutting mode to suppress periodic ripples, and locally superimposed cycloidal trajectories to compensate for random removal fluctuations. Then, based on Monte Carlo simulation, the cumulative effect of removal is evaluated, and a joint function is constructed with the optimization objectives of residual height uniformity and path smoothness. The polishing trajectory and overlap rate parameters are dynamically corrected according to the joint function.

4. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, In step 2, the parameters of the polishing trajectory include at least the trajectory spacing and the tool tilt angle.

5. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, In step 3, the constraints set for the robot's posture change sequence during the collision-free posture optimization process include at least the robot's grinding and polishing posture geometric compliance, joint angular velocity, and angular acceleration.

6. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, Step 3, the method for obtaining the robot posture change sequence includes: The robot's non-interference configuration space is converted into a weighted directed graph, and a mapping relationship is established between the weighted directed graph and the mapping model. The shortest distance of the weighted directed graph is calculated, and the shortest distance is used to generate the robot's posture change sequence through a swarm intelligence optimization algorithm.

7. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, In step 5, the method for optimizing the dwell time includes: inputting the amount of material removed into the inversion equation to solve for the tool dwell time, wherein the inversion equation includes: In the formula, F -1 (·) is the inversion function, h target Δt represents the amount of material removed. comp It is used to compensate for dwell time and to correct deviations caused by factors such as process parameters.

8. The multi-parameter collaborative optimization method for deterministic grinding and polishing of confined space robots as described in claim 1, characterized in that, In step 6, the morphology simulation result is the set of material removal amounts corresponding to all the discrete processing points.

Images