A path and speed coordination planning method and device for a laser galvanometer fly-cutting system

Abstract

The application provides a path and speed cooperative planning method and equipment of a laser galvanometer fly machining system, belongs to the technical field of laser galvanometer fly machining, and establishes a variable range of a macro platform relative to a target trajectory; determines a migration condition between state points to form a state transition constraint condition; iteratively generates a macro platform trajectory meeting kinematic constraints based on a steady state action space and the state transition constraint, and outputs an initial type value point sequence; extracts a survival kernel subset in the steady state action space, and embeds the survival kernel into a cooperative planning framework; fits the initial type value point by using a quintic B-spline curve to generate a high-precision macro platform trajectory meeting dynamic constraints; and point by point calculates a micro platform trajectory to ensure that the micro platform trajectory meets machining requirements within a maximum machining range. The laser galvanometer fly machining system can efficiently and accurately plan a macro-micro platform cooperative motion trajectory under the premise of meeting equipment physical constraints, and improves machining stability and product quality.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of laser galvanometer fly machining, and particularly relates to a path and speed coordination planning method and equipment of a laser galvanometer fly machining system. BACKGROUND

[0002] The laser galvanometer fly machining system is composed of a numerical control machine tool (macro platform) and a laser galvanometer (micro platform), and through macro-micro coordinated motion, the advantages of flexible numerical control machine tool, large working space and high-speed and high-response laser galvanometer can be fully exerted to realize high-speed, high acceleration and high-precision laser machining of complex profiles. In the laser galvanometer fly machining system with macro-micro coupling characteristics, the key to realizing efficient and high-precision machining lies in coordinating the motion of the macro and micro platforms.

[0003] However, in laser galvanometer fly machining, if the acceleration, direction angle change and the like of the macro platform motion are not constrained, the macro platform may appear super motor maximum acceleration and deceleration motion, causing motor overheating and damage, affecting the service life of the equipment; the instantaneous large change in the direction angle will cause abrupt inflection points in the machining trajectory, affecting the machining precision and reducing the system reliability. Moreover, the steady-state motion space of the macro platform contains a large number of state points, and if direct traversal planning is not filtered, the calculation amount will increase exponentially with the number of state points, resulting in extremely long planning time, which cannot meet the real-time or efficient planning requirements of laser machining.

[0004] For the target machining trajectory, the macro and micro platforms need to be coordinated, but if the motion relationship between the two is not clearly defined in the prior art, the micro platform may not be able to cooperate with the macro platform due to motion exceeding its machining range, or the macro and micro motion may deviate from the target trajectory after superposition, causing machining misalignment and omission, and failing to realize precise coordinated machining. SUMMARY

[0005] The application provides a path and speed coordination planning method for a laser galvanometer fly machining system, which includes the construction of a macro platform motion base space, state transition constraints, gradual optimization, survival kernel efficiency improvement, trajectory refinement and micro platform coordination. The laser galvanometer fly machining system can efficiently and accurately plan the macro and micro platform coordinated motion trajectory under the premise of meeting the physical constraints of the equipment, and improve the machining stability and product quality.

[0006] The method comprises the following steps:

[0007] Step S101: based on the variable range of the target trajectory and the machining area of the micro platform, a variable range of the macro platform relative to the target trajectory is established, and the variable range includes a set of feasible position points of the macro platform, a direction angle range and a speed constraint condition;

[0008] Step S102: According to the dynamic characteristics of the macro platform, determine the transition conditions between state points, including acceleration limit, direction angle change range and motion conversion rule within the interpolation period, form the state transition constraint condition;

[0009] Step S103: Through the progressive optimization strategy, decompose the overall planning into local sub-problems, based on the steady-state action space and state transition constraint, iteratively generate the macro platform trajectory that meets the kinematic constraint, output the initial type value point sequence;

[0010] Step S104: Extract the survival kernel subset in the steady-state action space, filter the state points that do not have the ability of continuous feasible transition, and embed the survival kernel into the cooperative planning framework;

[0011] Step S105: Fit the initial type value point with quintic B-spline curve, and combine genetic algorithm and sequential quadratic programming algorithm for global search and local optimization of the trajectory, to generate high-precision macro platform trajectory that meets the dynamic constraint;

[0012] Step S106: Based on the vector difference between the target trajectory and the macro platform trajectory, calculate the trajectory of the micro platform point by point, to ensure that the micro platform trajectory meets the processing requirements within the maximum processing range.

[0013] Further need to explain is that in step S101, the set represents the variation range of the macro platform relative to the target trajectory;

[0014] Among them, and represent the set of feasible position points in the feasible region;

[0015] is the width set corresponding to each center point on the center line of the feasible region;

[0016] The constraint condition satisfied by the macro platform feasible position point is:

[0017] ;

[0018] and represent the position coordinates of the macro platform, represent the trajectory set corresponding to the macro platform feasible position point.

[0019] Further need to explain is that in step S101, the macro platform motion speed is less than the maximum speed provided by the system , and satisfies the following constraint condition:

[0020]

[0021] represents the direction angle of the motion direction, representing the velocity;

[0022] All the steady state points satisfying the constraint conditions are combined to form a steady state action space ;

[0023] wherein the steady state action points in the steady state action space The unit direction vector of the steady state action points is obtained by a rotation matrix The rotation angle is obtained by the following formula,

[0024]

[0025] wherein, is the rotation angle.

[0026] It is further needed to be explained that in step S102, in the process of transforming from one state point to another state point within the unit time step , the acceleration and deceleration cannot exceed the maximum acceleration and maximum deceleration that can be provided by the driving device, satisfying the following conditions:

[0027]

[0028] wherein, and represent the velocities of the state point and the state point , respectively, represents the maximum acceleration;

[0029] The transformation from the state point to is the i-th segment of the search, the transformation from the state point to is the i+1-th segment of the search, the velocity direction in the last interpolation period of the i-th segment is the same, and the velocity direction in the first interpolation period of the i+1-th segment is the same as that of the i+1-th segment;

[0030] Let represent the unit direction of the line segment i, represent the unit direction vector of the line segment i+1, then:

[0031]

[0032] that is,

[0033]

[0034] wherein, is the interpolation period, is half of the feasible direction angle.

[0035] It needs to be further explained that in step S103, the overall planning is decomposed into local sub-problems, and the motion trajectory meeting the global optimization requirement is gradually approached through iterative calculation. The specific steps are as follows:

[0036] The single-step exploration process is performed, and the kth step planning state point is set as , that is, , considering the acceleration constraint of formula , all feasible speeds at the next time are traversed, and a state point at the next time is calculated. The calculation method is as follows:

[0037]

[0038] Among them, is the acceleration of state transition, and are the direction angle of the unit direction vector, is the displacement in a time step;

[0039] The nearest neighbor approximation strategy is used to find the grid point closest to the current calculation result in the established state space set. The out-of-bound detection is performed on the calculated state point according to formula , and the elements exceeding the micro-platform processing range are removed, and the single-step exploration is completed.

[0040] It needs to be further explained that the multi-step exploration and local planning are performed.

[0041] The local trajectory is generated through multiple single-step explorations, and the local optimal trajectory is selected based on the relative distance between the terminal state point and the starting point, forming the iteration basis of multi-step exploration and local planning;

[0042] The terminal state point of the local optimal trajectory is taken as the parent node, and the multi-step exploration and local planning are repeatedly performed, and the trajectory length is gradually expanded until the target path is covered, and the global planning of the macro-platform processing trajectory is completed.

[0043] In the multi-step exploration process, the feasibility of the local trajectory is evaluated in combination with the migration constraint condition, and the advantages and disadvantages of the trajectory are determined through index value comparison;

[0044] When the generated trajectory meets the preset end point traversal condition, the search process is terminated, and the final macro-platform processing trajectory is output.

[0045] It needs to be further explained that in step S104, the state quantity is , and the control quantity is ;

[0046] The path and speed cooperative operation method is defined as a function , the following expression is obtained:

[0047]

[0048] All the state points satisfying the constraints obtained by single-step exploration with a certain state point as the starting point are expressed as:

[0049]

[0050] Then is expressed as:

[0051]

[0052] The extracted survival kernel is integrated into the original path and speed cooperative planning algorithm framework, and the type value point is obtained through coarse planning to obtain a smooth trajectory.

[0053] Further, in step S105, a quintic B-spline curve is selected as the trajectory model for planning, and the generated type value point is fitted;

[0054] The initial trajectory is refined by a hybrid optimization algorithm combining genetic algorithm and sequential quadratic programming algorithm;

[0055] The refinement includes: expanding global search through genetic algorithm to solve the initial solution;

[0056] The initial solution is used as the starting point, and the SQP algorithm is used for local search to optimize the initial solution;

[0057] In the optimization process, the speed of each optimization point is solved, and a nonlinear constraint condition is set:

[0058]

[0059] According to the dynamic model of the macro-micro system, the driving force required by each trajectory point is solved, and the constraint condition is as follows:

[0060] .

[0061] Further, in step S106, the target machining trajectory is defined as the vector superposition of the macro platform and micro platform motion trajectories, i.e.

[0062]

[0063] Wherein, is the target trajectory vector, is the macro platform trajectory vector, is the micro platform trajectory vector;

[0064] Based on the given target trajectory, the trajectory of the macro platform is planned, the target trajectory and the planned macro platform trajectory are calculated by point-by-point vector difference, and the trajectory of the micro platform is planned;

[0065] Every interpolation period The operation mode is as follows:

[0066]

[0067] The constraint condition satisfied by the micro platform trajectory obtained by decomposition is:

[0068]

[0069] The maximum machining range of the micro platform.

[0070] According to another embodiment of the application, an electronic device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the path and speed collaborative planning method of the laser galvanometer fly machining system when executing the program.

[0071] From the above technical solutions, the application has the following advantages:

[0072] The path and speed collaborative planning method of the laser galvanometer fly machining system according to the application constructs a macro platform steady-state action space according to the target trajectory variable range and the micro platform machining area, clearly defines the constraints of feasible position, direction angle, and speed, and makes the macro platform motion always within the range of physical feasibility and meeting the machining requirements. Based on the dynamic characteristics of the macro platform, the acceleration, direction angle change, and interpolation period motion conversion rules of state point migration are constrained, so that the planned trajectory conforms to the actual motion ability of the equipment, and the trajectory feasibility and machining stability are improved.

[0073] Through iterative approximation of the global optimum, the single planning calculation amount is reduced, the complex macro platform trajectory planning is more easily implemented, and the engineering application scene is adapted. The survival kernel subset is extracted, the state points without sustained transfer ability are filtered, the invalid states that need to be traversed during planning are reduced, the calculation resource consumption is reduced, and the planning efficiency is improved; after embedding the collaborative planning framework, the algorithm focuses on effective state transfer, and accelerates the trajectory convergence. Combined with the quintic B-spline curve fitting value point, the smoothness and local adjustability are utilized to convert the discrete value point into continuous and smooth trajectory, solve the problem of insufficient precision of rough planning trajectory, and meet the requirements of laser machining on trajectory smoothness. Based on the vector superposition principle, the micro platform trajectory is calculated by the vector difference between the target trajectory and the macro platform trajectory, the motion division of the macro and micro platforms is defined, the motion of the two is coordinated, the final machining trajectory is ensured to meet the target requirements, the motion of the micro platform is avoided due to the out-of-bounds, and the complete execution of the machining task is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the present application, the drawings required to be used in the description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0075] Figure 1 A schematic diagram of the path and speed cooperative planning method is shown in FIG. 1.

[0076] Figure 2 A flow chart of the path and speed cooperative planning method of the laser galvanometer flying machining system is shown in FIG. 2.

[0077] Figure 3 A framework diagram of the path and speed cooperative planning method based on the survival theory is shown in FIG. 3.

[0078] Figure 4 A schematic diagram of the hybrid optimization method is shown in FIG. 4.

[0079] Figure 5 A schematic diagram of the electronic device is shown in FIG. 5. DETAILED DESCRIPTION

[0080] The path and speed cooperative planning method of the laser galvanometer flying machining system based on the target machining trajectory, establishes a steady-state action space, analyzes the state transition constraint condition, determines the multi-dimensional constraint strategy of the path and speed cooperative planning as shown in FIG. 1, and accordingly proposes the path and speed cooperative planning algorithm, which converts the original continuous trajectory planning problem into a discrete trajectory planning problem. Through multiple iterations, the globally optimal trajectory is planned. Figure 1

[0081] In order to reduce the algorithm complexity, the path and speed cooperative planning optimization method based on the survival theory is proposed, the survival kernel is extracted from the original state space, the state points without survival ability are eliminated, the extracted survival kernel is integrated into the original algorithm framework, and the path and speed cooperative planning method is optimized. In order to further improve the planning accuracy, the fine planning method based on the hybrid optimization algorithm is proposed, the quintic B-spline curve is taken as the trajectory model, and the hybrid optimization algorithm composed of the genetic algorithm (GA) and the sequential quadratic programming algorithm (SQP) is combined for further fine planning. The trajectory planning of the micro-platform is further carried out through the vector decomposition principle. In this way, the path and speed are cooperatively planned, the motion of the two platforms is reasonably distributed, the respective advantages of the macro and micro platforms are fully utilized, and the efficiency and quality of the laser machining are improved.

[0082] ​The path and speed coordination planning method of the laser galvanometer fly machining system involved in the present application will be described in detail below. For the purpose of illustration but not for the purpose of limitation, specific details such as specific system structures, techniques, etc. are presented in order to thoroughly understand the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details.

[0083] It should be understood that when used in the specification of the present application, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. The terms "comprising", "including", "having" and their variants mean "including but not limited to", unless otherwise specifically emphasized.

[0084] The phrase "one embodiment" or "some embodiments" appearing in the present application means that the specific feature, structure or characteristic described in the embodiment is included in one or more embodiments of the present application. Therefore, the phrases "in one embodiment", "in some embodiments", "in other some embodiments", "in yet some embodiments" appearing in different places in the present application do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized.

[0085] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative efforts fall within the scope of protection of the present application.

[0086] Please refer to Figure 2 The flowchart of the path and speed coordination planning method of the laser galvanometer fly machining system in a specific embodiment is shown, and the method comprises:

[0087] Step S101: Based on the variable range of the target trajectory and the machining area of the micro-platform, a steady-state action space of the macro-platform is established, which contains a set of feasible position points of the macro-platform, a range of direction angle and a speed constraint condition.

[0088] In some embodiments, the variable range of the target trajectory is determined, which is determined by the machining area of the micro-platform. The area covered by the micro-platform defines the boundary of the movable macro-platform. Then, the feasible domain of the macro-platform is divided into a set of state points by grid division, and the continuous feasible domain is discretized into individual state points.

[0089] The embodiment defines a steady-state action point, specifically including the position of the macro platform at a certain time, such as selected from a set of feasible position points, and also relates to speed and direction angle.

[0090] It should be noted that the set of feasible position points makes the position coordinates of the macro platform satisfy the requirement of covering the machining demand with the assistance of the micro platform, the direction angle is derived from the basic direction angle by means of a rotation matrix or the like to obtain other possible directions, and the speed is set not to exceed the maximum speed of the system to ensure the safety and controllability of the movement.

[0091] Step S102: According to the kinetic characteristics of the macro platform, the migration conditions between state points are determined, including acceleration limit, direction angle change range, and movement conversion rule within the interpolation period, to form state transition constraint conditions.

[0092] In some embodiments, according to the kinetic characteristics of the motor and transmission of the macro platform, such as the maximum acceleration and deceleration capability of the motor, it is determined that the acceleration (deceleration) speed cannot exceed the maximum acceleration (deceleration) speed that can be provided by the driving device when the state point is transferred, otherwise the equipment will be damaged or the movement cannot be stabilized.

[0093] In this embodiment, the change of the direction angle is determined by calculating the unit direction vectors of the two trajectory segments and combining the interpolation period to determine the feasible range of the change of the direction angle, so that the trajectory has physical feasibility and the stability and reliability of the machining process are improved.

[0094] Step S103: Through a progressive optimization strategy, the overall planning is decomposed into local sub-problems, and based on the steady-state action space and state transition constraints, the macro platform trajectory satisfying the kinematic constraints is iteratively generated, and an initial value point sequence is output.

[0095] In some embodiments, the progressive optimization strategy divides the overall macro platform trajectory planning into local sub-problems. In this embodiment, single-step exploration is performed, starting from the state point in the steady-state action space, and according to the state transition constraints, the possible position, speed, and direction at the next time are calculated. When the calculated state point is not in the discrete space, the nearest neighbor approximation is used to find the closest grid point, and points that exceed the machining range of the micro platform are also removed.

[0096] In this embodiment, multi-step exploration is based on the points obtained by single-step exploration as new starting points, and single-step exploration is repeated. After several times, a local trajectory is formed. By comparing the distance between the terminal state point and the starting point of the local trajectory, the trajectory that goes further in the same time is selected as the local optimal trajectory. Iteration is continuously performed until an initial value point sequence that can cover the machining process is generated. The value point is a key discrete point on the trajectory.

[0097] Step S104: Extract the surviving kernel in the steady-state action space, filter out the state points that do not have sustainable transfer ability, and embed the surviving kernel into the collaborative planning framework.

[0098] In some embodiments, the transfer of each state point in the steady-state action space is analyzed first to see if it can be transferred to other effective state points from a state point.

[0099] In this embodiment, the surviving theory is used to train and screen these state points, and a set of state points that can find at least one transfer target in the surviving kernel is extracted. Embedding these surviving kernels into the previous collaborative planning framework, only the state points in the surviving kernel are processed in subsequent planning, and the state points that are difficult to sustainably transfer are filtered out. Avoid wasting computing resources on invalid state points.

[0100] Step S105: Fit the initial type value points with a quintic B-spline curve, and combine genetic algorithm and sequential quadratic programming algorithm to perform global search and local optimization on the trajectory to generate a high-precision macro-platform trajectory that meets the dynamics constraints.

[0101] In some embodiments, the initial type value points obtained in step S103 are first fitted with a quintic B-spline curve, because the quintic B-spline curve has advantages in smoothness, continuity and local adjustability, and can convert discrete type value points into smooth trajectory curves. Then, a global search is performed using genetic algorithm, which simulates biological evolution and finds an initial solution close to the optimal solution in a large solution space through selection, crossover, and mutation operations. This initial solution may not be the most accurate, but it can cover a better area. Then, the initial solution obtained by genetic algorithm is used as the starting point for local search and optimization using the sequential quadratic programming algorithm (SQP), which is good at handling quadratic programming problems and can quickly converge to a local optimum, improving trajectory accuracy.

[0102] In the optimization process, speed constraints are also added to ensure that the speed does not exceed the maximum system speed, and dynamics constraints are also involved. The required driving force for the trajectory point can be calculated based on the macro-micro system dynamics model, so that the driving force is within the range that the device can withstand. This solves the problem of insufficient precision of the coarse planning trajectory and improves the quality of the processed products.

[0103] Step S106: Based on the vector difference between the target trajectory and the macro-platform trajectory, the trajectory of the micro-platform is calculated point by point to ensure that the micro-platform trajectory meets the processing requirements within the maximum processing range.

[0104] In some embodiments, considering that the target machining trajectory is a vector superposition of the macro-platform and micro-platform motion trajectories, after obtaining the macro-platform trajectory, the micro-platform trajectory vector can be calculated point by point by subtracting the macro-platform trajectory vector from the target trajectory vector. Then, it is checked whether the micro-platform trajectory is within the maximum machining range of the micro-platform, that is, the micro-platform position calculated in each interpolation period cannot exceed the area covered by the micro-platform, so as to ensure that the micro-platform can actually complete the corresponding motion, assist the macro-platform to realize the complete machining trajectory, and ensure the feasibility of micro-platform motion.

[0105] In this way, the macro-micro platform cooperative motion trajectory is completely planned, the micro-platform can cooperate with the macro-platform to realize the target machining trajectory, the machining accuracy of the entire laser galvanometer flying machining system is ensured, and the machining task can be completed according to the planned path and speed.

[0106] Further, as a refinement and extension of the embodiment of the path and speed cooperative planning method of the laser galvanometer flying machining system, in order to completely describe the specific implementation process in the path and speed cooperative planning method of the laser galvanometer flying machining system, the laser galvanometer flying machining system involved in the path and speed cooperative planning method of the laser galvanometer flying machining system is a redundant degree of freedom system, and the target path to be machined is the synthesis result of the macro-platform and micro-platform motion of the redundant degree of freedom system. The paths of the macro-micro platform are unknown.

[0107] The motion planning target of the embodiment is to plan a trajectory with the shortest machining time as possible on the basis of satisfying the kinematics and dynamics of the system, so as to complete the machining task more quickly and improve the production efficiency. Due to the macro-micro coupling characteristics of the system, for a specific target trajectory, the macro-micro platform can provide a plurality of different motion output combinations.

[0108] The motion of the macro-platform in the embodiment is not limited to strictly following the given target trajectory, but can move within a certain range, which is determined by the machining area covered by the micro-platform.

[0109] For the macro-platform, the motion planning problem can be expressed as follows:

[0110]

[0111] wherein, t represents a time variable, u represents a control variable of the macro-platform, u represents a control variable of the macro-platform, T represents the total machining time of the macro-platform from the start time to the end time , C represents a state variable constraint of the macro-platform, C represents the control constraints of the macro-platform, and C represents the dynamic constraints of the macro-platform. The specific definitions of the state vector and the control vector are as follows:

[0112]

[0113]

[0114] wherein and represent the position coordinates of the macro-platform represent the direction angle of the moving direction, represent the velocity, is the motor driving force.

[0115] For the macro-platform rough planning mode based on the survival theory of the embodiment, the steady state action space of the macro-platform is first constructed. Here, to establish the steady state action space containing a series of motion parameters of the macro-platform, the definition of the steady state action point is first determined, that is, at a certain moment, the position of the macro-platform, the velocity reached and the possible moving direction at the next moment.

[0116] The feasible region of the macro-platform is meshed to generate a set of discrete state points, and the steady state action space is established on this basis. The set of feasible position points of the macro-platform is constructed, and the macro-platform can move within the machining area that can be covered by the micro-platform.

[0117] The set represents the variable range of the macro-platform compared with the target trajectory, wherein and represent the set of apparent position points in the feasible region, that is, the given machining trajectory; and is the set of widths corresponding to each center point on the center line of the feasible region, that is, the feasible machining diameter of the micro-platform. The first element in the set sets the index of the starting point, and the subsequent elements are arranged in turn, and the index value has a dual function: it can reflect the relative distance of the current position from the starting point, and it can also reflect the search progress.

[0118] In the trajectory planning process, since the algorithm uses a constant time step and a fixed number of local planning steps, it is difficult to effectively judge the optimality of the trajectory simply by relying on the time parameter. Therefore, the present application proposes to calculate the shortest distance from the current position to the reference center line, and use the index value corresponding to the projection point to quantify the displacement amount in unit time.

[0119] Therefore, the feasible position points of the macro-platform need to satisfy the following constraints:

[0120]

[0121] wherein, and represent the position coordinates of the macro-platform Represents the trajectory set corresponding to the feasible position points of the macro platform; that is, the feasible domain related to the given processing trajectory, the position coordinates of the macro platform It can only be considered a feasible position point if it is within the range of the set determined by the processing trajectory, etc. In simple terms, it is the definition of the position set related to the trajectory that the macro platform can move.

[0122] During the actual machining process, the tool in the macro platform should always move in the predetermined direction to avoid unnecessary retraction or reverse movement. The tool movement speed in the macro platform should be less than the maximum speed that the system can provide. The following constraints must be met:

[0123]

[0124] According to the above conditions, all the steady-state points that meet the constraints are combined into a steady-state action space , where the steady-state action point in the steady-state action space is The unit direction vector of the steady-state action point can be obtained by the rotation matrix The formula is as follows:

[0125]

[0126] in, is the rotation angle, usually in radians.

[0127] This embodiment also analyzes state transitions and constraints. Specifically, the kinematic characteristics of the macro platform prevent it from arbitrarily switching between any two state points. Therefore, this embodiment proposes the migration characteristics between state points and the conditions required to achieve state transitions.

[0128] Since the steady-state state point in the steady-state action space is a discrete motion state, the unit time step Internal state point Transfer to Limited by the acceleration performance and direction angle change of the two feed axes of the macro platform. In the process of converting from one state point to another, the acceleration (deceleration) cannot exceed the maximum acceleration (deceleration) that the drive device can provide, that is, , satisfying the following conditions:

[0129]

[0130] in, and Represents the state points and speed, denotes the maximum acceleration.

[0131] In this embodiment, the feasible range of the direction angle is considered in combination with the velocity of the current point and the maximum acceleration related to the state point is transferred to The i-th segment of the search is transferred from the state point to The i+1-th segment of the search is transferred from the state point to The direction of the motion must be changed within one interpolation period. At the connection of the two segments, the magnitude of the velocity is kept unchanged. Let denote the unit direction vector of the segment i,

[0132] denote the unit direction vector of the segment i+1, then we have:

[0133] i.e.

[0134]

[0135] wherein is the interpolation period, is half of the feasible direction angle.

[0136] For the path-velocity cooperative planning mode, according to the constructed state space and the state transition constraint condition, the path-velocity cooperative planning method proposed by the application uses a progressive optimization strategy to divide the overall planning problem into multiple local optimization sub-problems, and gradually approaches the motion trajectory meeting the global optimization requirement through iterative calculation.

[0137] The specific steps are as follows:

[0138] (1) Single-step exploration

[0139] In the offline state, all stable action points in the stable action space are traversed to calculate the possible motion state in the next time step. It is assumed that when the k-th step is planned, the actual state point is i.e. , considering the acceleration constraint of formula , all feasible velocities at the next time are traversed to calculate a state point at the next time. The calculation method is as follows:

[0140]

[0141] wherein is the acceleration of the state transition, and is the direction angle a unit direction vector of the drink, is the displacement in a time step.

[0142] In order to facilitate analysis and calculation, the established steady-state action space is discretized by grid division, which leads to the fact that when the next state point is calculated from a state point, the calculated state point may not be strictly located in the established state space set. In view of this problem, the nearest neighbor approximation strategy is used to find the grid point closest to the current calculation result in the established state space set, and the state point is calculated according to the formula The calculated state point is detected for out-of-bound, and the elements exceeding the micro-platform processing range are removed. Thus, single-step exploration is completed, and a series of steady-state action points satisfying the kinematic constraint are searched.

[0143] (2) Multi-step exploration and local planning

[0144] Single-step exploration searches a series of steady-state action points from a state point, and all the steady-state action points obtained by single-step exploration are taken as new starting points for further single-step exploration. By analogy, after times of single-step exploration, a series of local trajectories connected by state points are obtained. From the th single-step exploration to the th single-step exploration is taken as a local planning with a single planning step number of After a complete local planning, the distance between each terminal state point and the starting point needs to be calculated. By comparing the index values of the nearest points on the center line corresponding to each terminal state point, the relative distance is determined. After the same time, the trajectory farther away from the starting point is the local optimal trajectory.

[0145] (3) Iterative optimization and global trajectory planning

[0146] When multi-step exploration is performed, each step is a single-step exploration of all state points obtained by the previous step of exploration. With the increase of the exploration step number, the number of state points increases exponentially. In order to improve the calculation efficiency, the invention adopts an iterative optimization planning strategy to realize global path planning. After completing a round of multi-step exploration, the terminal state point of the optimal trajectory of the local planning in this round is taken as a parent node for further multi-step exploration. After several searches, the trajectory may pass through the end point, and the search process is ended, that is, a complete processing trajectory of the macro platform is planned.

[0147] This embodiment also performs optimization based on the path and speed cooperative planning of the survival theory. This is based on the single-step search mechanism, and the global optimal trajectory planning is realized by iterative operation of multiple local optimal planning. Although a trajectory with excellent performance can be generated, the calculation complexity is high.

[0148] To solve the problem, the optimization method involved in the embodiment trains a steady-state action space under the constraint of a target trajectory based on survival theory, and extracts a survival kernel subset satisfying a continuous feasible transition condition. Figure 3 As shown in the accompanying drawings, the survival kernel can filter out state points without survival ability, avoid traversal calculation of invalid state points, and thus significantly improve planning efficiency.

[0149] The steady-state action space is constructed, the transition conditions between various state points are analyzed, and accordingly the transition of a state point to other state points can be completed, which can be regarded as a response to a certain input state point. In the discrete system, let the state quantity be , and the control quantity be .

[0150] The path and speed collaborative operation method introduced above is defined as a function , and the following expression can be obtained:

[0151]

[0152] Therefore, all state points satisfying the constraint obtained by single-step exploration with a certain state point as a starting point can be represented as:

[0153]

[0154] The system represented by can be represented as:

[0155]

[0156] By extracting the survival kernel from the constructed steady-state action space, it is ensured that any state point in the survival kernel space can at least explore a state point also located in the survival kernel space, that is, the survival kernel has completed single-step exploration of all state points in the original action space, and a large number of state points without survival ability are removed. The extracted survival kernel is integrated into the original path and speed collaborative planning algorithm framework, and the macro value point in the initial trajectory of the macro platform is obtained through coarse planning, and further planning is performed to obtain a smoother trajectory.

[0157] The path and speed collaborative planning method based on survival theory coarsely plans the macro value point in the initial trajectory of the macro platform, however, the trajectory generated by the method is insufficient in accuracy and is difficult to meet the demand of high-precision trajectory planning.

[0158] The present application further selects a quintic B-spline curve as a trajectory model for planning, utilizes the advantages of the quintic B-spline curve in terms of trajectory smoothness, continuity and local adjustability to fit the generated macro value point, and then performs further fine planning on the initial trajectory by a hybrid optimization algorithm combining a genetic algorithm (GA) and a sequential quadratic programming algorithm (SQP), as shown in the accompanying drawings.Figure 4 as shown.

[0159] The specific steps of fine programming can be divided into two stages: first, a global search is carried out through a genetic algorithm to obtain a group of results close to the optimal solution, which creates a good start for subsequent local optimization. Then, the high-quality initial solution obtained by the genetic algorithm is used as the starting point for local search using the SQP algorithm to further optimize the initial solution and improve the accuracy and quality of the solution. The SQP algorithm can quickly find a local optimal solution in the local search process, improving the efficiency and accuracy of the entire optimization process.

[0160] To correctly guide the convergence direction of the algorithm, reasonable constraints are formulated during the solution process of the hybrid optimization algorithm. In terms of speed constraints, during the programming process, the speed of each programming point is solved and the following nonlinear constraints are applied:

[0161]

[0162] Only kinematic constraints are considered in the coarse programming process, but in actual engineering applications, trajectories generated based solely on kinematic constraints often cannot meet the requirements of system characteristics. Therefore, in the fine programming stage, the invention further introduces dynamic constraints, solves the required driving force for each trajectory point according to the dynamics model of the macro-micro system, and the constraint conditions are as follows:

[0163]

[0164] The target machining trajectory can be considered as the vector superposition of the macro-platform and micro-platform motion trajectories. That is:

[0165]

[0166] where, is the target trajectory vector, is the macro-platform trajectory vector, is the micro-platform trajectory vector.

[0167] Based on the given target trajectory, the above algorithm is used to program the macro-platform trajectory, and the target trajectory and the programmed macro-platform trajectory are calculated point by point to obtain the micro-platform trajectory. For each interpolation period , the specific operation method is as follows:

[0168]

[0169] where, the decomposed micro-platform trajectory should satisfy the constraint:

[0170]

[0171] is the maximum machining range of the micro-platform.

[0172] The present application provides a full-process collaborative planning method from coarse planning to fine planning, from macro platform to micro platform, taking into account the motion feasibility, calculation efficiency, trajectory accuracy and processing range adaptability, and improving the intelligentization of the processing system.

[0173] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0174] As shown in Figure 5 The present application also provides an electronic device, which includes a display module 103, a memory 102, a processor 101, and a computer program stored in the memory and executable on the processor 101, and the processor 101 implements the steps of the path and speed collaborative planning method of the laser galvanometer fly-cutting system when executing the program.

[0175] In the embodiments of the present application, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workstation, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are merely examples, and are not intended to limit the implementation of the embodiments of the present application described and / or claimed herein.

[0176] In the embodiments of the present application, the processor 101 can be implemented by using at least one of a special-purpose integrated circuit, a programmable logic device, a field programmable gate array, a processor, a controller, a microcontroller, a microprocessor, an electronic unit designed to perform the functions described herein, and in some cases, such implementation can be implemented in a controller. For software implementation, the implementation of processes or functions can be implemented with separate software modules that allow at least one function or operation to be performed. The software code can be implemented by a software application (or program) written in any appropriate programming language, which can be stored in a memory and executed by a controller.

[0177] The display module 103 is used to display information input by the user or information provided to the user. The display module 103 can include a display panel, which can be configured in the form of a liquid crystal display, an organic light-emitting diode, etc.

[0178] Memory 102 can be used to store software programs as well as various data. Memory 102 can include high-speed random access memory, and can also include nonvolatile memory such as at least one magnetic disk storage device, flash memory device, or other nonvolatile solid-state memory device.

[0179] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Numerous modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A path and speed collaborative planning method for a laser galvanometer flight processing system, characterized in that: Methods include: S101: Based on the variable range of the target trajectory and the processing area of ​​the micro-platform, establish the variable range of the macro-platform relative to the target trajectory; The variation range includes the feasible position point set, direction angle range and speed constraint conditions of the macro platform; S102: Based on the dynamic characteristics of the macro platform, migration conditions between state points are determined, including acceleration limits, direction angle change ranges, and motion conversion rules within the interpolation period, to form state transfer constraints. S103: Decompose the overall planning into local sub-problems through a progressive optimization strategy. Based on the steady-state action space and state transition constraints, iteratively generate a macro-platform trajectory that satisfies the kinematic constraints and outputs an initial form value point sequence. Decompose the overall planning into local sub-problems and gradually approach the motion trajectory that meets the global optimal requirements through iterative calculation. The specific steps are as follows: Execute the single-step exploration process and set the k-th step planning state point to ,Right now , consideration The acceleration constraint is and Represents the state points With status point speed, Indicates the maximum acceleration; Traverse all feasible speeds at the next moment and calculate a certain state point at the next moment , the calculation method is as follows: in is the acceleration of state transition, and is the direction angle The unit direction vector of the drink, is the displacement within one time step; The nearest neighbor approximation strategy is used to find the grid point closest to the current calculation result in the established state space set. Perform out-of-bounds detection on the calculated state points, eliminate elements that exceed the processing range of the micro-platform, and complete single-step exploration; Perform multi-step exploration and local planning; Generate local trajectories through multiple single-step explorations, and select the local optimal trajectory based on the relative distance between the terminal state point and the starting point, forming the iterative basis for multi-step exploration and local planning; Taking the terminal state point of the local optimal trajectory as the parent node, repeatedly perform multi-step exploration and local planning, gradually expand the trajectory length until it covers the target path, and complete the global planning of the macro platform processing trajectory; When the generated trajectory meets the preset end point crossing condition, the search process is terminated and the final macro platform processing trajectory is output; S104: Extract the survival core subset in the steady-state action space, filter out the state points that do not have the ability to continuously transfer, and embed the survival core into the collaborative planning framework; S105: Using a quintic B-spline curve to fit the initial shape value points, and combining the genetic algorithm with the sequential quadratic programming algorithm to perform global search and local optimization of the trajectory, a high-precision macro-platform trajectory that meets the dynamic constraints is generated; S106: Based on the vector difference between the target trajectory and the macro platform trajectory, the trajectory of the micro platform is calculated point by point to ensure that the micro platform trajectory meets the processing requirements within the maximum processing range.

2. The path and speed collaborative planning method of the laser galvanometer on-the-fly machining system according to claim 1, characterized in that: In step S101, the Indicates the range of variation of the macro platform relative to the target trajectory; in, and Represents the set of visible position points in the feasible region; is the width set corresponding to each center point on the center line of the feasible region; The constraints satisfied by the feasible location points of the macro platform are: ； and Indicates the position coordinates of the macro platform, Represents the set of trajectories corresponding to the feasible position points of the macro platform.

3. The path and speed collaborative planning method of the laser galvanometer on-the-fly machining system according to claim 1, characterized in that: In step S101, the macro platform movement speed is less than the maximum speed provided by the system , and satisfy the following constraints: The direction angle indicating the direction of movement, Indicates speed; All the steady-state points that meet the constraints form a steady-state action space ; Among them, the steady-state action point in the steady-state action space , the unit direction vector of the steady-state action point is given by the rotation matrix It is obtained by the following formula: in, is the rotation angle.

4. The path and speed collaborative planning method of the laser galvanometer on-the-fly machining system according to claim 1, characterized in that: In step S102, In unit time step In the process of converting from one state point to another, the acceleration and deceleration cannot exceed the maximum acceleration and maximum deceleration that the drive device can provide, that is, , satisfying the following conditions: in, and Represents the state points With status point speed, Indicates the maximum acceleration; Defined by state point Transfer to For the i-th segment of the search, the state point Transfer to For the i+1 segment being searched, the velocity direction in the last interpolation cycle of the i segment is the same, and the velocity direction in the first interpolation cycle of the i+1 segment is the same as that of the i+1 segment; set up represents the unit direction of line segment i, Represents the unit direction vector of line segment i+1, then: Right now: in, is the interpolation period, is half of the feasible direction angle.

5. The path and speed collaborative planning method of the laser galvanometer on-the-fly machining system according to claim 1, characterized in that: Step S104 is still in the discrete system, and the state quantity is set to , the control quantity is ; The path and speed collaborative operation method is defined as function , we get the following expression: Taking a certain state point as the starting point, all state points that meet the constraints obtained by single-step exploration are expressed as: but Expressed as: The extracted survival kernel is integrated into the original path and speed collaborative planning algorithm framework, and the type value points are obtained through rough planning to obtain a smooth trajectory.

6. The path and speed collaborative planning method of the laser galvanometer on-the-fly machining system according to claim 1, characterized in that: In step S105, a quintic B-spline curve is selected as the planned trajectory model, and the generated shape value points are fitted; The initial trajectory is precisely planned using a hybrid optimization algorithm that combines genetic algorithm with sequential quadratic programming algorithm. Refined planning includes: conducting a global search using a genetic algorithm to find an initial solution; Taking the initial solution as the starting point, the SQP algorithm is used for local search to optimize the initial solution; During the optimization process, the velocity of each optimization point is solved and nonlinear constraints are set: According to the dynamic model of the macro-micro system, the driving force required for each trajectory point is solved, and the constraints are as follows: 。 7. The path and speed collaborative planning method of the laser galvanometer on-the-fly machining system according to claim 1, characterized in that: In step S106, the target processing trajectory is defined as the vector superposition of the macro-platform and micro-platform motion trajectories, that is: in, is the target trajectory vector, is the macro platform trajectory vector, is the micro-platform trajectory vector; Based on the given target trajectory, the trajectory of the macro platform is planned, and the point-by-point vector difference between the target trajectory and the planned macro platform trajectory is calculated to plan the trajectory of the micro platform. Each interpolation cycle The operation method is as follows: Among them, the constraints satisfied by the decomposed micro-platform trajectory are: It is the largest processing range of the micro platform.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the path and speed collaborative planning method of the laser galvanometer on-the-fly machining system as claimed in any one of claims 1 to 7 are implemented.

Images