An industrial mobile robot speed planning method considering stability and precision

Abstract

The application relates to an industrial mobile robot speed planning method considering stability and precision, comprising the following steps: multi-dimensional speed constraint construction, spatial domain acceleration closed loop iteration, end quasi-stop control and trajectory time and density adjustment; through iterative optimization under multi-dimensional constraints in the spatial domain and the speed square domain, stable transportation and high-precision quasi-stop control in a complex industrial scene are realized.

Description

Technical Field

[0001] This application relates to the field of autonomous mobile robots, and in particular to a speed planning method for industrial mobile robots that balances stability and accuracy. Background Technology

[0002] With the rapid development of intelligent manufacturing and smart logistics, industrial mobile robots (AGVs / AMRs) are widely used in automated factories and warehouses, undertaking tasks such as heavy-duty handling, precision parts transportation, and automated docking. In these application scenarios, speed planning, as the core module of the mobile robot navigation system, directly determines its operating efficiency, transportation safety, and final docking accuracy.

[0003] Typically, the control process for a mobile robot involves first generating a geometric path, and then a speed planner assigning timestamps along that path based on the robot's dynamic performance (such as maximum speed, acceleration, and jerk). An ideal speed planning scheme not only needs to fully utilize the physical performance of the chassis drive motors to improve efficiency, but also needs to ensure sufficiently smooth acceleration changes when handling high-center-of-gravity or fragile goods to prevent slippage or tipping. Simultaneously, when entering charging stations or under shelves, the robot must possess extremely high stopping accuracy to avoid end-effector impacts or positional overshoot.

[0004] Although existing velocity planning algorithms (such as traditional S-curve and trapezoidal profile planning) can meet the motion requirements of mobile robots to a certain extent, they still have the following significant drawbacks in complex industrial application environments: 1. Oversimplified physical constraint modeling, ignoring chassis geometry: Current mainstream algorithms often simplify the robot to a point mass model, limiting only the linear velocity and acceleration of the center point. This approach ignores the influence of structural parameters such as chassis wheelbase and track width on the drive wheel speed. In conditions such as narrow-radius turns, the outer drive wheel may have reached its speed limit, while the calculated value of the center point still shows normal, causing the motor to lose steps, report errors, or fail to track the trajectory due to overload.

[0005] 2. Difficulty in balancing safety and smoothness: When traditional algorithms smooth the speed limit profile (e.g., Gaussian filtering or mean filtering), the mathematical smoothing effect often raises the local speed limit value, causing the planned speed to exceed the original safe physical limit at certain locations. If the global speed limit is simply lowered to ensure safety, transportation efficiency will be significantly sacrificed.

[0006] 3. Time-domain planning leads to discontinuous motion control. Most existing technologies plan in the time domain. When the robot's speed fluctuates significantly, the fixed sampling time interval will result in extremely uneven spatial displacement steps. This makes the suppression of acceleration and jerk inconsistent in spatial distribution, especially during low-speed start-up and deceleration, which can easily generate significant mechanical shocks.

[0007] 4. Dynamic abrupt change in the final stopping phase: Existing deceleration control strategies (such as linear deceleration) often fail to naturally converge to zero at the moment of reaching the endpoint. This abrupt change in acceleration produces obvious "stop point oscillation" or "nodding" phenomena, resulting in a decrease in end-of-pipe alignment accuracy and failing to meet the requirements for millimeter-level precise stopping.

[0008] 5. Numerical integration cumulative error and point density imbalance: When converting the velocity profile into a time trajectory, the traditional Euler integration method exhibits significant cumulative error, leading to a discrepancy between the actual travel distance and the planned path length. Furthermore, existing methods often result in excessively sparse or dense trajectory point density when handling low-speed regions, causing instability in the command frequency received by the underlying servo drive system, which in turn can lead to motor vibration or control overshoot. Summary of the Invention

[0009] Based on the above-mentioned technical problems, this application provides a speed planning method for industrial mobile robots that balances stability and accuracy, including: multi-dimensional speed constraint construction, spatial domain acceleration closed-loop iteration, end-effector quasi-stop control, and trajectory temporalization and density adjustment; through iterative optimization under multi-dimensional constraints in the spatial domain and the square velocity domain, stable transportation and high-precision quasi-stop control in complex industrial scenarios are achieved. The multidimensional speed constraint is constructed by establishing an upper limit for each sampling point on the path and calculating the maximum allowed speed upper limit for each point. The spatial domain acceleration closed-loop iteration is to perform closed-loop optimization of the velocity curve in the spatial domain while satisfying the velocity constraints, thereby suppressing the jerk changes during the process. The end-point quasi-stop control is to perform quasi-stop control within a preset range from the endpoint to avoid mechanical impact caused by acceleration not returning to zero at the moment of stopping; The trajectory temporalization and density adjustment involves converting the planned spatial profile into a timestamped instruction stream executable by a PLC or driver using a strategy of global sparsification and local refinement at the end.

[0010] The construction of the multidimensional velocity constraint includes: Chassis dynamics limit construction: Based on chassis geometry parameters, path curvature, and drive wheel rotation limits, the projection is a center-point speed limit, with correction factors for the outer wheels. for: in, For path curvature, Wheelbase The wheelbase is the distance between the wheels. Limit the drive wheels Divide by To obtain the real-time speed limit of the center point : Segmented speed limits and smooth transitions: For speed limits set in different areas within the warehouse, transition lengths are defined before and after the speed limit switching points. Within these transition lengths, a smoothing function is used for a soft transition, thereby achieving... Based on this, the original speed limit value for each point is obtained. ; Conservative profile preprocessing: at each sampling point At the sampling point Centered on, radius Use a window to calculate the arithmetic mean of all original speed limits. : Pick and The minimum value is used as the final constraint speed. : .

[0011] The spatial domain acceleration closed-loop iteration includes, based on satisfying velocity constraints, performing closed-loop optimization of the velocity curve in the spatial domain to suppress jerk changes during the process, specifically: Forward acceleration transfer: Starting from the origin, based on the maximum acceleration limit, in, It is the acceleration at sampling point i. It is the velocity at sampling point i+1. It is the velocity at sampling point i. Reverse deceleration transmission: recursively calculating from the endpoint, constraining the velocity of the previous point according to the maximum deceleration limit. in, It is the deceleration at sampling point i. Bidirectional jerk constraint propagation: Jerk is detected by repeatedly checking the acceleration difference between adjacent points. Has the upper limit been exceeded? Among them, jerk The calculation formula is: Where it can be approximated as ,in If the average velocity is between two adjacent points, then we can further deduce: If calculated If the threshold is exceeded, the speed limit at that point is forcibly lowered to meet the jerk limit, and then the speed limit is updated. Acceleration domain closed-loop optimization iteration: Calculate acceleration by differentiating from the current velocity. Then, a smoothing filter is applied to the acceleration sequence, and the smoothed acceleration sequence is then... Integrating yields candidate velocity sequences : Introducing under-relaxation factor The newly generated smooth velocity is weighted and fused with the velocity from the previous round, and then gradually converged to the optimal smooth curve through multiple closed-loop iterations. .

[0012] The end-of-line stop control involves switching to precision stopping logic within a preset distance from the destination to avoid mechanical shock caused by acceleration not returning to zero at the moment of stopping, including: Define normalized displacement ,in The total length of the smooth segment, This is the current position. This is the starting position of the smooth segment; Construct a polynomial, where : The polynomial needs to be solved under the following constraints: Starting position velocity matching: Zero speed at the finish line: Starting position acceleration matching: Zero acceleration at the endpoint: Matching of starting position related items: Clear the relevant fields for the finish line location (Jerk): The following solution can be obtained by satisfying the above relationship: in, , , , , , The coefficients of each term in the fifth-order polynomial. The speed at the starting point, Acceleration at the starting point Represents the starting point , Represents the finish line .

[0013] In the trajectory temporalization and density adjustment, the end-point local refinement and supplementation involves detecting the arc length difference between adjacent points during the deceleration phase near the endpoint. If the arc length difference exceeds a preset threshold, second-order interpolation is performed based on spline curves and velocity profiles to supplement the points, ensuring that the control commands for the robot's low-speed quasi-stop phase are smooth and continuous.

[0014] In the trajectory temporalization and density adjustment, the global sparsity thinning involves initially screening trajectory points throughout the entire trajectory based on the configured minimum output spacing to ensure that no redundant points are output in the high-speed segment.

[0015] The trajectory time-sharing and density adjustment also includes using the fourth-order Runge-Kutta method for numerical integration to eliminate the cumulative displacement deviation caused by large changes in velocity.

[0016] Compared to existing technologies, this invention precisely projects the center point speed limit onto the drive wheel limit using chassis geometry parameters, which both maximizes motor performance and efficiency while absolutely ensuring no overload.

[0017] This invention introduces a conservative smoothing algorithm that "only decreases and does not increase," ensuring that while improving speed continuity, it never exceeds any preset safety threshold.

[0018] This invention is based on closed-loop iterative optimization in the spatial domain, which makes the suppression of jerk no longer affected by changes in linear velocity, thus achieving ultimate transportation stability.

[0019] This invention applies fifth-order Hermite interpolation in the velocity square domain to force the terminal acceleration and jerk to be zero, thereby achieving dynamic continuity and eliminating the physical impact at the moment of stopping.

[0020] This invention employs a fourth-order Runge-Kutta method combined with adaptive point density adjustment, which reduces the computing burden on the controller while eliminating cumulative displacement deviation and ensuring smooth and stable end control commands. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart based on an embodiment of this application; Figure 2 This is a schematic diagram of a speed curve example according to Embodiment 2 of this application. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0023] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0024] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0025] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0026] Example

[0027] A speed planning method for industrial mobile robots that balances stability and accuracy, such as... Figure 1 As shown, it includes: multidimensional velocity constraint construction, spatial domain acceleration closed-loop iteration, end-point quasi-stop control, and trajectory temporalization and density adjustment; through iterative optimization under multidimensional constraints in the spatial domain and velocity square domain, smooth transportation and high-precision quasi-stop control in complex industrial scenarios are achieved. The multidimensional speed constraint is constructed by establishing an upper limit for each sampling point on the path and calculating the maximum allowed speed upper limit for each point. The spatial domain acceleration closed-loop iteration is to perform closed-loop optimization of the velocity curve in the spatial domain while satisfying the velocity constraints, thereby suppressing the jerk changes during the process. The end-point quasi-stop control is to perform quasi-stop control within a preset range from the endpoint to avoid mechanical impact caused by acceleration not returning to zero at the moment of stopping; The trajectory temporalization and density adjustment involves converting the planned spatial profile into a timestamped instruction stream executable by a PLC or driver using a strategy of global sparsification and local refinement at the end.

[0028] The construction of the multidimensional velocity constraint includes: Chassis dynamics limit construction: Based on chassis geometry parameters, path curvature, and drive wheel rotation limits, the projection is a center-point speed limit, with correction factors for the outer wheels. for: in, For path curvature, Wheelbase The wheelbase is the distance between the wheels. Limit the drive wheels Divide by To obtain the real-time speed limit of the center point : Segmented speed limits and smooth transitions: For speed limits set in different areas within the warehouse, transition lengths are defined before and after the speed limit switching points. Within these transition lengths, a smoothing function is used for a soft transition, thereby achieving... Based on this, the original speed limit value for each point is obtained. ; Conservative profile preprocessing: at each sampling point At the sampling point Centered on, radius Use a window to calculate the arithmetic mean of all original speed limits. : Pick and The minimum value is used as the final constraint speed. : .

[0029] The spatial domain acceleration closed-loop iteration includes, based on satisfying velocity constraints, performing closed-loop optimization of the velocity curve in the spatial domain to suppress jerk changes during the process, specifically: Forward acceleration transfer: Starting from the origin, based on the maximum acceleration limit, in, It is the acceleration at sampling point i. It is the velocity at sampling point i+1. It is the velocity at sampling point i. Reverse deceleration transmission: recursively calculating from the endpoint, constraining the velocity of the previous point according to the maximum deceleration limit. in, It is the deceleration at sampling point i. Bidirectional jerk constraint propagation: Jerk is detected by repeatedly checking the acceleration difference between adjacent points. Has the upper limit been exceeded? Among them, jerk The calculation formula is: Where it can be approximated as ,in If the average velocity is between two adjacent points, then we can further deduce: If calculated If the threshold is exceeded, the speed limit at that point is forcibly lowered to meet the jerk limit, and then the speed limit is updated. Acceleration domain closed-loop optimization iteration: Calculate acceleration by differentiating from the current velocity. Then, a smoothing filter is applied to the acceleration sequence, and the smoothed acceleration sequence is then... Integrating yields candidate velocity sequences : Introducing under-relaxation factor The newly generated smooth velocity is weighted and fused with the velocity from the previous round, and then gradually converged to the optimal smooth curve through multiple closed-loop iterations. .

[0030] The end-of-line stop control involves switching to precision stopping logic within a preset distance from the destination to avoid mechanical shock caused by acceleration not returning to zero at the moment of stopping, including: Define normalized displacement ,in The total length of the smooth segment, This is the current position. This is the starting position of the smooth segment;

[0031] Construct a polynomial, where : The polynomial needs to be solved under the following constraints: Starting position velocity matching: Zero speed at the finish line:

[0032] Starting position acceleration matching:

[0033] Zero acceleration at the endpoint:

[0034] Matching of starting position related items:

[0035] Clear the relevant fields for the finish line location (Jerk):

[0036] The following solution can be obtained by satisfying the above relationship: in, , , , , , The coefficients of each term in the fifth-order polynomial. The speed at the starting point, Acceleration at the starting point Represents the starting point , Represents the finish line .

[0037] In the trajectory temporalization and density adjustment, the end-point local refinement and supplementation involves detecting the arc length difference between adjacent points during the deceleration phase near the endpoint. If the arc length difference exceeds a preset threshold, second-order interpolation is performed based on spline curves and velocity profiles to supplement the points, ensuring that the control commands for the robot's low-speed quasi-stop phase are smooth and continuous.

[0038] In the trajectory temporalization and density adjustment, the global sparsity thinning involves initially screening trajectory points throughout the entire trajectory based on the configured minimum output spacing to ensure that no redundant points are output in the high-speed segment.

[0039] The trajectory time-sharing and density adjustment also includes using the fourth-order Runge-Kutta method for numerical integration to eliminate the cumulative displacement deviation caused by large changes in velocity. Example

[0040] This embodiment uses a differential-driven industrial AMR as an example. Its chassis physical parameters are set as follows: wheelbase 0.8m, wheelbase 0 (drive wheel center coincides with rotation center). The robot needs to execute a composite path with a total length of 10m and a sampling interval of 0.02m.

[0041] The path topology and area speed limit settings are as follows: 0m - 3m section: straight path, speed limit 1.0m / s in the area; 3m - 6m section: Bezier curve, speed limit 0.5m / s in the area; 6m - 10m section: straight path, speed limit 1.0m / s.

[0042] Dynamics constraint settings: Maximum vehicle acceleration Maximum deceleration Maximum jerk Drive wheel limit speed .

[0043] 1. Construction of Multidimensional Velocity Constraints In the first stage, a clear upper limit is established for each sampling point on the path, and the maximum allowable speed limit is calculated for each point: Chassis dynamics limit projection: In the Bézier curve segment, set the maximum curve library of the sampling points. Limit the drive wheels Dividing by this factor yields the upper limit of the real-time velocity at the center point. : Segmented speed limiting and smooth transition: due to the maximum curvature point Therefore, we can directly take the minimum speed according to the speed limit of the path segment, and use a third-order smooth function to transition at the path connection.

[0044] Conservative profile preprocessing: Set smooth window radius At each sampling point, calculate the arithmetic mean of all original speed limits. : Pick and The minimum value is used as the final constraint speed. : 2. Spatial Domain Acceleration Closed-Loop Iteration In the second stage, in obtaining Then, global feasibility processing is performed: Acceleration and deceleration transmission: from the starting point Accelerate forward, from the finish line Reverse calculation of deceleration boundary, Take 0.01m.

[0045] Starting from the origin, based on the maximum acceleration limit.

[0046] Reverse deceleration transmission: Recursively calculate from the end point, constraining the speed of the previous point according to the maximum deceleration limit.

[0047] Bidirectional jerk constraint propagation: The algorithm detects jerk by checking the acceleration difference between adjacent points up to 12 times. Has the upper limit been exceeded? If the calculated If the threshold is exceeded, the speed limit at that point is forcibly lowered to meet the jerk limit, and then the speed limit is updated. For example, when , , .

[0048] Then it can be calculated Calculated If the threshold is exceeded, the speed limit at that point is forcibly lowered to meet the jerk limit, and then the speed limit is updated. Then, a smoothing filter is applied to the acceleration sequence, and the smoothed acceleration sequence is then... Integrating yields candidate velocity sequences : Introducing under-relaxation factor The newly generated smooth velocity is weighted and fused with the velocity from the previous round, and then gradually converged to the optimal smooth curve through multiple closed-loop iterations. 3. Terminal five Hermite quasi-stop control In the third stage, within 0.2m of the destination, the system switches to precision parking logic.

[0049] Define normalized displacement , This represents the current position. Assume the velocity at this moment... acceleration accelerometer

[0050] Construct a polynomial, where : The polynomial needs to be solved under the following constraints: Starting position velocity matching: Zero speed at the finish line:

[0051] Starting position acceleration matching:

[0052] Zero acceleration at the endpoint:

[0053] Matching of starting position related items:

[0054] Clear the relevant fields for the finish line location (Jerk):

[0055] The following solution can be obtained by satisfying the above relationship: Substituting into the polynomial, we get: 4. Trajectory temporalization and density adjustment In the final stage, to balance control accuracy and computational load, the algorithm executes a two-stage adaptive resampling strategy: Global sparsification thinning: In the entire trajectory, the trajectory points are initially filtered according to the configured minimum output spacing (0.01m here) to ensure that no redundant points are output in the high-speed section.

[0056] End-point local refinement and point supplementation: For the deceleration section near the endpoint (0.15m before the endpoint in this case), the arc length difference between adjacent points is detected. If the arc length difference exceeds a preset threshold (0.02m in this case), second-order interpolation is performed based on the spline curve and velocity profile to supplement the points, ensuring that the control commands for the robot during the low-speed quasi-stop phase are smooth and continuous.

[0057] During the time-varying process, the fourth-order Runge-Kutta method is used for numerical integration to eliminate the cumulative displacement deviation caused by large changes in velocity.

[0058] The results are as follows Figure 2 As shown, it achieves ultimate transportation stability.

[0059] This embodiment provides an electronic device including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0060] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0061] Optionally, the electronic device may further include a processor, memory, network interface, display screen, and input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a visual mapping and localization method that combines natural features with markers. The display screen may be a liquid crystal display (LCD) or an e-ink display. The input device may be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse.

[0062] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0063] Furthermore, in conjunction with the visual mapping and localization method combining natural features and markers in the above embodiments, this application embodiment can provide a storage medium for implementation. This storage medium stores a computer program; when executed by a processor, the computer program implements any of the visual mapping and localization methods combining natural features and markers in the above embodiments.

[0064] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods.

[0065] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0066] By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0067] Those skilled in the art should understand that the technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0068] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A speed planning method for industrial mobile robots that balances stability and accuracy, characterized in that, include: Multidimensional velocity constraint construction, spatial domain acceleration closed-loop iteration, end-stage quasi-stop control, and trajectory temporalization and density adjustment; By performing iterative optimization under multidimensional constraints in the spatial domain and the velocity square domain, smooth transportation and high-precision quasi-stop control can be achieved in complex industrial scenarios. The multidimensional speed constraint is constructed by establishing an upper limit for each sampling point on the path and calculating the maximum allowed speed upper limit for each point. The spatial domain acceleration closed-loop iteration is to perform closed-loop optimization of the velocity curve in the spatial domain while satisfying the velocity constraints, thereby suppressing the jerk changes during the process. The end-point quasi-stop control is to perform quasi-stop control within a preset range from the endpoint to avoid mechanical impact caused by acceleration not returning to zero at the moment of stopping; The trajectory temporalization and density adjustment involves converting the planned spatial profile into a timestamped instruction stream executable by a PLC or driver using a strategy of global sparsification and local refinement at the end.

2. The method according to claim 1, characterized in that, The construction of the multidimensional velocity constraint includes: Chassis dynamics limit construction: Based on chassis geometry parameters, path curvature, and drive wheel rotation limits, the projection is a center-point speed limit, with correction factors for the outer wheels. for: in, For path curvature, Wheelbase The wheelbase is the distance between the wheels. Limit the drive wheels Divide by To obtain the real-time speed limit of the center point : Segmented speed limits and smooth transitions: For speed limits set in different areas within the warehouse, transition lengths are defined before and after the speed limit switching points. Within these transition lengths, a smoothing function is used for a soft transition, thereby achieving... Based on this, the original speed limit value for each point is obtained. ; Conservative profile preprocessing: at each sampling point At the sampling point Centered on, radius Use a window to calculate the arithmetic mean of all original speed limits. ; Pick and The minimum value is used as the final constraint speed. ; 。 3. The method according to claim 2, characterized in that, The spatial domain acceleration closed-loop iteration includes, based on satisfying velocity constraints, performing closed-loop optimization of the velocity curve in the spatial domain to suppress jerk changes during the process, specifically: Forward acceleration transfer: Starting from the origin, based on the maximum acceleration limit, in, It is the acceleration at sampling point i. It is the velocity at sampling point i+1. It is the velocity at sampling point i. Reverse deceleration transmission: recursively calculating from the endpoint, constraining the velocity of the previous point according to the maximum deceleration limit. in, It is the deceleration at sampling point i. Bidirectional jerk constraint propagation: Jerk is detected by repeatedly checking the acceleration difference between adjacent points. Has the upper limit been exceeded? Among them, jerk The calculation formula is: Where it can be approximated as ,in If the average velocity is between two adjacent points, then we can further deduce: If calculated If the threshold is exceeded, the speed limit at that point is forcibly lowered to meet the jerk limit, and then the speed limit is updated. Acceleration domain closed-loop optimization iteration: Calculate acceleration by differentiating from the current velocity. Then, a smoothing filter is applied to the acceleration sequence. Integrating yields candidate velocity sequences : Introducing under-relaxation factor The newly generated smooth velocity is weighted and fused with the velocity from the previous round, and then gradually converged to the optimal smooth curve through multiple closed-loop iterations. 。 4. The method according to claim 3, characterized in that, The end-of-line stop control involves switching to precision stopping logic within a preset distance from the destination to avoid mechanical shock caused by acceleration not returning to zero at the moment of stopping, including: Define normalized displacement ,in The total length of the smooth segment, This is the current position. This indicates the starting position of the smooth segment. Construct a polynomial, where : ； The polynomial needs to be solved under the following constraints: Starting position velocity matching: Zero speed at the finish line: Starting position acceleration matching: Zero acceleration at the endpoint: Matching of starting position related items: Clear the relevant fields for the finish line location (Jerk): The following solution can be obtained by satisfying the above relationship: in, , , , , , The coefficients of each term in the fifth-order polynomial. The speed at the starting point, Acceleration at the starting point Represents the starting point , Represents the finish line .

5. The method according to claim 1, characterized in that, In the trajectory temporalization and density adjustment, the end-point local refinement and supplementation involves detecting the arc length difference between adjacent points during the deceleration phase near the endpoint. If the arc length difference exceeds a preset threshold, second-order interpolation is performed based on spline curves and velocity profiles to supplement the points, ensuring that the control commands for the robot's low-speed quasi-stop phase are smooth and continuous.

6. The method according to claim 1, characterized in that, In the trajectory temporalization and density adjustment, the global sparsity thinning involves initially screening trajectory points throughout the entire trajectory based on the configured minimum output spacing to ensure that no redundant points are output in the high-speed segment.

7. The method according to claim 1, characterized in that, The trajectory time-sharing and density adjustment also includes using the fourth-order Runge-Kutta method for numerical integration to eliminate the cumulative displacement deviation caused by large changes in velocity.

Images