Series-parallel hybrid redundant robot motion planning method and device, terminal and medium

Abstract

The invention provides a motion planning method and device for a series-parallel hybrid redundant robot, a terminal and a medium. The method comprises the steps that a performance space corresponding to a robot driving joint variable is constructed; in the performance space, motion trajectory planning is conducted on robot driving joint sets which are layered in advance through a forward layering planning strategy, and initial trajectories are obtained; a bias sampling strategy guided by a performance gradient is utilized in each layer of planning to guide bias sampling points to be concentrated towards a high-performance area; performing cross-layer rewiring optimization on the initial trajectory, and performing local refined search in a rewiring point neighborhood by using a virtual sampling strategy to obtain an optimized motion trajectory; and generating a target motion track of the robot based on the optimized motion track. According to the method, the joint configuration of the robot is mapped to the performance space, a strategy of combining forward hierarchical planning and reverse optimization is adopted, a virtual sampling point enhancement mechanism is introduced, and comprehensive optimization of multiple performance indexes is achieved.

Description

Technical Field

[0001] This invention relates to the field of robot motion planning technology, and in particular to a motion planning method, device, terminal and medium for a serial-parallel hybrid redundant robot. Background Technology

[0002] Currently, redundant robot systems significantly improve motion flexibility by increasing degrees of freedom, but they face challenges in motion planning, such as low efficiency in high-dimensional space search and difficulty in simultaneously achieving multiple performance metrics. Traditional motion planning methods, such as Restricted RRT and its improved algorithms (e.g., Informed RRT), address these challenges. Typically, random sampling and path searching are performed in a high-dimensional configuration space. This approach has inherent limitations: First, the "curse of dimensionality" causes sampling efficiency to drop sharply with increasing degrees of freedom, making it difficult to guarantee planning success, especially in complex obstacle environments. Second, due to the lack of modeling of the robot's own performance, it is difficult to simultaneously optimize multiple key indicators such as mechanism stiffness, force / torque bearing capacity, velocity / acceleration output performance, and motion accuracy during planning. These indicators are crucial for performing tasks such as precision assembly and heavy-duty operations. Although existing improved algorithms have improved planning efficiency to some extent, they have not fundamentally solved the multi-objective collaborative optimization problem of high-dimensional redundant systems. The fundamental reason is that these methods do not deeply integrate the robot's performance characteristics with the spatial search process, resulting in a somewhat blind search process.

[0003] Therefore, existing technologies have shortcomings and need to be improved and developed. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a motion planning method, device, terminal and medium for serial-parallel hybrid redundant robots, which addresses the above-mentioned defects of the prior art. The aim is to solve the problems of low sampling efficiency, poor trajectory quality and difficulty in balancing multiple performance indicators in the existing motion planning technology in high-dimensional redundant systems.

[0005] The technical solution adopted by this invention to solve the technical problem is as follows: In a first aspect, the present invention discloses a motion planning method for a serial-parallel hybrid redundant robot, wherein the method includes: Construct the performance space corresponding to the robot's drive joint variables; Within the performance space, the motion trajectory of the pre-layered robot drive joint groups is planned using a forward hierarchical planning strategy to obtain the initial trajectory; wherein, each layer of planning uses a performance gradient-guided bias sampling strategy to guide the bias sampling points to concentrate in the high-performance region; The initial trajectory is optimized by cross-layer rewiring, and a virtual sampling strategy is used to perform a local fine-grained search in the neighborhood of the rewiring point to obtain the optimized motion trajectory. The target motion trajectory of the robot is generated based on the optimized motion trajectory.

[0006] Optionally, the construction of the performance space corresponding to the robot-driven joint variables includes: The robot's drive joint variables are normalized into proportional parameters, and the proportional parameters are converted into performance space coordinates based on a pre-built performance space mapping function. Based on the performance space coordinates, a three-dimensional performance space is constructed using a preset scaling factor.

[0007] Optionally, within the performance space, the initial trajectory is obtained by using a forward hierarchical planning strategy to plan the motion trajectory of each pre-layered robot drive joint group, including: For the pre-layered basic layer robot drive joint group, a search tree is constructed in the first subspace of the performance space, and the first layer planning result is obtained by using the bias sampling strategy guided by the performance gradient. Based on the planning results of the first layer, for the second layer robot-driven joint group, motion trajectory planning is performed in the second subspace of the performance space, and the motion trajectory planning of other layers of robot-driven joint groups is performed in sequence to obtain the initial trajectory.

[0008] Optionally, the bias sampling point generation method of the bias sampling strategy is as follows: ; in, For bias sampling points, For random sampling points, For balance coefficient, For step size parameters, This represents the performance gradient.

[0009] Optionally, the step of optimizing the initial trajectory through cross-layer rewiring and using a virtual sampling strategy to perform a local fine-grained search in the neighborhood of the rewiring point to obtain the optimized motion trajectory includes: Starting from the bottom layer, perform cross-layer restepping operations layer by layer towards the base layer; For each of the cross-layer rewiring operations, multiple virtual sampling points are randomly generated within a preset radius centered on the rewiring point; Perform kinematic and dynamic constraint checks on each virtual sampling point to calculate the performance index of the trajectory segment passing through the virtual sampling point; The target virtual sampling point with the optimal performance index is determined, and the trajectory is updated using the target virtual sampling point to obtain the optimized motion trajectory.

[0010] Optionally, the calculation of the performance index of the trajectory segment passing through the virtual sampling points includes: The weighted sum of multiple sub-performance indices of the trajectory segment passing through virtual sampling points is calculated to obtain a balanced performance index, and the integral of the balanced performance index along the trajectory is maximized to evaluate the trajectory quality. The sub-performance indicators include at least two of the following: stiffness indicator, force / torque capacity indicator, speed / acceleration capacity indicator, and accuracy indicator.

[0011] Optionally, the motion planning method for the serial-parallel hybrid redundant robot further includes: The radius of the local fine-tuning search is dynamically adjusted based on the magnitude of the performance gradient; wherein the formula for calculating the radius is: ; in, This is the actual search radius. Based on the search radius, The gradient sensitivity coefficient, This is the descent gradient.

[0012] Secondly, the present invention also discloses a motion planning device for a serial-parallel hybrid redundant robot, wherein the device comprises: The performance space construction module is used to construct the performance space corresponding to the robot's drive joint variables; The motion trajectory planning module is used to plan the motion trajectory of the pre-layered robot drive joint groups within the performance space using a forward hierarchical planning strategy to obtain the initial trajectory; wherein, each layer of planning uses a performance gradient-guided bias sampling strategy to guide the bias sampling points to concentrate in the high-performance region; The optimization module is used to perform cross-layer rewiring optimization on the initial trajectory and to perform local fine-grained search in the neighborhood of the rewiring point using a virtual sampling strategy to obtain the optimized motion trajectory. The trajectory generation module is used to generate the target motion trajectory of the robot based on the optimized motion trajectory.

[0013] Thirdly, the present invention discloses a terminal, comprising: a memory, a processor, and a motion planning program for a serial-parallel hybrid redundant robot stored in the memory and executable on the processor, wherein the motion planning program for the serial-parallel hybrid redundant robot, when executed by the processor, implements the steps of the motion planning method for the serial-parallel hybrid redundant robot as described above.

[0014] Fourthly, the present invention discloses a computer-readable storage medium storing a computer program that can be executed to implement the steps of the motion planning method for a serial-parallel hybrid redundant robot as described above.

[0015] This invention provides a motion planning method, apparatus, terminal, and medium for a serial-parallel hybrid redundant robot. The motion planning method for the serial-parallel hybrid redundant robot includes: constructing a performance space corresponding to the robot's driven joint variables; within the performance space, using a forward hierarchical planning strategy to plan the motion trajectories of pre-layered robot driven joint groups to obtain initial trajectories; wherein, each layer of planning utilizes a performance gradient-guided bias sampling strategy to guide bias sampling points to concentrate in high-performance regions; performing cross-layer rewiring optimization on the initial trajectory, and using a virtual sampling strategy to perform a local fine-grained search in the neighborhood of the rewiring points to obtain the optimized motion trajectory; and generating the robot's target motion trajectory based on the optimized motion trajectory. Therefore, this invention transforms a high-dimensional planning problem into an efficient search in a low-dimensional space by constructing a performance space corresponding to the robot's driven joint variables. It adopts a strategy that combines forward hierarchical planning and inverse optimization. First, hierarchical bias sampling is performed in the performance space. Then, local fine-grained search is performed through cross-layer rewiring and virtual sampling strategies. This breaks through the limitations of discrete tree structures, achieves comprehensive optimization of multiple performance indicators, and effectively avoids singular configurations and obstacles. As a result, it can significantly improve sampling efficiency and trajectory quality in environments with various obstacle densities. Attached Figure Description

[0016] Figure 1 This is a flowchart of a preferred embodiment of the motion planning method for a serial-parallel hybrid redundant robot in this invention; Figure 2 This is a schematic diagram of a specific three-dimensional performance space structure disclosed in this invention; Figure 3 This is a schematic diagram of a specific motion planning hierarchy disclosed in this invention; Figure 4 This is a schematic diagram of a specific hierarchical planning and optimization process disclosed in this invention; Figure 5 This is a schematic diagram of a specific motion planning method for a serial-parallel hybrid redundant robot disclosed in this invention; Figure 6 This is a schematic diagram showing the distribution of a specific performance index in the performance space disclosed in this invention; Figure 7 This is a functional principle block diagram of a preferred embodiment of the motion planning device for a serial-parallel hybrid redundant robot in this invention; Figure 8 This is a functional principle block diagram of a preferred embodiment of the terminal in this invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0018] Please see Figure 1 , Figure 1 This is a flowchart of the motion planning method for the serial-parallel hybrid redundant robot in this invention. For example... Figure 1 As shown, the motion planning method for a serial-parallel hybrid redundant robot according to an embodiment of the present invention includes: Step S11: Construct the performance space corresponding to the robot's drive joint variables.

[0019] In this embodiment, robot-driven joint variables are normalized into proportional parameters, and these proportional parameters are converted into performance space coordinates based on a pre-constructed performance space mapping function. A three-dimensional performance space is then constructed based on these performance space coordinates and a preset scaling factor. Essentially, mapping the robot's joint configuration to a performance space that intuitively and comprehensively represents the system's performance—that is, a dimensionality-reduced space achieved by proportionally normalizing robot-driven joint variables—aims to simplify system performance analysis and reveal intrinsic relationships. Normalization into proportional parameters ensures that these parameters satisfy constraints. The proportional parameters are then converted into performance space coordinates through the mapping function, thus establishing the basic framework of the performance space. A scaling factor is introduced to construct a complete three-dimensional performance space. This performance space not only captures the proportional relationships between joint variables but also optimizes motion planning through an intuitive graphical representation. In other words, by normalizing the driven joint variables and constructing a performance space based on the normalized joint variable proportions, a high-dimensional planning problem is transformed into an efficient search in a low-dimensional space, improving search efficiency and significantly reducing computational complexity.

[0020] It should be noted that the dimension of the performance space is equal to the dimension of the robot's actual joint space. For example, see... Figure 2 As shown, the system is first initialized, including determining the initial and target states of the robotic arm, setting performance weight coefficients and algorithm parameters, and constructing a performance space mapping relationship. Next, in the performance space construction phase, the drive joint variables are normalized, and the proportional parameters are converted into performance space coordinates using a mapping function. Based on these performance space coordinates, a three-dimensional performance space is constructed using a preset scaling factor.

[0021] Step S12: Within the performance space, the motion trajectory of the pre-layered robot drive joint groups is planned using a forward hierarchical planning strategy to obtain the initial trajectory; wherein, each layer of planning uses a performance gradient-guided bias sampling strategy to guide the bias sampling points to concentrate in the high-performance region.

[0022] In this embodiment, within the performance space, a forward hierarchical planning strategy is used to plan the trajectories of different groups of drive joints. Each planning layer employs a performance gradient-guided bias sampling strategy to guide sampling points towards high-performance regions. It can be understood that the forward hierarchical planning uses a hierarchical strategy, see [link to previous section]. Figure 3 As shown, the drive joint is divided into multiple layers, and trajectory planning is performed layer by layer to optimize system performance.

[0023] Specifically, for the pre-defined base layer robot-driven joint groups, a search tree is constructed in the first subspace of the performance space, and a bias sampling strategy guided by performance gradients is used to obtain the first-layer planning result. Based on the first-layer planning result, for the second layer robot-driven joint groups, motion trajectory planning is performed in the second subspace of the performance space, and the motion trajectory planning for other layers of robot-driven joint groups is performed sequentially to obtain the initial trajectory. In other words, the first-layer planning involves constructing a search tree in the first subspace of the performance space for the first group of driven joints, using a bias sampling strategy guided by performance gradients; the second-layer planning involves fixing the first-layer planning result and performing trajectory planning in the second subspace of the performance space for the second group of driven joints; and the Nth-layer planning involves sequentially completing trajectory planning for all layers, where N is an integer greater than or equal to 2.

[0024] The bias sampling point generation method of the bias sampling strategy is as follows: ; in, For bias sampling points, For random sampling points, For balance coefficient, For step size parameters, This represents the performance gradient.

[0025] For example, the forward hierarchical planning is carried out layer by layer according to the hierarchical structure (L1, L2, L3). The L1 layer planning constructs a search tree in the performance space for the basic driving joint group and adopts a bias sampling strategy guided by performance gradient to improve efficiency. The L2 layer planning optimizes the intermediate layer joints based on the fixed L1 layer results to ensure motion continuity. The L3 layer planning completes the trajectory optimization of the final layer joints by solving the optimality equation.

[0026] Step S13: Optimize the initial trajectory by cross-layer rewiring and use a virtual sampling strategy to perform a local fine-grained search in the neighborhood of the rewiring point to obtain the optimized motion trajectory.

[0027] In this embodiment, a local fine-grained search is performed through cross-layer rewire routing and virtual sampling strategies. Specifically, the trajectory obtained from the forward hierarchical planning is optimized through cross-layer rewire routing, further improving global performance. The virtual sampling strategy, combined with local fine-grained search within the neighborhood of the optimization point, overcomes the limitations of the discrete tree structure. Specifically, starting from the bottom layer, cross-layer rewire routing operations are performed layer by layer towards the base layer. For each cross-layer rewire routing operation, multiple virtual sampling points are randomly generated within a preset radius centered on the rewire routing point. Kinematic and dynamic constraints are checked for each virtual sampling point to calculate the performance index of the trajectory segment passing through the virtual sampling point. The target virtual sampling point with the optimal performance index is determined, and the trajectory is updated using the target virtual sampling point to obtain the optimized motion trajectory.

[0028] Understandably, starting from the bottom layer, a better parent node is sought at the top layer. For each rewire operation, a virtual sampling strategy is used to generate multiple virtual sampling points around the rewire point. That is, multiple virtual sampling points are randomly generated within a preset radius centered on the rewire point. Kinematic and dynamic constraints are checked for each virtual sampling point to achieve collision detection and performance evaluation. Then, the performance index of the trajectory segment passing through the virtual sampling point is calculated. The virtual sampling point with the best performance index is selected as the new optimization result to update the trajectory and obtain the optimized motion trajectory.

[0029] In this embodiment, calculating the performance index of the trajectory segment passing through the virtual sampling points can specifically include: calculating a weighted sum of multiple sub-performance indices of the trajectory segment passing through the virtual sampling points to obtain a balanced performance index, and maximizing the integral of the balanced performance index along the trajectory to evaluate the trajectory quality; wherein, the sub-performance indices include at least two of the following: stiffness index, force / torque capacity index, velocity / acceleration capacity index, and accuracy index. The formula for calculating the balanced performance index is as follows: ; in, The total number of performance indicators. For the first The weighting coefficients of each performance indicator. For state The first Each performance metric value, For the first The maximum value of each performance metric.

[0030] Furthermore, the radius of the local fine-tuning search can be dynamically adjusted based on the magnitude of the performance gradient; the formula for calculating the radius is: ; in, This is the actual search radius. Based on the search radius, The gradient sensitivity coefficient, This is the descent gradient.

[0031] For example, see Figure 4 As shown, the inverse optimization process starts from layer L3 and proceeds upwards through cross-layer rewiring. A virtual sampling point enhancement strategy is employed to generate virtual sampling points around the rewiring points for local fine-grained searching, and the search radius is dynamically adjusted to accelerate convergence. Specifically, by searching for better parent node connections from the lower layers (e.g., L3) to the upper layers (e.g., L2 and L1), the trajectory is dynamically adjusted. Simultaneously, the virtual sampling point enhancement strategy generates virtual sampling points around the rewiring points for local fine-grained searching, overcoming the spatial limitations of the discrete tree structure. Furthermore, the adaptive search radius is dynamically adjusted based on the performance gradient, accelerating convergence and improving the quality of the solution. The trajectory generation and verification stage generates the final motion trajectory, performs collision detection and performance verification, and outputs the optimized joint trajectory to ensure practical feasibility.

[0032] Step S14: Generate the target motion trajectory of the robot based on the optimized motion trajectory.

[0033] In this embodiment, the robot's final motion trajectory is generated based on the optimized results. This trajectory can avoid singular configurations and obstacles.

[0034] As can be seen, in this embodiment of the invention, by constructing the performance space corresponding to the robot's driven joint variables, the high-dimensional planning problem is transformed into an efficient search in a low-dimensional space. By adopting a strategy that combines forward hierarchical planning and inverse optimization, hierarchical bias sampling is first performed in the performance space, and then local fine-grained search is performed through cross-layer rewiring and virtual sampling strategies. This breaks through the limitations of discrete tree structures, achieves comprehensive optimization of multiple performance indicators, and effectively avoids singular configurations and obstacles. As a result, sampling efficiency and trajectory quality can be significantly improved in various obstacle density environments.

[0035] For example, see Figure 5As shown, by constructing a performance space based on the normalized joint variable ratio, the high-dimensional planning problem is transformed into an efficient search in a low-dimensional space. A strategy combining forward hierarchical planning and inverse optimization is employed. First, hierarchical bias sampling and tree expansion are performed in the performance space to optimize the trajectories of parallel and serial mechanism joints respectively. Then, a local refined search is performed through cross-layer rewiring and virtual sampling strategies, overcoming the limitations of the discrete tree structure. Ultimately, comprehensive optimization of multiple performance indicators such as stiffness, force / torque load-bearing capacity, and velocity / acceleration output capacity is achieved, effectively avoiding singular configurations and obstacles. Sampling efficiency and trajectory quality are significantly improved in various obstacle density environments, making it suitable for robot operation tasks requiring high precision and high load capacity. The hierarchical planning and adaptive mechanism give the technical solution strong environmental adaptability, maintaining a high planning success rate in various obstacle density environments. Furthermore, through a built-in singularity detection and avoidance mechanism, singular configurations that may lead to system malfunction can be automatically identified and avoided, ensuring motion stability. In this study, the robot's joint configuration is mapped to a performance space that can intuitively and comprehensively represent the system's performance. A strategy combining forward hierarchical planning and inverse optimization is adopted, and a virtual sampling point enhancement mechanism is introduced to achieve comprehensive optimization of multiple performance indicators, rather than single-objective optimization. This results in better overall performance of the generated trajectory. Furthermore, thanks to the virtual sampling point enhancement strategy, the smoothness and continuity of the trajectory are effectively improved.

[0036] It should be noted that parameter settings include algorithm parameter configuration and performance weight configuration. In algorithm parameter configuration, the basic search radius is set based on the workspace size, the step size is adjusted according to the planning accuracy requirements, and the convergence threshold is set according to the accuracy requirements to balance computational efficiency and solution quality. Performance weight configuration adjusts the weight coefficients of each performance indicator according to the specific task requirements, such as... Figure 6 As shown, for example, high-precision tasks increase stiffness weight to improve stability, high-speed tasks increase speed capability weight to optimize motion performance, and heavy-load tasks focus on gravity / torque capacity weight to ensure output robustness, thereby achieving multi-objective optimization. Furthermore, special scenario handling addresses obstacle-dense environments and singular configuration avoidance. In obstacle-dense environments, complex constraints are addressed by increasing sampling density, decreasing step size parameters, and enhancing collision detection frequency, thereby improving planning success rate. For singular configuration avoidance, the determinant of the Jacobian matrix is ​​detected in real time, a safety threshold is set, and the trajectory is automatically adjusted to avoid singular regions, ensuring system control stability and motion safety.

[0037] In one embodiment, such as Figure 7 As shown, based on the above-described motion planning method for a serial-parallel hybrid redundant robot, the present invention also provides a motion planning device for a serial-parallel hybrid redundant robot, comprising: Performance space construction module 11 is used to construct the performance space corresponding to the robot drive joint variables; The motion trajectory planning module 12 is used to plan the motion trajectory of the pre-layered robot drive joint groups in the performance space using a forward hierarchical planning strategy to obtain the initial trajectory; wherein, each layer of planning uses a performance gradient-guided bias sampling strategy to guide the bias sampling points to concentrate in the high-performance region. The optimization module 13 is used to perform cross-layer rewiring optimization on the initial trajectory and to perform local fine-grained search in the neighborhood of the rewiring point using a virtual sampling strategy to obtain the optimized motion trajectory. The trajectory generation module 14 is used to generate the target motion trajectory of the robot based on the optimized motion trajectory.

[0038] Furthermore, it is worth noting that the working process of the motion planning device for a serial-parallel hybrid redundant robot provided in this embodiment is the same as the working process of the motion planning method for the serial-parallel hybrid redundant robot described above, and will not be repeated here. For details, please refer to the working process of the motion planning method for the serial-parallel hybrid redundant robot described above.

[0039] Figure 8 A schematic diagram of the structure of a terminal provided in an embodiment of this application. The terminal may include: The memory 501, the processor 502, and the computer program stored on the memory 501 and capable of running on the processor 502.

[0040] When the processor 502 executes the program, it implements the motion planning method for the serial-parallel hybrid redundant robot provided in the above embodiments.

[0041] Furthermore, the terminal also includes: Communication interface 503 is used for communication between memory 501 and processor 502.

[0042] The memory 501 is used to store computer programs that can run on the processor 502.

[0043] Memory 501 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0044] If the memory 501, processor 502, and communication interface 503 are implemented independently, they can be interconnected via a bus to communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, only one line is used in the diagram, but this does not imply that there is only one bus or one type of bus.

[0045] Optionally, in a specific implementation, if the memory 501, processor 502, and communication interface 503 are integrated on a single chip, then the memory 501, processor 502, and communication interface 503 can communicate with each other through an internal interface.

[0046] Processor 502 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0047] This embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the motion planning method for the serial-parallel hybrid redundant robot described above.

[0048] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the claims.

[0049] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0050] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can read and execute instructions from and from an instruction execution system, apparatus or device).

[0051] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0052] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for motion planning of a series-parallel hybrid redundant robot, characterized in that, The method comprises: constructing a performance space corresponding to robot driving joint variables; in the performance space, using a forward hierarchical planning strategy to respectively plan motion trajectories for robot driving joint groups hierarchically pre-divided to obtain initial trajectories; wherein each layer of planning uses a biased sampling strategy guided by a performance gradient to guide biased sampling points to concentrate in a high-performance area; cross-layer rewiring optimization is performed on the initial trajectories, and a virtual sampling strategy is used to perform local fine search in the neighborhood of rewiring points to obtain optimized motion trajectories; generating a target motion trajectory of the robot based on the optimized motion trajectories. 2.The motion planning method of the series-parallel hybrid redundant robot according to claim 1, wherein, The performance space corresponding to the robot driving joint variables is constructed, comprising: normalizing robot driving joint variables into scale parameters, and converting the scale parameters into performance space coordinates based on a pre-constructed performance space mapping function; constructing a three-dimensional performance space according to the performance space coordinates and using a pre-set scaling factor. 3.The motion planning method of the series-parallel hybrid redundant robot according to claim 1, wherein, The performance space corresponding to the robot driving joint variables is constructed, comprising: for the basic layer robot driving joint group hierarchically pre-divided, constructing a search tree in a first subspace of the performance space, and using a biased sampling strategy guided by a performance gradient to obtain a first layer planning result; based on the first layer planning result, for the second layer robot driving joint group, performing motion trajectory planning in a second subspace of the performance space, and sequentially performing motion trajectory planning for other layer robot driving joint groups to obtain initial trajectories.

4. The motion planning method of the series-parallel hybrid redundant robot according to claim 3, characterized in that, The biased sampling strategy has a biased sampling point generation mode: ； wherein, is a bias sampling point, is a random sampling point, is a balancing coefficient, is a step parameter, is a performance gradient. 5.The motion planning method of the series-parallel hybrid redundant robot according to claim 1, wherein, The cross-layer rewiring optimization is performed on the initial trajectories, and a virtual sampling strategy is used to perform local fine search in the neighborhood of rewiring points to obtain optimized motion trajectories, comprising: starting from the bottom layer, performing cross-layer rewiring operations layer by layer in the direction of the basic layer; for each cross-layer rewiring operation, a plurality of virtual sampling points are randomly generated within a preset radius centered on the rewiring point; performing kinematic constraint checking and dynamic constraint checking on each virtual sampling point to calculate the performance index of the trajectory segment passing through the virtual sampling point; determining a target virtual sampling point with the optimal performance index to update the trajectory using the target virtual sampling point to obtain the optimized motion trajectory. 6.The motion planning method of the series-parallel hybrid redundant robot according to claim 5, wherein, The performance index of the trajectory segment passing through the virtual sampling point is calculated, comprising: calculating the weighted sum of a plurality of sub-performance indexes of the trajectory segment passing through the virtual sampling point to obtain a balanced performance index, and maximizing the integral of the balanced performance index along the trajectory to evaluate the trajectory quality; wherein the sub-performance indexes include at least two of stiffness index, force / torque capacity index, speed / acceleration capability index and precision index.

7. The motion planning method of the series-parallel hybrid redundant robot according to any one of claims 1 to 6, characterized in that, Further comprising: dynamically adjusting the radius of the local fine search according to the size of the performance gradient; wherein the calculation formula of the radius is: ； wherein, is the actual search radius, is the base search radius, is the gradient sensitivity coefficient, is the step-down gradient.

8. A motion planning device for a series-parallel hybrid redundant robot, characterized in that, The device comprises: a performance space construction module for constructing a performance space corresponding to robot driving joint variables; The motion trajectory planning module is configured to perform motion trajectory planning on the robot driving joint groups in the performance space by using a forward hierarchical planning strategy to obtain initial trajectories, wherein each layer of planning uses a performance gradient guided biased sampling strategy to guide the biased sampling points to concentrate in a high performance area. The optimization module is configured to perform cross-layer rewiring optimization on the initial trajectories and perform local fine search in a neighborhood of a rewiring point by using a virtual sampling strategy to obtain optimized motion trajectories. The trajectory generation module is configured to generate target motion trajectories of the robot based on the optimized motion trajectories.

9. A terminal, characterized by comprising: The memory, the processor, and a motion planning program of a series-parallel hybrid redundant robot stored in the memory and executable on the processor, wherein the motion planning program of the series-parallel hybrid redundant robot, when executed by the processor, implements the steps of the motion planning method of the series-parallel hybrid redundant robot according to any one of claims 1 to 7. The computer readable storage medium stores a computer program, which can be executed to implement the steps of the motion planning method of the series-parallel hybrid redundant robot according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, ​

Images