A scheduling and path planning method and system for parallel collaborative work of multi-micro manipulator devices

Abstract

The application discloses a kind of multi-micro manipulator device parallel collaborative work scheduling and path planning method and system, comprising: according to system state space model, the task information of the task to be processed is obtained;According to the spatial distribution characteristics of target task in task information and task demand, the target task is grouped, and a plurality of task subgroups are obtained;Several task subgroups are assigned to corresponding micro manipulator device, and initial task allocation scheme is generated;In the process that micro manipulator device executes task, its execution state is monitored in real time, and according to the deviation of actual execution progress and expected planning, the remaining tasks are dynamically rescheduled, and updated task scheduling sequence is generated;According to the target position of each micro manipulator device, kinematic constraint and collision constraint, the optimal path without collision is planned for each micro manipulator device, and position calibration is carried out in the process of path execution.The method proposed in the application significantly improves the efficiency of multi-device parallel collaborative work.

Description

Technical Field

[0001] This invention belongs to the field of micromanipulation technology, and in particular relates to a scheduling and path planning method and system for multiple micromanipulation devices working in parallel and collaboratively. Background Technology

[0002] Micromanipulation devices (such as microneedles, microelectrodes, and micropipettes) are widely used in precision manipulation tasks such as cell injection, electrophysiological recording, and dye perfusion. These operations typically require precise positioning of micron-sized manipulation tools at the target cell location under microscopic observation to perform the specified operation. With the increasing demands of modern biomedical research for high-throughput screening, large-scale data acquisition, and high-efficiency experiments, the traditional serial operation mode, which relies on a single micromanipulation device to process multiple targets sequentially, is increasingly showing its limitations in terms of low throughput and long operation cycles, making it difficult to meet the efficiency requirements of practical applications.

[0003] To improve operational efficiency, existing technologies have proposed a scheme using multiple micromanipulators working in parallel and collaboratively. However, existing schemes still have the following technical shortcomings in practical applications: First, there is a lack of a unified mathematical modeling method for operating environments at the micrometer scale. Existing solutions struggle to systematically describe the key elements of multi-device collaborative operation systems, including: kinematic constraints of devices (such as velocity, acceleration, and direction of motion restrictions), spatial collision constraints between devices (such as minimum safe distance), obstacle avoidance constraints between devices and surrounding obstacles, and temporal dependencies between tasks. This lack of a unified model results in a lack of a reliable foundation for subsequent scheduling and planning.

[0004] Second, traditional scheduling algorithms struggle to adapt to the dynamic characteristics of micromanipulation systems. In actual operation, due to high micrometer-level precision requirements, complex operating environments, and fluctuations in device performance, deviations often occur between the actual execution progress and the planned schedule. Existing solutions lack effective real-time status monitoring and dynamic rescheduling mechanisms. When these deviations accumulate to a certain extent, it becomes difficult to adjust the execution sequence of remaining tasks online, resulting in poor system robustness and even operational failure.

[0005] Third, existing path planning methods struggle to balance computational efficiency and planning quality when micrometer-level accuracy is required. Micromanipulated devices need to move in complex three-dimensional spaces (such as tissue slices and culture dishes), requiring both micrometer-level positioning accuracy and avoiding collisions with other devices, obstacles, or target tissues. Existing methods often suffer from computationally intensive precise path planning, making it difficult to meet real-time requirements; while rapid planning methods frequently result in poor planning quality, with suboptimal paths or collision risks.

[0006] In summary, there is an urgent need for a method and system for collaborative scheduling and path planning of multiple micro-manipulators that can adapt to micrometer-scale operating environments, have dynamic adjustment capabilities, and balance planning efficiency and accuracy. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides a scheduling and path planning method for parallel collaborative operation of multiple micro-manipulation devices, comprising the following steps: Based on the system state-space model that includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters, obtain the task information of the task to be processed. Based on the spatial distribution characteristics and task requirements of the target tasks in the task information, the target tasks are grouped to obtain several task subgroups. Based on the spatial deployment location, functional characteristics, and kinematic constraints of each micromanipulator, the aforementioned task subgroups are assigned to the corresponding micromanipulators to generate an initial task allocation scheme. During the execution of tasks by the micromanipulator, its execution status is monitored in real time, and the remaining tasks are dynamically rescheduled based on the deviation between the actual execution progress and the expected plan, generating an updated task scheduling sequence. Based on the target position, kinematic constraints, and collision constraints of each micromanipulator, a collision-free optimal path is planned for each micromanipulator, and position calibration is performed during path execution.

[0008] Optionally, the device state parameters include the current spatial pose, kinematic parameters, functional attributes, and working status of each micromanipulation device; the target object parameters include the spatial coordinates, morphological features, and processing priority of the task or organization to be processed; the spatial constraint parameters include the minimum safe distance between devices, obstacle distribution, and accessibility boundaries; and the timing constraint parameters include the estimated execution time of a single task and the dependencies between tasks.

[0009] Optionally, based on the spatial distribution characteristics and task requirements of the target tasks in the task information, the target tasks are grouped to obtain several task subgroups. The specific process includes: A grouping model is constructed with the optimization objectives of maximizing the spatial clustering degree within the group, maximizing the feature difference between groups, and maximizing the difference of the grouping identifiers of adjacent tasks. The task to be processed is divided into several subgroups by solving a multi-objective optimization problem. The inter-group feature difference refers to the distinguishing index of the processing methods assigned to different groups.

[0010] Optionally, based on the spatial deployment location, functional characteristics, and kinematic constraints of each micromanipulator, the aforementioned task subgroups are assigned to the corresponding micromanipulators to generate an initial task allocation scheme. The specific process includes: A device-task allocation optimization model is constructed with the primary objective of minimizing task completion time. The device-task allocation optimization model comprehensively considers the geometric reachability score of each device to the target task group, the estimated total path length, the estimated total time cost, and the task load balance among devices. By solving the device-task allocation optimization model, each task group is allocated to the corresponding micromanipulation device.

[0011] Optionally, during the execution of tasks by the micromanipulator, its execution status is monitored in real time, and the remaining tasks are dynamically rescheduled based on the deviation between the actual execution progress and the expected plan, generating an updated task scheduling sequence. The specific process includes: The system collects execution data from each device in real time; it calculates a comprehensive deviation metric based on the execution data, taking into account time deviation, spatial deviation, and task completion deviation; when the comprehensive deviation metric exceeds a preset threshold, it re-solves the task scheduling optimization problem based on the current system state and with the optimization objective of maximizing the parallelism of subsequent tasks of each device.

[0012] Optionally, based on the target position, kinematic constraints, and collision constraints of each micromanipulator, a collision-free optimal path is planned for each micromanipulator, and position calibration is performed during path execution. The specific process includes: A sampling-based fast search algorithm is employed to generate an optimal path with micron-level accuracy that satisfies kinematic and collision constraints within a millimeter-level map space. The kinematic constraints include the maximum moving speed, maximum acceleration, and movement direction restrictions for each device. The collision constraints include safety distance constraints between devices, safety distance constraints between devices and static obstacles, and protection distance constraints between devices and target objects. During path execution, a visual feedback system is used at each planning node to detect the deviation between the actual position of the device and the target position and to generate compensating motion commands for micron-scale position calibration.

[0013] The present invention also provides a scheduling and path planning system for parallel collaborative operation of multiple micro-manipulation devices, used to implement the method, comprising: The system modeling module is used to perform mathematical modeling on a multi-micro-manipulation device collaborative operation system, and to establish a system state-space model that includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters. The task grouping module, connected to the system modeling module, is used to divide the target to be processed into several subgroups based on a multi-objective optimization algorithm. The device allocation module, connected to the task grouping module, is used to allocate each target group to the corresponding micromanipulation device based on a mixed integer programming model, and generate an initial task allocation scheme. The dynamic scheduling module, connected to the device allocation module, is used to monitor the system status in real time, generate the optimal task scheduling sequence, and trigger rescheduling when the deviation exceeds the threshold. The path planning module, connected to the dynamic scheduling module, is used to generate collision-free optimal paths that satisfy kinematic and collision constraints and to perform real-time error calibration.

[0014] Optionally, the dynamic scheduling module includes: The status monitoring submodule is used to collect the execution status data of each micro-manipulation device in real time; The deviation assessment submodule, connected to the status monitoring submodule, is used to calculate a comprehensive deviation metric from the expected plan based on the execution status data. The rescheduling trigger submodule, connected to the deviation evaluation submodule, is used to generate a trigger signal when the comprehensive deviation metric exceeds a preset threshold. The online optimization submodule, connected to the rescheduling trigger submodule, is used to re-solve the task scheduling optimization problem based on the current system state after receiving a trigger signal, and generate an updated task scheduling sequence.

[0015] Optionally, the path planning module includes: The path search submodule is used to generate a collision-free optimal path based on the target location, kinematic constraints, and collision constraints. The error calibration submodule, connected to the path search submodule, is used to detect the deviation between the actual position of the device and the target position through the visual feedback system and generate compensation motion commands to perform micron-scale position calibration during path execution.

[0016] The present invention also provides an electronic device, including a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0017] Compared with the prior art, the present invention has the following advantages and technical effects: First, this invention achieves a unified mathematical description of the micrometer-scale operating environment by constructing a system state-space model that includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters. On this basis, the complex multi-device collaborative scheduling problem is decoupled into four relatively independent sub-problems: task grouping, device allocation, dynamic scheduling, and path planning, and solved hierarchically. This significantly reduces the computational complexity of the problem and improves the real-time response capability of the system.

[0018] Second, based on the spatial distribution characteristics and task requirements of the target task, this invention constructs a grouping model with the optimization objectives of maximizing the spatial clustering degree within the group, maximizing the feature difference between groups, and maximizing the difference of the grouping identifiers of adjacent tasks. By solving this multi-objective optimization problem, the task to be processed is divided into several subgroups, so that spatially adjacent tasks can obtain differentiated processing identifiers, providing a high-quality raw data foundation for subsequent data analysis, image reconstruction, and signal differentiation.

[0019] Third, based on the spatial deployment location, functional characteristics and kinematic constraints of each micromanipulation device, the present invention constructs a device-task allocation optimization model with the main objective of minimizing task completion time. By comprehensively considering the geometric reachability score of each device to the target task group, the estimated total path length, the estimated total time cost and the task load balance among devices, the present invention achieves the optimal matching of tasks and devices, effectively shortens the overall task completion time and improves device utilization.

[0020] Fourth, by establishing a dynamic scheduling model for parallel operation of multiple devices, this invention monitors the execution status of each device in real time. When the deviation between the actual execution progress and the expected plan exceeds a preset threshold, the task scheduling optimization problem is re-solved based on the current system state. This effectively addresses the unavoidable execution errors and uncertainties in micro-manipulation systems, and improves the system's adaptability and robustness to dynamic environments.

[0021] Fifth, this invention constructs a path planning model that considers kinematic constraints and collision constraints for the target positions of each micro-manipulation device, and adopts a sampling-based fast search algorithm to generate a collision-free optimal path with micron-level accuracy in a millimeter-level map space. During the path execution process, micron-scale position calibration is performed at each node. Under the premise of ensuring collision-free safety, it can meet the dual requirements of real-time performance and accuracy of the micro-manipulation system in milliseconds.

[0022] In summary, by organically combining the above-mentioned technical means, this invention achieves multi-objective optimization with the shortest task completion time, the lowest device idle rate, the shortest movement path, and no collisions throughout the process, significantly improving the efficiency of parallel collaborative operation of multiple devices. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the system structure according to an embodiment of the present invention. Detailed Implementation

[0024] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0025] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0026] Example 1 like Figure 1 As shown, this embodiment provides a scheduling and path planning method for parallel collaborative operation of multiple micro-manipulation devices, including the following steps: Based on the system state-space model that includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters, obtain the task information of the task to be processed. Based on the spatial distribution characteristics and task requirements of the target tasks in the task information, the target tasks are grouped to obtain several task subgroups. Based on the spatial deployment location, functional characteristics, and kinematic constraints of each micromanipulator, the aforementioned task subgroups are assigned to the corresponding micromanipulators to generate an initial task allocation scheme. During the execution of tasks by the micromanipulator, its execution status is monitored in real time, and the remaining tasks are dynamically rescheduled based on the deviation between the actual execution progress and the expected plan, generating an updated task scheduling sequence. Based on the target position, kinematic constraints, and collision constraints of each micromanipulator, a collision-free optimal path is planned for each micromanipulator, and position calibration is performed during path execution.

[0027] As a specific implementation method, the method proposed in this embodiment is applied to the human brain neuron cell dye perfusion scenario, using four microbeams loaded with different fluorescent dyes to differentially label neurons in brain tissue slices.

[0028] It is feasible to mathematically model a multi-micromanipulation device collaborative operation system based on device state parameters, target object parameters, spatial constraint parameters, and temporal constraints, and establish a system state space model. The device state parameters include the current spatial pose, kinematic parameters, functional attributes, and working state of each micromanipulation device; the target object parameters include the spatial coordinates, morphological characteristics, and processing priority of the task or organization to be processed; the spatial constraint parameters include the minimum safe distance between devices, obstacle distribution, and reachability boundaries; and the temporal constraint parameters include the estimated execution time of a single task and the dependencies between tasks.

[0029] As one specific implementation method, the parameters for system modeling include: (1) Device state parameter modeling: Suppose the system contains M microneedles, denoted as set N = {N1, N2, ..., N}. M Each microneedle N i The state parameters are defined as follows: N i =(P i V i ,F i ,S i ); Where: P i =(x i ,y i ,z i ,θ i ,φ i (x) represents the spatial pose of the microneedle. i ,y i ,z i (θ) represents the three-dimensional spatial coordinates. i ,φ i ) represents the attitude angle; V i =(v max,i ,a max,i ) represents the kinematic parameters, namely maximum velocity and maximum acceleration; F i ∈{1,2,...,N} represents the dye type number loaded on the microneedles; S i The enumeration ∈{idle,moving,operating,retreating} represents the working state.

[0030] (2) Target object parameter modeling: Let the target cell set be C = {C1, C2, ..., C}. K}, each cell C j The parameter is defined as follows: C j =(Pos j Mor j ,Pri j ,Adj j ); Where: Pos j =(x j ,y j ,z j () represents the spatial coordinates of the cell body center; Mor j This is a vector representing cell morphological features, such as cell size and depth within the tissue; Pri j ∈ℝ⁺ indicates processing priority; Adj j={j1,j2,...} is the set of indices of spatially adjacent cells.

[0031] (3) Spatial constraint parameter modeling: The set of spatial constraint parameters S includes: minimum safe distance between devices: d safeNN Typical value is 120μm; minimum safe distance between device and obstacle: d safeNO Typical value is 100 μm; Device-to-target tissue protection distance: d protect Typical value is 100 μm; the working space boundary that each device can reach: W i ,i=1,2,...,M.

[0032] (4) Modeling of time-constrained parameters: The timing constraint parameters are defined as follows: Single-cell operation duration prediction T op Including positioning time t loc Puncture time t punc Infusion time t inj Withdrawal time t ret ; The task dependency matrix D∈{0,1} K×K D jk =1 indicates cell C j Must be in C k The operation was completed previously.

[0033] Implementably, based on the spatial distribution characteristics and task requirements of the target tasks in the task information, the target tasks are grouped to obtain several task subgroups. The specific process includes: A grouping model is constructed with the optimization objectives of maximizing the spatial clustering degree within a group, maximizing the feature difference between groups, and maximizing the difference in the grouping identifiers of adjacent tasks. The task to be processed is divided into several subgroups by solving a multi-objective optimization problem. The feature difference between groups refers to the distinguishing index of the processing methods assigned to different groups.

[0034] As a specific implementation method, the specific process of task grouping includes: Based on the spatial distribution characteristics of the target task group, a multi-objective optimization model is constructed to divide the K tasks into N groups, each corresponding to N different fluorescent dyes.

[0035] (1) Definition of decision variables: Define the grouping indicator variable g jn ∈{0,1}, where g jn =1 indicates task C j It is assigned to the nth group (using the nth dye).

[0036] (2) Constraints: Each task must be assigned to one group only: Σ n=1 N g jn =1,∀j∈{1,2,...,K}; (3) Optimization objective: Objective 1: Maximize the difference between adjacent task groups: Define the difference index between adjacent task cells: f1=Σ j=1 K Σ k∈Adjj Σ n=1 N (1−g jn ·g kn ); The goal is to assign different colored dyes to spatially adjacent neurons as much as possible.

[0037] Objective 2: Maximize intra-group spatial clustering: Define the spatial clustering degree of the nth group: f2=−Σ n=1 N Σ j:gjn=1 Σ k:gkn=1 Pos j -Pos k ||²; This objective allows cells within the same group to be spatially clustered, reducing the movement path during subsequent device allocation.

[0038] Objective 3: Balance of group size: f3=−Σ n=1 N |Σ j=1 K g jn −K / N|; Where K / N represents the average number of tasks that should be assigned to each group.

[0039] In practice, based on the spatial deployment location, functional characteristics, and kinematic constraints of each micromanipulator, the aforementioned task subgroups are assigned to the corresponding micromanipulators to generate an initial task allocation scheme. The specific process includes: A device-task allocation optimization model is constructed with the primary objective of minimizing task completion time. The device-task allocation optimization model comprehensively considers the geometric reachability score of each device to the target task group, the estimated total path length, the estimated total time cost, and the task load balance among devices. By solving the device-task allocation optimization model, each task group is allocated to the corresponding micromanipulation device.

[0040] As one specific implementation method, the device allocation process includes: Based on the spatial deployment location of each microneedle and the type of dye loaded, the cells are optimally matched with the microneedles.

[0041] (1) Definition of decision variables: Define the assignment indicator variable a ni ∈{0,1}, where a ni =1 indicates that the nth cell group is composed of microneedles N. i Responsible for handling this.

[0042] (2) Constraints: Dye type matching constraints: a ni =0,ifF i ≠n; That is, microneedles can only process cell groups that correspond to the type of dye they carry.

[0043] Each group is managed by a single microneedle: Σ i=1 M a ni =1,∀n∈{1,2,...,N}; (3) Optimization objective: The primary objective is to minimize the maximum completion time (makespan). minmax i∈{1,...,M} T i ; Where T i microneedle N i Estimated total time to complete all assigned tasks: T i =Σ n:ani=1 Σ j:gjn=1 (T op +T travel,ij ); T travel,ij microneedle N i The movement time between adjacent tasks. Auxiliary objectives include minimizing the total path length and load balancing.

[0044] Feasible practice is to monitor the execution status of the micromanipulator in real time during task execution, and dynamically reschedule the remaining tasks based on the deviation between the actual execution progress and the expected plan, generating an updated task scheduling sequence. The specific process includes: The system collects execution data from each device in real time; it calculates a comprehensive deviation metric based on the execution data, taking into account time deviation, spatial deviation, and task completion deviation; when the comprehensive deviation metric exceeds a preset threshold, it re-solves the task scheduling optimization problem based on the current system state and with the optimization objective of maximizing the parallelism of subsequent tasks of each device.

[0045] As a specific implementation method, the dynamic scheduling process includes: A dynamic scheduling model is established to monitor the execution status during system operation and trigger rescheduling when necessary. The scheduling employs an NSGAII-based optimization algorithm to quickly solve a pre-constructed multi-objective optimization function.

[0046] (1) Status monitoring: Real-time acquisition of the location information of each microneedle, the current task execution progress, and the remaining task queue.

[0047] (2) Deviation assessment: Calculate the overall deviation metric: Δ=w t ·Δ t +w s ·Δ s +w c ·Δ c ; Where: Δ t =|t actual -t planned | represents the time deviation; Δ s =‖P actual -P planned || represents spatial position deviation; Δ c =|progress actual -progress planned | represents the deviation in task completion; w t ,w s ,w c These are the weighting coefficients for each deviation.

[0048] (3) Rescheduling trigger: When Δ>Δ threshold When this occurs, a rescheduling mechanism is triggered, where Δ threshold This is the threshold for triggering rescheduling.

[0049] (4) Online optimization: During rescheduling, the task scheduling problem is resolved based on the current real-time state of the system: the completed and currently executing tasks are frozen, and the execution sequence of the remaining tasks is replanned; the optimization objective is to maximize the parallelism of subsequent tasks of each device.

[0050] Parallelism is defined as: Parallelism=Σ t (Number of devices operating simultaneously) / (Total time window × M).

[0051] Feasible, based on the target position, kinematic constraints, and collision constraints of each micromanipulator, a collision-free optimal path is planned for each micromanipulator, and position calibration is performed during path execution. The specific process includes: A sampling-based fast search algorithm is employed to generate an optimal path with micron-level accuracy that satisfies kinematic and collision constraints within a millimeter-level map space. The kinematic constraints include the maximum moving speed, maximum acceleration, and movement direction restrictions for each device. The collision constraints include safety distance constraints between devices, safety distance constraints between devices and static obstacles, and protection distance constraints between devices and target objects. During path execution, a visual feedback system is used at each planning node to detect the deviation between the actual position of the device and the target position and to generate compensating motion commands for micron-scale position calibration.

[0052] As a specific implementation method, the path planning process includes: For each microneedle's target location, a collision-free optimal path is generated and real-time calibration is performed.

[0053] (1) Path planning algorithm: Adopting an improved RRT The Rapidly-exploring RandomTreeStar algorithm performs path search within the three-dimensional workspace, comprehensively considering the geometric constraints, kinematic constraints, and obstacles of the microneedles. Algorithm pseudocode: Input: Starting position P _start Target location P _goal Obstacle set Obs; Output: Collision-free optimal path; 1. Initialize the search tree T, with P_start as the root node; 2.FORiter=1TOmax_iterationsDO; 3.P_rand←SampleRandomPoint(); 4.P_nearest←FindNearestNode(T,P_rand); 5.P_new←Steer(P_nearest,P_rand,step_size); 6.IFCollisionFree(P_nearest,P_new,Obs)THEN; 7.P_neighbors←FindNearbyNodes(T,P_new,radius); 8.P_parent←ChooseBestParent(P_neighbors,P_new); 9.AddNode(T,P_new,P_parent); 10.Rewire(T,P_neighbors,P_new); 11.ENDIF; 12. IFDistance(P_new, P_goal) <thresholdTHEN; 13.RETURNExtractPath(T,P_goal); 14. ENDIF; 15. ENDFOR.

[0054] (2) Collision detection: The CollisionFree function checks the following constraints: Distance between devices: ||P i (t)−P j (t)‖≥d safeNN ,∀i≠j; Distance between device and obstacle: min o∈Obs ||P i (t)−o‖≥d safeNO ; Distance between device and target tissue: For non-target areas, ||P i (t)−tissue‖≥d protect ; (3) Kinematic constraints: The path execution satisfies kinematic constraints: Maximum speed constraint: ‖Ṗ i (t)‖≤v max,i ; Maximum acceleration constraint: ||P̈ i (t)‖≤a max,i ; (4) Error calibration: During path execution, error calibration is performed at each planning node: The actual position P of the device is acquired through a visual feedback system. actual ; Calculate the positional deviation ε=P target -P actual If ||ε||>ε threshold Generate compensation motion commands, where ε thresholdThis is the position deviation tolerance threshold.

[0055] On the other hand, such as Figure 2 As shown, this embodiment also provides a scheduling and path planning system for parallel collaborative operation of multiple micro-manipulation devices, used to implement the method, including: The system modeling module is used to perform mathematical modeling on a multi-micro-manipulation device collaborative operation system, and to establish a system state-space model that includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters. The task grouping module, connected to the system modeling module, is used to divide the target to be processed into several subgroups based on a multi-objective optimization algorithm. The device allocation module, connected to the task grouping module, is used to allocate each target group to the corresponding micromanipulation device based on a mixed integer programming model, and generate an initial task allocation scheme. The dynamic scheduling module, connected to the device allocation module, is used to monitor the system status in real time, generate the optimal task scheduling sequence, and trigger rescheduling when the deviation exceeds the threshold. The path planning module, connected to the dynamic scheduling module, is used to generate collision-free optimal paths that satisfy kinematic and collision constraints and to perform real-time error calibration.

[0056] Implementable, the dynamic scheduling module includes: The status monitoring submodule is used to collect the execution status data of each micro-manipulation device in real time; The deviation assessment submodule, connected to the status monitoring submodule, is used to calculate a comprehensive deviation metric from the expected plan based on the execution status data. The rescheduling trigger submodule, connected to the deviation evaluation submodule, is used to generate a trigger signal when the comprehensive deviation metric exceeds a preset threshold. The online optimization submodule, connected to the rescheduling trigger submodule, is used to re-solve the task scheduling optimization problem based on the current system state after receiving a trigger signal, and generate an updated task scheduling sequence.

[0057] Implementable, the path planning module includes: The path search submodule is used to generate a collision-free optimal path based on the target location, kinematic constraints, and collision constraints. The error calibration submodule, connected to the path search submodule, is used to detect the deviation between the actual position of the device and the target position through the visual feedback system and generate compensation motion commands to perform micron-scale position calibration during path execution.

[0058] On the other hand, this embodiment also provides an electronic device, including a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0059] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A scheduling and path planning method for parallel collaborative operation of multiple micro-manipulation devices, characterized in that, Includes the following steps: Based on the system state space model, which includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters, obtain the task information of the task to be processed. Based on the spatial distribution characteristics and task requirements of the target tasks in the task information, the target tasks are grouped to obtain several task subgroups. Based on the spatial deployment location, functional characteristics, and kinematic constraints of each micromanipulator, the aforementioned task subgroups are assigned to the corresponding micromanipulators to generate an initial task allocation scheme. During the execution of tasks by the micromanipulator, its execution status is monitored in real time, and the remaining tasks are dynamically rescheduled based on the deviation between the actual execution progress and the expected plan, generating an updated task scheduling sequence. Based on the target position, kinematic constraints, and collision constraints of each micromanipulator, a collision-free optimal path is planned for each micromanipulator, and position calibration is performed during path execution. Device status parameters include the current spatial pose, kinematic parameters, functional attributes, and working status of each micromanipulator; target object parameters include the spatial coordinates, morphological features, and processing priority of the task or organization to be processed; spatial constraint parameters include the minimum safe distance between devices, obstacle distribution, and reachability boundaries; and timing constraint parameters include the estimated execution time of a single task and the dependencies between tasks. Based on the spatial distribution characteristics and task requirements of the target tasks in the task information, the target tasks are grouped to obtain several task subgroups. The specific process includes: A grouping model is constructed with the optimization objectives of maximizing the spatial clustering degree within the group, maximizing the feature difference between groups, and maximizing the difference of the grouping identifiers of adjacent tasks. The task to be processed is divided into several subgroups by solving a multi-objective optimization problem. The inter-group feature difference refers to the distinguishability index of the processing methods assigned to different groups; Based on the spatial deployment location, functional characteristics, and kinematic constraints of each micromanipulator, the aforementioned task subgroups are assigned to the corresponding micromanipulators to generate an initial task allocation scheme. The specific process includes: A device-task allocation optimization model is constructed with the primary objective of minimizing task completion time. The device-task allocation optimization model comprehensively considers the geometric reachability score of each device to the target task group, the estimated total path length, the estimated total time cost, and the task load balance among devices. By solving the device-task allocation optimization model, each task group is allocated to the corresponding micromanipulation device. During the execution of tasks by the micromanipulator, its execution status is monitored in real time, and the remaining tasks are dynamically rescheduled based on the deviation between the actual execution progress and the expected plan, generating an updated task scheduling sequence. The specific process includes: The system collects execution data from each device in real time; it calculates a comprehensive deviation metric based on the execution data, taking into account time deviation, spatial deviation, and task completion deviation; when the comprehensive deviation metric exceeds a preset threshold, it re-solves the task scheduling optimization problem based on the current system state and with the optimization objective of maximizing the parallelism of subsequent tasks of each device.

2. The method according to claim 1, characterized in that, Based on the target position, kinematic constraints, and collision constraints of each micromanipulator, a collision-free optimal path is planned for each micromanipulator, and position calibration is performed during path execution. The specific process includes: A sampling-based fast search algorithm is employed to generate an optimal path with micron-level accuracy that satisfies kinematic and collision constraints within a millimeter-level map space. The kinematic constraints include the maximum moving speed, maximum acceleration, and movement direction restrictions for each device. The collision constraints include safety distance constraints between devices, safety distance constraints between devices and static obstacles, and protection distance constraints between devices and target objects. During path execution, a visual feedback system is used at each planning node to detect the deviation between the actual position of the device and the target position and to generate compensating motion commands for micron-scale position calibration.

3. A scheduling and path planning system for parallel collaborative operation of multiple micro-manipulation devices, characterized in that, For implementing the method according to any one of claims 1-2, comprising: The system modeling module is used to perform mathematical modeling on a multi-micro-manipulation device collaborative operation system, and to establish a system state-space model that includes device state parameters, target object parameters, spatial constraint parameters, and timing constraint parameters. The task grouping module, connected to the system modeling module, is used to divide the target to be processed into several subgroups based on a multi-objective optimization algorithm. The device allocation module, connected to the task grouping module, is used to allocate each target group to the corresponding micromanipulation device based on a mixed integer programming model, and generate an initial task allocation scheme. The dynamic scheduling module, connected to the device allocation module, is used to monitor the system status in real time, generate the optimal task scheduling sequence, and trigger rescheduling when the deviation exceeds the threshold. The path planning module, connected to the dynamic scheduling module, is used to generate collision-free optimal paths that satisfy kinematic and collision constraints and to perform real-time error calibration.

4. The system according to claim 3, characterized in that, The dynamic scheduling module includes: The status monitoring submodule is used to collect the execution status data of each micro-manipulation device in real time; The deviation assessment submodule, connected to the status monitoring submodule, is used to calculate a comprehensive deviation metric from the expected plan based on the execution status data. The rescheduling trigger submodule, connected to the deviation evaluation submodule, is used to generate a trigger signal when the comprehensive deviation metric exceeds a preset threshold. The online optimization submodule, connected to the rescheduling trigger submodule, is used to re-solve the task scheduling optimization problem based on the current system state after receiving a trigger signal, and generate an updated task scheduling sequence.

5. The system according to claim 3, characterized in that, The path planning module includes: The path search submodule is used to generate a collision-free optimal path based on the target position, kinematic constraints, and collision constraints. The error calibration submodule, connected to the path search submodule, is used to detect the deviation between the actual position of the device and the target position through the visual feedback system and generate compensation motion commands to perform micron-scale position calibration during path execution.

6. An electronic device comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, characterized in that, When the processor executes the computing program, it implements the method described in any one of claims 1-2.

Images