A yard robot multi-task cooperative scheduling method and system

Abstract

The application provides a kind of yard robot multitask coordination scheduling method and system, it is related to yard service robot technical field.Construct four-layer system architecture of perception layer-decision layer-execution layer-coordination layer;Core uses multidimensional dynamic priority algorithm, in combination with task urgency, type weight, environment state, time constraint four dimensions real-time calculation task priority and generate execution sequence;Support carry intelligent task recommendation model based on LSTM+DQN, periodic state snapshot breakpoint recovery mechanism, based on auction algorithm+Hungarian algorithm Multi-robot coordination allocation strategy, while integrating task conflict detection, load balancing, whole-process monitoring and multi-modal user interaction module.The application can be self-adaptive according to weather, time period, site change Adjustment task order, interrupt task can continue from breakpoint, when multiple robots work cooperatively, automatically allocate tasks, avoid path conflict, effectively improve work efficiency, reduce resource consumption and artificial intervention frequency.

Description

Technical Field

[0001] This invention relates to the field of yard service robot technology, specifically to a multi-task collaborative scheduling method and system for yard robots. Background Technology

[0002] A yard service robot is an intelligent device capable of autonomously navigating, cleaning, maintaining, and monitoring tasks in a yard environment. With the acceleration of urbanization and the improvement of residents' living standards, the number of private and public yards is constantly increasing, and the demand for yard maintenance is also growing. As an efficient and intelligent yard maintenance solution, yard service robots have been widely used in residential yards, commercial properties, and public facilities. With the rapid development of artificial intelligence and sensor technologies, the intelligence level of yard service robots is constantly improving, and they can now perform multiple functions such as leaf sweeping, lawn mowing, irrigation and fertilization, and security monitoring.

[0003] Existing yard robots typically manage tasks using single-task sequential execution or fixed-priority scheduling. This scheduling method presents the following technical problems:

[0004] First, fixed priority scheduling cannot adapt to the complex and ever-changing courtyard environment. When environmental conditions change, the set priority cannot be dynamically adjusted, resulting in low task execution efficiency.

[0005] Second, traditional scheduling methods lack the ability to intelligently perceive environmental conditions and cannot proactively recommend and adjust tasks based on factors such as weather, season, and time.

[0006] Third, the single-task execution mode lacks an effective breakpoint resume mechanism when faced with interruptions such as insufficient power or obstacles, requiring the entire task to be restarted, resulting in a waste of time and resources.

[0007] Fourth, existing multi-robot collaborative scheduling schemes are relatively simple, lack global optimization and dynamic resource allocation capabilities, and cannot fully leverage the advantages of multi-robot collaborative operations.

[0008] Therefore, those skilled in the art provide a multi-task collaborative scheduling method and system for yard robots to solve the problems mentioned in the background art. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides a multi-task collaborative scheduling method and system for garden robots, solving problems such as fixed scheduling rules, lack of environmental adaptation, lack of breakpoint resume, inefficient multi-machine collaboration, frequent task conflicts, and low level of intelligence in existing technologies.

[0010] To achieve the above objectives, the present invention provides the following technical solution:

[0011] A method for multi-task collaborative scheduling of a garden robot includes the following steps:

[0012] Step 1: Obtain the perception data of the courtyard environment and the list of tasks to be performed;

[0013] Step 2: Based on the perceived data and the task list, establish a multi-task priority dynamic sorting algorithm model;

[0014] Step 3: Based on the multi-task priority dynamic sorting algorithm model, calculate the priority score of each task in the task list;

[0015] Step 4: Sort the task list according to the priority score of each task to generate a task execution sequence;

[0016] Step 5: Execute each task sequentially according to the task execution sequence.

[0017] Preferably, the multi-task priority dynamic ranking algorithm model includes an urgency assessment unit, a type weighting unit, a state awareness unit, and a time constraint unit; the urgency assessment unit is used to calculate an urgency score based on the urgency of the task; the type weighting unit is used to assign weight coefficients according to the task type; the state awareness unit is used to adjust the weights according to the current environmental state; the time constraint unit is used to calculate a time score based on the task's time constraints; the priority score is calculated using the following formula:

[0018] P = α × E + β × T + γ × S + δ × Time, where P is the priority score, E is the urgency score, T is the task type score, S is the environmental status score, Time is the time score, and α, β, γ, and δ are preset weight coefficients.

[0019] Preferably, adjusting the weights according to the current environmental state includes:

[0020] Increase the weight of leaf sweeping tasks when precipitation is detected;

[0021] When high temperatures are detected, increase the frequency of irrigation tasks;

[0022] When nighttime conditions are detected, priority is given to performing lighting maintenance and security patrols.

[0023] Preferably, the method further includes an intelligent task recommendation step: collecting historical operation data and environmental change data of the courtyard; constructing an intelligent task recommendation model based on the historical operation data and environmental change data; using the intelligent task recommendation model to predict potential cleaning areas and optimal operation times; and automatically generating recommended tasks based on the prediction results and adding them to the task list.

[0024] Preferably, the intelligent task recommendation model is built based on machine learning algorithms and includes a regional pollution level prediction sub-model and an operation timing optimization sub-model.

[0025] The regional pollution prediction sub-model is used to predict the degree of leaf accumulation in each area of ​​the courtyard.

[0026] The operation timing optimization sub-model is used to determine the optimal cleaning time for each area.

[0027] Preferably, the method further includes a task execution interruption and recovery step: during task execution, a snapshot of the current task execution state is periodically saved; when an external interruption event is detected, the interruption location and environmental map marker are recorded; when the interruption is recovered, the task continues to be executed from the interruption location based on the state snapshot and environmental map marker.

[0028] Preferably, the status snapshot includes: current task identifier, task execution progress percentage, robot current position coordinates, completed work areas, and uncompleted work areas.

[0029] Preferably, the external interruption events include: low battery warning, obstruction by obstacles, user manual pause, and communication interruption.

[0030] Preferably, the method further includes a multi-robot collaborative scheduling step: when multiple yard robots are detected in the same work area, a collaborative scheduling mode is activated; tasks are allocated according to the current position, task load and work capacity of each robot; a collaborative path plan is generated to avoid multi-robot path conflicts; and the work status of each robot is monitored in real time to dynamically allocate resources.

[0031] Preferably, the task allocation adopts a task allocation strategy based on an auction algorithm, including the following sub-steps:

[0032] S1: Each robot quotes a price for the task based on its own load and capabilities;

[0033] S2: After collecting all bids, the central scheduler calculates the optimal allocation scheme;

[0034] S3: Distribute the allocation results to each robot for execution.

[0035] Preferably, the multi-task collaborative scheduling system for the courtyard robot includes a perception layer, a decision-making layer, an execution layer, and a collaboration layer;

[0036] The perception layer is responsible for acquiring various types of perception data of the courtyard environment, including spatial distance information acquired by lidar, visual image information acquired by camera, meteorological environment information acquired by temperature and humidity sensor, obstacle information acquired by ultrasonic sensor, etc. The perception data is transmitted to the decision layer after preprocessing.

[0037] The multi-task priority dynamic ranking algorithm model in the decision-making layer calculates the priority score of each task based on these data and generates a task execution sequence.

[0038] The execution layer controls each actuator of the garden robot to complete the corresponding task according to the task execution sequence;

[0039] The collaboration layer is responsible for task allocation, path coordination, and resource scheduling in multi-robot scenarios.

[0040] This invention provides a multi-task collaborative scheduling method and system for a yard robot. It has the following beneficial effects:

[0041] 1. This invention provides a multi-task collaborative scheduling method and system for a garden robot. Through a multi-task priority dynamic sorting algorithm model, it realizes dynamic calculation and real-time adjustment of task priorities, improves the adaptability and flexibility of task scheduling, and enables the garden robot to adjust the task execution order in a timely manner according to environmental changes.

[0042] 2. This invention provides a multi-task collaborative scheduling method and system for yard robots. Through an intelligent task recommendation model, it predicts potential cleaning areas and optimal operation times based on historical data and environmental changes, thereby improving cleaning coverage and operation efficiency.

[0043] 3. This invention provides a multi-task collaborative scheduling method and system for a garden robot. Through a task execution interruption and recovery mechanism, it ensures stable operation under interruption conditions such as insufficient power or obstruction by obstacles, and avoids the waste of time and resources caused by repetitive operations.

[0044] 4. This invention provides a multi-task collaborative scheduling method and system for garden robots. Through a multi-robot collaborative scheduling architecture, it realizes task allocation, path coordination and resource optimization in multi-robot scenarios, giving full play to the advantages of large-scale operations. Attached Figure Description

[0045] Figure 1 This is a flowchart of the multi-task collaborative scheduling method for a garden robot according to the present invention;

[0046] Figure 2 This is a schematic diagram of the four-layer overall architecture of the multi-task collaborative scheduling system for the courtyard robot of the present invention;

[0047] Figure 3This is a schematic diagram of the multi-task priority algorithm module of the present invention. Detailed Implementation

[0048] The present invention will now be described in further detail 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 not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0049] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0050] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms "a," "the," and "the" as used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0051] To better understand the above technical solutions, the technical solutions of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0052] like Figures 1-3 As shown in some embodiments of this application, the overall architecture of the multi-task collaborative scheduling system for courtyard robots includes four main components: a perception layer, a decision-making layer, an execution layer, and a collaboration layer.

[0053] The perception layer is responsible for acquiring various perception data of the courtyard environment, including spatial distance information acquired through lidar, visual image information acquired through cameras, meteorological environmental information acquired through temperature and humidity sensors, obstacle information acquired through ultrasonic sensors, etc. The perception data is transmitted to the decision layer after preprocessing.

[0054] The multi-task priority dynamic ranking algorithm model in the decision-making layer calculates the priority score of each task based on these data and generates a task execution sequence.

[0055] The execution layer controls the various actuators of the yard robot to complete the corresponding tasks according to the task execution sequence.

[0056] The collaboration layer is responsible for task allocation, path coordination, and resource scheduling in multi-robot scenarios.

[0057] In some embodiments of this application, the multi-task priority dynamic ranking algorithm model is the core component of the system, and its design fully considers the complexity and variability of the courtyard environment. The algorithm model adopts a modular design, including four functional modules: an urgency assessment unit, a type weighting unit, a state perception unit, and a time constraint unit. The urgency assessment unit is responsible for assessing the urgency of each task, based on factors such as the task's lifecycle threshold, accumulated waiting time, and the triggering status of related events. The type weighting unit is responsible for assigning basic weight coefficients to different types of tasks; these coefficients can be configured according to the actual application scenario. The time constraint unit is responsible for calculating the task's time score, comprehensively evaluating it based on factors such as whether the task has a deadline, the proximity of the deadline, and the task's time window constraints.

[0058] In practice, the priority score is calculated using a weighted summation formula:

[0059] P = α × E + β × T + γ × S + δ × Time, where P is the overall priority score, E is the urgency score, T is the task type score, S is the environmental status score, Time is the time score, and α, β, γ, and δ are preset weight coefficients, satisfying α + β + γ + δ = 1.

[0060] In this embodiment, the default weight configuration is α=0.3, β=0.25, γ=0.3, δ=0.15, which can be adjusted according to actual needs. The urgency score E is calculated as E=min(T_w / T_t,1), where T_w is the cumulative waiting time of the task, and T_t is the task lifecycle threshold, ranging from 0 to 1. The task type score T is preset by the user or learned by the system based on historical data, with a typical value range of 0.1 to 1.0. The time score Time is calculated considering the task's deadline distance and whether it is within the allowed execution time window.

[0061] By adopting the above technical solution and using a multi-task priority dynamic sorting algorithm model, the dynamic calculation and real-time adjustment of task priorities are realized, enabling the garden robot to adjust the task execution order in a timely manner according to environmental changes and task characteristics, thereby improving the adaptability and flexibility of task scheduling.

[0062] In some preferred embodiments of this application, the state perception unit dynamically adjusts the weight coefficients of each task type based on the current environmental state. The environmental state is acquired through the fusion of multiple sensors, including a meteorological sensor to detect weather conditions, a light sensor to detect light intensity, and a sound sensor to detect environmental noise levels. When precipitation is detected, the state perception unit automatically increases the weight coefficient of the leaf sweeping task while decreasing the weight coefficient of the lawn mowing task, because fallen leaves are more easily washed away and accumulated during precipitation, requiring priority for cleaning. When high temperatures are detected, the unit automatically increases the execution frequency and weight of irrigation tasks while decreasing the priority of sweeping tasks to avoid sweeping operations that may generate dust during high-temperature periods. When a nighttime environment is detected, lighting maintenance and security patrol tasks are prioritized, while the priority of daytime-specific tasks is reduced.

[0063] In practical implementation, the state perception unit combines a rule-based conditional triggering mechanism with real-time monitoring of environmental perception data. The system pre-defines various mapping rules between environmental states and task weights. When a specific environmental state is detected, the corresponding weight adjustment is automatically triggered. For example, when the weather sensor detects precipitation, the system automatically increases the type weight T_leaves for the leaf sweeping task from 0.6 to 0.95, while decreasing the type weight T_mow for the lawn mowing task from 0.5 to 0.2. When the light sensor detects that the light intensity is below a preset threshold, the system automatically increases the type weight T_patrol for the security patrol task from 0.4 to 0.8. The weight adjustments employ a smooth transition method to avoid abrupt changes affecting scheduling stability.

[0064] By adopting the above technical solution, the real-time response of the state perception unit to environmental changes enables adaptive adjustment of task weights, allowing the garden robot to intelligently adjust task priorities according to actual environmental conditions, thereby improving environmental adaptability and operational efficiency.

[0065] In some preferred embodiments of this application, the intelligent task recommendation model is constructed based on machine learning algorithms, including a regional pollution level prediction sub-model and a job timing optimization sub-model. The regional pollution level prediction sub-model uses a combination of time series analysis and spatial interpolation algorithms to predict the degree of leaf accumulation and pollution level in different areas of the courtyard at different time points. The model inputs include historical cleaning data, weather data, seasonal data, tree distribution data, and area data, and the output is a pollution level score for each area. The job timing optimization sub-model uses reinforcement learning algorithms to learn the optimal job timing decision-making strategy through continuous interaction with the environment.

[0066] In the specific implementation process, the regional pollution level prediction sub-model is built based on a Long Short-Term Memory (LSTM) network. The network structure includes an input layer, multiple LSTM layers, a fully connected layer, and an output layer. Input features include cleaning records from the previous N days, meteorological data from the previous M days, seasonal identifiers, and regional identifiers. The model training uses the backpropagation algorithm, and the optimization objective is to minimize the prediction error. In this embodiment, N is 7, M is 14, the input feature dimension is 32, the number of LSTM layers is 2, and the number of hidden layer nodes is 128. The model output is a pollution level score for each region, ranging from 0 to 100, with higher scores indicating more severe pollution. The operation timing optimization sub-model uses a Deep Q-Network (DQN) algorithm. The state space includes the current time, current weather, current pollution level distribution, and robot state. The action space includes whether to perform or not perform a task. The reward function comprehensively considers cleaning effect, energy consumption cost, and time efficiency.

[0067] By adopting the above technical solution, the intelligent task recommendation model predicts potential cleaning areas and optimal operating times, automatically generates recommended tasks and adds them to the task list, thereby improving cleaning coverage and operating efficiency and reducing the need for manual intervention.

[0068] In some preferred embodiments of this application, the task execution interruption and recovery mechanism is an important component for ensuring stable system operation. The interruption recovery module periodically saves snapshots of the task execution state during task execution; the saving frequency can be adjusted according to the task type and execution duration. State snapshots are saved incrementally, only recording the differences from the previous save, reducing storage space usage and data transmission overhead. When an external interruption event is detected, the interruption recovery module immediately saves the current state and records the environmental map marker of the interruption location. Interruption events include, but are not limited to: low battery warning (triggered when battery power is below a preset threshold); obstacle obstruction (triggered when the robot cannot traverse a certain section within a preset time); user-manual pause (triggered when a user initiates a pause command via remote control or mobile app); and communication interruption (triggered when the communication connection with the central scheduler is lost for more than a preset time).

[0069] In the specific implementation process, the data structure of the state snapshot includes: a task identifier field, used to uniquely identify the currently executing task; a task type field, used to record task type information; an execution progress field, used to record the percentage of task completion, with a value range of 0 to 100; a robot position field, including X coordinate, Y coordinate, and heading angle; a completed area field, represented by a polygon coordinate list; an uncompleted area field, represented by a polygon coordinate list; an environment map marker field, used to record the coordinates of the interruption location on the environment map; and a timestamp field, used to record the snapshot saving time. The selection of environment map markers adopts a key point marking method, marking only locations that are meaningful for the continued execution of the task, such as area boundaries, obstacle edges, and path turning points.

[0070] By adopting the above technical solution and through the task execution interruption and recovery mechanism, stable operation is ensured in the event of interruptions such as insufficient power, obstruction by obstacles, and communication interruption. It realizes the function of continuing to execute the task from the breakpoint and avoids the waste of time and resources caused by restarting the entire task.

[0071] In some embodiments of this application, the multi-robot collaborative scheduling module is responsible for initiating a collaborative scheduling mode when multiple yard robots are detected working in the same area. The core functions of the collaborative scheduling mode include task allocation, path coordination, and resource allocation. Task allocation employs an auction-based algorithm, where each robot bids for a task based on its current location, task load, operational capacity, and energy consumption status. The bid comprehensively considers the task execution cost and benefits. After collecting all bids, the central scheduler calculates the optimal allocation scheme using either the Hungarian algorithm or a greedy algorithm and distributes the allocation result to each robot for execution. Path coordination employs a distributed collaborative path planning algorithm, where each robot independently plans its own path based on local perception and information sharing, while avoiding path conflicts with other robots.

[0072] In its implementation, the task allocation strategy based on the auction algorithm includes the following steps: First, the central scheduler publishes a task announcement, containing basic task information such as task type, task area, and expected completion time. Then, each robot bids for the task based on its own status. The bid calculation formula is: Bid_i = Cost_i × Performance_factor, where Bid_i is the bid of the i-th robot, Cost_i is the estimated cost of executing the task, and Performance_factor is the performance coefficient based on the robot's operational capabilities. Factors considered in the cost estimation include: travel distance cost, task execution time cost, energy consumption cost, and obstacle avoidance cost. After collecting all bids, the central scheduler uses a bipartite graph optimal matching algorithm to calculate the allocation scheme that minimizes the total cost. When multiple robots are detected operating in adjacent areas, the path coordination algorithm is activated. Each robot shares position and path information through a communication link and uses the velocity obstacle (VO) method or artificial potential field method to adjust local paths and avoid collisions.

[0073] By adopting the above technical solution and through the multi-robot collaborative scheduling module, task allocation, path coordination and dynamic resource allocation in multi-robot scenarios are realized, giving full play to the advantages of multi-robot collaborative operation and improving the efficiency and coverage of courtyard cleaning operations.

[0074] In some embodiments of this application, the system further includes a load balancing module for load-balanced scheduling among the robots. The load balancing module monitors the task queue length, current position, battery level, and work progress of each robot in real time. When it detects that some robots are overloaded while others are lightly loaded, it automatically migrates tasks. Triggering conditions for task migration include: the load difference exceeding a preset threshold, migration benefits exceeding migration costs, and migration not causing overload of other robots. The load balancing employs a strategy combining a threshold-based triggering mechanism and a utility-based optimization algorithm to maximize overall operational efficiency while ensuring system stability.

[0075] In implementation, the load balancing module adopts a control architecture combining centralized and distributed approaches. The centralized control unit resides in the central scheduler, responsible for collecting global information and calculating global optimization solutions; the distributed control unit resides in each robot, responsible for local decision-making and coordination. Load balancing is triggered when the difference in task queue length between any two robots exceeds a preset threshold, initiating a load balancing assessment. The assessment calculates migration benefits and costs; when the benefits outweigh the costs, task migration is executed. Task migration employs a gradual approach, migrating a limited number of tasks at a time to avoid impacting system stability.

[0076] By adopting the above technical solution and using the load balancing module, the task can be dynamically allocated and adjusted among multiple robots, avoiding the situation where some robots are overloaded while others are idle, thus improving the overall system's resource utilization and operational efficiency.

[0077] In some preferred embodiments of this application, the system further includes a task conflict detection and handling module for detecting and resolving conflicts between tasks. Task conflicts include three types: resource conflicts, time conflicts, and space conflicts. Resource conflicts occur when multiple tasks need to use the same resource, such as charging stations or garbage collection points. Time conflicts occur when the time windows of multiple tasks overlap, preventing simultaneous execution. Space conflicts occur when multiple tasks need to occupy the same physical space, such as simultaneously cleaning the same area. The task conflict detection module monitors the task queue in real time, detects potential conflict relationships, and takes appropriate handling strategies.

[0078] In practical implementation, resource conflict resolution strategies include resource reservation and resource competition. Resource reservation refers to reserving necessary resources in advance when allocating resources for tasks, ensuring resource availability during task execution. Resource competition refers to determining the resource allocation order based on task priority when multiple tasks compete for the same resource. Time conflict resolution strategies include time window adjustment and task splitting. Time window adjustment involves appropriately adjusting the task's time window to prevent overlap with other tasks without affecting task execution. Task splitting involves dividing a task into multiple sub-tasks, which are executed at different times. Spatial conflict resolution strategies include area division and sequential execution. Area division involves assigning conflicting areas to different robots for processing, while sequential execution involves arranging different robots to enter the conflicting area sequentially.

[0079] By adopting the above technical solution, the task conflict detection and handling module effectively solves various conflict problems that may occur during the parallel execution of multiple tasks, ensuring the smoothness and consistency of task execution.

[0080] In some embodiments of this application, the system further includes a task execution monitoring module for real-time monitoring of the execution status and effectiveness of each task. The task execution monitoring module evaluates the task execution effectiveness using sensor data and algorithm models to determine whether the task has achieved its expected goals. When a task's execution effectiveness is detected as substandard, a task re-execution or task adjustment mechanism is automatically triggered. Task execution effectiveness evaluation metrics include cleanliness, coverage, energy efficiency, and time efficiency. Cleanliness evaluation uses image recognition algorithms to analyze ground images before and after the operation to calculate the percentage improvement in cleanliness. Coverage evaluation calculates the percentage of coverage by comparing the work area with the actual work path.

[0081] In practice, the task execution monitoring module is deployed on the central scheduler and communicates in real time with the execution control modules of each robot. Monitoring data includes task execution progress, robot motion trajectory, sensor data streams, and execution effect evaluation results. When a task's execution effect is detected as substandard, the monitoring module generates a re-execution suggestion and decides whether to re-execute immediately or postpone it based on the current task queue status and system load. The priority of re-executed tasks is dynamically adjusted according to the degree of substandard performance; the higher the degree of substandard performance, the higher the re-execution priority. The monitoring module also generates task execution reports for users to view and analyze.

[0082] By adopting the above technical solution, the task execution monitoring module enables full monitoring of the task execution process and its effects, timely detection and handling of execution anomalies, and ensures the quality of task execution and stable system operation.

[0083] In some embodiments of this application, the system further includes a user interaction module for receiving user commands and providing system status feedback to the user. The user interaction module supports multiple interaction methods, including touchscreen interaction, voice interaction, remote control interaction, and mobile app interaction. Users can use the user interaction module to initiate task requests, adjust task parameters, pause or cancel tasks, and view task execution status. The user interaction module also provides an intelligent voice assistant function, supporting natural language understanding and speech synthesis, allowing users to control the robot to perform various operations via voice commands.

[0084] In its implementation, the user interaction module adopts a layered architecture, including a presentation layer, a business logic layer, and a data access layer. The presentation layer is responsible for displaying the user interface and capturing interactive events, supporting adaptation to various terminal devices. The business logic layer handles user requests, performing permission verification, parameter validation, and business logic processing. The data access layer is responsible for data interaction with backend services, including task management, data querying, and configuration updates. The user interaction module also supports personalized settings, allowing users to customize parameters such as task type, priority rules, and execution time, enabling the system to provide personalized services based on user preferences.

[0085] By adopting the above technical solution, a convenient user interaction method is provided through the user interaction module, enabling users to easily control and manage the operation tasks of the yard robot, thereby improving the user experience and the ease of use of the system.

[0086] This invention utilizes a multi-task priority dynamic sorting algorithm model to achieve dynamic calculation and real-time adjustment of task priorities, improving the adaptability and flexibility of task scheduling and enabling yard robots to adjust the task execution order in a timely manner according to environmental changes. Through an intelligent task recommendation model, it predicts potential cleaning areas and optimal operation times based on historical data and environmental changes, improving cleaning coverage and operational efficiency. A task execution interruption and recovery mechanism ensures stable operation under interruption conditions such as insufficient power or obstacles, avoiding time and resource waste caused by repetitive tasks. A multi-robot collaborative scheduling architecture realizes task allocation, path coordination, and resource optimization in multi-robot scenarios, fully leveraging the advantages of large-scale operations. Through the collaborative work of a load balancing module, a task conflict detection and handling module, a task execution monitoring module, and a user interaction module, a complete multi-task collaborative scheduling system for yard robots is constructed, possessing high practical value and market potential.

[0087] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0088] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

Claims

1. A multi-task collaborative scheduling method for a yard robot, characterized in that, Includes the following steps: Step 1: Obtain the perception data of the courtyard environment and the list of tasks to be performed; Step 2: Based on the perceived data and the task list, establish a multi-task priority dynamic sorting algorithm model; Step 3: Based on the multi-task priority dynamic sorting algorithm model, calculate the priority score of each task in the task list; Step 4: Sort the task list according to the priority score of each task to generate a task execution sequence; Step 5: Execute each task sequentially according to the task execution sequence.

2. The multi-task collaborative scheduling method for a yard robot according to claim 1, characterized in that, The multi-task priority dynamic sorting algorithm model in step 2 includes an urgency assessment unit, a type weight unit, a state awareness unit, and a time constraint unit. The urgency assessment unit is used to calculate an urgency score based on the urgency level of the task. The type weighting unit is used to assign weight coefficients according to the task type; The state-aware unit is used to adjust weights according to the current environmental state; The time constraint unit is used to calculate the time score based on the time constraints of the task. The priority score is calculated using the following formula: P = α × E + β × T + γ × S + δ × Time, where P is the priority score, E is the urgency score, T is the task type score, S is the environmental status score, Time is the time score, and α, β, γ, and δ are preset weight coefficients.

3. The multi-task collaborative scheduling method for a yard robot according to claim 2, characterized in that, The weight adjustment based on the current environmental state includes: Increase the weight of leaf sweeping tasks when precipitation is detected; When high temperatures are detected, increase the frequency of irrigation tasks; When nighttime conditions are detected, priority is given to performing lighting maintenance and security patrols.

4. The multi-task collaborative scheduling method for a yard robot according to claim 1, characterized in that, The method further includes an intelligent task recommendation step: collecting historical operation data and environmental change data of the courtyard; constructing an intelligent task recommendation model based on the historical operation data and environmental change data; using the intelligent task recommendation model to predict potential cleaning areas and optimal operation times; and automatically generating recommended tasks based on the prediction results and adding them to the task list.

5. The multi-task collaborative scheduling method and system for yard robots according to claim 4, characterized in that, The intelligent task recommendation model is built based on machine learning algorithms and includes a regional pollution level prediction sub-model and an operation timing optimization sub-model. The regional pollution prediction sub-model is used to predict the degree of leaf accumulation in each area of ​​the courtyard. The operation timing optimization sub-model is used to determine the optimal cleaning time for each area.

6. The multi-task collaborative scheduling method and system for yard robots according to claim 1, characterized in that, The method further includes a task execution interruption and recovery step: during task execution, a snapshot of the current task execution state is periodically saved; when an external interruption event is detected, the interruption location and environmental map marker are recorded; when the interruption is recovered, the task continues to be executed from the interruption location based on the state snapshot and environmental map marker.

7. The multi-task collaborative scheduling method and system for yard robots according to claim 6, characterized in that, The status snapshot includes: current task identifier, task execution progress percentage, robot current position coordinates, completed work areas, and uncompleted work areas.

8. The multi-task collaborative scheduling method and system for a yard robot according to claim 6, characterized in that, The external interruption events include: low battery warning, obstruction by obstacles, user manual pause, and communication interruption.

9. The multi-task collaborative scheduling method and system for a yard robot according to claim 1, characterized in that, The method also includes a multi-robot collaborative scheduling step: when multiple yard robots are detected in the same work area, a collaborative scheduling mode is activated; tasks are allocated according to the current position, task load and work capacity of each robot. Generate collaborative path planning to avoid path conflicts among multiple robots; monitor the operational status of each robot in real time and dynamically allocate resources; The task allocation adopts an auction-based task allocation strategy, including the following sub-steps: S1: Each robot quotes a price for the task based on its own load and capabilities; S2: After collecting all bids, the central scheduler calculates the optimal allocation scheme; S3: Distribute the allocation results to each robot for execution.

10. A multi-task collaborative scheduling system for a garden robot, characterized in that, It includes the perception layer, decision-making layer, execution layer, and coordination layer; The perception layer is responsible for acquiring various perception data of the courtyard environment, including spatial distance information acquired by lidar, visual image information acquired by camera, meteorological environment information acquired by temperature and humidity sensor, obstacle information acquired by ultrasonic sensor, etc. The perception data is transmitted to the decision layer after preprocessing. The multi-task priority dynamic ranking algorithm model in the decision-making layer calculates the priority score of each task based on these data and generates a task execution sequence. The execution layer controls each actuator of the garden robot to complete the corresponding task according to the task execution sequence; The collaboration layer is responsible for task allocation, path coordination, and resource scheduling in multi-robot scenarios.

Images