A multi-task dynamic scheduling method and device

Abstract

A multi-task dynamic scheduling method and device, the method comprising: sequentially performing dependency check, resource conflict check and hard constraint check on each detection task in a detection task set, generating a single-task scheduling plan for each detection task passing all checks; for each detection task not passing any of the checks, determining its corresponding associated detection task through resource competition relationship and task dependency chain, and generating a corresponding multi-task collaborative scheduling plan by combining a multi-objective optimization algorithm; and scheduling the detection tasks according to the single-task scheduling plan and / or the multi-task collaborative scheduling plan. The method realizes automatic and intelligent scheduling of detection tasks, significantly improves scheduling efficiency, effectively reduces detection task cost through a multi-objective optimization algorithm, and improves resource utilization by collaboratively scheduling detection tasks with resource competition and task dependency.

Description

Technical Field

[0001] This application relates to the field of task scheduling technology, specifically to a multi-task dynamic scheduling method and apparatus. Background Technology

[0002] Currently, the scheduling of testing tasks relies heavily on manual experience and uses static tools such as Excel and Gantt charts for scheduling. This traditional approach has the following limitations: Inefficient, manual scheduling is time-consuming and labor-intensive when faced with a massive number of tasks (task library) and complex resource constraints (such as personnel load, equipment load, site resource load, project dependencies, sample arrival date, etc.), and it is difficult to obtain the optimal solution.

[0003] Low resource utilization and the inability to accurately predict and balance resource load can easily lead to problems such as idle or overloaded key equipment and uneven workloads among personnel, resulting in resource waste.

[0004] The optimization goal is singular, usually focusing only on the delivery date. It is difficult to take into account multiple conflicting goals such as the shortest scheduling cycle, equipment utilization rate, testing cost and personnel load, resulting in high testing task costs.

[0005] The lack of collaboration makes it difficult for humans to coordinate globally in multi-task collaborative scheduling scenarios with project dependencies, which can easily lead to the "island" effect and affect overall efficiency. Summary of the Invention

[0006] This application provides a multi-task dynamic scheduling method and apparatus, which can solve the technical problems of low efficiency, high cost of detection tasks and low resource utilization in the current detection task scheduling technology.

[0007] To achieve the above objectives, in a first aspect, this application provides a multi-task dynamic scheduling method, the method comprising: For each detection task in the detection task set, dependency checks, resource conflict checks, and hard constraint checks are performed sequentially. For each detection task that passes all checks, a single task scheduling plan is generated.

[0008] For each detection task that fails any of the aforementioned checks, its corresponding associated detection task is determined through resource competition and task dependency chain, and a corresponding multi-task collaborative scheduling plan is generated by combining a multi-objective optimization algorithm.

[0009] The detection tasks are scheduled according to the single-task scheduling plan and / or the multi-task collaborative scheduling plan.

[0010] Furthermore, in one embodiment, the dependency check is to check whether all the prerequisite tasks of the current detection task have been completed; if so, the check passes.

[0011] The resource conflict check involves finding a shared available time window for all critical resources required by the current detection task within a desired time period. If such a shared available time window is found, the check passes. The desired time period is determined based on the task's delivery date, priority, or customer requirements.

[0012] The hard constraint check is to check whether the task schedule meets the hard constraints. If it does, the check passes. The hard constraints include the sample arrival date and weather conditions.

[0013] Furthermore, in one embodiment, the single-task scheduling plan includes: task ID and name, task start time and end time, resource list and plan status; the resource list includes equipment number, testing personnel and test site.

[0014] Furthermore, in one embodiment, the step of generating a corresponding multi-task cooperative scheduling plan by combining a multi-objective optimization algorithm includes: The detection tasks that fail any of the aforementioned checks and their associated detection tasks are grouped into a task group. A multi-objective optimization problem is constructed based on the task group, resource conflicts, and optimization objectives. The multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model to generate a corresponding multi-task collaborative scheduling plan.

[0015] The associated detection task is a detection task that has a resource competition relationship or task dependency relationship with the detection task that failed any of the checks.

[0016] The optimization objectives include scheduling cycle, total cost, resource utilization, and personnel load balance. The multi-task collaborative scheduling plan includes: a time schedule and resource allocation list for detection tasks, a global Gantt chart, a comparison of key indicators, and task change analysis.

[0017] Furthermore, in one embodiment, the step of solving the multi-objective optimization problem based on a genetic algorithm and a pre-established multi-objective optimization model to generate a corresponding multi-task cooperative scheduling plan includes: The multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model to obtain a non-dominated solution set.

[0018] Combining business, operational, and risk factors, an optimal solution is determined from the non-dominated solution set, and a multi-task collaborative scheduling plan is generated; the business factors include customer importance, contract terms, and costs; the operational factors include operational feasibility; and the risk factors include potential risks.

[0019] Furthermore, in one embodiment, the multi-objective optimization model is established based on detection task scheduling data, and includes decision variables, constraints, and objective functions.

[0020] The testing task scheduling data includes the testing equipment library, test site, load status, attributes of the testing tasks to be scheduled, and task priority; the decision variables include the start time of the testing tasks and resource allocation scheme; the constraints include resource capacity limitations and task dependencies; the objective function is a multi-objective weighted function, and the multi-objectives include scheduling cycle, resource utilization rate, and personnel load balance.

[0021] Furthermore, in one embodiment, if the associated detection task includes a detection task in a single task scheduling plan, and the detection task has a hard dependency relationship with the detection task that failed any of the checks, or meets the task group benefit conditions but is not within the exclusion range, then the detection task is included as an associated detection task in the generation of the task group.

[0022] The same detection task can only appear in one task group.

[0023] Furthermore, in one embodiment, the hard dependency relationship includes: the input of the subsequent task depends on the output result of the preceding task, there is data transfer or state dependency between tasks, and tasks share test samples or standard parts.

[0024] The task group benefit conditions include: the overall reduction in task group execution time is greater than or equal to a preset reduction threshold, the improvement in resource utilization is greater than or equal to a preset improvement threshold, and the reduction in total detection cost is greater than or equal to a preset reduction threshold.

[0025] The shortening threshold is set according to the ratio of the difference between the total duration of independent scheduling and the total duration of collaborative scheduling; the increasing threshold is set according to the ratio of the difference between the average resource utilization rate of single-task scheduling and the average resource utilization rate of multi-task collaborative scheduling; and the decreasing threshold is set according to the ratio of the difference between the total cost of single-task scheduling and the total cost of multi-task collaborative scheduling.

[0026] The exclusion scope includes critical path tasks, high-risk tasks, highest priority tasks, and / or tasks with special qualification requirements.

[0027] Furthermore, in one embodiment, during the scheduling of the detection task, the current status of each resource is collected in real time, and the task execution progress, resource load, and available time window are updated.

[0028] Secondly, this application provides a multi-task dynamic scheduling device, the device comprising: The first plan generation module is used to perform dependency checks, resource conflict checks, and hard constraint checks on each detection task in the detection task set in sequence, and generate a single task scheduling plan for each detection task that passes all checks.

[0029] The second plan generation module is used to determine the corresponding associated detection tasks for each detection task that fails any of the aforementioned checks through resource competition and task dependency chains, and generate a corresponding multi-task collaborative scheduling plan by combining a multi-objective optimization algorithm.

[0030] The scheduling module is used to schedule the detection tasks according to the single-task scheduling plan and / or the multi-task collaborative scheduling plan.

[0031] The beneficial effects of the technical solutions provided in this application include: By sequentially performing dependency checks, resource conflict checks, and hard constraint checks on each detection task in the detection task set, a single-task scheduling plan is generated for each detection task that passes all checks. For each detection task that fails any of the checks, its corresponding associated detection tasks are determined through resource competition relationships and task dependency chains. Combined with a multi-objective optimization algorithm, a corresponding multi-task collaborative scheduling plan is generated. Based on the single-task scheduling plan and / or the multi-task collaborative scheduling plan, the detection tasks are scheduled, thereby achieving automated and intelligent scheduling of detection tasks and significantly improving scheduling efficiency. By comprehensively considering multiple optimization objectives through the multi-objective optimization algorithm, the cost of detection tasks is effectively reduced. By collaboratively scheduling detection tasks with resource competition and task dependencies, resource utilization is improved. Attached Figure Description

[0032] Figure 1 This is a flowchart of a multi-task dynamic scheduling method according to an embodiment of this application.

[0033] Figure 2 This is a flowchart of a specific embodiment of the multi-task dynamic scheduling method of this application.

[0034] Figure 3 This is a schematic diagram of the scheduling platform business architecture in an embodiment of this application.

[0035] Figure 4 This is a block diagram of a multi-task dynamic scheduling device according to an embodiment of this application. Detailed Implementation

[0036] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0037] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0038] In a first aspect, embodiments of this application provide a multi-task dynamic scheduling method.

[0039] In one embodiment, see Figure 1 As shown, the above-mentioned multi-task dynamic scheduling method includes: S1. Perform dependency checks, resource conflict checks, and hard constraint checks on each detection task in the detection task set in sequence, and generate a single task scheduling plan for each detection task that passes all checks.

[0040] S2. For each detection task that fails any of the above checks, determine its corresponding associated detection task through resource competition and task dependency chain, and generate the corresponding multi-task collaborative scheduling plan by combining multi-objective optimization algorithm.

[0041] S3. Schedule the detection tasks according to the above single-task scheduling plan and / or the above multi-task collaborative scheduling plan.

[0042] In this embodiment, a preliminary scheduling of individual tasks is first performed, with the core being a rapid feasibility assessment. This assessment includes the following specific steps: resource availability matching, time window lookup, and hard constraint verification. If the assessment passes, a single-task plan is generated; if resource conflicts, time limit exceeding, or other issues are found, an early warning is immediately triggered. The purpose of the early warning is to expose risks in advance and provide a clear decision-making basis for whether to initiate multi-task collaborative scheduling.

[0043] For any of the above checks that fail (e.g., specific equipment, personnel, or venues are occupied, or task dependencies cannot be met), all associated check tasks are identified by analyzing resource competition relationships (e.g., competing for the same key resource) and task dependency chains (e.g., belonging to the same project process), and these tasks are grouped together to form a task group to be optimized.

[0044] The multi-objective optimization problem is solved using a multi-objective optimization algorithm to obtain scheduling adjustment suggestions, thereby generating a corresponding multi-task collaborative scheduling plan. For example, resource replacement suggestion: replace the equipment X used by conflicting tasks A and B with equipment Y with the same function so that task C can use equipment X as originally planned; timing adjustment suggestion: insert the new task C after task A and before task B for execution, and postpone task B by 2 days accordingly; task splitting suggestion: split task A into a preparation stage and a core testing stage, and execute the preparation stage in parallel in advance to shorten its critical resource occupation time.

[0045] This hierarchical, categorized, and collaboratively optimized scheduling strategy not only significantly shortens task scheduling time and reduces the cost of manual intervention, but also effectively balances the load of various resources, avoids the idleness or overload of key resources, and significantly improves the overall execution efficiency and resource utilization efficiency of detection tasks. At the same time, by taking into account multiple optimization objectives, it effectively controls detection costs and enhances the flexibility, adaptability, and intelligence of detection task scheduling, providing an efficient, economical, and reliable solution for task scheduling in complex detection environments.

[0046] Furthermore, in one embodiment, in step S1 above, the dependency check is to check whether all the prerequisite tasks of the current detection task have been completed; if so, the check passes.

[0047] Resource conflict checking identifies a shared available time window for all critical resources required for the current detection task within the expected time period. If a shared available time window is found, the check passes. The expected time period is determined based on the task's delivery date, priority, or customer requirements.

[0048] The hard constraint check verifies whether the task schedule meets the hard constraints; if so, the check passes. These hard constraints include the sample arrival date and weather conditions.

[0049] In this embodiment, the basic data required for dynamic scheduling of multiple tasks is first constructed, including a resource library: a library of fixed testing equipment, a library of mobile testing equipment, test sites, testing personnel, drivers, energy, test consumables, samples, and other resources; dynamic status data: real-time collected data on the current status, load, and available time windows of each resource; and a task and constraint library: attributes (such as type, duration, order date, and delivery date), task priority, project dependencies, and external influencing factors (such as weather constraints, sample arrival date, and logistics progress tracking) of all testing tasks to be scheduled. Based on this, by sequentially performing dependency checks, resource conflict checks, and hard constraint checks on the testing tasks, a rapid and accurate assessment of task feasibility is achieved, effectively distinguishing between simple and complex tasks, thus creating conditions for the implementation of a hierarchical scheduling strategy.

[0050] Furthermore, in one embodiment, the single-task scheduling plan in step S1 above includes: task ID and name, task start time and end time, resource list, and plan status. The resource list includes equipment number, testing personnel, and test site.

[0051] Furthermore, in one embodiment, in step S2 above, a corresponding multi-task cooperative scheduling plan is generated by combining a multi-objective optimization algorithm. The specific steps are as follows: The detection tasks that fail any of the above checks and their associated detection tasks are grouped into task groups. A multi-objective optimization problem is constructed based on the task groups, resource conflicts and optimization objectives. The multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model to generate a corresponding multi-task collaborative scheduling plan.

[0052] Among them, the associated detection task is a detection task that has a resource competition relationship or task dependency relationship with the detection task that failed any of the above checks.

[0053] The optimization objectives include scheduling cycle, total cost, resource utilization, and personnel load balance; the multi-task collaborative scheduling plan includes: a time schedule and resource allocation list for detection tasks, a global Gantt chart, a comparison of key indicators, and task change analysis.

[0054] In this embodiment, by grouping detection tasks that fail any check and their associated detection tasks with resource competition or task dependencies into task groups, accurate identification and task aggregation of complex scheduling scenarios are achieved, laying the foundation for multi-task collaborative optimization. A multi-objective optimization problem is constructed based on task groups, resource conflicts, and optimization objectives. Multiple conflicting optimization objectives, such as shortest scheduling cycle, lowest total cost, highest resource utilization, and most balanced personnel load, are unified into the optimization framework, overcoming the limitation of single optimization objectives in traditional scheduling methods. A genetic algorithm is used to solve the pre-established multi-objective optimization model, fully utilizing the global search capability and parallel processing advantages of the genetic algorithm in complex solution spaces to efficiently find a high-quality non-dominated solution set that satisfies all constraints, achieving an effective solution to the multi-objective optimization problem. Finally, a multi-task collaborative scheduling plan is generated, including a list of detection task time arrangements and resource allocations, a global Gantt chart, key indicator comparisons, and task change analysis.

[0055] It not only provides clear and specific operational guidelines for scheduling execution, but also enhances the interpretability and decision support capabilities of the scheme through visualization and quantitative comparison, thereby realizing global optimized scheduling in complex multi-task scenarios, significantly improving resource utilization and scheduling efficiency, effectively balancing multiple conflicting objectives, reducing the cost of detection tasks, and enhancing the level of intelligent scheduling decision-making and the executability of scheduling schemes.

[0056] Furthermore, in one embodiment, the above-mentioned multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model to generate a corresponding multi-task cooperative scheduling plan. The specific steps are as follows: The multi-objective optimization problem is solved by using a genetic algorithm and a pre-established multi-objective optimization model to obtain a non-dominated solution set.

[0057] By combining business, operational, and risk factors, the optimal solution is determined from the aforementioned non-dominated solution set, and a multi-task collaborative scheduling plan is generated. Business factors include customer importance, contract terms, and costs; operational factors include operational feasibility; and risk factors include potential risks.

[0058] In this embodiment, a multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model, yielding a non-dominated solution set. This enables efficient solving of multi-objective optimization problems under complex constraints, obtaining multiple non-dominated solutions and providing a rich set of alternatives for subsequent decision-making. Based on this, and considering business factors such as customer importance, contract terms, and costs, operational factors such as operational feasibility, and risk factors such as potential risks, the optimal solution is determined from the non-dominated solution set, and a multi-task collaborative scheduling plan is generated. This achieves an organic integration of technical feasibility and commercial value, ensuring that scheduling decisions not only meet technical optimization requirements but also align with comprehensive considerations of actual business scenarios. This significantly improves the practicality, acceptability, and execution success rate of the scheduling plan, effectively balancing multiple objectives of efficiency, cost, and risk, and enhancing the system's intelligence and scientific decision-making.

[0059] Furthermore, in one embodiment, the above-mentioned multi-objective optimization model is established based on detection task scheduling data, and includes decision variables, constraints, and objective functions.

[0060] The testing task scheduling data includes the testing equipment library, test site, load status, attributes of the testing tasks to be scheduled, and task priority; decision variables include the start time of the testing tasks and resource allocation scheme; constraints include resource capacity limitations and task dependencies; the objective function is a multi-objective weighted function, and the multiple objectives include scheduling cycle, resource utilization rate, and personnel load balance.

[0061] Furthermore, in one embodiment, if the aforementioned associated detection task includes a detection task in a single task scheduling plan, and the detection task has a hard dependency relationship with a detection task that has failed any of the aforementioned checks, or meets the task group benefit conditions but is not within the exclusion scope, then the detection task is included as an associated detection task in the generation of the task group.

[0062] Secondly, during the task group generation process, the same detection task appears in only one task group.

[0063] Hard dependencies include: the input of a subsequent task depends on the output of a preceding task, there is data transfer or state dependency between tasks, and tasks share test samples or standard parts.

[0064] The task group benefit conditions include: the overall reduction in task group execution time is greater than or equal to a preset reduction threshold, the improvement in resource utilization is greater than or equal to a preset improvement threshold, and the reduction in total detection cost is greater than or equal to a preset reduction threshold. Specifically, the reduction threshold is set based on the ratio of the difference between the total independent scheduling time and the total collaborative scheduling time; in this embodiment, it can be set to 15%. The improvement threshold is set based on the ratio of the difference between the average resource utilization rate of single-task scheduling and the average resource utilization rate of multi-task collaborative scheduling; in this embodiment, it can be set to 25%. The reduction threshold is set based on the ratio of the difference between the total cost of single-task scheduling and the total cost of multi-task collaborative scheduling; in this embodiment, it can be set to 10%.

[0065] The exclusion scope includes critical path tasks (tasks that affect project milestones), high-risk tasks (tasks with a failure rate >5%), highest priority tasks, and tasks with special qualification requirements.

[0066] In this embodiment, by setting hard dependency judgment rules, detection tasks with strong coupling relationships such as subsequent task input depending on the output result of previous task, data transfer or state dependency between tasks, and sharing of test samples or standard parts between tasks are identified as related detection tasks. This ensures that tasks with essential business relationships can be accurately included in the same task group for collaborative optimization, avoiding resource conflicts and process interruptions caused by task fragmentation.

[0067] By introducing task group benefit conditions, quantitative indicators such as the reduction in overall task group execution time, the improvement in resource utilization, and the reduction in total detection cost are used as economic thresholds for task group generation. Furthermore, reduction thresholds, improvement thresholds, and reduction thresholds are set to ensure that the multi-task collaboration mechanism is triggered only when collaborative scheduling can bring about significant benefit improvements. This effectively avoids the waste of computing resources and over-optimization caused by collaboration for the sake of collaboration.

[0068] By clearly defining the scope of exclusion, critical path tasks, high-risk tasks, highest priority tasks, and tasks with special qualification requirements were excluded from task group reorganization. This ensured the timely achievement of project milestones, reduced systemic risks introduced by collaborative scheduling, protected the rights and interests of high-priority clients, and ensured the compliant execution of tasks with special qualifications.

[0069] Meanwhile, by ensuring that "the same detection task appears in only one task group," the duplicate allocation of tasks and resource conflicts are eliminated, and the clarity and executability of the scheduling plan are guaranteed. This achieves the scientific, economical and secure generation of task groups, effectively controls scheduling risks and execution complexity while improving the efficiency of collaborative scheduling, and significantly enhances the practicality and reliability of the multi-task collaborative scheduling mechanism.

[0070] Furthermore, in one embodiment, during the scheduling of detection tasks, the current status of each resource is collected in real time, and the task execution progress, resource load, and available time window are updated.

[0071] In this embodiment, a dynamic data feedback mechanism is constructed by collecting the current status of each resource in real time during the scheduling process and synchronously updating the task execution progress, resource load, and available time window, thus realizing real-time synchronization between the scheduling plan and the actual execution status. By continuously tracking the dynamics of multi-dimensional resources, real-time and reliable data support is provided for the dynamic adjustment of subsequent tasks and plan optimization. By timely identifying task execution deviations and dynamically refreshing resource availability periods, resource over-allocation and conflict scheduling are effectively avoided. This significantly enhances the real-time response capability and environmental adaptability of the scheduling system, lays a data foundation for rapid rescheduling in case of emergencies, and improves the overall reliability of scheduling execution and the level of refinement of resource utilization.

[0072] See Figure 2 As shown, a specific embodiment of the above-mentioned multi-task dynamic scheduling method is given, and the steps are as follows: A1. Create a detection task and send it to the local scheduling platform. The platform schedules the resource library, task and constraint library to parse and match the task.

[0073] The specific steps for task parsing and matching are as follows: Resource demand matching: Match task attributes with the equipment capability set, personnel skill matrix, and consumable list in the resource library to initially filter out all available resources.

[0074] Constraint verification: Verify that the task meets all constraints defined in the constraint library, including: verifying whether the task meets its project dependencies (e.g., the preceding task must have been completed); checking whether the sample arrival date has been met; and assessing whether external constraints such as weather conditions allow the task to be executed.

[0075] Dynamic status awareness: Real-time query dynamic status data to obtain the real-time status, current load and future availability time window of all filtered available resources, and remove resources that are already occupied.

[0076] Figure 3This diagram illustrates the business architecture of the scheduling platform. Driven by target setting, the architecture optimizes factors such as scheduling cycle, equipment load rate, testing costs, personnel load rate, and net profit. It comprehensively considers multiple influencing factors, including project cycle (order date, delivery date), task priority, testing project dependencies, personnel load, equipment load, site resource load, testing consumables inventory, sample arrival date, logistics progress tracking, energy inventory, weather, and unforeseen circumstances. The scheduling platform, as the core processing unit, integrates four scheduling mechanisms: dynamic resource monitoring, single-task scheduling, multi-task collaborative scheduling, and manual scheduling. These mechanisms operate collaboratively in different scheduling scenarios, ultimately generating a static scheduling plan.

[0077] The platform's underlying layer relies on a basic data layer, including testing procedures (such as adjusting the dependencies between various testing items; prioritizing mandatory testing: whole vehicle testing > powertrain and other module testing > component and system testing; prioritizing development testing: component and system testing > powertrain and other module testing > whole vehicle testing) and basic information on testing items (covering personnel qualification matching, supporting venues, supporting fixed testing equipment (such as test benches, environmental chambers, etc.), supporting mobile testing equipment (such as sensors, dummies, etc.), supporting testing consumables, standard prices, standard costs, standard man-hour calculations, and other constraints (such as weather, temperature, test time, etc.), providing real-time support for the scheduling process in terms of procedure logic and operational status. The overall architecture is guaranteed by a resource library, comprehensively covering a testing personnel library, a fixed / mobile testing equipment library, a weather library, a testing consumables library, a testing venue library, a driver library, an energy library, a sample library, and a logistics library, forming full-element data support for the scheduling process, thereby realizing multi-task collaboration and dynamic optimization scheduling for intelligent testing scenarios.

[0078] A2. Determine whether the current detection task meets the single-task scheduling logic. If yes, proceed to step A3; otherwise, proceed to step A4.

[0079] The single-task scheduling logic includes: Dependency check: Verify if there are any incomplete mandatory prerequisite tasks for this task. If so, it is considered unsatisfactory. Resource conflict check: For all critical resources required by this task, find common available time windows within the expected time period. If none are found, it is considered unsatisfactory. Hard constraint check: Verify whether the task schedule meets all hard constraints such as sample arrival date and weather conditions. If any constraint is violated, it is considered unsatisfactory. Only when all three checks pass is the single-task scheduling logic considered satisfied.

[0080] A3. Generate a single-task scheduling plan. The specific steps are as follows: Window optimization: Select the optimal time window from multiple feasible time windows based on rules such as earliest start or resource balancing.

[0081] Resource specification: For each type of resource required by the task, specify the optimal specific object from the available instances (a single resource object that is specifically available in the resource library) (such as device A-01, engineer Zhang San).

[0082] Plan generation and locking: Combine the above information into a formal plan and store it in the database, and update the status of the relevant resources to locked within the time window.

[0083] View Update: Update this plan in the scheduling Gantt chart to create a visual representation.

[0084] A4. Generate multi-task collaborative scheduling suggestions, including suggestions for adjusting the task execution order, resource allocation, and even task scope and constraints.

[0085] The specific steps for generating multi-task collaborative scheduling suggestions are as follows: Root cause identification: Analyze the ultimate cause of single scheduling failure (such as missing critical resources, delay of preceding tasks, or deadlock of constraints).

[0086] Strategy matching: Based on the root cause, match the preset solution strategy (such as requesting new resources, negotiating an extension, or splitting tasks).

[0087] Content generation: Transform strategies into specific, actionable text recommendations (e.g., recommendation: request the activation of backup device B for task T100).

[0088] Impact Notes: To supplement the recommendations, please include a simple expected impact analysis (such as changes in cost and schedule).

[0089] A5. Manually assess multi-task collaborative scheduling suggestions to determine whether adjustments to task execution order or resource allocation are necessary. This assessment is not based on technical feasibility, but rather on a comprehensive consideration of business, operational, and risk factors. The specific steps are as follows: Comprehensive assessment: Operators review system recommendations and make a comprehensive judgment based on factors that the system has not fully quantified, such as customer importance, contract terms, actual costs, operational feasibility, and potential risks.

[0090] For example, if it is suggested to use backup equipment to save 2 days of construction time, but the cost increases by 5%, the operator will judge whether the customer is important enough to justify this cost and whether the project budget allows it.

[0091] Making decisions: The operator makes the final choice, adopting a suggestion, modifying it and then adopting it, or rejecting all suggestions and giving a new direction for processing.

[0092] Record the reasons: Record the results and main reasons for human decision-making to form a closed loop.

[0093] A6. Adjust the schedule: Adjust the task execution timetable based on the results of manual judgment.

[0094] A7. Generate a multi-task collaborative scheduling plan, including a new schedule and resource allocation list for all related detection tasks, a global Gantt chart, a comparison of key indicators of the optimization scheme (such as schedule and cost), and a change analysis of the affected tasks.

[0095] The specific steps for generating a multi-task collaborative scheduling plan are as follows: Problem modeling: The association detection task group, resource conflicts, and optimization objectives are formalized into a multi-objective optimization problem.

[0096] Algorithm solution: Heuristic algorithms such as genetic algorithms are used to search and calculate to find the globally optimal or satisfactory solution set that satisfies all constraints.

[0097] Solution generation: The optimal solution output by the algorithm is converted into a clear and visual scheduling plan, and multiple alternative solutions can be generated.

[0098] Impact assessment: Quantify the degree of improvement of the new plan compared to the original plan (such as the amount of time reduction) and identify the details of the plan changes for each task.

[0099] A8. Scheduling Plan Execution. The detection tasks are scheduled according to the single-task scheduling plan and / or the multi-task collaborative scheduling plan.

[0100] A9. Dynamic resource monitoring and emergency response.

[0101] During task execution, resource usage is dynamically monitored to ensure reasonable allocation and use of resources. The specific steps are as follows: Real-time status data collection: The current status of each resource (such as equipment operation / idle / fault, personnel working / available for dispatch, consumable inventory quantity, and current sample location) is collected in real time through equipment IoT sensors, personnel attendance system, and logistics tracking system.

[0102] Progress tracking and load updates: Real-time updates of task execution progress, and accordingly refreshes the load status and expected availability window of relevant resources.

[0103] Anomaly reporting: Automatically captures abnormal events such as equipment failure and task delays, and uses them as the basis for triggering dynamic scheduling.

[0104] Any unforeseen circumstances that arise during execution (such as equipment failure, personnel absence, emergency orders, or severe task timeouts) will be handled promptly. The handling method is not simply to stop the task, but rather to respond in a tiered manner based on the severity of the event, including: Local fine-tuning: For minor deviations (such as short-term task timeouts), the system automatically postpones subsequent affected tasks without triggering global rescheduling.

[0105] Triggering dynamic scheduling: For resource-level events (such as equipment failure), the system immediately identifies the scope of affected tasks and triggers a dynamic scheduling process (combining recovery time to determine whether the scheduling plan needs to be adjusted). Based on the principle of "executed tasks remain unchanged", a new scheduling plan is generated for the remaining tasks.

[0106] Emergency Pause and Alarm: For serious project-level events (such as sample damage), the system will pause related tasks and wait for senior human intervention to resolve the root cause.

[0107] A10. Determine whether the change in the execution status of the current task (such as delay, interruption, or advancement) affects other tasks. If yes, proceed to step A11; otherwise, proceed to step A12.

[0108] The determination is based on the dependencies between tasks and the shared resource plan. An impact is determined to occur if any of the following conditions are met: Direct dependency disruption: The delay or interruption of the current task causes its direct successor task to be unable to start as originally planned.

[0109] Indirect path impact: Changes in the current task's status extend the project's critical path, threatening the final delivery date.

[0110] Resource plan change: The current task has changed the end time of the critical resource's occupation, affecting all subsequent tasks that have reserved that resource.

[0111] Constraint triggering: The state of the current task changes, causing subsequent tasks to risk violating its hard constraints (such as deadlines).

[0112] A11. Adjust task execution. For situations affecting the execution of other tasks, adjust task execution to ensure all tasks can proceed smoothly. The core approach to adjustment is not simply stopping tasks, but triggering a dynamic rescheduling process for the "affected task set." The specific steps are as follows: Scope Lock: Identify all tasks that need adjustment (including the remainder of the current task and its subsequent tasks).

[0113] Trigger rescheduling: Take all tasks requiring adjustment as input and proceed to step A4. The scheduling engine will use the already executed portion as a rigid constraint to recalculate the optimal solution for the remaining work.

[0114] Plan Update: After the newly generated plan is manually confirmed, the global plan will be automatically updated, resources will be released and locked, and relevant personnel will be notified.

[0115] Continue execution: All tasks will continue to be executed according to the new, feasible plan.

[0116] A12. Complete this scheduling.

[0117] By implementing the above steps, we can ensure the efficient and orderly execution of the testing task, and respond quickly and handle any possible situations, thus guaranteeing the successful completion of the task.

[0118] Secondly, embodiments of this application also provide a multi-task dynamic scheduling device.

[0119] In one embodiment, reference is made to Figure 4 As shown, the aforementioned multi-task dynamic scheduling device includes a first plan generation module, a second plan generation module, and a scheduling module, specifically: The first plan generation module is used to perform dependency checks, resource conflict checks, and hard constraint checks on each detection task in the detection task set in sequence, and generate a single task scheduling plan for each detection task that passes all checks.

[0120] The second plan generation module is used to determine the corresponding associated detection tasks for each detection task that fails any of the above checks through resource competition and task dependency chains, and generate the corresponding multi-task collaborative scheduling plan by combining a multi-objective optimization algorithm.

[0121] The scheduling module is used to schedule the detection tasks according to the above-mentioned single-task scheduling plan and / or the above-mentioned multi-task collaborative scheduling plan.

[0122] In this application, by constructing a hierarchical, classified, and collaboratively optimized multi-task dynamic scheduling system, the intelligent, refined, and dynamic scheduling of detection tasks is realized, which significantly improves the operational efficiency and management level in complex detection environments.

[0123] Automated intelligent scheduling significantly reduces manual scheduling time. By optimizing algorithms to conduct integrated collaborative planning of tasks, personnel, and equipment, it significantly improves the utilization rate of key resources and reduces idle time and waiting.

[0124] By employing critical path simulation technology based on resource constraints and task dependencies, accurate and dynamic predictions of the project's total lifecycle can be achieved. This predictive capability, combined with robust dynamic adjustment capabilities, effectively addresses various unforeseen risks, thereby significantly improving the reliability of on-time project delivery.

[0125] By optimizing resource allocation (such as intelligently selecting the most cost-effective combination of equipment and personnel for tasks) and execution paths (such as changing sequential tasks to parallel ones), the wasted manpower and equipment costs of testing can be directly reduced. At the same time, by avoiding project delays, the risk of default and economic losses can be indirectly reduced.

[0126] Meanwhile, by taking into account multiple optimization objectives and building dynamic response capabilities, the testing cost is effectively controlled, and the flexibility, adaptability and intelligence of testing task scheduling are enhanced. This provides an efficient, economical and reliable full life cycle scheduling solution for complex testing scenarios such as third-party testing laboratories, R&D centers and production quality control departments.

[0127] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0128] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0129] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0130] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0131] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0132] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A multi-task dynamic scheduling method, characterized in that, The method includes: For each detection task in the detection task set, dependency checks, resource conflict checks, and hard constraint checks are performed in sequence. For each detection task that passes all checks, a single task scheduling plan is generated. For each detection task that fails any of the aforementioned checks, its corresponding associated detection task is determined through resource competition and task dependency chain, and a corresponding multi-task collaborative scheduling plan is generated by combining a multi-objective optimization algorithm. The detection tasks are scheduled according to the single-task scheduling plan and / or the multi-task collaborative scheduling plan.

2. The multi-task dynamic scheduling method as described in claim 1, characterized in that, The dependency check involves checking whether all prerequisite tasks for the current detection task have been completed; if so, the check passes. The resource conflict check involves finding a shared available time window for all critical resources required by the current detection task within a desired time period. If such a shared available time window is found, the check passes. The desired time period is determined based on the task's delivery date, priority, or customer requirements. The hard constraint check is to check whether the task schedule meets the hard constraints. If it does, the check passes. The hard constraints include the sample arrival date and weather conditions.

3. The multi-task dynamic scheduling method as described in claim 1, characterized in that, The single-task scheduling plan includes: task ID and name, task start time and end time, resource list and plan status; the resource list includes equipment number, testing personnel and test site.

4. The multi-task dynamic scheduling method as described in claim 1, characterized in that, The step of combining multi-objective optimization algorithms to generate corresponding multi-task collaborative scheduling plans includes: The detection tasks that fail any of the aforementioned checks and their associated detection tasks are grouped into a task group. A multi-objective optimization problem is constructed based on the task group, resource conflicts, and optimization objectives. The multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model to generate a corresponding multi-task collaborative scheduling plan. The associated detection task is a detection task that has a resource competition relationship or task dependency relationship with the detection task that failed any of the checks. The optimization objectives include scheduling cycle, total cost, resource utilization rate, and personnel load balance; the multi-task collaborative scheduling plan includes: a time schedule and resource allocation list for detection tasks, a global Gantt chart, a comparison of key indicators, and task change analysis.

5. The multi-task dynamic scheduling method as described in claim 4, characterized in that, The step of solving the multi-objective optimization problem based on a genetic algorithm and a pre-established multi-objective optimization model to generate a corresponding multi-task cooperative scheduling plan includes: The multi-objective optimization problem is solved using a genetic algorithm and a pre-established multi-objective optimization model to obtain a non-dominated solution set; Combining business, operational, and risk factors, an optimal solution is determined from the non-dominated solution set, and a multi-task collaborative scheduling plan is generated; the business factors include customer importance, contract terms, and costs; the operational factors include operational feasibility; and the risk factors include potential risks.

6. The multi-task dynamic scheduling method as described in claim 4, characterized in that, The multi-objective optimization model is established based on detection task scheduling data, and includes decision variables, constraints, and objective functions; The testing task scheduling data includes the testing equipment library, test site, load status, attributes of the testing tasks to be scheduled, and task priority; the decision variables include the start time of the testing tasks and the resource allocation scheme; the constraints include resource capacity limitations and task dependencies; the objective function is a multi-objective weighted function, and the multi-objectives include scheduling cycle, resource utilization rate, and personnel load balance.

7. The multi-task dynamic scheduling method as described in claim 4, characterized in that, If the associated detection task includes a detection task in a single task scheduling plan, and the detection task has a hard dependency relationship with the detection task that failed any of the checks, or meets the task group benefit conditions but is not within the exclusion range, then the detection task will be used as an associated detection task to participate in the generation of the task group. The same detection task can only appear in one task group.

8. The multi-task dynamic scheduling method as described in claim 7, characterized in that, The hard dependencies include: the input of a subsequent task depends on the output of a preceding task, there is data transfer or state dependency between tasks, and tasks share test samples or standard parts. The task group benefit conditions include: the overall reduction in task group execution time is greater than or equal to a preset reduction threshold, the improvement in resource utilization is greater than or equal to a preset improvement threshold, and the reduction in total detection cost is greater than or equal to a preset reduction threshold. The shortening threshold is set according to the ratio of the difference between the total duration of independent scheduling and the total duration of collaborative scheduling; the increasing threshold is set according to the ratio of the difference between the average resource utilization rate of single-task scheduling and the average resource utilization rate of multi-task collaborative scheduling; and the decreasing threshold is set according to the ratio of the difference between the total cost of single-task scheduling and the total cost of multi-task collaborative scheduling. The exclusion scope includes critical path tasks, high-risk tasks, highest priority tasks, and / or tasks with special qualification requirements.

9. The multi-task dynamic scheduling method as described in claim 1, characterized in that, During the scheduling of the detection tasks, the current status of each resource is collected in real time, and the task execution progress, resource load, and available time window are updated.

10. A multi-task dynamic scheduling device, characterized in that, The device includes: The first plan generation module is used to perform dependency checks, resource conflict checks, and hard constraint checks on each detection task in the detection task set in sequence, and generate a single task scheduling plan for each detection task that passes all checks. The second plan generation module is used to determine the corresponding associated detection tasks for each detection task that fails any of the aforementioned checks through resource competition and task dependency chains, and generate the corresponding multi-task collaborative scheduling plan by combining a multi-objective optimization algorithm. The scheduling module is used to schedule the detection tasks according to the single-task scheduling plan and / or the multi-task collaborative scheduling plan.

Images