Test task scheduling method and device, equipment and medium

Abstract

The application discloses a test task scheduling method, device, equipment and medium, and belongs to the field of charging and discharging test. The application determines predicted data of the change of charging and discharging power with time in the execution process based on the task information of the test task to be run, controls the starting time of the test task based on the predicted data, and adjusts the remaining time length of the adjustable idle period of the test task in the adjustable idle period according to the deviation between the actual total net power and the expected total net power of each test task being run. The scheme can master the power consumption and power feeding law of each task in advance, reasonably arrange the starting time of each test task to be run, reduce the local power grid power peak value when multiple tasks are concurrently run, and reduce the energy waste output to the external grid. The scheme adjusts the remaining time length of the idle period to correct the deviation, does not affect the effectiveness of the test result, improves the internal energy self-balancing rate and energy utilization efficiency of the local power grid, realizes intelligent peak regulation and energy-saving scheduling, and has the advantages of high practicability and high market value.

Description

Technical Field

[0001] This application belongs to the field of charge and discharge testing, and specifically relates to a method for scheduling test tasks, a device for scheduling test tasks, an electronic device, and a readable storage medium. Background Technology

[0002] With the rapid development of industries such as new energy batteries and energy storage devices, charge-discharge testing, as a core component for verifying the electrochemical performance, cycle life, and safety characteristics of devices, has seen a continuous expansion in testing scale. Dozens or even hundreds of charge-discharge testing devices are often deployed in a single testing area, and multi-task concurrent operation has become the norm. Consequently, issues such as high energy consumption in testing areas, large fluctuations in local power grid load, and significant pressure on power distribution facility capacity have gradually become important factors restricting the improvement of testing efficiency and the control of operating costs. The industry is continuously exploring various energy optimization and load management solutions.

[0003] Currently, energy management in charge-discharge testing systems is mainly developing along two directions: hardware topology upgrades and operational management optimization.

[0004] At the hardware topology level, early charge / discharge testing equipment mostly adopted dissipative topologies. During the discharge phase of batteries or energy storage devices, the electrical energy released was dissipated as heat through discharge resistors, resulting in extremely low energy utilization efficiency. This also required high-power cooling systems, further increasing equipment and operating costs. Building on this, testing equipment with energy feedback capabilities gradually became the industry mainstream. The electrical energy released during the discharge phase could be injected back into the local or external power grid via inverter circuits, reducing energy loss to some extent. However, in practical applications, each testing device typically operates independently, with its charging and discharging sequence entirely determined by its own preset test programs. The charging and discharging behavior of multiple devices lacks system-level coordination, and energy interaction relies on random superposition. When multiple devices simultaneously enter the charging phase, the instantaneous power demand of the local power grid surges, easily forming high power peaks. This places higher demands on the capacity configuration of power distribution facilities such as transformers and distribution cabinets, and may even trigger overload protection. When multiple devices simultaneously enter the discharging phase, a large amount of feedback energy is injected into the grid. If this feedback energy cannot be absorbed in time within the local area, it will result in actual waste, making it difficult to fully realize the energy-saving potential of feedback-type equipment.

[0005] To mitigate power fluctuations and improve energy efficiency, some sites employ hardware solutions involving the installation of independent energy storage devices. These devices absorb excess feedback energy through mediums such as supercapacitors and auxiliary energy storage batteries, releasing supplementary power during peak charging periods. While this approach can buffer power surges to some extent, it requires additional investment in hardware procurement and installation, occupies significant space, and the energy storage devices themselves have limited cycle life and energy conversion losses, resulting in high long-term maintenance costs. Therefore, it does not address the root cause of energy imbalance issues in the operation and management of testing tasks.

[0006] At the operational management level, some test management systems use manually preset fixed time intervals and polling start rules to stagger the start times of test tasks, avoiding the instantaneous impact caused by a large number of tasks starting simultaneously. However, these static rules rely on manual experience and cannot adapt to the differences in charging and discharging durations and power levels of different test programs, making it difficult to guarantee energy balance under any combination of tasks. Moreover, once the solution is determined, it lacks dynamic adjustment capabilities and is not flexible enough to deal with temporarily added test tasks. Directly starting new tasks can easily disrupt the existing operating power balance and cause new power fluctuations. In addition, general industrial energy management systems are also used in test workshops, mainly to realize functions such as total power consumption monitoring, over-limit alarms, and load protection. However, their control methods are relatively crude, mostly passive and reactive, and cannot be deeply integrated with the business characteristics of charging and discharging tests. Intervention actions often affect the normal execution progress and validity of test results.

[0007] In summary, existing energy management solutions for charge and discharge testing systems either rely on hardware investment, leading to high implementation costs, or are limited by independent operation, static rules, and passive response management models. These solutions struggle to address the core issues of excessively high local grid power peaks and low internal energy utilization efficiency during multi-task concurrent operation. They also fail to achieve stable control over power fluctuations during operation and lack the flexibility to manage dynamically added tasks. Consequently, they cannot achieve stable control of local grid power and efficient energy utilization while ensuring the normal execution of testing tasks and maintaining the validity of test results. There is still considerable room for optimization. Summary of the Invention

[0008] The purpose of this application is to provide a test task scheduling method, a test task scheduling device, an electronic device, and a readable storage medium, which can solve the problems of excessively high local power grid peak and low internal energy utilization efficiency when multiple charge and discharge test tasks are running concurrently, and at the same time improve the problems of difficult to smoothly control power fluctuations and insufficient flexibility in task management during operation.

[0009] To solve the above-mentioned technical problems, this application is implemented as follows: In a first aspect, embodiments of this application provide a method for scheduling test tasks, the method comprising: Based on the task information of the test task to be run, predictive data of the change of charging and discharging power over time during the execution of the test task is determined; wherein, the test process of the test task includes adjustable idle periods. Based on the predicted data, the timing of starting the test task to be run is controlled; Based on the deviation between the actual total net power and the expected total net power of each running test task, the remaining duration of the adjustable idle period for test tasks in the adjustable idle period is adjusted; wherein the expected total net power is determined based on the corresponding prediction data of each running test task.

[0010] Secondly, embodiments of this application provide a scheduling device for test tasks, the device comprising: The prediction data determination module is used to determine the prediction data of the change of charging and discharging power over time during the execution of the test task based on the task information of the test task to be run; wherein the test process of the test task includes adjustable idle periods. The start-up timing control module is used to control the start-up timing of the test task to be run based on the predicted data; The remaining duration adjustment module is used to adjust the remaining duration of the adjustable idle period of the test task in the adjustable idle period according to the deviation between the actual total net power and the expected total net power of each running test task; wherein the expected total net power is determined based on the corresponding prediction data of each running test task.

[0011] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0012] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0013] According to an embodiment of this application, based on the task information of a test task to be run, predicted data of the change in charging and discharging power over time during the execution of the test task to be run is determined; wherein, the test process of the test task includes an adjustable idle period; based on the predicted data, the start timing of the test task to be run is controlled; according to the deviation between the actual total net power and the expected total net power of each running test task, the remaining duration of the adjustable idle period of the test task in the adjustable idle period is adjusted; wherein, the expected total net power is determined based on the corresponding predicted data of each running test task. This solution pre-calculates power variation data for test tasks, enabling the system to understand the power consumption and feedback patterns of each task. By rationally scheduling the start-up time of each task, it guides the charging power consumption and discharging feedback of different tasks to cancel each other out over time, directly reducing the peak power of the local power grid during multi-task concurrency and minimizing energy waste from unabsorbed feedback power being output to the external grid. By comparing the deviation between actual and expected power, the remaining duration of idle periods is adjusted for correction, preventing the power deviation from continuously expanding and weakening the scheduling effect, while maintaining the accuracy of the test results. Ultimately, without the need for additional hardware, this solution improves the operational stability of the local power grid, enhances the energy utilization efficiency within the test system, reduces the capacity pressure on distribution facilities and operating electricity costs, and achieves intelligent peak shaving and energy-saving scheduling in multi-channel charging and discharging test scenarios. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the steps of a test task scheduling method provided in an embodiment of this application; Figure 2 This is a schematic diagram of an intelligent peak-shaving and energy-saving dispatching system; Figure 3 This is a structural block diagram of an embodiment of a test task scheduling device provided in another embodiment of the present invention; Figure 4 This is a structural block diagram of an electronic device for scheduling test tasks, according to an exemplary embodiment. Detailed Implementation

[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0016] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0017] The scheduling method for test tasks provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0018] The test task scheduling method, test task scheduling device, electronic device, and readable storage medium in the embodiments of this application can be executed by a local test management system, a dedicated scheduling server, a cloud energy management platform, or deployed in any other suitable hardware and software environment. The embodiments of this application do not impose any limitations on this. The test task scheduling method can be applied to a multi-channel charge-discharge test system.

[0019] In practical applications, this solution not only covers various electrochemical battery charge and discharge test scenarios, but can also be extended to multi-channel test scenarios with charge and discharge processes and resting periods, such as fuel cell testing, supercapacitor testing, and energy storage system testing. It has a wide range of applications and good scenario scalability and solution reusability.

[0020] In terms of technical implementation, this solution differs from the conventional approach of relying on external energy storage hardware buffers or general grid-side scheduling in traditional peak-shaving solutions. It fully leverages the operational characteristics of the testing process itself, and relies on the predictability of the testing program and the adjustable duration of the resting phase to achieve load peaking and energy efficiency improvement of the local power grid through pure software-level scheduling optimization. It does not require additional energy storage hardware devices, has flexible deployment methods, lower implementation and operation and maintenance costs, and the peak-shaving process will not interfere with the normal execution of the testing process or affect the validity of the test results.

[0021] Example 1 Reference Figure 1 The diagram illustrates a flowchart of a test task scheduling method provided in an embodiment of this application, which may specifically include the following steps: Step 101: Based on the task information of the test task to be run, determine the predicted data of the change of charging and discharging power over time during the execution of the test task to be run; wherein, the test process of the test task includes an adjustable idle period.

[0022] In this application embodiment, the test task refers to the charge-discharge performance verification task carried out for various energy storage devices, such as battery capacity testing, cycle life testing, and rate performance testing. This application embodiment does not limit this. Task information is the core parameter set that drives the operation of the test task, including current, voltage, cutoff conditions, number of cycles, etc., for each stage of the test process. It is usually carried in the form of a test program file and is the basic input data of the entire scheduling method. For example, the task information for the cycle life test task of a certain power battery cell corresponds to a preset test program file, which specifies the complete charge-discharge execution rules and stage parameters.

[0023] In this embodiment, the predicted charging and discharging power over time refers to time-series power data calculated in advance based on task information, covering the entire execution cycle of the test task. During the charging phase, the power is taken as a positive value, representing energy consumption from the local power grid; during the discharging phase, the power is taken as a negative value, representing energy fed back to the local power grid; and during the idle phase, the power is zero. This data transforms the unstructured test program into quantifiable and calculable structured data, providing a data foundation for subsequent startup timing planning and operational deviation verification. After the task transitions from a pending state to an active state, this predicted data can continue to be used for calculating the expected power.

[0024] In this embodiment, the adjustable idle period is the time during which the device is neither charged nor discharged and is in an open-circuit idle state during the test process. Its core characteristic is that the duration can be slightly adjusted within the range allowed by the process standard without changing the operating parameters of the charging and discharging phase, nor affecting the validity of the test results. The initial duration and allowable adjustment boundary of the adjustable idle period can be marked together in the predicted data. For example, the resting phase in the test process is a typical adjustable idle period, which serves to wait for the battery voltage to stabilize or the temperature to reach equilibrium. Adjusting the duration within the allowable range will not affect the accuracy of the test data.

[0025] In the embodiments of this application, there are various specific implementation methods for generating power prediction data based on task information, and the embodiments of this application do not limit this one.

[0026] In one specific implementation, the task information of the test task is first parsed in a structured manner to identify the stage type, running parameters and cutoff conditions of each step in the test process; then, for different types of stages, their power magnitude and expected duration are estimated, and the initial duration and allowable adjustment boundary of the idle stage are marked; finally, the power and timing data of all stages are spliced ​​together according to the execution order of the test process to generate complete prediction data.

[0027] In another specific implementation, for a test program containing loop execution logic, the power timing of a single loop is calculated first, and then the full-cycle prediction data is generated by splicing together the loop count.

[0028] For example, for a capacity test task of a lithium iron phosphate battery pack, the system reads its corresponding test program file and breaks it down into five standard stages: constant current charging, constant voltage charging, resting, constant current discharging, and resting. Combining the battery pack's rated voltage and test current parameters, the system estimates that the total power consumption of the charging stage is 60kW, expected to last 2.5 hours, and the total feedback power of the discharging stage is 58kW, expected to last 2.4 hours. The initial duration of the two resting stages is 20 minutes, with an allowable adjustment range of ±5 minutes. Finally, the system splices the data of each stage in the execution order to generate the power prediction data for the entire test cycle, and marks the location and adjustable boundaries of the two resting stages.

[0029] Step 102: Based on the predicted data, control the start time of the test task to be run.

[0030] In the embodiments of this application, controlling the start time of the test tasks to be run refers to actively arranging the start time of each test task so that the charging and power consumption phases and the discharging and feedback phases of different tasks complement each other in the time dimension, thereby reducing the power peaks and energy waste caused by the random superposition of multiple tasks from the root, and realizing peak shaving and valley filling of the local power grid.

[0031] In the embodiments of this application, there are various specific implementation methods for controlling the start-up timing based on predictive data, which can be adapted to different operating scenarios. The embodiments of this application do not limit this.

[0032] In one specific implementation, the initial batch startup scheduling method is suitable for scenarios where the system is initially running, there are currently no running test tasks, and all tasks are in a pending state, such as a scenario where a batch of test tasks are submitted after the test workshop is powered on. In this method, the system obtains the predicted data for all pending test tasks and maps the predicted data of all tasks to the same time axis. Within the allowed startup time range for each task, it iterates through different combinations of startup times, shifts the predicted data of each task according to the corresponding startup time, and then superimposes them moment by moment to obtain the local total net power curve under different combinations. With reducing the peak total net power as the core optimization objective, the globally optimal startup time combination is selected. Finally, according to the time corresponding to the optimal combination, startup commands are sequentially issued to each test device, controlling all pending tasks to start according to the planned time sequence.

[0033] For example, after a testing workshop is powered on, it submits eight power battery capacity testing tasks in batches. The peak power of each task during the charging phase is 80kW. If all of them start simultaneously, the local total net power peak will reach 640kW, far exceeding the normal load of the power distribution facilities. By scheduling the initial batch, the system staggers the start times of the eight tasks by 15-minute intervals, distributing the charging peaks of each task. Ultimately, the local total net power peak is controlled within 200kW, significantly reducing the instantaneous load impact.

[0034] In another specific implementation, the dynamic incremental access scheduling method is suitable for scenarios where some existing test tasks are already running and new test tasks are added to be run, such as when operators submit new test tasks one after another during test operation. In this method, the system obtains the predicted data corresponding to the running test tasks. This predicted data is generated when the corresponding task is in the waiting stage and continues to be used after the task starts. The predicted data of the running tasks and the predicted data of the newly added test tasks are mapped to the same time axis. Combined with the real-time progress of the currently running tasks, different access times are traversed within the allowed waiting time window for the new task, and the local total net power curves under different access times are superimposed. The optimal access time for the new task is selected with the minimum increase in the peak total net power after access as the optimization objective. When the system time reaches this optimal access time, a start command is sent to the corresponding device to control the start of the task to be run.

[0035] For example, in a test workshop's local power grid, three test cabinets are already in the discharge feedback phase, with a total feedback power of approximately 150kW, expected to last for 2 hours. At this time, two new charging test tasks are submitted, each with a charging power of 80kW. If started simultaneously, the combined power consumption of the two would reach a peak of 160kW, resulting in a net power consumption peak of 10kW even after offsetting. The system uses dynamic incremental scheduling to schedule the start times of the two new tasks at the beginning and middle of the existing discharge phase, respectively, so that the charging power consumption of the new tasks is offset as much as possible by the existing discharge feedback power. Ultimately, in the optimal solution, the total net power peak of the local power grid after the new tasks are added is reduced to below 10kW, without significantly impacting the existing power balance.

[0036] Step 103: Adjust the remaining duration of the adjustable idle period for test tasks in the adjustable idle period based on the deviation between the actual total net power and the expected total net power of each running test task; wherein the expected total net power is determined based on the corresponding prediction data of each running test task.

[0037] In this embodiment, the total net power refers to the algebraic sum of the power of all running test tasks within the local power supply network. A positive value represents the overall energy consumption from the external power grid, while a negative value represents the overall energy fed back to the external power grid. The expected total net power is calculated by superimposing the predicted data of each running task, while the actual total net power is obtained in real time through power acquisition devices. The deviation between the two is the core criterion for triggering dynamic adjustments.

[0038] In the embodiments of this application, adjusting the remaining duration of the adjustable idle period is based on utilizing the lossless adjustment characteristics of the adjustable idle period to correct the power deviation that occurs during operation in real time. This solves the problem that traditional solutions lack fine-grained control methods when there is a deviation between actual operation and prediction, and maintains the stability of scheduling effect.

[0039] In the embodiments of this application, there are various specific ways to adjust the idle time based on the power deviation, and the embodiments of this application do not limit this one.

[0040] In one specific implementation, real-time power data of the local power grid is collected at fixed intervals to obtain the actual total net power at the current moment. Simultaneously, based on the predicted data of each running test task, the expected total net power at the same moment is calculated. The two are compared to obtain the power deviation value. When the deviation exceeds the preset allowable range, the task currently in the adjustable idle period is selected from the running tasks, and the required duration adjustment is calculated based on the deviation direction and magnitude. Within the allowable adjustment boundary of the idle period, its remaining idle time is modified, thereby delaying or advancing the start of the next stage and correcting the total net power deviation.

[0041] In another specific implementation, different response levels can be set according to the magnitude of the deviation. When the deviation is small, the idle time of a single device is adjusted, and when the deviation is large, the idle time of multiple devices is adjusted.

[0042] For example, during testing, because the actual internal resistance of the battery was slightly lower than the design value, the actual discharge rate of a certain batch of discharge tasks was faster than predicted, resulting in the actual feedback power of the local power grid being higher than expected, and the negative deviation of the total net power exceeding the preset threshold. After detecting the deviation, the system selected 4 test tasks currently in the resting phase and calculated that each task needed to have its resting time extended by 3 minutes. After verifying that the 3-minute adjustment was within the allowable range of 25-35 minutes, the system issued instructions to the corresponding devices to extend the remaining time of their current resting phase, postpone the start of the subsequent discharge phase, offset the excess feedback power, and bring the total net power back to the expected range. The entire adjustment process did not change the core parameters of the charging and discharging phases and did not affect the validity of the test results.

[0043] According to an embodiment of this application, based on the task information of a test task to be run, predicted data of the change in charging and discharging power over time during the execution of the test task to be run is determined; wherein, the test process of the test task includes an adjustable idle period; based on the predicted data, the start timing of the test task to be run is controlled; according to the deviation between the actual total net power and the expected total net power of each running test task, the remaining duration of the adjustable idle period of the test task in the adjustable idle period is adjusted; wherein, the expected total net power is determined based on the corresponding predicted data of each running test task. This solution pre-calculates power variation data for test tasks, enabling the system to understand the power consumption and feedback patterns of each task. By rationally scheduling the start-up time of each task, it guides the charging power consumption and discharging feedback of different tasks to cancel each other out over time, directly reducing the peak power of the local power grid during multi-task concurrency and minimizing energy waste from unabsorbed feedback power being output to the external grid. By comparing the deviation between actual and expected power, the remaining duration of idle periods is adjusted for correction, preventing the power deviation from continuously expanding and weakening the scheduling effect, while maintaining the accuracy of the test results. Ultimately, without the need for additional hardware, this solution improves the operational stability of the local power grid, enhances the energy utilization efficiency within the test system, reduces the capacity pressure on distribution facilities and operating electricity costs, and achieves intelligent peak shaving and energy-saving scheduling in multi-channel charging and discharging test scenarios.

[0044] Example 2 In one optional embodiment of this application, a specific implementation of determining the predicted data of the change of charging and discharging power over time during the execution of the test task based on the task information of the test task to be run may include: dividing different functional time periods in the test process based on the task information, determining the power attribute and expected duration of each functional time period; generating the predicted data according to the power attribute and expected duration of each functional time period, and marking the position of the adjustable idle time period and its adjustable duration boundary in the predicted data.

[0045] Dividing the test process into functional time periods is the core process of structurally decomposing the execution flow of test tasks. By identifying the functional attributes of different stages in the process, the complete unstructured test process is broken down into multiple independent time periods with clear power characteristics, providing a structured data foundation for subsequent power prediction calculations. Functional time periods can be divided into power consumption time periods, energy depletion time periods, and idle time periods according to their power attributes. Among them, idle time periods correspond to adjustable idle time periods with the ability to adjust their duration.

[0046] There are various ways to determine the power attributes and expected duration of each functional period, and the embodiments of this application do not limit this.

[0047] In one specific implementation, the task information of the test task to be run is first read and parsed, and the type identifier, target operating parameters, and cutoff conditions of each execution step in the test process are identified and extracted. The type identifier distinguishes steps with different functions such as charging, discharging, and resting. Target operating parameters include target current and target voltage, and cutoff conditions include cutoff voltage, cutoff current, cutoff time, and number of cycles. Subsequently, quantitative estimations are performed for different types of steps: for charging steps, the average power consumption and expected duration are estimated based on the target charging current and preset operating voltage range; for discharging steps, the average feedback power and expected duration are estimated based on the target discharging current and corresponding operating voltage range; for resting steps, the preset initial duration is recorded, and the minimum and maximum allowable adjustment amounts under the corresponding test standard are matched as adjustable duration boundaries.

[0048] Generating forecast data and labeling adjustable idle time periods is the process of integrating discrete single-time power and time-series parameters into a complete full-cycle forecast dataset. At the same time, it reserves fast index information needed for subsequent dynamic adjustments to ensure the execution efficiency of subsequent scheduling and fine-tuning steps.

[0049] There are various ways to generate predicted data and labeled boundaries, and this application does not limit the specific implementation of these methods.

[0050] In one specific implementation, following the inherent execution order of the test process, the power data and timing parameters of all functional time periods are sequentially concatenated. With time as the horizontal axis and power as the vertical axis, this forms continuously varying charging and discharging power prediction timing data, i.e., prediction data. Power values ​​are positive for power consumption periods, representing energy consumption from the local grid; negative values ​​are used for power feedback periods, representing energy fed back to the local grid; and zero values ​​are used for idle periods. Subsequently, the corresponding information for all adjustable idle time periods is extracted from the prediction data, including the start position, initial duration, and upper and lower boundaries of the adjustable duration for each idle time period on the time axis. This forms the index information for adjustable idle time periods, which is then associated and annotated in the prediction data for direct use in subsequent scheduling optimization and dynamic adjustment steps.

[0051] For example, in a cycle life test task for a specific power battery cell, after reading the corresponding test program file, the system first parses and extracts the five execution steps for each cycle: "constant current charging, constant voltage charging, rest, constant current discharging, and rest," as well as the current, voltage, cutoff time parameters for each step and the total number of cycles. Then, it estimates the parameters for each step: the target current for the constant current charging stage is 1C, and based on the battery's rated voltage, the estimated average power consumption is 5kW, expected to last 58 minutes; the estimated average power consumption for the constant voltage charging stage is 2kW, expected to last 8 minutes; the initial duration for both rest stages is 30 minutes, with an adjustment range of ±5 minutes allowed according to the corresponding test standard, meaning the duration adjustment boundary is 25 to 35 minutes; the target current for the constant current discharging stage is 1C, the estimated average feedback power is 4.8kW, and expected to last 60 minutes.

[0052] After completing the parameter estimation for a single time period, the system splices all time period data of a single cycle in the execution order, and then expands to generate a power prediction time series curve covering the entire test cycle based on the total number of cycles. At the same time, it extracts the starting position, initial duration and adjustment boundary of the two resting phases on the time axis, generates a resting node index table, and associates and marks it in the prediction data to complete the generation of the entire set of prediction data.

[0053] By dividing the test process into functional time periods and quantifying the power and timing parameters of each time period, full-cycle power prediction data is generated. At the same time, the location and adjustment boundaries of adjustable idle time periods are marked. This transforms unstructured test procedures into quantifiable and calculable structured data, giving the system the ability to anticipate test energy consumption and providing a reliable data foundation for subsequent timing scheduling and deviation verification. Meanwhile, by pre-defining the adjustment boundaries of idle time periods, it ensures that subsequent dynamic adjustments are always performed within the range allowed by the test standards and do not affect the validity of the test results. Furthermore, the index information improves the response speed and execution efficiency of subsequent real-time fine-tuning.

[0054] Example 3 In one optional embodiment of this application, a specific implementation of controlling the start timing of the test task to be run based on the predicted data may include: determining a combination of start times for the test task to be run based on the predicted data, with the optimization objective of reducing the peak value of the total net power; and controlling the start of the test task to be run according to the combination of start times.

[0055] Determining the start time combination of the test tasks to be run is the core calculation link of global timing scheduling. By superimposing simulations of the power timing of multiple tasks, the optimal start time arrangement scheme is searched, which optimizes the total power distribution of the local power grid from the root, achieves the core goal of peak shaving and valley filling, and provides a scheduling basis for the subsequent task start execution.

[0056] There are various ways to determine the specific combination of startup times, and this application does not limit this one.

[0057] In one specific implementation, a multi-task power superposition calculation model is first constructed: the prediction data of all test tasks to be run are mapped to the same time axis. The prediction data can be discretized into a power value sequence according to a preset fixed time resolution to improve the calculation efficiency. For any set of candidate start times, the prediction data of each task is shifted according to the corresponding start time, and the power value is algebraically summed time by time to obtain the local total net power prediction curve under the candidate combination.

[0058] Then, the optimization objectives and constraints are set: minimizing the peak value of the total net power is taken as the core optimization objective, and other optimization sub-objectives and constraints can be added.

[0059] Finally, an appropriate solution strategy is selected based on the scale of concurrent tasks: when the number of concurrent tasks to be run is less than a preset threshold, a heuristic enumeration search strategy is adopted to traverse all candidate startup time combinations within the allowed startup time window and select the combination with the smallest total net power peak as the optimal startup time combination; when the number of concurrent tasks to be run is greater than or equal to the preset threshold, a greedy iterative strategy is adopted to sort all tasks to be run in descending order of power peak and include them in the scheduling scheme one by one. Each time, the startup time that minimizes the increase in the existing total net power peak is selected for the current task, and finally, a globally approximately optimal startup time combination is obtained.

[0060] Controlling the startup of test tasks based on startup time combinations is the step in translating scheduling calculations into actual execution actions. Through precise timed triggering and instruction issuance, it ensures that the optimal scheduling scheme can be executed accurately.

[0061] There are various ways to implement the control of the start of the test task, and this application does not limit the specific implementation of this method.

[0062] In one specific implementation, a corresponding startup timer is set for each test task to be run based on a determined combination of startup times. When the system clock reaches the startup time corresponding to a certain task, a startup execution command is sent to the test device to which the task belongs, triggering the device to start running according to the preset test procedure. Optionally, a response confirmation mechanism can be used to ensure the reliability of the command. After receiving the command, the device sends back a confirmation message. If no confirmation is received within the time limit, the command is automatically resent to ensure that the scheduling command is delivered accurately.

[0063] For example, after a testing workshop is turned on, it submits eight power battery capacity testing tasks in batches. The peak power of each task during the charging phase is 80kW. All tasks must be started within a two-hour time window after submission.

[0064] The system first discretizes the power prediction data of 8 tasks into a power value sequence with a fixed time resolution of 1 minute. Since the number of concurrent tasks is small, a heuristic enumeration search strategy is adopted to traverse the candidate start time combinations of each task within the allowed start window of 2 hours, calculate the total net power curve after superposition of each group, and select the combination with the smallest total net power peak as the optimal start time combination.

[0065] In the final optimal combination, the start times of each task are arranged sequentially at 15-minute intervals, reducing the total net power peak of 640kW when started simultaneously to less than 200kW. The system sets the start time for each task according to this start time combination. When the clock reaches the corresponding start time in sequence, the system sends a start command to the test equipment one by one, controlling the test tasks to start and run according to the planned sequence.

[0066] By using the predicted data from each test task, the optimal start-up time combination is solved with the goal of reducing the total net power peak. The task start-up is controlled according to the combination. It can actively guide the charging power consumption and discharging feedback of different tasks to complement each other in the time dimension, thereby dispersing the power load from the source, reducing the instantaneous peak, and reducing the capacity pressure and overload risk of local power distribution facilities. At the same time, the solution strategy that adapts to different concurrency scales can take into account both optimization effect and computational efficiency. Moreover, the whole process is implemented through software scheduling, without the need for additional energy storage hardware, which has good economic efficiency and engineering feasibility.

[0067] Example 4 In one optional embodiment of this application, when determining the combination of start-up times for the test tasks to be run, at least one of the following is further included: optimization sub-objective: reducing the cumulative amount of net electrical energy output to the external power grid during the period representing net feedback in the total net power; constraint condition: ensuring that the start-up time of each test task is within the allowable start-up time window; constraint condition: ensuring that the waiting start-up time of each test task does not exceed the allowable upper limit.

[0068] In the process of solving the global timing scheduling, based on reducing the total net power peak as the core optimization objective, by configuring differentiated optimization sub-objectives and constraints, the final output start time combination can simultaneously take into account the peak shaving effect, energy utilization efficiency and the scheduling requirements of actual test services, avoiding the problem of insufficient service adaptability caused by pure peak optimization, and improving the engineering practicality of the scheduling scheme.

[0069] When the total discharge feedback power of all test tasks within the local power grid exceeds the total charging power consumption, the total net power is negative, and the corresponding period is the net feedback period. During this period, excess feedback energy will be output to the external power grid. If the external power grid cannot fully absorb it or the feedback electricity price is low, it will result in ineffective energy waste. Reducing the accumulated energy during the net feedback period, i.e., through timing arrangement to allow charging power consumption to offset discharge feedback as much as possible, reduces ineffective energy output to the external grid and further improves the energy self-balancing rate within the local area.

[0070] In one specific implementation, when solving for the startup time combination, for the total net power curve corresponding to each group of candidate startup times, in addition to calculating the peak value of the total net power, the cumulative value of electrical energy during all net feedback periods is also counted. In the multi-objective optimization solution, this cumulative value is used as the secondary optimization objective. Among the candidate schemes with similar peak performance, the scheme with smaller net feedback electrical energy accumulation is selected first, so as to achieve dual optimization of peak shaving and improving self-balancing rate.

[0071] Constraints are used to limit the legal solution range of the scheduling scheme, ensuring that the final output start time combination conforms to the scheduling rules of the actual test business. They mainly include two types of constraints: The allowed start time window refers to the legal start time range set for each test task according to the test business scheduling requirements. The actual start time of the task must fall within this time range to match the overall test plan and workstation arrangement.

[0072] In one specific implementation, during the process of traversing candidate start times or iteratively solving, candidate times that exceed the allowed start time window of the corresponding task are pre-filtered, and the optimal solution is searched only within the legal time range to ensure that the final output start time combination meets the business scheduling requirements.

[0073] The waiting time for startup refers to the time interval from the submission of a test task to its official execution. Setting an allowable upper limit can prevent tasks from being delayed indefinitely in pursuit of optimal peak-shaving effects, thus ensuring the timely execution of regular test tasks.

[0074] In one specific implementation, a corresponding maximum waiting time is configured for each test task. During the solution process, the time difference between the candidate start time and the task submission time is checked, and candidate solutions that exceed the allowed upper limit are directly excluded. Optionally, a differentiated upper limit for the duration can be set based on the task priority, with higher priority tasks corresponding to a shorter allowed waiting upper limit, to ensure the timely execution of urgent test tasks.

[0075] For example, in the scheduling scenario of 8 power battery capacity testing tasks, if the only objective is to reduce the total net power peak, the optimal solution obtained would require the latest task to start to wait for 110 minutes, and the cumulative net output of electrical energy to the external power grid during the testing period would be approximately 120 kWh.

[0076] In this embodiment, two constraints are configured for all test tasks: the start time window is within 2 hours after the task is submitted, and the maximum waiting time for a single task to start is 90 minutes; at the same time, an optimization sub-objective of "reducing the net feedback energy accumulation" is added.

[0077] When the system re-solves the problem, candidate start-up times with waiting times exceeding 90 minutes are first filtered out. From the remaining valid combinations, the total net peak power and cumulative net feedback energy are calculated simultaneously. A comprehensive optimal solution is selected through multi-objective trade-offs. In the final start-up time combinations, the total net peak power is approximately 210kW, only slightly higher than the pure peak-optimal solution, but still far lower than the 640kW of simultaneous start-ups. The longest waiting time for a single task is 85 minutes, meeting the 90-minute upper limit requirement. Simultaneously, the cumulative net energy output to the external grid during the test period decreases to 82kWh, and the local energy self-balancing rate increases by approximately 30%, achieving a comprehensive optimization of peak-shaving effect, execution timeliness, and energy utilization.

[0078] By configuring additional optimization sub-objectives and multi-dimensional constraints during the solution process of startup time combinations, the internal energy self-balancing rate of the local power grid can be further improved while ensuring the core peak shaving effect and reducing the waste of ineffective feedback power. At the same time, through time window and waiting time constraints, the scheduling scheme can be fully adapted to the scheduling requirements of actual test services, taking into account both energy efficiency optimization and service execution timeliness, thereby improving the scenario adaptability and engineering implementation value of the scheduling algorithm.

[0079] Example 5 In one optional embodiment of this application, a specific implementation of the test tasks to be run including newly submitted test tasks may further include: obtaining prediction data corresponding to the newly submitted test tasks and adding it to the set of test tasks to be run; and in another specific implementation of controlling the start time of the test tasks to be run based on the prediction data, the implementation includes: determining the start time combination of the test tasks to be run in the set of test tasks to be run based on the prediction data of each currently running test task and the prediction data of the set of test tasks to be run, with the optimization objective of reducing the peak value of the total net power.

[0080] Acquiring new task prediction data and enqueuing it is a pre-processing step in dynamic scheduling scenarios. It is used to uniformly incorporate newly submitted test tasks into the global scheduling system, providing complete input data for subsequent joint optimization calculations and ensuring that the scheduling scope covers all tasks to be executed.

[0081] There are various ways to implement the new task enqueueing process, and this application embodiment does not limit this one.

[0082] In one specific implementation, after receiving a newly submitted test task, the system first parses the task information according to the aforementioned prediction data generation logic, generates corresponding prediction data on the change of charging and discharging power over time, then adds the new task to the set of test tasks to be run, marks it as pending scheduling, and triggers the global scheduling recalculation process to ensure that the new task can participate in unified timing optimization together with the original pending tasks.

[0083] The core scheduling logic in dynamic incremental scenarios is to jointly solve for the startup time combination by combining the predicted data of running tasks and tasks to be run. Unlike the initial batch scheduling which is based solely on the data of tasks to be run, in dynamic scenarios, existing tasks are already running, and the power changes of their remaining execution cycles directly affect the local total net power level. Only by incorporating the power timing of running tasks into the calculation can we ensure that the overall power balance is not disrupted after the new task is added, and avoid sudden power peak impacts.

[0084] There are various ways to jointly solve the startup time combination, and this application does not limit the specific implementation of this method.

[0085] In one specific implementation, the predicted data of all currently running test tasks are first obtained. Combined with the real-time progress of each task, the power timing data corresponding to its remaining execution cycles is extracted. Then, the remaining timing data of the running tasks and the complete predicted data of all tasks in the set of test tasks to be run are mapped to the same time axis. Within the allowable constraints of each task to be run, with the optimization objective of reducing the total net power peak, the updated combination of task start times is obtained, and the new scheme replaces the original scheduling scheme. The entire rescheduling process only adjusts the start times of tasks that have not yet started, without interfering with the normal execution of running tasks, ensuring the stability of the test process.

[0086] For example, in a testing workshop scheduling scenario, after one hour of operation of the initial batch of 8 power battery capacity testing tasks, 3 tasks have entered the discharge feedback phase, with a feedback power of approximately 50kW per task, totaling 150kW, and an estimated remaining discharge time of 1.5 hours. There are still 2 unstarted tasks in the pending test task set, with a peak charging power of 80kW per task. If the operator then submits 2 new power battery charging test tasks of the same specifications, the system first parses the test program parameters of the new tasks, generates corresponding charge and discharge power prediction data, adds the 2 new tasks to the pending test task set, and triggers a global rescheduling. If the 2 new tasks are started immediately, the combined local net power consumption will reach a peak of 160kW. Even after offsetting the existing feedback, there will still be a net peak power consumption of 10kW, and the peak value will further increase after being superimposed with the charging phase of the original pending tasks. The system performs a global solution by combining the remaining feedback timing of the running tasks with the predicted data of all tasks to be run. The startup times of the two new tasks are scheduled during the peak periods of the existing discharge feedback. At the same time, the startup timing of the original tasks to be run is optimized. In the end, the peak value of the local net power after the new tasks are connected is always controlled within 12kW, without causing a significant impact on the existing power balance, thus achieving a smooth connection of the new tasks.

[0087] By incorporating newly submitted test tasks into the set to be run and combining them with the prediction data of running tasks for global joint rescheduling, the optimal peak-shaving effect can be maintained continuously in scenarios where test tasks are dynamically added or removed. This avoids the random start of new tasks from disrupting the existing power balance and achieves smooth access for new tasks. At the same time, this mechanism is adapted to the business characteristics of tasks being submitted and changing dynamically in actual test production. It can automatically complete the rolling update of the scheduling scheme without manual intervention, improving the scenario adaptability and operational automation level of the scheduling system.

[0088] Example 6 In one optional embodiment of this application, a specific implementation of adjusting the remaining duration of the adjustable idle period of a test task in an adjustable idle period based on the deviation between the actual total net power and the expected total net power of each running test task may include: identifying a test task in the adjustable idle period when the absolute value of the deviation between the actual total net power and the expected total net power exceeds a first trigger threshold; and adjusting the remaining duration of the adjustable idle period of the identified test task based on the deviation.

[0089] Deviation trigger judgment is a prerequisite criterion for real-time fine-tuning. By setting a reasonable trigger threshold, it filters out small power fluctuations within the normal range, avoids frequent triggering of adjustment actions that interfere with the test process and increase system computing overhead, and ensures the stability of test operation and the operating efficiency of the scheduling system.

[0090] There are various ways to implement the deviation trigger judgment, and the embodiments of this application do not limit this one.

[0091] In one specific implementation, the system collects the actual total net power of the local power grid at a fixed sampling period, and simultaneously calculates the expected total net power at the same moment based on the predicted data of each running test task; the difference between the two is used to obtain the power deviation at the current moment, and the absolute value of the deviation is calculated; the absolute value of the deviation is compared with a preset first trigger threshold. If it exceeds the threshold, it is determined that a fine-tuning action needs to be triggered; if it does not exceed the threshold, it is determined to be a normal fluctuation, and no adjustment operation is performed.

[0092] Identifying adjustable idle tasks and adjusting the remaining time is the core execution step of the real-time fine-tuning process. By selecting tasks with adjustment capabilities as adjustment targets and matching the corresponding adjustment actions with the actual deviation, the power deviation is corrected without changing the core charging and discharging parameters, thus maintaining a stable local power state.

[0093] There are various ways to implement idle task identification and duration adjustment, and this application embodiment does not limit them.

[0094] In one specific implementation, after the fine-tuning action is triggered, the system iterates through all running test tasks, matches the current execution stage of each task with the time period division in the prediction data, and selects the tasks currently in the adjustable idle period as the adjustment targets. Then, based on the direction and magnitude of the deviation, the corresponding duration adjustment strategy and adjustment range are determined. Within the allowable adjustment boundary of the adjustable idle period, the remaining idle time of the target task is modified. By delaying or advancing the start time of the next stage, the total net power is driven back to the expected range.

[0095] For example, during the parallel operation of multiple power battery testing tasks in a testing workshop, the actual polarization voltage of the tested battery was lower than the design value, resulting in a slightly higher actual power consumption during the charging phase than the predicted value. This caused the actual net power in the local area to be 18kW higher than the expected net power. The system's preset first trigger threshold is 15kW. When the absolute value of the current deviation exceeds this threshold, a fine-tuning action is triggered. The system then iterates through all running test tasks and identifies four tasks currently in an adjustable idle period. Based on the magnitude of the positive deviation, the remaining idle time of these four tasks is extended, and the start time of the subsequent charging phase is postponed. After the adjustment, the actual net power in the local area gradually returns to the expected range. The entire adjustment process does not change the core test parameters such as current and voltage during the charging and discharging phases, and does not affect the accuracy of the final test results.

[0096] By setting a first trigger threshold as the starting criterion for real-time fine-tuning, minor power fluctuations within the normal range can be effectively filtered out, avoiding frequent adjustments that interfere with the testing process and reducing system operating costs. By accurately identifying tasks in adjustable idle periods and adjusting the duration based on the deviation, power deviations can be quickly corrected without affecting test quality, maintaining stable local power operation and ensuring the continuous and stable effect of pre-sequence scheduling.

[0097] Example 7 In one optional embodiment of this application, it may further include: when the absolute value of the deviation between the actual total net power and the expected total net power exceeds a second trigger threshold, updating the start timing of subsequent test tasks to be run based on the prediction data; wherein the second trigger threshold is higher than the first trigger threshold.

[0098] Setting high and low trigger thresholds and matching them with different levels of control measures is to construct a gradient power management system. This system employs appropriate response strategies for different degrees of power deviation, avoiding over- or under-control. The first trigger threshold corresponds to minor deviations, which is fine-tuned by matching the idle time of running tasks, resulting in a fast response and minimal impact on the testing process. The second trigger threshold corresponds to severe deviations, indicating that local fine-tuning alone is insufficient to effectively mitigate the deviation. It is necessary to adjust the startup plans of unstarted tasks at the global timing level to fundamentally correct the subsequent power superposition state. The two thresholds work together to form a hierarchical control logic of "minor fine-tuning and severe re-tuning," balancing scheduling efficiency and operational stability.

[0099] There are various ways to implement the graded deviation response, and the embodiments of this application do not limit this one.

[0100] In one specific implementation, the system calculates the absolute value of the power deviation at the current moment in real time and compares it with two preset trigger thresholds: if the absolute value of the deviation does not exceed the first trigger threshold, it is determined to be a normal power fluctuation, and no adjustment action is performed; if the absolute value of the deviation exceeds the first trigger threshold but does not exceed the second trigger threshold, it is determined to be a slight deviation, and the remaining duration of the adjustable idle period is adjusted to smooth out the deviation through local correction; if the absolute value of the deviation exceeds the second trigger threshold, it is determined to be a severe deviation, and while performing fine-tuning during the idle period, the startup timing update process of the tasks to be run is triggered, and subsequent power deviations are corrected from the root cause through global rescheduling.

[0101] There are various ways to update the timing of subsequent test tasks to be run, and this application embodiment does not limit this.

[0102] In one specific implementation, after triggering a severe deviation response, the system uses the current moment as the starting point, extracts the power timing data corresponding to the remaining execution cycles of all running tasks, and combines it with the predicted data of all tasks to be run. It then recalculates the optimal combination of startup times for the tasks to be run, with the optimization objective of reducing the total net peak power and minimizing power deviation. The new scheduling scheme replaces the original scheme, and the startup timings of all subsequent tasks to be run are updated synchronously. The entire rescheduling process only adjusts the startup plans of tasks that have not yet started, without interrupting or modifying the test parameters and processes of running tasks, ensuring the continuity of test services and the validity of test results.

[0103] For example, in a parallel testing scenario for power batteries, the system presets a first trigger threshold of 15kW and a second trigger threshold of 30kW. During the test, due to the consistency deviation in internal resistance among batteries of the same batch, the actual power consumption during the charging phase continuously exceeds the expected value, and the deviation of the local net power gradually expands to 35kW. The absolute value of the deviation exceeds the second trigger threshold of 30kW, and it is judged as a severe power deviation. At this time, simply adjusting the adjustable idle time of the currently running task can only offset about 20kW of the deviation at most, which cannot bring the total net power back to a reasonable range. The system then triggers the task start timing update process: based on the current time, combined with the power timing data of the remaining cycle of the currently running task, and the predicted data of the remaining 5 task tasks to be run, a global recalculation is performed. The 3 task tasks to be run that were originally planned to start successively within 10 minutes are postponed until after the current high power consumption phase ends, thus avoiding the power peak. After updating the startup timing, the peak deviation of the total net power in the local area gradually dropped to within 10kW. With fine-tuning during the idle period of the running task, the power deviation was finally stabilized below the first trigger threshold. This not only did not interrupt the ongoing test process, but also effectively corrected the severe power deviation at the global level.

[0104] By setting high and low trigger thresholds, a hierarchical control strategy of "local idle fine-tuning + global timing readjustment" is constructed. This strategy can quickly fine-tune routine minor deviations to minimize interference with the testing process, while also addressing power fluctuations at their source by updating the start timing of pending tasks when severe deviations occur, preventing the risk of power distribution overload caused by the continuous expansion of deviations. The two-level response mechanism works in tandem and progressively advances, balancing scheduling response speed, operational stability, and global optimization effects, significantly improving the system's adaptability and operational reliability in the face of varying degrees of power fluctuations.

[0105] Example 8 In one optional embodiment of this application, a specific implementation of adjusting the remaining duration of the adjustable idle period of the identified test task based on the deviation amount may include: selecting the test task in the adjustable idle period as the target task; calculating the duration adjustment amount of the adjustable idle period of the target task based on the deviation amount; and correcting the remaining duration according to the duration adjustment amount and the allowable adjustment boundary of the adjustable idle period of the target task.

[0106] Selecting the target task is the object screening step in the real-time fine-tuning process. By matching the deviation direction with the task stage attributes, the optimal adjustment object is selected to ensure that the adjustment action can accurately correct the power deviation and improve the effectiveness and response efficiency of fine-tuning.

[0107] There are various ways to select the target task, and the embodiments of this application do not limit this.

[0108] In one specific implementation, after triggering the fine-tuning action, the system first iterates through all running test tasks and filters out candidate tasks currently in an adjustable idle period. Then, based on the direction of the power deviation, it further filters tasks whose power attributes match the deviation direction in the next stage as target tasks. When the deviation is positive and a reduction in total net power is needed, tasks in the next power-consuming period are prioritized, and the start of the power-consuming stage is delayed by extending the idle time, directly offsetting the excess power consumption. When the deviation is negative and a increase in total net power is needed, tasks in the next power-consuming period are prioritized, and the power-consuming stage is started earlier by shortening the idle time, absorbing excess feedback energy. Optionally, tasks with larger remaining adjustable margins can be prioritized to achieve a larger deviation correction range with a single task, reducing the number of tasks requiring adjustment and minimizing the overall impact on the testing process.

[0109] Calculating the duration adjustment amount is the core quantitative step in the real-time fine-tuning process. By accurately calculating the required duration adjustment range through the deviation amount and the stage power parameters of the target task, it avoids under-adjustment or over-adjustment and ensures the accuracy of deviation correction.

[0110] There are various ways to calculate the duration adjustment amount, and this application does not limit this method.

[0111] In one specific implementation, the corresponding calculation logic is matched based on the deviation direction: When the deviation is positive and the current total net power needs to be reduced, select the target task as the next power-consuming stage and calculate the required extended idle time using the following formula: Required extended idle time = current deviation × expected duration of delayed power consumption phase ÷ expected power consumption phase of the task.

[0112] The current deviation is the difference between the actual total net power and the expected total net power. The expected duration of the delayed power consumption phase can be preset to a fixed value or adaptively matched according to the actual deviation.

[0113] When the deviation is negative and the current total net power needs to be increased, select the target task of the next stage as the power consumption stage, calculate the required reduction of idle time, and reduce the negative deviation by absorbing excess feedback power through early start of the power consumption stage.

[0114] Optionally, when there are multiple target tasks, the total adjustment requirements can be distributed across multiple tasks, with each task receiving a corresponding proportion of the adjustment amount. This reduces the adjustment range for each task and further weakens the impact on the testing process of a single task.

[0115] Adjusting the remaining time in conjunction with the allowable adjustment boundary is a key verification step to ensure that the adjustment is compliant and non-destructive. By limiting the adjustment range to within the limits allowed by the process standards, it is ensured that the time adjustment will not affect the validity of the test results, thus maintaining the bottom line of test quality.

[0116] There are various ways to modify the remaining time, and this application does not limit the specific implementation of these methods.

[0117] In one specific implementation, after calculating the duration adjustment amount for a single target task, the allowable adjustment boundary of the task's current adjustable idle period is retrieved, and it is verified whether the adjusted remaining duration falls within the allowable duration range. If the adjusted remaining duration does not exceed the upper or lower boundary, the remaining duration is directly corrected according to the calculated adjustment amount; if the adjustment amount exceeds the allowable boundary, the boundary value is used as the final adjustment range, and the remaining duration is corrected to the corresponding boundary value, ensuring that the adjustment is always performed within the range allowed by the test standard. After verification, a duration adjustment command is issued to the corresponding test device to update the remaining duration of the current idle period.

[0118] For example, in the aforementioned parallel testing scenario for power batteries, the system detected a positive deviation of 18kW in the local total net power, exceeding the first trigger threshold of 15kW, and triggered a fine-tuning action.

[0119] The system first filters candidate tasks, and after traversing all running test tasks, identifies 4 tasks that are currently in an adjustable idle period and whose next stage is a charging and power consumption stage as target tasks. The initial idle period of this batch of tasks is 30 minutes, and the adjustable boundary is 25 to 35 minutes. The remaining idle time of each task is 20 minutes.

[0120] The duration adjustment was then calculated: the expected duration of a single delayed charging phase was set at 3 minutes, and the expected power of the next charging phase for a single task was 80kW. The total adjustment requirement was calculated using the formula: 18kW × 3 minutes ÷ 80kW ≈ 0.675 minutes. The adjustment was distributed among the 4 target tasks, with each task extended by approximately 0.7 minutes, which could offset a power deviation of approximately 18.7kW, completely covering the current deviation.

[0121] Next, boundary verification was performed: the current remaining idle time for each task was 20 minutes. After extending it by 0.7 minutes, the remaining time was 20.7 minutes, which is still within the allowable adjustment range of 25 to 35 minutes and meets the test standard requirements.

[0122] Finally, the system issued instructions to the four test devices corresponding to the target tasks, extending the remaining time of their current resting phase by 0.7 minutes and postponing the start time of the subsequent charging phase. After the adjustment, the total net power of the local area gradually returned to the expected range. The core test parameters of the charging and discharging phases were not changed throughout the process, and the accuracy of the test results was not affected.

[0123] According to the embodiments of this application, by selecting the target task with the direction of the matching deviation, accurately calculating the duration adjustment amount in combination with power parameters, and verifying the correction range through the allowable adjustment boundary, accurate and controllable correction of power deviation can be achieved. This ensures both the correction efficiency and accuracy of the fine-tuning action, and ensures that all adjustments are performed within the allowable range of the process standard through boundary constraints, strictly guaranteeing the validity of the test results. At the same time, the quantitative adjustment calculation mechanism avoids the problems of over-correction or under-correction caused by blind adjustment, further improving the stability and reliability of real-time fine-tuning.

[0124] In a specific implementation scenario of this application, the test task scheduling method of this application can be implemented based on the intelligent peak-shaving and energy-saving scheduling system shown in Figure 2. The system consists of five core modules: ① test program parsing module, ② scheduling optimization engine, ③ real-time power consumption detection module, ④ control command issuance module, and ⑤ test task queuing module; each module is connected to multiple test cabinets (test cabinet 1 to test cabinet N) through a local energy bus (microgrid) to form a complete closed-loop scheduling architecture.

[0125] The test program parsing module is responsible for reading the test program files submitted by each test cabinet for the incoming test tasks. It parses the test process stages, power parameters, duration, and adjustable idle time boundaries. Based on this information, the module generates predicted charge / discharge power data for each test task, clearly marking the time-series distribution of power consumption, feedback, and idle time during the test, and synchronizing the data to the scheduling optimization engine. For example, after parsing a power battery cycle life test program, it generates a power time-series curve containing the entire cycle of "charging power consumption, idle time, discharge feedback, and idle time," with two idle stages marked as adjustable idle time periods, the duration of which can be adjusted within the range of 25-35 minutes.

[0126] Scheduling optimization engine. As the core computing unit of the system, the scheduling optimization engine receives prediction data from the test program parsing module and, with the goal of reducing the peak net local power, solves for the optimal combination of startup times for all tasks to be run.

[0127] In the initial batch scheduling scenario, the engine maps the power curves of all tasks to be run to the same time axis. By simulating and superimposing different combinations of startup times, it selects the startup sequence that can maximize peak shaving and valley filling. In dynamic incremental access scenarios, the engine combines the remaining power curve of running tasks to optimize the access time of newly submitted tasks, avoiding the random start of new tasks from disrupting the existing power balance.

[0128] For example, when eight power battery test tasks are submitted simultaneously, the engine optimizes the power peak of 640kW that would normally start simultaneously to less than 200kW by staggering the start times, and sends the optimal start time combination to the test task queuing module.

[0129] The test task queuing module and the control command issuing module are as follows: The test task queuing module receives the start time combination output by the scheduling optimization engine and sets the start time for each test task; the control command issuing module issues start signals to the corresponding test cabinets according to the timed trigger commands, controlling each task to start and run according to the planned sequence.

[0130] Real-time power consumption monitoring module and scheduling optimization engine. During test operation, the real-time power consumption monitoring module collects real-time power data from each test cabinet via the local energy bus, calculates the actual total net power of the local area, and synchronizes it to the scheduling optimization engine. The scheduling optimization engine simultaneously calculates the expected total net power based on the predicted data of each running task, and compares the deviation between the two: When the deviation exceeds the first trigger threshold, the engine identifies the test task that is currently in an adjustable idle period, calculates the required duration adjustment, and verifies whether the adjustment range is within the allowable boundary. Subsequently, an idle time adjustment command is generated and sent to the corresponding test cabinet through the control command distribution module to modify the remaining time of the static phase and correct the power deviation.

[0131] For example, when the actual total net power is 18kW higher than expected, the engine identifies 4 test tasks that are in the idle stage. By extending the idle time of each task, the excess power consumption is offset, and the local power returns to the expected range. Moreover, the adjustment process does not affect the core charging and discharging parameters of the test cabinet, ensuring the validity of the test results.

[0132] Through the coordinated operation of the above modules, the intelligent peak-shaving and energy-saving dispatching system can achieve closed-loop management of "pre-planning and dynamic correction during the process". Without adding new energy storage hardware or affecting the test quality, it can effectively reduce the peak power of the local power grid, improve energy utilization efficiency, and reduce the pressure on power distribution facilities and operating costs.

[0133] It should be noted that the scheduling method for test tasks involved in the above embodiments may include at least one of embodiments 1 to 8. For example, embodiment 1 may be implemented as an independent embodiment, embodiment 1+2 may be implemented as an independent embodiment, embodiment 1+2+3 may be implemented as an independent embodiment, embodiment 1+2+4 may be implemented as an independent embodiment, embodiment 1+2+3+4 may be implemented as an independent embodiment, embodiment 1+2+4+5 may be implemented as an independent embodiment, embodiment 1+2+3+4+5 may be implemented as an independent embodiment, embodiment 1+2+4+6 may be implemented as an independent embodiment, embodiment 1+2+3+4+6 may be implemented as an independent embodiment, embodiment 1+2+4+5+6 may be implemented as an independent embodiment, embodiment 1+2+3+4+5+6 may be implemented as an independent embodiment, etc., but it is not limited to these, and will not be listed in detail here.

[0134] It should be noted that the test task scheduling method provided in this application embodiment can be executed by a test task scheduling device, or a control module in the test task scheduling device for executing the method of loading test tasks. This application embodiment uses the method of loading test tasks executed by a test task scheduling device as an example to illustrate the test task scheduling method provided in this application embodiment.

[0135] Reference Figure 3 The diagram illustrates a structural block diagram of a test task scheduling device according to another embodiment of the present invention, which may specifically include: The prediction data determination module 201 is used to determine the prediction data of the change of charging and discharging power over time during the execution of the test task based on the task information of the test task to be run; wherein the test process of the test task includes adjustable idle periods. The start timing control module 202 is used to control the start timing of the test task to be run based on the predicted data; The remaining duration adjustment module 203 is used to adjust the remaining duration of the adjustable idle period of the test task in the adjustable idle period according to the deviation between the actual total net power and the expected total net power of each test task currently running; wherein the expected total net power is determined based on the corresponding prediction data of each test task currently running.

[0136] Optionally, the prediction data determination module includes: The information determination submodule is used to divide the test process into different functional time periods based on the task information, and determine the power attributes and expected duration of each functional time period. The data generation submodule is used to generate the prediction data based on the power attributes and expected duration of each functional time period, and to mark the location of the adjustable idle time period and its adjustable duration boundary in the prediction data.

[0137] Optionally, the startup timing control module includes: The first time combination determination submodule is used to determine the start time combination of the test tasks to be run based on the predicted data, with the optimization objective of reducing the peak value of the total net power. The startup control submodule is used to control the startup of the test task to be run according to the startup time combination.

[0138] Optionally, when determining the combination of start times for the test tasks to be run, at least one of the following is also included: Optimization sub-objective: Reduce the cumulative amount of net electrical energy output to the external power grid during the period representing net feedback in the total net power; Constraints: Ensure that the startup time of each test task falls within the allowed startup time window; Constraints: Ensure that the waiting time for each test task to start does not exceed the allowed upper limit.

[0139] Optionally, the test task to be run includes newly submitted test tasks, and the apparatus further includes: An add module is used to obtain the prediction data corresponding to the newly submitted test task and add it to the set of test tasks to be run; The startup timing control module includes: The second time combination determination submodule, based on the prediction data of each running test task and the prediction data of the set of test tasks to be run, determines the start time combination of the test tasks to be run in the set of test tasks to be run with the optimization objective of reducing the peak value of the total net power.

[0140] Optionally, the remaining duration adjustment module includes: The task identification submodule is used to identify test tasks in the adjustable idle period when the absolute value of the deviation between the actual total net power and the expected total net power exceeds the first trigger threshold. The duration adjustment submodule is used to adjust the remaining duration of the adjustable idle period of the identified test task based on the deviation amount.

[0141] Optionally, the device further includes: The timing update module is used to update the start timing of subsequent test tasks based on the prediction data when the absolute value of the deviation between the actual total net power and the expected total net power exceeds a second trigger threshold; wherein the second trigger threshold is higher than the first trigger threshold.

[0142] Optionally, the remaining duration adjustment module includes: The task selection submodule is used to select test tasks that are in the adjustable idle period as target tasks; The adjustment amount calculation submodule is used to calculate the duration adjustment amount of the adjustable idle period of the target task based on the deviation amount; The duration correction submodule is used to correct the remaining duration based on the duration adjustment amount and the allowable adjustment boundary of the adjustable idle time period of the target task.

[0143] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0144] Figure 4 This is a structural block diagram illustrating an electronic device 300 for scheduling test tasks according to an exemplary embodiment. For example, the electronic device 300 may be a computer.

[0145] Reference Figure 4 The electronic device 300 may include one or more of the following components: processing component 302, memory 304, power supply component 306, multimedia component 308, audio component 310, input / output (I / O) interface 312, sensor component 314, and communication component 316.

[0146] Processing component 302 typically controls the overall operation of electronic device 300, such as operations associated with display, telephone calls, data communication, camera operation, and recording. Processing component 302 may include one or more processors 320 to execute instructions to complete all or part of the steps of the methods described above. Furthermore, processing component 302 may include one or more modules to facilitate interaction between processing component 302 and other components. For example, processing component 302 may include a multimedia module to facilitate interaction between multimedia component 308 and processing component 302.

[0147] Memory 304 is configured to store various types of data to support the operation of device 300. Examples of this data include instruction messages, images, videos, etc., for any application or method operating on electronic device 300. Memory 304 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0148] Power supply component 306 provides power to various components of electronic device 300. Power supply component 306 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 300.

[0149] Multimedia component 308 includes a screen that provides an output interface between the electronic device 300 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation.

[0150] Audio component 310 is configured to output and / or input audio signals. For example, audio component 310 includes a microphone (MIC) configured to receive external audio signals when electronic device 300 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 304 or transmitted via communication component 316. In some embodiments, audio component 310 also includes a speaker for outputting audio signals.

[0151] I / O interface 312 provides an interface between processing component 302 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0152] Sensor assembly 314 includes one or more sensors for providing state assessments of various aspects of electronic device 300. For example, sensor assembly 314 may detect the on / off state of device 300, the relative positioning of components such as the display and keypad of electronic device 300, changes in position of electronic device 300 or a component of electronic device 300, the presence or absence of user contact with electronic device 300, orientation or acceleration / deceleration of electronic device 300, and temperature changes of electronic device 300. Sensor assembly 314 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 314 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 314 may also include an accelerometer, gyroscope, magnetometer, pressure sensor, or temperature sensor.

[0153] Communication component 316 is configured to facilitate wired or wireless communication between electronic device 300 and other devices. Electronic device 300 can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 316 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 316 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0154] In an exemplary embodiment, the electronic device 300 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0155] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 304 including instructions, which can be executed by a processor 320 of an electronic device 300 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0156] A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a terminal's processor, enables the terminal to perform a scheduling method for a test task, the method comprising: Based on the task information of the test task to be run, predictive data of the change of charging and discharging power over time during the execution of the test task is determined; wherein, the test process of the test task includes adjustable idle periods. Based on the predicted data, the timing of starting the test task to be run is controlled; Based on the deviation between the actual total net power and the expected total net power of each running test task, the remaining duration of the adjustable idle period for test tasks in the adjustable idle period is adjusted; wherein the expected total net power is determined based on the corresponding prediction data of each running test task.

[0157] Optionally, determining the predicted data of the change in charging and discharging power over time during the execution of the test task based on the task information of the test task to be run includes: Based on the task information, the test process is divided into different functional time periods, and the power attributes and expected duration of each functional time period are determined. Based on the power attributes and expected duration of each functional period, the prediction data is generated, and the location of the adjustable idle period and its adjustable duration boundary are marked in the prediction data.

[0158] Optionally, controlling the start timing of the test task to be run based on the predicted data includes: Based on the predicted data, with the optimization objective of reducing the peak value of total net power, the combination of start times for the test tasks to be run is determined. The test task to be run is started according to the aforementioned startup time combination.

[0159] Optionally, when determining the combination of start times for the test tasks to be run, at least one of the following is also included: Optimization sub-objective: Reduce the cumulative amount of net electrical energy output to the external power grid during the period representing net feedback in the total net power; Constraints: Ensure that the startup time of each test task falls within the allowed startup time window; Constraints: Ensure that the waiting time for each test task to start does not exceed the allowed upper limit.

[0160] Optionally, the test task to be run includes newly submitted test tasks, and the method further includes: Obtain the prediction data corresponding to the newly submitted test task and add it to the set of test tasks to be run; The step of controlling the start timing of the test task to be run based on the predicted data includes: Based on the predicted data of each running test task and the predicted data of the set of test tasks to be run, the start time combination of the test tasks to be run in the set of test tasks to be run is determined with the optimization objective of reducing the peak value of the total net power.

[0161] Optionally, adjusting the remaining duration of the adjustable idle period for test tasks in the adjustable idle period based on the deviation between the actual total net power and the expected total net power of each running test task includes: When the absolute value of the deviation between the actual total net power and the expected total net power exceeds the first trigger threshold, the test task in the adjustable idle period is identified. Based on the deviation, the remaining duration of the adjustable idle period of the identified test task is adjusted.

[0162] Optionally, the method further includes: When the absolute value of the deviation between the actual total net power and the expected total net power exceeds the second trigger threshold, the start timing of subsequent test tasks to be run is updated based on the predicted data; wherein the second trigger threshold is higher than the first trigger threshold.

[0163] Optionally, adjusting the remaining duration of the adjustable idle period of the identified test task based on the deviation includes: Select the test task that is in the adjustable idle period as the target task; Based on the deviation, calculate the adjustment amount of the adjustable idle time period for the target task; The remaining duration is adjusted based on the duration adjustment amount and the allowable adjustment boundary of the adjustable idle time period of the target task.

[0164] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts between the various embodiments, please refer to each other.

[0165] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0166] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0167] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0169] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments, as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0170] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0171] The present invention has provided a detailed description of a test task scheduling method, a test task scheduling device, an electronic device, and a readable storage medium. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for scheduling test tasks, characterized in that, The method includes: Based on the task information of the test task to be run, predictive data of the change of charging and discharging power over time during the execution of the test task is determined; wherein, the test process of the test task includes adjustable idle periods. Based on the predicted data, the timing of starting the test task to be run is controlled; Based on the deviation between the actual total net power and the expected total net power of each running test task, the remaining duration of the adjustable idle period for test tasks in the adjustable idle period is adjusted; wherein the expected total net power is determined based on the corresponding prediction data of each running test task.

2. The method according to claim 1, characterized in that, The method of determining predicted data on the change of charging and discharging power over time during the execution of the test task based on the task information of the test task to be run includes: Based on the task information, the test process is divided into different functional time periods, and the power attributes and expected duration of each functional time period are determined. Based on the power attributes and expected duration of each functional period, the prediction data is generated, and the location of the adjustable idle period and its adjustable duration boundary are marked in the prediction data.

3. The method according to claim 1 or 2, characterized in that, The step of controlling the start timing of the test task to be run based on the predicted data includes: Based on the predicted data, with the optimization objective of reducing the peak value of total net power, the combination of start times for the test tasks to be run is determined. The test task to be run is started according to the aforementioned startup time combination.

4. The method according to claim 3, characterized in that, When determining the combination of start times for the test tasks to be run, at least one of the following is also included: Optimization sub-objective: Reduce the cumulative amount of net electrical energy output to the external power grid during the period representing net feedback in the total net power; Constraints: Ensure that the startup time of each test task falls within the allowed startup time window; Constraints: Ensure that the waiting time for each test task to start does not exceed the allowed upper limit.

5. The method according to any one of claims 1-4, characterized in that, The test tasks to be run include newly submitted test tasks, and the method further includes: Obtain the prediction data corresponding to the newly submitted test task and add it to the set of test tasks to be run; The step of controlling the start timing of the test task to be run based on the predicted data includes: Based on the predicted data of each running test task and the predicted data of the set of test tasks to be run, the start time combination of the test tasks to be run in the set of test tasks to be run is determined with the optimization objective of reducing the peak value of the total net power.

6. The method according to any one of claims 1-5, characterized in that, The step of adjusting the remaining duration of the adjustable idle period for test tasks in the adjustable idle period based on the deviation between the actual total net power and the expected total net power of each running test task includes: When the absolute value of the deviation between the actual total net power and the expected total net power exceeds the first trigger threshold, the test task in the adjustable idle period is identified. Based on the deviation, the remaining duration of the adjustable idle period of the identified test task is adjusted.

7. The method according to claim 6, characterized in that, The method further includes: When the absolute value of the deviation between the actual total net power and the expected total net power exceeds the second trigger threshold, the start timing of subsequent test tasks to be run is updated based on the predicted data; wherein the second trigger threshold is higher than the first trigger threshold.

8. The method according to claim 6 or 7, characterized in that, The step of adjusting the remaining duration of the adjustable idle period of the identified test task based on the deviation includes: Select the test task that is in the adjustable idle period as the target task; Based on the deviation, calculate the adjustment amount of the adjustable idle time period for the target task; The remaining duration is adjusted based on the duration adjustment amount and the allowable adjustment boundary of the adjustable idle time period of the target task.

9. A scheduling device for test tasks, characterized in that, The device includes: The prediction data determination module is used to determine the prediction data of the change of charging and discharging power over time during the execution of the test task based on the task information of the test task to be run; wherein the test process of the test task includes adjustable idle periods. The start-up timing control module is used to control the start-up timing of the test task to be run based on the predicted data; The remaining duration adjustment module is used to adjust the remaining duration of the adjustable idle period of the test task in the adjustable idle period according to the deviation between the actual total net power and the expected total net power of each running test task; wherein the expected total net power is determined based on the corresponding prediction data of each running test task.

10. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the scheduling method for the test task as described in any one of claims 1-7.

11. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the scheduling method for the test task as described in any one of claims 1-7.

Images