Resource scheduling method, device and equipment for artificial intelligence teaching platform

Abstract

The application relates to the field of resource scheduling, in particular to a resource scheduling method, device and equipment for an artificial intelligence teaching platform, which comprises the following steps: obtaining a target training scene selected by a user; determining a target function module set based on the target training scene; obtaining the current state of each target function module in the target function module set; in the case that at least one target function module is in a non-idle state, determining the predicted idle time of the target function module; comparing the predicted idle time with a preset waiting time to obtain a comparison result; and determining a scheduling strategy based on the comparison result, wherein the scheduling strategy comprises at least one of waiting execution, executing a replacement training scene and reserving the target function module. Through the dynamic resource scheduling mechanism, the state and idle time of the function module are monitored and predicted in real time, resource allocation under multi-task concurrency is optimized, and therefore the resource utilization rate of the teaching platform is improved and the learning experience is improved.

Description

Technical Field

[0001] This application relates to the field of resource scheduling, and in particular to a resource scheduling method, apparatus and equipment for an artificial intelligence teaching platform. Background Technology

[0002] Existing teaching platforms typically adopt a modular and integrated design concept, integrating multiple functional modules such as visual recognition, voice interaction, robotic arm control, and sensor networks to provide a unified experimental and training environment for multiple disciplines such as computer science, automation, and mechatronics engineering.

[0003] However, modular integration platforms still have significant shortcomings in practical teaching applications, specifically in the resource scheduling of functional modules. On the one hand, high-performance functional modules in the platform are usually expensive and have a limited number of deployments. On the other hand, when students conduct experiments, they need to call the corresponding target functional modules based on specific training scenarios, such as robotic arm object grasping experiments and multi-sensor environmental monitoring experiments. When multiple students or experimental tasks request to use the same target functional module at the same time within a similar time period, the existing platform lacks an effective and intelligent resource scheduling mechanism.

[0004] Current scheduling methods are mostly limited to simple first-come, first-served rules or static scheduling tables, which have the following problems: First, they cannot obtain and provide feedback on the dynamic status of each target functional module in real time, making it difficult for students to judge the availability of functional modules in advance; Second, when a target functional module is not idle, its preset idle time cannot be accurately determined. If a functional module is occupied after a student submits an experiment request, the student can only wait passively without knowing the specific waiting time, which seriously disrupts the continuity of learning, further reduces the resource utilization rate of the teaching platform and the students' practical experience, and makes it difficult to meet the needs of efficient and flexible practical teaching. Summary of the Invention

[0005] This application provides a resource scheduling method, apparatus, and device for an artificial intelligence teaching platform. It can monitor and predict the status and idle time of functional modules in real time through an intelligent dynamic resource scheduling mechanism, optimize resource allocation under multi-task concurrency, thereby significantly improving the resource utilization rate of the teaching platform, ensuring the continuity of student experiments, and enhancing the learning experience.

[0006] On one hand, embodiments of this application provide a resource scheduling method for an artificial intelligence teaching platform, including:

[0007] Obtain the target training scenario selected by the user;

[0008] Determine the set of target functional modules based on the target training scenario;

[0009] Obtain the current state of each target functional module in the target functional module set, wherein the current state includes idle, busy, and paused use;

[0010] If at least one of the target functional modules is in a non-idle state, determine the predicted idle time of the target functional module;

[0011] The predicted idle time is compared with the preset waiting time to obtain the comparison result;

[0012] A scheduling strategy is determined based on the comparison results, wherein the scheduling strategy includes at least one of waiting to execute, executing an alternative training scenario, and reserving the target functional module.

[0013] Optionally, determining the predicted idle time of the target functional module includes:

[0014] Obtain the current training scenario to which the target functional module belongs;

[0015] Obtain historical training data for the current training scenario;

[0016] Calculate the average training time for the current training scenario based on the historical training data;

[0017] Obtain the training users and execution progress corresponding to the current training scenario;

[0018] Obtain the average completion time of each training scenario for the training users;

[0019] The predicted idle time is calculated based on the average completion time, average training time, and progress.

[0020] Optionally, calculating the predicted idle time based on the average completion time, average training time, and progress includes:

[0021] Calculate the time difference between the average training time and the corresponding average completion time for each training scenario;

[0022] The stability coefficient of the training user is determined based on multiple time differences.

[0023] Based on the stability coefficient, the average training time and average completion time are calculated to obtain the predicted completion time of the training user;

[0024] The predicted idle time is calculated based on the predicted completion time and the executed progress.

[0025] Optionally, determining the stability coefficient of the training user based on multiple time differences includes:

[0026] Calculate the standard deviation of the multiple time differences and compare the standard deviation with a preset stability threshold;

[0027] If the standard deviation is less than the preset stability threshold, the stability coefficient of the training user is determined to be a first preset value;

[0028] If the standard deviation is not less than the preset stability threshold, the stability coefficient of the training user is determined to be a second preset value;

[0029] Wherein, the first preset value is greater than the second preset value.

[0030] Optionally, the calculation of the average training time and average completion time based on the stability coefficient includes:

[0031] The predicted completion time = λ × average training time + (1-λ) × average completion time, where λ is an adjustment coefficient, the value of λ is in the range of [0,1], and the value of λ is determined based on the mapping relationship between λ and the stability coefficient.

[0032] Optionally, determining the scheduling strategy based on the comparison result includes:

[0033] If the predicted idle time is not greater than the preset waiting time, a first scheduling strategy is generated that includes waiting execution options and waiting time.

[0034] If the predicted idle time is greater than the preset waiting time, a second scheduling strategy is generated that includes recommended alternative training scenarios and reservation options.

[0035] Display the first scheduling strategy or the second scheduling strategy.

[0036] Optionally, generate recommended alternative training scenarios, including:

[0037] Select all functional modules that are in an idle state;

[0038] The functional modules based on the idle state determine the training scenario to be used;

[0039] Calculate the similarity between the training scenario to be used and the target training scenario in terms of teaching objectives and skill requirements;

[0040] Based on the similarity, a preset number of training scenarios are selected as alternative training scenarios.

[0041] Optional, generate appointment options, including:

[0042] Generate reservation information for the target training scenario, the reservation information including a set of target functional modules, reservation priority, and reservation validity period;

[0043] Add the appointment information to the appointment list;

[0044] When all functional modules in the target functional module set become idle, the execution order of the reservation information is determined based on the reservation priority and reservation time.

[0045] Based on the execution order, a reservation readiness notification is sent to the corresponding user, and within a preset time after receiving the user's confirmation instruction, the corresponding functional module is locked and the target training scenario is started.

[0046] On the other hand, embodiments of this application provide a resource scheduling device for an artificial intelligence teaching platform, the device comprising:

[0047] The first acquisition module is used to acquire the target training scenario selected by the user.

[0048] The first determining module is used to determine a set of target functional modules based on the target training scenario;

[0049] The second acquisition module is used to acquire the current state of each target functional module in the target functional module set, wherein the current state includes idle, busy, and paused use;

[0050] The second determining module is used to determine the predicted idle time of the target functional module when at least one of the target functional modules is in a non-idle state.

[0051] The comparison module is used to compare the predicted idle time with the preset waiting time to obtain a comparison result;

[0052] The third determining module is used to determine a scheduling strategy based on the comparison result, wherein the scheduling strategy includes at least one of waiting to execute, executing an alternative training scenario, and reserving the target functional module.

[0053] In another aspect, embodiments of this application provide an electronic device, the device including: a processor and a memory storing computer program instructions;

[0054] When the processor executes the computer program instructions, it implements a resource scheduling method for an artificial intelligence teaching platform as described in any one of the first aspects.

[0055] This application discloses a resource scheduling method, apparatus, and device for an artificial intelligence teaching platform. By acquiring the current status of target functional modules in real time and combining this with the calculation of predicted idle time, the method can dynamically assess resource availability, avoiding resource idleness or conflicts caused by static scheduling or first-come-first-served rules in existing technologies. By comparing the predicted idle time with preset waiting times, the method can intelligently recommend the optimal scheduling strategy, thereby optimizing the usage order and time allocation of functional modules, reducing idle time, significantly improving the overall resource utilization of the teaching platform, and providing clear feedback on predicted idle time. This allows students to understand the availability status of target functional modules and the expected waiting time in advance, avoiding the problems of passive waiting and inability to plan experiment time in existing technologies. Through a scheduling strategy based on the comparison results, students can choose to execute alternative training scenarios when the waiting time is long, ensuring that practical learning is not interrupted and maintaining learning continuity. This helps improve learning efficiency. Attached Figure Description

[0056] Figure 1 This is a flowchart illustrating a resource scheduling method for an artificial intelligence teaching platform in an embodiment of this application;

[0057] Figure 2 This is a structural block diagram illustrating a resource scheduling device for an artificial intelligence teaching platform in the embodiments of this application;

[0058] Figure 3 This is a structural block diagram illustrating an electronic device in the embodiments of this application; Detailed Implementation

[0059] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0060] It should be noted that, in this document, relational terms such as "first" and "second" are used merely 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 apparatus 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 apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0061] Existing teaching platforms typically employ a modular design, integrating multiple functional modules such as visual recognition, voice interaction, and robotic arm control to provide a unified experimental environment for multiple disciplines. However, when multiple students request the same functional module within a similar timeframe, existing scheduling mechanisms are limited to first-come, first-served rules or static scheduling tables, failing to provide real-time feedback on the module's dynamic status. This results in students passively waiting without predicting the waiting time, severely reducing resource utilization and learning continuity. For example, in a robotic arm object grasping experiment, if the target robotic arm module is occupied, students can only wait or repeatedly check its status, causing interruptions in the learning process and resource idleness.

[0062] To address the problems of existing technologies, embodiments of this application provide a resource scheduling method, apparatus, and device for an artificial intelligence teaching platform. The resource scheduling method for an artificial intelligence teaching platform provided by embodiments of this application will be described below first.

[0063] Figure 1 This illustration shows a flowchart of a resource scheduling method for an artificial intelligence teaching platform according to an embodiment of this application. Figure 1 As shown in the figure, a resource scheduling method for an artificial intelligence teaching platform provided in this application includes the following steps:

[0064] S101, Obtain the target training scenario selected by the user;

[0065] In this embodiment, the artificial intelligence teaching platform is first introduced. This platform includes, but is not limited to, a collaborative robotic arm, a 3D depth camera, a material sorting line, a vision / speech model, a sound pickup unit, an intelligent sorting module, an electrically controlled vehicle model, a smart home module, and an instrument unit. The collaborative robotic arm adopts an integrated design, making full use of production space and seamlessly integrating into the actual environment. It also features collision avoidance detection based on a precise dynamic model, enabling safe collaborative operation with humans. The 3D depth camera is the device for implementing artificial intelligence vision technology. Several material recognition modules serve as materials for the application of vision technology, providing support for related technical practices. The vision / speech model provides pre-trained recognition models. The sound pickup unit serves as a voice input / output carrier, ensuring smooth voice interaction. The sorting module is the application carrier of artificial intelligence vision technology, facilitating the realization of material sorting scenarios. The electrically controlled vehicle model serves as an application carrier for fatigue driving scenarios, simulating relevant scenarios for technical verification. The smart home module is the application carrier of artificial intelligence voice technology, promoting the implementation of voice control in home scenarios. The instrument unit serves as an application carrier for intelligent meter reading based on AI vision, providing a practical platform for intelligent meter reading technology.

[0066] Training scenarios include, but are not limited to, smart home, handwriting recognition, fatigue detection, material sorting, and visual intelligent meter reading. In this embodiment, the collaborative robotic arm, 3D depth camera, material sorting line, visual / voice model, sound pickup unit, intelligent sorting module, electric control vehicle model, smart home module, and instrument unit are used as functional modules.

[0067] In this embodiment, the user selects the training scenario they want to conduct practical training using input devices such as a mouse and keyboard on the electronic device. When the electronic device obtains the training scenario selected by the user, it uses the corresponding training scenario as the target training scenario.

[0068] S102, Determine the set of target functional modules based on the target training scenario.

[0069] When a target training scenario is acquired, the electronic device queries a pre-stored training scenario-functional module mapping table to determine the set of target functional modules required to complete the target training scenario. The set of target functional modules is the combination of functional modules required to complete the target training scenario. The electronic device stores a training scenario-functional module mapping table, which is pre-set according to the functional model required by the target scenario.

[0070] For example, material sorting scenarios require collaborative robotic arms, 3D depth cameras, and intelligent sorting modules; smart home scenarios require voice interaction modules (including amplified sound pickup units and visual / voice models) and smart home modules; and fatigue detection scenarios require electric vehicle models and visual models.

[0071] S103, obtain the current state of each target functional module in the target functional module set, where the current state includes idle, busy, and paused use.

[0072] In this embodiment, the current state of the target functional module can be obtained through the state monitoring interface of the target functional module. For physical device functional modules such as collaborative robotic arms and 3D depth cameras, a connection is established with the device controller through the Internet of Things protocol. Each functional module is equipped with a state sensor and periodically sends heartbeat signals and operating data to the electronic device. For software functional modules such as vision / speech models, the electronic device obtains the operating status through the API interface.

[0073] In other alternative embodiments, the determination can also be made through user interaction data with the AI ​​teaching platform. For example, when there is a material sorting training scenario being executed in the AI ​​teaching platform, the status of the associated collaborative robotic arm, 3D depth camera, and sorting module is marked as busy.

[0074] Each functional module has a usage time interval. When the time interval is not reached, it is marked as paused and the pause time is displayed. The time interval can be 1 minute or 5 minutes, without specific limitation.

[0075] S104, when at least one target functional module is in a non-idle state, determine the predicted idle time of the target functional module.

[0076] S104 specifically includes: obtaining the current training scenario to which the target functional module belongs; obtaining historical training data of the current training scenario; calculating the average training time of the current training scenario based on the historical training data; obtaining the training users and execution progress corresponding to the current training scenario; obtaining the average completion time of each training scenario for the training users; and calculating the predicted idle time based on the average completion time, average training time, and execution progress.

[0077] The process of calculating predicted idle time based on average completion time, average training time, and progress includes: calculating the time difference between the average training time and the corresponding average completion time for each training scenario; determining the stability coefficient of the training user based on multiple time differences; calculating the predicted completion time of the training user based on the stability coefficient and the average training time and average completion time; and calculating the predicted idle time based on the predicted completion time and progress.

[0078] In this embodiment, determining the stability coefficient includes: calculating the standard deviation of multiple time differences and comparing the standard deviation with a preset stability threshold; if the standard deviation is less than the preset stability threshold, determining the stability coefficient of the training user as a first preset value; if the standard deviation is not less than the preset stability threshold, determining the stability coefficient of the training user as a second preset value; wherein the first preset value is greater than the second preset value.

[0079] The average training time and average completion time are calculated based on the stability coefficient, including: predicted completion time = λ × average training time + (1-λ) × average completion time, where λ is the adjustment coefficient, the value of λ is in the range of [0,1], and the value of λ is determined based on the mapping relationship between λ and the stability coefficient.

[0080] In this embodiment, when the target functional module is in a non-idle state, the task scheduling data of the artificial intelligence teaching platform is queried to obtain the training task information currently occupying the target functional module. The training task information includes the task ID, the corresponding training scenario, i.e. the current training scenario. At this time, the electronic device obtains the relevant information of the current training scenario. The relevant information includes, but is not limited to, scenario type, such as material sorting, scenario complexity level such as primary, intermediate or advanced, and expected standard training duration range.

[0081] After obtaining the current training scenario, query the historical training data of the current training scenario in the historical training database, select the historical training time in the historical training data, and calculate the average training time of all historical training times. Obtain the current progress through the task scheduling table, such as the percentage of running time to the total task or the percentage of completed operations to the total operations.

[0082] To improve the accuracy of predicting idle time, historical data from training users is obtained. First, the average completion time for each training scenario is calculated. Then, the time difference between the average completion time and the average training time for the corresponding training scenario is calculated. Finally, the standard deviation is calculated using all time differences. A low standard deviation indicates that the user's completion time in different scenarios is very close to the global average level for each scenario, and their behavior pattern is stable and predictable. A high standard deviation indicates that the training user may be much faster than the average level in some scenarios and much slower than the average level in others, and their behavior pattern fluctuates greatly and is highly unpredictable.

[0083] For example, to obtain data on all training scenarios completed by the training users, the following two data points are obtained for each training scenario completed by the training user: the average completion time of the training user: the average time taken by the training user to complete training scenario X multiple times; if it is completed only once, then the time taken for that single completion is used; the average training time of the training scenario: the average time taken by all users to complete training scenario X. The above data is updated in real time.

[0084] Next, the time difference between the average completion time and the average training time of the training scenarios completed by the training users is calculated. This time difference is the absolute deviation between the individual performance of the user and the global average level. All historical records of the training users are obtained to obtain a set of time difference sequences: {ΔT1,ΔT2,...,ΔTN}, where N is the number of different training scenarios completed by the training. The standard deviation of the time difference sequence is calculated.

[0085] In this embodiment, the stability of training users is classified by a stability threshold. The standard deviation is compared with a preset threshold, and the stability coefficient is determined based on the comparison result. The preset stability threshold is used to judge the critical value of the stability of user training behavior. It can be set to a fixed value or dynamically adjusted interval value by analyzing the standard deviation distribution characteristics in historical data, and is used to distinguish between high-stability users and low-stability users.

[0086] If the standard deviation is less than the stability threshold, the training user is determined to be a stable user and assigned a higher first preset value, such as λ=0.8. If the standard deviation is greater than or equal to the stability threshold, the training user is determined to be a fluctuating user and assigned a lower second preset value, such as λ=0.4. The determined λ is input into the prediction completion time = λ × average training time + (1-λ) × average completion time to obtain the predicted idle time of the training user.

[0087] It should be noted that the stability coefficient is positively correlated with λ, and the mapping relationship between the stability coefficient and λ is predetermined.

[0088] S105, compare the predicted idle time with the preset waiting time to obtain the comparison result.

[0089] In this embodiment, the preset waiting time is the maximum acceptable waiting time threshold for the user, which can be determined by user settings or preset values. The comparison result includes being greater than the preset waiting time or not greater than the preset waiting time.

[0090] S106, determine the scheduling strategy based on the comparison results, wherein the scheduling strategy includes at least one of waiting to execute, executing alternative training scenarios, or reserving target functional modules.

[0091] S106 specifically includes: generating a first scheduling strategy that includes a waiting execution option and a waiting time when the predicted idle time is not greater than a preset waiting time; generating a second scheduling strategy that includes a recommended alternative training scenario and a reservation option when the predicted idle time is greater than the preset waiting time; and displaying the first scheduling strategy or the second scheduling strategy.

[0092] The process of generating recommended alternative training scenarios includes: selecting all functional modules that are in an idle state; determining training scenarios to be used based on the functional modules in the idle state; calculating the similarity between the training scenarios to be used and the target training scenario in terms of teaching objectives and skill requirements; and selecting a preset number of training scenarios to be used as alternative training scenarios based on the similarity.

[0093] The process of generating reservation options includes: generating reservation information for the target training scenario, which includes a set of target functional modules, reservation priority, and reservation validity period; adding the reservation information to the reservation list; determining the execution order of the reservation information based on the reservation priority and reservation time when all functional modules in the target functional module set become idle; sending a reservation ready notification to the corresponding user based on the execution order; and locking the corresponding functional module and starting the target training scenario within a preset time after receiving the user's confirmation instruction.

[0094] In this embodiment, if the predicted idle time is not greater than the preset waiting time, a waiting execution information with a countdown prompt is generated, and the user can choose to continue waiting or temporarily interrupt the operation according to their own plan; if the predicted idle time is greater than the preset waiting time, the idle functional module combination with the highest matching degree with the teaching objectives is selected, and the option of reserving the functional module is provided for the user to lock in future usage rights.

[0095] For the generation of alternative training scenarios: The training scenarios to be used are the set of training scenarios that the currently idle functional modules can support. This can be determined by querying the training scenario-functional module mapping table. The training scenario-functional module mapping table stores the training scenarios that each functional module can participate in and their associated teaching objectives. Similarity refers to the degree of matching between two training scenarios in terms of knowledge transfer and skill training. The vector space model can be used to vectorize the teaching objective text, and a weighted score can be calculated by combining the keyword matching algorithm of skill requirements.

[0096] In this embodiment, when the target functional module is not idle, the electronic device queries all idle functional modules, determines the training scenario to be used through the idle functional modules, extracts the teaching objective description text and skill requirement tags for each training scenario to be used, converts them into high-dimensional vectors through natural language processing technology, and calculates the cosine similarity with the vector of the target training scenario. At the same time, it matches the overlap ratio of skill keywords. Finally, the scenarios ranked in the top N by comprehensive similarity score are output as the recommendation result, ensuring that users can quickly select an alternative that is highly consistent with the original experimental objective when resources are limited.

[0097] For generating reservation options: Reservation information refers to the set of target functional modules, reservation priority, and reservation validity period. Reservation priority refers to the weight used to distinguish the processing order of different reservation requests, which can be implemented using a numerical scoring mechanism or classification tags. The numerical score can be calculated based on user identity, task urgency, or historical usage frequency. Reservation validity period refers to the time range during which the reservation request remains active, which can be implemented using a timestamp or countdown mechanism.

[0098] In this embodiment, when a user submits a request for a target training scenario, the electronic device generates reservation information containing the required set of functional modules, assigns priority and validity period parameters, and adds the reservation information to the reservation list. When it is detected that all functional modules involved in a certain reservation request have become idle, the electronic device determines the execution order of the request in the reservation list according to a preset sorting rule. For example, requests with higher priority can be processed first even if the reservation time is later, while requests with the same priority follow the first-come, first-served principle. If the user responds within a preset time, the relevant functional modules are locked and the training scenario is started; if no confirmation is received within the time limit, the functional modules are released and the reservation is marked as invalid.

[0099] Based on the resource scheduling method for an artificial intelligence teaching platform provided in the above embodiments, this application also provides a specific implementation of a resource scheduling device for an artificial intelligence teaching platform. Please refer to the following embodiments.

[0100] First see Figure 2 The resource scheduling device 200 for an artificial intelligence teaching platform provided in this application embodiment includes the following modules:

[0101] The first acquisition module 201 is used to acquire the target training scenario selected by the user;

[0102] The first determining module 202 is used to determine the set of target functional modules based on the target training scenario;

[0103] The second acquisition module 203 is used to acquire the current status of each target functional module in the target functional module set, wherein the current status includes idle, busy and paused use;

[0104] The second determining module 204 is used to determine the predicted idle time of a target functional module when at least one target functional module is in a non-idle state.

[0105] The comparison module 205 is used to compare the predicted idle time with the preset waiting time to obtain the comparison result;

[0106] The third determining module 206 is used to determine a scheduling strategy based on the comparison results, wherein the scheduling strategy includes at least one of waiting to execute, executing an alternative training scenario, or reserving a target function module.

[0107] As an optional implementation of this embodiment, the second determining module 204 includes:

[0108] The scene acquisition submodule is used to acquire the current training scene to which the target functional module belongs;

[0109] The data acquisition submodule is used to acquire historical training data for the current training scenario;

[0110] The first calculation submodule is used to calculate the average training time for the current training scenario based on historical training data.

[0111] The progress acquisition submodule is used to acquire the training users and execution progress corresponding to the current training scenario;

[0112] The time acquisition submodule is used to obtain the average completion time of each training scenario for the training users.

[0113] The second calculation submodule is used to calculate the predicted idle time based on the average completion time, average training time, and progress.

[0114] In this optional embodiment, the second computing submodule includes:

[0115] The first calculation unit is used to calculate the time difference between the average training time and the corresponding average completion time for each training scenario.

[0116] The determination unit is used to determine the stability coefficient of the training user based on multiple time differences;

[0117] The second calculation unit is used to calculate the average training time and average completion time based on the stability coefficient, so as to obtain the predicted completion time of the training user.

[0118] The third calculation unit is used to calculate the predicted idle time based on the predicted completion time and the progress already executed.

[0119] In this optional embodiment, the determining unit is specifically used for:

[0120] Calculate the standard deviation of multiple time differences and compare the standard deviation with a preset stability threshold; if the standard deviation is less than the preset stability threshold, determine the stability coefficient of the training user as the first preset value; if the standard deviation is not less than the preset stability threshold, determine the stability coefficient of the training user as the second preset value; wherein, the first preset value is greater than the second preset value.

[0121] In this optional embodiment, the second computing unit is specifically used for:

[0122] Predicted completion time = λ × average training time + (1-λ) × average completion time, where λ is an adjustment coefficient, the value of λ ranges from [0,1], and the value of λ is determined based on the mapping relationship between λ and the stability coefficient.

[0123] As an optional implementation of this embodiment, the third determining module 206 includes:

[0124] The first generation submodule is used to generate a first scheduling strategy that includes waiting execution options and waiting time, provided that the predicted idle time is not greater than the preset waiting time.

[0125] The second generation submodule is used to generate a second scheduling strategy that includes recommended alternative training scenarios and reservation options when the predicted idle time is greater than the preset waiting time.

[0126] The display submodule is used to display either the first scheduling strategy or the second scheduling strategy.

[0127] In this optional embodiment, the second generation submodule is specifically used for:

[0128] Select all functional modules that are in an idle state; determine the training scenarios to be used based on the functional modules in an idle state; calculate the similarity between the training scenarios to be used and the target training scenario in terms of teaching objectives and skill requirements; select a preset number of training scenarios to be used as alternative training scenarios based on the similarity.

[0129] In this optional embodiment, the second generation submodule is further specifically used for:

[0130] Generate reservation information for the target training scenario, including the target functional module set, reservation priority, and reservation validity period; add the reservation information to the reservation list; when all functional modules in the target functional module set become idle, determine the execution order of the reservation information based on the reservation priority and reservation time; send a reservation ready notification to the corresponding user based on the execution order, and lock the corresponding functional module and start the target training scenario within a preset time after receiving the user's confirmation instruction.

[0131] Figure 3 A schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application is shown.

[0132] An electronic device may include a processor 301 and a memory 302 storing computer program instructions.

[0133] Specifically, the processor 301 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0134] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. In one instance, memory 302 may include removable or non-removable (or fixed) media, or memory 302 may be non-volatile solid-state memory. Memory 302 may be internal or external to the integrated gateway disaster recovery device.

[0135] In one instance, memory 302 may be read-only memory (ROM). In one instance, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0136] Memory 302 may include read-only memory (ROM), random access memory (RAM), disk storage media device, optical storage media device, flash memory device, electrical, optical, or other physical / tangible memory storage device. Therefore, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to a resource scheduling method for an artificial intelligence teaching platform according to a first aspect of this disclosure.

[0137] The processor 301 reads and executes computer program instructions stored in the memory 302 to achieve... Figure 1 The illustrated embodiment presents a resource scheduling method for an artificial intelligence teaching platform.

[0138] In one example, the electronic device may also include a communication interface 303 and a bus 304. For example, Figure 3 As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 304 and complete communication with each other.

[0139] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0140] Bus 304 includes hardware, software, or both, that couples components of an electronic device together. For example, and not as a limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 304 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0141] The electronic device can execute a resource scheduling method for an artificial intelligence teaching platform according to an embodiment of this application, thereby achieving a combination of Figures 1-2 This invention describes a resource scheduling method and apparatus for an artificial intelligence teaching platform.

[0142] Furthermore, in conjunction with the resource scheduling method for an artificial intelligence teaching platform described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the resource scheduling methods for an artificial intelligence teaching platform described in the above embodiments.

[0143] In an optional embodiment, in conjunction with a resource scheduling method for an artificial intelligence teaching platform as described in the above embodiments, this application embodiment can provide a computer program product to implement the method. The instructions in the computer program product are executed by the processor of an electronic device, enabling the electronic device to implement any of the resource scheduling methods for an artificial intelligence teaching platform as described in the above embodiments.

[0144] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0145] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0146] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0147] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in 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, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0148] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A resource scheduling method for an artificial intelligence teaching platform, characterized in that, include: Obtain the target training scenario selected by the user; Determine the set of target functional modules based on the target training scenario; Obtain the current state of each target functional module in the target functional module set, wherein the current state includes idle, busy, and paused use; If at least one of the target functional modules is in a non-idle state, determine the predicted idle time of the target functional module; The predicted idle time is compared with the preset waiting time to obtain the comparison result; Based on the comparison results, a scheduling strategy is determined, wherein the scheduling strategy includes at least one of waiting to execute, executing an alternative training scenario, and reserving the target functional module; Determining the predicted idle time of the target functional module includes: Obtain the current training scenario to which the target functional module belongs; Obtain historical training data for the current training scenario; Calculate the average training time for the current training scenario based on the historical training data; Obtain the training users and execution progress corresponding to the current training scenario; Obtain the average completion time for each training scenario for the training users; The predicted idle time is calculated based on the average completion time, average training time, and progress. The calculation of the predicted idle time based on the average completion time, average training time, and progress includes: Calculate the time difference between the average training time and the corresponding average completion time for each training scenario; The stability coefficient of the training user is determined based on multiple time differences. Based on the stability coefficient, the average training time and average completion time are calculated to obtain the predicted completion time of the training user; The predicted idle time is calculated based on the predicted completion time and the executed progress.

2. The resource scheduling method for an artificial intelligence teaching platform according to claim 1, characterized in that: The determination of the stability coefficient of the training user based on multiple time differences includes: Calculate the standard deviation of the multiple time differences and compare the standard deviation with a preset stability threshold; If the standard deviation is less than the preset stability threshold, the stability coefficient of the training user is determined to be a first preset value; If the standard deviation is not less than the preset stability threshold, the stability coefficient of the training user is determined to be a second preset value; Wherein, the first preset value is greater than the second preset value.

3. The resource scheduling method for an artificial intelligence teaching platform according to claim 2, characterized in that: The calculation of the average training time and average completion time based on the stability coefficient includes: The predicted completion time = λ × average training time + × Average completion time, where λ is an adjustment coefficient, the value of λ ranges from [0,1], and the value of λ is determined based on the mapping relationship between λ and the stability coefficient.

4. The resource scheduling method for an artificial intelligence teaching platform according to claim 1, characterized in that: The step of determining the scheduling strategy based on the comparison result includes: If the predicted idle time is not greater than the preset waiting time, a first scheduling strategy is generated that includes waiting execution options and waiting time. If the predicted idle time is greater than the preset waiting time, a second scheduling strategy is generated that includes recommended alternative training scenarios and reservation options. Display the first scheduling strategy or the second scheduling strategy.

5. A resource scheduling method for an artificial intelligence teaching platform according to claim 4, characterized in that: Generate recommended alternative training scenarios, including: Select all functional modules that are in an idle state; The functional modules based on the idle state determine the training scenario to be used; Calculate the similarity between the training scenario to be used and the target training scenario in terms of teaching objectives and skill requirements; Based on the similarity, a preset number of training scenarios to be used are selected as alternative training scenarios.

6. The resource scheduling method for an artificial intelligence teaching platform according to claim 4, characterized in that: Generate appointment options, including: Generate reservation information for the target training scenario, the reservation information including a set of target functional modules, reservation priority, and reservation validity period; Add the appointment information to the appointment list; When all functional modules in the target functional module set become idle, the execution order of the reservation information is determined based on the reservation priority and reservation time. Based on the execution order, a reservation readiness notification is sent to the corresponding user, and within a preset time after receiving the user's confirmation instruction, the corresponding functional module is locked and the target training scenario is started.

7. A resource scheduling device for an artificial intelligence teaching platform, characterized in that, The device includes: The first acquisition module is used to acquire the target training scenario selected by the user. The first determining module is used to determine a set of target functional modules based on the target training scenario; The second acquisition module is used to acquire the current state of each target functional module in the target functional module set, wherein the current state includes idle, busy, and paused use; The second determining module is used to determine the predicted idle time of the target functional module when at least one of the target functional modules is in a non-idle state. The comparison module is used to compare the predicted idle time with the preset waiting time to obtain a comparison result; The third determining module is used to determine a scheduling strategy based on the comparison result, wherein the scheduling strategy includes at least one of waiting to execute, executing an alternative training scenario, and reserving the target functional module; The scene acquisition submodule is used to acquire the current training scene to which the target functional module belongs; The data acquisition submodule is used to acquire historical training data for the current training scenario; The first calculation submodule is used to calculate the average training time for the current training scenario based on historical training data. The progress acquisition submodule is used to acquire the training users and execution progress corresponding to the current training scenario; The time acquisition submodule is used to obtain the average completion time of each training scenario for the training users. The second calculation submodule is used to calculate the predicted idle time based on the average completion time, average training time and the progress already executed. The first calculation unit is used to calculate the time difference between the average training time and the corresponding average completion time for each training scenario. The determination unit is used to determine the stability coefficient of the training user based on multiple time differences; The second calculation unit is used to calculate the average training time and average completion time based on the stability coefficient, so as to obtain the predicted completion time of the training user. The third calculation unit is used to calculate the predicted idle time based on the predicted completion time and the progress already executed.

8. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements a resource scheduling method for an artificial intelligence teaching platform as described in any one of claims 1-6.

Images