A multi-dimensional constraint and data snapshot-based intelligent course scheduling system architecture and implementation method

Abstract

The application provides a kind of intelligent course arrangement system architecture and implementation method based on multidimensional constraint and data snapshot, applied to data processing technical field.The application obtains basic business data and multidimensional constraint parameters, and completes entity time sequence association, constraint structured modeling and resource overall planning through modular decoupling design, to generate standardized input data set.Through creating course arrangement task exclusive data snapshot to lock data state, combined with core algorithm and conflict detection mechanism to generate initial course arrangement scheme, to support manual fine-tuning and real-time verification optimization.Relying on snapshot traceability mechanism to associate historical data, to build closed-loop optimization system, finally through multidimensional timetable display model and snapshot archiving retrieval function, to output evaluation report and visual timetable containing course arrangement efficiency, constraint satisfaction rate and other indicators, to realize efficient, compliant and traceable intelligent course arrangement.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an intelligent scheduling system architecture and implementation method based on multi-dimensional constraints and data snapshots. Background Technology

[0002] Existing course scheduling systems generally suffer from the following technical deficiencies:

[0003] The data is highly coupled and difficult to maintain: basic information such as teachers, classes, and courses are closely tied to the scheduling logic. When a certain piece of information (such as the teacher's role) changes, it will directly affect the correctness and traceability of historical scheduling results, leading to data chaos in the system.

[0004] Weak constraint handling capabilities: The system can usually only handle simple time and location conflicts. It lacks effective modeling and management mechanisms for complex, role-related soft constraints (such as "not priority for class" or "must attend class") and resource group-level global constraints (such as laboratory capacity), resulting in low quality of automatic scheduling results and requiring a lot of manual adjustment.

[0005] Lack of time-series management and data isolation: The system cannot effectively manage data and scheduling tasks by semester, and data from different semesters interfere with each other. When creating new scheduling tasks, the system is easily affected by currently changed system settings, making it impossible to accurately reproduce the decision-making environment of historical scheduling.

[0006] Poor system scalability: Due to unclear module functional boundaries, adding a new constraint type or business function requires extensive system refactoring, resulting in high development and maintenance costs.

[0007] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0008] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0009] According to one aspect of this application, an intelligent scheduling method based on multi-dimensional constraints and data snapshots is provided, comprising: acquiring basic business data and multi-dimensional constraint parameters of the courses to be scheduled; processing the basic business data and multi-dimensional constraint parameters, and based on modular decoupling design, completing the temporal association of entity information, the structured modeling of constraint rules, and the global coordination of resource groups, constructing a basic data-constraint rule mapping model, and generating a standardized input dataset that can be adapted to the scheduling engine; processing the standardized input dataset, creating a data snapshot specific to the scheduling task, locking the current business data status and constraint rule system, and combining the one-click scheduling core algorithm to construct a constraint matching-conflict detection collaborative mechanism to generate an initial intelligent scheduling scheme that meets multi-dimensional constraints; and based on online editing interaction... The framework processes the initial scheduling scheme, introducing a manual fine-tuning instruction receiving module and a real-time conflict verification mechanism to dynamically adjust course time slot allocation, resource usage arrangement, and teacher teaching adaptation relationships, generating a human-machine collaborative optimized scheduling adjustment scheme. It then processes the scheduling adjustment scheme, using a snapshot tracing mechanism to link historical scheduling data in real time, constructing a scheduling effectiveness evaluation model and parameter iterative optimization mechanism to generate a closed-loop optimization system driven by both temporal consistency and constraint adaptability. Finally, it processes the standardized input dataset, initial scheduling scheme, scheduling adjustment scheme, and closed-loop optimization system to construct a multi-dimensional timetable display model and snapshot archiving and retrieval mechanism, generating a comprehensive evaluation report and a multi-dimensional visualized timetable that includes scheduling efficiency, constraint satisfaction rate, historical consistency, and resource utilization.

[0010] Another aspect of this application is an intelligent scheduling system architecture based on multi-dimensional constraints and data snapshots, comprising: the system being configured to execute the aforementioned intelligent scheduling method based on multi-dimensional constraints and data snapshots by executing executable instructions.

[0011] According to another aspect of this application, an electronic device includes: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the above-described intelligent scheduling method based on multi-dimensional constraints and data snapshots by executing the executable instructions.

[0012] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described intelligent scheduling method based on multi-dimensional constraints and data snapshots.

[0013] This application presents an intelligent scheduling system architecture and implementation method based on multi-dimensional constraints and data snapshots. Through a scheduling snapshot mechanism, this application decouples scheduling tasks from dynamically changing global business data, ensuring the immutability and traceability of historical scheduling results. It adopts a highly modular design, with each unit having a single responsibility and clear boundaries. When a new constraint type (such as adding a role) or resource is needed, it only requires extension within the corresponding management unit without modifying the core logic of the scheduling engine, greatly reducing system coupling and improving maintainability and scalability.

[0014] Through independent role management and resource group management units, the system can perform unified and structured modeling and management of soft constraints (preferences) and hard constraints (capacity), providing rich and accurate input for the core algorithm of one-click course scheduling, thereby improving the quality and rationality of the scheduling results from the source. The semester management unit connects all basic data, enabling multi-version management of data and effectively avoiding mutual interference between data from different semesters, allowing the system to better simulate the periodic changes in teaching plans in the real world.

[0015] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0016] Figure 1 The flowchart illustrates an intelligent course scheduling method based on multi-dimensional constraints and data snapshots provided in an embodiment of this application.

[0017] Figure 2 This illustration shows a schematic diagram of the architecture of an intelligent scheduling system based on multi-dimensional constraints and data snapshots, provided in an embodiment of this application. Detailed Implementation

[0018] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0019] The following is combined with Figure 1 This application describes an intelligent scheduling method based on multi-dimensional constraints and data snapshots, according to exemplary embodiments thereof. It should be noted that the following application scenarios are shown only to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited in any way. Rather, the embodiments of this application are applicable to any suitable scenario.

[0020] In one implementation, Figure 1 The diagram illustrates a flowchart of an intelligent scheduling method based on multi-dimensional constraints and data snapshots according to an embodiment of this application.

[0021] S101, obtain the basic business data and multi-dimensional constraint parameters of the courses to be scheduled.

[0022] In one implementation, basic business data is collected by classifying and collecting data from various units within a basic information management module cluster. Each unit performs its own function, and all data is associated with the target semester identifier, ensuring data timeliness and accuracy. Specifically, the basic attributes of various core entities are collected through the entity management unit. Specifically, the teacher management unit collects basic teacher identity information, teaching subjects, and available teaching times; the class management unit collects class number, grade level, and number of students; the course management unit collects course name, course type (required / elective), weekly class hours, and whether dedicated teaching resources (such as laboratories or computer labs) are required; the classroom management unit collects classroom number, capacity, campus location, and equipment configuration (such as whether multimedia and experimental equipment are provided); and the teaching and research office management unit collects the department to which the teaching and research office belongs and the scope of courses it is responsible for coordinating.

[0023] By collecting data through class scheduling plan units, the core information of each class's teaching arrangements in a specific semester is clarified, including the list of courses to be taken, the number of weekly class hours for each course, and the pre-assigned instructors. This data is the key link connecting various basic entity information with the scheduling engine.

[0024] The resource group management unit collects information on shared teaching resources such as laboratories and computer rooms, including resource names, unique numbers, capacity limits (such as the maximum number of classes that can be accommodated at the same time), available time periods, and exclusive usage rules (such as only being available for specific courses), providing data support for global resource constraint management.

[0025] Multi-dimensional constraint parameters are collected through corresponding functional units to form a structured rule system, specifically as follows: Collected through the role management unit, strategic scheduling rules are defined for teacher roles, including four core constraints: "cannot teach," "not a priority for teaching," "must teach," and "priority for teaching." For example, a constraint of "not a priority for teaching on Wednesday afternoons" is set for teachers undertaking teaching and research tasks, and a constraint of "must teach during the corresponding time slot of the core required courses" is set for lecturers of core courses. This unit transforms these abstract preferences into rule labels that the scheduling engine can recognize.

[0026] Based on the time-series division rules of semester management units, key time-series information such as the target semester corresponding to the course scheduling, the number of teaching weeks within the semester, holiday arrangements, and course start and end times are clearly defined to ensure that all course scheduling data is accurately linked to the semester cycle and to avoid data interference between different semesters from the source.

[0027] By combining the shared resource data collected by the resource group management unit, the global capacity limits of various resources are clarified, such as "Laboratory A can accommodate a maximum of 2 classes for experimental teaching at the same time" and "Computer Room B can support a maximum of 40 students to use it at the same time per class", which provides a basis for avoiding resource conflicts during the scheduling process.

[0028] Collect priority goals related to scheduling tasks, such as "prioritizing core courses for graduating classes in prime time slots (e.g., 9-11 am)" and "prioritizing dedicated classrooms or high-quality teaching resources for school-level excellent courses," to provide priority guidance for the rule adaptation of the scheduling algorithm.

[0029] S102 processes basic business data and multi-dimensional constraint parameters. Based on modular decoupling design, it completes the temporal association of entity information, the structured modeling of constraint rules, and the global coordination of resource groups. It constructs a basic data-constraint rule mapping model and generates a standardized input dataset that can be adapted to the scheduling engine.

[0030] In one implementation, the differences between the entity attributes and rule types of multi-dimensional constraint parameters in the basic business data and the functional boundary requirements of the modular decoupling design are compared. A modular hierarchical processing algorithm is used to perform semester-dimensional association of entity information, label conversion of constraint rules, and global capacity coordination of resource groups. This yields basic entity data, constraint rule vectors, and resource configuration parameters suitable for the scheduling logic. Dynamic responsibility division and data association rule matching are implemented for the module decoupling requirements. Combined with temporal isolation technology and constraint structured modeling mechanisms, single-module preprocessed data is generated. First, the differences between the basic business data, constraint parameters, and the functional boundaries of the modular decoupling design are compared. A modular hierarchical processing algorithm (which adopts a three-layer processing architecture of "entity-constraint-resource") is used to complete the initial data processing. The first layer uses an entity attribute clustering algorithm to classify and extract entity data such as teachers, classes, and courses. The second layer uses a constraint rule parsing algorithm to decompose constraint parameters such as role strategies, resource capacity, and semester temporal sequence. The third layer uses a global resource coordination algorithm to perform capacity calculation and time period sorting of shared resource data. Combined with temporal isolation and constraint structured modeling, single-module preprocessed data is generated.

[0031] By comparing the attribute types of each entity (teacher, class, course, etc.) in the basic business data with the rule categories of multi-dimensional constraint parameters (role strategy, resource capacity, semester sequence, etc.), and the functional boundary requirements of each unit in the basic information management module cluster, the data ownership and processing adaptation direction are clarified through the data ownership determination algorithm (based on module function keyword matching and data feature similarity calculation, setting function matching threshold and feature similarity threshold, when the data's function matching degree and feature similarity with a certain module are both higher than the threshold, it is determined to be the data belonging to that module). This avoids functional overlap and data confusion between modules.

[0032] The modular, hierarchical processing algorithm performs three core operations: First, it associates entity information by semester, binding entity data such as teachers, classes, and courses with target semester identifiers to ensure data time-series processing; second, it transforms constraint rules into labels, converting abstract policy constraints such as "cannot teach" and "priority teaching" into rule labels that the scheduling engine can recognize through the role management unit; and third, it coordinates the global capacity of resource groups, sorting out global constraints such as capacity limits and available time periods for shared resources such as laboratories and computer rooms through the resource group management unit.

[0033] Based on the module decoupling requirement, the responsibilities of each module are dynamically divided, clarifying the core processing scope of entity management units, role management units, resource group management units, etc. At the same time, data association rules are established to ensure that entity data, constraint rules, resource information and the functional requirements of the corresponding modules are accurately matched.

[0034] By combining time-series isolation technology, data from different semesters can be independently divided to avoid cross-semester data interference. Through a constraint-structured modeling mechanism, scattered constraint rules are organized into a structured system, and finally, preprocessed data for each module is generated, covering entity base data, constraint rule vectors, and resource configuration parameters that are adapted to the scheduling logic.

[0035] For preprocessed data in a single module, the data attributes are aligned with the module's functional requirements based on the collaborative standards of the basic information management module cluster. This alignment ensures the preprocessed data is linked through semester-related identifiers and role tags, generating a subset of entity data and constraint rules with synchronized attributes. Referring to the collaborative standards of the basic information management module cluster, the data attributes of each preprocessed module are verified to match the corresponding module's functional requirements. For example, the "subject taught" attribute in teacher data is matched with the "course affiliation" attribute in the course management unit to ensure consistent data format and semantic representation, meeting the collaborative needs between modules.

[0036] To align the preprocessed data, a unified semester association identifier and role label are added. Data can be associated across units through the identifier and label. For example, class data labeled "First Semester of the 2024-2025 Academic Year" can be bound to course data and teacher role data with the same identifier, so that entity data and corresponding constraint rules can form an association relationship.

[0037] Through the above association operations, data with synchronized attributes and valid associations are filtered out, and entity data subsets and constraint rule subsets are generated to ensure that each entity data subset is equipped with a corresponding constraint rule subset, and that the two are consistent in key attributes such as semester and role.

[0038] Based on attribute-synchronized entity data subsets and constraint rule subsets, cross-module data consistency verification and standardized business data format processing are performed. The bridging adaptation feature of the class scheduling plan unit ensures the effectiveness of data mapping, generating a standardized input dataset that the scheduling engine can directly adapt to. Cross-module consistency verification is conducted on the attribute-synchronized entity data subsets and constraint rule subsets to check the logical relationships and accuracy between data from different modules. For example, it verifies whether there are conflicts between the "available teaching time slots" of the teacher data subset and the "role strategy constraints" of the constraint rule subset, and whether the resource configuration parameters match the resource requirements of the entity data, eliminating contradictory data.

[0039] Following a standard format recognizable by the scheduling engine, the validated data undergoes standardized formatting, including field names, data types, and representation methods, ensuring consistent data format, well-structured structure, and compliance with the scheduling engine's input requirements. Leveraging the bridging and adaptability of class scheduling plan units, subsets of entity data and constraint rules are mapped to scheduling objectives, clearly defining the mapping relationship between core information such as each class's course list, weekly class hours, and instructors, and corresponding constraint rules and resource allocations, thus ensuring the effectiveness and accuracy of data mapping.

[0040] A standardized input dataset is integrated to form a scheduling engine that can be directly adapted. This dataset covers complete entity information, structured constraint rules, and precise resource configuration parameters. The data has a clear time sequence, accurate correlation, and uniform format, which can directly provide input support for the scheduling engine's scheduling calculation.

[0041] S103 processes the standardized input dataset, creates a dedicated data snapshot for the scheduling task, locks the current business data status and constraint rule system, and, combined with the one-click scheduling core algorithm, constructs a constraint matching-conflict detection collaborative mechanism to generate an initial intelligent scheduling solution that meets multi-dimensional constraints.

[0042] In one implementation, a streamlined scheduling scheme generation mechanism is established, encompassing snapshot creation, data locking, algorithm adaptation, collaborative mechanism construction, and scheme validity verification, to address the data integrity and constraint correlation of the standardized input dataset. When a user creates a new scheduling task, the scheduling task and snapshot management module initiate a snapshot creation process. This process copies and stores key business data from the standardized input dataset, such as class scheduling plans, teacher roles, and resource group capacities, forming a read-only snapshot data specific to that scheduling task. This ensures the traceability of the original data upon which the scheduling is based. By locking the current business data state and constraint rule system through the snapshot mechanism, subsequent scheduling calculations are unaffected by changes in global system data, ensuring that the scheduling logic executes based on a fixed data environment and avoiding scheduling deviations caused by data coupling.

[0043] The scheduling engine module calls the "one-click scheduling core algorithm," which adopts a hybrid architecture of "constraint priority-driven - dynamic resource allocation." Its core includes a constraint parsing sub-algorithm, a resource scheduling sub-algorithm, and a conflict pre-resolution sub-algorithm. The constraint parsing sub-algorithm uses a rule tree traversal mechanism to hierarchically decompose multi-dimensional constraints. The resource scheduling sub-algorithm uses a greedy strategy to achieve efficient matching of resources and time slots. The conflict pre-resolution sub-algorithm uses a predictive model to avoid high-frequency conflict scenarios in advance. The algorithm dynamically adapts parameters based on the characteristics of the standardized input dataset (such as the number of entities and constraint complexity): for scenarios with a large number of entities, the number of parallel computing threads and the data sharding threshold are adjusted to improve data processing throughput; for scenarios with high constraint complexity, the constraint weight calculation coefficient and conflict detection step size are optimized to enhance constraint satisfaction accuracy. Simultaneously, the number of resource allocation iterations and the conflict resolution threshold are adjusted to ensure the algorithm can efficiently handle the core requirements of the current scheduling task.

[0044] A constraint matching and conflict detection collaborative mechanism is constructed, enabling the scheduling engine to read constraint rules (role policies, resource capacity, etc.) from snapshots in real time during algorithm execution and simultaneously perform conflict detection, achieving coordinated progress in constraint matching and conflict avoidance. The initial generated scheduling scheme is validated to check whether it meets all constraint rules, whether the data logic is self-consistent, and whether resource allocation is reasonable, ensuring the scheme has basic feasibility.

[0045] Furthermore, a hierarchical functional allocation strategy is employed between the snapshot management module (responsible for data solidification) and the scheduling engine (responsible for scheme calculation). This, along with the step-by-step execution of static data snapshots and dynamic algorithm operations, forms a foundational system for generating initial scheduling schemes that covers all elements: data, algorithm, mechanism, and effectiveness. The hierarchical functional allocation strategy between the snapshot management module and the scheduling engine module ensures that the snapshot management module is specifically responsible for the solidification, storage, and retrieval of snapshot data, while the scheduling engine module focuses on the calculation of scheduling schemes, conflict detection, and initial optimization. This clearly defines the responsibilities of each module and improves processing efficiency.

[0046] The process is executed step-by-step in the order of "static data snapshot - dynamic algorithm operation": First, the snapshot management module copies key business data such as class scheduling plans, teacher roles, and resource group capacity from the standardized input dataset to a dedicated snapshot table through database transactions and binds it to the scheduling task ID, thus locking and retaining the static data and ensuring that the scheduling data baseline cannot be changed. Then, the scheduling engine module starts the one-click scheduling core algorithm with a hybrid architecture of "constraint priority-driven - dynamic resource allocation" based on the static snapshot data. It executes the constraint parsing sub-algorithm (rule tree traversal to decompose multi-dimensional constraints), resource scheduling sub-algorithm (greedy strategy to match resources and time periods), and conflict pre-resolution sub-algorithm (predictive model to avoid high-frequency conflicts) in sequence. By isolating the execution of static data and dynamic algorithms, the interference of global data dynamic changes on algorithm calculation is avoided. At the same time, the algorithm performs conflict detection by reading the constraint rules in the snapshot in real time, ensuring that the scheduling process is stable and controllable.

[0047] By integrating snapshot data (data layer), one-click scheduling core algorithm (algorithm layer), constraint matching-conflict detection collaborative mechanism (mechanism layer), and scheme validity verification standard (validity layer), a basic system for generating initial scheduling schemes covering all elements is formed, providing stable support for subsequent scheme optimization.

[0048] To address the core requirements of multi-dimensional constraints on the rationality of course scheduling and conflict avoidance, a course scheduling scheme generation and optimization model is constructed. A constraint precision matching layer is added to the model to accurately sort and refine the priority weights of different types of constraints and conflict detection response standards. The model aims to "improve the rationality of course scheduling and conflict avoidance capabilities," aligning with the core requirements of multi-dimensional constraints on course scheduling and achieving precise adaptation between constraint rules and the course scheduling algorithm.

[0049] A constraint precision matching layer is added to the model. The core function of this layer is to prioritize and refine response standards for different types of constraints: First, it clarifies the priority weight of constraints, setting the global capacity constraint of resource groups (hard constraints) as high priority and the role strategy constraint (soft constraints) as medium to high priority, ensuring that key constraints are satisfied first; Second, it formulates conflict detection and response standards, clarifying the detection threshold and response speed requirements for different types of conflicts (such as time conflicts and resource conflicts), thereby improving the pertinence of conflict handling.

[0050] A scheme generation-constraint satisfaction mapping model is constructed to transform the scheduling requirements of the standardized input dataset into quantifiable algorithm execution rules. A constraint satisfaction rate threshold is introduced to standardize and verify the performance of the schemes. An iterative conflict resolution mechanism is used to correct substandard schemes, generating an initial intelligent scheduling scheme that combines constraint adaptability and scheduling rationality. The scheme generation-constraint satisfaction mapping model transforms the scheduling requirements (such as class course lists and weekly class hour requirements) in the standardized input dataset into quantifiable algorithm execution rules, clarifying the computational logic and judgment criteria of the algorithm in resource allocation, time slot arrangement, and teacher adaptation, making the scheduling process systematic and predictable.

[0051] A threshold for constraint satisfaction rate is set as a standardized verification indicator for the performance of the scheme. This threshold is set based on the core requirements of the scheduling business and is used to measure the proportion of scheduling schemes that meet the constraint rules. Schemes that do not reach the threshold need to enter the correction process.

[0052] For scheduling schemes that fail to meet the standards, a conflict iteration resolution mechanism is used to correct them. For detected conflicts (such as multiple courses taught by the same teacher at the same time, or over-allocation of resources), adjustments are made in sequence according to the constraint priority weights, and the scheme is iteratively optimized until it meets the constraint satisfaction rate threshold.

[0053] After the above process optimization, an initial intelligent scheduling scheme with both constraint adaptability and scheduling rationality is generated. This scheme not only meets the multi-dimensional constraint rules, but also meets the actual execution needs of teaching business, providing a high-quality basic version for subsequent manual fine-tuning.

[0054] S104 processes the initial course scheduling scheme based on an online editing and interactive framework, introduces a manual fine-tuning instruction receiving module and a real-time conflict verification mechanism, dynamically adjusts course time allocation, resource usage arrangement, and teacher teaching adaptation relationship, and generates a human-machine collaborative optimized course scheduling adjustment scheme.

[0055] In one implementation, an initial scheduling scheme is loaded based on an online editing and interactive framework. A corresponding data snapshot of the scheduling task is associated with and constraints are locked. The manual fine-tuning instruction receiving module obtains adjustment requirements for time slot allocation, resource usage, teacher adaptation, and class-course matching. The online editing and interactive framework receives the initial intelligent scheduling scheme transmitted by the scheduling engine and presents information such as course time slot allocation, resource usage arrangements, teacher teaching adaptation relationships, and class-course matching logic in a visual format to the user, allowing for intuitive viewing and identification of the content requiring adjustment.

[0056] The system automatically associates the dedicated data snapshot corresponding to the scheduling task, reads all the constraint rules (including role strategy constraints, resource capacity constraints, semester time constraints, etc.) that this scheduling depends on through the snapshot, and sets these rules as unmodifiable constraint benchmarks to ensure that manual fine-tuning is always carried out within the original constraint system, avoiding deviations from the core requirements due to global data changes.

[0057] The module collects various adjustment requests from users through a manual fine-tuning instruction receiving module, ensuring accurate and comprehensive transmission of requirements. This module provides users with a convenient interactive entry point, allowing them to submit adjustment requests for key aspects of the scheduling plan. These include adjustments to course time allocation (e.g., changing a math class from Monday morning to Wednesday afternoon), resource allocation (e.g., changing a course's classroom from a regular classroom to a laboratory), teacher-teaching adaptation (e.g., changing a course's instructor), and adjustments to class-course matching logic (e.g., adding elective courses or removing duplicate courses for a class). The module captures and records these adjustment requests in real time, forming a structured set of adjustment instructions.

[0058] A real-time conflict verification mechanism is activated, which combines the constraint rules in the snapshot to perform real-time verification of the adjustment operation, ensuring that the adjustment process does not deviate from the original constraint system and data consistency requirements. Whenever a user submits an adjustment request, the system immediately triggers the real-time conflict verification mechanism. This mechanism is derived from the conflict detection function of the scheduling engine and can quickly identify various conflicts that may be caused by the adjustment operation.

[0059] The conflict verification mechanism reads the locked constraint rules in the snapshot and verifies the adjustment operation one by one: for example, whether the adjusted course time slot conflicts with the teacher's "cannot teach" role constraint, whether the newly allocated resources exceed the global capacity limit of the resource group, whether the adjusted class-course matching meets the semester sequence requirements, etc. At the same time, it checks whether the adjustment operation will cause data logic contradictions, ensuring that the adjustment process does not deviate from the original constraint system and data consistency requirements.

[0060] The system provides real-time feedback on the verification results to the user, clearly informing them whether the adjustment operation is compliant: if it complies with the rules, it allows the operation to continue; if there is a conflict, it details the type of conflict and the binding rules on which it is based, guiding the user to correct the adjustment requirements.

[0061] The system integrates and validates approved adjustments to generate a human-machine collaborative optimization plan for scheduling adjustments that combines the rationality of human intervention, the adaptability of constraint rules, and the correlation with historical data. The system filters all user-submitted adjustment requests, retaining only compliant operations that have passed real-time conflict verification, and removing invalid operations that conflict with or do not comply with constraint rules.

[0062] The effective adjustments selected are categorized and integrated according to dimensions such as courses, classes, and resources, and applied one by one to the initial course scheduling plan. The time slots, resources, teachers, classes, and other related information in the plan are updated synchronously to ensure that the logic of each part of the plan is consistent.

[0063] After integration and updates, a final human-machine collaborative optimization scheduling adjustment plan is generated. This plan not only reflects the user's willingness to intervene manually, but also fully complies with all the constraints locked in the snapshot. At the same time, by linking the snapshot with historical data, the historical data consistency of the adjusted plan is ensured, providing reliable support for subsequent closed-loop optimization and traceability.

[0064] S105 processes the course scheduling adjustment plan, uses a snapshot tracing mechanism to link historical course scheduling data in real time, constructs a course scheduling effect evaluation model and parameter iterative optimization mechanism, and generates a course scheduling closed-loop optimization system driven by time consistency and constraint adaptability.

[0065] In one implementation, a snapshot tracing mechanism is used to link the scheduling adjustment plan and historical scheduling data in real time, generating a time-series scheduling data association set. The scheduling task and snapshot management module activates the snapshot tracing mechanism, associating the current scheduling adjustment plan with its corresponding exclusive snapshot data through the scheduling task ID. Simultaneously, it retrieves historical scheduling task snapshots (including historical scheduling plans, constraints, resource configuration data, etc.) from the same or related semesters in the system. The core information of the current scheduling adjustment plan (adjusted time slot allocation, resource usage, teacher matching relationships, etc.) is linked with historical scheduling data in real time according to time, class, and course dimensions. Duplicate and redundant data are eliminated, generating a time-series scheduling data association set. This ensures data continuity and relevance, providing complete data support for subsequent effect evaluation.

[0066] A data adaptation analysis algorithm is used to extract features from the associated set of time-series scheduling data, and a scheduling effectiveness evaluation model is constructed. This algorithm integrates a time-series sequence analysis sub-algorithm and a constraint fit calculation sub-algorithm: the time-series sequence analysis sub-algorithm mines time-series correlation features through sliding window matching and trend fitting; the constraint fit calculation sub-algorithm extracts constraint fit features through rule matching quantification and compliance statistics. Two core types of features are extracted: first, time-series consistency features, including the overlap of historical scheduling periods for the same course in the same class (calculated by the percentage of time period intersections) and the consistency of resource usage patterns across different semesters (matched by the similarity of resource usage time-series sequences); second, constraint fit features, including constraint rule satisfaction rate (calculated by the ratio of the number of courses satisfying constraints to the total number of courses), resource capacity compliance rate (statistical analysis of the compliance ratio between actual resource usage and capacity limits), and role strategy fit accuracy rate (calculated by the percentage of times teacher role constraints match scheduling arrangements). The algorithm transforms these features into standardized quantitative indicators in the 0-1 range, ensuring that the scheduling effectiveness evaluation model can accurately identify the correlation between data features.

[0067] A multi-dimensional course scheduling effectiveness evaluation model is constructed. The model takes a time-series course scheduling data association set as input and takes the time consistency and constraint adaptability of the course scheduling scheme as the core evaluation objectives. The evaluation dimensions and calculation logic are set: the time consistency dimension is used to measure the degree of fit between the current scheme and the historical course scheduling pattern, and the constraint adaptability dimension is used to measure the degree to which the scheme meets the multi-dimensional constraints. The model outputs the quantitative evaluation results of course scheduling effectiveness, providing a basis for judgment on scheme optimization.

[0068] By combining temporal consistency standards and constraint adaptability requirements, the scheduling effectiveness evaluation model is iteratively optimized to generate a dynamic optimization parameter set. Temporal consistency standards (such as historical scheduling overlap rate thresholds and resource usage pattern continuity requirements) and constraint adaptability requirements (such as constraint satisfaction rate thresholds and zero-tolerance requirements for resource conflicts) serve as the core basis for model optimization, clarifying the weight allocation direction for each evaluation dimension of the model.

[0069] Based on the above criteria, the key parameters of the course scheduling effectiveness evaluation model are iteratively optimized, including adjusting the weight ratio of temporal consistency features and constraint adaptability features, optimizing the judgment threshold of evaluation indicators, and correcting the feature calculation logic. Through multiple rounds of iteration, the model can more accurately identify the optimization space in the course scheduling scheme, and finally generate a dynamic optimization parameter set to provide accurate guidance for scheme correction.

[0070] The scheduling engine module performs real-time adaptation and correction of the timetable adjustment scheme based on a dynamic optimization parameter set, generating dynamic adjustment instructions for timetable optimization. The module receives the dynamic optimization parameter set and, according to the optimization direction and correction standards within the parameter set, performs real-time adaptation and correction of the current timetable adjustment scheme. For example, if the model evaluation shows that a class's time slot conflicts with historical scheduling patterns (timing consistency not met), the time slot is adjusted to a time slot that conforms to historical patterns; if resource usage exceeds constraints (constraint adaptability not met), compliant resources are reallocated. Conflict detection is triggered simultaneously during the correction process to ensure the corrected scheme is compliant, ultimately generating dynamic adjustment instructions for timetable optimization, specifying the detailed adjustment content and execution standards.

[0071] The system collaboratively processes time-series correlated data, scheduling effectiveness evaluation results, and dynamic adjustment instructions to generate a closed-loop scheduling optimization system driven by both time-series consistency and constraint adaptability. Time-series correlated data provides the data foundation for the closed loop, evaluation results provide optimization directions, and dynamic adjustment instructions provide the execution path, forming a cyclical mechanism of "data collection - effectiveness evaluation - scheme correction - data update." This mechanism, driven by time-series consistency and constraint adaptability as its two core drivers, continuously monitors the quality of scheduling schemes, automatically or guiding users to make optimizations and adjustments, ultimately generating a closed-loop scheduling optimization system driven by both time-series consistency and constraint adaptability, ensuring the continuous optimization and compliance of scheduling schemes.

[0072] S106 processes the standardized input dataset, initial scheduling scheme, scheduling adjustment scheme, and closed-loop optimization system to build a multi-dimensional timetable display model and snapshot archiving and retrieval mechanism, generating a comprehensive evaluation report and a multi-dimensional visualized timetable that includes scheduling efficiency, constraint satisfaction rate, historical consistency, and resource utilization.

[0073] In one implementation, based on the entire course scheduling process data, a multi-dimensional integration strategy is adopted. First, the standardized input dataset, initial course scheduling plan, course scheduling adjustment plan, and closed-loop optimization system are correlated and formatted. Based on the multi-dimensional integration strategy, the entire process data is correlated and formatted uniformly, laying the foundation for subsequent processing: The core data of the standardized input dataset, initial course scheduling plan, course scheduling adjustment plan, and closed-loop optimization system are comprehensively correlated: data relationships are established according to key fields such as semester identifier, class number, course name, and resource ID, clarifying the corresponding logic of data at each stage; simultaneously, the format of data from different sources is standardized, unifying the field definitions, data types, and expression standards (such as unifying time period formats and constraint rule expression methods), eliminating duplicate and redundant data, ensuring that the integrated data is logically coherent and formatted uniformly, forming a complete course scheduling process dataset.

[0074] After data integration, a multi-dimensional timetable display model was constructed, supporting the visualization of classes, teachers, and resource groups, and establishing a retrieval mechanism that links snapshots with timetable results. The multi-dimensional timetable display model supports visualization from three core dimensions: class, teacher, and resource group. The class dimension displays the complete course schedule for a specific class (including time slot, classroom, and instructor); the teacher dimension displays the teaching schedule for a specific teacher (including class, course, and time slot); and the resource group dimension displays the occupancy status of resources such as laboratories and computer labs (including class, course, and time slot), meeting the viewing needs of different users.

[0075] Establish a mechanism to link snapshots with timetable results, associateing each final timetable with a unique data snapshot of the corresponding scheduling task, and assigning each timetable a unique snapshot retrieval identifier. When viewing the timetable, users can quickly retrieve the corresponding snapshot data using this identifier, view the business data status and constraint rules at the time of scheduling, and ensure the traceability of timetable results.

[0076] Based on core scheduling metrics, a quantitative evaluation algorithm was designed to calculate and summarize the results of the entire process data. The core scheduling metrics are defined as scheduling efficiency, constraint fulfillment rate, historical consistency, and resource utilization rate, all of which align with the beneficial effects of the technical disclosure document and business needs: scheduling efficiency reflects the time taken from initiation to completion of the scheduling task; constraint fulfillment rate reflects the proportion of timetables that meet multi-dimensional constraints; historical consistency reflects the degree to which the current timetable matches historical scheduling patterns; and resource utilization rate reflects the effective utilization rate of resources such as classrooms and laboratories.

[0077] A quantitative evaluation algorithm is designed, employing a "hierarchical calculation of indicators - normalized summary of results" architecture. This algorithm includes sub-algorithms for calculating time-series indicators, constraint indicators, and resource indicators. Using a standardized dataset of the entire scheduling process as input, it accurately calculates each core indicator according to pre-defined logic. Scheduling efficiency is achieved through the time-series indicator calculation sub-algorithm, employing a precise time difference calculation logic. The core formula is "scheduling completion time - scheduling start time," simultaneously eliminating non-algorithm execution time such as manual pauses during scheduling, outputting a precise quantitative value of scheduling time consumption. Constraint satisfaction rate is achieved through the constraint indicator calculation sub-algorithm, using rule-matching counting logic. It first iterates through and compares the entire scheduling process data using constraint rules, counting the "number of courses satisfying the constraints" and the "total number of courses," then calculates the result using the ratio of these two values, while simultaneously marking courses that do not satisfy the constraints and their corresponding constraint types.

[0078] Historical consistency is achieved through a time-series indicator calculation sub-algorithm. It employs a scheduling overlap matching logic to extract key dimensions such as time periods, resources, and teachers from the current and historical scheduling. It calculates the proportion of "current scheduling overlaps with historical scheduling" to the "total scheduling," outputting a quantitative result of historical consistency. Resource utilization is achieved through a resource-based indicator calculation sub-algorithm. It uses a resource occupancy duration statistics logic to calculate the "actual resource occupancy duration" and "total available resource duration" for each resource type. The ratio of these two values ​​yields the utilization rate for each resource type, which is then weighted by resource weights to calculate the overall resource utilization rate. Finally, the algorithm normalizes the quantitative results of each indicator to the 0-1 range, categorizes and summarizes them, and generates a detailed indicator calculation report, providing accurate data support for subsequent comprehensive evaluation.

[0079] The evaluation indicator results are categorized and logically verified, and a structured comprehensive evaluation report is generated according to importance weights to ensure data authenticity and traceability. The quantified indicator results are categorized and logically verified as follows: each indicator result is classified as "Excellent, Good, Satisfactory, Unsatisfactory"; the logical rationality between indicator results is verified (e.g., if the constraint satisfaction rate is unsatisfactory, is there an abnormally high resource utilization rate?), and contradictory or abnormal indicator results are eliminated to ensure the authenticity and reliability of the evaluation data.

[0080] The results of the indicators are ranked according to their importance weight (constraint satisfaction rate > historical consistency > resource utilization rate > scheduling efficiency) to generate a structured comprehensive evaluation report. The report includes basic information about the scheduling task, quantitative results and hierarchical explanations of each core indicator, explanations of indicator logic verification, snapshot retrieval identifiers, etc., to ensure that the report data is traceable and the conclusions are clear and explicit.

[0081] The system comprehensively evaluates the multi-dimensional timetable display, snapshot archiving retrieval efficiency, and assessment report completeness to generate a final comprehensive assessment report and a multi-dimensional visualized timetable. The system performs a final verification based on the multi-dimensional timetable display (clarity and intuitiveness, completeness of dimensions), snapshot archiving retrieval efficiency (retrieval response speed, accuracy of associations), and assessment report completeness (comprehensiveness of information, rigor of logic) to ensure that all deliverables meet business requirements. After verification, the final comprehensive assessment report and multi-dimensional visualized timetable are generated. Users can directly view and export the timetable, or trace the original scheduling data through snapshot markers in the report, meeting the needs for scheduling implementation and review in teaching management.

[0082] In one implementation, such as Figure 2 As shown, this application also provides an intelligent course scheduling system architecture based on multi-dimensional constraints and data snapshots, including:

[0083] The basic information collection module 201 is used to acquire basic business data and multi-dimensional constraint parameters of the courses to be scheduled.

[0084] The basic data processing module 202 is used to process basic business data and multi-dimensional constraint parameters. Based on modular decoupling design, it completes the temporal association of entity information, the structured modeling of constraint rules, and the global coordination of resource groups. It constructs a basic data-constraint rule mapping model and generates a standardized input dataset that can be adapted to the scheduling engine after cross-module consistency verification and standardized format processing.

[0085] The intelligent scheduling engine module 203 is used to process standardized input datasets, create exclusive data snapshots for scheduling tasks, lock the current business data status and constraint rule system, combine the one-click scheduling core algorithm, build a constraint matching-conflict detection collaborative mechanism, and generate an initial intelligent scheduling scheme that meets multi-dimensional constraints through data solidification and layered execution of algorithm operations.

[0086] The scheduling scheme optimization module 204 is used to load the initial scheduling scheme based on the online editing interaction framework, associate the corresponding data snapshot to lock the constraint benchmark, introduce a manual fine-tuning instruction receiving module and a real-time conflict verification mechanism, dynamically adjust the course time allocation, resource occupation arrangement, teacher teaching adaptation relationship, integrate compliant adjustment operations, and generate a human-machine collaborative optimized scheduling adjustment scheme.

[0087] The closed-loop optimization management module 205 is used to process the scheduling adjustment plan. It uses a snapshot tracing mechanism to link historical scheduling data in real time, extracts time-series consistency and constraint adaptability features, constructs a scheduling effect evaluation model and parameter iterative optimization mechanism, and generates a time-series consistency-constraint adaptability dual-driven closed-loop scheduling optimization system after multiple rounds of adaptation and correction.

[0088] The timetable output and display module 206 is used to process the standardized input dataset, initial timetable scheme, timetable adjustment scheme and closed-loop optimization system, build a multi-dimensional timetable display model and snapshot archiving and retrieval mechanism, design a quantitative evaluation algorithm to calculate core indicators, and generate a comprehensive evaluation report and a multi-dimensional visualized timetable that includes timetable efficiency, constraint satisfaction rate, historical consistency and resource utilization.

[0089] The computer-readable storage medium provided in the above embodiments of this application and the intelligent scheduling method based on multi-dimensional constraints and data snapshots provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0090] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for evaluating the intelligent scheduling method, electronic device, electronic device, and readable storage medium based on multi-dimensional constraints and data snapshots are basically similar to the embodiments of the intelligent scheduling method based on multi-dimensional constraints and data snapshots described above, and are therefore described simply. Relevant parts can be referred to in the descriptions of the embodiments of the intelligent scheduling method based on multi-dimensional constraints and data snapshots described above.

Claims

1. An intelligent course scheduling method based on multi-dimension constraints and data snapshots, characterized in that, include: Obtain basic business data and multi-dimensional constraint parameters for courses to be scheduled; The system processes basic business data and multi-dimensional constraint parameters. Based on modular decoupling design, it completes the temporal association of entity information, the structured modeling of constraint rules, and the global coordination of resource groups. It constructs a basic data-constraint rule mapping model and generates a standardized input dataset that can be adapted to the scheduling engine. The standardized input dataset is processed to create a dedicated data snapshot for the scheduling task, locking the current business data status and constraint rule system. Combined with the core algorithm of one-click scheduling, a constraint matching and conflict detection collaborative mechanism is constructed to generate an initial intelligent scheduling scheme that meets multi-dimensional constraints. This includes ensuring data integrity and constraint correlation for the standardized input dataset, establishing a streamlined process for generating scheduling schemes through snapshot creation, data locking, algorithm adaptation, collaborative mechanism construction, and scheme validity verification. Furthermore, a hierarchical functional allocation strategy is implemented, with the snapshot management module responsible for data solidification and the scheduling engine responsible for scheme calculation. The step-by-step execution of static data snapshots and dynamic algorithm calculations forms a comprehensive coverage... The system generates an initial course scheduling plan based on a comprehensive framework encompassing data, algorithms, mechanisms, and effectiveness. It addresses the core requirements of multi-dimensional constraints for course scheduling rationality and conflict avoidance by constructing an optimized course scheduling plan generation model. This model includes a constraint precision matching layer to accurately analyze and refine the priority weights of different types of constraints and conflict detection response standards. Furthermore, it constructs a plan generation-constraint satisfaction mapping model, transforming the scheduling requirements of the standardized input dataset into quantifiable algorithm execution rules. A constraint satisfaction rate threshold is introduced to standardize and validate plan performance. Combined with a conflict iteration resolution mechanism, substandard plans are corrected, resulting in an initial intelligent course scheduling plan that combines constraint adaptability with reasonable scheduling. The initial course scheduling scheme is processed based on an online editing and interactive framework. A manual fine-tuning instruction receiving module and a real-time conflict verification mechanism are introduced to dynamically adjust course time slot allocation, resource allocation, and teacher teaching adaptation relationships, generating a human-computer collaborative optimized course scheduling adjustment scheme. This includes loading the initial course scheduling scheme based on the online editing and interactive framework, associating it with a dedicated data snapshot of the corresponding scheduling task to lock the constraint benchmark, obtaining adjustment requirements for time slot allocation, resource allocation, teacher adaptation, and class-course matching through the manual fine-tuning instruction receiving module, initiating a real-time conflict verification mechanism, and verifying the adjustment operations in real time based on the constraint rules in the snapshot to ensure that the adjustment process does not deviate from the original constraint system and data consistency requirements, and integrating the verified adjustment operations to generate a human-computer collaborative optimized course scheduling adjustment scheme that combines the rationality of manual intervention, the adaptability of constraint rules, and the correlation with historical data. The scheduling adjustment plan is processed, and historical scheduling data is linked in real time through a snapshot tracing mechanism. A scheduling effect evaluation model and parameter iterative optimization mechanism are constructed to generate a scheduling closed-loop optimization system driven by time consistency and constraint adaptability. The standardized input dataset, initial scheduling scheme, scheduling adjustment scheme and closed-loop optimization system are processed to build a multi-dimensional timetable display model and snapshot archiving and retrieval mechanism, and generate a comprehensive evaluation report and a multi-dimensional visualized timetable that includes scheduling efficiency, constraint satisfaction rate, historical consistency and resource utilization.

2. The method as described in claim 1, characterized in that, The system processes basic business data and multi-dimensional constraint parameters. Based on a modular decoupling design, it completes the temporal association of entity information, structured modeling of constraint rules, and global coordination of resource groups. It constructs a basic data-constraint rule mapping model and generates a standardized input dataset that the scheduling engine can adapt to, including: By comparing the differences between the entity attributes and rule types of multi-dimensional constraint parameters in the basic business data and the functional boundary requirements of the modular decoupling design, the modular hierarchical processing algorithm is used to perform semester-dimensional association of entity information, label conversion of constraint rules, and global capacity coordination of resource groups. This process obtains basic entity data, constraint rule vectors, and resource configuration parameters that are compatible with the scheduling logic. Dynamic responsibility division and data association rule matching are implemented for the module decoupling requirements. Combined with temporal isolation technology and constraint structured modeling mechanism, preprocessed data for a single module is generated. For preprocessed data in a single module, the data attributes and module functional requirements are adapted and aligned based on the collaborative standard of the basic information management module cluster. This allows the aligned preprocessed data to be associated with the semester association identifier and role tag, generating a subset of entity data and a subset of constraint rules with synchronized attributes. Based on the entity data subset and constraint rule subset of attribute synchronization, through cross-module data consistency verification and business data standardization format processing, combined with the bridging adaptation characteristics of class scheduling plan unit to ensure the effectiveness of data mapping, a standardized input dataset that the scheduling engine can directly adapt to is generated.

3. The method as described in claim 1, characterized in that, The scheduling adjustment plan is processed, and historical scheduling data is linked in real time through a snapshot tracing mechanism. A scheduling effectiveness evaluation model and parameter iterative optimization mechanism are constructed to generate a scheduling closed-loop optimization system driven by both temporal consistency and constraint adaptability, including: Based on the snapshot traceability mechanism, the scheduling adjustment plan and historical scheduling data are linked in real time to generate a time-series scheduling data association set; By using data adaptation analysis algorithms, feature extraction is performed on the correlation set of time-series scheduling data to construct a scheduling effectiveness evaluation model; By combining the time-series consistency standard and constraint adaptability requirements, the parameters of the course scheduling effectiveness evaluation model are iteratively optimized to generate a dynamic optimization parameter set; Based on the dynamic optimization parameter set, the scheduling adjustment scheme is adapted and corrected in real time, and the dynamic adjustment instruction for scheduling optimization is generated. By collaboratively processing time-series correlated data, scheduling effectiveness evaluation results, and dynamic adjustment instructions, a closed-loop scheduling optimization system driven by both time-series consistency and constraint adaptability is generated.

4. The method as described in claim 1, characterized in that, The standardized input dataset, initial scheduling scheme, scheduling adjustment scheme, and closed-loop optimization system are processed to construct a multi-dimensional timetable display model and snapshot archiving and retrieval mechanism. This generates a comprehensive evaluation report and a multi-dimensional visualized timetable, including metrics such as scheduling efficiency, constraint satisfaction rate, historical consistency, and resource utilization. Based on the data from the entire course scheduling process, a multi-dimensional integration strategy is adopted. First, the standardized input dataset, the initial course scheduling plan, the course scheduling adjustment plan, and the closed-loop optimization system are correlated and formatted. After data integration, a multi-dimensional timetable display model is built to support the visualization of classes, teachers, and resource groups, and a retrieval mechanism that binds snapshots and timetable results is established. Based on the core indicators of course scheduling, a quantitative evaluation algorithm was designed to calculate the indicators and summarize the results of the entire process data; The evaluation indicators are classified and logically verified, and a structured comprehensive evaluation report is generated according to their importance weights to ensure that the data is authentic and traceable. By combining the multi-dimensional timetable display effect, snapshot archiving and retrieval efficiency, and evaluation report completeness, a final comprehensive evaluation report and multi-dimensional visualized timetable are generated.

5. An intelligent course scheduling system architecture based on multi-dimensional constraints and data snapshots, characterized in that, The system architecture is configured to execute the intelligent scheduling method based on multi-dimensional constraints and data snapshots as described in any one of claims 1 to 4 by executing executable instructions.

6. An electronic device, characterized in that, include: First processor; and memory for storing executable instructions of the first processor; The first processor is configured to execute the intelligent scheduling method based on multi-dimensional constraints and data snapshots as described in any one of claims 1 to 4 by executing the executable instructions.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the second processor, it implements the intelligent scheduling method based on multi-dimensional constraints and data snapshots as described in any one of claims 1 to 4.

Images