Intelligent educational administration management and automatic course arrangement system based on simulated annealing algorithm

Abstract

The invention relates to the technical field of educational administration management, in particular to an intelligent educational administration management and automatic course arrangement system based on a simulated annealing algorithm, which comprises a data and knowledge layer module, a constraint and target engine module, a simulated annealing optimization engine module, a business collaboration and interface module and a visualization module. The data and knowledge layer stores and manages cultivation schemes, teachers, classrooms, semester time models and historical class schedule data, and generates a teaching task quintuple including courses, teachers, classes, classrooms and time; defining and calculating hard and soft constraints by the constraint and target engine, and supporting scenarized configuration of constraint weights; the simulated annealing optimization engine optimizes the class schedule through initial solution construction, multi-neighborhood search and other units; the business collaboratively responds to the business change and triggers global or incremental optimization; and an interface and visual output course arrangement result and visual conflict are provided, and an API integrated external system is provided. The curriculum schedule is ensured to be feasible, and the curriculum arrangement quality and efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of academic affairs management technology, specifically to an intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm. Background Technology

[0002] In the current field of academic affairs management, automatic course scheduling systems face multiple technical challenges, making it difficult to meet the needs of intelligent academic affairs management in terms of efficiency, accuracy, and flexibility.

[0003] In terms of data management, the existing system cannot fully integrate training program data, teacher data, classroom data, semester time model and historical timetable data, nor can it generate a five-tuple of teaching tasks containing course, teacher, class, classroom and time according to standard procedures. This leads to incomplete or logically conflicting basic data for scheduling, which affects the accuracy of scheduling. In terms of constraint handling, the existing system has loopholes in the definition and inspection of hard constraints, which can easily lead to scheduling conflicts that violate the basic logic of academic affairs; it does not fully cover soft constraints and lacks the ability to configure constraint weights in a scenario-based manner, making it unable to adapt to the optimization needs of different academic affairs scenarios. In terms of timetable optimization, the existing system's initial timetable solution is difficult to satisfy all hard constraints. The iterative optimization search method is too simple and prone to local optima. Furthermore, it lacks a scientific temperature scheduling and convergence fine-tuning mechanism, which affects both optimization efficiency and timetable quality. At the business collaboration level, the existing system has a weak ability to respond to business changes such as curriculum updates, classroom applications, and temporary class rescheduling. Either global rescheduling consumes resources and disrupts the stability of the timetable, or local modifications are prone to causing new conflicts. In terms of output and system integration, the existing system's scheduling results are monotonous, making it difficult to intuitively display resource status and conflicts. Furthermore, it lacks standardized API interfaces, making it difficult to integrate efficiently with external academic affairs systems and affecting the overall efficiency of academic affairs management.

[0004] Therefore, to address the above problems, an intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm includes: The data and knowledge layer module is used to store and manage training program data, faculty data, classroom data, semester time models and historical timetable data, and generate teaching task quintuples, which include course, teacher, class, classroom and time. The Constraints and Objectives Engine module is used to define and compute hard and soft constraints. Hard constraints include the uniqueness of teachers at the same time, the uniqueness of classes, the uniqueness of classrooms, the matching of classroom capacity and type, and prerequisite course rules. Soft constraints include course uniformity, the same class on different days, minimizing the walking distance between teachers and students, and morning and evening preferences. It also supports the scenario-based configuration of constraint weights. The simulated annealing optimization engine module connects the data and knowledge layer module and the constraint and objective engine module. It is used for timetable optimization based on the simulated annealing algorithm and includes: The initial solution construction unit is used to generate an initial timetable solution that satisfies all hard constraints based on a heuristic greedy algorithm and a local repair algorithm. The multi-neighborhood search unit is used to perform one or more neighborhood operations during the simulated annealing iteration process, including time slot swapping, time slot shifting, classroom replacement, cross-building swapping, grouping, or splitting. The annealing scheduling unit is used to manage the initial temperature setting, segmented cooling and reheating strategies, and to control the acceptance and rejection of candidate solutions based on the Metropolis criterion. The convergence fine-tuning unit is used to fine-tune the unqualified regions by calling a local metaheuristic algorithm after the iteration meets the stopping condition; The business collaboration module connects to the simulated annealing optimization engine module to respond to external business change requests, including curriculum version updates, classroom application approvals and temporary class rescheduling, and triggers the optimization engine to perform global optimization or incremental rearrangement based on the influence domain. The interface and visualization module is used to output scheduling results, visualize conflicts and resource status, and provide APIs for data integration with external academic affairs systems.

[0007] As a preferred approach, the data and knowledge layer module performs the following operations when generating the five-tuple of teaching tasks: Based on the data analysis of the training program, the correspondence between courses and classes is analyzed to generate an initial task set containing courses and classes; Based on the teacher data, assign qualified teachers to each task in the initial task set, forming a course-teacher-class task subset; Based on classroom data, including classroom capacity and type, a matching classroom is assigned to each task in the course-teacher-class task subset, forming a course-teacher-class-classroom task extension set; Using the semester time model, time periods are allocated to each task in the course-teacher-class-classroom task extension set to generate a preliminary teaching task quintuple. Referencing historical timetable data, perform conflict checks and adjustments on the initial teaching task quintuples, and output the final teaching task quintuples.

[0008] As a preferred option, the constraint and objective engine module is used to define and compute hard and soft constraints, and supports scenario-based configuration of constraint weights, including: Define the rules for the uniqueness of teachers at the same time, the uniqueness of classes, the uniqueness of classrooms, the matching rules for classroom capacity and type, and the rules for prerequisite courses, forming a set of hard constraint rules; Define the course uniformity rule, the same course on different days rule, the rule of minimizing the walking distance between teachers and students rule, and the morning / evening preference rule to form a set of soft constraint rules; For a given timetable solution, conflict detection is performed based on the set of hard constraint rules. If any hard constraint rule is detected to be violated, the timetable solution is marked as infeasible. For the timetable solutions marked as feasible, the degree of violation of each soft constraint rule is calculated based on the weighted soft constraint rule set, and the weighted violation degrees are aggregated to obtain the total soft constraint evaluation score; Based on the results of the hard constraint checking step and the soft constraint evaluation step, the overall objective function value of the timetable solution is output for the iterative evaluation of the simulation annealing optimization engine.

[0009] As a preferred approach, the initial solution construction unit is used to generate an initial timetable solution that satisfies all hard constraints based on a heuristic greedy algorithm and a local patching algorithm, including: Based on the teaching task quintuples output by the data and knowledge layer module, a heuristic greedy algorithm is used to prioritize the allocation of time slots and classroom resources for each teaching task, generating a preliminary timetable solution. The heuristic greedy algorithm allocates resources based on resource availability rules, including classroom capacity matching, teacher time availability, and avoidance of class time conflicts. Hard constraint conflict detection is performed on the preliminary timetable solution to identify allocation positions that violate the rules of teacher uniqueness, class uniqueness, classroom uniqueness, classroom capacity and type matching, or prerequisite courses. For the identified hard constraint conflicts, a local patching algorithm is applied to iteratively adjust the conflict location, including reallocating time slots, changing classrooms, or modifying the task order, in order to eliminate individual conflicts. During the local repair process, the hard constraint satisfaction status of the timetable solution is continuously checked. If there are unrepaired conflicts, the conflict detection and local repair steps are repeated until all hard constraints are satisfied. The output is an initial timetable solution that satisfies all hard constraints, which is then used as input to the simulated annealing optimization engine.

[0010] As a preferred approach, the multi-neighborhood search unit is used in the simulated annealing iteration process to include: Based on the current timetable solution, a neighborhood operation type is randomly selected. Neighborhood operation types include time slot exchange, time slot shift, classroom replacement, cross-building exchange, grouping, or splitting. Based on the selected neighborhood operation type, identify one or more teaching tasks as operation objects from the current timetable solution; Apply the selected neighborhood operation to the object being operated on to generate a new candidate timetable solution; The new candidate timetable solution is output to the annealing scheduling unit for subsequent processing.

[0011] As a preferred embodiment, the annealing scheduling unit manages the initial temperature setting, staged cooling and reheating strategies, and controls the acceptance and rejection of candidate solutions based on the Metropolis criterion, including: The soft-constraint evaluation total score is based on the initial timetable solution. The initial temperature value is calculated by random perturbation sampling, so that the initial acceptance probability is higher than the preset threshold. Set the current temperature value to the initial temperature value, and initialize the current timetable solution to the initial timetable solution; Entering a segmented cooling cycle, the optimization process is divided into multiple temperature stages. Each temperature stage performs a predetermined number of iterations. Each iteration includes: receiving new candidate timetable solutions from the multi-neighborhood search unit, calculating the objective function value of the new candidate timetable solution, and comparing it with the objective function value of the current timetable solution to obtain the change in the objective function value. The change in the objective function value is processed based on the Metropolis criterion. If the change in the objective function value is less than zero, a new candidate timetable solution is accepted as the current timetable solution. If the change in the objective function value is greater than or equal to zero, a new candidate timetable solution is accepted as the current timetable solution with a certain probability. The acceptance probability is calculated based on the change in the objective function value and the current temperature value. After the iteration of the current temperature stage is completed, the current temperature value is reduced according to the segmented cooling strategy. The segmented cooling strategy sets a different cooling rate for each temperature stage, and the cooling rate decreases as the temperature stage increases. After lowering the current temperature value, monitor the convergence status. If it is detected that the objective function value of the current course solution has not improved for multiple consecutive temperature stages, trigger the retemperature strategy to reset the current temperature value to the starting temperature value of the previous temperature stage or a higher temperature value. Repeat the segmented cooling cycle until the current temperature value is lower than the termination temperature threshold or the maximum number of iterations is reached, then output the current timetable solution to the convergence fine-tuning unit.

[0012] As a preferred embodiment, the convergence fine-tuning unit is used to fine-tune the unqualified regions by invoking a local metaheuristic algorithm after the iteration meets the stopping condition, including: Based on the current timetable solution output by the simulated annealing optimization engine, regions that do not meet soft constraints or whose objective function values ​​do not meet the standards are identified as non-compliant regions. For the identified non-compliant areas, at least one local metaheuristic algorithm is selected based on the area characteristics. The local metaheuristic algorithm includes local search algorithm, greedy algorithm or variable neighborhood search algorithm. The selected local metaheuristic algorithm is applied to iteratively fine-tune the unqualified areas to generate the fine-tuned timetable solution; The fine-tuned timetable solution is verified by hard constraints to ensure that all hard constraints are still satisfied. Based on the hard constraint verification results, calculate the soft constraint evaluation total score of the fine-tuned timetable solution; Based on the comparison between the total soft constraint evaluation score and the total soft constraint evaluation score of the original timetable solution, a decision is made on whether to accept the fine-tuning result. If the total soft constraint evaluation score after fine-tuning is better, the fine-tuned timetable solution is accepted as the final output; otherwise, the original timetable solution is reverted to and other fine-tuning strategies are tried.

[0013] As a preferred solution, the business collaboration module is used to respond to external business change requests, including curriculum version updates, classroom application approvals, and temporary class rescheduling, and to trigger the optimization engine to perform global optimization or incremental reordering based on the impact domain, including: Receive external business change requests and parse the request type and content; Based on the parsed request type and content, determine the scope of impact of the change request. If the change request involves multiple teaching tasks or global resource adjustments, trigger global optimization; otherwise, trigger incremental rearrangement based on the impact domain. If an incremental rearrangement based on the impact domain is triggered, the impact domain is calculated based on the content of the change request. The impact domain includes courses, teachers, classes, classrooms, or time periods that are directly or indirectly affected by the change. Based on the calculated influence domain, the simulated annealing optimization engine is invoked to perform incremental rearrangement. The simulated annealing optimization engine only optimizes and adjusts the teaching tasks within the influence domain, while keeping the teaching tasks outside the influence domain unchanged. If global optimization is triggered, the simulated annealing optimization engine will be invoked to re-optimize the entire timetable; Output the optimized timetable results to the interface and visualization module.

[0014] As a preferred solution, the interface and visualization module are used to output scheduling results, visualize conflicts and resource status, and provide an API for data integration with external academic affairs systems, including: Receive the final scheduling results from the business collaboration module; Based on the received final scheduling results, the constraint and target engine module is invoked to evaluate hard and soft constraints, and generate conflict reports and resource status summaries. By utilizing conflict reports and resource status summaries, combined with the final scheduling results, a visual dataset is constructed, including timetable arrangements, resource allocation details, and conflict location information; Convert the visualized dataset into a graphical user interface renderable format to generate timetable views, resource usage heatmaps, and conflict highlight reports; The system outputs a timetable view, resource usage heatmap, and conflict highlight report through a graphical user interface for users to view interactively. The system receives data query requests from external academic affairs systems via API interfaces and returns query responses based on the final course scheduling results and visualized datasets. The system receives data update requests from external academic affairs systems via API interfaces and forwards these requests to the business collaboration module to trigger subsequent optimization processes.

[0015] As can be seen from the technical solution provided by the present invention above, the intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm provided by the present invention has the following beneficial effects: This invention stores and manages training program data, teacher data, classroom data, semester time models, and historical timetable data through a data and knowledge layer module, and generates teaching task quintuples according to a standardized process. During the generation process, multiple rounds of conflict checks and adjustments are performed, such as verifying the initial quintuples with reference to historical timetable data, to ensure that the final output teaching task quintuples are free of data conflicts and logically consistent. This avoids scheduling problems caused by data errors or contradictions from the source, and provides accurate and usable basic data for constraint calculation and timetable optimization. The constraint and objective engine module clearly defines hard and soft constraint rules. Hard constraints prevent violations of academic affairs logic, such as "repeated scheduling of teachers at the same time" and "mismatch of classroom capacity," through conflict detection, ensuring that the timetable has basic feasibility. Soft constraints quantify the quality of the timetable through weighted evaluation and support scenario-based configuration of constraint weights. The priority of soft constraints can be adjusted according to the needs of different departments and semesters. For example, science and engineering departments can increase the weight of minimizing the walking distance between teachers and students, while humanities departments can increase the weight of course uniformity, making the timetable quality evaluation more in line with actual business scenarios and taking into account both feasibility and personalized needs. The simulated annealing optimization engine module achieves efficient optimization through multi-unit collaboration; the initial solution construction unit combines heuristic greedy algorithms and local patching algorithms to quickly generate initial timetable solutions that satisfy hard constraints; the multi-neighborhood search unit expands the search space of solutions through diverse neighborhood operations, avoiding the fixation of optimization directions; the annealing scheduling unit balances the "exploration" and "utilization" of solutions through segmented cooling and reheating strategies, preventing getting trapped in local optima; the convergence fine-tuning unit performs local optimization on unqualified areas, further improving the quality of the timetable; the entire optimization process ensures the high quality of the timetable solutions while controlling optimization time through reasonable scheduling strategies, thus improving optimization efficiency. The business collaboration module can promptly receive and parse external business change requests such as curriculum version updates, classroom application approvals, and temporary class rescheduling. By determining the scope of the change's impact, it triggers different optimization methods. For changes affecting the entire system, it triggers global optimization to ensure that the overall timetable adapts to the new requirements. For local changes, it calculates the impact domain and only performs incremental rearrangement of teaching tasks within the impact domain, keeping the timetable outside the impact domain unchanged. This differentiated processing not only avoids timetable failure caused by business changes but also reduces unnecessary computational consumption, while ensuring the overall stability of the timetable and minimizing interference with the original teaching arrangements for teachers and students. The interface and visualization module transforms the optimized scheduling results into an intuitive timetable view, resource usage heatmap, and conflict highlighting report. Users can quickly view the timetable arrangement, resource usage, and conflict issues through the graphical user interface, reducing the difficulty of information acquisition. At the same time, the module provides an API interface for integration with external academic affairs systems, supporting external systems to query scheduling data and initiate update requests, realizing collaboration between the scheduling system and external academic affairs management processes, reducing data silos, and improving the overall efficiency of academic affairs management. The conflict highlighting function also makes it easier for administrators to quickly locate and handle conflicts, further ensuring the smooth execution of the timetable. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall structure of an intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific embodiments.

[0019] like Figure 1 As shown, this embodiment of the invention provides an intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm, including: The data and knowledge layer module is used to store and manage training program data, faculty data, classroom data, semester time models and historical timetable data, and generate teaching task quintuples, which include course, teacher, class, classroom and time. The Constraints and Objectives Engine module is used to define and compute hard and soft constraints. Hard constraints include the uniqueness of teachers at the same time, the uniqueness of classes, the uniqueness of classrooms, the matching of classroom capacity and type, and prerequisite course rules. Soft constraints include course uniformity, the same class on different days, minimizing the walking distance between teachers and students, and morning and evening preferences. It also supports the scenario-based configuration of constraint weights. The simulated annealing optimization engine module connects the data and knowledge layer module and the constraint and objective engine module. It is used for timetable optimization based on the simulated annealing algorithm and includes: The initial solution construction unit is used to generate an initial timetable solution that satisfies all hard constraints based on a heuristic greedy algorithm and a local repair algorithm. The multi-neighborhood search unit is used to perform one or more neighborhood operations during the simulated annealing iteration process, including time slot swapping, time slot shifting, classroom replacement, cross-building swapping, grouping, or splitting. The annealing scheduling unit is used to manage the initial temperature setting, segmented cooling and reheating strategies, and to control the acceptance and rejection of candidate solutions based on the Metropolis criterion. The convergence fine-tuning unit is used to fine-tune the unqualified regions by calling a local metaheuristic algorithm after the iteration meets the stopping condition; The business collaboration module connects to the simulated annealing optimization engine module to respond to external business change requests, including curriculum version updates, classroom application approvals and temporary class rescheduling, and triggers the optimization engine to perform global optimization or incremental rearrangement based on the influence domain. The interface and visualization module is used to output scheduling results, visualize conflicts and resource status, and provide APIs for data integration with external academic affairs systems.

[0020] In this embodiment, the data and knowledge layer module is the basic data hub of the intelligent academic affairs management and automatic course scheduling system based on the simulated annealing algorithm. Its core function is to store and manage all the core data required for course scheduling, and generate teaching task quintuples through a standardized process. This provides accurate and conflict-free data input for subsequent modules such as constraint calculation and timetable optimization, ensuring that the entire course scheduling process is reliable and effective from the data source. The Data and Knowledge Layer module is primarily responsible for storing and managing five types of core data: curriculum data, faculty data, classroom data, semester time models, and historical timetable data. Based on data management, this module gradually constructs five-tuples of teaching tasks, including course, teacher, class, classroom, and time, through hierarchical processing and association matching. During the generation process, the module performs multiple rounds of conflict checks and adjustments to ensure that the final output five-tuples of teaching tasks conform to the basic logic of timetable scheduling, without data conflicts or logical contradictions. This provides core data support for the hard and soft constraint calculations of the Constraint and Target Engine module, the initial timetable solution construction of the Simulated Annealing Optimization Engine module, and subsequent optimizations. It is a prerequisite for the normal operation of the entire timetable system. The Data and Knowledge Layer module includes: The curriculum is linked to the class unit: Training program data analysis: Receive and store training program data for each major and grade, extract key information such as course name, course nature, semester offered, and applicable classes, clarify the class range that each course needs to be offered in a specific semester, and establish the correspondence between courses and classes; Initial task set generation: Based on the course-class correspondence obtained from the parsing, an initial task set containing only two elements, course and class, is generated; each item in the task set is clearly labeled with course information and corresponding class information, forming the basic framework of the scheduling task; Teacher allocation unit: Faculty Data Management: Stores basic information about teachers, including their name, employee number, department, list of courses they can teach, time availability, etc., to ensure that faculty data is updated in real time and accurately reflects teachers' teaching qualifications and time availability. Teacher matching and task subset construction: For each task in the initial task set, qualified teachers are selected based on faculty data; the selection criteria include whether the teacher is qualified to teach the course and whether the teacher has available time during the course's scheduled period; after assigning qualified teachers to each task, the initial task set is expanded into a course teacher class task subset that includes the three elements of course teacher class. Based on classroom data, including classroom capacity and type, a matching classroom is assigned to each task in the course-teacher-class task subset, forming a course-teacher-class-classroom task extension set; Classroom matching unit: Classroom Data Management: Stores detailed information about classrooms, including classroom number, building location, capacity type, equipment configuration, usage restrictions, etc. Regularly checks classroom status to ensure that information such as classroom capacity type and equipment configuration is consistent with reality; Classroom allocation and task extension set construction: Based on the course teacher's class task subset, a matching classroom is assigned to each task in combination with course requirements and classroom data; the matching rules prioritize the matching of classroom capacity with class size, i.e., classroom capacity is not less than class size; secondly, the matching of classroom type with course requirements is required, such as assigning a laboratory to experimental classes and a multimedia classroom to multimedia courses; at the same time, it is ensured that there are no usage restrictions or conflicts with the classroom during the course schedule; after the allocation is completed, the task subset is expanded into a course teacher class classroom task extension set containing the four elements of course teacher, class, and classroom. Time period allocation unit: Semester Time Model Management: Stores semester time model data, which clearly defines the start and end dates of the semester, the number of teaching weeks, the number of teaching days per week, the division of daily teaching periods, holidays and class suspension periods, etc., providing a standard time framework for time allocation; Time allocation and preliminary quintuple generation: Using the semester time model, specific time slots are allocated to each task in the course teacher class classroom task extension set; during the allocation process, immediate conflicts are initially avoided, that is, the same teacher or the same class is not assigned multiple tasks in the same time slot; after the allocation is completed, a preliminary teaching task quintuple containing the five elements of course teacher class classroom time is generated. Historical data verification and adjustment unit: Historical timetable data management: Stores timetable data from previous semesters, including the five-tuple of history teaching tasks, records of history scheduling conflicts, and historical resource usage, providing a reference for conflict checking in the current scheduling. Conflict checking and five-tuple adjustment: Refer to historical timetable data to check for conflicts in the preliminary teaching task five-tuples; the check includes whether there are conflicts with fixed tasks that have not been adjusted in the historical timetable, and whether there are resource allocation combinations that have frequently caused problems in the past; if a conflict is detected, the preliminary teaching task five-tuples are adjusted, and the adjustment methods include changing the time period, changing the classroom or rematching the teacher, and finally outputting a conflict-free final teaching task five-tuple. The workflow of the data and knowledge layer module is as follows: Initialization phase: After the module starts, it first establishes connections with the curriculum database, faculty database, classroom database, semester time model database, and historical timetable database, and verifies the connection status and data readability of each database. Load basic data from various databases, perform format and integrity checks on the data, mark data with missing key information and report back to relevant management personnel to ensure that the basic data for subsequent data processing is complete and usable; Initial task set generation phase: The system calls up the training program and class-related units, reads the training program data, and extracts the correspondence between courses and classes through structured parsing technology. Based on the extracted correspondence, an initial task set containing course and class information is generated, and a unique identifier is assigned to each task item for easy subsequent tracking and management. Course teacher class task subset generation stage: Call the teacher allocation unit, read the teacher data, and for each task in the initial task set, filter teachers according to the order of their teaching qualification time availability. Match each task with a qualified teacher, expand the initial task set into a subset of course teacher class tasks that includes course teacher class information, and store it in a temporary data storage area. Course teacher class classroom task extension set generation stage: Call the classroom matching unit, read classroom data, and combine course requirements with classroom attributes to assign a matching classroom to each task in the course teacher's class task subset; Integrate classroom information into task subsets to form an extended set of course teacher class classroom tasks that includes course teacher class classroom information, and update temporary data storage content. Preliminary teaching task: Five-tuple generation stage Call the time period allocation unit, read the semester time model data, and build a time resource pool; Assign specific time slots to each task in the course teacher's classroom task extension set to avoid immediate conflicts and generate a preliminary teaching task quintuple. Final teaching task: Five-tuple output stage Call the historical data verification and adjustment unit, read the historical timetable data, and perform conflict checks on the preliminary teaching task quintuple; Adjustments are made to the detected conflicts, and after confirming that there are no conflicts, the final teaching task quintuple is output and synchronized to the constraint and objective engine module and the simulated annealing optimization engine module. Archiving stage: The final teaching task quintuple generated this time will be archived and stored in the designated database, recording the generation time and associated scheduling cycle; Disconnect from each database, release system resources, generate a data processing report, and provide feedback on information such as the number of tasks, conflicts, and adjustments made during this data processing.

[0021] In this embodiment, the constraint and objective engine module is a key evaluation module of the intelligent academic affairs management and automatic course scheduling system based on the simulated annealing algorithm. Its core function is to define the hard and soft constraint rules in the course scheduling process, calculate the degree to which the timetable solution satisfies these constraints, generate the overall objective function value of the timetable solution, and provide a judgment basis for the iterative optimization of the simulated annealing optimization engine module, so as to ensure that the course scheduling result not only meets the basic feasibility requirements, but also takes into account the teaching experience and resource utilization efficiency. The Constraint and Objective Engine module is primarily responsible for two core tasks: first, defining the hard and soft constraints required for timetable scheduling, forming a standardized set of constraint rules; second, for the timetable solutions input from the Simulated Annealing Optimization Engine module, performing hard constraint conflict detection and soft constraint evaluation, and calculating the overall objective function value of the timetable solution. Hard constraints are the fundamental rules that timetable scheduling must satisfy; if violated, the timetable solution is infeasible. Soft constraints are optimization rules that improve the quality of the timetable, reflecting the merits of the timetable solution through weighted evaluation. The module also supports scenario-based configuration of constraint weights, allowing adjustment of soft constraint weights according to the scheduling needs of different departments and semesters, making the evaluation results more closely aligned with actual business scenarios. The final output of the overall objective function value is directly used for iterative evaluation by the Simulated Annealing Optimization Engine, guiding the optimization direction of the timetable solution. This includes: Hard constraint rule management unit: Hard constraint rules are defined as follows: five types of rules that must be strictly followed during the course scheduling process. These are: the uniqueness of teachers in the same time slot (a teacher can only teach one course in the same time slot), the uniqueness of classes in the same time slot (a class can only teach one course in the same time slot), the uniqueness of classrooms in the same time slot (a classroom can only be used for one course in the same time slot), the matching rule between classroom capacity and type (the classroom capacity must not be less than the number of students in the class and the classroom type must meet the course requirements), and the prerequisite course rule (prerequisite courses must be scheduled before subsequent courses). Construction of hard constraint rule set: Organize the five defined hard constraint rules into a structured hard constraint rule set, clarify the judgment criteria and detection logic of each type of rule, and ensure that the rules can be recognized by the system and used for subsequent conflict detection; Hard constraint rule modeling transforms the basic feasibility requirements in timetable scheduling into computable logical rules. For each type of hard constraint, a corresponding logical judgment model is constructed. For example, the model for the uniqueness of teachers in the same time period will traverse all teaching tasks in the timetable solution for the same time period to check whether there is duplicate assignment of the same teacher. The model for the classroom capacity and type matching rule will compare the number of students in the class with the classroom capacity and the course type with the classroom type in the teaching task to determine whether they match. Through the step-by-step verification of the logical judgment model, the feasibility of the timetable solution is quickly identified, ensuring that the timetable results meet the most basic academic management requirements. Soft constraint rule management unit: Soft constraint rule definition: Four types of optimization rules are determined to improve the quality of the timetable, namely, the course uniformity rule, that is, the courses of the same class are evenly distributed within a week; the same course on different days rule, that is, the multiple teaching of the same course is arranged on different dates; the teacher and student walking distance minimization rule, that is, the teacher or class walks the shortest distance between classrooms between consecutive courses; and the morning and evening preference rule, that is, according to the needs of teachers or classes, the morning and evening time courses are prioritized or avoided. Construction of a set of soft constraint rules: The four types of defined soft constraint rules are organized into a structured set of soft constraint rules, and the evaluation dimensions and violation degree calculation methods of each type of rule are clarified, so as to provide a standardized basis for subsequent soft constraint evaluation; The quantitative assessment of soft constraints transforms the qualitative requirements of timetable quality into quantifiable evaluation indicators. For each type of soft constraint, a corresponding quantitative calculation method is designed. For example, the course uniformity rule quantifies uniformity by calculating the variance of the number of courses per day for the same class within a week (the smaller the variance, the higher the uniformity and the lower the degree of violation); the rule for minimizing the walking distance between teachers and students quantifies distance by calculating the difference in classroom location coordinates for consecutive courses (the smaller the coordinate difference, the closer the distance and the lower the degree of violation). Then, by assigning weights, the degree of violation of soft constraints in different dimensions is integrated into a unified total evaluation score, thereby achieving a quantitative measurement of timetable quality. Constraint weight scenario-based configuration unit: Initial configuration of constraint weights: Based on the needs of typical course scheduling scenarios, initial weights are set for each type of soft constraint rule. The weight reflects the importance of the soft constraint in the overall timetable quality assessment. For example, the course uniformity weight of core courses can be higher than that of non-core courses. Dynamic adjustment of constraint weights: Supports adjusting soft constraint weights according to different business scenarios. For example, science and engineering departments may pay more attention to classroom type matching for experimental courses (the weight of minimizing the walking distance between teachers and students can be adjusted accordingly), while humanities departments may pay more attention to the course uniformity weight. The module can receive scenario change instructions, update the soft constraint weight configuration, and apply it to subsequent evaluations. This technology constructs a weight adjustment mechanism based on scenario-based needs. The module presets weight templates for different scenarios. For example, in the "Department Key Course Template," course uniformity and morning / evening preference have higher weights, while in the "Laboratory Course Template," minimizing the walking distance between teachers and students has higher weights. It also supports manual adjustment of individual soft constraint weights. After adjustment, the system automatically updates the weight configuration file and applies the new weights to subsequent soft constraint evaluation calculations. This technology ensures that constraint evaluation can adapt to the scheduling needs of different business scenarios and improves the practicality of the evaluation results. Constraint evaluation calculation unit: Hard constraint conflict detection: Receives the timetable solution input from the simulated annealing optimization engine module, and checks the allocation of each teaching task in the timetable solution one by one according to the set of hard constraint rules; if any hard constraint rule is detected to be violated, such as the same teacher being assigned two courses in the same time period, the timetable solution is marked as infeasible; if no hard constraint rule is violated, the timetable solution is marked as feasible. Calculation of the degree of violation of soft constraints: For a timetable solution marked as feasible, the degree of violation of each type of soft constraint is calculated based on the set of soft constraint rules and the currently configured soft constraint weights; for example, the degree of violation of the course uniformity rule can be measured by the difference in the number of courses per day in the same class, and the degree of violation of the rule of the same course on different days can be measured by the number of times the same course is scheduled on the same day; Aggregation of soft constraint evaluation scores: Multiply the degree of violation of each type of soft constraint by its corresponding weight to obtain the weighted degree of violation of each type of soft constraint. Then sum the weighted degree of violation of all soft constraints to obtain the total soft constraint evaluation score of the timetable solution. The lower the total score, the higher the degree of satisfaction of the timetable solution with the soft constraints. Overall objective function value generation: The overall objective function value is generated by combining the hard constraint conflict detection results and the soft constraint evaluation total score; if the timetable solution is infeasible, an extremely high objective function value is assigned to indicate its unavailability; if the timetable solution is feasible, the soft constraint evaluation total score is used as the core parameter to generate an objective function value that reflects the overall quality of the timetable solution. This technology combines hard constraint feasibility with soft constraint quality to construct an objective function that guides optimization. For infeasible solutions, an objective function value far exceeding that of feasible solutions is assigned, causing them to be prioritized for elimination in simulated annealing iterations. For feasible solutions, the objective function value is constructed with the total score of soft constraint evaluation as the core; the lower the total score, the smaller the objective function value, indicating a higher quality solution. The logic for constructing the objective function value directly connects to the iteration logic of the simulated annealing optimization engine, providing a clear numerical basis for the engine to determine whether to accept candidate solutions. The workflow of the constraint and objective engine module includes: Initialization phase: After the module starts, it loads the preset set of hard constraint rules and soft constraint rules, and clarifies the judgment criteria and calculation methods for each type of rule; Based on the current course scheduling scenario, load the corresponding soft constraint weight configuration. If adjustments are needed, update the weights through the constraint weight scenario configuration unit to ensure that the weights meet the current business requirements. Establish a data connection with the simulated annealing optimization engine module to prepare for receiving the timetable solution data input by the module; Hard constraint conflict detection phase: Receive the solution to be evaluated from the simulated annealing optimization engine module; The logical judgment model of the hard constraint rule management unit is invoked to check whether each teaching task in the timetable solution violates the hard constraint rule; If any hard constraint rule is detected to be violated, the timetable solution is marked as infeasible; if no hard constraint rule is violated, the timetable solution is marked as feasible. Soft constraint assessment phase: Soft constraint evaluation is performed only on timetable solutions marked as feasible, and the quantitative calculation method of the soft constraint rule management unit is invoked to calculate the degree of violation of each type of soft constraint; By combining the currently configured soft constraint weights, the degree of violation of each type of soft constraint is multiplied by the corresponding weight to obtain the weighted degree of violation of each type of soft constraint; Summarize the weighted violation rates of all soft constraints to obtain the total soft constraint evaluation score for the solution to this timetable; Overall objective function value calculation stage: If the timetable solution is deemed infeasible, assign it an extremely high overall objective function value to indicate that it does not meet the basic timetable requirements. If the timetable solution is feasible, the total score is evaluated with soft constraints as the core, and an overall objective function value that reflects the overall quality of the timetable solution is generated. The lower the total score, the smaller the objective function value. Result output stage: The feasibility indicator of the timetable solution and the overall objective function value are synchronized to the simulated annealing optimization engine module, providing a basis for the module to judge whether the candidate solution is accepted and to guide iterative optimization; If the scheduling scenario changes in the future, the system receives a scenario change instruction, updates the soft constraint weights through the constraint weight scenario configuration unit, and returns to the initialization phase to reload the configuration, ensuring that subsequent evaluations are adapted to the new scenario.

[0022] In this embodiment, the simulated annealing optimization engine module is the core optimization unit of the intelligent academic affairs management and automatic course scheduling system based on the simulated annealing algorithm. One end of the module is connected to the data and knowledge layer module to obtain the five-tuple of teaching tasks, and the other end is connected to the constraint and target engine module to obtain the constraint evaluation results. Through the full-process optimization of the simulated annealing algorithm, starting from generating an initial timetable solution that satisfies the hard constraints, through multiple rounds of iterative search and temperature scheduling, a high-quality timetable solution is finally output through local fine-tuning, providing optimized course scheduling data support for the business collaboration module. The core task of the simulated annealing optimization engine module is to optimize the timetable solution based on the simulated annealing algorithm. It generates an initial timetable solution that satisfies all hard constraints through an initial solution construction unit, then generates diverse candidate timetable solutions during the iteration process through a multi-neighborhood search unit, controls the acceptance and rejection of candidate solutions through temperature management and the Metropolis criterion, balances the exploration and utilization of solutions to avoid local optima, and finally performs local optimization on the unmet requirements after the iteration stops, further improving the quality of the timetable solution. These four units work collaboratively to form a complete optimization chain of "initial solution generation - iterative optimization - local fine-tuning," ensuring that the final output timetable solution satisfies all hard constraints and maximizes compliance with soft constraints. This includes: Initial solution for constructing units: Preliminary timetable generation: Based on the five-tuple of teaching tasks output by the data and knowledge layer module, a heuristic greedy algorithm is used to prioritize the allocation of time slots and classroom resources for each teaching task. The allocation process follows resource availability rules, including matching classroom capacity with class size, prioritizing time slots with high teacher availability, and avoiding course conflicts between classes in the same time slot, thereby generating a preliminary timetable solution. Hard constraint conflict detection: For the generated preliminary timetable solution, call the hard constraint rules of the constraint and target engine module to perform conflict detection, identify the allocation positions that violate the rules of uniqueness of teachers, uniqueness of classes, uniqueness of classrooms, classroom capacity and type matching or prerequisite courses, and mark the conflict points. Local patching and adjustment: For the identified hard constraint conflict points, the local patching algorithm is applied for iterative adjustment; the adjustment methods include reallocating time slots for conflicting tasks, exchanging classroom resources for conflicting tasks, and modifying the arrangement order of conflicting tasks. Each adjustment only repairs a single conflict point; Hard constraint verification and initial solution output: During the local repair process, the hard constraint satisfaction status of the timetable solution is continuously checked; if there are unrepaired conflicts, the conflict detection and local repair steps are repeated; until all hard constraints are satisfied, the timetable solution is output as the initial timetable solution and passed to the annealing scheduling unit. The initial solution construction unit combines a heuristic greedy algorithm and a local patching algorithm to ensure the feasibility of the initial solution. The heuristic greedy algorithm prioritizes the allocation of high-quality resources based on resource availability rules to quickly build the initial solution framework. The local patching algorithm iteratively detects and repairs hard constraint conflicts, specifically resolves single conflict points, avoids overall solution reconstruction, and efficiently generates an initial timetable solution that satisfies all hard constraints, providing a qualified starting foundation for subsequent simulated annealing iterations. Annealing scheduling unit: Initial temperature setting: Based on the initial schedule solution output by the initial solution construction unit, the constraint and objective engine module is called to calculate its soft constraint evaluation total score, and the initial temperature value is calculated by random perturbation sampling; the calculation process must ensure that the acceptance probability of the candidate solution corresponding to the initial temperature is higher than the preset threshold, so as to provide a basis for the "exploration" capability of subsequent iterations; Initial state initialization: Set the current temperature value to the calculated initial temperature value, and initialize the current timetable solution to the initial timetable solution output by the initial solution construction unit, thus completing the initial preparation for optimization iteration; Segmented Cooling Cycle Start-up: The entire optimization process is divided into multiple temperature stages, and a segmented cooling cycle is entered; a predetermined number of iterations are performed within each temperature stage, and the single iteration process is as follows: receive the new candidate timetable solution output by the multi-neighborhood search unit, call the constraint and objective engine module to calculate the objective function value of the new candidate solution, compare it with the objective function value of the current timetable solution, and obtain the change in the objective function value; Metropolis Criterion Processing: The Metropolis criterion is applied based on the change in the objective function value. If the change in the objective function value is less than zero, it indicates that the new candidate solution is of higher quality, and the new candidate solution is directly accepted as the current timetable solution. If the change in the objective function value is greater than or equal to zero, the new candidate solution is accepted as the current timetable solution based on the probability calculated based on the change in the objective function value and the current temperature value. Temperature adjustment and rewarming: After the predetermined number of iterations for the current temperature stage is completed, the current temperature value is reduced according to the segmented cooling strategy. The segmented cooling strategy sets a different cooling rate for each temperature stage, and the cooling rate decreases as the temperature stage increases. After the temperature is reduced, the convergence status is monitored. If it is detected that the objective function value of the current course solution has not improved within multiple consecutive temperature stages, the rewarming strategy is triggered to reset the current temperature value to the starting temperature value of the previous temperature stage or a higher temperature value. Loop Termination and Solution Output: Repeat the segmented cooling loop until the current temperature value is lower than the preset termination temperature threshold or the maximum number of iterations is reached, then stop the loop and output the current timetable solution to the convergence fine-tuning unit; The annealing scheduling unit balances the "exploration" and "utilization" of solutions through temperature management and the Metropolis criterion. In the initial high-temperature stage, it accepts more poor candidate solutions to achieve a wide-range search. As the temperature decreases, the probability of accepting poor solutions gradually decreases, and it gradually focuses on the region of high-quality solutions, that is, "utilizing" existing high-quality solutions. The segmented cooling strategy rapidly explores in the high-temperature stage and finely optimizes in the low-temperature stage by decreasing the cooling rate. The reheating strategy raises the temperature again when the search stalls to restore the exploration capability and ensure that the optimization process does not get stuck in local optima. Convergence fine-tuning unit: Non-compliant area identification: Receive the current timetable solution output by the annealing scheduling unit, call the constraint and objective engine module to calculate its total soft constraint evaluation score and the degree of violation of each soft constraint, identify the areas that do not meet the soft constraint requirements or whose objective function values ​​do not meet the standards, and determine these areas as non-compliant areas; Local metaheuristic algorithm selection: Select at least one local metaheuristic algorithm based on the characteristics of the non-compliant area; the available algorithms include local search algorithms, greedy algorithms, or variable neighborhood search algorithms. The selection criteria include the characteristics of conflict type and resource distribution within the area. For example, a local search algorithm can be selected for areas where the course uniformity does not meet the standard. Iterative fine-tuning execution: The selected local metaheuristic algorithm is applied to iteratively fine-tune the non-compliant areas; during the fine-tuning process, only the teaching tasks within the area are adjusted, without changing the task arrangements outside the area, and the degree of soft constraint violation is reduced by adjusting time periods, classroom resources, etc. Hard constraint verification: For the fine-tuned timetable solution, call the constraint and target engine module again to perform hard constraint verification to ensure that the fine-tuning operation does not violate any hard constraint rules. If a hard constraint conflict occurs, revert to the state before fine-tuning and reselect the fine-tuning strategy. Evaluation and Result Determination: Based on the hard constraint verification results, calculate the total soft constraint evaluation score of the fine-tuned timetable solution and compare it with the total soft constraint evaluation score of the timetable solution before fine-tuning. If the total soft constraint evaluation score after fine-tuning is better, then accept the fine-tuned timetable solution as the final timetable solution; if it is not better, then revert to the timetable solution before fine-tuning and try other local metaheuristic algorithms to fine-tune it again until the optimal result is obtained. The convergence fine-tuning unit achieves refined optimization of the solution based on the local metaheuristic algorithm. After the iteration stops, an adaptation algorithm is selected for the local features of the unqualified area, and only the local area is adjusted to avoid fluctuations in the overall solution. The hard constraint verification step ensures that the fine-tuning does not destroy the feasibility of the solution. The optimization effect is determined by comparing the total score with the soft constraints, which further improves the degree to which the timetable solution meets the soft constraints, making the final solution more in line with the actual teaching needs. The workflow of the simulated annealing optimization engine module includes: Initialization and data reception phase: After the module starts, it establishes connections with the data and knowledge layer module and the constraint and goal engine module, receives the teaching task quintuple output by the data and knowledge layer module, and obtains the hard and soft constraint rules and evaluation methods of the constraint and goal engine module. The initial solution construction unit is invoked, and the teaching task quintuple is used as input to start the initial timetable solution generation process; Initial timetable generation stage: The initial solution construction unit uses a heuristic greedy algorithm to allocate time slots and classroom resources to generate a preliminary timetable solution; Hard constraint conflict detection is performed on the preliminary timetable solution, conflict points are marked, and a local repair algorithm is applied for iterative repair. Once all hard constraints are satisfied, output the initial timetable solution to the annealing scheduling unit. Simulated annealing iterative optimization stage: The annealing scheduling unit calculates the initial temperature value, initializes the current temperature and current timetable solution, and starts the segmented cooling cycle; Within each temperature stage, the process iterates a predetermined number of times: the multi-neighborhood search unit randomly selects the neighborhood operation type, identifies the operation object, and generates a new candidate timetable solution; the annealing scheduling unit calculates the change in the objective function value between the new candidate solution and the current solution, and determines whether to accept the new candidate solution according to the Metropolis criterion. Once a single temperature stage iteration ends, the temperature is reduced according to a segmented cooling strategy; the convergence status is monitored, and reheating is triggered if necessary. Repeat the segmented cooling cycle until the stopping condition is met, and output the current timetable solution to the convergence fine-tuning unit. Local fine-tuning and final solution output stage: The convergence fine-tuning unit identifies the unsatisfactory regions of the current timetable solution and selects a local metaheuristic algorithm for fine-tuning; After fine-tuning, perform hard constraint verification, calculate the total score for soft constraint evaluation, and compare it with the score before fine-tuning. Accept a better solution or roll back and retry until a final timetable solution that meets the requirements is obtained, and output it to the business collaboration module.

[0023] In this embodiment, the business collaboration module is the dynamic business processing unit of the intelligent academic affairs management and automatic course scheduling system based on the simulated annealing algorithm. One end of the module is connected to the simulated annealing optimization engine module, and the other end is connected to the external academic affairs business change requirements. Its core function is to receive and parse the external business change request, determine the scope of the change, and trigger the simulated annealing optimization engine to perform global optimization or incremental rescheduling based on the influence domain, so as to ensure that the scheduling results can adapt to the dynamic adjustment requirements in academic affairs management in real time and maintain the effectiveness and rationality of the timetable. The business collaboration module is primarily responsible for responding to three types of external business change requests: curriculum version updates, classroom application approvals, and temporary class rescheduling. The module analyzes the request type and content to determine the scope of impact: if the change involves multiple teaching tasks or global resource adjustments, it triggers the simulated annealing optimization engine module to perform global optimization; if the change only affects local tasks or resources, it calculates the impact domain and triggers the engine to perform incremental rescheduling based on the impact domain. Throughout this process, the module ensures that the optimized timetable results are promptly output to the interface and visualization module, while maintaining collaboration with external academic affairs business processes. This allows the scheduling system to flexibly respond to dynamic changes in academic affairs management, preventing timetable invalidation due to business changes. This includes: Change request receiving and parsing unit: Change Request Reception: Receive business change requests initiated by external academic affairs systems or administrators. Request forms include electronic application forms submitted within the system, change instructions transmitted through API interfaces, etc., to ensure that all external change requests can be captured in a timely manner. Request Type and Content Analysis: The received change requests are classified and analyzed to identify whether the request type is a classroom application approval for curriculum version update or a temporary class rescheduling; at the same time, the key content in the request is extracted, such as the course adjustment information involved in the curriculum update, the time and classroom number of the classroom application, and the course teacher classes and adjustment time periods involved in the temporary class rescheduling, so as to clarify the specific objects and requirements of the change. Unit for determining the scope of impact of changes: Analysis of influencing factors: Based on the parsed request content, analyze the teaching tasks and resources that may be affected by the changes. For example, updating the curriculum version may involve the addition or deletion of multiple courses or adjustments to class hours, while temporary class rescheduling may involve changes to the time or classroom of a single course. Impact Scope Determination: Determine the impact scope of the change request based on the identified influencing factors; if the change involves multiple teaching tasks (such as curriculum updates leading to adjustments in multiple courses) or requires global resource adjustments (such as changes to classroom usage rules across the entire school), then global optimization needs to be triggered; if the change only affects a single or a small number of related teaching tasks (such as a teacher temporarily rescheduling a single course), then incremental rearrangement based on the impact domain needs to be triggered. The change request structured parsing technology uses preset request type templates to perform structured parsing of different types of business change requests. For curriculum version update requests, the template includes fields such as adding, deleting, adjusting class hours, and changing applicable classes. For classroom application approval requests, the template includes fields such as the requested classroom number, the requested time slot, and the requested purpose. For temporary class rescheduling requests, the template includes fields such as course identifier, teacher identifier, class identifier, original time slot, and new time slot. By extracting key information through field matching, unstructured request content is transformed into structured data that the system can recognize, providing a clear basis for subsequent assessment of the scope of impact. Influence domain calculation unit: Directly Affected Task Identification: For change requests determined to be incremental rearrangements, identify the directly affected teaching tasks based on the content of the change. For example, the courses involved in temporary class rescheduling are directly affected tasks, and the original courses corresponding to the classrooms applied for in classroom application approval are directly affected tasks. Indirect Influence Task Tracing: Starting from the directly influencing task, trace other related teaching tasks. For example, if the teacher of the directly influencing task has other courses in the adjusted time period, then that course is an indirectly influencing task; if the class of the directly influencing task has other courses in the adjusted time period, then that course is also an indirectly influencing task. Influence domain determination: Integrate directly and indirectly affected tasks, identify the courses, teachers, classes, classrooms, or time periods that are directly or indirectly affected by the changes, and form a complete influence domain to provide a basis for adjustment scope in subsequent incremental rearrangement; The influence domain calculation unit is based on the relationship model between teaching tasks and resources to achieve dynamic tracing of the influence domain. The model stores multi-dimensional relationships such as course-teacher-class teacher-time class-time classroom-time. When a change occurs, the model traces back layer by layer through the relationship chain, starting from the changed object. For example, in the case of a temporary class rescheduling, starting from "course A (original time slot T1)," the model traces back other courses of teacher X in time slot T1 through the relationship chain "course A-teacher X", other courses of class Y in time slot T1 through the relationship chain "course A-classroom C", and other courses of classroom C in time slot T1 through the relationship chain "course A-classroom C". Finally, all related objects are integrated to form the influence domain, ensuring that the influence domain is complete and does not expand. Optimize triggering and execution units: Global optimization trigger: When it is determined that global optimization needs to be performed, a global optimization command is sent to the simulated annealing optimization engine module, transmitting the complete data after the changes (such as updated training program data and adjusted resource data), triggering the engine to re-optimize the entire timetable to ensure that all teaching tasks and resource allocations meet the changed requirements. Incremental Rearrangement Trigger: When incremental rearrangement is determined to be required, an incremental rearrangement command is sent to the simulated annealing optimization engine module. The influence domain information is calculated synchronously, and the engine only optimizes and adjusts the teaching tasks within the influence domain, while keeping the teaching tasks outside the influence domain unchanged, thus avoiding large fluctuations in the overall timetable. By setting influence domain boundary markers, the simulated annealing optimization engine can achieve precise control over the adjustment range. When incremental rearrangement is triggered, teaching tasks within the influence domain are marked as "adjustable," and teaching tasks outside the influence domain are marked as "non-adjustable." During the optimization process, the simulated annealing optimization engine only reallocates classroom resources for the time periods of "adjustable" tasks, while keeping the original allocation unchanged for "non-adjustable" tasks. At the same time, it ensures that the adjustment of "adjustable" tasks does not affect the normal arrangement of "non-adjustable" tasks, avoiding chaos in the overall timetable due to local adjustments, and achieving precise boundary control of the optimization range. Optimization process monitoring: During the optimization process executed by the simulated annealing optimization engine, the system receives real-time feedback on the optimization progress to ensure the optimization process proceeds normally; if an optimization anomaly occurs (such as failure to complete for an extended period of time), an alert is issued and the system assists the engine in resuming the optimization process; Optimization result output unit: Optimization result reception: Receive the optimized timetable results output by the simulated annealing optimization engine module, including the complete timetable after global optimization or the local timetable adjustment content after incremental rearrangement; Result verification: Perform basic verification on the received optimized timetable results to confirm that the results meet the hard constraints (such as no conflicts between teachers or classrooms at the same time) and ensure the feasibility of the results. Result forwarding: Output the optimized timetable results that have passed the verification to the interface and visualization module for subsequent display of timetable results, generation of conflict reports and integration of data from external academic affairs systems; This technology ensures that the complete data after the changes can be accurately synchronized to the simulated annealing optimization engine, providing a foundation for global optimization. When global optimization is triggered, the module first integrates all the core data after the changes (such as updated curriculum data, adjusted teacher data, and classroom data) to form a complete updated dataset. Then, the updated dataset is transmitted to the simulated annealing optimization engine module through the data synchronization interface, overwriting the engine's original basic data. Finally, the engine is triggered to rebuild the initial timetable solution based on the updated dataset and perform full-process optimization, ensuring that the global optimization results can fully adapt to the changed business requirements. The workflow of the business collaboration module includes: Initialization phase: After the module starts, it establishes connections with the simulated annealing optimization engine module and the interface with the visualization module, and verifies the connection status to ensure normal data transmission. Load the preset change request type templates (training program version update template, classroom application approval template, temporary class rescheduling template), initialize the association model required for impact domain calculation, and complete the preparatory work before business processing; Change request reception and parsing phase: All business change requests received from external sources, whether they are application forms submitted within the system or instructions transmitted through API interfaces, are uniformly connected to the request processing queue. The corresponding template is invoked according to the request type, the request content is parsed in a structured manner, key information (such as the object being changed, the content being changed, and the time requirement for the change) is extracted, and a parsing report is generated. Impact Scope Assessment Stage: Based on the analysis report, we can sort out the teaching tasks and resources involved in the changes and determine whether the changes affect multiple teaching tasks or require global resource adjustments. If the impact is global (e.g., the curriculum update involves multiple courses), then global optimization is triggered; if the impact is local (e.g., a single course is temporarily rescheduled), then the impact domain calculation phase begins. Influence domain calculation stage: Starting with the object of change, trace the tasks that are directly or indirectly affected by the change based on the relationship model; Integrate all affected course teachers' classes, classrooms, or time periods to determine the scope of the impact area and generate an impact area report; Optimize the triggering and execution phases: If global optimization is required, send a global optimization command to the simulated annealing optimization engine module to synchronize the changed complete dataset and trigger the engine to re-optimize the entire timetable. If incremental rearrangement is required, send an incremental rearrangement command to the simulated annealing optimization engine module, synchronize the influence domain report, and trigger the engine to adjust only the teaching tasks within the influence domain. Monitor the optimization progress in real time, receive feedback on the optimization status from the engine, and ensure that the optimization process is completed normally; Optimization results output stage: Receive the optimized timetable results output by the simulated annealing optimization engine module, and perform hard constraint basic verification on the results; If the verification passes, the optimization results will be forwarded to the interface and visualization modules for subsequent display and data integration; if the verification fails, the results will be sent back to the engine module to re-execute the optimization. Archiving stage: Archive and store information such as the parsing report, impact domain report (incremental reordering), and optimization results of this business change request, and record the change processing time and processing results; Generate a business change processing log and feed it back to the external academic affairs management system for administrators to trace and view.

[0024] In this embodiment, the interface and visualization module is the external interaction and result presentation unit of the intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm. One end of it connects to the business collaboration module to obtain the optimized course scheduling results, and the other end displays the course scheduling information to the user through a graphical user interface. It also integrates data with the external academic affairs system through the API interface. Its core function is to transform the course scheduling results inside the system into an intuitive and interactive form, while ensuring data interoperability with the external academic affairs system, so as to ensure that the course scheduling information can be used and managed efficiently. The Interface and Visualization module is primarily responsible for four core tasks: First, receiving the final scheduling results from the Business Collaboration module; second, calling the Constraint and Target Engine module to evaluate the scheduling results using hard and soft constraints, generating conflict reports and resource status summaries; third, constructing a visualization dataset based on the scheduling results and evaluation information, converting it into a graphical user interface-renderable format, and outputting a timetable view, resource occupancy heatmap, and conflict highlighting report; fourth, providing an API interface to interact with external academic affairs systems, handling external data query and update requests, and forwarding update requests to the Business Collaboration module to trigger subsequent optimizations. Through the process of "result reception - evaluation - visualization - external interaction," the module achieves effective output of scheduling results and data collaboration between systems, meeting users' needs for viewing and integration with external systems; including: Scheduling Results Reception and Evaluation Unit: Course scheduling result reception: Receive the final course scheduling result from the business collaboration module. This result includes course teacher class classroom time allocation information for all teaching tasks, ensuring complete acquisition of optimized course scheduling data; Constraint evaluation trigger: The received final scheduling results are passed to the constraint and target engine module to trigger the hard constraint and soft constraint evaluation process, check whether there are hard constraint conflicts in the scheduling results, and calculate the degree of violation of soft constraints; Evaluation Result Acquisition and Organization: Acquire the evaluation results returned by the constraint and objective engine module, and organize them into a conflict report and a resource status summary; the conflict report records the specific location and conflict type of hard constraint conflicts, and the resource status summary summarizes the occupancy and idle time periods of resources such as classrooms and teachers; The constraint assessment linkage is achieved through the data transfer mechanism between modules to link the scheduling results with the constraint assessment. When the scheduling result receiving and assessment unit transmits the scheduling data to the constraint and target engine module, it carries a unique result identifier to ensure that the assessment result can be accurately associated with the corresponding scheduling result. After the assessment is completed, the constraint and target engine module returns the assessment result according to the identifier. The unit organizes the corresponding conflict report and resource status summary by matching the identifier to avoid data confusion and ensure the consistency between the assessment result and the scheduling result. Visualizing dataset building blocks: Basic data integration: Integrate the time schedule and resource allocation details in the final scheduling results with the conflict location information in the conflict report to clarify the allocation information and associated conflict status of each teaching task; Visualized Data Modeling: Based on the integrated basic data, a visual dataset is constructed. This dataset must include the timeline class-course correspondence required for the timetable view, the resource usage frequency during different time periods required for the resource usage heatmap, and the conflict location identifiers and conflict reason explanations required for the conflict highlight report, ensuring the data supports subsequent visualization. This technology constructs a multi-dimensional data model based on user viewing needs. For the timetable view, a matrix model is constructed with time and class (or teacher) as two dimensions, filling teaching tasks into the corresponding dimension intersections. For the resource usage heatmap, a frequency statistics model is constructed with resources (classroom or teacher) and time as dimensions, counting the number of resource usages at each dimension intersection, using color intensity to represent frequency. For the conflict highlight report, a conflict identifier model is constructed, associating conflict locations with the dimension intersections of the timetable view, and using specific markers to associate conflict types, ensuring the visualized content intuitively reflects the data meaning. Visual presentation unit: Data format conversion: Convert the constructed visualization dataset into a format that can be rendered by a graphical user interface, such as converting timetable data into a table view format, resource usage data into a heatmap matrix format, and conflict data into a highlighted mark format; Visual content generation: Based on the converted format, three types of visual content are generated: the timetable view displays the course schedule by class or teacher, the resource usage heatmap displays the resource usage density by classroom or time, and the conflict highlight report highlights conflicting teaching tasks in the timetable view with specific icons and marks the conflict type. Visual content output: The generated timetable view resource usage heatmap and conflict highlight report are output to the graphical user interface for users to view through interactive operations, such as filtering timetables by class, zooming in to view resource usage for a certain time period, and clicking on conflict markers to view conflict details; API Interface Management Unit: Query Interface Processing: Receive data query requests from external academic affairs systems via API interface. Request content includes queries for class schedules of specific classes, teaching arrangements of specific teachers, and occupancy status of specific classrooms, etc. Extract corresponding data from the visualization dataset or final scheduling results based on the request content, organize it into a standardized response format, and return it to the external academic affairs system. Update Interface Processing: Receive data update requests from external academic affairs systems via API interface. Request content includes course time adjustments, teacher changes, classroom changes, etc.; perform format validation on update requests to ensure that the request information is complete and includes the teaching task identifier and change content required for the change. Update request forwarding: Forwards the verified data update request to the business collaboration module, triggering the business collaboration module to start the subsequent impact scope judgment and optimization engine call process, ensuring that the change requirements of the external system can trigger the internal re-optimization of the system; API data interaction adopts a standardized API protocol to achieve interaction with external academic affairs systems; the query interface is designed based on the RESTful protocol, defining unified query parameters (such as class identifier, teacher identifier, classroom identifier) ​​and response format (such as JSON format) to ensure that external systems can easily initiate queries and parse results; the update interface adopts a request verification mechanism to check the parameter integrity and format legality of received update requests to prevent invalid requests from entering the system; when forwarding requests, the interface request identifier is carried to ensure that the business collaboration module can trace the source of the request, which facilitates the feedback of subsequent optimization results.

[0025] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm, characterized in that: include: The data and knowledge layer module is used to store and manage training program data, faculty data, classroom data, semester time models and historical timetable data, and generate teaching task quintuples, which include course, teacher, class, classroom and time. The Constraints and Objectives Engine module is used to define and compute hard and soft constraints. Hard constraints include the uniqueness of teachers at the same time, the uniqueness of classes, the uniqueness of classrooms, the matching of classroom capacity and type, and prerequisite course rules. Soft constraints include course uniformity, the same class on different days, minimizing the walking distance between teachers and students, and morning and evening preferences. It also supports the scenario-based configuration of constraint weights. The simulated annealing optimization engine module connects the data and knowledge layer module and the constraint and objective engine module. It is used for timetable optimization based on the simulated annealing algorithm and includes: The initial solution construction unit is used to generate an initial timetable solution that satisfies all hard constraints based on a heuristic greedy algorithm and a local repair algorithm. The multi-neighborhood search unit is used to perform one or more neighborhood operations during the simulated annealing iteration process, including time slot swapping, time slot shifting, classroom replacement, cross-building swapping, grouping, or splitting. The annealing scheduling unit is used to manage the initial temperature setting, segmented cooling and reheating strategies, and to control the acceptance and rejection of candidate solutions based on the Metropolis criterion. The convergence fine-tuning unit is used to fine-tune the unqualified regions by calling a local metaheuristic algorithm after the iteration meets the stopping condition; The business collaboration module connects to the simulated annealing optimization engine module to respond to external business change requests, including curriculum version updates, classroom application approvals and temporary class rescheduling, and triggers the optimization engine to perform global optimization or incremental rearrangement based on the influence domain. The interface and visualization module is used to output scheduling results, visualize conflicts and resource status, and provide APIs for data integration with external academic affairs systems.

2. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The data and knowledge layer module performs the following operations when generating a five-tuple of teaching tasks: Based on the data analysis of the training program, the correspondence between courses and classes is analyzed to generate an initial task set containing courses and classes; Based on the teacher data, assign qualified teachers to each task in the initial task set, forming a course-teacher-class task subset; Based on classroom data, including classroom capacity and type, a matching classroom is assigned to each task in the course-teacher-class task subset, forming a course-teacher-class-classroom task extension set; Using the semester time model, time periods are allocated to each task in the course-teacher-class-classroom task extension set to generate a preliminary teaching task quintuple. Referencing historical timetable data, perform conflict checks and adjustments on the initial teaching task quintuples, and output the final teaching task quintuples.

3. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The constraint and objective engine module is used to define and calculate hard and soft constraints, and supports scenario-based configuration of constraint weights, including: Define the rules for the uniqueness of teachers at the same time, the uniqueness of classes, the uniqueness of classrooms, the matching rules for classroom capacity and type, and the rules for prerequisite courses, forming a set of hard constraint rules; Define the course uniformity rule, the same course on different days rule, the rule of minimizing the walking distance between teachers and students rule, and the morning / evening preference rule to form a set of soft constraint rules; For a given timetable solution, conflict detection is performed based on the set of hard constraint rules. If any hard constraint rule is detected to be violated, the timetable solution is marked as infeasible. For the timetable solutions marked as feasible, the degree of violation of each soft constraint rule is calculated based on the weighted soft constraint rule set, and the weighted violation degrees are aggregated to obtain the total soft constraint evaluation score; Based on the results of the hard constraint checking step and the soft constraint evaluation step, the overall objective function value of the timetable solution is output for the iterative evaluation of the simulation annealing optimization engine.

4. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The initial solution construction unit is used to generate an initial timetable solution that satisfies all hard constraints based on a heuristic greedy algorithm and a local repair algorithm, including: Based on the teaching task quintuples output by the data and knowledge layer module, a heuristic greedy algorithm is used to prioritize the allocation of time slots and classroom resources for each teaching task, generating a preliminary timetable solution. The heuristic greedy algorithm allocates resources based on resource availability rules, including classroom capacity matching, teacher time availability, and avoidance of class time conflicts. Hard constraint conflict detection is performed on the preliminary timetable solution to identify allocation positions that violate the rules of teacher uniqueness, class uniqueness, classroom uniqueness, classroom capacity and type matching, or prerequisite courses. For the identified hard constraint conflicts, a local patching algorithm is applied to iteratively adjust the conflict location, including reallocating time slots, changing classrooms, or modifying the task order, in order to eliminate individual conflicts. During the local repair process, the hard constraint satisfaction status of the timetable solution is continuously checked. If there are unrepaired conflicts, the conflict detection and local repair steps are repeated until all hard constraints are satisfied. The output is an initial timetable solution that satisfies all hard constraints, which is then used as input to the simulated annealing optimization engine.

5. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The multi-neighbor search unit is used to include, during the simulated annealing iteration process: Based on the current timetable solution, a neighborhood operation type is randomly selected. Neighborhood operation types include time slot exchange, time slot shift, classroom replacement, cross-building exchange, grouping, or splitting. Based on the selected neighborhood operation type, identify one or more teaching tasks as operation objects from the current timetable solution; Apply the selected neighborhood operation to the object being operated on to generate a new candidate timetable solution; The new candidate timetable solution is output to the annealing scheduling unit for subsequent processing.

6. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The annealing scheduling unit is used to manage the initial temperature setting, segmented cooling and reheating strategies, and controls the acceptance and rejection of candidate solutions based on the Metropolis criterion, including: The soft-constraint evaluation total score is based on the initial timetable solution. The initial temperature value is calculated by random perturbation sampling, so that the initial acceptance probability is higher than the preset threshold. Set the current temperature value to the initial temperature value, and initialize the current timetable solution to the initial timetable solution; Entering a segmented cooling cycle, the optimization process is divided into multiple temperature stages. Each temperature stage performs a predetermined number of iterations. Each iteration includes: receiving new candidate timetable solutions from the multi-neighborhood search unit, calculating the objective function value of the new candidate timetable solution, and comparing it with the objective function value of the current timetable solution to obtain the change in the objective function value. The change in the objective function value is processed based on the Metropolis criterion. If the change in the objective function value is less than zero, a new candidate timetable solution is accepted as the current timetable solution. If the change in the objective function value is greater than or equal to zero, a new candidate timetable solution is accepted as the current timetable solution with a certain probability. The acceptance probability is calculated based on the change in the objective function value and the current temperature value. After the iteration of the current temperature stage is completed, the current temperature value is reduced according to the segmented cooling strategy. The segmented cooling strategy sets a different cooling rate for each temperature stage, and the cooling rate decreases as the temperature stage increases. After lowering the current temperature value, monitor the convergence status. If it is detected that the objective function value of the current course solution has not improved for multiple consecutive temperature stages, trigger the retemperature strategy to reset the current temperature value to the starting temperature value of the previous temperature stage or a higher temperature value. Repeat the segmented cooling cycle until the current temperature value is lower than the termination temperature threshold or the maximum number of iterations is reached, then output the current timetable solution to the convergence fine-tuning unit.

7. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The convergence fine-tuning unit is used to fine-tune the unqualified regions by calling a local metaheuristic algorithm after the iteration meets the stopping condition, including: Based on the current timetable solution output by the simulated annealing optimization engine, regions that do not meet soft constraints or whose objective function values ​​do not meet the standards are identified as non-compliant regions. For the identified non-compliant areas, at least one local metaheuristic algorithm is selected based on the area characteristics. The local metaheuristic algorithm includes local search algorithm, greedy algorithm or variable neighborhood search algorithm. The selected local metaheuristic algorithm is applied to iteratively fine-tune the unqualified areas to generate the fine-tuned timetable solution; The fine-tuned timetable solution is verified by hard constraints to ensure that all hard constraints are still satisfied. Based on the hard constraint verification results, calculate the soft constraint evaluation total score of the fine-tuned timetable solution; Based on the comparison between the total soft constraint evaluation score and the total soft constraint evaluation score of the original timetable solution, a decision is made on whether to accept the fine-tuning result. If the total soft constraint evaluation score after fine-tuning is better, the fine-tuned timetable solution is accepted as the final output; otherwise, the original timetable solution is reverted to and other fine-tuning strategies are tried.

8. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The business collaboration module is used to respond to external business change requests, including curriculum version updates, classroom application approvals, and temporary class rescheduling, and to trigger the optimization engine to perform global optimization or incremental reordering based on the influence domain, including: Receive external business change requests and parse the request type and content; Based on the parsed request type and content, determine the scope of impact of the change request. If the change request involves multiple teaching tasks or global resource adjustments, trigger global optimization; otherwise, trigger incremental rearrangement based on the impact domain. If an incremental rearrangement based on the impact domain is triggered, the impact domain is calculated based on the content of the change request. The impact domain includes courses, teachers, classes, classrooms, or time periods that are directly or indirectly affected by the change. Based on the calculated influence domain, the simulated annealing optimization engine is invoked to perform incremental rearrangement. The simulated annealing optimization engine only optimizes and adjusts the teaching tasks within the influence domain, while keeping the teaching tasks outside the influence domain unchanged. If global optimization is triggered, the simulated annealing optimization engine will be invoked to re-optimize the entire timetable; Output the optimized timetable results to the interface and visualization module.

9. The intelligent academic affairs management and automatic course scheduling system based on simulated annealing algorithm according to claim 1, characterized in that: The interface and visualization module are used to output scheduling results, visualize conflicts and resource status, and provide an API for data integration with external academic affairs systems, including: Receive the final scheduling results from the business collaboration module; Based on the received final scheduling results, the constraint and target engine module is invoked to evaluate hard and soft constraints, and generate conflict reports and resource status summaries. By utilizing conflict reports and resource status summaries, combined with the final scheduling results, a visual dataset is constructed, including timetable arrangements, resource allocation details, and conflict location information; Convert the visualized dataset into a graphical user interface renderable format to generate timetable views, resource usage heatmaps, and conflict highlight reports; The system outputs a timetable view, resource usage heatmap, and conflict highlight report through a graphical user interface for users to view interactively. The system receives data query requests from external academic affairs systems via API interfaces and returns query responses based on the final course scheduling results and visualized datasets. The system receives data update requests from external academic affairs systems via API interfaces and forwards these requests to the business collaboration module to trigger subsequent optimization processes.

Images