Intelligent invigilator distribution management method based on multi-dimensional examination room analysis

Abstract

The invention discloses an invigilation personnel intelligent distribution management method based on multi-dimensional examination room analysis. The method comprises the steps of data acquisition, data preprocessing, multi-dimensional examination room analysis, obtaining of a multi-dimensional constraint invigilation personnel distribution model, hyper-parameter search optimization and intelligent invigilation personnel distribution. The invention relates to the technical field of education informatization, in particular to an invigilation personnel intelligent distribution management method based on multi-dimensional examination room analysis, which innovatively provides a multi-dimensional examination room analysis strategy and effectively solves the problem of intelligent distribution of invigilation personnel in various scenes. A new distribution activation function and a multi-dimensional constraint distribution loss function are designed, and good balance is realized between the model precision and the calculation efficiency, so that the accuracy and robustness of the model are improved; a diversity exploration mechanism and an adaptive replacement strategy are adopted to improve an algorithm for obtaining optimal parameters of a distribution model, so that an optimal parameter combination is obtained, and the accuracy of a model result is improved.

Description

Technical Field

[0001] The present invention relates to the field of educational information technology, and in particular to an intelligent allocation and management method for invigilators based on multi-dimensional examination room analysis. Background Art

[0002] In the field of education, examinations and tests are important means of evaluating candidates' learning outcomes. Traditional invigilator allocation usually relies on manual arrangements or simple random allocation systems. With the progress of society and the growing demand for examinations, this traditional allocation method has gradually failed to meet the needs of modern education. In order to improve the efficiency and quality of invigilation work, it is necessary to continuously optimize and improve the allocation rules of invigilators; however, the traditional invigilator allocation method has technical problems such as uncontrollable invigilation sessions and inability to adapt to diversified invigilation rules, which may result in the invigilator allocation results not meeting specific invigilation requirements; the existing intelligent allocation model for invigilators has certain limitations in the activation function and loss function, which leads to deviations in the allocation results, thereby affecting the robustness and accuracy of the allocation model; the existing intelligent allocation model has inappropriate built-in parameter settings, which leads to inaccurate model prediction results. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis. In view of the technical problems of uncontrollable examination sessions and inability to adapt to diversified invigilation rules in traditional examination room invigilator allocation methods, which may result in invigilator allocation results that may not meet specific invigilation requirements, this solution innovatively proposes a multi-dimensional examination room analysis strategy. Through multi-dimensional analysis of the examination room, flexible examination room constraint rules are established, which can adjust the invigilation sessions according to actual needs and adapt to the restrictions on invigilation subjects in different scenarios, thereby effectively solving the problem of intelligent allocation of invigilators in various scenarios; in view of the limitations of activation functions and loss functions in existing models suitable for intelligent allocation of invigilators, the invigilation results in invigilators being divided into different groups. In order to solve the technical problem that the allocation results are biased, thus affecting the robustness and accuracy of the allocation model, this solution innovatively designs a new allocation activation function and a multi-dimensional constrained allocation loss function, which can accurately measure the difference between the model prediction value and the true value, and achieve a good balance between model accuracy and computational efficiency, thereby significantly improving the accuracy and stability of the model and solving the problem of inaccurate intelligent allocation results of invigilators; in view of the technical problem that the existing intelligent allocation model has inappropriate built-in parameter settings, which leads to inaccurate model prediction results, this solution improves the algorithm for obtaining the optimal parameters of the prediction model through a diversity exploration mechanism and an adaptive replacement strategy, so as to obtain the optimal parameter combination and improve the accuracy of the intelligent allocation model results.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides an intelligent allocation management method for invigilators based on multi-dimensional examination room analysis, the method comprising the following steps:

[0005] Step S1: data collection;

[0006] Step S2: data preprocessing;

[0007] Step S3: Multi-dimensional examination room analysis;

[0008] Step S4: Obtaining a multi-dimensional constraint invigilator allocation model;

[0009] Step S5: hyperparameter search optimization;

[0010] Step S6: Intelligent allocation of invigilators.

[0011] Furthermore, in step S1, the data collection is specifically to obtain the original data of the invigilator allocation from the teaching affairs system by collection;

[0012] The invigilator allocation raw data includes examination room data, invigilator data and historical invigilator allocation data; the examination room data includes examination room location, capacity, examination subjects, examination time, examination date and examination room historical question record data; the invigilator data includes name, position, department, invigilation experience, available time period, subjects allowed to be invigilated, and past examination room performance.

[0013] Furthermore, in step S2, the data preprocessing is used to perform data cleaning, standardization and feature selection on the original data of invigilator allocation to obtain preliminary data of invigilator allocation; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to extract relevant features from the original data, including invigilator characteristics, examination room characteristics and interaction characteristics between invigilators and examination rooms.

[0014] Furthermore, in step S3, the multi-dimensional examination room analysis specifically includes the following steps:

[0015] Step S31: Analysis of the subjects to be invigilated in the examination room. Specifically, the invigilators can only be assigned to the examination rooms for the subjects that are allowed to be invigilated, ensuring the matching of the subjects with the invigilator’s background, and establishing the constraint formula for the invigilated subjects. The formula used is as follows:

[0016] ;

[0017] In the formula, Indicates the constraint value of the invigilation subject, which is used to analyze whether the invigilator can be assigned to the examination room. A value of 1 indicates that the invigilator can be assigned, and a value of 0 indicates that the invigilator cannot be matched. Indicates the invigilator, Indicates the invigilation examination room;

[0018] Step S32: Analysis of the examination room invigilation time, specifically, the examination room invigilation time of the invigilator needs to be matched according to the available time period of the invigilator, and the principle of clustering the invigilation time is given priority in the allocation, the time utilization rate of the invigilator is optimized, and the invigilation time constraint formula is established; the formula used is as follows:

[0019] ;

[0020] In the formula, Indicates the invigilation time constraint value, which is used to analyze whether the invigilator's time arrangement in the examination room is reasonable. A value of 1 indicates reasonableness, and a value of 0 indicates unreasonableness. Indicates the time period constraint value, which is used to analyze whether the invigilator is available during the time period of the examination room. A value of 1 indicates that the invigilator is available, and a value of 0 indicates that the invigilator is not available. It indicates the score of the invigilator's time cohesion. The value range is [0,1]. A value of 1 indicates that the time is completely cohesive, and a value of 0 indicates that the time is not cohesive at all. represents the time clustering weight factor, which is used to control the importance of time clustering;

[0021] Step S33: Analyzing the number of invigilators in the examination room, specifically evaluating the demand for invigilators in the examination room according to the number of examinees in the examination room. The evaluation method is that when the number of examinees in the examination room reaches 15, one invigilator needs to be arranged to ensure that the number of invigilators allocated to the examination room meets the demand, and establishes a formula for constraining the number of invigilators; the formula used is as follows:

[0022] ;

[0023] In the formula, Indicates the examination room The number of invigilators required, Indicates the examination room The number of candidates, Represents the invigilator weight factor, which is used to control the number of invigilators;

[0024] Step S34: Examination room invigilation difficulty analysis, specifically matching the invigilation experience of the invigilator according to the examination room invigilation difficulty, the examination room invigilation difficulty is specifically quantitatively evaluated by the examination format, examination time, examination room layout and examination room environment; high-difficulty examination rooms must give priority to the allocation of experienced invigilators to ensure the quality of invigilation, and establish an invigilation difficulty constraint formula; the formula used is as follows:

[0025] ;

[0026] In the formula, Indicates the invigilation difficulty constraint value, which is used to analyze whether the invigilator's experience can match the difficulty of the examination room. A value of 1 indicates a match, and a value of 0 indicates a mismatch. Indicates the invigilator's invigilation experience. Indicates the difficulty level of the invigilation test room.

[0027] Further, in step S4, the multi-dimensional constraint invigilator allocation model is obtained, specifically by establishing a multi-dimensional constraint invigilator allocation model and training the model with historical data to obtain a trained multi-dimensional constraint invigilator allocation model; the steps include:

[0028] Step S41: Establishing a multi-dimensional constraint invigilator allocation model, specifically including the following steps:

[0029] Step S411: Initialize the model structure, including the input layer, hidden layer and output layer;

[0030] Step S412: Calculate the hidden layer output using the following formula:

[0031] ;

[0032] In the formula, represents the output of the hidden layer, represents the hidden layer input data, represents the connection weight from the input layer to the hidden layer, m represents the number of features of the input layer, Represents the bias parameter of the hidden layer;

[0033] Step S413: Design a distribution activation function; the formula used is as follows:

[0034] ;

[0035] In the formula, represents the allocation activation function, x represents the input variable of the allocation activation function, Represents the parameter that adjusts the overall nonlinear strength of the activation function, represents the parameter used to adjust the nonlinear strength of the middle region, Represents the stability factor, which is used to make smooth transitions between different intervals. Used to enhance the dynamic range of the activation function in the positive and negative regions;

[0036] Step S414: Calculate the output value of the output layer; the formula used is as follows:

[0037] ;

[0038] In the formula, represents the output value of the output layer, represents the connection weight from the hidden layer to the output layer, h represents the number of neurons in the hidden layer, and k represents the neuron index of the output layer. Represents the bias parameter of the output layer;

[0039] Step S415: Design a multi-dimensional constraint allocation loss function; the formula used is as follows:

[0040] ;

[0041] ;

[0042] ;

[0043] In the formula, Loss represents the multi-dimensional constraint allocation loss function value, Represents the hard constraint loss function value, which is strictly satisfied by adding a large penalty value to the loss function. represents the soft constraint loss function value, represents the penalty value for violating the invigilation subject matching constraint, i represents the i-th examination room, j represents the j-th invigilator, represents the weight factor of the invigilation time constraint, represents the weight factor for the number of invigilators, The weight factor representing the difficulty constraint of the proctoring;

[0044] Step S42: training the multi-dimensional constraint invigilator allocation model, specifically using the historical invigilator allocation data to train the multi-dimensional constraint invigilator allocation model to obtain a trained multi-dimensional constraint invigilator allocation model.

[0045] Furthermore, in step S5, the hyperparameter search optimization specifically includes the following steps:

[0046] Step S51: Initialize parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N, dimension d of search space, maximum number of iterations T, manager ratio P N And the elite ratio S N ;

[0047] Step S52: Initialize the population, specifically, all search individuals are randomly initialized in the feasible search space; the formula used is as follows:

[0048] ;

[0049] ;

[0050] ;

[0051] In the formula, represents the position of the jth dimension of the i-th individual, Indicates between A random number uniformly distributed in the range, and Respectively represent the search lower and upper limits of dimension j;

[0052] Step S53: Population division, calculation of individual fitness values ​​f in the population i ; The performance of the multi-dimensional constraint invigilator allocation model established based on individual positions is used as the individual fitness value; individuals are sorted from best to worst according to their fitness values, and divided into managers, elites and workers;

[0053] Step S54: Individual class learning stage, specifically including the following steps:

[0054] Step S541: Obtain the individual class knowledge gap, the formula used is as follows:

[0055] ;

[0056] ;

[0057] ;

[0058] ;

[0059] In the formula, represents the individual with the best fitness value, represents a random individual among managers, represents a random individual among the workers, and represents two random individuals in the population, Represents the knowledge gap between managers and elites, represents the knowledge gap between managers and workers, Represents the knowledge gap between the elite and the workers, represents the knowledge gap between two random individuals;

[0060] Step S542: Obtain the learning factor, the formula used is as follows:

[0061] ;

[0062] In the formula, represents the learning factor, Indicates the gap between individuals;

[0063] Step S543: Obtain the self-awareness factor, the formula used is as follows:

[0064] ;

[0065] ;

[0066] In the formula, Represents the self-awareness factor, Indicates the difference between individuals. represents the individual fitness function, represents the learning resistance of the j-th dimension of the i-th individual, Represents the learning resistance of the individual with the worst individual fitness value in the population;

[0067] Step S544: Acquire learning knowledge, the formula used is as follows:

[0068] ;

[0069] In the formula, represents the amount of knowledge acquired by the kth learning, which is used to adjust the individual position;

[0070] Step S545: Individual learning process update, specifically, individual learning process update is performed through a diversity exploration mechanism, and the formula used is as follows:

[0071] ;

[0072] In the formula, represents the j-th dimension position of the i-th individual in the t+1-th generation population, represents the j-th dimension position of the i-th individual in the t-th generation population, Indicates between A random number uniformly distributed in the range, and Indicates between A random number uniformly distributed in the range, Indicates the current iteration number;

[0073] Step S55: Individual reflection phase update, the formula used is as follows:

[0074] ;

[0075] In the formula, , , and Indicates between A random number uniformly distributed in the range, represents a random individual among managers;

[0076] Step S56: Searching for a boundary control mechanism, the formula used is as follows:

[0077] ;

[0078] Step S57: Perform a replacement strategy, specifically, perform an adaptive replacement strategy by introducing probability and randomness, and the formula used is as follows:

[0079] ;

[0080] In the formula, Indicates between A random number uniformly distributed within a range;

[0081] Step S58: search determination, specifically, by constructing a search termination condition, searching and determining the optimal individual position, and obtaining the optimal individual position data setting;

[0082] The search termination conditions include threshold termination and iteration termination;

[0083] The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed;

[0084] The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached;

[0085] The optimal individual position includes multi-dimensional constraint invigilator allocation model parameters.

[0086] Furthermore, in step S6, the intelligent allocation of invigilators is to use the examination room data and the invigilator data as input data of the multi-dimensional constraint invigilator allocation model after hyperparameter search optimization, to intelligently allocate invigilators, to obtain invigilator allocation data, and to record the allocation plan of the invigilators.

[0087] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0088] (1) In order to address the technical problems of the traditional examination room invigilator allocation method, such as the uncontrollable number of examination sessions and the inability to adapt to diversified invigilation rules, which may result in the invigilator allocation results not meeting the specific invigilation requirements, this solution innovatively proposes a multi-dimensional examination room analysis strategy. Through the multi-dimensional analysis of the examination room, flexible examination room constraint rules are established. The number of invigilation sessions can be adjusted according to actual needs, and the restrictions on invigilation subjects in different scenarios can be adapted, thereby effectively solving the problem of intelligent invigilator allocation in various scenarios.

[0089] (2) In order to address the technical problem that the activation function and loss function in the existing intelligent allocation model for invigilators have certain limitations, which leads to deviations in the allocation results and thus affects the robustness and accuracy of the allocation model, this solution innovatively designs a new allocation activation function and a multi-dimensional constrained allocation loss function, which can accurately measure the difference between the model prediction value and the true value, and achieve a good balance between model accuracy and computational efficiency, thereby significantly improving the accuracy and stability of the model and solving the problem of inaccurate results in the intelligent allocation of invigilators.

[0090] (3) In order to address the technical problem that the existing intelligent allocation models have inappropriate built-in parameter settings, which leads to inaccurate model prediction results, this solution improves the algorithm for obtaining the optimal parameters of the prediction model through a diversity exploration mechanism and an adaptive replacement strategy, thereby obtaining the optimal parameter combination and improving the accuracy of the intelligent allocation model results. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] Figure 1 A flowchart of an intelligent allocation management method for invigilators based on multi-dimensional examination room analysis provided by the present invention;

[0092] Figure 2 is a schematic diagram of the process of step S3;

[0093] Figure 3 is a schematic flow chart of step S4;

[0094] Figure 4 is a flow chart of step S41;

[0095] Figure 5 is a schematic flow chart of step S5;

[0096] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0097] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0098] In the description of the present invention, it is necessary to understand that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred system or element must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0099] Example 1, see Figure 1 The technical solution adopted by the present invention is as follows: The present invention provides an intelligent allocation management method for invigilators based on multi-dimensional examination room analysis, the method comprising the following steps:

[0100] Step S1: Data collection, by collecting examination room data, invigilator data and historical invigilator allocation data from the teaching affairs system to obtain the original invigilator allocation data;

[0101] Step S2: Data preprocessing, by performing data cleaning, standardization and feature selection on the collected raw data, to obtain preliminary data on the allocation of invigilators;

[0102] Step S3: Multi-dimensional examination room analysis: by analyzing the examination subjects, examination time, number of invigilators and examination difficulty, a corresponding constraint formula is established to ensure that the invigilators match the examination room requirements;

[0103] Step S4: obtaining a multi-dimensional constraint invigilator allocation model, establishing a multi-dimensional constraint invigilator allocation model, designing an allocation activation function and a multi-dimensional constraint allocation loss function, and obtaining a trained allocation model through model training;

[0104] Step S5: Hyperparameter search optimization, by using the diversity exploration mechanism and adaptive replacement strategy to optimize the hyperparameters of the model and obtain the optimal parameter combination of the model;

[0105] Step S6: Intelligent allocation of invigilators. Through the optimized multi-dimensional constraint invigilator allocation model, invigilators are intelligently allocated to obtain invigilator allocation data, and the allocation plan of invigilators is recorded.

[0106] Example 2, see Figure 1 ,This embodiment is based on the above embodiment. In step S1, the data collection is specifically to obtain the original data of the invigilator allocation from the teaching affairs system by collection;

[0107] The invigilator allocation raw data includes examination room data, invigilator data and historical invigilator allocation data; the examination room data includes examination room location, capacity, examination subjects, examination time, examination date and examination room historical question record data; the invigilator data includes name, position, department, invigilation experience, available time period, subjects allowed to be invigilated, and past examination room performance.

[0108] Example 3, see Figure 1 This embodiment is based on the above embodiment. In step S2, the data preprocessing is used to perform data cleaning, standardization processing and feature selection on the original data of invigilator allocation to obtain preliminary data of invigilator allocation; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the standardization processing is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to extract relevant features from the original data, including invigilator features, examination room features and interaction features between invigilators and examination rooms.

[0109] Example 4, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S3, the multi-dimensional examination room analysis specifically includes the following steps:

[0110] Step S31: Analysis of the subjects to be invigilated in the examination room. Specifically, the invigilators can only be assigned to the examination rooms for the subjects that are allowed to be invigilated, ensuring the matching of the subjects with the invigilator’s background, and establishing the constraint formula for the invigilated subjects. The formula used is as follows:

[0111] ;

[0112] In the formula, Indicates the constraint value of the invigilation subject, which is used to analyze whether the invigilator can be assigned to the examination room. A value of 1 indicates that the invigilator can be assigned, and a value of 0 indicates that the invigilator cannot be matched. Indicates the invigilator, Indicates the invigilation examination room;

[0113] Step S32: Analysis of the examination room invigilation time, specifically, the examination room invigilation time of the invigilator needs to be matched according to the available time period of the invigilator, and the principle of clustering the invigilation time is given priority in the allocation, the time utilization rate of the invigilator is optimized, and the invigilation time constraint formula is established; the formula used is as follows:

[0114] ;

[0115] In the formula, Indicates the invigilation time constraint value, which is used to analyze whether the invigilator's time arrangement in the examination room is reasonable. A value of 1 indicates reasonableness, and a value of 0 indicates unreasonableness. Indicates the time period constraint value, which is used to analyze whether the invigilator is available during the time period of the examination room. A value of 1 indicates that the invigilator is available, and a value of 0 indicates that the invigilator is not available. It indicates the score of the invigilator's time cohesion. The value range is [0,1]. A value of 1 indicates that the time is completely cohesive, and a value of 0 indicates that the time is not cohesive at all. represents the time clustering weight factor, which is used to control the importance of time clustering;

[0116] Step S33: Analyzing the number of invigilators in the examination room, specifically evaluating the demand for invigilators in the examination room according to the number of examinees in the examination room. The evaluation method is that when the number of examinees in the examination room reaches 15, one invigilator needs to be arranged to ensure that the number of invigilators allocated to the examination room meets the demand, and establishes a formula for constraining the number of invigilators; the formula used is as follows:

[0117] ;

[0118] In the formula, Indicates the examination room The number of invigilators required, Indicates the examination room The number of candidates, Represents the invigilator weight factor, which is used to control the number of invigilators;

[0119] Step S34: Examination room invigilation difficulty analysis, specifically matching the invigilation experience of the invigilator according to the examination room invigilation difficulty, the examination room invigilation difficulty is specifically quantitatively evaluated by the examination format, examination time, examination room layout and examination room environment; high-difficulty examination rooms must give priority to the allocation of experienced invigilators to ensure the quality of invigilation, and establish an invigilation difficulty constraint formula; the formula used is as follows:

[0120] ;

[0121] In the formula, Indicates the invigilation difficulty constraint value, which is used to analyze whether the invigilator's experience can match the difficulty of the examination room. A value of 1 indicates a match, and a value of 0 indicates a mismatch. Indicates the invigilator's invigilation experience. Indicates the difficulty level of the invigilation test room.

[0122] By performing the above operations, this solution innovatively proposes a multi-dimensional examination room analysis strategy to address the technical problems of uncontrollable examination sessions and inability to adapt to diversified invigilation rules in the traditional examination room invigilator allocation method, which may result in the invigilator allocation results not meeting specific invigilation requirements. Through multi-dimensional analysis of the examination room, flexible examination room constraint rules are established, which can adjust the invigilation sessions according to actual needs and adapt to the invigilation subject restrictions in different scenarios, thereby effectively solving the problem of intelligent invigilator allocation in various scenarios.

[0123] Example 5, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. Further, in step S4, the multi-dimensional constraint invigilator allocation model is obtained, specifically by establishing a multi-dimensional constraint invigilator allocation model and training the model with historical data to obtain a trained multi-dimensional constraint invigilator allocation model; the following steps are included:

[0124] Step S41: Establishing a multi-dimensional constraint invigilator allocation model, specifically designing an allocation activation function and a multi-dimensional constraint allocation loss function, and establishing a multi-dimensional constraint invigilator allocation model;

[0125] Step S42: training the multi-dimensional constraint invigilator allocation model, specifically using the historical invigilator allocation data to train the multi-dimensional constraint invigilator allocation model to obtain a trained multi-dimensional constraint invigilator allocation model.

[0126] Example 6, see Figure 1 , Figure 3 and Figure 4 Based on the above embodiment, this embodiment, in step S41: establishing a multi-dimensional constraint invigilator allocation model, specifically includes the following steps:

[0127] Step S411: Initialize the model structure, including the input layer, hidden layer and output layer;

[0128] Step S412: Calculate the hidden layer output using the following formula:

[0129] ;

[0130] In the formula, represents the output of the hidden layer, represents the hidden layer input data, represents the connection weight from the input layer to the hidden layer, m represents the number of features of the input layer, Represents the bias parameter of the hidden layer;

[0131] Step S413: Design a distribution activation function; the formula used is as follows:

[0132] ;

[0133] In the formula, represents the allocation activation function, x represents the input variable of the allocation activation function, Represents the parameter that adjusts the overall nonlinear strength of the activation function, represents the parameter used to adjust the nonlinear strength of the middle region, Represents the stability factor, which is used to make smooth transitions between different intervals. Used to enhance the dynamic range of the activation function in the positive and negative regions;

[0134] Step S414: Calculate the output value of the output layer; the formula used is as follows:

[0135] ;

[0136] In the formula, represents the output value of the output layer, represents the connection weight from the hidden layer to the output layer, h represents the number of neurons in the hidden layer, and k represents the neuron index of the output layer. Represents the bias parameter of the output layer;

[0137] Step S415: Design a multi-dimensional constraint allocation loss function; the formula used is as follows:

[0138] ;

[0139] ;

[0140] ;

[0141] In the formula, Loss represents the multi-dimensional constraint allocation loss function value, Represents the hard constraint loss function value, which is strictly satisfied by adding a large penalty value to the loss function. represents the soft constraint loss function value, represents the penalty value for violating the invigilation subject matching constraint, i represents the i-th examination room, j represents the j-th invigilator, represents the weight factor of the invigilation time constraint, represents the weight factor for the number of invigilators, The weight factor representing the proctoring difficulty constraint.

[0142] By performing the above operations, in order to solve the technical problem that the activation function and loss function in the existing intelligent allocation model for invigilators have certain limitations, resulting in deviations in the allocation results, thereby affecting the robustness and accuracy of the allocation model, this solution innovatively designs a new allocation activation function and a multi-dimensional constraint allocation loss function, which can accurately measure the difference between the model prediction value and the true value, and achieve a good balance between model accuracy and computational efficiency, thereby significantly improving the accuracy and stability of the model and solving the problem of inaccurate results of intelligent allocation of invigilators.

[0143] Embodiment 7, see Figure 1 and Figure 5 This embodiment is based on the above embodiment. In step S5, the hyperparameter search optimization specifically includes the following steps:

[0144] Step S51: Initialize parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N, dimension d of search space, maximum number of iterations T, manager ratio P N And the elite ratio S N ;

[0145] Step S52: Initialize the population, specifically, all search individuals are randomly initialized in the feasible search space; the formula used is as follows:

[0146] ;

[0147] ;

[0148] ;

[0149] In the formula, represents the position of the jth dimension of the i-th individual, Indicates between A random number uniformly distributed in the range, and Respectively represent the search lower and upper limits of dimension j;

[0150] Step S53: Population division: Calculate the individual fitness value f in the population i ; The performance of the multi-dimensional constraint invigilator allocation model established based on individual positions is used as the individual fitness value; individuals are sorted from best to worst according to their fitness values, and divided into managers, elites and workers;

[0151] Step S54: Individual class learning stage, specifically including the following steps:

[0152] Step S541: Obtain the individual class knowledge gap, the formula used is as follows:

[0153] ;

[0154] ;

[0155] ;

[0156] ;

[0157] In the formula, represents the individual with the best fitness value, represents a random individual among managers, represents a random individual among the workers, and represents two random individuals in the population, Represents the knowledge gap between managers and elites, represents the knowledge gap between managers and workers, Represents the knowledge gap between the elite and the workers, represents the knowledge gap between two random individuals;

[0158] Step S542: Obtain the learning factor, the formula used is as follows:

[0159] ;

[0160] In the formula, represents the learning factor, Indicates the gap between individuals;

[0161] Step S543: Obtain the self-awareness factor, the formula used is as follows:

[0162] ;

[0163] ;

[0164] In the formula, Represents the self-awareness factor, Indicates the difference between individuals. represents the individual fitness function, represents the learning resistance of the jth dimension of the i-th individual, Represents the learning resistance of the individual with the worst individual fitness value in the population;

[0165] Step S544: Acquire learning knowledge, the formula used is as follows:

[0166] ;

[0167] In the formula, represents the amount of knowledge acquired by the kth learning, which is used to adjust the individual position;

[0168] Step S545: Individual learning process update, specifically, individual learning process update is performed through a diversity exploration mechanism, and the formula used is as follows:

[0169] ;

[0170] In the formula, represents the j-th dimension position of the i-th individual in the t+1-th generation population, represents the j-th dimension position of the i-th individual in the t-th generation population, Indicates between A random number uniformly distributed in the range, and Indicates between A random number uniformly distributed in the range, Indicates the current iteration number;

[0171] Step S55: Individual reflection phase update, the formula used is as follows:

[0172] ;

[0173] In the formula, , , and Indicates between A random number uniformly distributed in the range, represents a random individual among managers;

[0174] Step S56: Search for a boundary control mechanism, using the following formula:

[0175] ;

[0176] Step S57: Perform a replacement strategy, specifically, perform an adaptive replacement strategy by introducing probability and randomness, and the formula used is as follows:

[0177] ;

[0178] In the formula, Indicates between A random number uniformly distributed within a range;

[0179] Step S58: search determination, specifically, by constructing a search termination condition, searching and determining the optimal individual position, and obtaining the optimal individual position data setting;

[0180] The search termination conditions include threshold termination and iteration termination;

[0181] The threshold termination is specifically to set the fitness threshold, when the individual fitness value fi When it is above the fitness threshold, the hyperparameter search is completed;

[0182] The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached;

[0183] The optimal individual position includes multi-dimensional constraint invigilator allocation model parameters.

[0184] By performing the above operations, in order to address the technical problem that the existing intelligent allocation model has inappropriate built-in parameter settings, resulting in inaccurate model prediction results, this solution improves the algorithm for obtaining the optimal parameters of the prediction model through a diversity exploration mechanism and an adaptive replacement strategy, thereby obtaining the optimal parameter combination and improving the accuracy of the intelligent allocation model results.

[0185] Embodiment 8, see Figure 1 This embodiment is based on the above embodiment. In step S6, the intelligent allocation of invigilators is to use the examination room data and the invigilator data as input data of the multi-dimensional constraint invigilator allocation model after hyperparameter search optimization, intelligently allocate invigilators, obtain invigilator allocation data, and record the allocation plan of invigilators.

[0186] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0187] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0188] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An intelligent allocation management method for invigilators based on multi-dimensional examination room analysis, characterized by: The method comprises the following steps: Step S1: Data collection, by collecting examination room data, invigilator data and historical invigilator allocation data from the teaching affairs system to obtain the original invigilator allocation data; Step S2: Data preprocessing, by performing data cleaning, standardization and feature selection on the collected raw data, to obtain preliminary data on the allocation of invigilators; Step S3: Multi-dimensional examination room analysis: by analyzing the examination subjects, examination time, number of invigilators and examination difficulty, a corresponding constraint formula is established to ensure that the invigilators match the examination room requirements; Step S4: obtaining a multi-dimensional constraint invigilator allocation model, establishing a multi-dimensional constraint invigilator allocation model, designing an allocation activation function and a multi-dimensional constraint allocation loss function, and obtaining a trained allocation model through model training; Step S5: Hyperparameter search optimization, by using the diversity exploration mechanism and adaptive replacement strategy, optimize the hyperparameters of the model to obtain the optimal parameter combination of the model; Step S6: Intelligent allocation of invigilators. Through the optimized multi-dimensional constraint invigilator allocation model, invigilators are intelligently allocated to obtain invigilator allocation data, and the allocation plan of invigilators is recorded.

2. According to claim 1, a method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis is characterized by: In step S3, the multi-dimensional examination room analysis specifically includes the following steps: Step S31: Analysis of the subjects to be invigilated in the examination room. Specifically, the invigilators can only be assigned to the examination rooms for the subjects that are allowed to be invigilated, ensuring the matching of the subjects with the invigilator’s background, and establishing the constraint formula for the invigilated subjects. The formula used is as follows: ； In the formula, Indicates the constraint value of the invigilation subject, which is used to analyze whether the invigilator can be assigned to the examination room. A value of 1 indicates that the invigilator can be assigned, and a value of 0 indicates that the invigilator cannot be matched. Indicates the invigilator, Indicates the invigilation examination room; Step S32: Analysis of the examination room invigilation time, specifically, the examination room invigilation time of the invigilator needs to be matched according to the available time period of the invigilator, and the principle of clustering the invigilation time is given priority in the allocation, the time utilization rate of the invigilator is optimized, and the invigilation time constraint formula is established; the formula used is as follows: ； In the formula, Indicates the invigilation time constraint value, which is used to analyze whether the invigilator's time arrangement in the examination room is reasonable. A value of 1 indicates reasonableness, and a value of 0 indicates unreasonableness. Indicates the time period constraint value, which is used to analyze whether the invigilator is available during the time period of the examination room. A value of 1 indicates that the invigilator is available, and a value of 0 indicates that the invigilator is not available. It indicates the score of the invigilator's time cohesion. The value range is [0,1]. A value of 1 indicates that the time is completely cohesive, and a value of 0 indicates that the time is not cohesive at all. represents the time clustering weight factor, which is used to control the importance of time clustering; Step S33: Analyzing the number of invigilators in the examination room, specifically evaluating the demand for invigilators in the examination room according to the number of examinees in the examination room. The evaluation method is that when the number of examinees in the examination room reaches 15, one invigilator needs to be arranged to ensure that the number of invigilators allocated to the examination room meets the demand, and establishes a formula for constraining the number of invigilators; the formula used is as follows: ； In the formula, Indicates the examination room The number of invigilators required, Indicates the examination room The number of candidates, Represents the invigilator weight factor, which is used to control the number of invigilators; Step S34: Examination room invigilation difficulty analysis, specifically matching the invigilation experience of the invigilator according to the examination room invigilation difficulty, the examination room invigilation difficulty is specifically quantitatively evaluated by the examination format, examination time, examination room layout and examination room environment; high-difficulty examination rooms must give priority to the allocation of experienced invigilators to ensure the quality of invigilation, and establish an invigilation difficulty constraint formula; the formula used is as follows: ； In the formula, Indicates the invigilation difficulty constraint value, which is used to analyze whether the invigilator's experience can match the difficulty of the examination room. A value of 1 indicates a match, and a value of 0 indicates a mismatch. Indicates the invigilator's invigilation experience. Indicates the difficulty level of the invigilation test room.

3. According to claim 1, a method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis is characterized in that: Further, in step S4, the multi-dimensional constraint invigilator allocation model is obtained, specifically by establishing a multi-dimensional constraint invigilator allocation model and training the model with historical data to obtain a trained multi-dimensional constraint invigilator allocation model; the steps include: Step S41: Establishing a multi-dimensional constraint invigilator allocation model, specifically including the following steps: Step S411: Initialize the model structure, including the input layer, hidden layer and output layer; Step S412: Calculate the hidden layer output using the following formula: ； In the formula, represents the output of the hidden layer, represents the hidden layer input data, represents the connection weight from the input layer to the hidden layer, m represents the number of features of the input layer, Represents the bias parameter of the hidden layer; Step S413: Design a distribution activation function; the formula used is as follows: ； In the formula, represents the allocation activation function, x represents the input variable of the allocation activation function, Represents the parameter that adjusts the overall nonlinear strength of the activation function, represents the parameter used to adjust the nonlinear strength of the middle region, Represents the stability factor, which is used to make smooth transitions between different intervals. Used to enhance the dynamic range of the activation function in the positive and negative regions; Step S414: Calculate the output value of the output layer; the formula used is as follows: ； In the formula, represents the output value of the output layer, represents the connection weight from the hidden layer to the output layer, h represents the number of neurons in the hidden layer, and k represents the neuron index of the output layer. Represents the bias parameter of the output layer; Step S415: Design a multi-dimensional constraint allocation loss function; the formula used is as follows: ； ； ； In the formula, Loss represents the multi-dimensional constraint allocation loss function value, Represents the hard constraint loss function value, which is strictly satisfied by adding a large penalty value to the loss function. represents the soft constraint loss function value, represents the penalty value for violating the invigilation subject matching constraint, i represents the i-th examination room, j represents the j-th invigilator, represents the weight factor of the invigilation time constraint, represents the weight factor for the number of invigilators, The weight factor representing the difficulty constraint of the proctoring; Step S42: training the multi-dimensional constraint invigilator allocation model, specifically using the historical invigilator allocation data to train the multi-dimensional constraint invigilator allocation model to obtain a trained multi-dimensional constraint invigilator allocation model.

4. According to claim 1, a method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis is characterized in that: In step S5, the hyperparameter search optimization specifically includes the following steps: Step S51: Initialize parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N, dimension d of search space, maximum number of iterations T, manager ratio P N And the elite ratio S N ; Step S52: Initialize the population, specifically, all search individuals are randomly initialized in the feasible search space; the formula used is as follows: ； ； ； In the formula, represents the position of the jth dimension of the i-th individual, Indicates between A random number uniformly distributed in the range, and Respectively represent the search lower and upper limits of dimension j; Step S53: Population division: Calculate the individual fitness value f in the population i ; The performance of the multi-dimensional constraint invigilator allocation model established based on individual positions is used as the individual fitness value; individuals are sorted from best to worst according to their fitness values, and divided into managers, elites and workers; Step S54: Individual class learning stage, specifically including the following steps: Step S541: Obtain the individual class knowledge gap, the formula used is as follows: ； ； ； ； In the formula, represents the individual with the best fitness value, represents a random individual among managers, represents a random individual among the workers, and represents two random individuals in the population, Represents the knowledge gap between managers and elites, represents the knowledge gap between managers and workers, Represents the knowledge gap between the elite and the workers, represents the knowledge gap between two random individuals; Step S542: Obtain the learning factor, the formula used is as follows: ； In the formula, represents the learning factor, Indicates the gap between individuals; Step S543: Obtain the self-awareness factor, the formula used is as follows: ； ； In the formula, Represents the self-awareness factor, Indicates the difference between individuals. represents the individual fitness function, represents the learning resistance of the jth dimension of the i-th individual, Represents the learning resistance of the individual with the worst individual fitness value in the population; Step S544: Acquire learning knowledge, the formula used is as follows: ； In the formula, represents the amount of knowledge acquired by the kth learning, which is used to adjust the individual position; Step S545: Individual learning process update, specifically, individual learning process update is performed through a diversity exploration mechanism, and the formula used is as follows: ； In the formula, represents the j-th dimension position of the i-th individual in the t+1-th generation population, represents the j-th dimension position of the i-th individual in the t-th generation population, Indicates between A random number uniformly distributed in the range, and Indicates between A random number uniformly distributed in the range, Indicates the current iteration number; Step S55: Individual reflection phase update, the formula used is as follows: ； In the formula, , , and Indicates between A random number uniformly distributed in the range, represents a random individual among managers; Step S56: Searching for a boundary control mechanism, the formula used is as follows: ； Step S57: Perform a replacement strategy, specifically, perform an adaptive replacement strategy by introducing probability and randomness, and the formula used is as follows: ； In the formula, Indicates between A random number uniformly distributed within a range; Step S58: search determination, specifically, by constructing a search termination condition, searching and determining the optimal individual position, and obtaining the optimal individual position data setting; The search termination conditions include threshold termination and iteration termination; The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed; The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached; The optimal individual position includes multi-dimensional constraint invigilator allocation model parameters.

5. According to claim 1, a method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis is characterized in that: In step S1, the data collection is specifically to obtain the original data of the invigilator allocation from the teaching affairs system; The invigilator allocation raw data includes examination room data, invigilator data and historical invigilator allocation data; the examination room data includes examination room location, capacity, examination subjects, examination time, examination date and examination room historical question record data; the invigilator data includes name, position, department, invigilation experience, available time period, subjects allowed to be invigilated, and past examination room performance.

6. The method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis according to claim 1 is characterized in that: In step S2, the data preprocessing is used to perform data cleaning, standardization and feature selection on the original data of invigilator allocation to obtain preliminary data of invigilator allocation; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to extract relevant features from the original data, including invigilator characteristics, examination room characteristics and interaction characteristics between invigilators and examination rooms.

7. The method for intelligent allocation and management of invigilators based on multi-dimensional examination room analysis according to claim 1 is characterized in that: In step S6, the intelligent allocation of invigilators is to use the examination room data and invigilator data as input data of the multi-dimensional constraint invigilator allocation model after hyperparameter search optimization, intelligently allocate invigilators, obtain invigilator allocation data, and record the allocation plan of invigilators.

Images