Online personalized learning resource recommendation method and system based on improved firefly algorithm

Abstract

The invention discloses an online personalized learning resource recommendation method and system based on an improved firefly algorithm. The method comprises the steps of obtaining a user portrait and learning resource characteristics of a learner; the user portrait and the learning resource features are input into a pre-trained learning resource recommendation model, and the learning resource recommendation model constructs a multi-dimensional matching fitness function between the learner and the learning resources according to the user portrait and the learning resource features of the learner, and solving the multi-dimensional matching fitness function through an improved firefly algorithm and outputting recommended learning resources. According to the online personalized learning resource recommendation method and system based on the improved firefly algorithm, by analyzing the multi-dimensional features of the learner and the learning resources, an online learning resource recommendation model for deep analysis of the multi-dimensional feature difference of the learner and the learning resources is constructed, and by solving the improved firefly algorithm, the online personalized learning resource recommendation method and system based on the improved firefly algorithm are obtained. And more personalized learning paths and resource recommendation meeting requirements are provided for learners.

Description

Technical Field

[0001] The present invention relates to an online personalized learning resource recommendation method and system based on an improved firefly algorithm, belonging to the technical field of smart education. Background Art

[0002] In the intelligent learning ecosystem, learning resources undoubtedly occupy a core position, and their richness and quality are the key cornerstones to ensure the efficient operation of online education. With the rapid development of information technology, although it has greatly enabled the research and development and deployment of online learning resources, this process is also accompanied by new challenges: how to accurately match learners' personalized needs, including their learning goals, ability levels and learning styles, in the vast ocean of online resources has become an urgent problem to be solved. The diversified, distributed and fragmented characteristics of resources have aggravated the cognitive burden of learners and led to frequent information overload. In particular, the contradiction between the surge in resources and the lack of individual effective screening ability has seriously restricted the efficiency of online learning. Therefore, it is particularly important to build an intelligent learning system that can rely on Internet technology and artificial intelligence technology to achieve intelligent analysis of learning situations, deep understanding of learner characteristics and highly customized learning services.

[0003] As an innovative practice in the field of education, the core goal of personalized learning resource customization service is to dynamically construct and push highly targeted learning resources based on the unique individual characteristics of learners, including but not limited to their potential learning ability, pre-set learning goals, and specific attributes of resources. The original intention of the design of this service mechanism is to effectively promote significant improvement in learning outcomes by accurately matching learning needs with resource supply. Research in this field has not only deepened the exploration of the scientific nature of online learning, but also opened up new paths to achieve intelligent and personalized learning experiences.

[0004] As a new branch of artificial intelligence, computational intelligence has attracted much attention in the field of swarm intelligence. As a soft simulation of group behavior in nature, swarm intelligence has demonstrated its powerful ability to solve complex problems through the collaboration of simple individuals, giving problem-solving methods higher robustness and flexibility. With its high efficiency and easy implementation, swarm intelligence algorithms have demonstrated excellent problem optimization capabilities in the field of intelligent learning ([1] Li Haojun, Zhang Guang, Wang Wanliang, et al. Personalized learning resource recommendation method based on multi-dimensional feature differences [J]. Systems Engineering - Theory and Practice, 2017, 37(11): 2995-3005.), such as personalized resource recommendation, intelligent planning of learning paths, generation of adaptive assessment materials, and optimization of learning platforms. It has accumulated fruitful research results and injected new impetus into the development of intelligent education.

[0005] In recent years, inspired by the unique luminescence and attraction mechanism of fireflies in nature, the firefly algorithm has been successfully proposed as a novel swarm intelligence optimization method and has received widespread attention. The algorithm cleverly constructs a self-organizing optimization search framework by precisely simulating the dynamic interaction between fireflies based on brightness differences and spatial distances. This framework not only demonstrates the simplicity and efficiency of its algorithm logic, but also shows excellent adaptability and robustness in a variety of optimization scenarios, providing a new perspective and powerful tool for solving complex optimization problems. The introduction of the firefly algorithm not only enriches the theoretical system of swarm intelligence algorithms, but also opens up new avenues for optimization research in computer science, engineering technology and other related fields. Although the firefly algorithm has shown its potential application value in many fields, its core attraction mechanism and the intrinsic optimization mechanism of the randomness model need to be further analyzed. In particular, how to finely adjust parameters during the operation of the algorithm to balance the ability of global exploration and local development has become a key challenge facing current research. This problem is not only related to the further improvement of the algorithm performance, but also directly affects the applicability and efficiency of the algorithm in actual complex optimization problems.

[0006] It can be seen that in order to generate a learning resource sequence that meets the learners' ability level and interest preferences according to their personalized characteristics and learning needs, there is an urgent need for an online personalized learning resource recommendation method and system based on the improved Firefly algorithm. Summary of the invention

[0007] The purpose of the present invention is to overcome the deficiencies in the prior art and provide an online personalized learning resource recommendation method and system based on the improved firefly algorithm. By deeply analyzing the multidimensional characteristics of learners and learning resources, an online learning resource recommendation model that deeply analyzes the differences in multidimensional characteristics of learners and learning resources is constructed. In view of the defect that the traditional firefly algorithm is prone to premature convergence due to over-reliance on the optimal solution during the search process, the attraction mechanism of the firefly algorithm is optimized and transformed. The optimization strategy randomly selects suboptimal individuals from the neighboring areas of the optimal individuals in the population as search leaders based on shape similarity information to lead the swarm to conduct global search. This measure effectively utilizes the information differences in different neighborhoods and enhances the diversity and effectiveness of the algorithm search path, thereby providing learners with more personalized and demand-oriented learning paths and resource recommendations.

[0008] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0009] In a first aspect, the present invention provides an online personalized learning resource recommendation method based on an improved firefly algorithm, comprising:

[0010] Obtain learners’ user profiles and learning resource characteristics;

[0011] The user portrait and learning resource characteristics are input into a pre-trained learning resource recommendation model, and the learning resource recommendation model constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics. The multi-dimensional matching fitness function is solved by the improved firefly algorithm and the recommended learning resources are output, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

[0012] Furthermore, the user portrait includes the learner's ability level, media type preference, content type preference and upper limit of learning time, and the learning resource characteristics include learning resource difficulty, media type, content type and time required for learning resources.

[0013] Furthermore, solving the multi-dimensional matching fitness function by improving the firefly algorithm and outputting the recommended learning resources includes:

[0014] Randomly initialize the positions and chaotic sequences of firefly populations;

[0015] Execute the iterative steps until a termination condition is met, wherein the termination condition includes that the multidimensional matching fitness function value is lower than a preset threshold or the algorithm reaches a preset maximum number of iterations;

[0016] When the termination condition is met, the recommended learning resources are determined and output;

[0017] The iterative steps include:

[0018] Adopt non-mandatory decision element updating method to determine recommended learning resources;

[0019] Update the adaptive convergence factor;

[0020] Update the optimal solution neighborhood information based on multi-dimensional space similarity theory;

[0021] Update the chaotic sequence according to the inverse sine map;

[0022] Compare the individual fitness value of fireflies with the average fitness value of the group, and select the individual update method based on the comparison results;

[0023] Implement the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update individuals;

[0024] Update individuals based on chaotic local search strategy.

[0025] Furthermore, the multidimensional matching fitness function includes six secondary fitness functions, wherein:

[0026] (1)

[0027] is the first-level secondary fitness function, which represents the difference between the learner's understanding ability and the difficulty of the recommended resources;

[0028] (2)

[0029] is the secondary fitness function, which represents the difference between the learner's media type preference and the recommended resource media type; is an indicator function used to determine the sequence and In the Whether the elements at the positions match, if ,but ,otherwise , For two sequences in The weight of the indicator function result at each position;

[0030] (3)

[0031] is a three-level secondary fitness function, which is used to evaluate the balance between the learner's media type preference and the recommended resource media type; It represents the mean of the difference between the learner's media type preference and the recommended resource media type when learning each knowledge point;

[0032] (4)

[0033] is a four-level secondary fitness function, which represents the difference between the learner’s content type preference and the recommended resource content type. For two sequences in The weight of the indicator function result at each position;

[0034] (5)

[0035] is a five-level secondary fitness function, which is used to evaluate the balance between the learner's content type preference and the recommended resource content type; It represents the mean of the difference between learners’ content type preference and recommended resource content type when learning each knowledge point;

[0036] (6)

[0037] It is a six-level secondary fitness function, which is used to evaluate whether the time required to complete the learning resources exceeds the upper limit of the learner's acceptable learning time;

[0038] (7)

[0039] (7) is the multi-dimensional matching fitness function, where , , , , and is the weight coefficient of the secondary fitness function, and the matching threshold is set ; The smaller the value, the higher the quality of the recommended learning resources, and the more suitable they are for learners. The higher the value, the less suitable the learning resource is for recommending to learners;

[0040] (8)

[0041] is the mean fitness of the population, is the population size;

[0042] Random Generation between [0,1] The dimensional vector is used as the initial position of the firefly population, and the same number of random numbers in the interval [0,1] are generated as the initial values ​​of the chaotic sequence, which can be expressed as X _ Cho h j = [ x _ Cho h 、 1 j , … , x _ Cho h 、 l j , … , x _ Cho h 、 L j ] ;

[0043] in, For learners Ability to understand knowledge points; m ∈ [ 1 , … , M ] , is the number of knowledge points, Indicates Resources The difficulty of each knowledge point, l ∈ [ 1 , … , L ] , is the number of resources, Indicates Resources The media type of each knowledge point, B l 、 m = [ b l 、 m 、 1 , … , b l 、 m 、 k , … , b l 、 m 、 K ] , k ∈ [ 1 , … , K ] , is the number of media types, 1 means the Resources Knowledge points are If it is 0, it means it is not presented in this type. Indicates that learners accept media type preference when learning a knowledge point, C m = [ c m 、 1 , … , c m 、 k , … , c m 、 K ] , 1 means the learner accepts Knowledge points are If it is 0, it means that the learner refuses to present the knowledge point in this media type. Indicates Resources The content type of each knowledge point, D l 、 m = [ d l 、 m 、 1 , … , d l 、 m 、 q , … , d l 、 m 、 Q ] , q ∈ [ 1 , … , Q ] , is the number of content types, 1 means the Resources Knowledge points are If it is 0, it means that the content type is not presented. Indicates that learners accept Content type preference when consuming knowledge points, E m = [ e m 、 1 , … , e m 、 q , … , e m 、 Q ] , 1 means the learner accepts Knowledge points are content type, otherwise 0 means that the learner refuses to present the knowledge point in this content type; Represents learning The time used by a resource, Indicates the upper limit of the learning time that learners can accept; is a decision sequence, where NET = [ Net 1 , … , Net l , … , Net L ] , is a decision-making element. If it is 1, it means resources are recommended to learners if If it is 0, it means that it is not recommended to the learner. is the maximum number of iterations, is the current iteration number.

[0044] Furthermore, the non-mandatory decision element updating method is used to determine the recommended learning resources, including:

[0045] Map the element values ​​in the continuous position space to the binary discrete space. Learning Resource Decision Sequence The corresponding firefly individual position vector The mapping produces, where

[0046] X h j = [ x h 、 1 j , … , x h 、 l j , … , x h 、 L j ] , h ∈ [ 1 , … , H ] , combined with After mapping the probability value, the learning resource decision sequence is expressed as follows:

[0047] (9)

[0048] (9) In the formula, is a random number in the range [0,1].

[0049] Combined with the decision sequence results in formula (9), the fitness function value of the learning resource corresponding to the decision sequence is calculated according to the matching fitness function formula (7).

[0050] Sort the fitness function values ​​of each learning resource from small to large. The fitness function value corresponding to the recommended learning resource must be the smallest. The individual corresponding to its decision sequence is the optimal individual of the firefly population, expressed as ;

[0051] Update statistics, Do deal with.

[0052] Furthermore, the formula for updating the adaptive convergence factor includes:

[0053] (10)

[0054] , Convergence factors The maximum and minimum values ​​of is the smoothing factor of the convergence curve;

[0055] The updating of the optimal solution neighborhood information based on multidimensional space similarity theory includes:

[0056] Determine the historical optimal individual in the iteration process based on the matching fitness function , X HOpt = [ x 1 HOpt , … , x l HOpt , … , x L HOpt ] , randomly selected with Individuals with higher comprehensive similarity are , X j 、 Approx · best = [ x 1 j 、 Approx · best , … , x l j 、 Approx · best , … , x L j 、 Approx · best ] ;

[0057] The calculation formula for the similarity evaluation standard between individuals in the population and the historical best individual is as follows:

[0058] (11)

[0059] in, The calculation formula is as follows:

[0060] (12)

[0061] The calculation formula is as follows:

[0062] (13)

[0063] The calculation formula is as follows:

[0064] (14)

[0065] Standard parameters for similarity assessment;

[0066] According to formula (11), the comprehensive similarity between the individuals in the population and the optimal individual is calculated, and the similarity values ​​are sorted from small to large. The first H% of individual positions are selected as candidate solutions, and a solution is randomly selected from them as .

[0067] Furthermore, the chaotic sequence is updated according to the inverse sine mapping, including:

[0068] (15)

[0069] The updating method of selecting an individual according to the comparison result includes:

[0070] like , select and implement the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update the individual, For individuals The corresponding fitness function value; if , implement the update of individuals based on chaotic local search strategy;

[0071] The implementation of the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update individuals includes:

[0072] (16)

[0073] is a random number in the range [0,1]. It follows a Gaussian distribution with a mathematical expectation of 0 and a variance of 1;

[0074] The updating of individuals based on the chaotic local search strategy includes:

[0075] (17)

[0076] in It is obtained by updating the chaotic sequence according to the inverse sine map.

[0077] In a second aspect, the present invention provides an online personalized learning resource recommendation system based on an improved firefly algorithm, comprising:

[0078] Acquisition module: used to obtain learners' user profiles and learning resource characteristics;

[0079] Output module: used to input the user portrait and learning resource characteristics into a pre-trained learning resource recommendation model, which constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics, solves the multi-dimensional matching fitness function through the improved firefly algorithm and outputs the recommended learning resources, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

[0080] In a third aspect, the present invention provides a computer device, including a processor and a storage medium;

[0081] The storage medium is used to store instructions;

[0082] The processor is used to operate according to the instructions to execute the steps of the method according to the first aspect.

[0083] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0084] Compared with the prior art, the present invention has the following beneficial effects:

[0085] The online personalized learning resource recommendation method based on the improved firefly algorithm provided by the present invention constructs an online learning resource recommendation model that deeply analyzes the differences in the multi-dimensional characteristics of learners and learning resources. In view of the defect that the traditional firefly algorithm is prone to premature convergence due to over-reliance on the optimal solution during the search process, the attraction mechanism of the firefly algorithm is optimized and transformed. The optimization strategy randomly selects suboptimal individuals from the neighboring areas of the optimal individuals in the population as search leaders based on shape similarity information to lead the swarm to conduct global search. This measure effectively utilizes the information differences in different neighborhoods and enhances the diversity and effectiveness of the algorithm search path, thereby providing learners with more personalized and demand-oriented learning paths and resource recommendations.

[0086] The online personalized learning resource recommendation method based on the improved firefly algorithm provided by the present invention introduces a dynamically adjusted adaptive convergence factor formula to break through the limitations of the firefly algorithm population search and improve the distribution of the solution set. This formula gives the population a strong global exploration ability in the early stage of iteration, ensuring a wide coverage of the solution space. In addition, in order to stabilize the convergence of the population and enhance its ability to escape the local optimal trap, the present invention integrates a search mechanism based on an improved chaotic mapping, which significantly broadens the search boundary and improves the global search performance of the algorithm through the complexity and ergodicity of the chaotic sequence. Simulation experiments have shown that the improved firefly algorithm proposed in the present invention has shown good application potential in the field of personalized online learning resource recommendation. Its convergence and recommendation matching are better than the existing learning resource recommendation method based on the swarm intelligence optimization algorithm, providing more efficient and robust technical support for the accurate recommendation of online educational resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 It is the algorithm flow chart of the present invention;

[0088] Figure 2 It is a comparison of convergence curves of different algorithms in the experimental environment of the embodiment of the present invention;

[0089] Figure 3 It is a comparison of convergence curves of different algorithms in the environment of Experiment 2 in the embodiment of the present invention;

[0090] Figure 4 It is a comparison of convergence curves of different algorithms in the third experimental environment of the embodiment of the present invention;

[0091] Figure 5 It is a comparison of convergence curves of different algorithms in the fourth experimental environment of the embodiment of the present invention;

[0092] Figure 6 It is a flow chart of an online personalized learning resource recommendation method based on an improved firefly algorithm provided by the present invention. DETAILED DESCRIPTION

[0093] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0094] The term "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0095] Embodiment 1:

[0096] Figure 6 This is a flowchart of an online personalized learning resource recommendation method based on an improved firefly algorithm in Embodiment 1 of the present invention. The online personalized learning resource recommendation method based on an improved firefly algorithm provided in this embodiment can be applied to a terminal and can be executed by an online personalized learning resource recommendation system based on an improved firefly algorithm. The system can be implemented by software and / or hardware. The system can be integrated in a terminal, for example: any smart phone, tablet computer or computer device with communication function. Figure 6 , the method of this implementation includes:

[0097] Obtain learners’ user profiles and learning resource characteristics;

[0098] The user portrait and learning resource features are input into a pre-trained learning resource recommendation model, the learning resource recommendation model constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource features, solves the multi-dimensional matching fitness function through an improved firefly algorithm and outputs the recommended learning resources, wherein the improved firefly algorithm uses shape similarity information to randomly select sub-optimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search;

[0099] like Figure 1 As shown, the method specifically includes:

[0100] Step A: Construct a multidimensional matching fitness function between learners and learning resources based on learner ability level, media type preference, content type preference, and upper limit of learning time, learning resource difficulty, media type, content type, and learning resource required time. Set matching threshold, population size, and maximum number of algorithm iterations, and randomly initialize the firefly population position and chaotic sequence.

[0101] Step B: Use the non-mandatory decision element update method to determine the recommended learning resources;

[0102] Step C: Update the adaptive convergence factor;

[0103] Step D: Update the optimal solution neighborhood information based on multidimensional space similarity theory;

[0104] Step E: Update the chaotic sequence according to the inverse sine mapping;

[0105] Step F: Compare the individual fitness value of fireflies with the average fitness value of the group, and select the individual update method based on the comparison result;

[0106] Step G: Implement the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update individuals;

[0107] Step H: Update individuals based on chaotic local search strategy;

[0108] Step I: Determine whether the multidimensional matching function value is lower than the threshold or whether the algorithm has reached the maximum number of iterations. If any of the conditions are met, execute step J, otherwise execute step B;

[0109] Step J: Identify recommended learning resources.

[0110] Step Aa: For learners The ability to understand a knowledge point. m ∈ [ 1 , … , M ] , is the number of knowledge points, Indicates Resources The difficulty of each knowledge point, l ∈ [ 1 , … , L ] , is the number of resources, Indicates Resources The media type of each knowledge point, B l 、 m = [ b l 、 m 、 1 , … , b l 、 m 、 k , … , b l 、 m 、 K ] , k ∈ [ 1 , … , K ] , is the number of media types, 1 means the Resources Knowledge points are If it is 0, it means it will not be presented in that type. Indicates that learners accept media type preference when learning a knowledge point, C m = [ c m 、 1 , … , c m 、 k , … , c m 、 K ] , 1 means the learner accepts Knowledge points are If it is 0, it means the learner refuses to present the knowledge point in this media type. Indicates Resources The content type of each knowledge point, D l 、 m = [ d l 、 m 、 1 , … , d l 、 m 、 q , … , d l 、 m 、 Q ] , q ∈ [ 1 , … , Q ] , is the number of content types, 1 means the Resources Knowledge points are If it is 0, it means that the content type is not presented. Indicates that learners accept Content type preference when consuming knowledge points, E m = [ e m 、 1 , … , e m 、 q , … , e m 、 Q ] , 1 means the learner accepts Knowledge points are If it is 0, it means the learner refuses to present the knowledge point in this content type. Represents learning The time used by a resource, Indicates the upper limit of the learning time that learners can accept. is a decision sequence, where NET = [ Net 1 , … , Net l , … , Net L ] , is a decision-making element. If it is 1, it means resources are recommended to learners if If it is 0, it means that it is not recommended to the learner. In order to make personalized learning resources match learner characteristics and improve the quality of personalized learning resource generation services, 6 secondary fitness functions are established, as shown below:

[0111] (1)

[0112] Represents the difference between the learner's comprehension ability and the difficulty of the recommended resources.

[0113] (2)

[0114] Represents the difference between the learner's media type preference and the recommended resource media type. is an indicator function used to determine the sequence and In the Whether the elements at the positions match, if ,but ,otherwise , For two sequences in The positions indicate the weight of the function result.

[0115] (3)

[0116] To evaluate the balance between learners’ media type preferences and the media types of recommended resources. Represents the mean of the difference between learners’ media type preferences and the media types of recommended resources when learning each knowledge point.

[0117] (4)

[0118] Indicates the difference between learners’ content type preferences and recommended resource content types. For two sequences in The positions indicate the weight of the function result.

[0119] (5)

[0120] To evaluate the balance between the differences between learners’ content type preferences and recommended resource content types. Represents the mean of the difference between learners’ content type preferences and recommended resource content types when learning each knowledge point.

[0121] (6)

[0122] To assess whether the time required to complete the learning resources exceeds the upper limit of the learner's acceptable learning time.

[0123] (7)

[0124] (7) is the multi-dimensional matching fitness function, where , , , , and is the weight coefficient of the secondary fitness function, and the matching threshold is set . The smaller the value, the higher the quality of the recommended learning resources, and the more suitable they are for learners. The higher the value, the less suitable the learning resource is for recommending to learners.

[0125] (8)

[0126] is the mean fitness of the population, For population size.

[0127] Step Ab: Random Generation between [0,1] The dimensional vector is used as the initial position of the firefly population, and the same number of random numbers in the interval [0,1] are generated as the initial values ​​of the chaotic sequence, which can be expressed as X _ Cho h j = [ x _ Cho h 、 1 j , … , x _ Cho h 、 l j , … , x _ Cho h 、 L j ] . Set the maximum number of iterations , is the current iteration number.

[0128] Step Ba: The specific operation of the non-mandatory decision element update method is as follows: Map the element value in the continuous position space to the binary discrete space. Learning Resource Decision Sequence The corresponding firefly individual position vector The mapping is generated, where X h j = [ x h 、 1 j , … , x h 、 l j , … , x h 、 L j ] , h ∈ [ 1 , … , H ] , combined with After mapping the probability value, the learning resource decision sequence is expressed as follows:

[0129] (9)

[0130] (7) In the formula, is a random number in the range [0,1].

[0131] Combined with the decision sequence results in formula (9), the fitness function value of the learning resource corresponding to the decision sequence is calculated according to the matching fitness function formula (7).

[0132] Step Bb: The specific operation of determining the recommended learning resources is as follows: Sort the fitness function values ​​of each learning resource from small to large, and the fitness function value corresponding to the recommended learning resource must be the smallest. The individual corresponding to its decision sequence is the optimal individual of the firefly population, expressed as .

[0133] Step Bc: Update the statistics. Do deal with.

[0134] Step Ca: The updating formula of the adaptive convergence factor is as follows:

[0135] (10)

[0136] , Convergence factors The maximum and minimum values ​​of is the smoothing factor of the convergence curve.

[0137] Step Da: The specific operation of updating the optimal solution neighborhood information based on multidimensional space similarity theory is as follows:

[0138] Determine the historical optimal individual in the iteration process based on the matching fitness function , X HOpt = [ x 1 HOpt , … , x l HOpt , … , x L HOpt ] , randomly selected with Individuals with higher comprehensive similarity are , X j 、 Approx · best = [ x 1 j 、 Approx · best , … , x l j 、 Approx · best , … , x L j 、 Approx · best ] .

[0139] Step Db: Based on the multidimensional space similarity theory, the similarity evaluation standard calculation formula between individuals in the population and the historical best individual is as follows:

[0140] (11)

[0141] in, The calculation formula is as follows:

[0142] (12)

[0143] The calculation formula is as follows:

[0144] (13)

[0145] The calculation formula is as follows:

[0146] (14)

[0147] Standard parameters for similarity evaluation.

[0148] Step Dc: Calculate the comprehensive similarity between the population individuals and the optimal individual according to formula (11), sort the similarity values ​​from small to large, select the first H% of individual positions as candidate solutions, and randomly select a solution from them as .

[0149] Step Ea: The formula for updating the chaotic sequence based on the inverse sine mapping is as follows:

[0150] (15)

[0151] Step Fa: The specific operation of selecting the update method of an individual is as follows:

[0152] like , select and implement the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update the individual, For individuals The corresponding fitness function value. , and update individuals based on chaotic local search strategy.

[0153] Step Ga: The formula for the individual to update the optimal solution neighborhood information-guided strategy based on the adaptive adjustment mechanism is as follows:

[0154] (16)

[0155] in From step 3, From step 4, is a random number in the range [0,1]. It follows a Gaussian distribution with a mathematical expectation of 0 and a variance of 1.

[0156] Step Ha: The formula for updating the individual based on the chaotic search strategy is as follows:

[0157] (17)

[0158] in Obtained from step Ea.

[0159] Step Ia: Determine whether the multidimensional matching function value is lower than the threshold. The specific operation is as follows: Combined with formula (7), calculate the multidimensional matching fitness function value corresponding to the individual, and determine whether the minimum function value of the population is lower than the matching threshold. .

[0160] Step Ib: Determine whether the algorithm has reached the maximum number of iterations. Is it equal to the maximum number of iterations? .

[0161] Step Ja: The specific operation of determining the recommended learning resources is as follows: According to formula (7), the multi-dimensional matching fitness function value corresponding to the decision sequence of the recommended learning resources must be the smallest.

[0162] S10: Construct multi-dimensional matching fitness function and set population size

[0163] For learners The ability to understand a knowledge point. m ∈ [ 1 , … , M ] , is the number of knowledge points, Indicates Resources The difficulty of each knowledge point, l ∈ [ 1 , … , L ] , is the number of resources, Indicates Resources The media type of each knowledge point, B l 、 m = [ b l 、 m 、 1 , … , b l 、 m 、 k , … , b l 、 m 、 K ] , k ∈ [ 1 , … , K ] , is the number of media types, 1 means the Resources Knowledge points are If it is 0, it means it will not be presented in that type. Indicates that learners accept media type preference when learning a knowledge point, C m = [ c m 、 1 , … , c m 、 k , … , c m 、 K ] , 1 means the learner accepts Knowledge points are If it is 0, it means the learner refuses to present the knowledge point in this media type. Indicates Resources The content type of each knowledge point, D l 、 m = [ d l 、 m 、 1 , … , d l 、 m 、 q , … , d l 、 m 、 Q ] , q ∈ [ 1 , … , Q ] , is the number of content types, 1 means the Resources Knowledge points are If it is 0, it means that the content type is not presented. Indicates that learners accept Content type preference when consuming knowledge points, E m = [ e m 、 1 , … , e m 、 q , … , e m 、 Q ] , 1 means the learner accepts Knowledge points are If it is 0, it means the learner refuses to present the knowledge point in this content type. Represents learning The time used by a resource, Indicates the upper limit of the learning time that learners can accept. is a decision sequence, where NET = [ Net 1 , … , Net l , … , Net L ] , is a decision-making element. If it is 1, it means resources are recommended to learners if If it is 0, it means that it is not recommended to the learner. In order to make personalized learning resources match learner characteristics and improve the quality of personalized learning resource generation services, 7 secondary fitness functions are established, as shown below:

[0164] (1)

[0165] Represents the difference between the learner's comprehension ability and the difficulty of the recommended resources.

[0166] (2)

[0167] Represents the difference between the learner's media type preference and the recommended resource media type. is an indicator function used to determine the sequence and In the Whether the elements at the positions match, if ,but ,otherwise , For two sequences in The positions indicate the weight of the function result.

[0168] (3)

[0169] To evaluate the balance between learners’ media type preferences and the media types of recommended resources. Represents the mean of the difference between learners’ media type preferences and the media types of recommended resources when learning each knowledge point.

[0170] (4)

[0171] Indicates the difference between learners’ content type preferences and recommended resource content types. For two sequences in The positions indicate the weight of the function result.

[0172] (5)

[0173] To evaluate the balance between the differences between learners’ content type preferences and recommended resource content types. Represents the mean of the difference between learners’ content type preferences and recommended resource content types when learning each knowledge point.

[0174] (6)

[0175] To assess whether the time required to complete the learning resources exceeds the upper limit of the learner's acceptable learning time.

[0176] (7)

[0177] (6) is the multi-dimensional matching fitness function, where , , , , and is the weight coefficient of the secondary fitness function, and the matching threshold is set . The smaller the value, the higher the quality of the recommended learning resources, and the more suitable they are for learners. The higher the value, the less suitable the learning resource is for recommending to learners.

[0178] (8)

[0179] is the mean fitness of the population, For population size.

[0180] S11: Set the matching threshold, the maximum number of algorithm iterations, and randomly initialize the firefly population position and chaotic sequence

[0181] Random Generation between [0,1] The dimensional vector is used as the initial position of the firefly population, and the same number of random numbers in the interval [0,1] are generated as the initial values ​​of the chaotic sequence, which can be expressed as X _ Cho h j = [ x _ Cho h 、 1 j , … , x _ Cho h 、 l j , … , x _ Cho h 、 L j ] . Set the maximum number of iterations , is the current iteration number.

[0182] S12: Using non-mandatory decision element update method to determine recommended learning resources

[0183] Map the element values ​​in the continuous position space to the binary discrete space. Learning Resource Decision Sequence The corresponding firefly individual position vector The mapping is generated, where X h j = [ x h 、 1 j , … , x h 、 l j , … , x h 、 L j ] , h ∈ [ 1 , … , H ] , combined with After mapping the probability value, the learning resource decision sequence is expressed as follows:

[0184] (9)

[0185] (7) In the formula, is a random number in the range [0,1].

[0186] Combined with the decision sequence results in formula (9), the fitness function value of the learning resource corresponding to the decision sequence is calculated according to the matching fitness function formula (7).

[0187] Sort the fitness function values ​​of each learning resource from small to large. The fitness function value corresponding to the recommended learning resource must be the smallest. The individual corresponding to its decision sequence is the optimal individual of the firefly population, expressed as . Update statistics, Do deal with.

[0188] S13: Update adaptive convergence factor

[0189] The updating formula of the adaptive convergence factor is as follows:

[0190] (10)

[0191] , Convergence factors The maximum and minimum values ​​of is the smoothing factor of the convergence curve.

[0192] S14: Update the optimal solution neighborhood information based on multidimensional space similarity theory

[0193] The specific operations of updating the optimal solution neighborhood information based on multidimensional space similarity theory are as follows:

[0194] Determine the historical optimal individual in the iteration process based on the matching fitness function , X HOpt = [ x 1 HOpt , … , x l HOpt , … , x L HOpt ] , randomly selected with Individuals with higher comprehensive similarity are , X j 、 Approx · best = [ x 1 j 、 Approx · best , … , x l j 、 Approx · best , … , x L j 、 Approx · best ] .

[0195] Based on the multidimensional space similarity theory, the calculation formula for the similarity evaluation standard between individuals in the population and the historical best individual is as follows:

[0196] (11)

[0197] in, The calculation formula is as follows:

[0198] (12)

[0199] The calculation formula is as follows:

[0200] (13)

[0201] The calculation formula is as follows:

[0202] (14)

[0203] Standard parameters for similarity evaluation.

[0204] According to formula (11), the comprehensive similarity between the individuals in the population and the optimal individual is calculated, and the similarity values ​​are sorted from small to large. The first H% of individual positions are selected as candidate solutions, and a solution is randomly selected from them as .

[0205] S15: Update the chaotic sequence based on the inverse sine map

[0206] The formula for updating the chaotic sequence based on the inverse sine map is as follows:

[0207] (15)

[0208] S16: Determine whether the individual fitness value is greater than or equal to the average fitness value of the population

[0209] Determine individual fitness value Is it greater than or equal to the average fitness of the population? .

[0210] S17: Implement the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update individuals

[0211] The formula for updating the optimal solution neighborhood information-guided strategy based on the adaptive adjustment mechanism is as follows:

[0212] (16)

[0213] in From step S13, From step S14, is a random number in the range [0,1]. It follows a Gaussian distribution with a mathematical expectation of 0 and a variance of 1.

[0214] S18: Update individuals based on chaotic local search strategy

[0215] The formula for updating the individual based on the chaotic search strategy is as follows:

[0216] (17)

[0217] in Obtained from step S15.

[0218] S19: Determine whether the multidimensional matching function value is lower than the threshold or whether the algorithm has reached the maximum number of iterations

[0219] Combined with formula (7), calculate the multidimensional matching fitness function value corresponding to the individual, and determine whether the minimum function value of the population is lower than the matching threshold ,judge Is it equal to the maximum number of iterations? .

[0220] S20: Output recommended learning resources

[0221] According to formula (7), the multi-dimensional matching fitness function value corresponding to the decision sequence of the recommended learning resource must be the smallest.

[0222] Experiment: The examples in the reference ([1] Yang JM , Chen YP , Horng JT, et al. Applying family competition to evolution strategies for constrained optimization [C] / / Evolutionary Programming VI, 6th International Conference, EP97, Indianapolis, Indiana, USA, April 13-16, 1997, Proceedings. DBLP, 1997. DOI: 10.1007 / BFb0014812. [2] Li Haojun, Zhang Guang, Wang Wanliang, et al. Personalized learning resource recommendation method based on multi-dimensional feature differences [J]. Systems Engineering - Theory & Practice, 2017, 37(11): 2995-3005.) are selected for comparison with the results of the present invention.

[0223] In order to verify the performance of the modified Firefly Algorithm (MFA) in the personalized learning resource recommendation problem, four groups of experiments were implemented. The unified parameter settings are shown in Table 1, and the specific learner characteristics and learning resource characteristics are shown in Table 2.

[0224] Table 1 Parameter settings of the improved firefly population algorithm in the present invention

[0225]

[0226] Table 2 Learner characteristics and learning resource characteristics

[0227]

[0228]

[0229] After 50 independent experiments, the mean and variance data of each algorithm after convergence are shown in Table 3. The convergence curve corresponding to the recommended learning resources is shown in Figure 2-5As shown in Table 3, under different learner characteristics and learning resource characteristics, the MFA algorithm in the present invention has the best convergence accuracy and stability. From the convergence curve in the figure, it can be seen that when the MFA algorithm solves the problem of personalized learning resource recommendation, the multi-dimensional matching degree corresponding to the recommended learning resources is better than that of the competing algorithms.

[0230] Table 3 Convergence mean and variance

[0231]

[0232] Embodiment 2:

[0233] An online personalized learning resource recommendation system based on an improved firefly algorithm, comprising:

[0234] Acquisition module: used to obtain learners' user profiles and learning resource characteristics;

[0235] Output module: used to input the user portrait and learning resource characteristics into a pre-trained learning resource recommendation model, which constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics, solves the multi-dimensional matching fitness function through the improved firefly algorithm and outputs the recommended learning resources, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

[0236] The online personalized learning resource recommendation system based on the improved firefly algorithm provided in the embodiment of the present invention can execute the online personalized learning resource recommendation method based on the improved firefly algorithm provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0237] Embodiment three:

[0238] An embodiment of the present invention further provides a computer device, including a processor and a storage medium;

[0239] The storage medium is used to store instructions;

[0240] The processor is used to operate according to the instructions to perform the steps of the following method:

[0241] Obtain learners’ user profiles and learning resource characteristics;

[0242] The user portrait and learning resource characteristics are input into a pre-trained learning resource recommendation model, and the learning resource recommendation model constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics. The multi-dimensional matching fitness function is solved by the improved firefly algorithm and the recommended learning resources are output, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

[0243] For the details of each step in this embodiment, please refer to the first embodiment, which will not be described in detail here. Since this embodiment and the first embodiment adopt the same technical concept, it also has the technical effects described in the first embodiment.

[0244] Embodiment 4:

[0245] The embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the following method are implemented:

[0246] Obtain learners’ user profiles and learning resource characteristics;

[0247] The user portrait and learning resource characteristics are input into a pre-trained learning resource recommendation model, and the learning resource recommendation model constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics. The multi-dimensional matching fitness function is solved by the improved firefly algorithm and the recommended learning resources are output, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

[0248] For the details of each step in this embodiment, please refer to the first embodiment, which will not be described in detail here. Since this embodiment and the first embodiment adopt the same technical concept, it also has the technical effects described in the first embodiment.

[0249] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0250] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or multiple boxes.

[0251] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0252] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0253] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An online personalized learning resource recommendation method based on an improved firefly algorithm, characterized in that: include: Obtain learners’ user profiles and learning resource characteristics; Inputting the user portrait and learning resource characteristics into a pre-trained learning resource recommendation model, wherein the learning resource recommendation model constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics; The multi-dimensional matching fitness function is solved by an improved firefly algorithm and recommended learning resources are output, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

2. The online personalized learning resource recommendation method based on the improved firefly algorithm according to claim 1 is characterized in that: The user portrait includes the learner's ability level, media type preference, content type preference and upper limit of learning time, and the learning resource characteristics include learning resource difficulty, media type, content type and time required for learning resources.

3. The online personalized learning resource recommendation method based on the improved firefly algorithm according to claim 2 is characterized in that: Solving the multi-dimensional matching fitness function by improving the firefly algorithm and outputting recommended learning resources, including: Randomly initialize the positions and chaotic sequences of firefly populations; Execute the iteration step to obtain the final individual position until the termination condition is met, wherein the termination condition includes that the multi-dimensional matching fitness function value is lower than a preset threshold or the algorithm reaches a preset maximum number of iterations; When the termination condition is met, the final individual position is output as the recommended learning resource; The iterative steps include: The recommended learning resources are determined by adopting a non-mandatory decision element updating method, and the fitness function value of the recommended learning resources is obtained; Based on the fitness function value of the recommended learning resources, the adaptive convergence factor is updated; According to the fitness function value of the recommended learning resources and the updated adaptive convergence factor, the optimal solution neighborhood information is updated based on the multi-dimensional space similarity theory; Based on the updated optimal solution neighborhood information, the chaotic sequence is updated according to the inverse sine mapping; According to the updated chaotic sequence, the individual fitness value of the fireflies is compared with the average fitness value of the group, and the individual update method is selected according to the comparison result; According to the updating method of individuals, the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism is implemented to update individuals and obtain the updated individual positions; According to the updated individual position, the individual is updated based on the chaotic local search strategy to obtain the final individual position.

4. The online personalized learning resource recommendation method based on the improved firefly algorithm according to claim 3 is characterized in that: The multidimensional matching fitness function includes six secondary fitness functions, which are respectively expressed as follows: (1) is the first-level secondary fitness function, which represents the difference between the learner's understanding ability and the difficulty of the recommended resources; (2) is the secondary fitness function, which represents the difference between the learner's media type preference and the recommended resource media type; is an indicator function used to determine the sequence and In the Whether the elements at the positions match, if ,but ,otherwise , For two sequences in The weight of the indicator function result at each position; (3) is a three-level secondary fitness function, which is used to evaluate the balance between the learner's media type preference and the recommended resource media type; It represents the mean of the difference between the learner's media type preference and the recommended resource media type when learning each knowledge point; (4) is a four-level secondary fitness function, which represents the difference between the learner’s content type preference and the recommended resource content type. For two sequences in The weight of the indicator function result at each position; (5) is a five-level secondary fitness function, which is used to evaluate the balance between the learner's content type preference and the recommended resource content type; It represents the mean of the difference between learners’ content type preference and recommended resource content type when learning each knowledge point; (6) It is a six-level secondary fitness function, which is used to evaluate whether the time required to complete the learning resources exceeds the upper limit of the learner's acceptable learning time; (7) (7) is the multi-dimensional matching fitness function, where , , , , and is the weight coefficient of the secondary fitness function, and the matching threshold is set ; The smaller the value, the higher the quality of the recommended learning resources, and the more suitable they are for learners. The higher the value, the less suitable the learning resource is for recommending to learners; (8) is the mean fitness of the population, is the population size; Random Generation between [0,1] The dimensional vector is used as the initial position of the firefly population, and the same number of random numbers in the interval [0,1] are generated as the initial values ​​of the chaotic sequence, which can be expressed as ; in, For learners Ability to understand knowledge points; , is the number of knowledge points, Indicates Resources The difficulty of each knowledge point, , is the number of resources, Indicates Resources The media type of each knowledge point, , , is the number of media types, 1 means the Resources Knowledge points are If it is 0, it means it is not presented in this type. Indicates that learners accept media type preference when learning a knowledge point, , 1 means the learner accepts Knowledge points are If it is 0, it means that the learner refuses to present the knowledge point in this media type. Indicates Resources The content type of each knowledge point, , , is the number of content types, 1 means the Resources Knowledge points are If it is 0, it means that the content type is not presented. Indicates that learners accept Content type preference when consuming knowledge points, , 1 means the learner accepts Knowledge points are content type, otherwise 0 means that the learner refuses to present the knowledge point in this content type; Represents learning The time used by a resource, Indicates the upper limit of the learning time that learners can accept; is a decision sequence, where , is a decision-making element. If it is 1, it means resources are recommended to learners if If it is 0, it means that it is not recommended to the learner. is the maximum number of iterations, is the current iteration number.

5. The online personalized learning resource recommendation method based on the improved firefly algorithm according to claim 4 is characterized in that: The non-mandatory decision element updating method is used to determine the recommended learning resources, including: Map the element values ​​in the continuous position space to the binary discrete space. Learning Resource Decision Sequence The corresponding firefly individual position vector The mapping produces, where , , combined with After mapping the probability value, the learning resource decision sequence is expressed as follows: (9) (9) In the formula, is a random number in the range [0,1]; Combined with the decision sequence results in formula (9), the fitness function value of the learning resource corresponding to the decision sequence is calculated according to the matching fitness function formula (7); Sort the fitness function values ​​of each learning resource from small to large. The fitness function value corresponding to the recommended learning resource must be the smallest. The individual corresponding to its decision sequence is the optimal individual of the firefly population, expressed as ; Update statistics, Do deal with.

6. The online personalized learning resource recommendation method based on the improved firefly algorithm according to claim 5 is characterized in that: The formula for updating the adaptive convergence factor includes: (10) , Convergence factors The maximum and minimum values ​​of is the smoothing factor of the convergence curve; The updating of the optimal solution neighborhood information based on multidimensional space similarity theory includes: Determine the historical optimal individual in the iteration process based on the matching fitness function , , randomly selected with Individuals with higher comprehensive similarity are , ; The calculation formula for the similarity evaluation standard between individuals in the population and the historical best individual is as follows: (11) in, The calculation formula is as follows: (12) The calculation formula is as follows: (13) The calculation formula is as follows: (14) Standard parameters for similarity assessment; According to formula (11), the comprehensive similarity between the individuals in the population and the optimal individual is calculated, and the similarity values ​​are sorted from small to large. The first H% of individual positions are selected as candidate solutions, and a solution is randomly selected from them as .

7. The online personalized learning resource recommendation method based on the improved firefly algorithm according to claim 6 is characterized in that: Update the chaotic sequence according to the inverse sine map, including: (15) The updating method of selecting an individual according to the comparison result includes: like , select and implement the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update the individual, For individuals The corresponding fitness function value; if , implement the update of individuals based on chaotic local search strategy; The implementation of the optimal solution neighborhood information guidance strategy based on the adaptive adjustment mechanism to update individuals includes: (16) is a random number in the range [0,1]. It follows a Gaussian distribution with a mathematical expectation of 0 and a variance of 1; The updating of individuals based on the chaotic local search strategy includes: (17) in It is obtained by updating the chaotic sequence according to the inverse sine map.

8. An online personalized learning resource recommendation system based on an improved firefly algorithm, characterized in that: include: Acquisition module: used to obtain learners' user profiles and learning resource characteristics; Output module: used to input the user portrait and learning resource characteristics into a pre-trained learning resource recommendation model, which constructs a multi-dimensional matching fitness function between the learner and the learning resource according to the learner's user portrait and learning resource characteristics, solves the multi-dimensional matching fitness function through the improved firefly algorithm and outputs the recommended learning resources, wherein the improved firefly algorithm uses shape similarity information to randomly select suboptimal individuals from the neighboring area of ​​the optimal individual in the population as search leaders to perform global search.

9. A computer device, characterized in that: including processor and storage medium; The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 7 are implemented.

Images