Personalized traffic safety education method and system and medium

Abstract

The invention discloses a personalized traffic safety education method, system and method. The method comprises the following steps: collecting and processing learning behavior data of a learner in real time; optimizing fuzzy rules and parameters of the adaptive neural fuzzy inference system by using a particle swarm optimization algorithm, constructing a classification model, and dividing the processed learning behavior data into different types of learners according to the processed learning behavior data; according to the type of the learner, recommending a personalized learning scheme; and executing a personalized learning decision, and implementing pushing and interaction operation of traffic safety education contents. According to the method, intelligent optimization and fuzzy reasoning technologies are fused, analysis and classification are carried out according to behavior characteristics of the learners, accurate learner type division is realized, accurate learning suggestions are provided, the traffic safety education effect is improved, and the problem that a traditional traffic safety education mode is lack of individuation is effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of traffic safety education, and particularly relates to a personalized traffic safety education method, system and medium. Background Art

[0002] The high incidence of traffic accidents has a significant impact on social safety and economic development, and the safety awareness and behavior of drivers are key factors. Most of the existing traffic safety education adopts unified training modes such as classroom education, driver's manuals, and simulated driving, which are difficult to meet the personalized needs of learners, resulting in limited educational effects. In view of the personalized characteristics of learners' learning time, entertainment time, and the correct rate of answering questions, using intelligent technologies to classify them and provide targeted education is an important direction to improve the effect of traffic safety education.

[0003] The Adaptive Network-based Fuzzy Inference System (ANFIS) combines the advantages of fuzzy logic and neural networks and is suitable for modeling complex problems; the Particle Swarm Optimization algorithm (PSO) can optimize the system performance. How to achieve more efficient dissemination and practical application of traffic safety knowledge through data analysis and classification to improve the safety awareness of practitioners and reduce the incidence of traffic accidents has become an urgent problem to be solved in the traffic industry. Summary of the Invention

[0004] Object of the Invention: To solve the above technical problems, the present invention provides a personalized traffic safety education method, system and medium. By combining PSO and ANFIS, accurate modeling and classification of learners' characteristics can be achieved, providing technical support for personalized education. Through data analysis and classification, more efficient dissemination and practical application of traffic safety knowledge can be realized.

[0005] Technical Solution: A personalized traffic safety education method provided by the present invention includes the following steps:

[0006] (1) Collect and process the learning behavior data of learners in real time;

[0007] (2) Use the Particle Swarm Optimization algorithm to optimize the fuzzy rules and parameters of the Adaptive Neuro-Fuzzy Inference System, construct a classification model, and divide the processed learning behavior data into different types of learners according to it;

[0008] (3) Recommend personalized learning plans according to the learner types;

[0009] (4) Execute personalized learning decisions and implement the push and interactive operations of traffic safety education content.

[0010] Further, the learning behavior data of the learners in step (1) includes the learning time, entertainment time, and the correct rate of answering questions in stage tests of the learners.

[0011] Furthermore, the learning behavior data collected in step (1) also needs to be preprocessed. The preprocessing steps include:

[0012] (11) Data filling: To ensure that the data is on a complete time scale, missing values are filled;

[0013] (12) Data standardization: The data is converted to a unified dimension for subsequent processing;

[0014] (13) Data extraction: Extract features from the original data that have an important impact on personalized traffic safety education.

[0015] Furthermore, the data filling method in step (11) is as follows: When the data at a certain moment is missing, the activity at the previous moment is used to fill it. For example, if the activity of the learner at the previous moment is in the learning area and the activity at the next moment is in the entertainment area, then the time in the learning area is used to fill it; when a period of time is missing, the two activities before and after this period are used, and the two activity times are weighted and averaged. For example, if the previous activity is in the learning area and the next activity is in the entertainment area, then each of the two activities accounts for half of the missing time; for example, if the previous activity and the next activity are the same, then the missing time of this period is also filled with the same activity.

[0016] Furthermore, the processing in step (1) is to scale the learning behavior data of the learner in the minimum-maximum normalization method, and its formula is:

[0017]

[0018] Where:

[0019] X s : The data after normalization;

[0020] X: The original data;

[0021] X max : The maximum value in the data;

[0022] X min : The minimum value in the data.

[0023] Furthermore, the particle swarm optimization algorithm in step (2) adopts the following formula:

[0024] PSO particle optimization speed update formula:

[0025]

[0026] PSO particle optimization position update formula:

[0027]

[0028] Among them,

[0029] the velocity of particle i in the (t + 1)-th generation;

[0030] ω: Inertia weight, which controls the balance between the global solution search and the local solution search of the particle;

[0031] c1, c2: Learning factors (acceleration constants), which respectively control the speed at which the particle approaches its own and the global optimal solutions;

[0032] r1, r2: Random numbers (range [0, 1]), which introduce randomness and enhance the exploration ability of the algorithm;

[0033] the historical optimal position of particle i in the t-th generation;

[0034] G t : Global optimal position, which is the position of the optimal solution in the current iteration of the t-th generation among all particles;

[0035] the current position of particle i in the t-th generation.

[0036] Furthermore, the learner classification formula of the adaptive neuro-fuzzy inference system described in step (2) is as follows

[0037] Learner-Type = ANFIS(study-time, amusing-time, accuracy-rate)

[0038] Among them,

[0039] Learner-Type: Learner type;

[0040] ANFIS: Adaptive neural network;

[0041] study-time: Personal time;

[0042] amusing-time: Personal entertainment time;

[0043] accuracy-rate: Personal question-solving accuracy rate.

[0044] Furthermore, the specific process of step (2) is that the adaptive neuro-fuzzy inference system first receives the data of the learner, conducts fuzzy inference, generates preliminary classification rules, and then the particle swarm optimization algorithm conducts global search and improvement on the preliminary classification rules to determine the optimal classification rules.

[0045] The present invention provides a personalized traffic safety education system, including:

[0046] A data collection unit for collecting the learning time, entertainment time, and the correct rate of stage test answers of learners;

[0047] A learner classification unit that uses the particle swarm optimization algorithm to optimize the fuzzy rules and parameters of the adaptive neuro - fuzzy inference system, constructs a classification model, and classifies learners into different types according to their learning behavior characteristics;

[0048] A personalized learning recommendation unit that provides customized learning paths and recommended content for different types of learners according to the learner classification results;

[0049] An execution unit for executing the personalized learning decisions of the control unit and implementing the push and interactive operations of traffic safety education content.

[0050] The present invention also provides a computer - readable storage medium, on which a personalized traffic safety education program is stored. The personalized traffic safety education program is designed to execute the method of personalized traffic safety education as described above.

[0051] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following significant advantages: Using the PSO - ANFIS model to analyze and classify according to the behavior characteristics of learners, it realizes accurate classification of learner types, provides accurate learning suggestions, improves the effect of traffic safety education, and effectively solves the problem of lack of personalization in the traditional traffic safety education mode. The system dynamically monitors the behavior data of learners, evaluates the learning effect through stage tests, and adjusts the recommended content in real time to ensure the learning progress and effect. The present invention integrates intelligent optimization and fuzzy inference technologies, can realize dynamic and personalized learning path planning for different learning types, and is applicable to traffic training institutions, driving schools, and public traffic safety education platforms. Description of the Drawings

[0052] Figure 1 It is a schematic flow chart of the method for personalized traffic safety education of the present invention;

[0053] Figure 2 It is a reasoning rule diagram of the classification model of the present invention;

[0054] Figure 3 It is a reasoning explanation diagram of the classification model of the present invention. Detailed Embodiments

[0055] The following further details the technical solutions of the present invention through specific embodiments. However, it is necessary to point out that the following embodiments are only for explaining the technical concept and characteristics of the present invention. The purpose is to enable those who are familiar with this technology to understand the content of the present invention and implement it accordingly, and it cannot be used to limit the protection scope of the present invention.

[0056] A personalized traffic safety education system, comprising:

[0057] A data collection unit for collecting the learning time, entertainment time and the correct rate of answering questions in stage tests of learners;

[0058] A learner classification unit that uses the particle swarm optimization algorithm to optimize the fuzzy rules and parameters of the adaptive neuro-fuzzy inference system, constructs a classification model, and divides learners into different types according to the learning behavior characteristics of learners;

[0059] A personalized learning recommendation unit that provides customized learning paths and recommended content for different types of learners according to the learner classification results;

[0060] An execution unit for executing the personalized learning decisions of the control unit and implementing the push and interaction operations of traffic safety education content.

[0061] Among them, the data collection frequency is once per minute.

[0062] A personalized traffic safety education method provided by the present invention includes the following steps:

[0063] (1) Collect and process the learning behavior data of learners in real time. The learning behavior data includes the learning time, entertainment time and the correct rate of answering questions in stage tests of learners;

[0064] (2) Use the particle swarm optimization algorithm to optimize the fuzzy rules and parameters of the adaptive neuro-fuzzy inference system, construct a classification model, and divide learners into different types according to the processed learning behavior data;

[0065] (3) Recommend personalized learning plans according to the learner types;

[0066] (4) Execute personalized learning decisions and implement the push and interaction operations of traffic safety education content.

[0067] Among them, the learning behavior data collected in step (1) needs to be preprocessed. The preprocessing steps include:

[0068] (11) Data filling: To ensure that the data is on a complete time scale, missing data is filled. The data filling method is as follows: When the data at a certain moment is missing, the activity at the previous moment is used for filling. For example, if the activity of the learner at the previous moment is in the learning area and the activity at the next moment is in the entertainment area, the time in the learning area is used for filling; when a certain period of time is missing, according to the two activities before and after this period of time, the two activity times are weighted and averaged. For example, if the previous activity is in the learning area and the next activity is in the entertainment area, each of the two activities accounts for half of the missing time of this period; for example, if the previous activity and the next activity are the same, the missing time of this period is also filled with the same activity.

[0069] (12) Data standardization: Convert the data into a unified dimension for subsequent processing.

[0070] (13) Data extraction: Extract features from the original data that have an important impact on personalized traffic safety education.

[0071] In step (1), the learning behavior data of the learner is scaled in the min-max normalization method, and its formula is:

[0072]

[0073] Where:

[0074] X s : The data after normalization;

[0075] X: The original data;

[0076] X max : The maximum value in the data;

[0077] X min : The minimum value in the data.

[0078] In step (2), the particle swarm optimization algorithm described adopts the following formula:

[0079] The velocity update formula for PSO particle optimization:

[0080]

[0081] The position update formula for PSO particle optimization:

[0082]

[0083] Where,

[0084] The velocity of particle i in the (t + 1)-th generation;

[0085] ω: Inertia weight, controlling the balance between the global solution search and the local solution search of the particle;

[0086] c1, c2: Learning factors (acceleration constants), respectively controlling the speed at which the particle approaches its own and the global optimal solutions;

[0087] r1, r2: Random numbers (range [0, 1]), introducing randomness to enhance the exploration ability of the algorithm;

[0088] The historical optimal position of particle i in the t-th generation;

[0089] G t:Global optimal position, which is the position of the optimal solution among all particles in the current iteration of the t-th generation;

[0090] The current position of particle i in the t-th generation.

[0091] The classification formula of the learner for the adaptive neuro-fuzzy inference system is as follows

[0092] Learner-Type = ANFIS(study-time, amusing-time, accuracy-rate)

[0093] Where,

[0094] Learner-Type: Learner type;

[0095] ANFIS: Adaptive neural network;

[0096] study-time: Personal time;

[0097] amusing-time: Personal entertainment time;

[0098] accuracy-rate: Personal question-solving accuracy rate.

[0099] The classification model uses the PSO algorithm to optimize the fuzzy rules of ANFIS. According to the processed learning behavior data, the specific process of dividing different types of learners is as follows: First, the adaptive neuro-fuzzy inference system receives the data of the learners, performs fuzzy inference, and generates preliminary classification rules. Then, the particle swarm optimization algorithm performs global search and improvement on the preliminary classification rules to determine the optimal classification rules.

[0100] In the present invention, the following fuzzy rules are followed:

[0101] 1. Learner type 1: If the learning time is long, and the entertainment time is long, and the answering correct rate is high;

[0102] 2. Learner type 2: If the learning time is long, and the entertainment time is long, and the answering correct rate is low;

[0103] 3. Learner type 3: If the learning time is long, and the entertainment time is short, and the answering correct rate is high;

[0104] 4. Learner type 4: If the learning time is short, and the entertainment time is long, and the answering correct rate is high;

[0105] 5. Learner type 5: If the learning time is short, and the entertainment time is long, and the answering correct rate is high;

[0106] 6. Learner Type 6: If the learning time is short, the entertainment time is long, and the answering accuracy rate is low;

[0107] 7. Learner Type 7: If the learning time is long, the entertainment time is short, and the answering accuracy rate is high;

[0108] 8. Learner Type 8: If the learning time is short, the entertainment time is short, and the answering accuracy rate is low.

[0109] The rules are expressed in function form as:

[0110]

[0111] Among them, Learner Type 1 represents time - management learners, Learner Type 2 represents scattered learners, Learner Type 3 represents top - student learners, Learner Type 4 represents efficient learners, Learner Type 5 represents entertainment - oriented learners, Learner Type 6 represents potential learners, Learner Type 7 represents inefficient learners, Learner Type 8 represents confused learners, st represents learning time, at represents entertainment time, and ar represents answering accuracy rate.

[0112] Furthermore, according to the learner types recommended, the system will recommend different learning modes of different styles for them. The following are the recommendations:

[0113] Time - management learners: Balanced learning mode. Adopt a combination of fixed learning periods and rest periods, appropriately reduce entertainment time, improve the efficiency of in - depth learning, and at the same time provide test questions with slightly higher difficulty.

[0114] Scattered learners: Basic improvement mode. Reduce the complexity of learning tasks, mainly focus on basic traffic safety knowledge, learn in modules, and increase learning participation through a reward mechanism.

[0115] Top - student learners: Advanced learning mode. Provide high - difficulty challenges and comprehensive test questions, guide learners to expand the breadth and depth of knowledge, and at the same time appropriately arrange entertainment and relaxation time to avoid over - learning.

[0116] Efficient learners: Interest - driven mode. Attract attention through gamified learning content or competition forms, increase the learning duration of traffic safety knowledge and provide content that combines entertainment and learning.

[0117] Entertainment - oriented learners: Basic improvement mode. Reduce the complexity of learning tasks, mainly focus on basic traffic safety knowledge, learn in modules, and increase learning participation through a reward mechanism.

[0118] Potential learners: Recommended intensive learning tasks, increased efficient knowledge expansion and testing to help learners fully explore their potential, and encourage appropriate extension of learning time.

[0119] Inefficient learners: Optimize the arrangement of learning tasks and reduce inefficient repetitive content. Recommend more concise traffic safety learning materials, and at the same time strengthen result feedback and learning skill guidance.

[0120] Confused learners: Set interesting content to attract attention, introduce basic courses, gradually cultivate learners' interest and confidence in traffic safety knowledge, and adopt a reward mechanism to enhance learning motivation.

[0121] The execution unit in step (4) is used to execute the personalized learning decisions of the control unit and implement the push and interaction operations of traffic safety education content. The system execution unit consists of multiple sub-modules, mainly including pushing customized traffic safety learning content for different types of learners according to the classification results of the control unit; providing a real-time testing function to evaluate learners' mastery of traffic safety knowledge; monitoring learners' learning processes and recording the status of completed tasks and uncompleted tasks.

[0122] The operation steps of the execution unit are as follows:

[0123] S01: Control instruction parsing: The execution unit receives the learning plan instructions generated by the control unit, parses them to clarify the learning content push type, test task requirements, and progress tracking method.

[0124] S02: Learning content customization and push: According to the learner type, push through short videos, animation cases, simulation scenarios, etc. to the learner's device side to adapt to their learning characteristics.

[0125] S03: Interactive testing and feedback: Embed real-time interactive testing during the learning process. The system automatically generates personalized test questions, records the learner's mastery level based on their answering situation, and feedbacks it to the control unit for optimizing the next learning strategy.

[0126] S04: Learning process synchronization and coordination: According to the proportion of learners' learning time and entertainment time, coordinate the execution of learning tasks and rest tasks to avoid task conflicts.

[0127] S05: Data monitoring and feedback: Real-time monitor learners' learning progress, answering accuracy rate, and participation, collect behavior data, and feedback the analysis results to the control unit to provide a basis for subsequent optimization.

[0128] S06: Abnormality detection and handling: The execution unit has an abnormality detection function to identify problems such as learners' learning interruption and abnormal answering rate, and automatically take measures such as supplementary content push or notifying the system administrator.

[0129] S07: User Interface Interaction: Provide a user interface where learners can view their learning progress, completed tasks, and learning data analysis results; administrators can monitor the overall operation status of the learning system, manually adjust the push strategy, or learning content.

[0130] S08 Security and Reliability Design: Ensure the security of the execution unit, including data encryption during transmission, backup of learning records, and high-availability design for content push, to prevent data loss and system interruption.

[0131] S09 Execution Effect Evaluation: Regularly evaluate the operation effect of the execution unit, including the completion rate of learning content, test accuracy, and learner feedback.

[0132] S10: Execution Record and Audit: Record all operation histories of the execution unit, including pushed content, test results, and learning behavior data, to provide detailed basis for subsequent optimization and improvement, and at the same time support audit and problem tracking functions.

[0133] The present invention also provides a computer-readable storage medium, on which a personalized traffic safety education program is stored, and the personalized traffic safety education program is designed to execute the method of personalized traffic safety education as described above.

[0134] The personalized traffic safety education method based on the Particle Swarm Optimization Algorithm (PSO) and Adaptive Neuro-Fuzzy Inference System (ANFIS) provided by the present invention provides accurate learning suggestions according to the behavioral characteristics of learners, improving the effect of traffic safety education. Dynamically monitor the behavioral data of learners, evaluate the learning effect through stage tests, and adjust the recommended content in real time to ensure the learning progress and effect. The present invention integrates intelligent optimization and fuzzy inference technologies, can achieve dynamic and personalized learning path planning for different learning types, and is applicable to traffic training institutions, driving schools, and public traffic safety education platforms.

Claims

1. A personalized traffic safety education method, characterized in that: The following steps are involved: (1) Collect and process learners’ learning behavior data in real time; (2) Using the particle swarm optimization algorithm to optimize the fuzzy rules and parameters of the adaptive neural fuzzy inference system, a classification model is constructed to classify learners into different types based on the processed learning behavior data; (3) Recommend personalized learning plans based on learner types; (4) Implement personalized learning decisions and implement the push and interactive operations of traffic safety education content.

2. The personalized traffic safety education method according to claim 1, characterized in that: The learner's learning behavior data in step (1) includes the learner's learning time, entertainment time and the correct rate of answering questions in the stage test.

3. The personalized traffic safety education method according to claim 1, characterized in that: The learning behavior data collected in step (1) is also preprocessed. The preprocessing steps include: (11) Data filling: To ensure that the data is complete in time scale, missing data are filled; (12) Data standardization: converting data into a unified dimension to facilitate subsequent processing; (13) Data extraction: Extract features from the raw data that have important implications for personalized traffic safety education.

4. The personalized traffic safety education method according to claim 3, characterized in that: The data filling method in step (11) is as follows: when the data at a certain moment is missing, the activity at the previous moment is used to fill it; when the data for a certain period of time is missing, the two activity times are weighted averaged based on the two activities before and after the period of time.

5. The personalized traffic safety education method according to claim 1, characterized in that: The processing described in step (1) is to scale the learner's learning behavior data using the minimum-maximum normalization method, and the formula is: in: X s : Normalized data; X: original data; X max : The maximum value in the data; X min : The minimum value in the data.

6. The personalized traffic safety education method according to claim 1, characterized in that: The particle swarm optimization algorithm described in step (2) adopts the following formula: The speed update formula of PSO particle optimization is: PSO particle optimization position update formula: in, The velocity of particle i in generation t+1; ω: Inertia weight, which controls the balance between particle search for global solution and local solution; c1, c2: learning factors (acceleration constants), which control the speed at which particles approach their own and the group's optimal solutions respectively; r1, r2: random numbers (range [0, 1]), which introduce randomness and enhance the algorithm's exploration ability; The historical optimal position of particle i in generation t; G t : The global optimal position is the position of the optimal solution of the current iteration t generation among all particles; The current position of particle i at generation t.

7. The personalized traffic safety education method according to claim 1, characterized in that: The learner classification formula of the adaptive neuro-fuzzy inference system in step (2) is as follows: Learner-Type=ANFIS(study-time,amusing-time,accuracy-rate) in, Learner-Type: learner type; ANFIS: Adaptive Neural Networks; study-time: personal time; amusing-time: personal recreational time; accuracy-rate: The accuracy rate of individual test-taking.

8. The personalized traffic safety education method according to claim 1, characterized in that: The specific process of step (2) is that the adaptive neural fuzzy inference system first receives the learner's data, performs fuzzy inference, generates preliminary classification rules, and then uses the particle swarm optimization algorithm to globally search and improve the preliminary classification rules to determine the optimal classification rules.

9. A personalized traffic safety education system, characterized in that: include: Data collection unit, used to collect learners' study time, entertainment time and correct answer rate of stage test questions; The learner classification unit uses the particle swarm optimization algorithm to optimize the fuzzy rules and parameters of the adaptive neural fuzzy inference system, builds a classification model, and divides learners into different types according to their learning behavior characteristics; Personalized learning recommendation unit, which provides customized learning paths and recommended content for different types of learners based on learner classification results; The execution unit is used to execute the personalized learning decisions of the control unit and implement the push and interactive operations of traffic safety education content.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores a personalized traffic safety education program, and the personalized traffic safety education program is designed to execute the personalized traffic safety education method as described in any one of claims 1-8.

Images