Electric power safety training management system and method
Through linear planning and genetic algorithms, the learning bottlenecks in power safety training are identified and optimized, and personalized management of power safety training is realized, learning efficiency and effect are improved, and the existing system is solved in the fixed training plan and lack of dynamic optimization. The problem.
Patent Information
- Application Number
- CN202411138562.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2025-01-07
AI Technical Summary
When managing and implementing training plans, the existing power safety training system lacks consideration of individual students' differences, resulting in poor training results and the inability to adjust the training plan in real time according to the students' learning progress, resulting in students' possible encounters during the learning process. Bottleneck without timely help.
Linear planning and genetic algorithms are used to determine the training courses, time and resource allocation, establish a learning curve model and a linear regression model, identify learning bottlenecks and optimize them, dynamically adjust the training plan, and ensure students' personalized learning planning.
Through personalized training plans, learning efficiency and effectiveness are improved, students' bottlenecks in the learning process are timely identified and solved, resource allocation is optimized, training costs are reduced, and employees' safety operation skills and risk awareness are improved.
Smart Images
Figure CN119272974A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of training management, and more specifically, to an electric power safety training management system and method. Background Art
[0002] In the power industry, employee safety training is crucial. The power generation and transmission and distribution process involves many high-risk operations. Employees' safe operation skills and risk awareness are directly related to the stable operation of the power system and personal safety. With the increasing complexity and automation of the power system, the training requirements for employees are becoming increasingly stringent. However, the existing power safety training system often faces the following problems when managing and implementing training plans:
[0003] Fixed training plans: Traditional power safety training plans are usually based on fixed course settings and time arrangements, and lack consideration for the differences of individual trainees. This training method is difficult to meet the needs of different trainees, resulting in unsatisfactory training results.
[0004] Lack of dynamic optimization capabilities: Existing training systems usually evaluate training effectiveness based on static data and are unable to adjust training plans in real time based on trainees’ learning progress and feedback. This causes trainees to encounter bottlenecks during the learning process without receiving timely help and adjustments, thus affecting learning outcomes.
[0005] Therefore, an electric power safety training management system and method are proposed to solve the above problems. Summary of the invention
[0006] To achieve the above object, the present invention provides the following technical solutions:
[0007] A power safety training management method comprises the following steps:
[0008] Use linear programming to determine the training courses, training time, resource allocation, and coaching times for each trainee, ensure a reasonable training sequence, avoid resource waste and work conflicts; generate training forms;
[0009] Conduct student training according to the training form and regularly collect training information of each trainee;
[0010] Establish a learning curve model based on training information, identify and mark the training courses where the target trainees have learning bottlenecks, and obtain the training course set that the target trainees need to optimize;
[0011] According to the training information and the training course set that needs to be optimized, a linear regression model is established for each training course that needs to be optimized to identify the degree of influence of the influencing factors of the target trainees on the trainees' performance;
[0012] Based on the training course set that the target trainees need to optimize, the genetic algorithm is used to optimize the personal learning plan of the target trainees.
[0013] In a preferred embodiment, the training form includes the courses each trainee attends, the time schedule of each course, resource allocation, and the number of tutoring sessions.
[0014] In a preferred embodiment, the logic of linear programming is:
[0015] Step 1: Determine the objective function: The goal is to maximize the learning effect of the students, and the objective function is expressed as: E ij represents the expected learning effect of student i in course j, X ij is a binary variable, indicating whether trainee i participates in course j (1 for participation, 0 for non-participation), N represents the total number of trainees participating in the training, and M represents the total number of courses;
[0016] Step 2: Determine the constraints: satisfy the following constraint formula: λ j Indicates the constraint type. When its value is 1, it indicates a time constraint in the constraint formula, i.e., the total learning time of a student on the course cannot exceed his / her available time. When its value is 2, it indicates a resource constraint in the constraint formula, i.e., the total resource consumption of a course cannot exceed its resource upper limit. When its value is 3, it indicates a tutoring times constraint in the constraint formula, i.e., the tutoring times of a course cannot exceed its upper limit. It means that the constraint formula is valid for every course j;
[0017] Step 3: Planning and solving: Use linear programming solving tools (such as Python's PuLP or SciPy) to solve the results that meet steps 1 and 2. The result is a training form that contains training courses, time, resource allocation, and number of tutoring sessions.
[0018] In a preferred embodiment, the influencing factors of trainees participating in the training include learning behavior factors and training course factors.
[0019] Learning behavior factors include learning frequency, participation, homework completion, and number of exercises; training course factors include course difficulty, relevance of course content, and teaching methods.
[0020] In a preferred embodiment, establishing a learning curve model based on training information refers to:
[0021] The students' scores in each learning stage and the time to complete the task are obtained from the training information. The changes in the scores of each student in each course over time are fitted to generate a learning curve. The learning curve is obtained by building a linear model. The slope of the learning curve in the linear model is the learning progress rate, which represents the increase in the students' scores per unit time in the course.
[0022] In a preferred embodiment, identifying and marking training courses in which target trainees have learning bottlenecks refers to:
[0023] The learning progress rate of the target trainees in the target course is obtained, and the average learning progress rate of all trainees in the target course is calculated. The learning bottleneck threshold is set to p times the average progress rate of all trainees, where p is greater than zero and less than one. For each trainee i, if his learning progress rate in the target course is lower than the learning bottleneck threshold, it is considered that the student has a learning bottleneck in the course and is marked. All marked courses are summarized to obtain the set of training courses that need to be optimized for the trainee i.
[0024] In a preferred embodiment, according to the training information and the training course set to be optimized, establishing a linear regression model for each training course to be optimized refers to:
[0025] Obtain the learning performance, corresponding data of learning behavior factors, and corresponding data of training course factors in the training information of each training course that needs to be optimized, and then establish a linear regression model:
[0026] S ij =β ij +β1F ij +β2P ij +β3A ij +β4L ij +β5D ij +β6R ij +β7T ij +∈ ij ; S ij represents the learning performance of student i in course j, β ij Represents the intercept term in the linear regression model established by student i for course j, ∈ ij represents the error term in the linear regression model established for student i on course j. β1, β2, β3, β4, β5, β6, and β7 are regression coefficients, which respectively represent the influence of learning frequency, participation, homework completion, number of exercises, course difficulty, course content relevance, and teaching method on learning performance. F ij represents the learning frequency of student i in course j, P ij represents the participation of student i in course j, Aij represents the completion of student i’s homework in course j, L ij represents the number of exercises that student i has done in course j, D ij represents the course difficulty of course j, R ij represents the relevance of the course content of course j, T ij represents the teaching method of course j;
[0027] Use the least squares method to estimate the regression coefficients of the linear regression model and ensure that the application performance of the linear regression model meets the requirements.
[0028] In a preferred embodiment, using a genetic algorithm to optimize the individual learning plan of the target trainee includes:
[0029] Calculate the learning bottleneck percentage value BJ of each training course that needs to be optimized. The learning bottleneck percentage value BJ is obtained by calculating the difference between the learning bottleneck threshold of the target training course and the actual learning progress rate, and then dividing the difference by the percentage value of the learning bottleneck threshold of the target training course. Then, the learning bottleneck percentage value BJ is calculated by ratio with the preset proportional coefficient R and rounded up to obtain the k value. The learning bottleneck percentage value BJ is the numerator. According to the k value, the learning behavior factors corresponding to the top k regression coefficients of the target training course and the vector composed of the training course factors are obtained, that is, the genes in the initialized genetic algorithm population in the genetic algorithm.
[0030] The learning plan of the target trainee is represented as a chromosome, each chromosome contains multiple genes, and each gene represents a combination of learning behavior and training course factors;
[0031] Fitness evaluation: Calculate the difference between the learning progress rate after optimization and the learning progress rate before optimization of all training courses that need to be optimized and sum them up to get the comprehensive improvement value. The fitness function is expressed as: the fitness is the comprehensive improvement value minus the total amount of resources required for the optimization operation;
[0032] Selection operation: Use the roulette wheel selection method to select offspring as new parents;
[0033] Crossover operation: randomly exchange data in chromosomes of different parents;
[0034] Mutation operation: randomly select data of different offspring chromosomes for adjustment;
[0035] Iteration and termination conditions: Repeat the selection, crossover and mutation operations, and iterate until the preset termination conditions are met, such as the fitness value convergence or the maximum number of iterations is reached, to obtain the optimal learning plan.
[0036] In a preferred embodiment, a power safety training management system includes:
[0037] The planning module is used to use linear programming to determine the training courses, training time, resource allocation, and coaching times for each trainee, and to generate training forms;
[0038] The training information collection module is used to train trainees according to the training form and regularly collect the training information of each trainee participating in the training;
[0039] A marking module is used to establish a learning curve model based on the training information, identify and mark the training courses where the target trainees have learning bottlenecks, and obtain the training course set that the target trainees need to optimize;
[0040] An identification module is used to establish a linear regression model for each training course that needs to be optimized based on the training information and the training course set that needs to be optimized, and to identify the degree of influence of the influencing factors of the target trainees on the trainees' performance;
[0041] The optimization module is used to optimize the individual learning plan of the target trainees based on the training course set that needs to be optimized for the target trainees using a genetic algorithm.
[0042] Technical effects and advantages of the present invention:
[0043] The present invention can formulate a personalized learning plan for each student based on the specific circumstances of each student, such as learning progress rate, learning bottlenecks and key influencing factors. This personalized training plan can better meet the actual needs of students and improve learning efficiency and effect. By monitoring the learning progress of students in real time, the present invention can dynamically adjust the training plan and promptly identify and solve the bottleneck problems encountered by students in the learning process. This can ensure that students always maintain a good learning state throughout the training process and avoid lagging or stagnating learning progress. The present invention uses linear programming and genetic algorithms to optimize the allocation of training resources and time management to avoid resource waste and work conflicts. The students' learning time, training resources and number of tutoring sessions are reasonably allocated, thereby reducing training costs while improving training effects.
[0044] Through iterative optimization of genetic algorithms, the present invention can find the globally optimal learning plan and resource allocation plan. Compared with the local optimization of traditional methods, the present invention can more effectively improve the overall training effect and ensure that the learning progress rate of trainees in each training course is maximized. By optimizing the management of power safety training, the present invention improves the safe operation skills and risk awareness of employees and reduces safety accidents caused by operational errors. This provides a strong guarantee for the stable operation of the power system and personal safety. The system design of the present invention has strong scalability and can be adjusted and expanded according to the needs of different enterprises. It is suitable for power enterprises of different sizes. Whether it is the basic training of new employees or the improvement of professional skills of senior employees, the present invention can provide effective training support. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] In order to facilitate the understanding of technical students in this field, the present invention is further described below with reference to the accompanying drawings;
[0046] Figure 1 The schematic diagram is a schematic diagram of a power safety training management method in the present invention.
[0047] Figure 2 The figure is a schematic diagram of an electric power safety training management system in the present invention. DETAILED DESCRIPTION
[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technical students in this field without making creative work are within the scope of protection of the present invention.
[0049] Reference Figure 1 - Figure 2 The following examples are obtained:
[0050] Embodiment 1: A method for managing power safety training comprises the following steps:
[0051] Use linear programming to determine the training courses, training time, resource allocation, and number of tutoring sessions for each trainee, ensuring that the training sequence is arranged reasonably to avoid resource waste and work conflicts; generate training forms; training courses refer to specific learning units set for trainees, which are designed to teach specific knowledge or skills. Determine the training courses that each trainee needs to attend and ensure that the course content matches the trainee's needs and skill level. Training time refers to the specific time arrangement of each course, including the start time, end time, and training duration. The training time is arranged reasonably to ensure that the trainee completes all courses within the specified time and avoid conflicts with other work arrangements. Resource allocation refers to the allocation of required resources for each course, such as training venues, equipment, teaching materials, and human resources (such as lecturers and tutors), ensuring that all resources required for training are in place to support efficient learning and teaching activities. The number of tutoring sessions refers to the number of additional interactions between trainees and lecturers or tutors during the training process, such as individual tutoring, answering questions, or practical guidance, etc., to provide additional learning support to help trainees overcome learning difficulties or understand complex content and ensure learning effectiveness. The training form is a detailed schedule that includes all the courses that trainees attend, the time schedule for each course, resource allocation, and the number of tutoring sessions, which means that the training form includes all the courses that trainees attend, the time schedule for each course, resource allocation, and the number of tutoring sessions.
[0052] Train trainees according to the training form, and regularly collect the training information of each trainee; establish a learning curve model based on the training information, identify and mark the training courses where the target trainees have learning bottlenecks, and obtain the training course set that needs to be optimized for the target trainees; establish a linear regression model for each training course that needs to be optimized based on the training information and the training course set that needs to be optimized, and identify the degree of influence of the influencing factors of the target trainees on the trainees' performance; based on the training course set that needs to be optimized for the target trainees, use a genetic algorithm to optimize the individual learning plan of the target trainees.
[0053] The logic of linear programming is:
[0054] Step 1: Determine the objective function: The goal is to maximize the learning effect of the students, and the objective function is expressed as: E ij represents the expected learning effect of student i in course j, X ij is a binary variable, indicating whether student i participates in course j (1 means participating, 0 means not participating), N represents the total number of trainees participating in the training, and M represents the total number of courses; Maximize learning effect: By optimizing the participation of each student in each course, the overall learning effect of the students is maximized. This means that with limited resources, students can get the most benefit from the training;
[0055] Step 2: Determine the constraints: satisfy the following constraint formula: λ j Indicates the constraint type. When its value is 1, it indicates a time constraint in the constraint formula, that is, the total learning time of a student on the course cannot exceed his / her available time. For example, the maximum available time of a student is 8 hours. If this time is exceeded, the student may not be able to complete other tasks or face the problem of insufficient energy. When its value is 2, it indicates a resource constraint in the constraint formula, that is, the total resource consumption of a course cannot exceed its resource upper limit. For example, if a course has only 20 seats, then at most 20 students can be arranged to participate in the course at the same time. When its value is 3, it indicates a tutoring frequency constraint in the constraint formula, that is, the tutoring frequency of a course cannot exceed its upper limit. For example, lecturers usually have limited time, so it is necessary to control the tutoring frequency of each course in order to better allocate the lecturer's time and energy. It means that the constraint formula is valid for every course j;
[0056] Step 3. Planning and solving: Use a linear programming solving tool (such as Python's PuLP or Sci Py) to solve the results that satisfy steps one and two. Use the solving tool to build a mathematical model of the objective function and constraints. Use a linear programming algorithm (such as the simplex method) to solve this mathematical model to obtain the optimal course schedule, time allocation, resource usage, and number of tutoring sessions for the students. The result is a training form that contains training courses, time, resource allocation, and number of tutoring sessions. This training form is used to guide subsequent training execution to ensure that all training activities are carried out smoothly under reasonable time and resource conditions, and that students can achieve the best learning results.
[0057] The factors influencing trainees’ participation in training include learning behavior factors and training course factors. Learning behavior factors include learning frequency, participation, homework completion, and number of exercises; training course factors include course difficulty, relevance of course content, and teaching methods.
[0058] Learning performance: Collect the learning performance of each student i in the training course j that needs to be optimized, such as test scores, homework scores, practical test results, etc.
[0059] Learning behavior factors: Learning frequency: the frequency of student i's training in course j, which may be measured by attendance rate or number of participations. Participation: student i's participation in course j, such as the number of questions asked, participation in class discussions, etc. Homework completion: student i's completion of homework in course j, such as homework completion rate or average score. Number of exercises: the number of exercises student i has in course j, such as the number of simulation operations or experiments.
[0060] Training course factors: Course difficulty: The difficulty of course j, generally evaluated by the course designer or experts, usually a quantitative score. Course content relevance: The relevance of course j to the trainees' current knowledge or work, which can also be evaluated by experts or obtained through questionnaires. Teaching method: The teaching method of course j, such as lectures, discussions, practical exercises, etc. It can be represented by a categorical variable (such as 0 = lectures, 1 = practical exercises, 2 = discussion).
[0061] Building a learning curve model based on training information means:
[0062] The students’ scores and task completion time in each learning stage are obtained from the training information. The changes of each student’s scores in each course over time are fitted to generate a learning curve. The learning curve is obtained by establishing a linear model. The formula of the linear model is: S i (t) = β0 + β1t + ∈; S i (t) is the score of student i at time t, β0 is the intercept, β1 is the learning progress rate (slope), and ∈ is the error term. The slope of the learning curve in the linear model is the learning progress rate, which represents the increase in the student's score per unit time in the course; the learning curve model can intuitively show the student's learning progress in a specific course by analyzing the student's score at each time point, which helps to understand the student's progress speed in the learning process and how the student performs at different learning stages. Calculate the learning progress rate: By fitting the learning curve through the linear regression model, the student's learning progress rate (slope β1) can be calculated. This slope represents the student's score growth per unit time in a specific course, which is an important indicator for evaluating the student's learning efficiency. Through the learning curve, it is possible to identify which students have difficulties in the learning process at an early stage. This early warning function allows training managers to intervene in time, provide additional support or adjust training strategies, thereby preventing students from falling behind expected goals.
[0063] By analyzing the learning curves of multiple trainees, an average benchmark for learning progress rate can be established. These benchmark values can be used to compare the performance of different trainees, thereby identifying which trainees are making slower progress in certain courses. Specifically, identify and mark the training courses where the target trainees have learning bottlenecks: obtain the learning progress rate of the target trainees in the target course, and calculate the average learning progress rate of all trainees in the target course. Set the learning bottleneck threshold to p times the average progress rate of all trainees, where p is greater than zero and less than one, for example, it can be 0.8 or 0.6. For each trainee i, if his learning progress rate in the target course is lower than the learning bottleneck threshold, the trainee is considered There is a learning bottleneck in the course and it is marked. All marked courses are summarized to obtain the training course set that needs to be optimized for the student i. The learning bottleneck refers to the situation that the student's learning progress rate in a certain course is significantly lower than the average level. Specifically, when the student's learning progress rate in a certain course is lower than the set learning bottleneck threshold (such as p times the average progress rate), the student is considered to have a learning bottleneck in the course. Subsequently, optimization measures for the learning bottleneck courses can be taken to significantly improve the students' learning effects in these courses and ensure that the students can successfully master the required knowledge and skills. By marking the learning bottleneck courses, training managers can give priority to optimizing these courses and the students' learning investment strategies.
[0064] According to the training information and the training course set to be optimized, a linear regression model is established for each of the training courses to be optimized:
[0065] Obtain the learning performance, corresponding data of learning behavior factors, and corresponding data of training course factors in the training information of each training course that needs to be optimized, and then establish a linear regression model:
[0066] S ij =β ij +β1F ij +β2P ij +β3A ij +β4L ij +β5D ij +β6R ij +β7T ij +∈ ij ; S ij represents the learning performance of student i in course j, β ij Represents the intercept term in the linear regression model established by student i for course j, ∈ ijrepresents the error term in the linear regression model established for student i on course j. β1, β2, β3, β4, β5, β6, and β7 are regression coefficients, which respectively represent the influence of learning frequency, participation, homework completion, number of exercises, course difficulty, course content relevance, and teaching method on learning performance. F ij represents the learning frequency of student i in course j, P ij represents the participation of student i in course j, A ij represents the completion of student i’s homework in course j, L ij represents the number of exercises that student i has done in course j, D ij represents the course difficulty of course j, R ij represents the relevance of the course content of course j, T ij Represents the teaching method of course j; use the least squares method to estimate the regression coefficients of the linear regression model, and ensure that the application performance of the linear regression model meets the requirements. The size of each regression coefficient reflects the influence of the corresponding parameter on the student's performance.
[0067] Using genetic algorithms to optimize the individual learning plans of trainees participating in the target training means:
[0068] Calculate the learning bottleneck percentage value BJ of each training course that needs to be optimized. The learning bottleneck percentage value BJ is obtained by calculating the difference between the learning bottleneck threshold of the target training course and the actual learning progress rate, and then dividing the difference by the percentage value of the learning bottleneck threshold of the target training course. Then the learning bottleneck percentage value BJ is calculated by the ratio of the preset proportional coefficient R and rounded up to get the k value. The learning bottleneck percentage value BJ is the numerator. According to the k value, the learning behavior factors corresponding to the first k regression coefficients of the target training course are obtained, and the vector composed of the training course factors is the genes in the initialized genetic algorithm population in the genetic algorithm; by calculating the k value, the factors that have the greatest impact on the learning progress rate of the trainees can be preferentially selected for optimization. The larger the regression coefficient, the more significant the impact of the factor on the learning performance. Therefore, selecting the factors corresponding to the first k regression coefficients can ensure that the optimization process focuses on the key factors that are most likely to improve the learning effect of the trainees. The learning bottleneck and progress rate of each trainee may be different. Therefore, the calculated k value and the corresponding key influencing factors will vary from person to person. In this way, the genetic algorithm can formulate personalized optimization strategies for each trainee instead of taking a one-size-fits-all approach. This personalized optimization helps to develop the most effective learning plan for each student's specific needs. In genetic algorithms, the more factors are considered, the higher the complexity and computational cost of the algorithm. By selecting the factors corresponding to the first k regression coefficients, the number of variables that need to be considered can be effectively reduced, thereby reducing the complexity of optimization and improving the efficiency of the algorithm. By focusing on the most important learning behavior factors and training course factors, genetic algorithms can find optimization paths more quickly and improve learning progress rates. Because these factors have the greatest impact on student performance, optimizing these factors often achieves significant results more quickly, thereby accelerating the convergence of the algorithm.
[0069] The learning bottleneck percentage value BJ reflects the severity of the learning bottleneck of students in a certain course. By combining the learning bottleneck percentage value BJ with the proportional coefficient R to calculate the k value, the genetic algorithm can dynamically adjust the optimization intensity. The more serious the bottleneck (that is, the larger the learning bottleneck percentage value BJ value), the more key influencing factors are selected (the larger the k value), and the stronger the optimization intensity. This dynamic adjustment mechanism ensures the reasonable allocation of resources and optimization intensity.
[0070] In the initial population of the genetic algorithm, each gene represents a combined factor in the student's learning plan. By selecting the factors corresponding to the first k regression coefficients, it is ensured that the initial population contains those high-value genes that are most likely to improve fitness. This selection not only ensures the diversity of the population, but also improves the overall fitness of the initial population, laying a good foundation for the final optimization result of the algorithm.
[0071] The learning plan of the target trainee is represented as a chromosome, each chromosome contains multiple genes, and each gene represents a combination of learning behavior and training course factors; the structure of the chromosome can be represented as: Chromosome = {G1, G2, ..., G n}; where each gene G j Represents a combination of key influencing factors of a training course, and selects the combination of the first k factors: G j =[f1, f2, ..., f k ]; f1, f2, ..., f k represents the top k influencing factors of training course j.
[0072] Fitness evaluation: The fitness function is used to evaluate the quality of each chromosome and measure the effectiveness of the optimized learning plan. Specifically, the fitness value is equal to the sum of the difference between the learning progress rate of all training courses after optimization and the learning progress rate before optimization (i.e., the comprehensive improvement value), minus the total amount of resources required for the optimization operation, and the difference between the learning progress rate of all training courses that need to be optimized after optimization and the learning progress rate before optimization is calculated and summed to obtain the comprehensive improvement value; the fitness function not only takes into account the improvement of the learning progress rate, but also balances the cost of resource input, ensuring that the final optimization result will not consume excessive resources while improving the learning effect, and realize the rational use of resources. Through fitness evaluation, it is possible to identify which learning plans are most likely to bring the greatest benefits, thereby guiding the selection and optimization process.
[0073] Selection operation: Use the roulette wheel selection method to select offspring as new parents; use the roulette wheel selection method to select individuals (chromosomes) from the current population as parents of the next generation. The roulette wheel selection method assigns selection probabilities based on the fitness values of the individuals. The higher the fitness value of the chromosome, the greater the probability of being selected as a parent. The roulette wheel selection method tends to select chromosomes with higher fitness, which means that learning plans with good optimization effects are more likely to be passed on to the next generation. In this way, the algorithm gradually approaches the optimal solution.
[0074] Crossover operation: randomly exchange data in chromosomes of different parents; crossover operation generates new chromosomes through gene recombination and explores new learning planning schemes, which can combine the excellent characteristics of different parents, thus potentially producing offspring that are better than the parents.
[0075] Mutation operation: randomly select data from different offspring chromosomes for adjustment; randomly select one or more genes in the chromosome and randomly adjust their values, such as changing the learning frequency or course difficulty. The mutation operation can fine-tune the individual's learning plan, thereby further improving the individual's fitness and providing more possibilities for subsequent optimization.
[0076] Iteration and termination conditions: Repeat the selection, crossover and mutation operations, and iterate until the preset termination conditions are reached, such as the fitness value convergence or the maximum number of iterations, to obtain the optimal learning plan. When the fitness value converges, it means that the best individual in the population is close to the global optimal solution, and further iterations are difficult to achieve significant improvement. By setting the maximum number of iterations, the algorithm is prevented from running too long and computing resources are saved. At the end of the algorithm, the chromosome with the highest fitness in the population is selected as the final optimal learning plan and applied to the target students. After multiple optimization operations, the final learning plan has fully considered the students' personalized needs and resource constraints, and can maximize the students' learning effect.
[0077] Embodiment 2: An electric power safety training management system, comprising:
[0078] The planning module is used to use linear programming to determine the training courses, training time, resource allocation, and coaching times for each trainee, and to generate training forms;
[0079] The training information collection module is used to train trainees according to the training form and regularly collect the training information of each trainee participating in the training;
[0080] A marking module is used to establish a learning curve model based on the training information, identify and mark the training courses where the target trainees have learning bottlenecks, and obtain the training course set that the target trainees need to optimize;
[0081] An identification module is used to establish a linear regression model for each training course that needs to be optimized based on the training information and the training course set that needs to be optimized, and to identify the degree of influence of the influencing factors of the target trainees on the trainees' performance;
[0082] The optimization module is used to optimize the individual learning plan of the target trainees based on the training course set that needs to be optimized for the target trainees using a genetic algorithm.
[0083] The above formulas are all dimensionless and numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technical students in this field according to actual conditions.
[0084] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0085] Students of ordinary skill in the art can appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical students can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0086] Technical students in the relevant field can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0087] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technical student familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A power safety training management method, characterized in that: The following steps are involved: Use linear programming to determine the training courses, training time, resource allocation, and coaching times for each trainee, and generate a training form; Conduct student training according to the training form and regularly collect training information of each trainee; Establish a learning curve model based on training information, identify and mark the training courses where the target trainees have learning bottlenecks, and obtain the training course set that the target trainees need to optimize; According to the training information and the training course set that needs to be optimized, a linear regression model is established for each training course that needs to be optimized to identify the degree of influence of the influencing factors of the target trainees on the trainees' performance; Based on the training course set that the target trainees need to optimize, the genetic algorithm is used to optimize the personal learning plan of the target trainees.
2. The power safety training management method according to claim 1 is characterized in that: The training form includes the courses each student attends, the time schedule for each course, resource allocation, and the number of coaching sessions.
3. The power safety training management method according to claim 2 is characterized in that: The logic of linear programming is: Step 1: Determine the objective function: The goal is to maximize the learning effect of the students, and the objective function is expressed as: E ij represents the expected learning effect of student i in course j, X ij is a binary variable, indicating whether trainee i participates in course j, N represents the total number of trainees participating in the training, and M represents the total number of courses; Step 2: Determine the constraints: satisfy the following constraint formula: λ j Indicates the constraint type. When its value is 1, it indicates a time constraint in the constraint formula; when its value is 2, it indicates a resource constraint in the constraint formula; when its value is 3, it indicates a tutoring times constraint in the constraint formula; It means that the constraint formula is valid for every course j; Step 3: Planning and solving: Use the linear programming solving tool to solve the results that meet steps 1 and 2. The result is a training form that includes training courses, time, resource allocation, and number of tutoring sessions.
4. The power safety training management method according to claim 3 is characterized in that: The factors influencing trainees’ participation in training include learning behavior factors and training course factors.
5. The power safety training management method according to claim 4 is characterized in that: Learning behavior factors include learning frequency, participation, homework completion, and number of exercises; training course factors include course difficulty, relevance of course content, and teaching methods.
6. The power safety training management method according to claim 5 is characterized in that: Building a learning curve model based on training information means: The students' scores in each learning stage and the time to complete the task are obtained from the training information. The changes in the scores of each student in each course over time are fitted to generate a learning curve. The learning curve is obtained by building a linear model. The slope of the learning curve in the linear model is the learning progress rate, which represents the increase in the students' scores per unit time in the course.
7. The power safety training management method according to claim 6, characterized in that: Identifying and marking training courses where target trainees have learning bottlenecks means: The learning progress rate of the target trainees in the target course is obtained, and the average learning progress rate of all trainees in the target course is calculated. The learning bottleneck threshold is set to p times the average progress rate of all trainees, where p is greater than zero and less than one. For each trainee i, if his learning progress rate in the target course is lower than the learning bottleneck threshold, it is considered that the student has a learning bottleneck in the course and is marked. All marked courses are summarized to obtain the set of training courses that need to be optimized for the trainee i.
8. The power safety training management method according to claim 7, characterized in that: According to the training information and the training course set to be optimized, a linear regression model is established for each of the training courses to be optimized: Obtain the learning performance, corresponding data of learning behavior factors, and corresponding data of training course factors in the training information of each training course that needs to be optimized, and then establish a linear regression model: S ij =b ij +β1F ij +β2P ij +β3A ij +β4L ij +β5D ij +β6R ij +β7T ij +∈ ij ; S ij represents the learning performance of student i in course j, β ij Represents the intercept term in the linear regression model established by student i for course j, ∈ ij represents the error term in the linear regression model established for student i on course j. β1, β2, β3, β4, β5, β6, and β7 are regression coefficients, which respectively represent the influence of learning frequency, participation, homework completion, number of exercises, course difficulty, course content relevance, and teaching method on learning performance. F ij represents the learning frequency of student i in course j, P ij represents the participation of student i in course j, A ij represents the completion of student i’s homework in course j, L ij represents the number of exercises that student i has done in course j, D ij represents the course difficulty of course j, R ij represents the relevance of the course content of course j, T ij represents the teaching method of course j; Use the least squares method to estimate the regression coefficients of the linear regression model and ensure that the application performance of the linear regression model meets the requirements.
9. The power safety training management method according to claim 8, characterized in that: Using genetic algorithms to optimize the individual learning plans of trainees participating in the target training means: Calculate the learning bottleneck percentage value BJ of each training course that needs to be optimized. The learning bottleneck percentage value BJ is obtained by calculating the difference between the learning bottleneck threshold of the target training course and the actual learning progress rate, and then dividing the difference by the percentage value of the learning bottleneck threshold of the target training course. Then, the learning bottleneck percentage value BJ is calculated by ratio with the preset proportional coefficient R and rounded up to obtain the k value. The learning bottleneck percentage value BJ is the numerator. According to the k value, the learning behavior factors corresponding to the top k regression coefficients of the target training course and the vector composed of the training course factors are obtained, that is, the genes in the initialized genetic algorithm population in the genetic algorithm. The learning plan of the target trainees is represented as chromosomes, each chromosome containing multiple genes; Fitness evaluation: Calculate the difference between the learning progress rate after optimization and the learning progress rate before optimization of all training courses that need to be optimized and sum them up to get the comprehensive improvement value. The fitness function is expressed as: the fitness is the comprehensive improvement value minus the total amount of resources required for the optimization operation; Selection operation: Use the roulette wheel selection method to select offspring as new parents; Crossover operation: randomly exchange data in chromosomes of different parents; Mutation operation: randomly select data of different offspring chromosomes for adjustment; Iteration and termination conditions: Repeat the selection, crossover and mutation operations, and iterate until the preset termination conditions are reached to obtain the optimal learning plan.
10. An electric power safety training management system, characterized in that: A power safety training management method according to any one of claims 1 to 9, comprising: The planning module is used to use linear programming to determine the training courses, training time, resource allocation, and coaching times for each trainee, and to generate training forms; The training information collection module is used to train trainees according to the training form and regularly collect the training information of each trainee participating in the training; A marking module is used to establish a learning curve model based on the training information, identify and mark the training courses where the target trainees have learning bottlenecks, and obtain the training course set that the target trainees need to optimize; An identification module is used to establish a linear regression model for each training course that needs to be optimized based on the training information and the training course set that needs to be optimized, and to identify the degree of influence of the influencing factors of the target trainees on the trainees' performance; The optimization module is used to optimize the individual learning plan of the target trainees based on the training course set that needs to be optimized for the target trainees using a genetic algorithm.