CN-122023083-A - Method and system for predicting capacity migration in industrial skill training

CN122023083ACN 122023083 ACN122023083 ACN 122023083ACN-122023083-A

Abstract

A method and a system for predicting capacity migration in industrial skill training relate to the field of professional skill assessment, and the method comprises the following steps: firstly, the constraint condition type of the virtual environment simplified relative to real equipment is obtained, and the operation performance difference value is recorded as constraint condition dependency strength by setting a comparison group and an experiment group virtual environment containing constraint and unconstrained. And then analyzing constraint conditions related to each link of the real machine task, calculating single capacity loss based on the dependence strength and the virtual-real difference multiple, and carrying out weighted summation by considering coupling coefficients among the constraint conditions to obtain a total capacity loss predicted value. And finally, converting the predicted value into real machine expected data, and comparing the real machine expected data with the actual data to generate a virtual-real capacity migration evaluation result of the target student. By implementing the method, the refinement degree of the migration evaluation of the deficiency-excess capacity can be improved.

Inventors

ZHAN FUXING
CHEN XIANGYANG
ZENG LE

Assignees

武汉厚溥数字科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260407

Claims (10)

1. A method for predicting capacity migration in industrial skill training, which is applied to a capacity migration prediction system, the method comprising: Obtaining constraint condition types of the virtual environment simplified relative to real equipment, wherein the constraint condition types comprise physical space constraint, information integrity constraint, operation feedback time delay constraint and error tolerance constraint; Setting a comparison virtual environment containing the constraint condition type and an experimental virtual environment for removing the constraint condition type for each constraint condition type, recording the operation performance index difference value of a target student under the comparison virtual environment and the experimental virtual environment, and taking the operation performance index difference value as the dependence intensity of the target student on the corresponding constraint condition type; extracting constraint condition types related to each operation link in a real machine task, determining a single-item capacity loss value based on the dependence intensity corresponding to each constraint condition type and the difference multiple of a real machine and a virtual environment, calculating coupling coefficients among different constraint condition types in the operation link, and carrying out weighted summation on all the single-item capacity loss values and the corresponding coupling coefficients to obtain a total capacity loss predicted value of the operation link; And converting the total energy loss predicted value into expected data of the real machine task, collecting actual data of the target student in a real machine environment, and generating a virtual-real capacity migration evaluation result value of the target student based on deviation data of the expected data and the actual data.
2. The method according to claim 1, wherein the step of extracting constraint condition types related to each operation link in the real machine task and determining a single capacity loss value based on the dependency strength corresponding to each constraint condition type and the difference multiple of the real machine and the virtual environment specifically comprises: acquiring an operation flow of the real machine task, and dividing the operation flow into a plurality of operation links; For each operation link, analyzing one or more constraint condition types in physical space constraint, information integrity constraint, operation feedback time delay constraint and error tolerance constraint related to the operation link; acquiring a reference value of each constraint condition type in the real machine environment and a corresponding actual value in the virtual environment; calculating the difference multiple of each constraint condition type according to the ratio of the reference standard value to the actual value; Calculating according to the dependence intensity and the difference multiple to obtain a preliminary loss value; and normalizing the preliminary loss value according to the standard completion time of the operation link to obtain the single capacity loss value.
3. The method according to claim 1, wherein the step of calculating coupling coefficients between different constraint condition types in the operation link, and performing weighted summation on all the single capacity loss values and the corresponding coupling coefficients to obtain the total capacity loss predicted value of the operation link specifically includes: Recording a first operation performance index of the target student in an experimental virtual environment in which two constraint condition types are removed simultaneously for each pair of constraint condition types in the operation link; comparing the first operation performance index with a second operation performance index when the single constraint condition type is removed respectively, and calculating a coupling coefficient between the two constraint condition types; constructing a coupling matrix containing all constraint condition types in the operation link; And carrying out weighted summation on all single energy loss values in the operation link based on the coupling coefficients in the coupling matrix to obtain the total energy loss predicted value.
4. The method according to claim 1, wherein the step of generating the virtual-actual ability migration evaluation result value of the target learner based on deviation data of the expected data and the actual data specifically includes: calculating the average value and standard deviation of the total energy loss predicted value of each operation link, and marking the operation links exceeding the average value preset multiple standard deviation as links to be analyzed; Calculating the deviation ratio of the expected data to the actual data for the links to be analyzed, and determining the operation links with the deviation ratio exceeding a preset threshold and the maximum number of the related constraint condition types as weak links; extracting a single capacity loss value corresponding to each constraint condition type in the capacity weak link, and carrying out normalization processing after sequencing the single capacity loss value according to the numerical value size to obtain the influence degree of the constraint condition type; and generating a virtual-actual capability migration evaluation result value of the target student based on the distribution position of the capability weak links and the influence degree of the constraint condition type.
5. The method according to claim 4, wherein the step of generating the virtual-real capability migration evaluation result value of the target learner based on the distribution position of the capability weak links and the influence degree of the constraint condition type specifically includes: Calculating the distribution density of the weak links in all operation links, and marking the continuously distributed weak links as important attention areas; According to the influence degree of the constraint condition types, calculating the accumulated influence value of each constraint condition type in the important attention interval; Generating an interval capacity loss coefficient based on a weighted combination of the distribution density of the important attention interval and the accumulated influence value; and carrying out weighted summation on the interval capacity loss coefficient and the total capacity loss predicted value to obtain the virtual-actual capacity migration evaluation result value.
6. The method according to claim 1, wherein after the step of generating the virtual-to-actual ability migration evaluation result value of the target learner based on deviation data of the expected data and the actual data, the method further comprises: constructing a virtual training scene library, and setting training scenes with a plurality of difficulty levels according to each constraint condition type; Based on the virtual-real capacity migration evaluation result value, matching a training scene corresponding to the difficulty level for the target student; Setting a progressive constraint condition adjustment mechanism in the training scene, wherein the constraint condition adjustment mechanism comprises gradually increasing physical space limitation, reducing operation prompt information, shortening operation feedback time and reducing error tolerance; And recording the adaptation process of the target trainee under different difficulty level scenes, and generating a personalized training advanced path.
7. The method according to claim 6, wherein the step of constructing a virtual training scene library, setting training scenes of a plurality of difficulty levels for each constraint type, specifically comprises: Dividing the training scene into a basic action training scene, a single skill training scene and a comprehensive skill training scene according to the operation property; setting physical space constraint and operation feedback time delay constraint in the basic action training scene; Setting information integrity constraint and error tolerance constraint in the single skill training scene; And setting all constraint condition types in the comprehensive skill training scene, and dynamically adjusting parameters of all constraint condition types according to the training performance of the target trainee.
8. A capability migration prediction system comprising one or more processors and memory coupled to the one or more processors, the memory for storing computer program code comprising computer instructions that the one or more processors invoke to cause the capability migration prediction system to perform the method of any of claims 1-7.
9. A computer readable storage medium comprising instructions which, when run on a capacity migration prediction system, cause the capacity migration prediction system to perform the method of any one of claims 1-7.
10. A computer program product, characterized in that the computer program product, when run on a capacity migration prediction system, causes the capacity migration prediction system to perform the method according to any of claims 1-7.

Description

Method and system for predicting capacity migration in industrial skill training Technical Field The application relates to the field of professional skill assessment, in particular to a method and a system for predicting capacity migration in industrial skill training. Background With the deep advancement of vocational education informatization, a virtual training system is widely applied in the technical training field. The virtual training environment simulates the operation scene of real equipment through a computer simulation technology, provides repeated training opportunities for students, and reduces training cost and safety risk. However, there is an objective difference between the virtual environment and the real device, and the learner performs well in the virtual environment and cannot fully guarantee that it is equally successful in the operation of a real machine. In the related art, the ability migration effect may be evaluated by comparing the operation performances of the trainee in the virtual environment and the real machine environment. The method is characterized in that after the virtual training is finished, real machine assessment is arranged, actual operation performance of a student on the real machine is recorded, and comparison analysis is carried out on the real machine results and the virtual training results, so that the effectiveness of the virtual training is judged. Some systems can be classified into macroscopic categories such as unskilled operation, poor environmental adaptability and the like as reference basis for subsequent training improvement in the typical error types of the statistics students in the real machine environment in the comparison process. However, with the increasing demand for fine management of training, the above-described scheme exposes significant limitations in practical applications. The evaluation conclusion obtained through the whole score comparison is rough, true machine errors of students are attributed to macroscopic categories such as unskilled operation or poor environmental adaptability, the capability of which degree of the students can be reduced in which operation links cannot be specified, and the reasons of the differences of the true machine performances of different students under the same virtual training score cannot be explained. This results in insufficient pertinence of the assessment results to subsequent training guidelines, reducing the practical value of the capacity migration assessment. Disclosure of Invention The application provides a capability migration prediction method and a capability migration prediction system in industrial skill training, which are used for improving the refinement degree of virtual capability migration evaluation. The method comprises the steps of obtaining constraint condition types of a virtual environment simplified relative to real equipment, wherein the constraint condition types comprise physical space constraint, information integrity constraint, operation feedback time delay constraint and error fault tolerance constraint, setting a comparison virtual environment containing constraint condition types and an experiment virtual environment removing constraint condition types for each constraint condition type, recording operation performance index difference values of a target student under the comparison virtual environment and the experiment virtual environment, taking the operation performance index difference values as the dependence intensity of the target student on the corresponding constraint condition types, extracting constraint condition types related to each operation link in a real machine task, determining single-item capacity loss values based on the dependence intensity corresponding to each constraint condition type and the difference multiple of the real machine and the virtual environment, calculating coupling coefficients among different constraint condition types in the operation links, carrying out weighted summation on all single-item capacity loss values and the corresponding coupling coefficients to obtain total capacity loss prediction values of the operation links, converting the total capacity loss prediction values into expected data of the real machine task, collecting the expected data of the target student in the environment, and generating actual data of the target student on the actual data of the target student, and evaluating the actual data of the target student in the real machine based on the expected data. In the embodiment, four types of constraint condition types simplified by the virtual environment are obtained, the dependency strength of a contrast environment and an experimental environment on each constraint condition is set for measuring a student, the constraint condition types related to each operation link of a real machine task are extracted, a coupling coefficient is calculated, a total energy loss predicted value is determined bas