An intelligent evaluation method and system for emergency safety training effects

Abstract

The present invention discloses an intelligent evaluation method and system for emergency safety training effectiveness, which relates to the technical field of emergency safety training effectiveness evaluation and aims to solve the technical problem of inaccurate and incomplete evaluation results in the prior art. The method comprises the following steps: S1, basic data preparation: standardizing and normalizing the behavior sequences in the emergency safety rescue plan to obtain standard action data, dividing the training set and the validation set to determine the standard values ​​of the standard actions, and marking the standard sample library; S2, evaluation model construction: setting the emergency safety training effectiveness evaluation model, dividing the sample subsets, using the stochastic gradient descent method to train the model and determine the parameters; S3, data processing and evaluation calculation: obtaining training and exercise behavior data, constructing a standard sample library and a model training library, training and revising the model, and calculating the safety training standardization score, the response score, and the efficiency score. The present invention has the advantages of improving the accuracy and comprehensiveness of emergency safety training effectiveness evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of emergency safety training effect evaluation, and more particularly to an emergency safety training effect intelligent evaluation method and system. Background Art

[0002] In the field of emergency safety training, with the development of digital technology, some institutions have begun to experiment with using data to evaluate training effectiveness. Currently, common evaluation methods rely primarily on manual observation and simple data recording. Manual observation often involves experienced instructors scoring trainees' performance based on their subjective judgment. This method is significantly influenced by subjective factors, and evaluation criteria vary between instructors, resulting in a lack of accuracy and consistency in evaluation results. Simple data recording, such as basic information like training completion time and number of operations, fails to fully and thoroughly reflect trainees' performance in key areas such as standardization of movements and adaptability.

[0003] Among data-based evaluation technologies, existing methods have many flaws when processing emergency safety training data. On the one hand, data processing is not sophisticated enough and fails to fully tap the potential value of behavioral sequences in emergency safety and rescue plans. In the process of converting plans into evaluation data, behavioral actions are not standardized and normalized, resulting in a lack of comparability between different action data and an inability to accurately construct a standard sample library for evaluation, which seriously affects the training effect and evaluation accuracy of the evaluation model. On the other hand, existing evaluation models are often simple in structure and cannot effectively extract key features from complex training data. For example, many models cannot simultaneously take into account multi-dimensional evaluation indicators such as action standardization, adaptability, and training efficiency, making it difficult to fully reflect the actual effect of emergency safety training. As a result, trainers and organizers are unable to accurately locate problems in training, which in turn affects the quality improvement of emergency safety training and the achievement of training goals.

[0004] In view of this, we propose an intelligent evaluation method and system for emergency safety training effectiveness. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent evaluation method and system for emergency safety training effects, so as to solve the technical problem of inaccurate and incomplete evaluation results in the prior art.

[0006] To solve the above technical problems, the present invention provides the following technical solution: a method for intelligently evaluating the effectiveness of emergency safety training, comprising the following steps:

[0007] S1. Basic data preparation: standardize and normalize the behavior sequences in the emergency safety and rescue plan to obtain standard action data, divide the training set and validation set to determine the standard action values, and mark the standard sample library;

[0008] S2. Evaluation model construction: setting up the emergency safety training effect evaluation model, dividing the sample subsets, using the stochastic gradient descent method to train the model and determine the parameters;

[0009] S3: Data processing and evaluation calculation, obtaining training and exercise behavior data, building a standard sample library and model training library, training and correcting the model, and calculating the safety training standardization score , strain score , efficiency score , substitute into the evaluation model to calculate the final evaluation score .

[0010] Preferably, the method for standardizing and normalizing the behavior sequence in S1 is:

[0011] S1.1, analyze the collected emergency safety and rescue plans in detail, identify each behavior action, construct a behavior sequence, clarify the key features of each behavior action and quantify them, and standardize the behavior sequence by removing the dimensional differences between different action data. For the standardized emergency safety standard action data, use the minimum-maximum normalization method to map the data value range to the interval , making data of different actions comparable;

[0012] S1.2. Divide the standard action data into a training set and a validation set. Calculate the mean and variance of the action data for the training set and validation set, respectively. The mean reflects the average level of the data, and the variance reflects the degree of data dispersion. Combine the mean and variance of the two and use the weighted average method to determine the standard value of the emergency safety standard action. ;

[0013] S1.3, randomly group the emergency safety standard action data into multiple sub-datasets, further divide each sub-dataset into a training set and a validation set, calculate the mean deviation and variance deviation between the training set and the validation set, and when the mean deviation and variance deviation are both less than the preset threshold, mark it as the emergency safety standard action standard sample library.

[0014] Preferably, the setting method of the evaluation model in S2 is:

[0015] S2.1. The evaluation model is designed as a multi-layer neural network structure. The input layer receives the processed emergency safety training data, the hidden layer extracts data features through nonlinear transformation, and the output layer outputs the corresponding results according to the evaluation task. The hidden layer uses the ReLU activation function, and the output layer uses the Softmax function to convert the output into a probability distribution to facilitate classification decision-making;

[0016] S2.2: Detailed division of the labeled emergency safety action standard sample library into multiple sample subsets containing action samples and their corresponding labels. Each action sample represents a specific action in emergency safety training, and the label clearly indicates whether the action is correct or belongs to a certain category.

[0017] S2.3, use the stochastic gradient descent method to train the set emergency safety training evaluation model, use the cross entropy loss function to measure the difference between the model prediction results and the true label, the model minimizes the loss function by continuously adjusting the parameters, and the parameter update step is controlled by the learning rate. At the same time, the trained model is verified with the verification samples in the sample subset, and the model accuracy is calculated. When the model accuracy reaches the pre-set threshold, it indicates that the model has learned the characteristics and rules of emergency safety standard actions well, determines the model parameters and outputs the final evaluation model.

[0018] Preferably, in S3, the method of obtaining training and exercise behavior data, building a standard sample library and a model training library, and training and correcting the model is:

[0019] S3.1. During emergency safety training drills, use data collection equipment to obtain comprehensive, real-time behavioral data on trainees. This data covers the trainees' posture, timing, and force, accurately reflecting their actual performance during the drill.

[0020] S3.2. Convert the emergency safety and rescue plan into a behavioral sequence diagram. Based on the plan's action flow and logical relationships, divide it into several task nodes containing unique behavioral actions. Each task node represents a specific action or operation step in the emergency safety training. The emergency safety standard action value is set. An action threshold is set for each task node to determine whether the trainee's actions meet the standard requirements. Based on the task nodes and action thresholds, a standard sample library and a model training library are constructed. The standard sample library provides correct reference examples, and the model training library stores actual training data.

[0021] S3.3, initialize the emergency safety training evaluation model, input the data in the standard sample library into the model for training, and use batch normalization technology to accelerate the model convergence speed and improve the training stability. According to the difference between the model output results and the standard sample library labels, the model parameters are continuously adjusted, and the training and correction process is repeated.

[0022] Preferably, the safety training standardization score The calculation method is:

[0023] The emergency safety training data to be evaluated is divided into different action units or time periods. The trained model is used to analyze the training data in each action unit or time period in detail, and the corresponding standard action value and action execution value are calculated. A weight function is introduced to measure the degree of difference between the two. The weight function values ​​of each action unit or time period are summed to obtain the safety training standardization score.

[0024] After dividing the data to be evaluated, calculate the standard action value and , weight function , ;

[0025] in, is a constant, is the fault tolerance coefficient, is the number of divided action units or time periods, It is The weight function value of an action unit or time period.

[0026] Preferably, the safety training strain score The calculation method is:

[0027] The collected training data is thoroughly analyzed and evaluated from three aspects: behavioral accuracy, behavioral effectiveness, and behavioral matching. Behavioral accuracy is measured by calculating the cosine similarity between the trainee's movements and the standard movements. Behavioral effectiveness is evaluated based on whether the movements achieve the expected results. Behavioral matching is measured by calculating the edit distance between the trainee's behavioral sequence and the emergency safety rescue plan's behavioral sequence. The scores of these three aspects are weighted and summed to obtain the safety training response score.

[0028] Calculating behavioral accuracy , behavioral matching , according to the formula Calculate the score, where is the trainer's motion vector, is the standard action vector, the effectiveness of the behavior Evaluate whether the action achieves the expected effect. is the edit distance, is the length of the behavior sequence, 、 、 is the weight parameter, and .

[0029] Preferably, the safety training efficiency score The calculation method is:

[0030] Based on the collected training data, the training efficiency is evaluated from two aspects: time and operation. The time score is calculated based on the ratio of the time required for the trainee to complete the task to the standard time. The operation score is calculated based on the ratio of the number of operation errors of the trainee to the total number of operations. The time score and operation score are weighted and summed to obtain the safety training efficiency score.

[0031] Calculate time score , operation score , according to the formula Calculate the score, where is the time required for trained personnel to complete the task, It's standard time. is the number of operational errors, is the total number of operations, and is the weight parameter, and .

[0032] Preferably, the emergency safety training assessment score The calculation method is:

[0033] Substitute the calculated safety training standardization score, safety training response score, and safety training efficiency score into the evaluation model, and sum them up according to the weights to obtain the emergency safety training evaluation score;

[0034] Will 、 、 Substitute into the evaluation model Calculate the final score, where 、 、 is the weight parameter, and .

[0035] An intelligent evaluation system for emergency safety training effects, comprising:

[0036] Data collection module: collects behavioral data on trainees’ movements, posture, timing, and strength during training exercises;

[0037] Data processing and sample construction module: Process emergency safety and rescue plan data, determine standard action values, and build a standard sample library and model training library;

[0038] Model building and training module: set up a multi-layer neural network evaluation model, divide the sample subsets to train the model, and determine and output the model parameters;

[0039] Evaluation and calculation module: train and modify the model, calculate the scores of standardization, adaptability, and efficiency, and substitute them into the model to obtain the overall evaluation score.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention standardizes and normalizes the behavioral sequences in emergency safety and rescue plans, constructs a scientific and reasonable standard sample library, and designs an evaluation model with a multi-layer neural network structure, which can accurately extract data features. This improvement effectively solves the problem of inaccurate and incomplete evaluation results in the existing technology, greatly improves the accuracy and comprehensiveness of emergency safety training effect evaluation, makes the evaluation results more valuable for reference, and helps trainers and organizers understand the training situation more accurately.

[0042] 2. Based on accurate evaluation results, the present invention enables trainees to clearly recognize their own strengths and weaknesses in terms of movement standardization, adaptability, and training efficiency. For example, through standardization scores, trainees can improve non-standard movements in a targeted manner; based on adaptability scores, emergency response training can be strengthened, which enables trainees to conduct more targeted training, improve training effects, and further enhance the quality of emergency safety training, solving the problem in existing technologies that cannot provide trainees with accurate training guidance.

[0043] 3. The comprehensive and accurate evaluation results of the present invention provide a strong basis for optimizing training programs. Organizers can adjust the training content, duration and methods according to the evaluation data of different trainees and reasonably allocate training resources. For example, the training duration can be increased for actions with generally poor standardization, and special training can be designed for links with weak adaptability. This helps to improve the utilization efficiency of training resources, solves the problems of lack of pertinence and waste of training resources in the existing technology, and makes emergency safety training more scientific and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Schematic diagram of the method flow of the present invention;

[0045] Figure 2 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0046] To facilitate those skilled in the art to understand the technical solution of the present invention, the technical solution of the present invention is further described with reference to the accompanying drawings.

[0047] Example 1, as Figure 1 As shown, the present invention provides an intelligent evaluation method for emergency safety training effects, comprising the following steps:

[0048] S1. Basic data preparation: standardize and normalize the behavior sequences in the emergency safety and rescue plan to obtain standard action data, divide the training set and validation set to determine the standard action values, and mark the standard sample library;

[0049] S2. Evaluation model construction: setting up the emergency safety training effect evaluation model, dividing the sample subsets, using the stochastic gradient descent method to train the model and determine the parameters;

[0050] S3: Data processing and evaluation calculation, obtaining training and exercise behavior data, building a standard sample library and model training library, training and correcting the model, and calculating the safety training standardization score , strain score , efficiency score , substitute into the evaluation model to calculate the final evaluation score .

[0051] In an embodiment of the present invention, the method for standardizing and normalizing the behavior sequence in S1 is:

[0052] S1.1. Analyze the collected emergency safety and rescue plans in detail, identify each action, and construct a behavior sequence. For each action, clarify its key characteristics (such as the start and end conditions, amplitude, speed, etc.) and quantify them. By removing the dimensional differences between different action data, standardize the behavior sequence. For the standardized emergency safety standard action data, use the minimum-maximum normalization method to map the data value range to the interval , making different action data comparable and avoiding the impact of data scale differences on subsequent model training;

[0053] Assume that the behavior sequence in the emergency safety rescue plan is , the standardized emergency safety standard action data is , using the minimum-maximum normalization formula , the value range is Standard action data ;

[0054] in, Indicates the first A behavioral action, Represents the standardized Emergency safety standard action data, Representing a collection The minimum value in Representing a collection The maximum value in ;

[0055] S1.2. Divide the standard action data into a training set and a validation set according to a pre-set ratio to facilitate effective verification and adjustment during model training and improve the model's generalization ability. Calculate the mean and variance of the action data for the training set and validation set, respectively. The mean reflects the average level of the data, while the variance reflects the degree of data dispersion. Combine the mean and variance of the two, and use a weighted average method to determine the standard value of the emergency safety standard action. This will serve as an important reference for subsequent evaluation of training effectiveness.

[0056] Standard action data according to Divide into training set and validation set ,in, is the pre-set division ratio, the value range is ;

[0057] Training set mean , training set variance ;

[0058] Validation set mean , validation set variance ;

[0059] By weighted average formula Determine the standard value of emergency safety standard action ;

[0060] in, is the number of data in the training set, The first data, is the number of data in the validation set, The validation set data, is the weight coefficient, and its value range is ;

[0061] S1.3. Randomly group the emergency safety action data into multiple sub-datasets. Each sub-dataset is independent and representative enough to fully cover different types of emergency safety action. Each sub-dataset is further divided into a training set and a validation set. The mean deviation and variance deviation between the training set and the validation set are calculated. When both the mean deviation and variance deviation are less than a preset threshold, it indicates that the training set and validation set of the sub-dataset have good consistency and stability. The sub-dataset is marked as a standard sample library for emergency safety action and used for subsequent model training.

[0062] Randomly group the emergency safety standard action data into Sub-datasets , each sub-dataset Proportional Divide into training set and validation set ;

[0063] Calculating mean deviation , variance bias ;

[0064] when and , mark the sub-dataset as the emergency safety specification action standard sample library;

[0065] in, is the number of sub-datasets, Is a sub-dataset The mean of the training set, Is a sub-dataset The mean of the validation set, Is a sub-dataset The variance of the training set, Is a sub-dataset The variance of the validation set, and is the set threshold.

[0066] In an embodiment of the present invention, the setting method of the evaluation model in S2 is:

[0067] S2.1. Thoroughly study the specific requirements and objectives of emergency safety and rescue plans, comprehensively consider the standardization of emergency safety training, adaptability, and training efficiency, and establish an emergency safety training effectiveness evaluation model. Design the evaluation model as a multi-layer neural network structure. The input layer receives processed emergency safety training data, the hidden layer extracts data features through nonlinear transformation, and the output layer outputs the corresponding results based on the evaluation task. The hidden layer uses the ReLU activation function to alleviate the gradient vanishing problem and speed up model training. The output layer uses the Softmax function to convert the output into a probability distribution to facilitate classification and decision-making.

[0068] Evaluating the Model It is a multi-layer neural network structure, and the input layer has neurons, and the hidden layer has neurons, and the output layer has neurons, ReLU activation function , Softmax function ;

[0069] in, The output layer The output probability of a neuron is The output layer The input value of a neuron, is a natural constant;

[0070] S2.2: Detailed division of the labeled emergency safety action standard sample library into multiple sample subsets containing action samples and their corresponding labels. Each action sample represents a specific action in emergency safety training. The label clearly indicates whether the action is correct or belongs to a specific category. This helps improve the efficiency and accuracy of model training and enables the model to better learn the characteristics of different types of actions.

[0071] Divide the labeled emergency safety standard action sample library into multiple sample subsets containing action samples and their corresponding labels , each action sample is , the corresponding label is ;

[0072] in, is the number of sample subsets, Indicates the number of the sample subset, Indicates the number of the action sample in the sample subset;

[0073] S2.3. Use stochastic gradient descent to train the pre-set emergency safety training and evaluation model. Use the cross-entropy loss function to measure the difference between the model's prediction results and the true label. The model minimizes the loss function by continuously adjusting parameters. The parameter update step is controlled by the learning rate. At the same time, the trained model is verified using validation samples from the sample subset. The model accuracy is calculated. When the model accuracy reaches a pre-set threshold, it indicates that the model has well learned the characteristics and patterns of emergency safety standard actions. The model parameters are determined and the final evaluation model is output.

[0074] Cross Entropy Loss Function , the parameter update formula is , calculate the accuracy of the model ,when The value is 1 when , otherwise it is 0. Reaching the threshold Determine the model parameters and output the model;

[0075] in, is the number of training samples, are the parameters of the model, is the learning rate, is the loss function with respect to the parameters The gradient, is the number of validation samples, The model is The predicted labels of the validation samples, The model is The true labels of the validation samples, is the indicator function.

[0076] In an embodiment of the present invention, in S3, the method of obtaining training and exercise behavior data, building a standard sample library and a model training library, and training and correcting the model is as follows:

[0077] S3.1. During emergency safety training drills, utilize a variety of advanced data acquisition equipment, including high-precision sensors, high-definition cameras, and professional recorders, to obtain comprehensive, real-time behavioral data on trainees. This data covers the trainees' posture, timing, and force, accurately reflecting their actual performance during the drill.

[0078] Assume that the collected behavioral data information is ,in, Indicates the collected Personal behavioral data information;

[0079] S3.2. Convert the emergency safety and rescue plan into a behavioral sequence diagram. Based on the plan's action flow and logical relationships, divide it into several task nodes containing unique behavioral actions. Each task node represents a specific action or operation step in the emergency safety training, facilitating detailed analysis of the training process. Emergency safety action standard values ​​are specified. Action thresholds are set for each task node to determine whether the trainee's actions meet the standard requirements. Based on the task nodes and action thresholds, a standard sample library and a model training library are constructed. The standard sample library provides correct reference examples, and the model training library stores actual training data to support model training and optimization.

[0080] Convert the emergency safety rescue plan into a behavior sequence diagram and divide it into several task nodes ,in, Indicates the Task nodes, combined with standard action values For each task node Setting the action threshold , based on which a standard sample library is constructed and model training library ;

[0081] S3.3. Initialize the emergency safety training evaluation model and input data from the standard sample library into the model for training. To accelerate model convergence and improve training stability, use batch normalization technology. Based on the difference between the model output and the standard sample library labels, continuously adjust the model parameters. Repeat the training and correction process until the model accuracy meets the pre-set requirements. At this point, the model can better adapt to the actual situation of emergency safety training and provide accurate prediction results for subsequent evaluation calculations.

[0082] Initialize the model and use batch normalization technology for input data , the formula is , for output data , the formula is ,in, is the normalized data, is the mean of the current batch of data, is the variance of the current batch of data, is a constant, and To learn the parameters.

[0083] In an embodiment of the present invention, the safety training standardization score The calculation method is:

[0084] The emergency safety training data to be evaluated is divided into different action units or time periods. The trained model is used to analyze the training data in each action unit or time period in detail, and the corresponding standard action value and action execution value are calculated. A weight function is introduced to measure the degree of difference between the two. The weight function values ​​of each action unit or time period are summed to obtain the safety training standardization score. This score reflects the degree of standardization of the trainees' actions. The lower the score, the more consistent the action is with the standard requirements.

[0085] After dividing the data to be evaluated, calculate the standard action value and , weight function , ;

[0086] in, is a constant, is the fault tolerance coefficient, is the number of divided action units or time periods, It is The weight function value of an action unit or time period.

[0087] In an embodiment of the present invention, the safety training response score The calculation method is:

[0088] Through in-depth analysis of the collected training data, we evaluate it from three perspectives: behavioral accuracy, behavioral effectiveness, and behavioral match. Behavioral accuracy is measured by calculating the cosine similarity between the trainee's movements and standard movements. Behavioral effectiveness is assessed based on whether the movements achieve the expected results. Behavioral match is measured by calculating the edit distance between the trainee's behavioral sequence and the behavioral sequence of the emergency safety and rescue plan. The weighted sum of these three scores yields a safety training response score, which reflects the trainee's ability to respond and handle problems in emergency situations.

[0089] Calculating behavioral accuracy , behavioral matching , according to the formula Calculate the score, where is the trainer's motion vector, is the standard action vector, the effectiveness of the behavior Evaluate whether the action achieves the expected effect. is the edit distance, is the length of the behavior sequence, 、 、 is the weight parameter, and .

[0090] In an embodiment of the present invention, the safety training efficiency score The calculation method is:

[0091] Based on the collected training data, the training efficiency is evaluated from two aspects: time and operation. The time score is calculated based on the ratio of the time required for the trainee to complete the task to the standard time. The operation score is calculated based on the ratio of the number of operational errors of the trainee to the total number of operations. The time score and the operation score are weighted and summed to obtain the safety training efficiency score, which reflects the efficiency level of emergency safety training.

[0092] Calculate time score , operation score , according to the formula Calculate the score, where is the time required for trained personnel to complete the task, It's standard time. is the number of operational errors, is the total number of operations, and is the weight parameter, and .

[0093] In an embodiment of the present invention, the emergency safety training evaluation score The calculation method is:

[0094] Substitute the calculated safety training standardization score, safety training response score, and safety training efficiency score into the evaluation model and sum them up according to the weights to obtain the emergency safety training evaluation score. This score comprehensively reflects the overall effect of emergency safety training and provides a quantitative indicator for evaluating training quality.

[0095] Will 、 、 Substitute into the evaluation model Calculate the final score, where 、 、 is the weight parameter, and .

[0096] Example 2, as Figure 2 As shown, an intelligent evaluation system for emergency safety training effects includes:

[0097] Data collection module: collects behavioral data on trainees’ movements, posture, timing, and strength during training exercises;

[0098] Data processing and sample construction module: Process emergency safety and rescue plan data, determine standard action values, and build a standard sample library and model training library;

[0099] Model building and training module: set up a multi-layer neural network evaluation model, divide the sample subsets to train the model, and determine and output the model parameters;

[0100] Evaluation and calculation module: train and modify the model, calculate the scores of standardization, adaptability, and efficiency, and substitute them into the model to obtain the overall evaluation score.

[0101] The embodiments disclosed in the present invention are preferred embodiments, but are not limited to them. Ordinary technicians in this field can easily understand the spirit of the present invention based on the above embodiments and make different extensions and changes. As long as they do not deviate from the spirit of the present invention, they are all within the scope of protection of the present invention.

Claims

1. An intelligent evaluation method for emergency safety training effects, characterized in that: The following steps are involved: S1. Basic data preparation: standardize and normalize the behavior sequences in the emergency safety and rescue plan to obtain standard action data, divide the training set and validation set to determine the standard action values, and mark the standard sample library; S2. Evaluation model construction: setting up the emergency safety training effect evaluation model, dividing the sample subsets, using the stochastic gradient descent method to train the model and determine the parameters; S3: Data processing and evaluation calculation, obtaining training and exercise behavior data, building a standard sample library and model training library, training and correcting the model, and calculating the safety training standardization score , strain score , efficiency score , substitute into the evaluation model to calculate the final evaluation score ; The setting method of the evaluation model in S2 is: S2.

1. The evaluation model is designed as a multi-layer neural network structure. The input layer receives the processed emergency safety training data, the hidden layer extracts data features through nonlinear transformation, and the output layer outputs the corresponding results according to the evaluation task. The hidden layer uses the ReLU activation function, and the output layer uses the Softmax function to convert the output into a probability distribution to facilitate classification decision-making; S2.2: Detailed division of the labeled emergency safety action standard sample library into multiple sample subsets containing action samples and their corresponding labels. Each action sample represents a specific action in emergency safety training, and the label clearly indicates whether the action is correct or belongs to a certain category. S2.

3. Use stochastic gradient descent to train the pre-set emergency safety training and evaluation model. Use the cross-entropy loss function to measure the difference between the model's prediction results and the true label. The model minimizes the loss function by continuously adjusting parameters. The parameter update step is controlled by the learning rate. At the same time, the trained model is verified using validation samples from the sample subset. The model accuracy is calculated. When the model accuracy reaches a pre-set threshold, it indicates that the model has well learned the characteristics and patterns of emergency safety standard actions. The model parameters are determined and the final evaluation model is output. In S3, the method of obtaining training and exercise behavior data, building a standard sample library and a model training library, and training and correcting the model is as follows: S3.

1. During emergency safety training drills, use data collection equipment to obtain comprehensive, real-time behavioral data on trainees. This data covers the trainees' posture, timing, and force, accurately reflecting their actual performance during the drill. S3.

2. Convert the emergency safety and rescue plan into a behavioral sequence diagram. Based on the plan's action flow and logical relationships, divide it into several task nodes containing unique behavioral actions. Each task node represents a specific action or operation step in the emergency safety training. The emergency safety standard action value is set. An action threshold is set for each task node to determine whether the trainee's actions meet the standard requirements. Based on the task nodes and action thresholds, a standard sample library and a model training library are constructed. The standard sample library provides correct reference examples, and the model training library stores actual training data. S3.3, initialize the emergency safety training and evaluation model, input the data from the standard sample library into the model for training, and use batch normalization technology to accelerate model convergence and improve training stability. Based on the difference between the model output results and the standard sample library labels, continuously adjust the model parameters and repeat the training and correction process; The safety training normative score The calculation method is: The emergency safety training data to be evaluated is divided into different action units or time periods. The trained model is used to analyze the training data in each action unit or time period in detail, and the corresponding standard action value and action execution value are calculated. A weight function is introduced to measure the degree of difference between the two. The weight function values ​​of each action unit or time period are summed to obtain the safety training standardization score. After dividing the data to be evaluated, calculate the standard action value and , weight function , ; in, is a constant, is the fault tolerance coefficient, is the number of divided action units or time periods, It is The weight function value of an action unit or time period; The Safety Training Strain Score The calculation method is: The collected training data is thoroughly analyzed and evaluated from three aspects: behavioral accuracy, behavioral effectiveness, and behavioral matching. Behavioral accuracy is measured by calculating the cosine similarity between the trainee's movements and the standard movements. Behavioral effectiveness is evaluated based on whether the movements achieve the expected results. Behavioral matching is measured by calculating the edit distance between the trainee's behavioral sequence and the emergency safety rescue plan's behavioral sequence. The scores of these three aspects are weighted and summed to obtain the safety training response score. Calculating behavioral accuracy , behavioral matching , according to the formula Calculate the score, where is the trainer's motion vector, is the standard action vector, the effectiveness of the behavior Evaluate whether the action achieves the expected effect. is the edit distance, is the length of the behavior sequence, 、 、 is the weight parameter, and ; The safety training efficiency score The calculation method is: Based on the collected training data, the training efficiency is evaluated from two aspects: time and operation. The time score is calculated based on the ratio of the time required for the trainee to complete the task to the standard time. The operation score is calculated based on the ratio of the number of operation errors of the trainee to the total number of operations. The time score and operation score are weighted and summed to obtain the safety training efficiency score. Calculate time score , operation score , according to the formula Calculate the score, where is the time required for trained personnel to complete the task, It's standard time. is the number of operational errors, is the total number of operations, and is the weight parameter, and .

2. The intelligent evaluation method for emergency safety training effect according to claim 1, characterized in that: The method for standardizing and normalizing the behavior sequence in S1 is: S1.1, analyze the collected emergency safety and rescue plans in detail, identify each behavior action, construct a behavior sequence, clarify the key features of each behavior action and quantify them, and standardize the behavior sequence by removing the dimensional differences between different action data. For the standardized emergency safety standard action data, use the minimum-maximum normalization method to map the data value range to the interval , making data of different actions comparable; S1.

2. Divide the standard action data into a training set and a validation set. Calculate the mean and variance of the action data for the training set and validation set, respectively. The mean reflects the average level of the data, and the variance reflects the degree of data dispersion. Combine the mean and variance of the two and use the weighted average method to determine the standard value of the emergency safety standard action. ; S1.3, randomly group the emergency safety standard action data into multiple sub-datasets, further divide each sub-dataset into a training set and a validation set, calculate the mean deviation and variance deviation between the training set and the validation set, and when the mean deviation and variance deviation are both less than the preset threshold, mark it as the emergency safety standard action standard sample library.

3. The intelligent evaluation method for emergency safety training effect according to claim 2, characterized in that: The emergency safety training evaluation score The calculation method is: Substitute the calculated safety training standardization score, safety training response score, and safety training efficiency score into the evaluation model, and sum them up according to the weights to obtain the emergency safety training evaluation score; Will 、 、 Substitute into the evaluation model Calculate the final score, where 、 、 is the weight parameter, and .

4. A system for the intelligent evaluation method of emergency safety training effect as claimed in claim 3, characterized in that: include: Data collection module: collects behavioral data on trainees’ movements, posture, timing, and strength during training exercises; Data processing and sample construction module: Process emergency safety and rescue plan data, determine standard action values, and build a standard sample library and model training library; Model building and training module: set up a multi-layer neural network evaluation model, divide the sample subsets to train the model, and determine and output the model parameters; Evaluation and calculation module: train and modify the model, calculate the scores of standardization, adaptability, and efficiency, and substitute them into the model to obtain the overall evaluation score.

Images