Task performance determination method and device, electronic equipment and storage medium

Abstract

The invention provides a task performance determination method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring an electrocardiogram characteristic parameter, a blood oxygen characteristic parameter and a workload when a user executes an instruction response task; inputting the electrocardio characteristic parameters and the blood oxygen characteristic parameters into a fatigue state estimation model to obtain a fatigue state grade output by the fatigue state estimation model; determining the task performance when the user executes the instruction response task according to the fatigue state level and the workload; wherein the fatigue state estimation model is obtained by taking a fatigue state evaluation scale as a label and training based on an electrocardiogram characteristic parameter sample, a blood oxygen characteristic parameter sample and a fatigue state grade sample. According to the method, the fatigue state grade with high accuracy can be obtained through the fatigue state estimation model on the basis of the electrocardiogram characteristic parameters and the blood oxygen characteristic parameters which are highly related to the fatigue state, and then the task performance with high accuracy is obtained in combination with the workload.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a task performance determination method, device, electronic device and storage medium. Background Art

[0002] Fatigue is a psychological and physiological state in which cognitive performance declines due to the human body performing high-intensity cognitive activities for a long time. When the human body enters the fatigue state, the body temperature will rise and the oxygen consumption will increase, which will affect the body's normal metabolism to a certain extent, increase the sense of fatigue, and seriously affect the work efficiency and ability of the operators when performing tasks.

[0003] Traditional task performance determination methods often use subjective or objective evaluation methods to determine the fatigue state of operators and then determine the corresponding task performance. However, since the former is based on the operator's subsequent recollection, the subjective evaluation method is highly subjective and easily affected by the operator's memory, making it difficult to monitor and evaluate the fatigue state of operators in real time and accurately during the operation process; while the latter is mainly based on a large number of physiological parameter characteristics of operators (such as eye movement, electromyography, heart rate, pulse, blood oxygen saturation, and electroencephalogram characteristics). If a physiological parameter characteristic is not accurate enough, it will also lead to the inability to accurately evaluate the fatigue state of operators during the operation process. In this way, neither the subjective evaluation method nor the objective evaluation method can determine the fatigue state with high accuracy, resulting in the final determined task performance being inaccurate. Summary of the invention

[0004] The present invention provides a task performance determination method, device, electronic device and storage medium, which are used to solve the defect in the prior art that neither the subjective evaluation method nor the objective evaluation method can determine the fatigue state with high accuracy, resulting in the task performance finally determined to be inaccurate. The present invention realizes an electrocardiogram characteristic parameter and a blood oxygen characteristic parameter with high correlation with the fatigue state, and through a trained fatigue state estimation model, it is possible to obtain a fatigue state level with high accuracy while simplifying data calculation, and then obtain a task performance with high accuracy in combination with the workload.

[0005] The present invention provides a method for determining task performance, comprising the following steps.

[0006] Collect the user's ECG characteristic parameters, blood oxygen characteristic parameters and workload when performing command response tasks; Inputting the electrocardiogram characteristic parameter and the blood oxygen characteristic parameter into a fatigue state estimation model to obtain a fatigue state level output by the fatigue state estimation model; determining, according to the fatigue state level and the workload, the task performance of the user when performing the instruction response task; Among them, the fatigue state estimation model is trained with a fatigue state evaluation scale as a label based on electrocardiogram feature parameter samples, blood oxygen feature parameter samples, and fatigue state level samples.

[0007] According to a task performance determination method provided by the present invention, determining the task performance of the user when executing the instruction response task according to the fatigue state level and the workload includes: determining a reaction time ratio influence coefficient, an error rate ratio influence coefficient, and a workload critical value corresponding to abnormal operations of the user at the fatigue state level according to the fatigue state level; determining the reaction time of the user when executing the instruction response task according to the workload and the reaction time ratio influence coefficient; determining the error rate of the user when executing the instruction response task according to the workload, the error rate ratio influence coefficient, and the workload critical value; and determining the task performance according to the reaction time and the error rate.

[0008] According to a task performance determination method provided by the present invention, determining the reaction time of the user when executing the instruction response task according to the workload and the reaction time ratio influence coefficient includes: determining the reaction time according to a reaction time calculation formula; wherein, the reaction time calculation formula is: t = a×l work +b; t represents the reaction time; a represents a first coefficient in the reaction time ratio influence coefficient; l work represents the workload; b represents a second coefficient in the reaction time ratio influence coefficient.

[0009] According to a task performance determination method provided by the present invention, determining the error rate of the user when executing the instruction response task according to the workload, the error rate ratio influence coefficient, and the workload critical value includes: determining the error rate according to an error rate calculation formula; wherein, the error rate calculation formula is: ; r represents the error rate; α represents a first coefficient in the error rate ratio influence coefficient; Q represents the workload critical value; β represents a second coefficient in the error rate ratio influence coefficient.

[0010] According to a task performance determination method provided by the present invention, determining the task performance according to the reaction time and the error rate includes: determining the task performance according to a task performance calculation formula; wherein, the task performance calculation formula is: ; wherein, represents the task performance; Stired represents the fatigue state level; T represents the maximum reaction time threshold required for the user to complete the instruction response task; C represents all the operation times when the user completes the instruction response task.

[0011] According to a task performance determination method provided by the present invention, the fatigue state estimation model is trained based on the following steps: obtaining the electrocardiogram feature parameter samples, blood oxygen feature parameter samples, and the fatigue state evaluation scale; inputting the electrocardiogram feature parameter samples and the blood oxygen feature parameter samples into a support vector machine model to obtain the fatigue state level samples output by the support vector machine model; updating the model parameters of the support vector machine model according to the fatigue state evaluation scale and the fatigue state level samples to obtain the trained fatigue state estimation model.

[0012] According to a task performance determination method provided by the present invention, the electrocardiogram feature parameters at least include: the standard deviation, total power value calculated from heart rate variability data, and the ratio of low-frequency range power to high-frequency range power, and the mean, standard deviation, maximum value, minimum value, standardized mean, standardized maximum value, standardized minimum value, and standardized standard deviation calculated from heart rate data; the blood oxygen feature parameters at least include: the mean, standard deviation, maximum value, minimum value, standardized mean square deviation, standardized maximum value, and standardized minimum value calculated from blood oxygen data.

[0013] The present invention also provides a task performance determination device, including the following modules: A data acquisition device, configured to acquire electrocardiogram feature parameters, blood oxygen feature parameters, and workload of a user when performing an instruction response task; A fatigue assessment device, configured to input the electrocardiogram feature parameters and the blood oxygen feature parameters into a fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; A data processing device, configured to determine the task performance of the user when performing the instruction response task according to the fatigue state level and the workload; wherein, the fatigue state estimation model is trained based on electrocardiogram feature parameter samples, blood oxygen feature parameter samples, and fatigue state level samples with the fatigue state evaluation scale as the label.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the computer program, it implements the task performance determination method as described in any one of the above.

[0015] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the task performance determination method described in any one of the above is implemented.

[0016] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the task performance determination method described in any one of the above is implemented.

[0017] The task performance determination method, device, electronic device and storage medium provided by the present invention collect the electrocardiogram characteristic parameters, blood oxygen characteristic parameters and workload of a user when executing an instruction response task; input the electrocardiogram characteristic parameters and the blood oxygen characteristic parameters into a fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; determine the task performance of the user when executing the instruction response task according to the fatigue state level and the workload; wherein, the fatigue state estimation model is trained based on electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples and fatigue state level samples with a fatigue state evaluation scale as labels. This method is based on electrocardiogram characteristic parameters and blood oxygen characteristic parameters that have a high correlation with the fatigue state. Through the trained fatigue state estimation model, while simplifying data calculation, a fatigue state level with high accuracy can be obtained, and then combined with the workload, a task performance with high accuracy can be obtained. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a flowchart of the task performance determination method provided by the present invention.

[0020] Figure 2 It is a simple schematic diagram of a pilot and a machine respectively executing an instruction response task provided by the present invention.

[0021] Figure 3 It is a schematic diagram of a scenario of human-machine collaborative execution of an instruction response task provided by the present invention.

[0022] Figure 4 It is an overall schematic diagram of the task performance determination method provided by the present invention.

[0023] Figure 5 It is a schematic diagram of the structure of the task performance determination device provided by the present invention.

[0024] Figure 6 It is a schematic structural diagram of the electronic device provided by the present invention. Specific embodiments

[0025] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0026] To better understand the embodiments of the present invention, first, the background art will be elaborated in detail: In the case where the operator is a flight crew member, the fatigue state of the flight crew member will be affected by factors such as the operation task, operation time, and resource reserve. When the flight crew member is performing an operation task, a series of complex human-machine interaction operations will be performed, thus continuously consuming the physical strength and cognitive resources of the flight crew member. As the operation time increases, the fatigue degree of the flight crew member is also continuously increasing. When the flight crew member is in a fatigue state, the physical strength and cognitive resources can be restored through rest or sleep, and even special intervention and regulation are required to restore the working ability. When the resource reserve is sufficient, the human body can withstand more consumption and recover faster. When the resource reserve is insufficient due to lack of sleep, the human body will enter a fatigue state faster and recover more slowly.

[0027] In the air combat mission scenario, the flight crew member is often in an operation environment with high mobility, high noise, high vibration, information overload, long flight duration, and night flight that violates biological laws, which is extremely likely to induce the flight crew member to generate a cognitive fatigue state. Cognitive fatigue is a state in which an individual's cognitive ability fails and is exhausted after a long time of performing tasks or activities with a high cognitive load. It is a sense of fatigue caused by the brain continuously concentrating, paying attention, and thinking for a long time. However, a certain degree of cognitive fatigue can be transformed into physical fatigue, causing a double burden on physiology and psychology. In the fatigue state, the short-term memory of the human body is prone to disorders. For example, an overly fatigued flight crew member may forget the numbers displayed on the instrument he just saw or forget the conversation with other flight crew members a few minutes ago. In addition, the flight crew member will have a slower reaction speed, a phenomenon of chaotic thinking logic, and may even cause a decline in physiological motor skills and be unable to maintain normal operations, thereby increasing the decision-making error rate and the task performance deviating greatly from the expectation. Therefore, the fatigue state mainly affects the execution ability of the flight crew member.

[0028] Traditional task performance determination methods often use subjective evaluation methods or objective evaluation methods to determine the fatigue state of operators, and then determine the corresponding task performance. Among them, the former mainly evaluates the fatigue state in the form of an evaluation scale, usually using the Karolinska Sleepiness Scale, the Stanford Sleepiness Scale, etc. Since the subjective evaluation method is based on the operator's retrospective memory, the subjective evaluation method has strong subjectivity and is easily affected by the operator's memory, making it difficult to monitor and evaluate the fatigue state of operators during the operation in real time and accurately; while the latter is mainly based on a large number of physiological parameter characteristics of the operator. If a certain physiological parameter characteristic is not accurate enough, it will also lead to the inability to accurately evaluate the fatigue state of the operator during the operation.

[0029] Due to the complexity of the combat system and combat patterns, the scene situation information faced by pilots in the battlefield has increased exponentially. While the complexity of the task itself has increased, the uncertainty of the battlefield operation environment and the variability of the collaborative operation mode have also simultaneously caused pilots to be overly fatigued.

[0030] Therefore, how to effectively improve the accuracy of determining the fatigue state and then improve the accuracy of determining the task performance has become a technical problem that urgently needs to be solved.

[0031] To solve the above technical problems, in the task performance determination method provided by the present invention, the electrocardiogram feature parameters, blood oxygen feature parameters and workload of the user during the execution of the instruction response task are collected; the electrocardiogram feature parameters and the blood oxygen feature parameters are input into the fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; according to the fatigue state level and the workload, the task performance of the user during the execution of the instruction response task is determined; wherein, the fatigue state estimation model is trained based on electrocardiogram feature parameter samples, blood oxygen feature parameter samples and fatigue state level samples with a fatigue state evaluation scale as the label. This method is based on electrocardiogram feature parameters and blood oxygen feature parameters with a high correlation with the fatigue state. Through the trained fatigue state estimation model, while simplifying the data calculation, it can obtain a relatively accurate fatigue state level, and then combine the workload to obtain a relatively accurate task performance. In addition, for the air combat mission scenario where the pilot is located, this task performance can be used to support the deployment of the pilot's combat mission and the planning of the comprehensive pre-war mission design in the air combat mission.

[0032] Taking the operator as a pilot as an example, before implementing the task performance determination method provided by the present invention, the following processes are also included: Process 1: Design a comprehensive fatigue induction experiment.

[0033] The fatigue state of flight personnel is mainly affected by factors such as sleep, work intensity, mood, and stress. To ensure that flight personnel reach different levels of fatigue states before performing command response tasks, certain test subjects need to be pre-set to actively induce the fatigue state of flight personnel, which serves as a precondition for subsequent tests. Generally, the means that can be adopted include but are not limited to: sleep deprivation, strenuous exercise, etc.

[0034] Process 2: Construct a test module that can be used to evaluate the fatigue state of flight personnel.

[0035] The evaluation of the fatigue state of flight personnel mainly includes subjective evaluation methods or objective evaluation methods.

[0036] Among them, the subjective evaluation method: evaluate the fatigue state in the form of an evaluation scale. Usually, the Karolinska Sleepiness Scale, the Stanford Sleepiness Scale, etc. are used. The fatigue state evaluation scale used in subjective evaluation combines the characteristics of the "Subjective Fatigue Assessment Scale", the "Athlete Psychological Fatigue Questionnaire", the "Visual Analogue Scale", etc., and is formulated according to the actual situation of flight personnel. It is filled in item by item by flight personnel, and they circle "yes" or "no" according to their own state. Record the score of each item of the flight personnel and form a subjective evaluation of the flight personnel's own fatigue state.

[0037] Objective evaluation method: Through electrocardiogram characteristic parameter samples and blood oxygen characteristic parameter samples that are highly correlated with the fatigue state, study the model for monitoring the fatigue state of flight personnel by fusing the characteristics of these two types of samples. Using the fatigue state evaluation scale as a label, perform training on the Support Vector Machine (SVM) model to obtain a trained fatigue state estimation model.

[0038] Process 3: Extract test characteristic data, that is, collect electrocardiogram characteristic parameter samples and blood oxygen characteristic parameter samples.

[0039] Perform time-domain and frequency-domain characteristic analysis on the electrocardiogram characteristic parameter samples and blood oxygen characteristic parameter samples recorded during the test. The analyzed characteristics are mainly divided into two categories. One category is the segmented data directly measured during the test, and the other category is the standardized data relative to the resting state. At this time, the analyzed characteristics can be expressed as: f={f1,f2,⋯,f i ,⋯,f 18}, where f i is the i-th dimensional physiological characteristic.

[0040] Specifically, the electrocardiogram feature parameters may at least include: the standard deviation f1, total power value f2, and the ratio f3 of the power in the low-frequency range to the power in the high-frequency range calculated from the heart rate variability data (also known as heart rate variability), and the mean f4, standard deviation f5, maximum value f6, minimum value f7, standardized mean f8, standardized maximum value f9, and standardized minimum value f calculated from the heart rate data 10 and the standardized standard deviation f 11 . The blood oxygen feature parameters may at least include: the mean f 12 , standard deviation f 13 , maximum value f 14 , minimum value f 15 , standardized mean square error f 16 , standardized maximum value f 17 and the standardized minimum value f 18 .

[0041] Process 4: Construct a fatigue state evaluation model.

[0042] The present invention performs a three-classification process on the fatigue state to obtain the fatigue state level. Specifically, according to the characteristics of the test data samples (i.e., the above-mentioned electrocardiogram feature parameter samples and blood oxygen feature parameter samples), a support vector machine model is selected as the training model, which can effectively avoid overfitting for small samples of the test and has good generalization ability. The core of using the support vector machine model lies in the construction of the kernel function. Optionally, the commonly used kernel functions may include the Gaussian kernel function, linear kernel function, polynomial kernel function, and Sigmoid kernel function. The fatigue state classification based on multiple physiological feature parameters belongs to a typical non-linear high-dimensional space data classification problem. Therefore, the present invention constructs a model based on the Gaussian kernel function, verifies and trains the support vector machine model by the K-fold cross-validation method, makes full use of the data, avoids the contingency of the performance of the support vector machine model under a certain specific data segmentation state, improves the generalization performance of the support vector machine model, and uses the grid search method to optimize the penalty factor (C) and the kernel function width (γ).

[0043] Exemplarily, in combination with the above Process 2, Process 3, and Process 4, the fatigue state estimation model is trained based on the following steps: obtaining electrocardiogram feature parameter samples, blood oxygen feature parameter samples, and a fatigue state evaluation scale; inputting the electrocardiogram feature parameter samples and blood oxygen feature parameter samples into the support vector machine model to obtain the fatigue state level samples output by the support vector machine model; and updating the model parameters of the support vector machine model according to the fatigue state evaluation scale and the fatigue state level samples to obtain the trained fatigue state estimation model.

[0044] Among them, the trained fatigue state estimation model is a three-classification evaluation model for the fatigue state.

[0045] During the process of training a support vector machine model, an electronic device can first obtain samples of electrocardiogram feature parameters, samples of blood oxygen feature parameters, and a fatigue state evaluation scale. Then, the electronic device uses the samples of electrocardiogram feature parameters and samples of blood oxygen feature parameters as input data and inputs them into the model to be trained, that is, into the support vector machine model. The support vector machine model analyzes and classifies the input data to obtain prediction data, that is, samples of fatigue state levels. Next, the electronic device uses the fatigue state evaluation scale as a label and combines it with the samples of fatigue state levels for error analysis to update the model parameters of the support vector machine model and obtain a trained fatigue state estimation model. This fatigue state estimation model does not require a large amount of sample data for training. It only depends on samples of electrocardiogram feature parameters and samples of blood oxygen feature parameters that have a high correlation with the fatigue state, and can obtain a fatigue state estimation model with high accuracy for subsequent output of the fatigue state level.

[0046] Process 5: Design a method for evaluating the task performance of flight personnel.

[0047] For an instruction response task, the task performance mainly depends on the performance of the flight personnel's reaction execution ability, as well as the response delay of the machine system and the performance of the logical timing function. It should be noted that the reaction execution ability is an inherent characteristic of the human body, and there are differences in ability performance among different individuals. For the same individual, the performance of the reaction execution ability remains relatively stable and only changes due to factors such as practice and forgetting. However, in a specific task scenario, the ability performance of flight personnel will be significantly and dynamically affected by their physiological and mental states.

[0048] Among them, the reaction execution ability refers to the ability of an individual to quickly make a reaction after receiving a stimulus, which is very important for activities such as handling emergencies, coping with unexpected events, and performing high-speed movements, such as temporary instruction tasks. From the coupling relationship of the human-machine-environment system, it can be seen that the task performance of the human-machine-environment system is ultimately reflected in the response to task requirements. Therefore, under the condition that the abilities of humans and machines are measurable and controllable, the performance of the operation ability can be characterized by task performance, and the reaction execution ability can be manifested as the result of the corresponding task performance of the corresponding instruction response.

[0049] In the process of designing the task performance method, in the task scenario of fatigue estimation, the task performance of flight personnel can be calculated using the task completion quality. Specifically, indicators such as operation accuracy rate and behavioral reaction time can be used for quantitative evaluation, which can reflect the quality of the interaction operation of flight personnel when completing a specific task (such as an instruction response task), and the reaction execution ability of the flight personnel can be represented by relevant indicators. By evaluating the reaction time of flight personnel from receiving the instructions of the ground commander to performing the operation, and the proportion of abnormal operations (that is, ineffective operations and / or incorrect operations) in all operations when performing a specific task, a comprehensive evaluation of the task performance can be carried out.

[0050] Optionally, the initial calculation formula for task performance is as follows: ; Wherein, represents task performance; t represents reaction time; r represents error rate; T represents the maximum reaction time threshold required for the user to complete the instruction response task; C represents all the operation times when the user completes the instruction response task.

[0051] Process 6, Variable Selection and Definition of Instruction Response Task.

[0052] For the instruction response task, the performance indicators corresponding to the two dimensions of reaction time t and error rate r are mainly concerned (such as task performance ).

[0053] Exemplarily, as Figure 2 shown, it is a schematic diagram of the scenario where the pilot and the machine respectively execute the instruction response task provided by the present invention. In Figure 2 , the performance indicator of the pilot executing the instruction response task can be represented by P Ha (t, r); the performance indicator of the machine executing the instruction response task can be represented by P Ma (t, r).

[0054] In the actual application scenario, a complete instruction response task generally has both the link executed by the pilot and the link executed by the machine. Among them, the operation link of the instruction response task can be represented by Stp a .

[0055] Exemplarily, as Figure 3 shown, it is a schematic diagram of the scenario where the human-machine collaboratively executes the instruction response task provided by the present invention. In Figure 3 , the requirement of the instruction response task can be represented by P damand (t, r). It can be seen from Figure 3 that there will be multiple paths from the task instruction to the response execution in an instruction response task. The operation process path can be represented by E work-path . Each execution link will bring corresponding reaction time and error rate. The comprehensive performance after multiple execution links in series is the cumulative result of the reaction time and error rate brought by all execution links. In Figure 3 , the orange dashed box exemplifies the path where all links are executed by the pilot. The comprehensive task performance can be represented by Pa(t, r)=P Ha1 +P Ha2 +P Ha3 ; the green dashed box exemplifies the path where the pilot executes the first two execution links and the machine executes the last link. The comprehensive task performance can be represented by Pa(t, r)=P Ma1 +PMa2 +P Ma3 is represented as

[0056] As the performance of the pilot's response task to corresponding instructions changes under different fatigue state level results S tired and the processing performance of the machine for different instruction signals changes under different working conditions, by adjusting the operation links and processes, on the one hand, the workload I borne by the pilot can be changed work and on the other hand, the cumulative results of the reaction time and error rate in the task performance will be changed

[0057] The task performance determination method provided by the present invention will be elaborated in detail below

[0058] Figure 1 is a schematic flowchart of the task performance determination method provided by the present invention. As Figure 1 shown, the method includes Step 101, collect the electrocardiogram characteristic parameters, blood oxygen characteristic parameters and workload of the user when performing the instruction response task

[0059] Step 102, input the electrocardiogram characteristic parameters and blood oxygen characteristic parameters into the fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model

[0060] Among them, the fatigue state estimation model is trained based on electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples and fatigue state level samples with the fatigue state evaluation scale as the label

[0061] Since the fatigue state estimation model has been pre-trained, after the electronic device collects the physiological index data of the user when performing the instruction response task, that is, the electrocardiogram characteristic parameters and blood oxygen characteristic parameters, the electrocardiogram characteristic parameters and blood oxygen characteristic parameters can be directly used as input data and input into the trained fatigue state estimation model to output a relatively accurate fatigue state level

[0062] It should be noted that since the fatigue state estimation model can perform three-classification processing on the fatigue state, the fatigue state level output by the fatigue state estimation model is one of the low-level fatigue state, medium-level fatigue state and high-level fatigue state

[0063] The whole process does not require a large number of physiological parameter characteristics, only depends on the electrocardiogram characteristic parameters and blood oxygen characteristic parameters with a high correlation with the fatigue state, and through the trained fatigue state estimation model, a relatively accurate fatigue state level can be obtained. At the same time, the model calculation process is simplified and the data processing efficiency is improved

[0064] Step 103: Determine the task performance of the user when executing the instruction response task according to the fatigue state level and the workload.

[0065] In some embodiments, for the electronic device to determine the task performance of the user when executing the instruction response task according to the fatigue state level and the workload, it may include: The electronic device determines the reaction time ratio impact coefficient, the error rate ratio impact coefficient, and the workload critical value corresponding to abnormal operations of the user at the fatigue state level according to the fatigue state level; the electronic device determines the reaction time of the user when executing the instruction response task according to the workload and the reaction time ratio impact coefficient; the electronic device determines the error rate of the user when executing the instruction response task according to the workload, the error rate ratio impact coefficient, and the workload critical value; the electronic device determines the task performance according to the reaction time and the error rate.

[0066] In the process of the electronic device determining the task performance of the user when executing the instruction response task, since the corresponding reaction time ratio impact coefficient, error rate ratio impact coefficient, and workload critical value are different for different fatigue state levels, the electronic device can determine the corresponding reaction time ratio impact coefficient, error rate ratio impact coefficient, and workload critical value based on the fatigue state level output by the fatigue state estimation model; then, calculate the workload and the reaction time ratio impact coefficient to obtain the reaction time. At the same time, calculate the workload, the error rate ratio impact coefficient, and the workload critical value to obtain the error rate, and further combine the reaction time to determine the task performance of the user.

[0067] In some embodiments, for the electronic device to determine the reaction time of the user when executing the instruction response task according to the workload and the reaction time ratio impact coefficient, it may include: The electronic device determines the reaction time according to the reaction time calculation formula.

[0068] Wherein, the reaction time calculation formula is: t = a×l work +b; t represents the reaction time; a represents the first coefficient in the reaction time ratio impact coefficient; l work represents the workload; b represents the second coefficient in the reaction time ratio impact coefficient.

[0069] Based on the above reaction time calculation formula, the electronic device can obtain the reaction time of the user with higher accuracy when executing the instruction response task.

[0070] In some embodiments, for the electronic device to determine the error rate of the user when executing the instruction response task according to the workload, the error rate ratio impact coefficient, and the workload critical value, it may include: The electronic device determines the error rate according to the error rate calculation formula.

[0071] Among them, the error rate calculation formula is: ; r represents the error rate; α represents the first coefficient in the error rate proportional impact coefficient; Q represents the workload critical value; β represents the second coefficient in the error rate proportional impact coefficient.

[0072] Based on the above reaction time calculation formula, the electronic device can obtain the error rate of a user with high accuracy when performing an instruction response task.

[0073] That is to say, for different fatigue state levels, the first and second coefficients in the reaction time proportional impact coefficient, the first and second coefficients in the error rate proportional impact coefficient, and the workload critical value are different.

[0074] In some embodiments, the electronic device determines the task performance according to the reaction time and the error rate, including: determining the task performance according to the task performance calculation formula.

[0075] Among them, the task performance calculation formula is: ; Among them, represents the task performance; S tired represents the fatigue state level; T represents the longest reaction time critical value required for the user to complete the instruction response task; C represents all the operation times when the user completes the instruction response task.

[0076] Based on the above reaction time calculation formula, the electronic device can obtain the task performance of a user with high accuracy when performing an instruction response task.

[0077] In the embodiments of the present invention, the electrocardiogram characteristic parameters, blood oxygen characteristic parameters and workload of the user when performing the instruction response task are collected; the electrocardiogram characteristic parameters and blood oxygen characteristic parameters are input into the fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; the task performance of the user when performing the instruction response task is determined according to the fatigue state level and the workload; among them, the fatigue state estimation model is trained based on electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples and fatigue state level samples with the fatigue state evaluation scale as the label. This method is based on electrocardiogram characteristic parameters and blood oxygen characteristic parameters with high correlation with the fatigue state. Through the trained fatigue state estimation model, while simplifying the data calculation, it can obtain a fatigue state level with high accuracy, and then combine the workload to obtain a task performance with high accuracy.

[0078] To better understand the task performance determination method provided by the present invention, the task performance determination method can be described by an overall example, such as Figure 4 shown, which is the overall schematic diagram of the task performance determination method provided by the present invention. FromFigure 4 As can be seen, in the case of generating an instruction response task requirement, an instruction task response scenario can be preset. At this time, the electronic device can evaluate the fatigue state of the user when performing the instruction response task. Specifically, using the fatigue state evaluation scale as a label, training is performed based on the electrocardiogram feature parameter samples, blood oxygen feature parameter samples, and fatigue state level samples to obtain a fatigue state estimation model. Then, the electronic device inputs the collected electrocardiogram feature parameters and blood oxygen feature parameters into the fatigue state estimation model, obtains the fatigue state level output by the fatigue state estimation model, and combines the task assessment index data (such as reaction time and error rate). Through the above task performance calculation formula, the task performance of the user when performing the instruction response task is obtained.

[0079] Throughout the process, based on the electrocardiogram feature parameters and blood oxygen feature parameters that have a relatively high correlation with the fatigue state, through the trained fatigue state estimation model, while simplifying the data calculation, a fatigue state level with relatively high accuracy can be obtained, and then a task performance with relatively high accuracy can be obtained.

[0080] To better understand the task performance determination method provided by the present invention, taking the task of continuous dispatch with high intensity as an example, the above processes 1 - 6 and Figure 1 the task performance determination method shown can be specifically exemplified and described, which may include the following steps: Step 1: Design a ground simulation test.

[0081] In the comprehensive fatigue induction test, to simulate the state impact brought by the long - time continuous operation of flight crew in the air, 11 professional flight crew first perform two running exercises on a treadmill, each lasting 7 minutes (min). Different fatigue states are induced by setting the slope level of running. Then, they enter the flight simulation system based on the cockpit simulation platform for a 7 - minute flight task. To reduce the influence of external factors on the collection of physiological parameters, a flexible electrocardiogram sensor and an ear - mounted blood oxygen monitoring system are used to collect the electrocardiogram signal and blood oxygen signal of the test flight crew respectively, and the subjective evaluation of the flight crew in the fatigue state is collected before and after performing the instruction response task. Before the start of each group of tests, all flight crew sit still for 7 minutes and keep their breathing steady, and the physiological signals during this process are collected as the resting - state control group.

[0082] Step 2: Design a subjective scale, that is, a fatigue state evaluation scale.

[0083] The fatigue state evaluation scale used for subjective evaluation combines the characteristics of the "Subjective Fatigue Assessment Scale", the "Athlete Psychological Fatigue Questionnaire", the "Visual Analogue Scale", etc., and is formulated according to the actual situation of flight personnel. The flight personnel fill it out item by item, circle "yes" or "no" according to their own state. Except for items 10, 13, and 14 in the subjective evaluation questionnaire which are reverse scored, that is, answering "yes" is scored 0 points and answering "no" is scored 1 point, other items are forward scored, that is, answering "yes" is scored 1 point and answering "no" is scored 0 points. The scores of items 1 to 8 are added together to obtain the physical fatigue score, and the scores of items 9 to 14 are added together to obtain the cognitive fatigue score. The sum of the two is the total fatigue score. The highest total score is 14. The higher the score, the more serious the fatigue degree of the flight personnel is reflected. The initially set division thresholds are: 3 points and 9 points, that is, a score of 0 to 3 points can be considered not fatigued, 4 to 9 points is moderate fatigue, and 10 to 14 points is severe fatigue, so as to obtain the subjective fatigue feeling degree of the subjects after the operation.

[0084] Exemplarily, the problem descriptions of the subjective evaluation questionnaire are shown in Table 1.

[0085] Table 1: As can be seen from Table 1, the subjective evaluation questionnaire has 14 questions.

[0086] Step 3: Collect and process data.

[0087] Perform fatigue significance analysis on the 18 characteristic parameters involved in the above process 3. Among them, P < 0.05 is significant, and the analysis results are shown in Table 2.

[0088] Table 2: As can be seen from Table 2: The total power value f2, mean value f4, and standardized mean value f8 included in the electrocardiogram characteristic parameters, and the mean value f 12 , minimum value f 15 , standardized mean square deviation f 16 and standardized minimum value f 18 of the blood oxygen characteristic parameters all show an upward trend with the aggravation of the fatigue state; the ratio f3 of the low-frequency range power to the high-frequency range power, standard deviation f5, and standardized standard deviation f 11 included in the electrocardiogram characteristic parameters all first decrease and then increase with the change of the fatigue state eigenvalue; the minimum value f7 and standardized minimum value f 10 included in the electrocardiogram characteristic parameters all first increase and then decrease with the change of the fatigue state eigenvalue.

[0089] Among them, the standard deviation f1, maximum value f6, and standardized maximum value f9 included in the electrocardiogram characteristic parameters, and the standard deviation f of the blood oxygen characteristic parameters13 and the maximum value f 14 and the normalized maximum value f 17 both have no significant correlation with the change of fatigue state. That is to say, 12 out of the 18 characteristic parameters have a significant correlation with the change of fatigue state of fighting fish.

[0090] Step 4: Construct a fatigue state evaluation model.

[0091] Using the results of the fatigue state evaluation scale as labels, the support vector machine model takes 12 significant characteristic parameters as the input data for model training. Based on the K-fold cross-validation method, 8-fold cross-validation training is performed on 33 groups of samples of comprehensive fatigue respectively. The test results are shown in Table 3, with the unit of %.

[0092] Among them, H represents the sample data set for model training, H1 represents using the first group of data as the test group of the model, and other data groups (i.e., H2 - H8) as the training group. The model is constructed through the support vector machine model, and its value represents the accuracy of the model when using the first group of data as the test group.

[0093] Table 3: It can be seen from Table 3 that the average comprehensive fatigue recognition accuracy of the model among different subject groups is 82.6%, that is, the constructed fatigue state estimation model has good applicability to the identification of comprehensive fatigue state in flight tasks. This fatigue state estimation model provides a feasible solution for effectively identifying the fatigue state level of flight personnel, can support the real-time and effective evaluation of the fatigue state of flight personnel based on electrocardiogram signals and blood oxygen signals, and provides methods and means for the intervention and recovery of abnormal states.

[0094] Step 5: Flight test and data collection.

[0095] Taking continuous high-intensity sortie flight as the test scenario, typical flight personnel are selected for continuous tracking. During the command response task, the blood oxygen saturation, heart rate, and electrocardiogram parameters of the flight personnel are collected during the flight, and a questionnaire survey on the fatigue degree of the flight personnel is conducted. During the test process, considering the impact of high-intensity flight on safety, routine tasks that have been fully trained are selected to keep the stress state of the flight personnel normal. By setting up ground command and guidance and not adopting dynamic confrontation and other measures, the situation awareness state of the flight personnel is kept high.

[0096] The fatigue state evaluation model is used to evaluate the fatigue state level of flight personnel, and it is compared with the subjective fatigue evaluation results of the subjective evaluation questionnaire, as shown in Table 4.

[0097] Table 4: To eliminate the influence of individual differences, the same pilot was selected to perform the above command response tasks. By analyzing the data shown in Table 4, it can be seen that: For the same pilot, when the sleep was sufficient on the day before the flight, the fatigue state level was consistent with the subjective fatigue evaluation result, both being in a low fatigue state; when the sleep was insufficient due to performing a night flight mission on the day before the flight, the fatigue state level was from a low fatigue state to a moderate fatigue state, which was consistent with the subjective fatigue evaluation result; the fatigue state level of the second flight was consistent with the subjective fatigue evaluation result, being in a moderate fatigue state; when the flight duration increased, the fatigue state level was from a moderate fatigue state to a high fatigue state, which was consistent with the subjective fatigue evaluation result. That is, the fatigue state assessment model can accurately distinguish the fatigue state level of the pilot, has good consistency with the subjective fatigue evaluation result, and the model evaluation is effective.

[0098] By comparing the fatigue state level obtained from the real-time assessment of the fatigue state with the subjective fatigue evaluation result, it can be seen that the fatigue state assessment model can accurately identify the changes in the fatigue state of the pilot during the flight. However, due to the effect of retrospective recall in the subjective evaluation method, it mainly reflects the final state of the pilot's fatigue and has hysteresis. Therefore, the fatigue state assessment model constructed in the present invention has good accuracy and real-time performance, and can perform real-time assessment of the fatigue state of the pilot during the mission process.

[0099] It can be seen from this that by regulating factors such as resource reserve and operation time, the human fatigue state can be affected. As the operation time continues, the fatigue state changes from low to medium to high, indicating that the large-intensity continuous flight scenario is effective. When the operation time is long and the resource reserve is poor, the model evaluation results are in a moderate fatigue state and a high fatigue state. The above fatigue state assessment model can correctly distinguish the fatigue state and maintain good consistency with the subjective fatigue evaluation result, and the model evaluation is effective.

[0100] That is to say, the leave-one-out method is used to cross-validate the model, and the rationality and reliability of the model are confirmed by mutual verification with the subjective fatigue evaluation result of the subjective evaluation questionnaire.

[0101] Step 6: Analysis results of flight test data.

[0102] In the flight test scenario, the tasks ordered by the ground command post and executed by the aircrew all belong to command response tasks. Five typical command response tasks of flight tests are selected for analysis in the present invention, and the workload and fatigue state levels are shown in Tables 5 - 9 respectively.

[0103] Table 5: Table 6: Table 7: Table 8: Table 9: Among them, T represents the instruction response task; Q1 represents the visual channel; Q2 represents the auditory channel; Q3 represents the cognitive channel; Q4 represents the fine motor channel; Q5 represents the speech channel; Q6 represents the gross motor channel; Q7 represents the time resource.

[0104] Step 7: Task performance calculation.

[0105] The task performance corresponding to different workloads and fatigue state levels is shown in Table 10.

[0106] Table 10: Step 8: Parameter estimation of the task performance corresponding to the instruction response task.

[0107] For the first coefficient a and the second coefficient b in the influence coefficient of the fatigue state level on the reaction time ratio in the reaction time calculation formula, based on the large-intensity flight test data above, the fatigue state levels S tired are selected as the moderate fatigue state and the high fatigue state, and the workload l work is the statistical result of the reaction time under low, medium, and high conditions for parameter identification. Among them, for the moderate fatigue state, a = 0.0161 and b = 0.7407; for the high fatigue state, a = 0.0377 and b = 1.6411. Based on this, using the above reaction time calculation formula, we can obtain: t1 = 0.0161×l work + 0.7407, R 2 = 0.9008, where t1 represents the reaction time corresponding to the moderate fatigue state; t2 = 0.0377×l work + 1.6411, R 2 = 0.8086, where t1 represents the reaction time corresponding to the high fatigue state.

[0108] For the first coefficient α and the second coefficient β in the influence coefficient of the fatigue state level on the error rate ratio in the reaction time calculation formula, based on the large-intensity flight test data above, the fatigue state levels S tiredParameter identification is performed on the error rate statistical results in the moderate fatigue state and the high fatigue state. The workload critical value Q corresponding to the moderate fatigue state is 41.1; the Q corresponding to the high fatigue state is 21.5; α = 0.0009, β = 0.0061. Based on this, using the above error rate calculation formula, we can obtain: r = 0.0009×l work -0.0061, R 2 = 1.

[0109] Based on the above analysis, the influence law of workload - fatigue state level - task performance of the instruction response task is as follows: The influence of the fatigue state level on the workload - task performance curve shows an approximate proportional adjustment effect. The higher the fatigue state level, the higher the degree of task performance decline and the faster it decreases with the workload. The reaction time shows an approximate proportional increase trend as the fatigue state increases. As the workload increases, the increase amplitude of the reaction time intensifies; abnormal operations start to appear in the high workload area of the moderate fatigue state, and the error rate in the medium and high workload areas of the high fatigue state increases significantly.

[0110] The influence of the fatigue state level on task performance has a cumulative effect. In the current three - classification fatigue state, the decline effect on task performance caused by the reaction time effect will intensify in the same - level fatigue state. The reason for this phenomenon is that the fatigue state is a continuously changing state. As the working time accumulates, the fatigue state changes continuously. However, the state assessment is classified. Therefore, in the high fatigue state, as the working time increases and the fatigue state continues to increase, the task performance will continue to decline.

[0111] The task performance determination method provided by the present invention has been preliminarily applied to flight tests, and the scientific validity of the method has been verified in practical applications. It has achieved great results in the field of engineering technology and has the value of popularization.

[0112] Next, the task performance determination device provided by the present invention will be described. The task performance determination device described below can be correspondingly referred to the task performance determination method described above.

[0113] Figure 5 is the structural schematic diagram of the task performance determination device provided by the present invention, as Figure 5 shown. The device includes: A data acquisition device 501, configured to collect electrocardiogram characteristic parameters, blood oxygen characteristic parameters, and workload of a user when performing an instruction response task; A fatigue assessment device 502, configured to input the electrocardiogram characteristic parameters and the blood oxygen characteristic parameters into a fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; A data processing device 503 is configured to determine the task performance of the user when executing the instruction response task according to the fatigue state level and the workload; wherein, the fatigue state estimation model is trained based on electrocardiogram feature parameter samples, blood oxygen feature parameter samples, and fatigue state level samples with a fatigue state evaluation scale as the label.

[0114] Optionally, the data processing device 503 is specifically configured to determine a reaction time ratio influence coefficient, an error rate ratio influence coefficient, and a workload critical value corresponding to abnormal operations of the user at the fatigue state level according to the fatigue state level; determine the reaction time of the user when executing the instruction response task according to the workload and the reaction time ratio influence coefficient; determine the error rate of the user when executing the instruction response task according to the workload, the error rate ratio influence coefficient, and the workload critical value; and determine the task performance according to the reaction time and the error rate.

[0115] Optionally, the data processing device 503 is specifically configured to determine the reaction time according to a reaction time calculation formula; wherein, the reaction time calculation formula is: t = a×l work +b; t represents the reaction time; a represents the first coefficient in the reaction time ratio influence coefficient; l work represents the workload; b represents the second coefficient in the reaction time ratio influence coefficient.

[0116] Optionally, the data processing device 503 is specifically configured to determine the error rate according to an error rate calculation formula; wherein, the error rate calculation formula is: ; r represents the error rate; α represents the first coefficient in the error rate ratio influence coefficient; Q represents the workload critical value; β represents the second coefficient in the error rate ratio influence coefficient.

[0117] Optionally, the data processing device 503 is specifically configured to determine the task performance according to a task performance calculation formula; wherein, the task performance calculation formula is: ; wherein, represents the task performance; S tired represents the fatigue state level; T represents the longest reaction time critical value required for the user to complete the instruction response task; C represents all the operation times when the user completes the instruction response task.

[0118] Optionally, the fatigue state estimation model is trained based on the following steps: obtaining the electrocardiogram (ECG) feature parameter samples, blood oxygen feature parameter samples, and the fatigue state evaluation scale; inputting the ECG feature parameter samples and the blood oxygen feature parameter samples into a support vector machine model to obtain the fatigue state level samples output by the support vector machine model; updating the model parameters of the support vector machine model according to the fatigue state evaluation scale and the fatigue state level samples to obtain the trained fatigue state estimation model.

[0119] Optionally, the ECG feature parameters at least include: the standard deviation, total power value, and ratio of low-frequency range power to high-frequency range power calculated from heart rate variability data, and the mean, standard deviation, maximum value, minimum value, standardized mean, standardized maximum value, standardized minimum value, and standardized standard deviation calculated from heart rate data; the blood oxygen feature parameters at least include: the mean, standard deviation, maximum value, minimum value, standardized mean square deviation, standardized maximum value, and standardized minimum value calculated from blood oxygen data.

[0120] Figure 6 An example of the physical structure diagram of an electronic device is shown as Figure 6 As shown, the electronic device may include: a processor 610, a communication interface 620, a memory 630, and a communication bus 640. Among them, the processor 610, the communication interface 620, and the memory 630 complete mutual communication through the communication bus 640. The processor 610 can call the logical instructions in the memory 630 to execute the task performance determination method, which includes: collecting the ECG feature parameters, blood oxygen feature parameters, and workload of the user when performing an instruction response task; inputting the ECG feature parameters and the blood oxygen feature parameters into the fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; determining the task performance of the user when performing the instruction response task according to the fatigue state level and the workload; where the fatigue state estimation model is trained based on the ECG feature parameter samples, blood oxygen feature parameter samples, and fatigue state level samples with the fatigue state evaluation scale as the label.

[0121] In addition, when the logical instructions in the above-mentioned memory 630 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0122] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the task performance determination method provided by the above-mentioned various methods. The method includes: collecting electrocardiogram characteristic parameters, blood oxygen characteristic parameters, and workload of a user when performing an instruction response task; inputting the electrocardiogram characteristic parameters and the blood oxygen characteristic parameters into a fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; determining the task performance of the user when performing the instruction response task according to the fatigue state level and the workload; wherein, the fatigue state estimation model is trained based on electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples, and fatigue state level samples with a fatigue state evaluation scale as a label.

[0123] In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the task performance determination method provided by the above-mentioned various methods. The method includes: collecting electrocardiogram characteristic parameters, blood oxygen characteristic parameters, and workload of a user when performing an instruction response task; inputting the electrocardiogram characteristic parameters and the blood oxygen characteristic parameters into a fatigue state estimation model to obtain the fatigue state level output by the fatigue state estimation model; determining the task performance of the user when performing the instruction response task according to the fatigue state level and the workload; wherein, the fatigue state estimation model is trained based on electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples, and fatigue state level samples with a fatigue state evaluation scale as a label.

[0124] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0125] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or equivalently replace some of the technical features. However, these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for determining task performance, characterized in that: include: Collect the user's ECG characteristic parameters, blood oxygen characteristic parameters and workload when performing command response tasks; Inputting the electrocardiogram characteristic parameter and the blood oxygen characteristic parameter into a fatigue state estimation model to obtain a fatigue state level output by the fatigue state estimation model; determining, according to the fatigue state level and the workload, the task performance of the user when performing the instruction response task; The fatigue state estimation model is obtained by training based on the fatigue state evaluation scale as a label and based on the electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples and fatigue state level samples.

2. The method according to claim 1, characterized in that Determining the task performance of the user when performing the instruction response task according to the fatigue state level and the workload includes: According to the fatigue state level, determining a reaction time proportional influence coefficient, an error rate proportional influence coefficient, and a workload critical value corresponding to abnormal operation of the user at the fatigue state level; Determining the reaction time of the user when performing the instruction response task according to the workload and the reaction time ratio influence coefficient; Determining the error rate of the user when performing the instruction response task according to the workload, the error rate ratio influence coefficient and the workload critical value; The task performance is determined based on the reaction time and the error rate.

3. The method according to claim 2, characterized in that The determining, according to the workload and the reaction time ratio influence coefficient, the reaction time of the user when performing the instruction response task includes: Determine the reaction time according to a reaction time calculation formula; Wherein, the reaction time calculation formula is: t=a×l work +b; t represents the reaction time; a represents the first coefficient in the reaction time ratio influence coefficient; l work represents the workload; b represents the second coefficient in the reaction time ratio influence coefficient.

4. The method according to claim 3, characterized in that The determining, according to the workload, the error rate ratio influence coefficient and the workload critical value, the error rate of the user when performing the instruction response task includes: Determine the error rate according to an error rate calculation formula; The error rate calculation formula is: ; r represents the error rate; α represents the first coefficient in the error rate ratio influence coefficient; Q represents the workload critical value; β represents the second coefficient in the error rate ratio influence coefficient.

5. The method according to claim 4, characterized in that Determining the task performance according to the reaction time and the error rate includes: Determine the task performance according to the task performance calculation formula; The task performance calculation formula is: ； in, Indicates the task performance; S tired represents the fatigue state level; T represents the longest reaction time critical value required for the user to complete the instruction response task; C represents the total number of operations when the user completes the instruction response task.

6. The method according to any one of claims 1 to 5, characterized in that: The fatigue state estimation model is trained based on the following steps: Acquire the electrocardiogram characteristic parameter sample, the blood oxygen characteristic parameter sample, and the fatigue state evaluation scale; Inputting the ECG characteristic parameter samples and the blood oxygen characteristic parameter samples into a support vector machine model to obtain the fatigue state level samples output by the support vector machine model; The model parameters of the support vector machine model are updated according to the fatigue state evaluation scale and the fatigue state grade samples to obtain a trained fatigue state estimation model.

7. The method according to any one of claims 1 to 5, characterized in that: The ECG characteristic parameters include at least: standard deviation, total power value, and ratio of low-frequency range power to high-frequency range power calculated from heart rate variability data, and mean, standard deviation, maximum value, minimum value, standardized mean value, standardized maximum value, standardized minimum value, and standardized standard deviation calculated from heart rate data; The blood oxygen characteristic parameters include at least: mean, standard deviation, maximum value, minimum value, standardized mean square error, standardized maximum value and standardized minimum value calculated from the blood oxygen data.

8. A task performance determination device, characterized in that: include: A data acquisition device, used to collect the user's electrocardiogram characteristic parameters, blood oxygen characteristic parameters and workload when performing a command response task; A fatigue assessment device, used for inputting the electrocardiogram characteristic parameter and the blood oxygen characteristic parameter into a fatigue state estimation model to obtain a fatigue state level output by the fatigue state estimation model; A data processing device is used to determine the task performance of the user when executing the instruction response task based on the fatigue state level and the workload; wherein the fatigue state estimation model is obtained by training based on the fatigue state evaluation scale as a label and based on electrocardiogram characteristic parameter samples, blood oxygen characteristic parameter samples and fatigue state level samples.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the task performance determination method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the task performance determination method according to any one of claims 1 to 7 is implemented.

Images