Psychological assessment management method for power employees and related equipment

Abstract

The invention discloses an electric power employee psychological assessment management method and related equipment. The method comprises the following steps: acquiring physiological information of an electric power employee, an acquisition instruction of the physiological information being triggered based on work order information of the electric power employee, and determining an acquisition time range and an acquisition type of the physiological information based on the work order information; under the condition that psychological assessment is carried out on the electric power employees, based on the physiological information of the electric power employees, an assessment problem is generated through a multi-mode psychological assessment large model; and based on the physiological information of the power employees, the assessment problem and the feedback content of the power employees for the assessment problem, generating an assessment result through the multi-modal psychological assessment large model. The method can solve the problems that a current large model shows strong learning and reasoning capabilities in multiple fields, but still faces certain challenges in the aspect of mental health assessment, and the accuracy needs to be further improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computers. More specifically, the present invention relates to a method and related equipment for psychological assessment and management of power company employees. Background Art

[0002] The mental health assessment and intervention of power company employees are crucial because employees face high pressure and complex environments at work, and mental health problems may affect their work performance and quality of life. Accurate and reasonable psychological assessment and intervention measures require a comprehensive analysis of the multimodal information of employees. This information not only includes physiological signals of employees, such as heart rate, blood pressure, etc., but also includes language, behavior, and emotional data related to mental state. Therefore, a model capable of comprehensively processing this multimodal information is particularly crucial. Current large models have demonstrated powerful learning and reasoning abilities in multiple fields, but in the aspect of mental health assessment and intervention, they still face certain challenges. Especially when dealing with the individualized mental states of employees, how to effectively integrate data in multiple aspects such as physiology, emotion, and language for accurate analysis and intervention remains an urgent problem to be solved. Summary of the Invention

[0003] A series of simplified concepts are introduced in the Summary of the Invention section, which will be further elaborated in the Detailed Description section. The Summary of the Invention section of the present invention does not mean to attempt to define the key features and essential technical features of the claimed technical solution, nor does it mean to attempt to determine the protection scope of the claimed technical solution.

[0004] To solve the problem that current large models have demonstrated powerful learning and reasoning abilities in multiple fields, but in the aspect of mental health assessment, they still face certain challenges and the accuracy needs to be further improved. In a first aspect, the present invention proposes a method for psychological assessment and management of power company employees, and the method includes: Obtain the physiological information of the power company employee, and the acquisition instruction of the physiological information is triggered based on the work order information of the power company employee to determine the acquisition time range and acquisition type of the physiological information based on the work order information; When conducting a psychological assessment of the power company employee, generate assessment questions based on the physiological information of the power company employee through a multimodal psychological assessment large model; Generate an assessment result through the multimodal psychological assessment large model based on the physiological information of the power company employee, the assessment questions, and the feedback content of the power company employee regarding the assessment questions.

[0005] Optionally, it further includes: Establish a mapping between physiological information and doctor's questions as the first labeled data to train the psychological assessment large model, where the doctor's questions are the doctor's assessment questions given based on the psychological information; Establish a mapping between physiological information, doctor's assessment questions, training content of the power employee's feedback based on the doctor's assessment questions, and doctor's assessment results as the second labeled data to train the psychological assessment large model.

[0006] Optionally, it further includes: Obtain the model output quality ranking information of the multi-modal psychological assessment large model for multiple groups of output result information generated during an assessment process, where each group of output result information includes assessment questions and / or assessment results; Optimize and train the multi-modal psychological assessment large model based on the model output quality ranking information through the instruction fine-tuning method of comparing prompt words.

[0007] Optionally, the optimizing and training of the multi-modal psychological assessment large model based on the model output quality ranking information through the instruction fine-tuning method of comparing prompt words includes: Based on the model output quality ranking information, add quality level labels to the sample data composed of the output result information and associated model input information of the multi-modal psychological assessment large model during an assessment process to generate template sample data that conforms to at least two quality-level prompt word templates, and the output result information further includes intervention suggestion information; Optimize and train the multi-modal psychological assessment large model through the comparison between the template sample data that conforms to at least two quality-level prompt word templates.

[0008] Optionally, the physiological information is obtained through a wearable intelligent terminal device, and the method further includes: Obtain the action information of the power employee through the wearable intelligent terminal device; When the action information matches the target production operation scenario indicated by the work order information, trigger the acquisition instruction of the physiological information, and the target production operation scenario includes high-voltage live working scenarios, power failure repair scenarios, cable tunnel confined space inspection scenarios, and power dispatching center duty scenarios.

[0009] Optionally, it further includes: When receiving the acquisition instruction of the physiological information, evaluate the influence coefficient of the action information on the physiological information of the power employee as the action influence coefficient; Filter the physiological information based on the influence coefficient to predict the influence coefficient of the target generation operation scenario pressure on the physiological information of the power worker as the pressure influence coefficient, and use the physiological information and the associated pressure influence coefficient together as the target physiological information of the power worker; In the case of conducting a psychological assessment on the power worker, generate assessment questions based on the target physiological information of the power worker through a multi-modal psychological assessment large model; Generate an assessment result through the multi-modal psychological assessment large model based on the target physiological information of the power worker, the assessment questions, and the feedback content of the power worker for the assessment questions.

[0010] Optionally, it further includes: Obtain the coordinated operation workers who handle the same work order and the same production operation scenario as the power worker; Horizontally compare the pressure influence coefficients of the power worker and all the coordinated operation workers to classify the pressure influence coefficients; Use the physiological information, the associated pressure influence coefficient, and the classification information together as the target physiological information of the power worker.

[0011] In a second aspect, the present invention also proposes a power worker psychological assessment management device, including: An acquisition unit for acquiring the physiological information of a power worker, and the acquisition instruction of the physiological information is triggered based on the work order information of the power worker to determine the acquisition time range and acquisition type of the physiological information based on the work order information; A generation unit for generating assessment questions based on the physiological information of the power worker through a multi-modal psychological assessment large model in the case of conducting a psychological assessment on the power worker; An assessment unit for generating an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power worker, the assessment questions, and the feedback content of the power worker for the assessment questions.

[0012] In a third aspect, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program stored in the memory, it implements the steps of the power worker psychological assessment management method according to any one of the first aspects above.

[0013] In a fourth aspect, the present invention also proposes a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the power worker psychological assessment management method according to any one of the first aspects above.

[0014] In summary, the proposed psychological assessment management method for power workers in this application obtains the physiological information of power workers, and the acquisition instruction of the physiological information is triggered based on the work order information of the power workers, so as to determine the acquisition time range and acquisition type of the physiological information based on the work order information; In the case of psychological assessment of the power worker, assessment questions are generated through a multi-modal psychological assessment large model based on the physiological information of the power worker; an assessment result is generated through the multi-modal psychological assessment large model based on the physiological information of the power worker, the assessment questions, and the feedback content of the power worker for the assessment questions. Traditional psychological assessment methods usually rely on subjective questionnaires, while this method triggers physiological data collection through work orders, enabling psychological assessment to be based on objective physiological data, reducing the influence of human subjective factors on the assessment results. In addition, personalized assessment questions are generated based on physiological data fluctuations, making the psychological test more accurate and able to more accurately reflect the actual psychological state of the workers. This method comprehensively utilizes physiological data, work order information, and natural language feedback information, and conducts in-depth analysis through a multi-modal psychological assessment large model, overcoming the limitations of single physiological signal or questionnaire assessment. By combining time series data analysis and natural language processing, the accuracy of psychological state recognition is improved, making the assessment results more comprehensive.

[0015] In summary, the present invention triggers physiological data collection through work orders, combines personalized assessment questions, a multi-modal psychological assessment large model, and personalized psychological intervention measures, providing a precise and efficient mental health management solution for power workers, greatly enhancing the scientificity and practicality of mental health management.

[0016] For the psychological assessment management method of power workers of the present invention, other advantages, objectives, and features of the present invention will be partially reflected by the following description, and partially will also be understood by those skilled in the art through the research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to limit this specification. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings: Figure 1 It is a schematic flow chart of a psychological assessment management method for power workers provided by an embodiment of the present application; Figure 2 It is a schematic structural diagram of a psychological assessment management device for power workers provided by an embodiment of the present application; Figure 3 It is a schematic structural diagram of an electronic device for psychological assessment management of power workers provided by an embodiment of the present application. Detailed implementation manners

[0018] In the description and claims of this application and the above-mentioned drawings, terms such as "first", "second", "third", "fourth", etc. (if any) are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices. The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments.

[0019] To solve the problem of misdiagnosis or treatment delay that is likely to occur due to the large differences in the diagnosis and treatment levels and concepts between different regions and the relatively insufficient social attention to rheumatic immune diseases, please refer to Figure 1 , which is a schematic flowchart of a psychological assessment management method for power employees provided by an embodiment of this application, and specifically may include: steps S110 to S130.

[0020] S110, obtain the physiological information of the power employee, and the acquisition instruction of the physiological information is triggered based on the work order information of the power employee to determine the acquisition time range and acquisition type of the physiological information based on the work order information.

[0021] S120, in the case of performing a psychological assessment on the power employee, generate assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power employee.

[0022] S130, generate an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power employee, the assessment questions, and the feedback content of the power employee for the assessment questions.

[0023] It is understandable that by comprehensively analyzing the physiological data, work order information, and personalized assessment questions of power workers through a multi-modal psychological assessment large model, the accuracy of psychological assessment can be improved, and targeted intervention plans can be provided. Traditional psychological assessment methods usually rely on fixed questionnaires or single physiological indicators. In the above-mentioned solution, first, the work order information is used to automatically trigger the collection of physiological data to ensure that the data can truly reflect the physiological stress state of workers in the actual working environment. Second, when conducting psychological assessment, this method does not directly adopt traditional fixed questionnaires, but dynamically generates personalized assessment questions according to the fluctuations of physiological data, making the assessment more accurate. Finally, this method integrates multi-modal data through deep learning technology, including the analysis of time-series physiological signals (such as heart rate, blood pressure, EDA) and natural language processing (text information answered by workers), comprehensively assesses the psychological state of workers, and provides personalized intervention plans.

[0024] Exemplarily, the physiological information collection mechanism of this method is not to collect temporarily during psychological assessment, but is triggered based on work orders to ensure that the collected data can truly reflect the physiological stress state of workers in actual work. The work order information includes operation types (such as substation inspection, high-voltage equipment maintenance), environmental conditions (such as high temperature, high humidity, strong noise), task duration (such as night shift, continuous operation), task risk level (such as general maintenance, emergency accident handling), etc. The system automatically determines the collection type and time range of physiological data according to the work order information. For example, when working in a high-temperature environment, the system will collect heart rate, blood pressure, and skin conductance (EDA) before, during, and after the operation; during night shift operations, it will monitor heart rate variability (HRV) and electroencephalogram (EEG); when working at height, it will collect blood oxygen level and respiratory rate. This way can avoid human intervention in data collection, ensure the objectivity and authenticity of the data, and provide high-quality input for psychological assessment.

[0025] Exemplarily, when the system needs to conduct a psychological assessment of workers, it will not use a fixed questionnaire, but dynamically generate assessment questions based on the physiological data of the workers. Specifically, the system will analyze the fluctuations of physiological data and automatically generate personalized assessment questions that conform to the current state of the workers. For example, if it is detected that the worker's heart rate has been maintained at a high level for a long time after working in a high-temperature environment, the system may generate the question: "During the just task process, did you feel nervous or have palpitations?" If the HRV of the worker continues to decline after continuous night shifts, it may ask: "Have you felt lack of energy or difficulty concentrating recently?" This way of dynamically adjusting assessment questions makes the psychological test more in line with the actual psychological state of individuals, avoids the limitations of traditional questionnaire assessments, and improves the accuracy of the assessment.

[0026] Exemplarily, after the employee answers the assessment questions, the system will input the employee's physiological data, work order information, assessment questions and their feedback content into the multi-modal psychological assessment large model for analysis. This model combines deep learning technology to fuse and process data of different modalities. For example, the time-series physiological signal analysis part (such as a large language model) is used to process the trends of heart rate, blood pressure, and EDA, and extract potential stress patterns; the natural language processing part analyzes the text information of the employee's answers to identify their emotional state and psychological distress. After integrating the information of these two parts, the system generates a psychological assessment report. For example, if it is detected that the employee has a low HRV and a drastic change in EDA, and at the same time the feedback is "irritable and difficult to concentrate at work recently", the system may assess that the employee is in a state of chronic stress and recommend psychological intervention; if the physiological indicators recover quickly and the feedback is "adapting to stress", it may be assessed that the stress is moderate and no additional intervention is required.

[0027] In summary, the power employee psychological assessment management method provided by the embodiments of the present application generates assessment questions through the multi-modal psychological assessment large model based on the physiological information of the power employee; generates an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power employee, the assessment questions, and the feedback content of the power employee for the assessment questions. Traditional psychological assessment methods usually rely on subjective questionnaires, while this method triggers physiological data collection through work orders, making the psychological assessment based on objective physiological data, reducing the influence of human subjective factors on the assessment results. In addition, personalized assessment questions are generated based on physiological data fluctuations, making the psychological test more accurate and able to more accurately reflect the actual psychological state of the employee. This method comprehensively utilizes physiological data, work order information, and natural language feedback information, and conducts in-depth analysis through the multi-modal psychological assessment large model, overcoming the limitations of single physiological signal or questionnaire assessment. By combining time-series data analysis and natural language processing, the accuracy of psychological state recognition is improved, making the assessment results more comprehensive.

[0028] According to some embodiments, it further includes: Establishing a mapping between physiological information and doctor's questions as the first labeled data to train the psychological assessment large model, where the doctor's questions are the doctor's assessment questions given based on the psychological information; Establishing a mapping between physiological information, doctor's assessment questions, and the training content of the power employee's feedback based on the doctor's assessment questions and the doctor's assessment results as the second labeled data to train the psychological assessment large model.

[0029] It is understandable that for the collection of multi-modal psychological assessment and intervention data, the core task of this part is to collect real psychological assessment and intervention data, including the physiological signals of employees, psychological-related questions designed based on physiological signals, the responses of subjects, psychological assessment results, and intervention measures. The scenarios for data collection include medical clinics, psychological counseling, and other environments. The specific data composition includes: Physiological signal data: such as heart rate, blood pressure, respiration, blood oxygen saturation, etc.; Personalized questions and answers: record the questions and answers between the subject and the psychologist, especially psychological-related questions; Psychological assessment results: such as the stress situation of the subject and its main incentives; Intervention measures: suggestions given by the doctor based on the psychological assessment results. Based on these data, two types of labeled data sets can be constructed: The associated data set of physiological signals and personalized questions: The label of this data set is the questions proposed by the doctor; The associated data set of physiological signals, doctor's questions, and subject's answers: The label of this data set is the psychological assessment results and intervention measures given by the doctor. These data will provide a basis for the training of multi-modal large models.

[0030] Exemplarily, first, a mapping between physiological information and doctor's questions can be established to construct the first labeled data set. At this stage, the doctor provides the psychological assessment questions that he usually asks based on the physiological data of the employee (such as HRV, EDA, heart rate, blood pressure, etc.). For example, when the doctor sees a decrease in HRV and an increase in EDA, he may ask "Have you had any anxiety or sleep problems recently?" This process constructs a mapping between physiological signals and doctor's questions, enabling the large model to learn the assessment logic of the doctor under different physiological data patterns, so as to generate more practical assessment questions in actual applications.

[0031] Exemplarily, second, a mapping between physiological information, doctor's assessment questions, employee feedback, and doctor's assessment results can be established to construct the second labeled data set. At this stage, the doctor not only provides assessment questions, but also conducts a psychological assessment based on the employee's feedback information and gives a final diagnosis (such as mild anxiety, moderate stress, severe fatigue, etc.). The system establishes a mapping relationship between physiological data, questions proposed by the doctor, the employee's answers, and the doctor's final assessment results, enabling the large model to learn the complete psychological assessment process, the full-link reasoning from physiological signals, question feedback to the final psychological assessment results. This process greatly enhances the assessment accuracy of the model, enabling it to more accurately identify the psychological state of employees in actual applications.

[0032] In some examples, it also includes: Obtaining the model output quality ranking information of the multi-group output result information generated by the multi-modal psychological assessment large model during an assessment process, where each group of output result information includes assessment questions and / or assessment results; Optimize and train the multi-modal psychological assessment large model through an instruction fine-tuning method that compares prompt words based on the quality ranking information output by the model.

[0033] It can be understood that for the training of the multi-modal large model, in this part, the multi-modal large model can be trained according to the collected data. The model training will be divided into three stages: The first stage: the mapping of physiological signals and doctor's questions. In this stage, we will use the physiological signal data and label data of the subjects for training. The task of the model is to generate relevant psychological assessment questions based on the physiological signals of the subjects. Through the training of this stage, the model will be able to automatically generate questions to be asked according to real-time physiological signals (such as abnormal changes in physiological indicators such as heart rate, blood pressure, and respiration). The goal is to enable the model to identify which changes in physiological signals may be related to mental health problems and generate appropriate questions to further evaluate the mental state of employees. The second stage: the mapping from physiological signals and Q&A data to assessment results and interventions. In the training of this stage, the input data of the model includes not only the physiological signals of the subjects and the questions raised by the doctor, but also the answers of the subjects. The output label is the doctor's psychological assessment result and intervention measures. Through the training of this stage, the model will be able to synthesize physiological signals, question content, and the answers of the subjects to generate personalized psychological assessment results for each employee and provide corresponding intervention suggestions. The training goal of this stage is to enable the model to understand complex multi-modal inputs and provide accurate mental health assessments and customized intervention measures for employees based on these inputs. Expert feedback optimization and comparison prompt word fine-tuning. In order to further improve the quality of the model in psychological assessment and intervention, an expert feedback mechanism is introduced on the basis of the second stage. In this stage, the model will generate multiple assessment and intervention suggestions for the same input data. However, due to the differences in the quality of the suggestions generated by the model, we need to use the professional knowledge of psychological experts or psychiatrists to evaluate these outputs to determine which suggestions are more in line with actual needs, theoretical basis, and expert judgment. Specifically, experts will evaluate and rank multiple personalized questions, psychological assessment results, and intervention suggestions generated by the model based on the physiological signals of the same person. The purpose of the expert ranking is to screen out those suggestions with high quality, high accuracy, and practical operability, and remove those outputs with poor effects or unreasonable ones. This process provides important guidance for subsequent fine-tuning.

[0034] In some examples, the optimizing and training of the multi-modal psychological assessment large model through the instruction fine-tuning method that compares prompt words based on the quality ranking information output by the model includes: Adding quality level labels to the sample data composed of the output result information of the multi-modal psychological assessment large model in an evaluation process and the associated model input information based on the quality ranking information output by the model to generate template sample data that conforms to at least two quality levels, where the output result information further includes intervention recommendation information; Optimally training the multi-modal psychological assessment large model by comparing the template sample data that conforms to at least two quality levels of the prompt templates.

[0035] Exemplarily, based on the expert's ranking feedback, an instruction fine-tuning method for contrastive prompting words can be adopted. The specific operations of the method are as follows: Definition of high-quality and low-quality labels: Through expert feedback, "high-quality" and "low-quality" labels are defined, corresponding to the output content that experts consider to be of high quality and unreasonable, respectively. For each set of inputs (i.e., the physiological signals of the subjects), two prompting word templates will be generated: High-quality prompting word template: This template includes personalized questions, evaluation results, and intervention suggestions that experts consider to be of high quality. The example template is as follows: "For the psychological signal xxx, a set of high-quality personalized psychological questions is xxx, a reasonable evaluation result is xxx, and a reasonable intervention suggestion is xxx."; Low-quality prompting word template: This template includes personalized questions, evaluation results, and intervention suggestions that experts consider to be of low quality. The example template is as follows: "For the psychological signal xxx, a set of low-quality personalized psychological questions is xxx, an unreasonable evaluation result is xxx, and an unreasonable intervention suggestion is xxx." Among them, during the fine-tuning process, the model will receive these contrastive prompting word templates for learning and optimization. By comparing the high-quality and low-quality outputs, the model can learn which features and strategies are effective, and which outputs may lead to misleading or interventions that do not meet the actual needs of employees. Thus, through expert annotation and quality ranking information, the reasoning ability of the large model in psychological assessment is optimized, making the questions, evaluation results, and intervention suggestions generated by it more in line with the actual situation of employees, and improving the credibility of psychological assessment. Compared with the model that solely relies on data training, this method introduces expert quality ranking and contrastive learning, enabling the model to effectively identify low-quality outputs and optimize towards high quality, thereby reducing ineffective or misleading psychological assessments and intervention suggestions. The psychological states and working environments of different employees are different, and traditional fixed questionnaires are difficult to adapt to individual differences. This method enables the model to adapt to the personalized needs of different employees through dynamic prompting word templates, improving the applicability of psychological assessment. Since the system optimizes the rationality of intervention suggestions through expert annotation, the finally generated intervention measures are more in line with psychological and medical principles, improving the actual effect of employees' mental health management. By introducing quality level labels and contrastive learning, the model can not only be applicable to the current psychological assessment needs of power employees, but also be extended to other high-pressure industries (such as medical, aviation) in the future for psychological assessment and intervention, improving the practical value of the large model.

[0036] In some examples, the physiological information is obtained through a wearable intelligent terminal device, and the method further includes: Obtaining the motion information of power employees through the wearable intelligent terminal device; Trigger the acquisition instruction of the physiological information when the action information matches the target production operation scenario indicated by the work order information, where the target production operation scenario includes high-voltage live working scenarios, power failure emergency repair scenarios, cable tunnel confined space inspection scenarios, and power dispatching center duty scenarios.

[0037] It can be understood that by using wearable intelligent terminal devices to collect the physiological information and action information of power workers in real time and combining with work order information to intelligently trigger the psychological assessment process, compared with the traditional psychological assessment methods based on fixed time or manual triggering, this method can dynamically adjust the assessment timing according to the actual work scenarios of power workers, thereby improving the timeliness and pertinence of psychological assessment. By combining physiological information, action information, work order information with the multi-modal psychological assessment large model, it is possible to more accurately identify the psychological states of workers under special working conditions such as high-voltage operation, emergency accident handling, and confined space inspection, and provide precise intervention suggestions.

[0038] Exemplarily, in this method, the intelligent terminal devices worn by power workers (such as smart watches, smart helmets or smart protective clothing) can monitor their physiological information in real time, including but not limited to heart rate (HR), heart rate variability (HRV), blood oxygen saturation (SpO2), skin electrical activity (EDA), respiratory rate, body temperature, etc. These physiological indicators can reflect the psychological stress level of workers during work. For example: an increase in heart rate and a decrease in HRV may indicate that the worker is in a tense state; a significant increase in EDA may indicate that the worker is experiencing an acute stress response; a decrease in blood oxygen level may indicate the risk of hypoxia when the worker is working in a confined space. To avoid interfering with the normal work of workers, this method does not perform physiological data collection in all cases, but combines work order information and action information to ensure that physiological data collection occurs in specific work scenarios.

[0039] Exemplarily, the intelligent terminal devices are also equipped with motion sensors (such as accelerometers, gyroscopes, GPS modules) to detect the activity status of workers. For example: if the smart helmet worn by a worker detects that the head posture remains at a low angle for a long time, and at the same time the smart watch detects small-amplitude hand shaking, it may indicate that the worker is performing fine operations, such as high-voltage live working; if the GPS position matches the substation maintenance area, and at the same time the intelligent terminal records large-amplitude arm movements, it may indicate that the worker is performing equipment maintenance or high-voltage line connection operations; if the worker is inside the cable tunnel, the environmental sensor detects a low oxygen level, and at the same time the physiological data indicates a significant decrease in HRV and an increase in EDA, it may indicate that the worker is experiencing anxiety or hypoxia discomfort in the confined environment. These action information can be used to judge the actual operation scenario of the worker and ensure that the psychological assessment is triggered at the appropriate time point without interfering with the normal work process.

[0040] Exemplarily, the collection of physiological information is not manually triggered by the employee, but automatically triggered by the matching of work order information and action information. The work order information includes the employee's work tasks, working environment, work order execution time, etc. For example, an employee's work order task is "perform high-voltage live working at the 110 kV substation from 10:00 to 12:00". The action information reflects the employee's actual behavior. If the intelligent terminal detects that the current action characteristics of this employee conform to the typical mode of high-voltage live working (such as the arm fixed at a certain angle, slight tremors of the hand, and stable head posture), then the physiological data collection instruction is triggered. In this way, it can be ensured that the collection of physiological information occurs in key work scenarios, such as: High-voltage live working scenario: When the employee is operating high-voltage electrical equipment, they are highly concentrated and under great psychological pressure, which may affect the operation stability; Power failure repair scenario: In case of an emergency that needs to be handled urgently, the employee's psychological pressure surges, which may affect the judgment ability and emergency response; Cable tunnel confined space inspection scenario: Long-term stay in a closed environment may cause anxiety or physical discomfort, affecting work safety; Power dispatching center duty scenario: Long-term duty and high-intensity decision-making work may lead to the accumulation of chronic stress, affecting the mental health of employees. After collecting the employee's physiological information, the system will input it into the multi-modal psychological assessment large model. The large model will automatically generate personalized psychological assessment questions based on physiological data, work order information, the employee's past psychological assessment records, etc. For example: For high-voltage operation personnel, if the HRV decreases and the EDA increases, it may ask: "During the current task, do you feel strong psychological pressure? Does this affect your operation stability?"; For fault repair personnel, if the heart rate increases significantly, it may ask: "When dealing with emergencies, do you feel nervous and do you need to adjust the task assignment?"; For confined space inspection personnel, if a decrease in blood oxygen and an increase in EDA are detected, it may ask: "During the inspection process, have you experienced short-term chest tightness or anxiety?" After analyzing the employee's physiological data and response feedback, the large model will generate a psychological assessment result and provide personalized intervention suggestions based on the principles of psychology and behavioral science. For example: For employees with a short-term increase in stress, it is recommended to carry out relaxation training (such as deep breathing or meditation) and reasonably arrange work and rest times; For employees working in a long-term high-pressure environment, it is recommended to rotate positions regularly, reduce consecutive high-pressure tasks, or arrange for mental health counseling; For employees who experience hypoxia or anxiety during the inspection, it is recommended to shorten the inspection time, increase the ventilation frequency, or provide psychological counseling resources.

[0041] It is understandable that by combining work order information and action information to trigger physiological information collection, it is ensured that psychological assessment occurs in key work scenarios rather than random collection, thereby enhancing the pertinence and timeliness of the assessment. Compared with the traditional method of triggering psychological assessment at fixed times or manually, this method uses an intelligent triggering mechanism to ensure the assessment quality without affecting the normal work process of employees. The large model can generate targeted assessment questions based on the physiological state and working environment of employees instead of using a fixed questionnaire, thus improving the accuracy of the assessment and the user experience. By combining psychological theories to provide personalized intervention suggestions, it helps employees better cope with work pressure, reduce mental health risks, and improve work safety and efficiency.

[0042] In some examples, it further includes: When receiving the collection instruction of the physiological information, based on the action information, evaluate the influence coefficient of the action information on the physiological information of the power employee as the action influence coefficient; Based on the influence coefficient, filter the physiological information to predict the influence coefficient of the target generated working scenario pressure on the physiological information of the power employee as the pressure influence coefficient, and use the physiological information and the associated pressure influence coefficient together as the target physiological information of the power employee; When conducting a psychological assessment on the power employee, generate assessment questions through the multi-modal psychological assessment large model based on the target physiological information of the power employee; Generate an assessment result through the multi-modal psychological assessment large model based on the target physiological information of the power employee, the assessment questions, and the feedback content of the power employee for the assessment questions.

[0043] Exemplarily, use wearable intelligent terminal devices to collect the physiological information and action information of power employees in real time, and combine work order information to intelligently trigger the psychological assessment process. Compared with the traditional psychological assessment method based on fixed time or manual triggering, this method can dynamically adjust the assessment timing according to the actual work scenario of power employees, thereby improving the real-time and pertinence of psychological assessment. By combining physiological information, action information, work order information with the multi-modal psychological assessment large model, it is possible to more accurately identify the psychological state of employees under special working conditions such as high-voltage operation, emergency accident handling, and confined space inspection, and provide precise intervention suggestions.

[0044] Exemplarily, the intelligent terminal devices worn by power workers (such as smart watches, smart helmets or smart protective clothing) can monitor their physiological information in real time, including but not limited to heart rate (HR), heart rate variability (HRV), blood oxygen saturation (SpO2), skin electrical activity (EDA), respiratory rate, body temperature, etc. These physiological indicators can reflect the psychological stress level of workers during work. For example: an increase in heart rate and a decrease in HRV may indicate that the worker is in a tense state; a significant increase in EDA may indicate that the worker is experiencing an acute stress response; a decrease in blood oxygen level may suggest that the worker has an oxygen deficiency risk when working in a confined space. To avoid interfering with the normal work of workers, this method does not collect physiological data in all cases, but combines work order information and motion information to ensure that physiological data collection occurs in specific work scenarios.

[0045] Exemplarily, the intelligent terminal devices are also equipped with motion sensors (such as accelerometers, gyroscopes, GPS modules) to detect the activity status of workers. For example: if the smart helmet worn by a worker detects that the head posture remains at a low angle for a long time, and at the same time the smart watch detects small hand movements, it may indicate that the worker is performing fine operations, such as live high-voltage work. If the GPS location matches the substation maintenance area, and at the same time the intelligent terminal records large arm movements, it may indicate that the worker is performing equipment maintenance or high-voltage line connection work. If the worker is inside a cable tunnel, the environmental sensor detects a low oxygen level, and at the same time the physiological data indicates a significant decrease in HRV and an increase in EDA, it may indicate that the worker is experiencing anxiety or hypoxia discomfort in a confined environment. These motion information can be used to judge the actual work scenario of the worker and ensure that the psychological assessment is triggered at an appropriate time point without interfering with the normal work process.

[0046] Exemplarily, the collection of physiological information is not manually triggered by employees, but automatically triggered by the matching of work order information and action information. The work order information includes the work tasks of employees, the working environment, the work order execution time, etc. For example, the work order task of an employee is "conduct high-voltage live working at the 110 kV substation from 10:00 to 12:00". The action information reflects the actual behavior of the employee. If the intelligent terminal detects that the current action characteristics of this employee conform to the typical mode of high-voltage live working (such as the arm fixed at a certain angle, slight tremors of the hand, and stable head posture), then a physiological data collection instruction is triggered. In this way, it can be ensured that the collection of physiological information occurs in key work scenarios. For example, in the high-voltage live working scenario, when employees are operating high-voltage electrical equipment, they are highly concentrated and under great psychological pressure, which may affect the operation stability; in the power failure emergency repair scenario, sudden accidents need to be handled urgently, and the psychological pressure of employees surges, which may affect the judgment ability and emergency response; in the cable tunnel confined space inspection scenario, long-term stay in a closed environment may lead to anxiety or physical discomfort, affecting operation safety; in the power dispatching center duty scenario, long-term duty and high-intensity decision-making work may lead to the accumulation of chronic stress, affecting the mental health of employees. After collecting the physiological information of employees, the system will input it into the multi-modal psychological assessment large model. The large model will automatically generate personalized psychological assessment questions based on physiological data, work order information, the past psychological assessment records of employees, etc. For example, for high-voltage operation personnel, if the HRV decreases and the EDA increases, it may ask: "During the current task, do you feel strong psychological pressure? Does this affect your operation stability?"; for emergency repair personnel, if the heart rate increases significantly, it may ask: "When dealing with sudden accidents, do you feel nervous and do you need to adjust the task assignment?"; for confined space inspection personnel, if a decrease in blood oxygen and an increase in EDA are detected, it may ask: "During the inspection process, have you experienced short-term chest tightness or anxiety?" After analyzing the physiological data of employees and the response feedback, the large model will generate psychological assessment results and provide personalized intervention suggestions based on the principles of psychology and behavioral science. For example, for employees with a short-term increase in stress, it is recommended to carry out relaxation training (such as deep breathing or meditation) and reasonably arrange work and rest time; for employees working in a long-term high-pressure environment, it is recommended to rotate positions regularly, reduce continuous high-pressure tasks, or arrange mental health counseling; for employees who experience hypoxia or anxiety during the inspection process, it is recommended to shorten the inspection time, increase the ventilation frequency, or provide psychological counseling resources. Thus, the accuracy and timeliness of psychological assessment are improved: by combining work order information and action information to trigger the collection of physiological information, ensuring that psychological assessment occurs in key work scenarios rather than random collection, thereby enhancing the pertinence and timeliness of the assessment. Compared with the traditional method of triggering psychological assessment at regular intervals or manually, this method, through an intelligent triggering mechanism, ensures the assessment quality without affecting the normal operation process of employees.Large models can generate targeted assessment questions based on the physiological state of employees and the working environment, rather than using a fixed questionnaire, thus improving the accuracy of assessment and the user experience. Provide personalized intervention suggestions in combination with psychological theories to help employees better cope with work pressure, reduce mental health risks, and improve work safety and efficiency.

[0047] In some examples, it further includes: Obtain coordinating employees who handle the same work order and are in the same production operation scenario as the power employees; Horizontally compare the stress impact coefficients of the power employees and all the coordinating employees to classify the stress impact coefficients; Based on the physiological information, the associated stress impact coefficient, and the classification information together as the target physiological information of the power employees.

[0048] It can be understood that since it is a horizontal comparison of employees with the same scenario influencing factors, the differences in task risk difficulty, task type, and environmental factors they face can be not considered. The differences in the horizontal comparison mainly come from the differences in the sensitivity of employees to stress. Focus on the physiological response comparison of different employees in the same scenario, so as to more accurately evaluate the individual psychological stress level and provide personalized psychological intervention suggestions.

[0049] Exemplarily, for n employees in the same operation scenario: Among them, Represents the average stress impact coefficient of all employees in this scenario, as the reference value.

[0050] And the stress sensitivity PS is used to measure the sensitivity of an employee's physiological response to stress compared to others in the same operation scenario: Among them: Represents the stress sensitivity of employee i; Is the standard deviation of the stress impact coefficient.

[0051] The Z-score calculated by this formula represents the deviation degree of the employee's stress impact coefficient compared to the average level. Based on the stress sensitivity PS, employees can be divided into different levels to determine their response differences to the same scenario stress. For example, low sensitivity (L1): (insensitive to pressure); medium sensitivity (L2): normal pressure response; high sensitivity (L3): highly sensitive to pressure. Thus, by making a horizontal comparison, the sensitivity of individual employees to pressure can be identified, improving the individual accuracy of psychological assessment. It can optimize the target physiological information, ensure more accurate psychological intervention measures, and improve the scientific nature of the mental health management of power employees.

[0052] Please refer to Figure 2 , an embodiment of the mental assessment management device for power employees in the embodiment of the present application may include: An acquisition unit 21, configured to acquire the physiological information of the power employee, and the acquisition instruction of the physiological information is triggered based on the work order information of the power employee, so as to determine the acquisition time range and acquisition type of the physiological information based on the work order information; A generation unit 22, configured to generate assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power employee when performing a psychological assessment on the power employee; An assessment unit 23, configured to generate an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power employee, the assessment questions, and the feedback content of the power employee for the assessment questions.

[0053] In summary, the mental assessment management device for power employees provided in the embodiment of the present application generates assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power employee; and generates an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power employee, the assessment questions, and the feedback content of the power employee for the assessment questions. Traditional psychological assessment methods usually rely on subjective questionnaires, while this method triggers physiological data collection through work orders, making psychological assessment based on objective physiological data, reducing the influence of human subjective factors on the assessment result. In addition, generating personalized assessment questions based on physiological data fluctuations makes the psychological test more accurate and can more accurately reflect the actual mental state of employees. This method comprehensively utilizes physiological data, work order information, and natural language feedback information, and performs in-depth analysis through a multi-modal psychological assessment large model, overcoming the limitations of single physiological signal or questionnaire assessment. By combining time series data analysis and natural language processing, the accuracy of mental state recognition is improved, making the assessment result more comprehensive.

[0054] As Figure 3 shown, the embodiment of the present application further provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored on the memory 320 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of any of the above methods for mental assessment management of power employees: Obtain the physiological information of power workers, and the acquisition instruction of the physiological information is triggered based on the work order information of the power workers, so as to determine the acquisition time range and acquisition type of the physiological information based on the work order information; In the case of conducting a psychological assessment on the power worker, generate assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power worker; Generate an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power worker, the assessment questions, and the feedback content of the power worker for the assessment questions.

[0055] Since the electronic device introduced in this embodiment is the device adopted for a power worker psychological assessment management device in an embodiment of the present application, based on the method introduced in the embodiment of the present application, those skilled in the art can understand the specific implementation manner and various variation forms of the electronic device in this embodiment. Therefore, the specific implementation of how this electronic device implements the method in the embodiment of the present application will not be introduced in detail here. As long as the device adopted by those skilled in the art to implement the method in the embodiment of the present application belongs to the scope protected by the present application.

[0056] In the specific implementation process, when the computer program 311 is executed by the processor, it can implement Figure 1 Any one of the corresponding embodiments: Obtain the physiological information of power workers, and the acquisition instruction of the physiological information is triggered based on the work order information of the power workers, so as to determine the acquisition time range and acquisition type of the physiological information based on the work order information; In the case of conducting a psychological assessment on the power worker, generate assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power worker; Generate an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power worker, the assessment questions, and the feedback content of the power worker for the assessment questions.

[0057] It should be noted that in the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0058] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0059] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 block or multiple blocks.

[0060] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 block or multiple blocks.

[0061] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 block or multiple blocks.

[0062] Embodiments of the present application also provide a computer program product, which includes computer software instructions. When the computer software instructions run on a processing device, the processing device is caused to execute as Figure 1 the process of psychological assessment management of power industry employees in the corresponding embodiment.

[0063] A computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present application are all or partially generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from a website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or a wireless manner (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be stored by a computer or a data storage device such as a server or a data center that includes one or more integrated available media. The available medium may be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)), etc.

[0064] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0065] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in an electrical, mechanical, or other form.

[0066] The units described as separate components may or may not be physically separated, and the components displayed 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 units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0067] In addition, in each embodiment of the present application, each functional unit may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0068] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, may be embodied in the form of a software product. The 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 in the various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0069] The above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application 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 recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A psychological assessment and management method for power industry employees, characterized in that, Including: Obtain the physiological information of power employees, where the acquisition instruction of the physiological information is triggered based on the work order information of the power employees, so as to determine the acquisition time range and acquisition type of the physiological information based on the work order information; In the case of conducting a psychological assessment on the power employee, generate assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power employee; Generate an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power employee, the assessment questions, and the feedback content of the power employee regarding the assessment questions.

2. The method according to claim 1, wherein It also includes: Establish a mapping between physiological information and doctor's questions as the first labeled data to train the psychological assessment large model, where the doctor's questions are the doctor's assessment questions given based on the psychological information; Establish a mapping between physiological information, doctor's assessment questions, and the training content of the power employee's feedback based on the doctor's assessment questions and the doctor's assessment results as the second labeled data to train the psychological assessment large model.

3. The method according to claim 1, wherein It also includes: Obtain the model output quality ranking information of the expert for multiple groups of output result information generated by the multi-modal psychological assessment large model during an assessment process, where each group of output result information includes assessment questions and / or assessment results; Optimize and train the multi-modal psychological assessment large model through the instruction fine-tuning method of comparing prompt words based on the model output quality ranking information.

4. The method according to claim 3, wherein The optimizing and training of the multi-modal psychological assessment large model through the instruction fine-tuning method of comparing prompt words based on the model output quality ranking information includes: Based on the model output quality ranking information, add quality level labels to the sample data composed of the output result information and the associated model input information of the multi-modal psychological assessment large model during an assessment process to generate template sample data that conforms to the prompt word templates of at least two quality levels, and the output result information also includes intervention suggestion information; Optimize and train the multi-modal psychological assessment large model through the comparison between the template sample data that conforms to the prompt word templates of at least two quality levels.

5. The method according to any one of claims 1 to 4, characterized in that, The physiological information is obtained through a wearable intelligent terminal device, and the method also includes: Obtain the action information of power employees through the wearable intelligent terminal device; Trigger the acquisition instruction of the physiological information when the action information matches the target production operation scenario indicated by the work order information, and the target production operation scenario includes high-voltage live working scenarios, power failure repair scenarios, cable tunnel confined space inspection scenarios, and power dispatching center on-duty scenarios.

6. The method according to claim 5, characterized in that, It also includes: In the case of receiving the acquisition instruction of the physiological information, evaluate the influence coefficient of the action information on the physiological information of the power employee as the action influence coefficient based on the action information; Filter the physiological information based on the influence coefficient to predict the influence coefficient of the target production operation scenario pressure on the physiological information of the power employee as the pressure influence coefficient, so as to use the physiological information and the associated pressure influence coefficient together as the target physiological information of the power employee. In the case of conducting a psychological assessment on the power worker, assessment questions are generated through a multi-modal psychological assessment large model based on the target physiological information of the power worker; Based on the target physiological information of the power worker, the assessment questions, and the feedback content of the power worker regarding the assessment questions, an assessment result is generated through the multi-modal psychological assessment large model.

7. The method according to claim 6, wherein It further includes: Obtain coordinated operation workers who handle the same work order and are in the same production operation scenario as the power worker; Horizontally compare the stress influence coefficients of the power worker and all the coordinated operation workers to classify the stress influence coefficients; Based on the physiological information, the associated stress influence coefficient, and the classification information together as the target physiological information of the power worker.

8. A psychological assessment and management device for power industry employees, characterized in that, It includes: An acquisition unit for acquiring the physiological information of a power worker, and the acquisition instruction of the physiological information is triggered based on the work order information of the power worker to determine the acquisition time range and acquisition type of the physiological information based on the work order information; A generation unit for generating assessment questions through a multi-modal psychological assessment large model based on the physiological information of the power worker in the case of conducting a psychological assessment on the power worker; An assessment unit for generating an assessment result through the multi-modal psychological assessment large model based on the physiological information of the power worker, the assessment questions, and the feedback content of the power worker regarding the assessment questions.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to implement the steps of the power worker psychological assessment management method according to any one of claims 1-7 when executing the computer program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program, when executed by the processor, implements the power worker psychological assessment management method according to any one of claims 1-7.

Images