An augmented reality assisted assembly system considering human factors

Abstract

The present invention discloses an augmented reality assisted assembly system that takes human factors into consideration, which belongs to the technical field of process assembly and includes a scene model and assembly model import module, a real-time assisted assembly module, and a human physiological state monitoring and evaluation module. After collecting human physiological parameters through the human physiological state monitoring module, the present invention combines the current posture of the human body and the task parameters, and imports them into the evaluation model for automatic analysis, and finally gives the current physiological and psychological state of the assembler, and gives prompts in time, fully improving the assembly load of the assembler and reducing human assembly errors, which is worthy of being promoted and used.

Description

Technical Field

[0001] The present invention relates to the technical field of process assembly, and in particular to an augmented reality assisted assembly system taking human factors into consideration. Background Art

[0002] As the demand for manufacturing industry continues to develop from fixed and single function to flexible and multi-functional, in terms of product design, the product structure is more complex, the number of parts is increasing, and the assembly quality is constantly improving; in terms of product manufacturing, it has shifted from the past single variety, large batch and planned production to multi-variety, small batch and customized production. These two reasons make it impossible or difficult for manufacturing companies to adopt costly fully automated assembly production lines, so highly flexible manual assembly is still needed to cope with complex product structure design and lengthy assembly processes. At present, mainstream assembly instructions mainly rely on textual guidance documents such as assembly instruction manuals. The assembly instructions are not intuitive enough and the actual guidance effect is greatly reduced. At the same time, if there is a need for adjustment in the assembly process, all relevant guidance documents need to be re-made, and the scalability of assembly instructions is poor. Under the diversified products and short development cycles, the cognitive ability of on-site assemblers is heavily burdened, and the assembly pressure increases, which seriously hinders the improvement of the quality and efficiency of manual assembly.

[0003] In order to improve the quality and efficiency of manual assembly, human-computer interaction technology based on virtual reality and augmented reality has gradually been widely used in the field of auxiliary assembly due to its convenient implementation and flexible guidance characteristics. Among them, the auxiliary assembly system based on virtual reality is mostly used for assembly training before on-site assembly, increasing the proficiency of on-site assemblers in the assembly process, and timely improving and adjusting the existing assembly process; the auxiliary assembly system based on augmented reality is mainly used for on-site assembly guidance, reducing the cognitive load of assemblers, and improving assembly efficiency and quality.

[0004] However, in the current assisted assembly system based on augmented reality technology, although the psychological cognitive load of assemblers has been significantly reduced, for complex equipment, the characteristics of multiple assembly processes, large assembly workload, and high repetitiveness still put great physiological pressure on assemblers. Especially when the assembly space is insufficient and the parts structure is complex and dense, assemblers may need to assemble for a long time in bad postures such as bending, kneeling, and crawling. Working in such an assembly environment and pressure for a long time, the number of human errors increases rapidly, and the assembly quality and efficiency still need to be improved, which will cause long-term damage to the assemblers' bodies. Therefore, it is also necessary to add monitoring and evaluation of the physiological state of the human body in the augmented reality assisted assembly system, so as to prevent human errors and further improve the assembly quality and efficiency while ensuring the physiological and psychological comfort of the assemblers as much as possible. Therefore, an augmented reality assisted assembly system considering human factors is proposed. Summary of the invention

[0005] The technical problem to be solved by the present invention is: how to solve the problem that in the current assisted assembly system based on augmented reality technology, although the psychological cognitive load of assemblers is significantly reduced, for complex equipment, the characteristics of many assembly processes, large assembly workload and high repetitiveness still put great physiological pressure on assemblers, especially when the assembly space is insufficient and the parts structure is complex and dense, assemblers may need to assemble for a long time in bad postures such as bending, kneeling and crawling. Working in such an assembly environment and pressure for a long time will rapidly increase the number of human errors, and the assembly quality and efficiency still need to be improved, which will cause long-term damage to the assemblers' bodies. An augmented reality assisted assembly system taking human factors into consideration is provided. The system can detect the physiological state of the human body during the assisted assembly process, evaluate its physiological state from the human factors perspective, and then make timely adjustments to the specific assembly rhythm.

[0006] The present invention solves the above technical problems through the following technical solutions, which include a scene model and assembly model import module, a real-time auxiliary assembly module, and a human physiological state monitoring and evaluation module;

[0007] The scene model and assembly model import module is used to import and process three-dimensional models, extract assembly information and set multi-mode guidance information;

[0008] The real-time auxiliary assembly module is used to perform object visual processing and human-computer interaction to provide real-time guidance to the assembler;

[0009] The human physiological state monitoring and evaluation module is used to collect human parameters, collect human postures, set data synchronization and evaluate human state, obtain the current physiological and psychological state of the assembler, and give prompts in time.

[0010] Furthermore, the scene model and assembly model import module includes a three-dimensional model processing submodule, an assembly information extraction submodule, and a multi-mode guidance information submodule; the three-dimensional model processing submodule is used to provide the system with an import interface for the environment model, tool model and assembly model, and to perform lightweight settings on the data after importing the model; the assembly information extraction submodule is used to extract the assembly relationship, hierarchy, part information and assembly path, and generate an assembly model sequence and a model tree; the multi-mode guidance information submodule is used to annotate the assembly model tree and model sequence obtained by assembly information extraction.

[0011] Furthermore, the annotation content of the multi-mode guidance information submodule includes voice prompt entry of operation steps, tool information annotation and equipment name annotation.

[0012] Furthermore, the real-time auxiliary assembly module includes an object vision processing submodule and a human-computer interaction submodule; the object vision processing submodule is used for three-dimensional scene reconstruction and target recognition and tracking; the human-computer interaction submodule is used for augmented reality scene display and personnel manipulation feedback.

[0013] Furthermore, the object vision processing submodule includes a three-dimensional scene reconstruction unit and a target recognition and tracking unit; the three-dimensional scene reconstruction unit is used to realize multi-directional three-dimensional scene reconstruction as the environmental input of the real-time auxiliary assembly module; the target recognition and tracking unit is realized based on augmented reality glasses, and adopts simultaneous positioning and map construction technology. When the assembler moves continuously in the indoor environment, a series of continuous depth maps at different angles are obtained through the augmented reality glasses, and different depth maps are accumulated and analyzed continuously, and the analysis and integration are calculated to establish an accurate environmental model.

[0014] Furthermore, the human-computer interaction submodule includes an augmented reality scene display unit and an assembler manipulation feedback unit; the augmented reality scene display unit is used to import the scene model and the tooling tools in the assembly model import module, and solve the registered posture of the virtual guide tool such as indicator arrows, highlight symbols, etc., integrate it with the real scene, and import it into the augmented reality glasses; the assembler manipulation feedback unit is used to prompt the assembler's current assembly status and actions through the system's interference detection function and human body state assessment function.

[0015] Furthermore, the human physiological state monitoring and evaluation module includes a human parameter acquisition submodule, a human posture acquisition submodule, a human state evaluation submodule and a data synchronization setting submodule; the human parameter acquisition submodule is used for acquisition initialization, data processing and conversion, and data display; the human posture acquisition submodule is used for equipment connection, posture calibration, and motion capture; the human state evaluation submodule is used for comfort analysis, fatigue analysis and lumbar force analysis; the data synchronization setting submodule is used to synchronize the data of the human parameter acquisition submodule and the human posture acquisition submodule.

[0016] Furthermore, the human body parameter acquisition submodule includes an acquisition initialization unit, a data processing and conversion unit, and a data display unit; the acquisition initialization unit is used to install the sensor at the position where measurement is required; the data processing and conversion unit is used to perform eye tracking analysis, HRV heart rate variability analysis, EDA galvanic skin response analysis, EMG electromyography analysis, and RESP respiratory analysis; the data display unit is used to display the data acquired by the data processing and conversion unit, and illustrate the current state of the human body.

[0017] Furthermore, when conducting eye tracking analysis, an eye tracker is used to measure eye movement data signals, measure and analyze eye-related information and data, process and analyze eye movement data signals, and present statistical characteristics of eye movement data changing over time; when conducting HRV heart rate variability, an electrocardiograph is used to process and analyze PPG signals offline; when conducting EDA galvanic skin response, data is collected through an galvanic skin acquisition device, and then SC skin conductance data analysis, SCR event-related galvanic skin analysis, and SCL phase-related galvanic skin analysis are performed; when conducting EMG electromyography analysis, data is collected through an electromyography acquisition device to obtain statistical characteristics of the electromyography signal in the time domain, and then periodic force data is obtained to realize electromyography analysis; when conducting RESP breathing analysis, data is collected through a respiratory information acquisition device, and the respiratory behavior within a set time period is analyzed in a full-process and staged data information processing manner, and the time domain and frequency domain data reports of the respiratory signal are statistically analyzed and exported.

[0018] Furthermore, the human posture acquisition submodule includes a device connection unit, a posture calibration unit, and a motion capture unit; the device connection unit is used to turn on sensors, place sensors, connect sensors, and wear sensors; the posture calibration unit is used to perform standing posture, half-squatting posture, and unfolded posture calibration; the motion capture unit is used to set scenes, contact points, contact constraint tolerance, heading angles, pitch angle parameters, and perform motion recording.

[0019] Furthermore, the human body state assessment submodule includes a comfort analysis unit, a fatigue analysis unit and a lumbar vertebrae force analysis unit; the comfort analysis unit, fatigue analysis unit and lumbar vertebrae force analysis unit are used to assess the human body's comfort, fatigue and lumbar vertebrae force conditions respectively.

[0020] Compared with the prior art, the present invention has the following advantages: the augmented reality assisted assembly system that takes human factors into consideration collects human physiological parameters through the human physiological state monitoring module, combines the current human posture and work task parameters, and imports them into the evaluation model for automatic analysis, and finally gives the current physiological and psychological state of the assembler, and gives prompts in time, which fully improves the assembly load of the assembler and reduces human-induced assembly errors, and is worthy of popularization and use. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a schematic structural diagram of an augmented reality assisted assembly system taking human factors into consideration in an embodiment of the present invention;

[0022] Figure 2 It is a schematic diagram of the implementation process of the augmented reality assisted assembly system considering human factors in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0024] like Figure 1 As shown, this embodiment provides a technical solution: an augmented reality assisted assembly system considering human factors, including a scene model and assembly model import module, a real-time assisted assembly module, and a human physiological state monitoring and evaluation module;

[0025] The scene model and assembly model import module is used to provide functions such as 3D model import and processing, assembly information extraction and multi-mode guidance information setting;

[0026] Real-time auxiliary assembly module, which is used to provide object visual processing and human-computer interaction functions, and provide real-time guidance to assemblers;

[0027] The human physiological state monitoring and evaluation module is used to provide functions such as human parameter collection, human posture collection, data synchronization setting and human state evaluation. It can provide the current physiological and psychological state of the assembler and give prompts in time to fully improve the assembly load of the assembler and reduce human assembly errors.

[0028] like Figure 2 As shown, it is an implementation flow chart of the augmented reality assisted assembly system considering human factors in this embodiment.

[0029] In the scene model and assembly model import module, the 3D model import and processing functions are implemented based on the 3D model processing submodule, which provides the system with environment models, tool models and assembly models, and supports commonly used neutral CAD models, including Step, IGES, FBX and other 3D model import interfaces. After importing the model, the data can be lightweighted, including setting feature categories and size thresholds, and then automatically lightweighted. Lightweighting mainly uses the common Instancing instantiation algorithm to save only one copy of the same model and render other identical models repeatedly in real time.

[0030] The assembly information extraction function is implemented based on the assembly information extraction submodule, including the extraction of assembly relationships, hierarchical structures, part information and assembly paths. The assembly model input to this submodule is required to be a .CATProcess assembly model generated from the Replay dynamic assembly simulation method of the Delmia / DPM module. This submodule is developed by Dassault Systemes based on the CAA (Component Application Architecture, CAA) secondary development platform. A CATSession is created in the background and the information file containing the path of the assembly model is loaded. The document containing the path information to be read is obtained and loaded from the file, and a pointer to the document is given. The document pointer is used to obtain the path information of the corresponding assembly parts, and the assembly model sequence and model tree are generated for real-time assembly guidance.

[0031] The multi-mode guidance information setting function is implemented based on the multi-mode guidance information sub-module, and is annotated in the assembly model tree and model sequence obtained through assembly information extraction. The annotation content includes voice prompt entry of operation steps, tool information annotation and equipment name annotation, which are called and displayed as the auxiliary assembly process proceeds.

[0032] In the real-time assisted assembly module, the object vision processing function is implemented based on the object vision processing submodule, which mainly includes 3D scene reconstruction and target recognition and tracking. Among them, the 3D scene reconstruction function can realize multi-directional, high-precision and fast 3D scene reconstruction, which is the environmental input of the real-time assisted assembly module. This function consists of 8 industrial cameras forming a camera array system of 6 pairs of binocular vision reconstruction. First, each pair of binocular cameras is calibrated for binocular vision, and then the entire camera array is calibrated to obtain the internal parameters and rotation and translation parameters of the camera. It should be noted that the entire camera array system will not be allowed to move and adjust after calibration. Then, the dense matching algorithm parameters are set, and after obtaining the binocular disparity map, the depth map is obtained by combining the depth camera feature extraction function and using the binocular camera calibration parameters to further reconstruct the 3D point coordinates. Then, the 3D point cloud reconstructed from each perspective is roughly spliced ​​by the calibration parameters of the camera array, and optimized using the ICP (Iterative Closest Point) algorithm; finally, the color is fused into the point cloud to form a full-view 3D scene model.

[0033] The target recognition and tracking function is developed based on augmented reality glasses, including 4 environmental perception cameras, 1 depth camera, and 1 IMU (inertial measurement unit). Using simultaneous localization and mapping (SLAM) technology, when the assembler moves continuously in the indoor environment, the augmented reality glasses can obtain a series of continuous depth maps at different angles, continuously iterate, accumulate and analyze different depth maps, and analyze and integrate to calculate and establish an accurate environmental model. The main steps include: (1) converting the depth map format, converting the depth measurement unit into meters, and calculating the coordinate information of each vertex in the depth map and the direction of the surface; (2) calculating the camera position and orientation information in the world coordinate system, and using the iterative alignment algorithm to track the six degrees of freedom values ​​(the movement of the object along the three rectangular coordinate axes of x, y, and z in space and the rotation around these three coordinate axes) in real time to ensure that the system knows how much the augmented reality glasses have changed compared to the initial posture at the current moment; (3) continuously tracking the posture of the augmented reality glasses camera, and real-time fusion of the depth information of the known posture sequence frames into a single world three-dimensional MineCraft space. Therefore, as more new depth measurement information is gradually obtained, the grid elements of the space are gradually supplemented, the details are improved, and the model accuracy is more complete and accurate; (4) After the accurate environmental model is established, the recognition and tracking interference of the target model in the environmental model has been greatly reduced; the specific recognition and tracking can adopt a mature neural network algorithm - Faster Regional Convolutional Neural Network (Faster R-CNN), by using the real picture of the target for training in advance, to obtain the target recognition model, which can automatically extract and track the target from the video stream of the augmented reality glasses.

[0034] The human-computer interaction function is implemented based on the human-computer interaction submodule, including the augmented reality scene display function and the assembly personnel manipulation feedback function. The augmented reality scene display function imports the scene model and the tooling tools in the assembly model import module, and solves the registered posture of the virtual guidance tools such as the indicator arrows and highlight symbols, and then integrates them with the real scene and imports them into the augmented reality glasses. The assembly personnel manipulation feedback function, that is, through the interference inspection function and human body state evaluation function of the system, prompts the current assembly status and actions of the assembler, and the prompting methods include voice and text.

[0035] The human body parameter acquisition function is implemented based on the human body parameter acquisition submodule. The acquisition initialization function requires that the sensors be installed at the locations where they need to be measured, such as physiological parameter sensors installed at various parts of the human body, and environmental parameter sensors installed in the working environment. After the sensors are installed, they need to be connected to the computer interface so that the system can collect and process data; the data processing and conversion functions include eye tracking analysis, HRV heart rate variability analysis, EDA skin electrodermal response analysis, EMG electromyography analysis, and RESP respiratory analysis.

[0036] Among them, eye tracking analysis uses an eye tracker to measure eye tracking data, measure and analyze information and data related to the eyes. It mainly includes pupil diameter analysis (Pupil Analysis), blink recognition analysis (BlinkAnalysis), eye movement visualization analysis (Visualization), area of ​​interest analysis (Gaze AOI), eye movement data statistics, eye movement data guide block, eye movement statistics analysis report and other functions. It can process and analyze eye movement data signals. It can present the statistical characteristics of eye movement data changing over time, including pupil diameter, eye state, and speed. It can also zoom in and out on different types of eye movement data signals for easy viewing of data.

[0037] Among them, HRV heart rate variability analysis uses an electrocardiogram recorder to process and analyze PPG signals offline. On the basis of presenting the data as a whole, different segments can be processed and analyzed. PPG signals can be zoomed in and out for easy viewing of data; raw data, processed data and analysis can be exported in ASCII format; and report sheets can be exported.

[0038] Among them, EDA galvanic skin response analysis is combined with galvanic skin acquisition equipment for collection, which can perform SC skin conductance data analysis, SCR event-related galvanic skin analysis, and SCL phase-related galvanic skin analysis. It can export original data, processed data, and analyzed data in ASCII format; and can export visual analysis reports.

[0039] Among them, EMG electromyography analysis uses electromyography acquisition equipment to collect data, and can obtain some statistical characteristics of electromyography signal time domain, including: mean (Mean), median (Median), standard deviation (STD), maximum / minimum (Max / Min), variance (Variance), root mean square (RMS), mean absolute value (MeanAbsolute Value), integrated electromyography (iEMG); can obtain cycle force data, including: cycle start time (StartTime), cycle end time (EndTime), root mean square (RMS), mean absolute value (MeanAbsolute Value), integrated electromyography (iEMG), median frequency (Median Frequency), mean frequency (Mean Frequency). Finally, a variety of data can be exported, including raw data Raw Data, processed data Processed, normalized data (Normalized), PSD data.

[0040] Among them, RESP respiratory analysis collects data through respiratory information collection equipment. With a full-process and phased data information processing method, it can more accurately analyze the respiratory behavior within a certain period of time, and can count and export the time domain and frequency domain data reports of the respiratory signal. The specific data includes time domain data AVRESP, SD, Max, Min, Rang; frequency domain data Power, Peak value, and visualized power spectrum diagram, etc.

[0041] The human posture acquisition function is implemented based on the human posture acquisition submodule, and is connected using the human posture acquisition device. It includes device connection, posture calibration and motion capture functions. The device connection function includes turning on the sensor, placing the sensor, connecting the sensor, and wearing the sensor. Among them, the device should be kept as far away from the ground magnetic interference and the desktop magnetic interference as possible. When all devices are successfully connected, the system prompts that the device is successfully connected and proceeds to the next step, that is, posture calibration. The posture calibration content includes standing posture, half-squatting posture, and unfolding posture. After the calibration is completed, the position of the sensor cannot be moved. Because the algorithm has recorded the position of the sensor and the body at this time, if the sensor position is accidentally moved, it needs to be recalibrated. The motion capture function includes setting the scene (such as flat ground, climbing stairs), contact points (such as hand contact, foot contact or hip contact), contact constraint tolerance (the smaller the value, the more firmly the footsteps contact the ground; the larger the value, the easier it is for the footsteps to slide in contact with the ground), heading angle (adjust the yaw direction of the virtual human model), pitch angle (adjust the pitch angle of the virtual human model) and other parameters, and can record actions.

[0042] The human body state assessment function is implemented based on the human body state assessment submodule, including comfort analysis, fatigue analysis and lumbar force analysis functions. This submodule first synchronizes the data of the human body parameter acquisition and human body posture acquisition submodules, that is, the current time recorded in the eye tracking data, HRV heart rate variability analysis data, EDA skin electrodermal response analysis data, EMG electromyography analysis data, and RESP breathing analysis data is compared with the time recorded in the posture acquisition, and the corresponding / similar time data is imported into the secondary development Delmia / Ergonomics module for evaluation to obtain the human body state evaluation, such as RULA analysis results, lifting and lowering analysis results, and biomechanical single action analysis results. The results are associated with the preset status prompt voice and text in the operation guidance information, and corresponding prompts are given according to the status.

[0043] The comfort analysis unit, fatigue analysis unit and lumbar force analysis unit are used to evaluate the comfort, fatigue and lumbar force of the human body, respectively. They mainly call commercial analysis software such as Delmia to obtain analysis results. First, call Delmia for initialization, including software startup status and human body model, etc.; then correspond the human posture in the augmented reality glasses with the human posture in Delmia, and create the posture of the Delmia human body model in the software; finally, by calling the accessibility analysis, visibility analysis, RULA analysis and biomechanical single action analysis functions of the human activity analysis module in Delmia, the human comfort, fatigue and lumbar force analysis results are derived and displayed in the augmented reality glasses for prompts.

[0044] To sum up, the augmented reality assisted assembly system taking human factors into consideration in the above-mentioned embodiment, after collecting human physiological parameters through the human physiological state monitoring module, combines the current human posture and work task parameters, and imports them into the evaluation model for automatic analysis, and finally gives the current physiological and psychological state of the assembler, and gives prompts in time, which fully improves the assembly load of the assembler and reduces human-induced assembly errors, and is worthy of promotion and use.

[0045] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. An augmented reality assisted assembly system considering human factors, characterized in that: include: Scene model and assembly model import module, real-time auxiliary assembly module, human physiological state monitoring and evaluation module; The scene model and assembly model import module is used to import and process three-dimensional models, extract assembly information and set multi-mode guidance information; The scene model and assembly model import module includes a three-dimensional model processing submodule, an assembly information extraction submodule, and a multi-mode guidance information submodule; the three-dimensional model processing submodule is used to provide the system with an import interface for the environment model, tool model, and assembly model, and to perform lightweight settings on the data after the model is imported; the assembly information extraction submodule is used to extract the assembly relationship, hierarchical structure, part information, and assembly path, and generate an assembly model sequence and a model tree; the multi-mode guidance information submodule is used to mark the assembly model tree and model sequence obtained by extracting the assembly information; The real-time auxiliary assembly module is used to perform object visual processing and human-computer interaction to provide real-time guidance to the assembler; The real-time auxiliary assembly module includes an object visual processing submodule and a human-computer interaction submodule; the object visual processing submodule is used to perform three-dimensional scene reconstruction and target recognition and tracking; the human-computer interaction submodule is used to perform augmented reality scene display and personnel control feedback; The object vision processing submodule includes a 3D scene reconstruction unit and a target recognition and tracking unit; the 3D scene reconstruction unit is used to realize multi-directional 3D scene reconstruction as the environmental input of the real-time auxiliary assembly module; the target recognition and tracking unit is realized based on augmented reality glasses, and adopts simultaneous positioning and map building technology. When the assembler moves continuously in the indoor environment, the augmented reality glasses are used to obtain a series of continuous depth maps at different angles, continuously iterate, accumulate and analyze different depth maps, integrate and calculate to establish an accurate environmental model; The human-computer interaction submodule includes an augmented reality scene display unit and an assembly personnel manipulation feedback unit; the augmented reality scene display unit is used to import the scene model and the tooling tools in the assembly model import module, and solve the registered posture of the virtual guide tool, integrate it with the real scene, and import it into the augmented reality glasses; The assembler manipulation feedback unit is used to prompt the assembler's current assembly status and actions through the system's interference detection function and human body status assessment function; The human physiological state monitoring and evaluation module is used to collect human parameters, human posture, synchronize data and evaluate human state, obtain the current physiological and psychological state of the assembler, and give prompts; The human physiological state monitoring and evaluation module includes a human body parameter acquisition submodule, a human body posture acquisition submodule, a human body state evaluation submodule and a data synchronization setting submodule; The human body parameter acquisition submodule is used for acquisition initialization, data processing and conversion, and data display; The human body posture acquisition submodule is used for device connection, posture calibration, and motion capture; the human body state assessment submodule is used for comfort analysis, fatigue analysis, and lumbar vertebrae force analysis; the data synchronization setting submodule is used for time synchronization of the data of the human body parameter acquisition submodule and the human body posture acquisition submodule.

2. The augmented reality assisted assembly system considering human factors according to claim 1, characterized in that: The human body parameter acquisition submodule includes an acquisition initialization unit, a data processing and conversion unit, and a data display unit; the acquisition initialization unit is used to install the sensor at the position where measurement is required; the data processing and conversion unit is used to perform eye tracking analysis, HRV heart rate variability analysis, EDA galvanic skin response analysis, EMG electromyography analysis, and RESP respiratory analysis; the data display unit is used to display the data acquired by the data processing and conversion unit, and illustrate the current state of the human body.

3. The augmented reality assisted assembly system considering human factors according to claim 2, characterized in that: When conducting eye tracking analysis, the eye movement data signal is measured through an eye tracker, the information and data related to the eyes are measured and analyzed, the eye movement data signal is processed and analyzed, and the statistical characteristics of the eye movement data changing over time are presented; When conducting HRV heart rate variability, the PPG signal is processed and analyzed offline through an electrocardiogram recorder; when conducting EDA skin conductance response, data is collected through skin conductance acquisition equipment, and then SC skin conductance data analysis, SCR event-related skin conductance analysis, and SCL phase-related skin conductance analysis are performed; When performing EMG electromyography analysis, data is collected through electromyography acquisition equipment to obtain statistical characteristics of electromyography signals in the time domain, and then periodic force data is obtained to achieve electromyography analysis. When performing RESP respiratory analysis, data is collected through respiratory information acquisition equipment, and the respiratory behavior within the set time period is analyzed in a full-process and staged data information processing method, and the time domain and frequency domain data reports of the respiratory signals are counted and exported.

4. The augmented reality assisted assembly system considering human factors according to claim 3, characterized in that: The human body posture acquisition submodule includes a device connection unit, a posture calibration unit, and a motion capture unit; the device connection unit is used to turn on the sensor, place the sensor, connect the sensor, and wear the sensor; The posture calibration unit is used to calibrate the standing posture, half-squatting posture, and unfolding posture; the motion capture unit is used to set the scene, contact point, contact constraint tolerance, heading angle, pitch angle parameters, and perform motion recording.

5. The augmented reality assisted assembly system considering human factors according to claim 4, characterized in that: The human body state assessment submodule includes a comfort analysis unit, a fatigue analysis unit and a lumbar vertebrae force analysis unit; the comfort analysis unit, the fatigue analysis unit and the lumbar vertebrae force analysis unit are used to assess the human body's comfort, fatigue and lumbar vertebrae force conditions respectively.

Images