A method for evaluating and analyzing personnel fatigue intensity in human-machine collaborative operation systems

Abstract

A method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system includes the following steps: S1. Establishing the state equation and observation equation of the human-machine collaborative system, and determining the picking efficiency and fatigue state of the picker as state variables; S2. Constructing a picker fatigue model, comprehensively considering physiological and psychological factors, including perceived workload and actual workload, as well as shoulder and arm muscle load; S3. Applying the extended Kalman filter algorithm, combined with actual order data, to dynamically estimate the picker's fatigue state; S4. Designing numerical experiments, setting parameters and simulation conditions, and analyzing sensitivity factors such as picking intensity and order input frequency to study their impact on system performance; S5. Analyzing the numerical experiment results to evaluate the system performance and picker fatigue state under different human-machine collaborative modes. The present invention can accurately evaluate the picker's fatigue state, optimize the order picking process, improve system efficiency, and reduce the physical burden on pickers.

Description

Technical Field

[0001] The present invention relates to the field of human-machine collaborative systems and intelligent warehousing technology, and in particular to a method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system, which is used to evaluate and optimize the fatigue status of pickers in a human-machine collaborative order picking system. Background Art

[0002] With the rise of Industry 4.0, intelligent warehousing systems are becoming a growing trend in the logistics industry. Human-robot collaborative order picking, an emerging operational model within intelligent warehousing systems, effectively improves picking efficiency and accuracy through close collaboration between pickers and robots. However, pickers are prone to fatigue during prolonged, repetitive tasks, which not only affects picking efficiency but can also lead to health issues such as musculoskeletal disorders. Therefore, how to scientifically assess picker fatigue and optimize operational processes has become a pressing issue in the intelligent warehousing field.

[0003] While research has focused on ergonomic issues in warehouse operations, most studies focus on purely manual picking systems. Fatigue assessment and management strategies for human-robot collaborative systems are relatively limited. Furthermore, existing fatigue assessment methods often rely on questionnaires or subjective reports, lacking objectivity and real-time performance. For human-robot collaborative systems, a more accurate and automated fatigue assessment method is needed to enable real-time monitoring and dynamic adjustment of picker performance.

[0004] It should be noted that the information disclosed in the above background technology section is only used to understand the background of this application, and therefore may include information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0005] The main purpose of the present invention is to solve the problems existing in the above-mentioned background technology and to provide a method for evaluating and analyzing the fatigue strength of personnel in a human-machine collaborative operation system.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system comprises the following steps: S1. According to the characteristics of the human-machine collaborative system, a state equation and an observation equation of the system are established, and the picking efficiency and fatigue state of the picker are determined as state variables; S2. Based on the state variables determined in step S1, a picker fatigue model is established, and the model takes into account physiological and psychological factors, including actual workload and perceived workload, to quantify the picker's fatigue degree; S3. Based on the fatigue model constructed in step S2 and the state equation in step S1, an extended Kalman filter algorithm is used to dynamically estimate the picker's fatigue state according to actual order data; S4. Under the guidance of the dynamic estimation result obtained in step S3, the parameters and simulation conditions of the numerical experiment are designed, and a sensitivity analysis is performed to study the influence of different parameters such as picking intensity, order input frequency, etc. on the system performance and the picker's fatigue state; S5. The numerical experiment results of step S4 are analyzed to evaluate the system performance and the picker's fatigue state under different human-machine collaborative modes.

[0008] Furthermore, in step S1, based on the characteristics of the human-machine collaborative order picking system, the system's state equation and observation equation are first established. The state variable is set to the picker's picking efficiency, because picking efficiency directly reflects the system's operating status and the picker's working status. When constructing the state equation, considering that picking efficiency decreases with the accumulation of fatigue index, a dynamic equation is used to describe this relationship. At the same time, the original order is used as the observation equation, combined with the actual order completion status to estimate and update the state of picking efficiency. This setting enables the system to reflect changes in the picker's working status in real time.

[0009] Furthermore, in the process of establishing the state equation and observation equation, an extended Kalman filter (EKF) model was used to address the nonlinear factors in the system. By linearizing the nonlinear model, the EKF enables the traditional Kalman filter algorithm to be applied to this scenario. This approach makes the system model more realistic and improves the accuracy of state estimation.

[0010] Furthermore, the present invention considers multiple factors in its model of picker fatigue, including but not limited to the physical intensity of the picking task, the picker's psychological burden, the work environment, and the picker's physiological and psychological state. By integrating these factors, the present invention can more comprehensively assess the picker's fatigue state.

[0011] Furthermore, in step S2, a fatigue model based on workload and muscle load was established to comprehensively assess the picker's fatigue status. This model not only considers physical factors such as actual workload, such as the weight and distance of goods moved, but also incorporates the psychological dimension of perceived workload to comprehensively assess the picker's fatigue level during the operation. This comprehensive fatigue characterization method helps to more accurately understand the picker's fatigue status and provides a basis for subsequent fatigue management.

[0012] Furthermore, in step S3, the extended Kalman filter algorithm is used to dynamically estimate the picker's fatigue state. This algorithm continuously integrates system observation data (such as order completion status) with state equation prediction results to update the picker's fatigue state in real time. This method effectively captures the dynamic changes in the picker's fatigue state, providing managers with real-time fatigue status information, thereby supporting more scientific decision-making.

[0013] Furthermore, the proposed method dynamically estimates the fatigue state of the picker through the extended Kalman filter algorithm, which can handle the nonlinear characteristics of the system. Time step to state transition function And the observation function is linearized to achieve the picking Accurate tracking of personnel fatigue status.

[0014] Furthermore, in step S3, the extended Kalman filter algorithm employed by the present invention is particularly well-suited for systems with nonlinear dynamic characteristics, such as human-machine collaborative order picking systems. This algorithm linearizes the state prediction model by calculating the Jacobian matrix, and then updates the state estimate at each time step, ensuring its accuracy and reliability.

[0015] Furthermore, when using the extended Kalman filter algorithm for fatigue state estimation, it is necessary to adjust and verify the algorithm parameters based on actual order data. By comparing and analyzing the fatigue state estimation results under different order data, the algorithm performance can be further optimized, and the accuracy and reliability of fatigue state estimation can be improved.

[0016] Furthermore, in step S4, the present invention designs a detailed numerical experimental scheme, simulating different working conditions and system parameters through simulation experiments to study their impact on picker fatigue and system performance. The experimental results will provide important data support for the design and optimization of the actual system.

[0017] Furthermore, in step S4, numerical experiment parameters and simulation conditions were designed to study the impact of different parameters on system performance. These parameters include picking intensity, order input frequency, and picker input frequency. Different configurations of these parameters directly affect system efficiency and picker fatigue. Through simulation experiments, the impact of these parameters on system performance can be systematically evaluated, providing theoretical support for practical production management.

[0018] Furthermore, a sensitivity analysis of the experimental results was conducted. By varying key parameters such as picking intensity and order entry frequency, the impact of these changes on picker fatigue and system performance was evaluated. This helped identify the factors most critical to system performance, thus providing guidance for system optimization.

[0019] Furthermore, in step S5, analysis of the numerical experimental results allows evaluation of system performance and picker fatigue under different human-robot collaborative modes. The experimental results show that different collaborative modes and system strategies significantly impact picker fatigue and overall system performance. For example, the single-picker, multi-robot model excels in reducing picker fatigue, while appropriate rest arrangements and order allocation strategies can also effectively reduce picker fatigue. These findings provide important insights for optimizing the design and operation of human-robot collaborative order picking systems.

[0020] Furthermore, the present invention also proposes system optimization suggestions based on the simulation results, including adjusting the ratio of pickers and robots, improving picking path planning, optimizing order allocation strategies, etc., to reduce picker fatigue and improve the overall efficiency of the system.

[0021] The present invention has the following beneficial effects:

[0022] This paper proposes a method for assessing and analyzing fatigue intensity in human-machine collaborative systems. This method establishes a picker fatigue model and dynamically estimates it using actual order data. The effectiveness of the method is verified through numerical experiments. This method provides a new technical approach for managing picker fatigue in intelligent warehousing systems, helping to improve the efficiency and safety of warehousing operations.

[0023] This invention's picker fatigue analysis method, based on an extended Kalman filter, effectively addresses the shortcomings of existing human-machine collaborative order picking systems in terms of worker fatigue assessment and management, improving picking efficiency while also reducing the physical and psychological fatigue and associated health risks associated with long work hours. By accurately assessing picker fatigue and dynamically adjusting work and rest schedules, this invention optimizes the operational processes of human-machine collaborative systems, improving overall system performance and meeting the high demands for immediacy, efficiency, and employee well-being in modern intelligent warehousing systems.

[0024] By establishing state and observation equations for a human-robot collaborative order picking system and using the picker's picking efficiency as the state variable, this invention accurately reflects the working status of both the picker and the robot in real time, thereby improving picking efficiency and accuracy. Furthermore, by integrating a model of picker fatigue, taking into account physiological and psychological factors, and further dynamically estimating the picker's fatigue state through an extended Kalman filter algorithm, this approach provides a scientific basis for system optimization. These measures help reduce fatigue-related errors, improve overall operational efficiency, and optimize warehouse management.

[0025] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system according to an embodiment of the present invention;

[0027] Figure 2 This is a schematic diagram of the linear Kalman filter principle according to an embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of the extended Kalman filter principle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope of the present invention and its application.

[0030] like Figure 1 As shown, an embodiment of the present invention provides a method for evaluating and analyzing fatigue strength of personnel in a human-machine collaborative operation system, comprising the following steps S1 to S5:

[0031] S1. According to the characteristics of the human-machine collaborative system, establish the system's state equation and observation equation, and determine the state variables.

[0032] The state equation is used to describe the dynamic changes in the picker's fatigue state over time, as well as the relationship between picking efficiency and fatigue state. The observation equation is used to correct and update the picker's state estimate based on actual order data. The specific steps include S11 to S12:

[0033] S11. First, based on the characteristics of the human-machine collaborative system, the system's state equation and observation equation are established. The state equation describes the dynamic changes in the picker's fatigue state over time, as well as the relationship between picking efficiency and fatigue state. The observation equation uses actual order data to correct and update the picker's state estimate.

[0034] The state equation can be expressed as: k+1 =δ k-α·δ k Workload k +w k Among them, α is the fatigue impact factor, and Workload k is the workload of stage k, w k is process noise. The observation equation relates picking efficiency to actual order fulfillment: k =H·δ k +∈ k , H is the observation matrix, Z k is the observed order completion, ∈ k is the observation noise.

[0035] S12. Determine the following state variables: 1. Picking efficiency (δ k ):Indicates the picker's ability to complete orders per unit time, and this efficiency varies with the picker's fatigue state. 2. Fatigue state: Quantifies the degree of physical and mental fatigue accumulated by the picker due to long working hours. This can be assessed through physiological measurements (such as heart rate, electromyography) and psychological questionnaires. 3. Rest need: Dynamically adjust the rest time according to the fatigue state and picking efficiency to maintain the picker's work performance and health. Rest need can be expressed by the following formula: Rest k+1 =Rest k +β·(δ k -Threshold) where β is the rest adjustment factor, Threshold

[0036] is the preset fatigue threshold. Specifically, it includes the following steps:

[0037] S121. Define initial conditions for state variables, such as initial fatigue level and initial work efficiency, which are based on the state of the picker before the start of work; S122. According to the characteristics of the picking task and the work performance of the picker, establish a mathematical model of the change of state variables over time; S123. Use appropriate mathematical tools and algorithms to solve the state equation to ensure that the fatigue state and picking efficiency of the picker can be tracked and predicted in real time; S124. Use the observation equation to combine the actual order data with the state estimate, and use the Kalman filter algorithm to correct the state of the picker in real time to improve the accuracy of the state estimate; S125. Consider special requirements and constraints such as workstation layout, picking path and order processing process to ensure that the state equation and observation equation can truly reflect the working conditions of the human-machine collaborative system.

[0038] S2. Construct a comprehensive picker fatigue assessment model, which quantifies and analyzes the picker's fatigue status at work through two main aspects.

[0039] First, based on the workload fatigue characterization method, the perceived workload and actual completed workload of the pickers are quantified through questionnaires, behavioral observations, and physiological indicator monitoring, thereby evaluating their overall workload. Secondly, based on the fatigue model of muscle load, physiological measurement techniques such as electromyography (EMG) are used to collect activity data of the shoulder and arm muscles during the picking process, analyze the relationship between muscle load and fatigue level, and establish a corresponding mathematical model. In addition, the effects of the working environment, workstation design, and picking task characteristics on muscle load are also considered, and a real-time monitoring and adjustment algorithm is developed to ensure that the pickers work within a safe fatigue threshold. Finally, by combining the two fatigue models, a comprehensive method for evaluating the fatigue status of pickers is provided, which provides a basis for formulating reasonable work arrangements and rest. The plan provides a scientific basis for optimizing human-machine collaborative systems Workflow, improve efficiency and ensure The health of the pickers. The specific steps include S21~S22:

[0040] S21. The following detailed methods are used to characterize the workload-based fatigue of pickers: Definition of perceived workload: The subjective feelings of pickers about their work tasks, including task difficulty, complexity, time pressure, etc., are quantified into numerical indicators through questionnaires; Quantification of actual workload: The number of orders actually completed and the weight of goods handled by pickers within a certain period of time are recorded, and the data are automatically collected through the data recording system of the workstation; Quantification method: Combine questionnaires, behavioral observations and physiological indicators (such as heart rate, blood pressure) monitoring, and use statistical methods to convert perceived workload and actual workload into comparable numerical values; Establishment of relationship model: Apply linear regression analysis to establish a relationship model between perceived workload and actual workload, for example: Fatigue Index =a·Perceived Workload+b·Actual Workload+c, where a, b, and c are regression coefficients obtained by data fitting.

[0041] S22. Use the following methods to establish a fatigue model based on muscle load: Application of physiological measurement technology: Use electromyography (EMG) equipment to collect shoulder and arm muscle activity data during the picking process and analyze muscle force; Muscle load analysis: Evaluate the muscle load when lifting, carrying, and placing goods, considering the frequency and duration of muscle contraction; Construction of mathematical model: Establish a mathematical model between muscle load and fatigue level, considering maximum contraction force (MVC), continuous working time, and recovery time, for example: Muscle Fatigue (t) = MVC·(1-e -λ·t ), where λ is the fatigue growth rate and t is the working time; Biomechanical simulation: Use biomechanical principles to simulate the muscle fatigue accumulation process, predict the muscle fatigue threshold, and ensure that the work schedule is within a safe range. The specific steps include S221 to S224:

[0042] S221. Consider the impact of the following factors on picker fatigue: Workstation design: Evaluate how shelf height, picking tools, workstation ergonomics, etc. affect the picker's muscle load.

[0043] S222. Adjust the muscle load model to reflect actual working conditions: Consider factors such as order diversity, cargo weight distribution, picking path layout, etc., and adjust the muscle load model to more accurately reflect actual working conditions.

[0044] S223. Develop an algorithm to monitor and adjust the picker's muscle load in real time: Develop an algorithm to monitor muscle load in real time to ensure that the picker works within a safe fatigue threshold.

[0045] S224. Combining two fatigue models, a method for comprehensively evaluating the fatigue status of pickers is provided: combining the fatigue characterization based on workload and the fatigue model results based on muscle load to provide a scientific basis for work arrangements and rest plans.

[0046] S3, such as Figure 2 、 Figure 3 As shown in Figure 1, by initializing the picker's fatigue state and error covariance matrix, an extended Kalman filter algorithm is established and applied to dynamically predict and update the picker's fatigue state. The specific steps include S31 to S310. The symbols of the Kalman filter model are shown in Table 1:

[0047] Table 1 Kalman filter model symbol description

[0048]

[0049] S31: Set the initial estimated value δ0 of the picker's fatigue state and the initial error covariance matrix P0. These initial values ​​are based on the picker's state before the start of work, such as the initial fatigue level and initial work efficiency.

[0050] S32. Construct a mathematical model to describe the evolution of the picker's fatigue state over time. The model considers the impact of the picking task on fatigue and the effect of rest on fatigue relief. The model takes the following form: k|k-1 =δ k-1 -α·Workload k-1 +Rest k-1 Among them, α is the fatigue impact factor, and Workload k-1 is the workload of the last moment, Rest k-1 It is the amount of fatigue relief from rest.

[0051] S33, using the state of the previous moment to estimate δ k-1 and the prediction model to calculate the fatigue state prediction value δ at the current moment k|k-1This step makes predictions based only on previous information and does not involve actual observed data.

[0052] S34. Calculate the Jacobian matrix F for the nonlinear state prediction model k To perform linearization. The Jacobian matrix contains the first-order partial derivatives of the state prediction model with respect to the state variable, and its calculation formula is as follows:

[0053] S35. Build an observation model based on actual order data to correlate the picker's fatigue status with the observable order completion status. The observation model can be in the following form: k =h(δ k )+∈ k Among them, h is the observation function, Z k is the observed order completion, ∈ k is the observation noise.

[0054] S36, according to the current prediction error covariance P k|k-1 The Kalman gain K is calculated by the error covariance R of the observation model k The Kalman gain determines the weight of the observation data in the state update, and its calculation formula is as follows: K k =P k|k-1 H T (HP k|k-1 H T +R) -1 .

[0055] S37, combined with the predicted fatigue state δ k|k-1 and the actual order data Z k , through the Kalman gain K k Correct the predicted state to obtain the updated fatigue state estimate δ k :δ k =δ k|k-1 +K k (Z k -Hδ k|k-1 ).

[0056] S38. Update the error covariance matrix P k To reflect the uncertainty of the new state estimate: P k =(IK k H)P k|k-1 .

[0057] S39. Evaluate the prediction accuracy of the model and adjust the model parameters as needed to improve the accuracy of fatigue state estimation. This may include adjusting the fatigue influence factor α, the observation function h, or the observation noise covariance R.

[0058] S310: Repeat steps S32 to S39 at each time step or after completing a certain number of orders. This process dynamically tracks and estimates the fatigue state of the picker, ensuring the continuity and real-time nature of the estimation.

[0059] S4. Design and implement numerical experiments to evaluate the impact of picker fatigue in a human-robot collaborative order picking system. This includes determining the experimental objectives, setting up the simulation environment, selecting key parameters (such as the number of pickers and robots, order characteristics, work intensity, etc.), initializing the simulation model, running the simulation process, and collecting and recording key performance data. Through this data, analyze the system performance and picker fatigue status under different parameter configurations to provide a basis for system optimization. The specific steps include S41 to S45:

[0060] S41. Design a numerical experiment plan based on the characteristics of the human-machine collaborative system. This includes clarifying the experimental objectives, such as evaluating the impact of different parameters on picker fatigue and system performance; setting up the simulation environment, determining the system states and variables to be simulated in the experiment; and establishing evaluation criteria for the expected results to measure whether the experimental results meet the expected objectives.

[0061] S42. Set key parameters in the simulation experiment based on the experimental design. These parameters include the number of pickers, which simulates the needs of different scale operations; the number of robots, which affects the effectiveness of human-machine collaborative operations; rate and fatigue level; order arrival pattern, Affects the workload of the system; and the complexity of the picking task Complexity reflects the diversity and difficulty of order processing.

[0062] S43: Initialize the simulation model using the state equation and observation equation established in step S1 and the fatigue model proposed in step S2. This includes setting the picker's initial fatigue state and picking efficiency to provide a realistic starting point for the simulation experiment.

[0063] S44. Start the simulation software or a custom simulation program and run the simulation process of the human-machine collaborative order picking system according to the set parameters and initialization conditions. This step is the core of the experimental design and simulates the actual system operation through simulation.

[0064] S45. During the simulation process, key performance indicator data is collected in real time. This data includes picker fatigue status, order completion status, system throughput, etc., providing a basis for subsequent data analysis and result evaluation.

[0065] The parameters in step S4 include: number of pickers: set the number of pickers in the experiment to simulate the work requirements of different scales; number of robots: determine the number of robots working with the pickers, which will affect the work efficiency and the pickers' fatigue level; order characteristics: including the number, type, size, etc. of orders, which affect the complexity and workload of the picking task; work intensity: the amount of work completed by the picker in unit time, which affects the speed of fatigue accumulation; order input frequency: the speed at which new orders enter the system, which affects the picker's work rhythm and rest time arrangement; picker input frequency: the frequency at which the picker participates in the task, which is related to workload and rest time; fatigue parameter: set according to the picker's physiological and psychological state, used to evaluate the impact of fatigue on work efficiency; rest strategy: determine the length and frequency of the rest time after the picker completes the task, which affects the fatigue recovery effect; partition strategy: if applicable, including the division method of the picking area and the work distribution of the picker in different areas.

[0066] S5. Perform sensitivity analysis and statistical processing on the data collected from the simulation experiment to evaluate the impact of key parameter changes on picker fatigue and system performance. This step aims to identify the most influential factors, refine optimization strategies, and form practical conclusions that will help improve system efficiency and protect picker well-being. Specific steps include S51-S52:

[0067] S51. Conduct a sensitivity analysis to identify the impact of key parameters on system performance and picker fatigue. This involves varying parameters such as order input intensity and picker rest frequency, and observing how these changes affect picking efficiency, order completion time, and accumulated picker fatigue. The goal of this step is to determine which factors are most sensitive to the system, thereby providing optimization guidance for system design and operation.

[0068] S52. Conduct in-depth statistical analysis on the data collected from the simulation experiment. This includes calculating statistical indicators such as mean, standard deviation, probability distribution, etc. to evaluate the effects of different strategies on system efficiency (such as throughput). volume, order processing time) and picker health (e.g. fatigue level, rest need). Based on these analyses, we have drawn conclusions that are helpful for actual system optimization, such as the best human-machine ratio and the optimal rest frequency.

[0069] In summary, the present invention's fatigue intensity assessment and analysis method for human-machine collaborative work systems offers an innovative solution to the fatigue management needs of pickers in intelligent warehousing systems. By establishing state and observation equations, this method captures pickers' picking efficiency and fatigue status in real time, providing a data foundation for accurate assessment. Furthermore, by constructing a fatigue model that comprehensively considers physiological and psychological factors, the present invention can quantify a picker's fatigue level, thereby providing a more comprehensive understanding of the impact of fatigue on work efficiency and health.

[0070] Utilizing the extended Kalman filter algorithm, the present invention achieves dynamic estimation of picker fatigue. This approach not only improves the accuracy of the assessment, but also enhances its real-time and automated nature, effectively addressing the shortcomings of traditional methods that rely on questionnaires or subjective reporting. Through designed numerical experiments and sensitivity analysis, the present invention further investigates the impact of key parameters such as picking intensity and order entry frequency on system performance and picker fatigue, providing a scientific basis for system optimization.

[0071] This invention can monitor and dynamically adjust the work status of pickers in real time, optimizing the workflow of the human-machine collaborative system and significantly improving picking efficiency and accuracy. Furthermore, by accurately assessing and dynamically adjusting work and rest schedules, it reduces physical and mental fatigue and associated health risks for pickers, thus meeting the high demands of modern intelligent warehousing systems for immediacy, efficiency, and employee well-being. Furthermore, the application of this invention is not limited to intelligent warehousing; its principles and methods have the potential to be extended to other fields, possessing broad practical application value and social benefits.

[0072] The above content is a further detailed description of the present invention in conjunction with specific / preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, without departing from the concept of the present invention, they can also make several substitutions or modifications to these described embodiments, and these substitutions or modifications should be regarded as belonging to the scope of protection of the present invention. In the description of this specification, the reference terms "one embodiment", "some embodiments", "preferred embodiments", "examples", "specific examples", or "some examples" and the like mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in an appropriate manner in any one or more embodiments or examples. Without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples. Although detailed descriptions have been made, Described The embodiments of the present invention and its advantages, but it should be understood that without departing from the protection of the patent application Various changes, substitutions, and alterations can be made herein without departing from the scope thereof.

Claims

1. A method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system, characterized in that: The following steps are involved: S1. Based on the characteristics of the human-machine collaborative system, the state equation and observation equation of the system are established, and the picking efficiency and fatigue state of the picker are determined as state variables; S2. Based on the state variables determined in step S1, a picker fatigue model is established, wherein the fatigue model takes into account physiological and psychological factors, including actual workload and perceived workload, to quantify the picker's fatigue level; S3. Based on the fatigue model constructed in step S2 and the state equation in step S1, the extended Kalman filter algorithm is used to dynamically estimate the picker's fatigue state according to the actual order data; S4. Guided by the dynamic estimation results obtained in step S3, design the parameters and simulation conditions of the numerical experiment, conduct sensitivity analysis, and study the effects of different parameters, including picking intensity and order input frequency, on system performance and picker fatigue status; S5. Analyze the numerical experimental results of step S4 and evaluate the system performance and picker fatigue status under different human-machine collaboration modes.

2. The method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system according to claim 1, wherein: Step S1 specifically includes: S11. Based on the characteristics of the human-machine collaborative system, establish the system's state equation and observation equation, wherein the state equation is used to describe the dynamic changes in the picker's fatigue state over time, as well as the relationship between picking efficiency and fatigue state; the observation equation is used to correct and update the picker's state estimate based on actual order data; S12. Determine state variables, including: The picking efficiency of the picker refers to the ability to complete orders per unit time. This picking efficiency varies with the picker's fatigue state. The fatigue status of the pickers, which quantifies the degree of physical and mental fatigue accumulated by the pickers due to long working hours; The picker's rest needs are dynamically adjusted according to fatigue status and picking efficiency to maintain the picker's work performance and health.

3. The method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system according to claim 2, wherein: Step S12 specifically includes: S121, further determining initial conditions of the state variables based on the state of the picker before the start of work, including initial fatigue level and initial work efficiency; S122. Define the evolution rules of the state variables and establish a mathematical model of the changes of the state variables over time based on the characteristics of the picking task and the picker's work performance; S123. Use appropriate mathematical tools and algorithms to solve the state equation to ensure that the picker's fatigue state and picking efficiency can be tracked and predicted in real time; S124. Using the observation equation to combine the actual order data with the state estimate, the Kalman filter algorithm is used to perform real-time correction of the picker's state to improve the accuracy of the state estimate. S125. Consider the special requirements and constraints of the system, including workstation layout, picking path and order processing flow, to ensure that the state equation and observation equation can truly reflect the working conditions of the human-machine collaborative system.

4. The method for evaluating and analyzing fatigue strength of personnel in a human-machine collaborative operation system according to any one of claims 1 to 3, wherein: Step S2 specifically includes: S21. Perform workload-based fatigue characterization, including: Define perceived workload, which is the picker's subjective feeling about the work task, including the difficulty, complexity and time pressure of the task; Define the actual workload completed, that is, the number of orders actually completed by the picker within a certain period of time and the weight of goods handled; Quantify the perceived and actual workload of pickers through questionnaires, behavioral observations, or physiological indicator monitoring; Using statistical analysis methods, a relationship model between perceived workload and actual workload was established to evaluate the overall workload of pickers. S22. Establish a fatigue model based on muscle load, focusing on the load on the shoulder and arm muscles during the picking process. The fatigue model based on muscle load includes: Physiological measurement techniques, including electromyography (EMG), are used to collect data on shoulder and arm muscle activity during the picking process. Analyze the impact of picking actions on muscle load, including muscle exertion when lifting, carrying, and placing goods; Establish a mathematical model between muscle load and fatigue level, taking into account the maximum contraction force, continuous working time and recovery time of the muscle; Using the principles of biomechanics, the accumulation process of muscle fatigue of pickers under different working conditions is simulated, and the threshold of muscle fatigue is predicted.

5. The method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system according to claim 4, wherein: Step S22 specifically includes: S221. Consider the impact of the work environment and workstation design on picker fatigue, including shelf height, use of picking tools, and ergonomic design of workstations; S222. Consider the characteristics of picking tasks, including order diversity, weight distribution of goods, and layout of picking routes, and adjust the fatigue model based on muscle load to more accurately reflect actual working conditions. S223. Real-time monitoring and adjustment of fatigue models based on muscle load to ensure that pickers work within a safe fatigue threshold and reduce the risk of musculoskeletal disorders; S224. Combining workload-based fatigue characterization and muscle load-based fatigue models, a method for comprehensively evaluating the fatigue status of pickers is developed to provide a basis for the formulation of work arrangements and rest plans.

6. The method for evaluating and analyzing fatigue strength of personnel in a human-machine collaborative operation system according to any one of claims 1 to 3, wherein: In step S3, the extended Kalman filter algorithm includes a state-space model and an observation model; the state-space model is used to describe the evolution of the system state over time. In the state-space model, the state of the system is defined by a set of first-order differential equations, and the first-order differential equations describe the dynamic behavior of the system state; The observation model cooperates with the state space model to define how to obtain actual observation data from the system state. The observation model contains observation noise, which reflects the deviation between the actual observation and the ideal state. The working process of the extended Kalman filter includes: Prediction step: At the beginning of each time step, based on the current state estimate and the known system dynamic model, the state at the next moment is predicted. This step does not rely on new observations but is based on previous information; Update step: Once new observation data is obtained, the extended Kalman filter compares the data with the predicted state, calculates the residual, and updates the state estimate by an optimal estimate, the Kalman gain, which minimizes the sum of squares of the prediction errors; The extended Kalman filter (EKF) linearizes the nonlinear function at each time step, making the Kalman filter applicable to nonlinear systems. Among them, the linearization processing of nonlinear functions includes: EKF converts the nonlinear problem into a linear problem by calculating the Jacobian matrix, that is, the first-order partial derivatives of the state transfer function and the observation function with respect to the state variable, thereby allowing the filter to update its estimate at each time step to adapt to the nonlinear characteristics of the system.

7. The method for evaluating and analyzing fatigue strength of personnel in a human-machine collaborative operation system according to claim 6, wherein: Step S3 includes: S31. Initialization state: setting an initial estimate of the picker's fatigue state and an initial error covariance matrix. The initial estimate is based on the picker's state before work begins, including an initial fatigue level and initial work efficiency. S32. State prediction model establishment: Construct a mathematical model that describes the evolution of the picker's fatigue state over time. This state prediction model takes into account the impact of the picking task on the picker's fatigue and the effect of rest on fatigue relief; S33, state prediction: using the state estimation and state prediction model of the previous moment, calculate the fatigue state prediction value of the current moment. This step does not involve actual observation data and is based only on previous information. S34. Calculate Jacobian matrix: For nonlinear state prediction models, calculate the Jacobian matrix for linearization. The Jacobian matrix contains the first-order partial derivatives of the state prediction model with respect to the state variables. S35. Observation model establishment: Based on actual order data, an observation model is established, which associates the picker's fatigue status with the observable order completion status; S36, Kalman gain calculation: Calculate the Kalman gain based on the current prediction error covariance and the error covariance of the observation model. The Kalman gain determines the weight of the observation data in the state update; S37, state update: combining the predicted fatigue state and the actual order data, correcting the predicted state through the Kalman gain to obtain an updated fatigue state estimate; S38, Error covariance update: Update the error covariance matrix to reflect the uncertainty of the new state estimate; S39, Model Verification and Adjustment: Evaluate prediction accuracy and adjust parameters of the state prediction model and observation model as needed to improve the accuracy of fatigue state estimation; S310, loop execution: at each time step or after completing a certain number of orders, repeat steps S32 to S39 to achieve dynamic tracking and estimation of the picker's fatigue state.

8. The method for evaluating and analyzing fatigue strength of personnel in a human-machine collaborative operation system according to any one of claims 1 to 3, wherein: The specific process of step S4 includes: S41. Experimental Design: Design a numerical experiment plan based on the characteristics of the human-computer collaborative system, including the experimental objectives, simulation environment settings, and evaluation criteria for expected results; S42. Parameter setting: Based on the experimental design, set the key parameters in the simulation experiment, including the number of pickers, the number of robots, the order arrival pattern, and the complexity of the picking task; S43, model initialization: using the state equation and observation equation established in step S1, and the fatigue model established in step S2, the simulation model is initialized, and the initial fatigue state and picking efficiency of the picker are set; S44, simulation operation: starting the simulation software or the custom simulation program, and running the simulation process of the human-machine collaborative order picking system according to the set parameters and initialization conditions; S45. Data collection: During the simulation process, key performance indicators of the pickers are collected in real time, including fatigue status data, order completion status, and system throughput.

9. The method for evaluating and analyzing fatigue intensity of personnel in a human-machine collaborative operation system according to claim 8, wherein: The parameters in step S4 include: Number of pickers: The number of pickers set in the experiment is used to simulate the needs of operations of different scales; Number of robots: The number of robots working with pickers affects the efficiency and fatigue level of human-robot collaborative work; Order characteristics: including the quantity, type, and size of the order. Order characteristics affect the complexity of the picking task and the workload of the picker; Work intensity: The amount of work a picker completes per unit time, which affects the rate at which the picker's fatigue accumulates; Order entry frequency: The speed at which new orders enter the system affects the picker's work pace and rest time arrangements; Picker input frequency: the frequency with which pickers participate in picking tasks, which is related to the picker’s workload and rest time; Fatigue parameters: Parameters set based on the picker’s physiological and psychological state to assess the impact of fatigue on work efficiency; Rest strategy: The length and frequency of rest taken by pickers after completing a task affects fatigue recovery. Zoning strategy: includes how the picking areas are divided and how the pickers are assigned to different areas.

10. The method for evaluating and analyzing fatigue strength of personnel in a human-machine collaborative operation system according to any one of claims 1 to 3, wherein: Step S5 includes: S51. Sensitivity analysis: By changing key parameters, including order input intensity and picker rest frequency, observe the impact of changes in key parameters on picker fatigue status and system performance; S52. Result processing: Conduct statistical analysis on the data collected from the simulation experiment, evaluate the impact of different strategies on the efficiency of the human-machine collaborative system and the health of pickers, and draw conclusions that will help optimize the actual system.

Images