A quantitative characterization method of labor efficiency considering the influence of multiple factors

Abstract

A method for quantitatively characterizing labor efficiency that considers the influence of multiple factors includes: collecting worker physical condition data, environmental condition data, and construction progress data; annotating the collected data according to label information; identifying worker workload and calculating labor efficiency based on the collected construction progress data, and synchronously annotating the label information of the construction progress data and the labor efficiency data; aligning data with the same label information according to the label information, using the label information as an entity and "worker physical condition data, environmental condition data-labor efficiency data" as an attribute to establish an "influencing factor-labor efficiency" knowledge graph; based on the "influencing factor-labor efficiency" knowledge graph, using deep learning to establish a mapping relationship between worker physical condition, environmental factors, and labor efficiency. The present invention can scientifically evaluate the correlation between influencing factors and labor efficiency, and quantitatively characterize the impact of influencing factors on labor efficiency fluctuations.

Description

Technical Field

[0001] The present invention relates to the field of engineering construction, and in particular to a labor efficiency quantitative characterization method taking into account the influence of multiple factors. Background Art

[0002] Engineering construction involves three major phases: survey, design, and construction. Construction is the most expensive and crucial phase of the project, requiring the highest investment. Therefore, safe and efficient construction operations are crucial for the safe and high-quality completion of the project. Scientific and rational construction scheduling is crucial to ensuring the smooth progress of construction operations.

[0003] Construction scheduling primarily involves the coordination of resources such as personnel, materials, and equipment. Currently, when scheduling construction on-site, managers typically calculate the planned number of construction workers based on the project's total construction schedule, the project's workload, and labor efficiency quotas. However, labor efficiency is not fixed in real-world production and is susceptible to fluctuations due to factors such as environmental factors and workers' physical condition. Therefore, using quotas to predict labor numbers cannot accurately account for changes in labor efficiency due to these various factors. This inevitably leads to two scenarios: overestimating expected labor efficiency, resulting in insufficient labor input and delays; or underestimating expected labor efficiency, resulting in excessive labor input and wasted resources. Therefore, the existing use of quotas in construction organization results in a mismatch between the planned labor force and actual demand, leading to overstaffing and understaffing in actual construction, ultimately causing delays and wasted resources.

[0004] Existing research examining the correlation between influencing factors and labor efficiency typically involves only a single variable or a subset of these variables. Few studies have combined the physical condition of construction workers with environmental factors. Furthermore, a comprehensive measurement method is currently unavailable, requiring separate collection of environmental and labor efficiency data during research, consuming significant human and material resources. Summary of the Invention

[0005] The present invention provides a method for quantitatively characterizing labor efficiency that takes into account the influence of multiple factors. It can scientifically evaluate the correlation between influencing factors and labor efficiency, and quantitatively characterize the impact of influencing factors on labor efficiency fluctuations, breaking through the existing extensive construction scheduling and improving the industry level of construction scheduling.

[0006] A method for quantitatively characterizing labor efficiency considering the influence of multiple factors includes the following steps:

[0007] Step 1: Data collection: Collect workers' physical condition data, environmental status data, and construction progress data;

[0008] Step 2: Label the collected worker physical condition data, environmental condition data, and construction progress data according to label information, including collection time, collection location, and construction type;

[0009] Step 3: Identify the worker workload and calculate labor efficiency based on the collected construction progress data, and simultaneously annotate the label information of the construction progress data with the labor efficiency data;

[0010] Step 4: Align the worker's physical condition data, environmental condition data, and labor efficiency data with the same label information according to the label information. Use the label information as the entity and "worker's physical condition data, environmental condition data - labor efficiency data" as the attribute to build an "influencing factors - labor efficiency" knowledge graph;

[0011] Step 5: Based on the "influencing factors-labor efficiency" knowledge graph, use deep learning to establish a mapping relationship between workers' physical condition, environmental factors and labor efficiency.

[0012] Furthermore, the worker's physical condition data is collected through electronic files and integrated electronic bracelets; the environmental condition data is collected through an integrated sensing platform; and the construction progress data is obtained through video images collected by a mobile wireless motion camera.

[0013] Furthermore, the worker's physical condition data includes basic condition data, vital sign condition data, sleep quality data and motion status data. The worker's basic condition data is collected using electronic files, and the basic condition data includes the worker's height, weight, gender, age, type of work, proficiency, length of service, and health status; the worker's vital sign condition data is collected using an integrated electronic bracelet, and the vital sign condition data includes the worker's body temperature, blood oxygen, blood pressure, and heart rate; the worker's sleep quality data is collected the night before work, and the motion status data is continuously measured during the construction process, and the motion status data includes the worker's cumulative working hours, energy consumption per unit time, and the motion amplitude of the construction operation.

[0014] Furthermore, the worker's physical condition data includes basic condition data, vital sign condition data, sleep quality data and motion status data. The worker's basic condition data is collected using electronic files, and the basic condition data includes the worker's height, weight, gender, age, type of work, proficiency, length of service, and health status; the worker's vital sign condition data is collected using an integrated electronic bracelet, and the vital sign condition data includes the worker's body temperature, blood oxygen, blood pressure, and heart rate; the worker's sleep quality data is collected the night before work, and the motion status data is continuously measured during the construction process, and the motion status data includes the worker's cumulative working hours, energy consumption per unit time, and the motion amplitude of the construction operation.

[0015] Furthermore, in step 3, based on the collected construction progress data, the worker workload is identified and the labor efficiency is calculated, which specifically includes:

[0016] Step 31: Using target detection, identify various construction activities in the video image;

[0017] Step 32: for the identified construction activities, image segmentation is used to achieve pixel-level progress estimation;

[0018] Step 33: By comparing the progress at different times, solve the worker's workload per unit time and calculate the labor efficiency.

[0019] Furthermore, step 31 utilizes target detection to identify various construction activities in the video image, specifically including:

[0020] ① Establish a dedicated dataset for construction activities

[0021] For each construction activity in the video image, a sufficient number of images are selected, manually labeled, divided into construction activity zones, and activity types are labeled to build a dedicated dataset for construction activities to meet the training needs of the YOLO model;

[0022] ② Grid division

[0023] First, the input image is divided into a finite number of grids. Each grid is responsible for detecting a part of the object in the image. The size of each grid is:

[0024]

[0025] Where CS represents the size of each grid, W and H represent the length and width of the image respectively, and S represents the number of grids in the W and H directions;

[0026] ③Boundary prediction

[0027] For image data, the YOLO model can predict multiple bounding boxes and corresponding object category probabilities for each grid. For each grid i, it outputs B bounding boxes. The information of each bounding box includes:

[0028] x i ,y i : The coordinates of the center of the bounding box relative to the grid;

[0029] w i ,h i : The width and height of the bounding box, relative to the scale of the entire image;

[0030] c i,j : Category probability, indicating the probability of an object existing in the grid;

[0031] For each bounding box, predict the following information:

[0032] P(Objective): The confidence level of whether the grid contains an object;

[0033] P(Class k ): The probability that the object in the grid belongs to category k;

[0034] ④Model training

[0035] The YOLO model was trained using a dataset dedicated to construction activities. A loss function was used to measure the difference between the model's predicted value and the actual value. The loss function value was reduced by optimizing the model parameters. The loss function consists of three parts: classification loss, which is used to measure the difference between the predicted category and the actual category; localization loss, which is used to measure the difference between the center point and size of the predicted bounding box and the true value; and confidence loss, which is used to measure the difference between the predicted probability of the object existing and the actual value.

[0036]

[0037] Where L cls represents the classification loss, Is an indicator function, which is 1 when the grid (i, j) contains an object, otherwise it is 0; P(Class k )and Represent the predicted category probability and the actual category probability respectively;

[0038]

[0039] Where L coord represents the position loss, λ coord is the weight coefficient used to balance the positioning loss;

[0040]

[0041] Where L conf represents the confidence loss, P(Objective) and ] represent the confidence of whether the prediction contains an object and the confidence of whether the actual object is contained;

[0042] L=L cls +L coord +L conf Formula (5)

[0043] Where L represents the total loss function;

[0044] ⑤Duplicate elimination

[0045] When performing target detection, there may be multiple detection frames that overlap and repeat. The non-maximum suppression algorithm is used to remove duplicate detection frames. By calculating the overlap between bounding boxes, those with high overlap with other boxes are removed, and only the bounding box with the highest score is retained. Finally, the above steps are repeated until all detection frames are processed.

[0046] ⑥ Output results

[0047] The category, location, and confidence of the object in each image will be output, where the category probability is calculated by the Softmax function, and the bounding box is calculated by x i ,y i ,w i ,h i Describes the location of an object.

[0048] Furthermore, step 32 implements pixel-level progress estimation for the identified construction activities through image segmentation, specifically including:

[0049] ① Image preprocessing

[0050] The images of the bounding boxes corresponding to the construction activities identified by target detection are used as input. Each input image is preprocessed, including image resizing and normalization, to adapt it to the input requirements of the network model.

[0051] ② Feature extraction

[0052] A pre-trained convolutional neural network is used as the backbone network to extract high-level features from the input image. For each target region, accurate feature extraction including edge, texture, and shape is performed through RoI operation.

[0053] ③ Object mask generation

[0054] Based on each target area, a pixel-level mask of the target is generated through a fully convolutional network. The mask represents the precise shape of the object. The area with a pixel value of 1 in the mask image corresponds to the object, and the area with a pixel value of 0 corresponds to the background.

[0055] ④Object segmentation and area calculation

[0056] Through the mask output, the area of ​​each target is calculated, and the segmented area of ​​each object is obtained by counting the number of pixels with a value of 1 in the mask;

[0057]

[0058] Where M k is the mask of object k, Mask(x,y) is the mask image, which represents the object pixel area, with a value of 1 representing the object area and 0 representing the background;

[0059] Then, based on the area of ​​the segmented region and the calibration ratio between the size of the video image data corresponding to the type of construction activity and the actual size, the corresponding progress value is calculated:

[0060] W k =Area(M k )·I k Formula (7)

[0061] Where W k Indicates the progress value, that is, the current construction progress, I k It represents the calibration ratio between the size of the image data and the actual size, I k Obtained through calibration when laying out video measurement points.

[0062] Furthermore, step 33 compares the progress at different times to solve the worker's workload per unit time and calculate the labor efficiency, which specifically includes:

[0063] Based on the image acquisition time, acquisition location, construction type, and quantified construction progress, the labor efficiency corresponding to each activity is calculated:

[0064]

[0065] Where t s represents the starting time of labor efficiency monitoring, then t f Indicates the end time of monitoring. and points Specifically, it represents the construction activity k at time t s and t f The corresponding progress value.

[0066] Furthermore, step 4 of establishing the influencing factor-labor efficiency mapping relationship specifically includes:

[0067] Based on the knowledge graph of "influencing factors-labor efficiency", deep learning is used to establish the relationship between workers' physical condition, environmental condition and labor efficiency L k The mapping relationship between them is:

[0068] L k =f(E,R) Formula (9)

[0069] Where E represents the environmental state; R represents the physical condition of the worker.

[0070] The present invention has the following beneficial effects:

[0071] (1) A wireless data acquisition framework based on a bracket-mounted detector is proposed. This framework can realize the data collection of internal and external influencing factors and labor efficiency. It has the characteristics of portability, strong adaptability, high integration, and remote control. It saves manpower consumption in the detection process and realizes the rapid collection and long-term monitoring of relevant raw data at a low cost.

[0072] (2) A labor efficiency recognition method integrating machine vision and a method for constructing the "influencing factor-labor efficiency" mapping relationship combining knowledge graph and deep learning are proposed. This method can realize the intelligent analysis of large amounts of labor efficiency collected data and the accurate characterization of labor efficiency changes, thereby realizing more scientific construction scheduling and avoiding construction delays and resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a schematic diagram of the modules of the acquisition framework of the present invention;

[0074] Figure 2 This is a schematic diagram of a worker wearing a terminal for collecting physical status of a worker according to the present invention;

[0075] Figure 3 It is a schematic diagram of the arrangement of the sensor bracket of the present invention;

[0076] Figure 4 It is a schematic diagram of the arrangement of the acquisition framework of the present invention;

[0077] Figure 5 This is a flow chart of a method for quantitatively characterizing labor efficiency that takes into account the influence of multiple factors.

[0078] In the figure: 1- integrated electronic bracelet; 2- anemometer; 3- wind pressure gauge; 4- thermometer and hygrometer; 5- illuminance meter; 6- solar radiometer; 7- sound intensity meter; 8- mobile smart phone; 9- multi-channel Bluetooth transceiver; 10- solar power panel; 11- mobile wireless action camera; 12- construction work object. DETAILED DESCRIPTION

[0079] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0080] See also Figure 1-5 , an embodiment of the present invention provides a method for quantitatively characterizing labor efficiency taking into account the influence of multiple factors, comprising the following steps:

[0081] Step 1: Data collection: Collect workers' physical condition data, environmental status data, and construction progress data;

[0082] Data collection is mainly used to collect factors that affect workers' construction efficiency during the construction process and construction progress data including workers' workload information. The influencing factors include internal influencing factors (workers' physical condition data) and environmental influencing factors (environmental status data).

[0083] like Figure 1 As shown, workers' physical status data is collected through electronic files and integrated electronic wristbands; environmental status data is collected through an integrated sensing platform; and construction progress data is obtained by collecting video images.

[0084] Specifically, electronic files are used to collect basic worker data, primarily recording information such as height, weight, gender, age, type of work, proficiency, length of service, and health status. These files are recorded when workers join the company and are updated regularly based on the construction cycle.

[0085] like Figure 2 As shown, an integrated electronic wristband is used to collect workers' vital status data, including real-time measurement of workers' body temperature, blood oxygen, blood pressure, heart rate and other data; and the workers' sleep quality data from the previous night is collected before they start working. During the construction process, the workers' cumulative working hours, energy consumption per unit time, construction work movement amplitude and other motion status data are continuously measured.

[0086] An integrated sensing platform is used to collect environmental status data. The integrated sensing platform includes a sensor bracket and several sensors. The sensor bracket is a telescopic bracket. The sensors are fixed to the sensor bracket by snaps. The number of sensor snaps can be adaptively adjusted according to different influencing factors of concern to achieve adjustment of the type and quantity of sensors.

[0087] like Figure 3 As shown, based on the study of sensitive factors affecting labor efficiency, the sensors in this embodiment include an anemometer 2, an anemometer 3, a thermometer and hygrometer 4, an illuminance meter 5, a solar radiometer 6, and a sound intensity meter 7. The above sensors can all be connected via Bluetooth and have independent batteries. They are connected to a smart mobile phone 8 via a multi-channel Bluetooth transceiver 9 fixed on the sensor bracket to achieve collaborative control through the smart mobile phone 8. In addition, the smart mobile phone 8 can be controlled through remote multi-terminals. In order to ensure uninterrupted power supply to each sensor, a solar panel 10 is used as a backup power supply. The environmental status data includes wind speed, wind pressure, temperature, humidity, light intensity, solar radiation, noise, etc. collected by the sensors,

[0088] like Figure 4As shown, construction progress data is collected by several mobile wireless motion cameras 11 installed at the construction site. The video images captured by the mobile wireless motion cameras 11 are transmitted in real time to a smart phone 8 via Bluetooth. The location and number of mobile wireless motion cameras 11 should ensure that all workers in the construction scene are within their field of view.

[0089] Step 2: The collected worker health data, environmental data, and construction progress data are annotated based on the collection time, collection location, and construction type, and the data sets are stored in the database. The data is categorized and stored as structured data (worker health data, environmental data) and unstructured image data (construction progress data).

[0090] Step 3: Identify the workers' workload and calculate the labor efficiency based on the collected construction progress data, and synchronously mark the label information of the construction progress data and the labor efficiency data.

[0091] Specifically, based on the collected construction progress data, target detection and progress estimation are used to identify the workload of workers and calculate labor efficiency, such as Figure 5 The specific steps are as follows:

[0092] Step 31, target detection: Use target detection to identify various construction activities in the video image; it mainly includes the steps of establishing a dedicated dataset for construction activities, grid division, boundary prediction, model training, duplicate elimination, and outputting results.

[0093] Through manual labeling, various construction activities in the video images are annotated to establish a dedicated dataset for construction activities. The images to be identified are partitioned through grid division, and the YOLO model is used to predict the boundaries of the targets to be identified. Considering that the general YOLO model is not applicable to the identification of construction progress targets, the present invention uses a dedicated dataset for construction progress targets to train the YOLO model. After training, the non-maximum suppression algorithm is used to eliminate the phenomenon of cross-repetition of detection frames. Finally, the YOLO model can accurately output the output of the construction progress target.

[0094] This embodiment of the present invention utilizes an end-to-end object detection algorithm that simultaneously completes object classification, localization, and bounding box regression tasks in an image through a single forward propagation. This algorithm transforms the object detection problem into a regression problem, using the YOLO model to predict the object's category and location information. Object detection can be used to identify various construction activities and corresponding work areas at a construction site using video images. Step 21 specifically includes:

[0095] ① Establish a dedicated dataset for construction activities

[0096] For each construction activity in the video image, a sufficient number of images are selected, and manual annotation is used to divide the images into construction activity zones and label the activity types to construct a special dataset for construction activities to meet the training requirements of the YOLO model.

[0097] ② Grid division

[0098] First, the input image is divided into a finite number of grids, each of which is responsible for detecting a portion of the object in the image. The size of each grid is:

[0099]

[0100] Where CS represents the size of each grid, W and H represent the length and width of the image respectively, and S represents the number of grids in the W and H directions.

[0101] ③Boundary prediction

[0102] For image data, the YOLO model can predict multiple bounding boxes and corresponding object category probabilities for each grid. For each grid i, the algorithm can output B bounding boxes. The information of each bounding box includes:

[0103] x i ,y i : The coordinates of the center of the bounding box relative to the grid (in the range [0,1])

[0104] w i ,h i : The width and height of the bounding box, relative to the scale of the entire image (range [0,1])

[0105] c i,j : Category probability, indicating the probability of an object existing in the grid.

[0106] For each bounding box, the algorithm predicts the following information:

[0107] P(Objective): The confidence level of whether the grid contains an object (this value is multiplied by the category probability to indicate the confidence level of the object).

[0108] P(Class k ): The probability that the object in the grid belongs to category k.

[0109] ④Model training

[0110] The YOLO model was trained using a dataset dedicated to construction activities. A loss function was used to measure the difference between the model's predictions and the actual values. The loss function was then optimized to reduce the value of the loss function. The loss function consists of three components: classification loss (the difference between the predicted category and the actual category); localization loss (the difference between the center point and size of the predicted bounding box and the true value); and confidence loss (the difference between the predicted probability of an object and the actual value).

[0111]

[0112] Where L cls represents the classification loss, Is an indicator function, which is 1 when the grid (i, j) contains an object, otherwise it is 0. k )and represent the predicted category probability and the actual category probability respectively.

[0113]

[0114] Where L coord represents the position loss, λ coord is the weight coefficient used to balance the positioning loss.

[0115]

[0116] Where L conf represents the confidence loss, P(Objective) and ] represent the confidence of whether the prediction contains an object and the actual confidence of whether the object is contained.

[0117] L=L cls +L coord +L conf Formula (5)

[0118] Where L represents the total loss function.

[0119] ⑤Duplicate elimination

[0120] During object detection, multiple detection boxes may overlap and overlap. This algorithm uses non-maximum suppression to remove duplicate detection boxes. By calculating the overlap between bounding boxes, it removes those with high overlap with other boxes, retaining only the ones with the highest scores. It sorts all detection results by predicted score. It then selects the bounding box with the highest score. It then removes any other boxes whose overlap with that box exceeds a threshold. Finally, it repeats these steps until all detection boxes have been processed.

[0121] ⑥ Output results

[0122] The algorithm will output the category, location and confidence of each object in the image. The category is calculated by the Softmax function to get the category probability. The bounding box is calculated by x i ,y i ,w i ,h i Describes the location of an object.

[0123] Step 32: For the identified construction activities, image segmentation is used to achieve progress estimation based on the image pixel level.

[0124] The present invention uses a deep learning algorithm to achieve pixel-level object segmentation based on target detection. By taking the image of the bounding box range corresponding to each construction activity identified by target detection as input, high-level features (such as edges, textures, shapes, etc.) in the image are extracted, and then a pixel-level mask is accurately generated for each object, and finally the progress is estimated through the mask. Through this operation, the progress information of each construction activity in the video image can be obtained, providing data support for the subsequent calculation of labor efficiency. Step 22 specifically includes:

[0125] ① Image preprocessing

[0126] Since the size of each input image may be different, it is preprocessed, including adjusting the image size and normalization, to adapt it to the input requirements of the network model.

[0127] ② Feature extraction

[0128] A pretrained convolutional neural network is used as the backbone network to extract high-level features from the input image. For each target region, precise feature extraction is performed using RoI operations. RoIs use bilinear interpolation to accurately calculate the value within each subregion, avoiding quantization errors caused by max pooling and preserving detailed information about the target region.

[0129] ③ Object mask generation

[0130] Based on each target region, a pixel-level mask of the target is generated through a fully convolutional network. The mask represents the precise shape of the object. The areas with pixel values ​​of 1 in the mask image correspond to the object, and the areas with pixel values ​​of 0 correspond to the background.

[0131] ④Object segmentation and area calculation

[0132] The mask output can be used to calculate the area of ​​each target. By counting the number of pixels with a value of 1 in the mask, the area of ​​the segmented region of each object can be obtained.

[0133]

[0134] Where Mk is the mask of object k, Mask(x,y) is the mask image, which represents the object pixel area. The value 1 represents the object area, and 0 represents the background.

[0135] Based on the area of ​​the segmented region and the calibration ratio between the size of the video image data and the actual size corresponding to the type of construction activity, the corresponding progress value can be calculated:

[0136] W k =Area(M k )·I k Formula (7)

[0137] Where W k Indicates the progress value, that is, the current construction progress. k It represents the calibration ratio between the size of the image data and the actual size. This data is obtained through calibration when laying out the video measurement points.

[0138] Step 33: By comparing the progress at different times, solve the worker's workload per unit time and calculate the labor efficiency.

[0139] Because video image monitoring points are fixed, the difference between progress values ​​at different times represents the worker's workload, and by comparing this with time, the corresponding labor efficiency can be calculated. Based on the quantified construction progress, the labor efficiency corresponding to each activity can be calculated based on the image acquisition time, location, and construction type.

[0140]

[0141] Where t s represents the starting time of labor efficiency monitoring, then t f Indicates the end time of monitoring. and points Specifically, it represents the construction activity k at time t s and t f The corresponding progress value.

[0142] Through the above process, the unstructured video images are converted into structured labor efficiency data, and the collection time, collection location, construction type and other annotation information of the original image video data (construction progress data) are synchronously annotated with the labor efficiency data.

[0143] Step 3: Align the data time scale and build the "influencing factors-labor efficiency" knowledge graph

[0144] Worker physical condition data, environmental condition data, and labor efficiency data with identical labeling information were aligned based on collection time, location, and construction type. Furthermore, using the label information as entities and "worker physical condition data, environmental condition data - labor efficiency data" as attributes, a knowledge graph of "influencing factors - labor efficiency" was constructed.

[0145] Step 4: Establish a mapping relationship between influencing factors and labor efficiency

[0146] Based on the knowledge graph of "influencing factors-labor efficiency", deep learning is used to establish the relationship between workers' physical condition, environmental condition and labor efficiency L k The mapping relationship between them.

[0147] L k =f(E,R) Formula (9)

[0148] Where E represents the environmental conditions, including wind speed, wind pressure, temperature, humidity, light intensity, solar radiation, and noise; R represents the worker's physical condition, including basic status data, vital sign status data, sleep quality data, and exercise status data.

[0149] The present invention establishes a labor efficiency measurement process, forms a labor efficiency measurement method that takes into account the physical condition of workers and the influence of the external environment, proposes an integrated multi-influencing factor data collection framework, realizes the synchronous collection of labor efficiency and influencing factor data, and proposes an influencing factor-labor efficiency mapping relationship characterization method based on the combination of knowledge graph and deep learning, which provides a methodological support for the scientific quantification of labor efficiency changes in actual construction.

[0150] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for quantitatively characterizing labor efficiency considering the influence of multiple factors, characterized by: The steps include: Step 1: Data collection: Collect workers' physical condition data, environmental status data, and construction progress data; Step 2: Label the collected worker physical condition data, environmental condition data, and construction progress data according to label information, including collection time, collection location, and construction type; Step 3: Identify the worker workload and calculate labor efficiency based on the collected construction progress data, and simultaneously annotate the label information of the construction progress data with the labor efficiency data; Step 4: Align the worker's physical condition data, environmental condition data, and labor efficiency data with the same label information according to the label information. Use the label information as the entity and "worker's physical condition data, environmental condition data - labor efficiency data" as the attribute to create an "influencing factors - labor efficiency" knowledge graph. Step 5: Based on the "influencing factors-labor efficiency" knowledge graph, use deep learning to establish a mapping relationship between workers' physical condition, environmental factors, and labor efficiency.

2. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 1, characterized in that: The workers' physical condition data is collected through electronic files and integrated electronic wristbands; the environmental condition data is collected through an integrated sensing platform; and the construction progress data is obtained through video images collected by a mobile wireless motion camera.

3. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 2, characterized in that: The worker's physical condition data includes basic condition data, vital sign condition data, sleep quality data and motion condition data. The worker's basic condition data is collected using electronic files. The basic condition data includes the worker's height, weight, gender, age, job type, proficiency, length of service and health status; the worker's vital sign condition data is collected using an integrated electronic bracelet. The vital sign condition data includes the worker's body temperature, blood oxygen, blood pressure and heart rate; the worker's sleep quality data is collected the night before work, and the motion condition data is continuously measured during the construction process. The motion condition data includes the worker's cumulative working hours, energy consumption per unit time and the motion amplitude of the construction operation.

4. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 2, characterized in that: The integrated sensing platform includes a sensor bracket and several sensors. The sensor bracket is a telescopic bracket. The sensors are fixed to the sensor bracket by snaps. The sensors include an anemometer, an anemometer, a thermometer and a hygrometer, an illuminometer, a solar radiometer, and a sound intensity meter. The sensors are connected to a smart mobile phone through a multi-channel Bluetooth transceiver fixed on the sensor bracket to achieve collaborative control through the smart mobile phone. The environmental status data includes wind speed, wind pressure, temperature, humidity, light intensity, solar radiation, and noise collected by the sensors.

5. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 1, characterized in that: In step 3, the worker workload is identified and labor efficiency is calculated based on the collected construction progress data, which specifically includes: Step 31: Using target detection, identify various construction activities in the video image; Step 32: for the identified construction activities, image segmentation is used to achieve pixel-level progress estimation; Step 33: By comparing the progress at different times, solve the worker's workload per unit time and calculate the labor efficiency.

6. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 5, characterized in that: Step 31 uses target detection to identify various construction activities in the video image, including: ① Establish a dedicated dataset for construction activities For each construction activity in the video image, a sufficient number of images are selected, manually labeled, divided into construction activity zones, and activity types are labeled to build a dedicated dataset for construction activities to meet the training needs of the YOLO model; ② Grid division First, the input image is divided into a finite number of grids. Each grid is responsible for detecting a part of the object in the image. The size of each grid is: Where CS represents the size of each grid, W and H represent the length and width of the image respectively, and S represents the number of grids in the W and H directions; ③Boundary prediction For image data, the YOLO model can predict multiple bounding boxes and corresponding object category probabilities for each grid. For each grid i, it outputs B bounding boxes. The information of each bounding box includes: x i ,y i : The coordinates of the center of the bounding box relative to the grid; w i ,h i : The width and height of the bounding box, relative to the scale of the entire image; c i,j : Category probability, indicating the probability of an object existing in the grid; For each bounding box, predict the following information: P(Objective): The confidence level of whether the grid contains an object; P(Class k ): The probability that the object in the grid belongs to category k; ④Model training The YOLO model was trained using a dataset dedicated to construction activities. A loss function was used to measure the difference between the model's predicted value and the actual value. The loss function value was reduced by optimizing the model parameters. The loss function consists of three parts: classification loss, which is used to measure the difference between the predicted category and the actual category; localization loss, which is used to measure the difference between the center point and size of the predicted bounding box and the true value; and confidence loss, which is used to measure the difference between the predicted probability of the object existing and the actual value. Where L cls represents the classification loss, Is an indicator function, which is 1 when the grid (i, j) contains an object, otherwise it is 0; P(Class k )and Represent the predicted category probability and the actual category probability respectively; Where L coord represents the position loss, λ coord is the weight coefficient used to balance the positioning loss; Where L conf represents the confidence loss, P(Objective) and Respectively represent the confidence of whether the prediction contains an object and the confidence of the actual inclusion of the object; L=L cls +L coord +L conf Formula (5) Where L represents the total loss function; ⑤Duplicate elimination When performing target detection, there may be multiple detection frames that overlap and repeat. The non-maximum suppression algorithm is used to remove duplicate detection frames. By calculating the overlap between bounding boxes, those with high overlap with other boxes are removed, and only the bounding box with the highest score is retained. Finally, the above steps are repeated until all detection frames are processed. ⑥ Output results The category, location, and confidence of the object in each image will be output, where the category probability is calculated by the Softmax function, and the bounding box is calculated by x i ,y i ,w i ,h i Describes the location of an object.

7. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 5, characterized in that: Step 32 implements pixel-level progress estimation for the identified construction activities through image segmentation, specifically including: ① Image preprocessing The images of the bounding boxes corresponding to the construction activities identified by target detection are used as input. Each input image is preprocessed, including image resizing and normalization, to adapt it to the input requirements of the network model. ② Feature extraction A pre-trained convolutional neural network is used as the backbone network to extract high-level features from the input image. For each target region, accurate feature extraction including edge, texture, and shape is performed through RoI operation. ③Object mask generation Based on each target area, a pixel-level mask of the target is generated through a fully convolutional network. The mask represents the precise shape of the object. The area with a pixel value of 1 in the mask image corresponds to the object, and the area with a pixel value of 0 corresponds to the background. ④Object segmentation and area calculation Through the mask output, the area of ​​each target is calculated, and the segmented area of ​​each object is obtained by counting the number of pixels with a value of 1 in the mask; Where M k is the mask of object k, Mask(x,y) is the mask image, which represents the object pixel area, with a value of 1 representing the object area and 0 representing the background; Then, based on the area of ​​the segmented region and the calibration ratio between the size of the video image data corresponding to the type of construction activity and the actual size, the corresponding progress value is calculated: W k = Area(M k )·I k Equation (7) Where W k Indicates the progress value, that is, the current construction progress, I k It represents the calibration ratio between the size of the image data and the actual size, I k Obtained through calibration when laying out video measurement points.

8. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 7, characterized in that: Step 33 compares the progress at different times to determine the worker's workload per unit time and calculate labor efficiency, specifically including: Based on the image acquisition time, acquisition location, construction type, and quantified construction progress, the labor efficiency corresponding to each activity is calculated: Where t s represents the starting time of labor efficiency monitoring, then t f Indicates the end time of monitoring. and points Specifically, it represents the construction activity k at time t s and t f The corresponding progress value.

9. The method for quantitatively characterizing labor efficiency considering the influence of multiple factors according to claim 8, characterized in that: Step 4: Establishing the mapping relationship between influencing factors and labor efficiency includes: Based on the knowledge graph of "influencing factors-labor efficiency", deep learning is used to establish the relationship between workers' physical condition, environmental condition and labor efficiency L k The mapping relationship between them: L k =f(E,R) Formula (9) Where E represents the environmental state; R represents the physical condition of the worker.

Images