Safety monitoring equipment based on building construction site and monitoring system thereof

Abstract

The invention discloses safety monitoring equipment based on a building construction site and a monitoring system thereof, and relates to the field of building construction.The safety monitoring equipment comprises a user side, a monitoring side and a terminal, the user side is used for obtaining position data of building constructors, and the monitoring side is used for collecting and obtaining image data of the construction site; and the terminal is in communication connection with the user side and the monitoring side. According to the invention, the working safety risk coefficients of different building constructors can be obtained by analyzing the working environments of different building constructors, then the data information of each building constructor in the building construction site is collected and obtained in real time, and the data information of each building constructor in the building construction site is obtained by combining the working safety risk coefficients of the building constructors. According to the method, the danger when different building constructors are located in different risk areas and make different dangerous behavior postures is analyzed, and then early warning instructions of different grades are made according to the danger, so that the effect of monitoring the dynamic states of the building constructors is achieved, and the safety of building construction operation is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of building construction safety monitoring, and in particular to a safety monitoring device based on a building construction site and a monitoring system thereof. Background Art

[0002] Safety monitoring is required during construction to ensure the safety of workers during construction; For example, the authorization announcement number is CN113691719B, the authorization announcement date is 2023.03.28, and the name is a construction safety monitoring system, including a support frame and a monitoring camera, the support frame is fixedly provided with a monitoring camera, and the monitoring camera is provided with a protective component outside for protection...; As in the above application, in the existing safety monitoring equipment for construction sites, most of them simply use cameras to monitor the construction sites, and then manually observe and monitor, and manually issue early warnings when construction workers exhibit dangerous behaviors. This early warning method is not only ineffective, but also has many interference factors, which not only wastes manpower, but also affects the monitoring effect. In addition, in the existing safety monitoring system, the monitoring standards for all personnel are unified and poorly targeted. Summary of the invention

[0003] 1. Purpose of the invention In view of this, the purpose of the present invention is to propose a safety monitoring device and a monitoring system based on a construction site, which can achieve the effect of "automatic monitoring, targeted monitoring, and automatic early warning", which not only reduces labor costs, but also improves monitoring accuracy and intelligence, and greatly improves the safety of construction operations.

[0004] (II) Technical solution In order to achieve the above technical objectives, the present invention provides a safety monitoring device based on a construction site, which includes a user end, a monitoring end and a terminal, wherein: The configuration of the user end is: carried by construction workers to obtain location data of construction workers and issue early warnings; The configuration of the monitoring terminal is as follows: it is installed at the construction site to collect and obtain image data of the construction site; The terminal is configured to: be communicatively connected with the user terminal and the monitoring terminal; The terminal is further configured to: divide the construction site into risk areas of multiple safety levels; Based on the risk areas divided at the construction site and combined with the collected work data of construction workers, the work safety risk coefficient of construction workers is generated; Based on the work safety risk coefficient of construction workers and their current behavior, early warning instructions are generated.

[0005] As a further description of the above technical solution: the method in which the terminal divides the construction site into risk areas of multiple safety levels is as follows: Retrieve construction site drawings; The danger level of each location in the construction site map is evaluated. Based on the danger level of each location in the construction site map, the entire construction site map is divided into risk areas with three safety levels, namely: low-risk area, medium-risk area and high-risk area, among which the danger levels of low-risk area, medium-risk area and high-risk area gradually increase.

[0006] As a further description of the above technical solution: the method for generating the work safety risk coefficient of construction workers by the terminal based on the risk area divided by the construction site and combining the collected work data of the construction workers is as follows: Collect historical work data of construction workers, including the average daily working hours of construction workers and the average daily time that construction workers are in each risk area; Based on the historical work data of construction workers, the work safety risk coefficient of construction workers is analyzed and evaluated as follows: ; in, is the work safety risk factor for construction workers, is the average daily working hours of construction workers. are the average daily time that construction workers spend in high-risk areas and medium-risk areas, respectively. is the weight coefficient, Both are greater than 0.

[0007] As a further description of the above technical solution: the terminal generates a warning instruction based on the work safety risk coefficient of the construction personnel and the current behavior of the construction personnel in the following specific steps: Collect the behavior postures of construction workers and determine whether the current behavior posture is a dangerous behavior posture; When the current behavior posture is determined to be a dangerous behavior posture, a danger value of the dangerous behavior posture is calculated; Generate safety hazard values ​​based on the risk values ​​of dangerous behavior postures, the work safety risk coefficients of construction workers, and the areas where construction workers are located; Generate corresponding early warning instructions based on the safety hazard value.

[0008] As a further description of the above technical solution: the method for judging whether the current behavior posture of the construction worker is a dangerous behavior posture is as follows: Take photos of construction workers in their current postures; The photo is uploaded to the first machine learning model to identify and determine whether the current behavior posture of the construction worker is a dangerous behavior posture.

[0009] As a further description of the above technical solution: the training method of the first machine learning model is as follows: X groups of historical photos are collected, and labels are set for the photos, where the labels are that the behavior postures of construction workers in the photos are dangerous behavior postures; the photos and the labels corresponding to the photos constitute a group of training data; the X groups of training data constitute a sample set, where X is an integer greater than 1; and the sample set is divided into a training set and a test set; the photos in the training set are used as input of a first machine learning model, and the labels in the training set are used as output of the first machine learning model; the first machine learning model is trained to obtain a dangerous behavior posture judgment model, and minimizing the sum of prediction errors is used as the training goal; the dangerous behavior posture judgment model is evaluated using the test set, and the dangerous behavior posture judgment model when the sum of prediction errors reaches convergence is used as the constructed first machine learning model.

[0010] As a further description of the above technical solution: when the current behavior posture is judged to be a dangerous behavior posture, the method for calculating the danger value of the dangerous behavior posture is as follows: The dangerous behavior posture photo containing the dangerous behavior posture is uploaded to the second machine learning model to calculate the danger value of the dangerous behavior posture, wherein the training method of the second machine learning model is as follows: X groups of historical dangerous behavior posture photos are collected, and labels are set for the dangerous behavior posture photos, wherein the labels are the danger values ​​of the dangerous behavior postures of the construction workers in the photos; the dangerous behavior posture photos and the labels corresponding to the dangerous behavior posture photos constitute a group of training data; the X groups of training data constitute a sample set, where X is an integer greater than 1, and the sample set is divided into a training set and a test set; the photos in the training set are used as input of a second machine learning model, and the labels in the training set are used as output of the second machine learning model, and the second machine learning model is trained to obtain a dangerous behavior posture danger value calculation model, wherein minimizing the sum of prediction errors is the training goal, and the dangerous behavior posture danger value calculation model is evaluated using the test set, and the dangerous behavior posture danger value calculation model when the sum of prediction errors reaches convergence is used as the constructed second machine learning model.

[0011] As a further description of the above technical solution: Based on the risk value of the dangerous behavior posture, the work safety risk coefficient of the construction workers, and the area where the construction workers are located, the method for generating the safety hazard value is as follows: ; in, is the safety risk value, is the correction factor, is the risk value corresponding to the dangerous behavior posture, It is the work safety risk factor for construction workers.

[0012] As a further description of the above technical solution: Based on the potential safety hazard value, the method for generating an early warning instruction is as follows: The preset safety hazard value threshold is ,in ; like , a first warning instruction is generated, and the user terminal issues an alarm according to the first warning instruction; like , a second warning instruction is generated, the user terminal issues an alarm according to the second warning instruction, and the terminal records and uploads the dangerous behavior posture of the construction personnel; like , a third early warning instruction is generated, the user end issues an alarm according to the third early warning instruction, the terminal records and uploads the dangerous behavior posture of the construction personnel, and notifies the superior management department of the construction personnel at the same time.

[0013] Specifically, the present invention further discloses a monitoring system based on safety monitoring equipment of a construction site, which comprises: Data acquisition module: used to collect image data of the construction site and location data of construction workers; Data processing module: used to parse and process the image data of the construction site and the location data of the construction workers, obtain the behavior posture of the construction workers based on the analysis results, and generate early warning instructions based on the danger of the dangerous behavior posture when the behavior posture of the construction workers is a dangerous behavior posture; Early warning module: used to issue early warnings based on generated early warning instructions.

[0014] In the above technical scheme, the present invention provides a safety monitoring device and a monitoring system based on a construction site, which analyzes the working environment of different construction workers to obtain the work safety risk coefficients of different construction workers, and then collects and obtains the data information of each construction worker in the construction site in real time, that is, the location information and behavior posture of the construction workers, and analyzes the dangers of different construction workers in different risk areas and when they make different dangerous behavior postures by combining the work safety risk coefficients of the construction workers, and then makes different levels of early warning instructions according to the degree of danger. It can be seen that the monitoring system has a strong personnel targeting, thereby achieving the effect of "automatic monitoring, targeted monitoring, and automatic early warning", which not only reduces labor costs, but also improves monitoring accuracy and intelligence, thereby greatly improving the safety of construction operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0016] Figure 1 An overall schematic diagram of a safety monitoring device based on a construction site provided by the present invention; Figure 2 A specific flow chart of generating early warning instructions based on a safety monitoring device at a construction site provided by the present invention; Figure 3 A system block diagram of a monitoring system based on safety monitoring equipment at a construction site provided by the present invention. DETAILED DESCRIPTION

[0017] The following description is merely exemplary in nature and is not intended to limit the present disclosure, application, and use. It should be understood that in all of these figures, the same or similar reference numerals indicate the same or similar parts and features. The various drawings only schematically represent the concepts and principles of the embodiments of the present disclosure, and do not necessarily show the specific dimensions and proportions of the various embodiments of the present disclosure. Specific parts in specific drawings may be exaggerated to illustrate the relevant details or structures of the embodiments of the present disclosure.

[0018] Example 1 Reference Figure 1 , Figure 2 :This embodiment provides a technical solution: a safety monitoring device based on a construction site, which includes a user end, a monitoring end and a terminal, wherein: The configuration of the user end is: carried by construction workers, used to obtain location data of construction workers, and equipped with an early warning device, such as a buzzer alarm, used to issue an early warning when construction workers exhibit dangerous behavior (for example, it can be located by GPS, and optionally, built into the user's mobile phone in the form of software); The configuration of the monitoring terminal is as follows: installed at the construction site, used to collect image data of the construction site (by collecting and obtaining images of various locations at the construction site through multiple high-definition cameras installed at the construction site); The terminal is configured to: be communicatively connected with the user terminal and the monitoring terminal; The terminal is further configured to: divide the construction site into risk areas of multiple safety levels; Based on the risk areas divided at the construction site and combined with the collected work data of construction workers, the work safety risk coefficient of construction workers is generated; Based on the work safety risk coefficient of construction workers and their current behavior, early warning instructions are generated.

[0019] This embodiment provides a safety monitoring device based on a construction site and a monitoring system thereof. Specifically, the terminal divides the construction site into risk areas of multiple safety levels as follows: Retrieve construction site drawings; Assess the danger level of each location in the construction site map, and based on the danger level of each location in the construction site map (the danger level assessment method for each location in the construction site map is as follows: mark the dangerous points and dangerous areas in the construction site map, where the dangerous points include but are not limited to: foundation pits, collision-prone points, processing points; the dangerous areas include but are not limited to: high-altitude fall coverage areas, areas with falling risks, and the danger level gradually decreases as the distance from the dangerous points and dangerous areas increases), divide the entire construction site map into three safety levels of risk areas, namely: low-risk areas, medium-risk areas, and high-risk areas (mark low-risk areas, medium-risk areas, and high-risk areas in the construction site map with different colors, such as high-risk areas within 1 meter from dangerous points and dangerous areas, medium-risk areas within 1-3 meters from dangerous points and dangerous areas, and low-risk areas outside 3 meters from dangerous points and dangerous areas), where the danger levels of low-risk areas, medium-risk areas, and high-risk areas gradually increase.

[0020] This embodiment provides a safety monitoring device based on a construction site and a monitoring system thereof. Specifically, the terminal generates a work safety risk coefficient of construction workers based on the risk areas divided at the construction site and the collected work data of the construction workers as follows: Collect historical work data of construction workers, including the average daily working hours of construction workers and the average daily time that construction workers are in each risk area; Based on the historical work data of construction workers, the work safety risk coefficient of construction workers is analyzed and evaluated as follows: ; in, It is the work safety risk coefficient of construction workers, which is used to measure the work vigilance of construction workers. is the average daily working hours of construction workers. are the average daily time that construction workers spend in high-risk areas and medium-risk areas, respectively. is the weight coefficient, All are greater than 0, where the weight factor is obtained by a technician in this field who collects multiple sets of comprehensive parameters and sets corresponding weight factors for each set of comprehensive parameters. The set weight factors and the collected comprehensive parameters are substituted into the formula. Any three formulas constitute a three-variable linear equation system. The calculated weight factors are screened and averaged to obtain The mean of , optionally, =1, is 0.52, It is 0.48.

[0021] This embodiment provides a safety monitoring device and a monitoring system based on a construction site. Specifically, the terminal generates a warning instruction based on the work safety risk coefficient of the construction personnel and the current behavior of the construction personnel. The specific steps are as follows: Collect the behavior and posture of construction workers to determine whether the current behavior and posture are dangerous. Specifically, the dangerous behavior and posture include but are not limited to: not wearing personal protective equipment (such as not wearing a hard hat, safety shoes, goggles, earplugs, gloves, etc.), using open flames, overextending or leaning the body, operating heavy tools with one hand, keeping the head down or bent for a long time, sitting on unstable or inappropriate surfaces (such as edges, boxes, etc.); When the current behavior posture is determined to be a dangerous behavior posture, a danger value of the dangerous behavior posture is calculated; Generate safety hazard values ​​based on the risk values ​​of dangerous behavior postures, the work safety risk coefficients of construction workers, and the areas where construction workers are located; Generate corresponding early warning instructions based on the safety hazard value.

[0022] This embodiment provides a safety monitoring device and a monitoring system based on a construction site. Specifically, a method for determining whether the current behavior posture of a construction worker is a dangerous behavior posture is as follows: Take photos of construction workers in their current postures; The photo is uploaded to the first machine learning model to identify and determine whether the current behavior posture of the construction worker is a dangerous behavior posture.

[0023] This embodiment provides a safety monitoring device based on a construction site and a monitoring system thereof. Specifically, the training method of the first machine learning model is as follows: X groups of historical photos are collected, and labels are set for the photos, where the labels are that the behavior postures of construction workers in the photos are dangerous behavior postures; the photos and the labels corresponding to the photos constitute a group of training data; the X groups of training data constitute a sample set, where X is an integer greater than 1; and the sample set is divided into a training set and a test set; the photos in the training set are used as input of a first machine learning model, and the labels in the training set are used as output of the first machine learning model; the first machine learning model is trained to obtain a dangerous behavior posture judgment model, and minimizing the sum of prediction errors is used as the training goal; the dangerous behavior posture judgment model is evaluated using the test set, and the dangerous behavior posture judgment model when the sum of prediction errors reaches convergence is used as the constructed first machine learning model, and the first machine learning model is a CNN model.

[0024] This embodiment provides a safety monitoring device and a monitoring system based on a construction site. Specifically, when the current behavior posture is judged to be a dangerous behavior posture, the method for calculating the danger value of the dangerous behavior posture is as follows: The dangerous behavior posture photo containing the dangerous behavior posture is uploaded to the second machine learning model to calculate the danger value of the dangerous behavior posture, wherein the training method of the second machine learning model is as follows: X groups of historical dangerous behavior posture photos are collected, and labels are set for the dangerous behavior posture photos, wherein the labels are the danger values ​​of the dangerous behavior postures of the construction workers in the photos; the dangerous behavior posture photos and the labels corresponding to the dangerous behavior posture photos constitute a group of training data; the X groups of training data constitute a sample set, where X is an integer greater than 1, and the sample set is divided into a training set and a test set; the photos in the training set are used as the input of a second machine learning model, and the labels in the training set are used as the output of the second machine learning model, and the second machine learning model is trained to obtain a dangerous behavior posture danger value calculation model, and minimizing the sum of prediction errors is the training goal, and the dangerous behavior posture danger value calculation model is evaluated using the test set, and the dangerous behavior posture danger value calculation model when the sum of prediction errors reaches convergence is used as the constructed second machine learning model, and the second machine learning model is a CNN model.

[0025] This embodiment provides a safety monitoring device and a monitoring system based on a construction site. Specifically, based on the danger value of dangerous behavior posture, the work safety risk coefficient of construction workers, and the area where the construction workers are located, a method for generating a safety hazard value is as follows: ; in, is the safety risk value, is the correction factor, is the risk value corresponding to the dangerous behavior posture, is the work safety risk coefficient of the construction workers, where the work safety risk coefficient of the construction workers is inversely proportional to the safety hazard value, that is, the higher the work safety risk coefficient of the construction workers, the higher the work vigilance of the construction workers. Therefore, relative to those with lower work safety risk coefficients, their safety hazard values ​​are lower.

[0026] It should be noted that the correction factor The correction factor is set by technicians in this field after comprehensive consideration in order to quantify the risk of different dangerous behaviors in low-risk areas, medium-risk areas, and high-risk areas. For example, dangerous behavior: not wearing personal protective equipment. Since the risk of this dangerous behavior in a low-risk area is low, the correction factor is The value of is small, optionally, =0.55, and the risk of the risk behavior in the medium-risk area increases. At this time, the correction factor The value of increases accordingly, for example, using open flames, etc., optionally, =0.85. Similarly, the risk of this dangerous behavior in the high-risk area is further increased. At this time, the correction factor The value of increases, optionally, =1.

[0027] This embodiment provides a safety monitoring device and a monitoring system based on a construction site. Specifically, a method for generating a warning instruction based on a safety hazard value is as follows: The preset safety hazard value threshold is ,in The preset safety hazard value threshold is the basis for dividing and generating the first warning instruction, the second warning instruction and the third warning instruction. The setting of the safety hazard value threshold is that the technical personnel in this field collect multiple groups of different dangerous behaviors in historical situations, and combine the safety hazard values ​​corresponding to the dangerous behaviors, so that the safety hazard value can affect the severity setting of the warning instruction; like , a first warning instruction is generated, and the user terminal issues an alarm according to the first warning instruction; like , a second warning instruction is generated, the user terminal issues an alarm according to the second warning instruction, and the terminal records and uploads the dangerous behavior posture of the construction personnel; like , a third early warning instruction is generated, the user end issues an alarm according to the third early warning instruction, the terminal records and uploads the dangerous behavior posture of the construction personnel, and notifies the superior management department of the construction personnel at the same time.

[0028] To sum up: in this embodiment, the safety monitoring equipment of the construction site can analyze the working environment of different construction workers, obtain the work safety risk coefficient of different construction workers, and then collect and obtain the data information of each construction worker in the construction site in real time, that is, the location information and behavior posture of the construction workers, and analyze the danger of different construction workers in different risk areas and when they make different dangerous behavior postures by combining the work safety risk coefficient of the construction workers, and then make different levels of early warning instructions according to the degree of danger, thereby achieving the effect of monitoring the dynamics of construction workers and greatly improving the safety of construction operations.

[0029] Example 2 Reference Figure 3 :This embodiment provides a technical solution: a monitoring system based on safety monitoring equipment at a construction site, which includes a data acquisition module, a data processing module and an early warning module, wherein the data acquisition module is used to collect image data of the construction site and location data of construction personnel, the data processing module is used to parse and process the image data of the construction site and location data of construction personnel, and obtain the behavior posture of the construction personnel based on the analysis results, and when the behavior posture of the construction personnel is a dangerous behavior posture, a warning instruction is generated based on the dangerousness of the dangerous behavior posture, and the early warning module is used to issue an early warning according to the generated early warning instruction.

[0030] The exemplary implementation scheme of the present disclosure is described in detail above with reference to the preferred embodiments. However, it can be understood by those skilled in the art that, without departing from the concept of the present disclosure, various modifications and variations can be made to the above-mentioned specific embodiments, and various technical features and structures proposed in the present disclosure can be combined in various ways without exceeding the protection scope of the present disclosure, which is determined by the attached claims.

Claims

1. A safety monitoring device based on a construction site, characterized in that: It includes user end, monitoring end and terminal, among which: The configuration of the user end is: carried by construction workers to obtain location data of construction workers and issue early warnings; The configuration of the monitoring terminal is as follows: it is installed at the construction site to collect and obtain image data of the construction site; The terminal is configured to: be communicatively connected with the user terminal and the monitoring terminal; The terminal is further configured to: divide the construction site into risk areas of multiple safety levels; Based on the risk areas divided at the construction site and combined with the collected work data of construction workers, the work safety risk coefficient of construction workers is generated; Based on the work safety risk coefficient of construction workers and their current behavior, early warning instructions are generated.

2. A safety monitoring device based on a construction site according to claim 1, characterized in that: The method by which the terminal divides the construction site into risk areas of multiple safety levels is as follows: Retrieve construction site drawings; The danger level of each location in the construction site map is evaluated. Based on the danger level of each location in the construction site map, the entire construction site map is divided into risk areas with three safety levels, namely: low-risk area, medium-risk area and high-risk area, among which the danger levels of low-risk area, medium-risk area and high-risk area gradually increase.

3. A safety monitoring device based on a construction site according to claim 1, characterized in that: The method for generating the work safety risk coefficient of construction workers by the terminal based on the risk areas divided by the construction site and combining the collected work data of the construction workers is as follows: Collect historical work data of construction workers, including the average daily working hours of construction workers and the average daily time that construction workers are in each risk area; Based on the historical work data of construction workers, the work safety risk coefficient of construction workers is analyzed and evaluated as follows: ； in, is the work safety risk factor for construction workers. is the average daily working hours of construction workers. are the average daily time that construction workers spend in high-risk areas and medium-risk areas, respectively. is the weight coefficient, Both are greater than 0.

4. The safety monitoring device based on a construction site according to claim 1 is characterized in that: The specific steps of generating the early warning instruction by the terminal based on the work safety risk coefficient of the construction personnel and the current behavior of the construction personnel are as follows: Collect the behavior postures of construction workers and determine whether the current behavior posture is a dangerous behavior posture; When the current behavior posture is determined to be a dangerous behavior posture, a danger value of the dangerous behavior posture is calculated; Generate safety hazard values ​​based on the risk values ​​of dangerous behavior postures, the work safety risk coefficients of construction workers, and the areas where construction workers are located; Generate corresponding early warning instructions based on the safety hazard value.

5. A safety monitoring device based on a construction site according to claim 4, characterized in that: The method for judging whether the current behavior posture of the construction personnel is a dangerous behavior posture is as follows: Take photos of construction workers in their current postures; The photo is uploaded to the first machine learning model to identify and determine whether the current behavior posture of the construction worker is a dangerous behavior posture.

6. A safety monitoring device based on a construction site according to claim 5, characterized in that: The training method of the first machine learning model is as follows: X groups of historical photos are collected, and labels are set for the photos, where the labels are that the behavior postures of construction workers in the photos are dangerous behavior postures; the photos and the labels corresponding to the photos constitute a group of training data; the X groups of training data constitute a sample set, where X is an integer greater than 1; and the sample set is divided into a training set and a test set; the photos in the training set are used as input of a first machine learning model, and the labels in the training set are used as output of the first machine learning model; the first machine learning model is trained to obtain a dangerous behavior posture judgment model, and minimizing the sum of prediction errors is used as the training goal; the dangerous behavior posture judgment model is evaluated using the test set, and the dangerous behavior posture judgment model when the sum of prediction errors reaches convergence is used as the constructed first machine learning model.

7. The safety monitoring device based on a construction site according to claim 5 is characterized in that: When the current behavior posture is judged to be a dangerous behavior posture, the method for calculating the danger value of the dangerous behavior posture is as follows: The dangerous behavior posture photo containing the dangerous behavior posture is uploaded to the second machine learning model to calculate the danger value of the dangerous behavior posture, wherein the training method of the second machine learning model is as follows: X groups of historical dangerous behavior posture photos are collected, and labels are set for the dangerous behavior posture photos, wherein the labels are the danger values ​​of the dangerous behavior postures of the construction workers in the photos; the dangerous behavior posture photos and the labels corresponding to the dangerous behavior posture photos constitute a group of training data; the X groups of training data constitute a sample set, where X is an integer greater than 1, and the sample set is divided into a training set and a test set; the photos in the training set are used as input of a second machine learning model, and the labels in the training set are used as output of the second machine learning model, and the second machine learning model is trained to obtain a dangerous behavior posture danger value calculation model, wherein minimizing the sum of prediction errors is the training goal, and the dangerous behavior posture danger value calculation model is evaluated using the test set, and the dangerous behavior posture danger value calculation model when the sum of prediction errors reaches convergence is used as the constructed second machine learning model.

8. The safety monitoring device based on a construction site according to claim 1 is characterized in that: Based on the hazard value of dangerous behavior posture, the work safety risk factor of construction workers, and the area where construction workers are located, the method for generating safety hazard values ​​is as follows: ； in, is the safety risk value, is the correction factor, is the risk value corresponding to the dangerous behavior posture, It is the work safety risk factor for construction workers.

9. A safety monitoring device based on a construction site according to claim 8, characterized in that: Based on the potential safety hazard value, the method for generating early warning instructions is as follows: The preset safety hazard value threshold is ,in ; like , a first warning instruction is generated, and the user terminal issues an alarm according to the first warning instruction; like , a second warning instruction is generated, the user terminal issues an alarm according to the second warning instruction, and the terminal records and uploads the dangerous behavior posture of the construction personnel; like , a third early warning instruction is generated, the user end issues an alarm according to the third early warning instruction, the terminal records and uploads the dangerous behavior posture of the construction personnel, and notifies the superior management department of the construction personnel at the same time.

10. A monitoring system for a safety monitoring device based on a construction site as claimed in any one of claims 1 to 9, characterized in that: It includes: Data acquisition module: used to collect image data of the construction site and location data of construction workers; Data processing module: used to parse and process the image data of the construction site and the location data of the construction workers, obtain the behavior posture of the construction workers based on the analysis results, and generate early warning instructions based on the danger of the dangerous behavior posture when the behavior posture of the construction workers is a dangerous behavior posture; Early warning module: used to issue early warnings based on generated early warning instructions.

Images