Construction environment inspection method and system and storage medium

Abstract

The invention discloses a construction environment inspection method and system and a storage medium, and the method comprises the steps: obtaining a construction risk coefficient, personnel position information and camera information of at least one target person in a construction environment in each inspection period, and inputting the data into a pre-trained inspection planning model; the method comprises the steps that a target camera is started to output the inspection probability of each target person, the target person corresponding to the maximum value is selected from all the inspection probabilities to serve as the inspection target of the inspection period, and then the focusing direction of the target camera is controlled to be aligned with the inspection target in the inspection period till inspection is completed in all the inspection periods. By adopting the embodiment of the invention, inspection resources can be utilized most effectively, blind inspection is avoided, and the pertinence and efficiency of inspection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of big data processing, and in particular to a construction environment inspection method, system and storage medium. Background Art

[0002] The existing safety supervision usually adopts the manual + camera inspection mode. The common safety inspection process is: the construction team enters the corresponding information at the construction site according to the work ticket and installs a camera that can be connected to the safety supervision center. The on-duty personnel of the monitoring center select the construction site, rotate the camera remotely, and zoom in the monitoring screen to conduct construction safety compliance inspection. This inspection method will consume a lot of manpower costs. Since it does not fully consider the individual risk differences of construction workers, the inspection resources cannot be effectively utilized. Some high-risk personnel may not be inspected enough, and some low-risk personnel may be inspected blindly, resulting in some safety hazards not being discovered and handled in time. Summary of the invention

[0003] The purpose of the embodiments of the present invention is to provide a construction environment inspection method, system and storage medium, which can make the most effective use of inspection resources, avoid blind inspections, and improve the pertinence and efficiency of inspections.

[0004] To achieve the above object, an embodiment of the present invention provides a method for inspecting a construction environment, comprising: In each inspection cycle, the construction risk factor and personnel location information of at least one target person in the construction environment are obtained, and the camera information of the target camera in the construction environment is obtained; Inputting the construction risk coefficient, the personnel location information and the camera information into the inspection planning model, so that the inspection planning model outputs the inspection probability of each target personnel; Select the target person corresponding to the maximum value from all inspection probabilities as the inspection target of this inspection cycle; In this inspection cycle, the focus direction of the target camera is controlled to be aligned with the inspection target until all inspection cycles are completed.

[0005] As an improvement of the above scheme, the construction risk coefficient is calculated based on the risk level value corresponding to the personnel information of the target personnel; wherein the personnel information includes at least one of basic information, team information and construction information, and each personnel information corresponds to a risk level value.

[0006] As an improvement of the above solution, the basic information includes attribute information and first historical violation information; when the personnel information includes basic information, the method for determining the first risk level value corresponding to the basic information includes: Divide historical data into positive and negative samples; where positive samples include data records where no safety accidents have occurred, and negative samples include data records where safety accidents have occurred; For the attribute information in the negative samples, clustering analysis is performed using clustering algorithms to obtain several attribute categories; Calculate the contribution rate of safety accidents in each attribute category, and integrate the contribution rates of all attribute categories to obtain a contribution rate set; Acquire a target contribution rate that matches the target person from the contribution rate set, and acquire a violation risk coefficient corresponding to the first historical violation information; The sum of the target contribution rate and the violation risk coefficient is taken as the first risk level value.

[0007] As an improvement of the above solution, the calculation of the contribution rate of each attribute category to the occurrence of safety accidents includes: For each attribute category, counting the number of positive samples belonging to the current attribute category among the positive samples, and counting the number of negative samples belonging to the current attribute category among the negative samples; The contribution rate of the safety incident occurring in the current attribute category is determined according to the number of positive samples and the number of negative samples.

[0008] As an improvement of the above solution, the first historical violation information includes a first time interval between the target person's violation time and the current time, and the violation risk coefficient is inversely proportional to the first time interval.

[0009] As an improvement of the above solution, the team information includes second historical violation information and historical working hours, and the second historical violation information is a second time interval between the violation time of any person in the team and the current time; when the personnel information includes team information, the method for determining the second risk level value corresponding to the team information includes: Calculating a first risk coefficient according to the second time interval, and calculating a second risk coefficient according to the historical working hours; wherein the second time interval and the first risk coefficient are inversely proportional, and the historical working hours and the second risk coefficient are inversely proportional; The sum of the first risk coefficient and the second risk coefficient is taken as the second risk degree value.

[0010] As an improvement of the above solution, the construction information includes the construction type and its corresponding construction risk level; when the personnel information includes construction information, the third risk level value corresponding to the construction information is directly proportional to the construction risk level.

[0011] As an improvement of the above scheme, the inspection planning model is a reinforcement learning model, a reward function is provided in the inspection planning model, and the inspection planning model takes the maximum value of the reward function output as the optimization target; wherein the reward function is related to the construction risk coefficient.

[0012] To achieve the above purpose, an embodiment of the present invention further provides a construction environment inspection system, comprising: An information acquisition module, used to acquire the construction risk factor and personnel location information of at least one target person in the construction environment in each inspection cycle, and to acquire the camera information of the target camera in the construction environment; An inspection probability generation module, used for inputting the construction risk coefficient, the personnel location information and the camera information into the inspection planning model, so that the inspection planning model outputs the inspection probability of each target personnel; Inspection target determination module, used to select the target personnel corresponding to the maximum value from all inspection probabilities as the inspection target of this inspection cycle; The inspection control module is used to control the focus direction of the target camera to align with the inspection target in this inspection cycle until all inspection cycles are completed.

[0013] To achieve the above-mentioned purpose, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the inspection method of the construction environment as described in any of the above-mentioned embodiments.

[0014] Compared with the prior art, the inspection method, system and storage medium of the construction environment disclosed in the present invention can accurately select the highest-risk personnel from a large number of uninspected personnel as inspection targets by obtaining the construction risk coefficient of the target personnel, combining the personnel location information and camera information, and using the inspection planning model to output the inspection probability, so as to make the most effective use of inspection resources, avoid blind inspections, and improve the pertinence and efficiency of inspections. When determining the inspection targets, the personnel location information and camera information are also taken into account, which helps to plan a more reasonable inspection route, reduce time waste and distance consumption during the inspection process, improve the overall inspection efficiency, and shorten the time required to complete all inspection cycles. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a system topology diagram of the inspection method provided by the embodiment of the present invention; Figure 2 It is a flow chart of a method for inspecting a construction environment provided by an embodiment of the present invention; Figure 3is another flow chart of a method for inspecting a construction environment provided by an embodiment of the present invention; Figure 4 is a schematic diagram of positioning a target person by a positioning device provided in an embodiment of the present invention; Figure 5 is a workflow diagram of the inspection planning model provided by an embodiment of the present invention; Figure 6 It is a structural block diagram of a construction environment inspection system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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 creative work are within the scope of protection of the present invention.

[0017] See also Figure 1 , Figure 1 This is a system topology diagram of the inspection method provided by the embodiment of the present invention. It can be divided into three layers from bottom to top, corresponding to the intelligent inspection device at the power construction site (field layer), the Internet (transmission layer) and the security management platform layer. The field layer mainly includes edge computing devices, cameras, sound and light alarms and positioning devices.

[0018] Edge computing devices are used to provide computing and storage capabilities at infrastructure construction sites, deploy trained reinforcement learning inference models, and process input data including but not limited to video data uploaded by cameras, team member data, sensor data, etc., and determine inspection targets and execute inspection strategies based on these data, as well as upload inspection images that meet the violation identification accuracy to the security management platform, and receive alarm instructions issued by the management platform.

[0019] The camera is the control object of the reinforcement learning model. First, it can upload its own device status (including the current focus angle and focal length of the camera) and the captured video frame data. Second, it can receive and execute control commands (including rotating the camera and zooming).

[0020] The sound and light alarm is connected to the edge computing device via wired or wireless means. When the violation identification model of the safety management platform finds any illegal construction behavior at the construction site, it will send an alarm command to the edge computing device, thereby triggering the sound and light alarm function of the sound and light alarm.

[0021] The positioning device can accurately locate personnel in the construction environment. For example, the positioning device is a UWB (Ultra WideBand) positioning system, which includes two parts: a UWB base station and a UWB tag. The UWB tag needs to be worn on the hands of the target personnel. At the same time, four UWB base stations need to be deployed at the construction site to achieve ranging positioning based on TDOA (Time Difference of Arrival).

[0022] It should be noted that during the system initialization phase (such as when construction starts every day), the above equipment needs to be checked for health and loaded with configuration. Among them, the health check mainly includes the network link status of the edge computing device and the security management platform, the computing / storage status of the edge computing device, the camera status, the positioning device, and the network link status between the camera, positioning device and edge computing device, whether the sound and light alarm can work normally, etc. Configuration loading mainly loads some algorithm models, such as inspection planning models, etc.

[0023] See also Figure 2 , Figure 2 1 is a flowchart of a method for inspecting a construction environment provided by an embodiment of the present invention. The method for inspecting a construction environment is implemented by an edge computing device. The method for inspecting a construction environment includes: S1. In each inspection cycle, the construction risk factor and personnel location information of at least one target person in the construction environment are obtained, and the camera information of the target camera in the construction environment is obtained; S2, inputting the construction risk coefficient, the personnel location information and the camera information into the inspection planning model, so that the inspection planning model outputs the inspection probability of each target personnel; S3. Select the target person corresponding to the maximum value from all inspection probabilities as the inspection target of this inspection cycle; S4. In this inspection cycle, the focus direction of the target camera is controlled to be aligned with the inspection target until all inspection cycles are completed.

[0024] Exemplarily, the inspection planning model is deployed on the edge computing device. The inspection planning model can determine the target personnel who need to be monitored during this inspection cycle, thereby controlling the target camera deployed on site to periodically inspect the construction status of the target personnel. Not only can the lens rotation be completed automatically, but also the target personnel can occupy more than 50% of the radiation in the shooting picture through automatic docking, and the data is transmitted back to the security management platform deployed on the cloud to complete the detection of violations. In addition, due to the different types of work and the uneven comprehensive quality of team members at the construction site, the safety risk of each target personnel is different. The present invention will calculate the risk coefficient of each target personnel based on the personnel information of the target personnel, and based on the position of each target personnel, use the inspection planning model to calculate the overall risk cost of personnel inspection within a cycle. The ultimate goal is to obtain the maximum global benefit (that is, the overall risk cost is small), and then screen out inspection targets with higher risk levels for targeted inspections.

[0025] See also Figure 3 , Figure 3 is another flow chart of a method for inspecting a construction environment provided by an embodiment of the present invention, combined with Figure 3 The above steps S1 to S4 are described in detail.

[0026] Specifically, in step S1, the construction risk coefficient is calculated based on the risk level value corresponding to the personnel information of the target personnel; wherein the personnel information includes at least one of basic information, team information and construction information, and each personnel information corresponds to a risk level value. The details of each personnel information are as follows: 1) Basic information, including attribute information and the first historical violation information. The basic information part represents the basic physical quality of the construction worker, and reflects the strength of the construction worker's safety awareness to a certain extent through historical violations. The attribute information includes gender, age, length of service, etc. Each piece of data in the first historical violation information includes three dimensions, namely, the violation time (the first time interval between the target person's violation time and the current time), severity, and violation content.

[0027] 2) Team information, which includes the second historical violation information and historical working hours. Team information reflects the management level of the team to which the target person belongs. The second historical violation information also includes the violation time (the second time interval between the violation time of any person in the team and the current time), severity, and violation content. Historical working hours represent the experience of the team.

[0028] 3) Construction information, including the construction type and its corresponding construction risk level, which records in detail the environmental conditions faced by the target personnel in this operation. Construction types include welding, cutting, wiring, etc., each of which has a corresponding construction risk level. In addition, the construction risk level can also further consider the influence of construction time, construction location, meteorological indicators, etc.

[0029] For the above three types of personnel information, the process of determining the corresponding risk level values ​​is explained respectively: Case 1: When the personnel information includes basic information, the method for determining the first risk level value corresponding to the basic information includes: dividing the historical data into positive and negative samples; wherein the positive samples include data records where no safety accidents have occurred, and the negative samples include data records where safety accidents have occurred; for the attribute information in the negative samples, clustering analysis is performed using a clustering algorithm to obtain several attribute categories; the contribution rate of each attribute category to the occurrence of safety accidents is calculated, and the contribution rates of all attribute categories are integrated to obtain a contribution rate set; the target contribution rate matching the target personnel is obtained from the contribution rate set, and the violation risk coefficient corresponding to the first historical violation information is obtained; the sum of the target contribution rate and the violation risk coefficient is taken as the first risk level value. Wherein, the calculation of the contribution rate of each attribute category to the occurrence of safety accidents includes: for each attribute category, counting the number of positive samples belonging to the current attribute category in the positive samples, and counting the number of negative samples belonging to the current attribute category in the negative samples, and determining the contribution rate of the current attribute category to the occurrence of safety accidents according to the number of positive samples and the number of negative samples.

[0030] For example, all historical data are divided into positive and negative samples according to whether a safety accident occurs, that is, ,in represents positive samples, that is, data records where no safety accidents have occurred. represents negative samples, that is, data records where safety accidents have occurred. The attribute information (gender, age and length of service) in the basic information in the set is clustered and analyzed using a clustering algorithm. For example, the clustering algorithm used in the embodiment of the present invention is the K-Means algorithm, and the k value is determined by the silhouette coefficient method. The k value represents the number of clusters into which the data is to be divided. The attribute information of is represented by a data to be clustered, such as gender, age, and length of service, which can be expressed as , assuming that the total amount of data of the target person is n, satisfying , then the value range of k is set to .

[0031] Iterate clustering for different k values ​​and use the following formula to calculate the target personnel The silhouette coefficient of the data to be clustered: (1); in, The target person The average distance between the data to be clustered (sample points) and all other points in its cluster, The target person The average distance between the data to be clustered (sample points) and all the points in the nearest cluster. For a k value, the silhouette coefficient satisfy: (2); but , that is, take the k value that minimizes the silhouette coefficient. So far, the data to be clustered is divided into k categories through the k-means algorithm, which is recorded as .

[0032] Then, the contribution rate of each cluster to whether an accident can be judged is traversed, recorded as , the calculation process of the target personnel's contribution rate is as follows: 1. Statistics The attribute category in the collection The number of records is recorded as ; 2. Statistics The attribute category in the collection The number of records is recorded as ; 3. Calculate contribution rate ,satisfy: (3); 4. Traversal , repeat the above process to get .

[0033] Finally, the contribution rates of all attribute categories are integrated to obtain the contribution rate set , obtain the target contribution rate that matches the target person from the contribution rate set. In the screening process, for each target person, traverse , find the most similar attribute category, such as obtaining a similarity index through a similarity calculation method, select the target attribute category corresponding to the maximum value, and then index the contribution rate corresponding to the target attribute category as the target contribution rate.

[0034] Exemplarily, after obtaining the target contribution rate of the target person and the violation risk coefficient corresponding to the first historical violation information, the sum of the target contribution rate and the violation risk coefficient is taken as the first risk level value, satisfying the following formula: (4); in, For target personnel The target contribution rate of is the risk coefficient that represents the cluster category to which it belongs; For target personnel The risk factor of violation.

[0035] Specifically, the first historical violation information includes a first time interval between the target person's violation time and the current time, and the violation risk coefficient is in inverse proportion to the first time interval.

[0036] For example, the attenuation concept is used to calculate the proportion of historical violations. Assuming that the target person The first time interval of the historical violation time from the current month is expressed as , Indicates the months after the passage of time, the target person The violation risk factor meets the following requirements: (5).

[0037] Case 2: When the personnel information includes team information, the method for determining the second risk level value corresponding to the team information includes: calculating the first risk coefficient based on the second time interval, and calculating the second risk coefficient based on the historical working hours; wherein, the second time interval and the first risk coefficient are inversely proportional, and the historical working hours and the second risk coefficient are inversely proportional; and the sum of the first risk coefficient and the second risk coefficient is the second risk level value.

[0038] Exemplary, target personnel The second historical violation information is calculated in the same way as formula (5) in case 1, and the sum of the reciprocals of the monthly span is taken as the first risk coefficient of this part, which is recorded as The team's historical work is calculated based on the inverse of the cumulative months as the second risk factor for this part, recorded as . Then the target person The second risk level value corresponding to the team information satisfies: (6); Furthermore, in order to maintain consistency in order of magnitude with the basic information, the second risk level values ​​of all target personnel in this construction may be normalized. The calculation process for normalizing the data may refer to the prior art, and the present invention does not make any specific limitation on this.

[0039] Case 3: When the personnel information includes construction information, the third risk level value corresponding to the construction information is directly proportional to the construction risk level.

[0040] For example, the sample division method is the same as that in case 1, and the negative samples are counted The number of all construction types in the set. For all construction types that have had safety accidents, sort them from large to small according to their number to obtain the construction risk values ​​corresponding to different construction types. These construction risk values ​​can be further normalized. The construction risk levels are divided into four categories: major risk, relatively large risk, general risk, and low risk. The construction risk levels correspond to different construction types. As the weight of the construction type, it can be determined based on experience. For example, the construction risk level is recorded as After obtaining the construction type of the target personnel, the construction risk value and the construction risk level are multiplied as the third risk level value corresponding to the construction information. The third risk level value satisfies: (7).

[0041] Furthermore, in order to maintain consistency in order of magnitude with the basic information part, a normalized calculation is performed on the operational risk part of all personnel in this construction. The calculation process for normalizing the data can refer to the existing technology, and the present invention does not make specific limitations on this.

[0042] Furthermore, assuming that the target person The personnel information includes the above three types of information (basic information, team information and construction information), then the target personnel Construction risk factor The risk level values ​​corresponding to the three types of information are added together, if the following formula is satisfied: (8).

[0043] Specifically, each target person is provided with a positioning device, and the position information of each target person is determined by receiving positioning information from the positioning device.

[0044] For example, see Figure 4 , Figure 4It is a schematic diagram of the positioning device provided in an embodiment of the present invention for positioning the target person. The acquisition of the position of the target person is achieved through a UWB positioning system, which includes two parts: a UWB base station and a UWB tag. Four UWB base stations are deployed at the construction site, and positioning is performed based on TDOA using the arrival time difference. The UWB tag worn by the target person sends a UWB signal to the outside once, and all UWB base stations within the wireless coverage range of the tag will receive the wireless signal. If two UWB base stations with known coordinates receive the signal, and the intervals between the UWB tag and the two UWB base stations are different, then the time points when the two UWB base stations receive the signal are different, thus deriving the concept of "arrival time difference".

[0045] According to the above principle, assuming that the target person The three-dimensional coordinates of , the locations of the four UWB base stations are , , , . It is known that the signal propagation speed is a fixed value denoted by , the time it takes for the tag signal to arrive at the four base stations is , , , , the distance from the 4 base stations is , , , .use Indicates the target person The distance difference between base stations m and n can be obtained by the following formula: (9); in, , , , and the speed of propagation is a known scalar, then the result of formula (9) can be directly calculated.

[0046] Furthermore, since the coordinates of the four base stations are known, the following equations can be obtained, and the target personnel can be obtained by solving them: The three-dimensional coordinates of .

[0047] (10).

[0048] At this point, the three-dimensional coordinates of the target person based on the UWB positioning system are solved.

[0049] Specifically, in step S2, the inspection planning model is a reinforcement learning model, a reward function is provided in the inspection planning model, and the inspection planning model takes the maximum value of the reward function output as an optimization target; wherein the reward function is related to the construction risk coefficient.

[0050] For example, see Figure 5 , Figure 5 is a workflow diagram of an inspection planning model provided by an embodiment of the present invention, wherein the inspection planning model is a DDQN (Dual Deep Q Network) model. Represents the state of the environment at time t, which is the input of the reinforcement learning model. Includes the three-dimensional coordinates of each target person (such as A and B in the figure) and construction risk factor , and further includes a mark indicating whether the target person has completed the inspection. The existence of this mark can avoid repeatedly selecting the same target person as the inspection target. If the target person A has completed the inspection, it is marked as 1, otherwise it is 0. This inspection mark can be recorded as .besides, It also includes the camera information of the target camera, such as the 3D coordinates of the target camera and the direction vector of the current focus direction Therefore, the state description of the infrastructure construction site can be recorded as: .

[0051] It should be noted that, at the initial stage of the task, the spatial position coordinates of the target camera are camera_position=(0,0,0), and the unit space vector corresponding to the initial field of view center of the target camera is camera_focus=(0,0,1). In the embodiment of the present invention, a target camera can be arranged to monitor the inspection target. Of course, a construction environment can also have multiple cameras, and the remaining cameras except the target camera work normally according to the set program.

[0052] During the task execution, the physical position coordinates of the target camera remain unchanged. The actions that the target camera can perform include camera rotation and autofocus. The rotation makes the center of the lens field of view align with the target to be captured, and adjusts the lens focal length to a suitable position so that the target to be captured occupies more than 50% of the frame in the captured image. The action space of the target camera is the target personnel who have not been inspected during this inspection cycle.

[0053] It should be noted that when the inspection task is started, the target camera is located at the initialization position. After inspecting the target, the direction vector of the target camera's field of view is updated to , now proceed to the next goal The inspection, the action required at this time consists of two parts: one is the target camera turning: record the time required , this value is and The ratio of the angle between the two and the angular velocity of the target camera. The second is to change the focal length: record the time required ,The focusing time is proportional to the distance between the target camera and the inspection target, and the value is the ratio of the distance between the inspection target and the coordinate origin to the focusing speed of the target camera.

[0054] Furthermore, in the problem scenario to be solved by the present invention, since the start of the inspection task, all inspection targets may be at risk due to the previous waiting time, and the probability of risk is proportional to its risk coefficient and waiting time. Representatives The inspection time required for each inspection target, Representative The construction risk factor of each inspection target, Representative The risk cost of waiting for an inspection target to complete the inspection is: (11); Therefore, the overall risk cost from the start of the inspection task to the end of the task satisfies: (12); in, is the total risk cost, is the total number of inspection cycles, They are the inspection duration corresponding to each inspection cycle.

[0055] The above formula (12) can be expressed more generally as: (13).

[0056] For DDQN, The reward obtained at the moment is the discounted accumulation of the rewards obtained from the current moment to the end of the entire scene. The reward function in the embodiment of the present invention should minimize the total risk cost. as the goal.

[0057] For example, the input state In the inspection planning model, the inspection planning model will consider the target personnel that can be photographed within the coverage range of the target camera (determined by the personnel position information. If there is a target person who is far away from the target camera, this target person will not be selected as the inspection target. This is because the distance is too far and the target camera cannot track it accurately. Even if it is focused, the image is not clear), as well as the construction risk factor of the target person. Factors such as the inspection probability between 0 and 1 are output for each target person. = .

[0058] In the embodiment of the present invention, when it comes to target personnel with different construction risk factors, the overall risk cost needs to be considered. For example, the risk factor of work points such as high-altitude work and live work is higher, while the risk factor of ground non-live work is relatively low. At this time, the optimal path is sought in combination with the position of the target camera and the construction risk factor. Especially when the position of the on-site personnel moves, because the position and field of view of the target camera are recorded in the state matrix, according to the properties of the Markov chain, the deployed inspection planning model can directly calculate the optimal strategy in combination with the current on-site status without the need to recalculate all paths, meeting the real-time requirements of the control system, especially for complex environmental conditions.

[0059] Specifically, in steps S3 to S4, after obtaining the inspection probabilities of all target persons, the target person corresponding to the maximum value is taken as the inspection target, the focus direction of the target camera is controlled to aim at the inspection target, and the shooting of the inspection image is completed until all inspection cycles are completed.

[0060] Exemplarily, the safety management platform is deployed in a cloud platform, on which various types of violation detection models can be deployed, such as identification of not wearing a helmet, identification of crossing a fence, etc. The safety management platform receives the inspection image uploaded in step S4 to perform a violation identification operation, and if a violation operation of the inspection target occurs, a prompt message is issued, such as issuing a command to trigger an audible and visual alarm.

[0061] Compared with the prior art, the inspection method of the construction environment disclosed in the present invention has the following beneficial effects: 1. Aiming at the problem that the camera focus tracking operation in the prior art depends on the safety supervisor, which leads to high skill requirements for the supervisor and limited real-time performance, the present invention monitors the construction personnel through high-precision image extraction technology, thereby providing a more reliable guarantee for construction safety. Through automated and intelligent image processing technology, the present invention not only reduces the workload of safety supervisors, but also enhances the real-time performance and accuracy of construction site safety management.

[0062] 2. By obtaining the construction risk coefficient of the target personnel, combining the personnel location information and camera information, and using the inspection planning model to output the inspection probability, it is possible to accurately select the highest-risk personnel from the many uninspected personnel as the inspection target, so that inspection resources can be used most effectively, avoiding blind inspections and improving the pertinence and efficiency of inspections.

[0063] 3. Taking into account the location of personnel and camera information, it helps to plan a more reasonable inspection route, reduce time waste and distance consumption during the inspection process, improve overall inspection efficiency, and shorten the time required to complete all inspection cycles.

[0064] 4. Carry out key inspections on personnel with high construction risk factors, so as to promptly discover their safety hazards and violations during the construction process, take timely measures to correct and deal with them, reduce the probability of construction accidents, and ensure the safety of construction personnel and the smooth progress of construction projects.

[0065] 5. Based on multi-source data such as construction risk factors, personnel location information, and camera information, intelligent inspection management is achieved through analysis and decision-making through the inspection planning model. This data-driven approach can formulate inspection strategies more objectively and accurately, avoiding interference and subjectivity from human factors.

[0066] See also Figure 6 , Figure 6 It is a structural frame of a construction environment inspection system 100 provided in an embodiment of the present invention. The construction environment inspection system 100 includes: The information acquisition module 11 is used to obtain the construction risk factor and personnel location information of at least one target person in the construction environment in each inspection cycle, and obtain the camera information of the target camera in the construction environment; An inspection probability generation module 12 is used to input the construction risk coefficient, the personnel location information and the camera information into the inspection planning model, so that the inspection planning model outputs the inspection probability of each target personnel; Inspection target determination module 13, used to select the target person corresponding to the maximum value from all inspection probabilities as the inspection target of this inspection cycle; The inspection control module 14 is used to control the focus direction of the target camera to aim at the inspection target in this inspection cycle until all inspection cycles are completed.

[0067] Specifically, the construction risk coefficient is calculated based on the risk level value corresponding to the personnel information of the target personnel; wherein the personnel information includes at least one of basic information, team information and construction information, and each piece of personnel information corresponds to a risk level value.

[0068] Specifically, the basic information includes attribute information and first historical violation information; when the personnel information includes basic information, a method for determining a first risk level value corresponding to the basic information includes: Divide historical data into positive and negative samples; where positive samples include data records where no safety accidents have occurred, and negative samples include data records where safety accidents have occurred; For the attribute information in the negative samples, clustering analysis is performed using clustering algorithms to obtain several attribute categories; Calculate the contribution rate of safety accidents in each attribute category, and integrate the contribution rates of all attribute categories to obtain a contribution rate set; Acquire a target contribution rate that matches the target person from the contribution rate set, and acquire a violation risk coefficient corresponding to the first historical violation information; The sum of the target contribution rate and the violation risk coefficient is taken as the first risk level value.

[0069] Specifically, the calculation of the contribution rate of each attribute category to the occurrence of safety accidents includes: For each attribute category, counting the number of positive samples belonging to the current attribute category among the positive samples, and counting the number of negative samples belonging to the current attribute category among the negative samples; The contribution rate of the safety incident occurring in the current attribute category is determined according to the number of positive samples and the number of negative samples.

[0070] Specifically, the first historical violation information includes a first time interval between the target person's violation time and the current time, and the violation risk coefficient is in inverse proportion to the first time interval.

[0071] Specifically, the team information includes second historical violation information and historical working hours, and the second historical violation information is a second time interval between the violation time of any person in the team and the current time; when the personnel information includes team information, the method for determining the second risk level value corresponding to the team information includes: Calculating a first risk coefficient according to the second time interval, and calculating a second risk coefficient according to the historical working hours; wherein the second time interval and the first risk coefficient are inversely proportional, and the historical working hours and the second risk coefficient are inversely proportional; The sum of the first risk coefficient and the second risk coefficient is taken as the second risk degree value.

[0072] Specifically, the construction information includes the construction type and its corresponding construction risk level; when the personnel information includes construction information, the third risk level value corresponding to the construction information is in direct proportion to the construction risk level.

[0073] Specifically, the inspection planning model is a reinforcement learning model, a reward function is provided in the inspection planning model, and the inspection planning model takes the maximum value of the reward function output as an optimization target; wherein the reward function is related to the construction risk coefficient.

[0074] It is worth noting that the working process of each module in the inspection system 100 for the construction environment described in the embodiment of the present invention can refer to the working process of the inspection method for the construction environment described above, and will not be repeated here.

[0075] Furthermore, an embodiment of the present invention also provides a computer-readable storage medium, which stores a computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the inspection method of the construction environment as described in any of the above embodiments.

[0076] The present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0077] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for inspecting a construction environment, characterized in that: include: In each inspection cycle, the construction risk factor and personnel location information of at least one target person in the construction environment are obtained, and the camera information of the target camera in the construction environment is obtained; Inputting the construction risk coefficient, the personnel location information and the camera information into the inspection planning model, so that the inspection planning model outputs the inspection probability of each target personnel; Select the target person corresponding to the maximum value from all inspection probabilities as the inspection target of this inspection cycle; In this inspection cycle, the focus direction of the target camera is controlled to be aligned with the inspection target until all inspection cycles are completed.

2. The inspection method for a construction environment as claimed in claim 1, characterized in that: The construction risk coefficient is calculated based on the risk level value corresponding to the personnel information of the target personnel; wherein the personnel information includes at least one of basic information, team information and construction information, and each piece of personnel information corresponds to a risk level value.

3. The inspection method for a construction environment as claimed in claim 2, characterized in that: The basic information includes attribute information and first historical violation information; when the personnel information includes basic information, a method for determining a first risk level value corresponding to the basic information includes: Divide historical data into positive and negative samples; where positive samples include data records where no safety accidents have occurred, and negative samples include data records where safety accidents have occurred; For the attribute information in the negative samples, clustering analysis is performed using clustering algorithms to obtain several attribute categories; Calculate the contribution rate of safety accidents in each attribute category, and integrate the contribution rates of all attribute categories to obtain a contribution rate set; Acquire a target contribution rate that matches the target person from the contribution rate set, and acquire a violation risk coefficient corresponding to the first historical violation information; The sum of the target contribution rate and the violation risk coefficient is taken as the first risk level value.

4. The inspection method for a construction environment as claimed in claim 3, characterized in that: The calculation of the contribution rate of each attribute category to the occurrence of safety accidents includes: For each attribute category, count the number of positive samples belonging to the current attribute category in the positive samples, and count the number of negative samples belonging to the current attribute category in the negative samples; The contribution rate of the safety incident occurring in the current attribute category is determined according to the number of positive samples and the number of negative samples.

5. The inspection method for a construction environment as claimed in claim 3, characterized in that: The first historical violation information includes a first time interval between the target person's violation time and the current time, and the violation risk coefficient is inversely proportional to the first time interval.

6. The inspection method for a construction environment as claimed in claim 2, characterized in that: The team information includes second historical violation information and historical working hours, wherein the second historical violation information is a second time interval between the violation time of any person in the team and the current time; When the personnel information includes team information, the method for determining the second risk level value corresponding to the team information includes: Calculating a first risk coefficient according to the second time interval, and calculating a second risk coefficient according to the historical working hours; wherein the second time interval and the first risk coefficient are inversely proportional, and the historical working hours and the second risk coefficient are inversely proportional; The sum of the first risk coefficient and the second risk coefficient is taken as the second risk degree value.

7. The inspection method for a construction environment as claimed in claim 2, characterized in that: The construction information includes the construction type and its corresponding construction risk level; when the personnel information includes construction information, the third risk level value corresponding to the construction information is in direct proportion to the construction risk level.

8. The inspection method for a construction environment as claimed in claim 1, characterized in that: The inspection planning model is a reinforcement learning model, in which a reward function is provided. The inspection planning model takes the maximum value of the reward function output as an optimization target; wherein the reward function is related to the construction risk coefficient.

9. A construction environment inspection system, characterized in that: include: An information acquisition module, used to acquire the construction risk factor and personnel location information of at least one target person in the construction environment in each inspection cycle, and to acquire the camera information of the target camera in the construction environment; An inspection probability generation module, used for inputting the construction risk coefficient, the personnel location information and the camera information into the inspection planning model, so that the inspection planning model outputs the inspection probability of each target personnel; Inspection target determination module, used to select the target person corresponding to the maximum value from all inspection probabilities as the inspection target of this inspection cycle; The inspection control module is used to control the focus direction of the target camera to align with the inspection target in this inspection cycle until all inspection cycles are completed.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the inspection method for the construction environment as described in any one of claims 1 to 8.

Images