A whole-process monitoring system for offshore project demolition operation

Abstract

The application provides a kind of offshore project demolition operation whole process monitoring system, the system includes data acquisition module, work ticket generation module, demolition operation path planning module, demolition data collection and case module, illegal behavior identification module, risk analysis module, visualization module;Realize in considering complex offshore environment, monitor the whole process of demolition operation, generate work operation ticket using work ticket generation module;Using demolition operation path planning module generates operation path planning graph;Using illegal behavior identification module to monitor the illegal behavior in the process of demolition operation for accurate identification, and using risk analysis module to identify and feedback the risk type that illegal behavior may correspond to, while analyzing and identifying the potential risk in its demolition operation process content;Through effective monitoring, the safety and efficiency of operation can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of offshore operation monitoring, and particularly relates to a whole-process monitoring system for offshore project demolition operation. BACKGROUND

[0002] At present, an offshore operation platform is a land-like area built in seawater for facilitating offshore operation, and is used for production operation or other activities, such as offshore oil extraction and offshore construction operation. With the combination of three-dimensional visualization and the offshore operation platform, it is more convenient for construction personnel and supervisors to control the production operation process.

[0003] However, there are problems of information island and slow reaction in traditional project management, which will reduce the safety and efficiency of offshore demolition operation. In the process of realizing whole-process intelligent monitoring and management of offshore demolition operation, the prior art still has deficiencies in data acquisition, information processing, rule violation identification and risk analysis. Therefore, the present application provides a whole-process monitoring system for offshore project demolition operation, which improves the safety, efficiency and engineering management level of demolition operation. SUMMARY

[0004] The present application aims to provide a whole-process monitoring system for offshore project demolition operation to solve the problems in the background.

[0005] The present application is achieved by the following technical scheme: a whole-process monitoring system for offshore project demolition operation, comprising a data acquisition module, a work ticket generation module, a demolition operation path planning module, a demolition material collection module, a rule violation identification module, a risk analysis module and a visualization module.

[0006] The data acquisition module uses a drone to perform three-dimensional scanning on the demolition area to obtain high-precision three-dimensional data, uses a combination of a laser radar and a high-definition camera to obtain video image data and create a three-dimensional model of the area, and collects demolition team personnel information, main mechanical and ship information.

[0007] The work ticket generation module obtains demolition area information, demolition team personnel information, demolition project information, demolition object information, demolition sequence information and demolition estimated completion time, and generates a work operation ticket.

[0008] The demolition operation path planning module constructs a high-precision three-dimensional visualization model based on the obtained three-dimensional data, allows users to intuitively view the structure and potential risks of the demolition area in a virtual environment, and generates an operation path planning graph based on a path planning algorithm of reinforcement learning in combination with the spatial position of the object and the demolition sequence.

[0009] The demolition information collection and case filing module is used to collect all information related to the demolition work, including the content of the work operation ticket, video images, and form a complete job file;

[0010] The violation behavior identification module analyzes the real-time video stream of the obtained video image data, and automatically identifies the violation behavior of the work personnel by combining a multi-channel time convolution network with a self-attention mechanism.

[0011] The risk analysis module is used to feedback the risk type that the violation behavior may correspond to, and simultaneously analyze and identify the potential risks in the demolition work process.

[0012] The visualization module is used to provide a visual interface, support real-time updating of mobile and computer terminals, and monitor and track the demolition progress, whether the object is demolished, and whether the demolition sequence is correct in real time and display on the interface.

[0013] Specifically, high-precision three-dimensional data is obtained for data cleaning to remove noise and duplicate points, and the processed three-dimensional data is used to generate a three-dimensional mesh model, which is smoothed using a triangulation algorithm to form a continuous and smooth surface. A texture mapping algorithm is used to add material properties to the model, and the structure features and potential hazard locations on the model are labeled.

[0014] Specifically, in the process of generating a three-dimensional mesh model, a triangulation algorithm is used for smoothing to form a continuous and smooth surface, as follows:

[0015] For each vertex, identify the neighborhood triangles around it and calculate the average position of the vertex. The neighborhood triangles are set to a fixed radius or a number of neighborhood points to ensure that the processed results are smooth without losing structure.

[0016] The coordinates of the current point are updated by averaging the coordinates of the neighborhood points, and the normal vector consistency is maintained during the smoothing process to prevent surface inversion.

[0017] Specifically, a texture mapping algorithm is used to add material properties to the three-dimensional network model, and the structure features and potential hazard locations on the model are labeled. The texture selection is as follows:

[0018] The calibrated and processed three-dimensional data is inserted into the three-dimensional network model, and normal mapping and displacement mapping are used to enhance the three-dimensional effect and material texture. Normal mapping can generate a detailed surface under lighting conditions.

[0019] Key structural features and potential hazard areas are identified and labeled in the three-dimensional network model, and the key structures in the model are extracted to ensure that they are linked to the key points of the actual demolition work.

[0020] The potential hazard area is color divided: red represents heavy object hanging area, to warn potential safety risk; yellow represents damaged area, to prompt the place that needs special attention; blue represents the height of the hazard area, including the gap between the upper and lower layers, so that workers can understand the space of work.

[0021] Specifically, the process of generating the work ticket by the work ticket generation module specifically includes:

[0022] In the operation ticket, the name, skill, health status, and responsibility division information of each member in the operation are listed in detail;

[0023] The object to be removed is automatically identified, including the type of object, material characteristics, potential hazards, and the history information of the removal project is displayed, and the precise marking of the position of the object is displayed;

[0024] In the generated operation ticket, the real-time updated three-dimensional model is embedded, the position of the removed object is clearly marked, and color coding is used to represent different danger levels;

[0025] Combined with the data analysis and on-site investigation in the early stage, the system automatically generates recommended removal steps and provides operation specifications, including necessary safety measures and emergency plans;

[0026] According to historical cases and real-time data, the duration of each removal step is predicted, and the estimated completion time is generated in the operation ticket.

[0027] Specifically, the process of generating the work ticket by the work ticket generation module specifically includes:

[0028] The update formula of reinforcement learning is:

[0029] Q(s t , a t )←Q(s t , a t )+α((R(s t , a t )+γmax Q(s t+1 , a)-Q(s t , a t ))

[0030] Where Q is the optimal action to obtain the maximum total reward, Q(s t , a t ) represents the expected long-term cumulative reward of performing action a t in state s t , and max Q(s t+1 , a) represents the state s t+1The maximum value of the Q value of all possible actions a in the next state s t is the state at time t, a t is the action at time t, alpha is the learning rate, R is the reward function, gamma is the discount factor, s t+1 is the state at time t+1.

[0031] to maximize the expected reward of the final path P, then:

[0032] P=arg max Q(s t , a t )

[0033] The trust degree of path selection is introduced, and the reputation value C(P) is defined to evaluate the reliability of different paths P, and the formula is as follows:

[0034] C(P)=ω c ·C t (P)+ω r ·R(P)

[0035] Wherein, C t (P) is the path feasibility record based on historical data, omega c and omega r are reputation evaluation weights, and R(P) is the safety score of path P, reflecting the potential risk of the path.

[0036] The final optimization objective function is updated as:

[0037]

[0038] Wherein, Gamma(P) is the path optimization objective function, omega1 is the time-related weight, affecting the speed of path selection, T(P) is the total time of path P, measuring the efficiency of the path, omega2 is the risk-related weight, affecting the speed of path selection, R(P) is the safety score of path P, reflecting the potential risk of the path, and omega3 is the reputation value-related weight.

[0039] Specifically, the violation behavior recognition module adopts a combination of multi-channel time convolution network and self-attention mechanism to automatically identify the violation behavior of the work personnel, and the process specifically includes:

[0040] First, set the time window to T, and extract the feature vector x t ∈R d , wherein t represents the time index, and d is the feature dimension; The video frame sequence is represented as X=[x1, x2, x3, …, x T ]∈R T×d

[0041] Second, perform a convolution operation on the input feature sequence, with the parameterized convolution kernel being W∈R. K×d Where K is the kernel size, the convolution output is calculated as follows:

[0042] Z t =σ(W*x t +b)

[0043] Where σ is the activation function, b is the bias, and Z is the bias value. t The output value of the convolutional layer;

[0044] Third, global features G are obtained through max pooling operations:

[0045] G = MaxPool(Z) t )

[0046] Fourth, after obtaining the global features, a self-attention mechanism is introduced to enhance the representational power of the features. For any g... i ∈G, calculate attention weights:

[0047]

[0048] Among them, A ij It is a global feature g i For global feature g j Attention weights, g i It is the i-th global feature, g j It is the j-th global feature, n is the number of global features, g i T This indicates that for global feature g i transpose;

[0049] Fifth, attention weights are applied to global features to form a weighted feature G. att :

[0050]

[0051] Sixth, the weighted features are mapped through a fully connected layer to output the probability of the violation:

[0052] y = softmax(W fc G att +b fc )

[0053] Among them, W fc and b fc These are all parameters of the fully connected layer, and softmax represents normalization.

[0054] Specifically, the risk analysis module includes calculating the risk value R for each type of violation h. h :

[0055] R h = f (P h , S, E)

[0056] Wherein, P h is the probability of violation behavior h, S is environmental factors, E is external event influence factor.

[0057] Compared with the prior art, the present application has the following beneficial effects:

[0058] The offshore project demolition operation whole process monitoring system provided by the present application provides solid guarantee for efficient and safe demolition in complex offshore environment, avoids the problems of information silos and slow reaction in traditional project management, greatly improves the safety and efficiency of offshore demolition operation, and has wide application prospect. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only preferred embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0060] Figure 1 The offshore project demolition operation whole process monitoring system provided by the present application provides solid guarantee for efficient and safe demolition in complex offshore environment, avoids the problems of information silos and slow reaction in traditional project management, greatly improves the safety and efficiency of offshore demolition operation, and has wide application prospect. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical scheme and advantages of the present application more obvious, the following will describe the example embodiments according to the present application with reference to the drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described here. Based on the embodiments of the present application described in the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the protection scope of the present application.

[0062] In the following description, a large number of specific details are given in order to provide a more thorough understanding of the present application. However, it is obvious for those skilled in the art that the present application can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present application, some technical features known in the art are not described.

[0063] It is to be understood that the application can assume various alternative embodiments, and should not be limited to the examples described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the application to those skilled in the art.

[0064] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising", when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. As used herein the term "and / or" includes any and all combinations of associated items.

[0065] For a thorough understanding of the application, a detailed description will be made in the following description with reference to the following structures, so as to illustrate the technical solutions presented by the application. The alternative embodiments of the application are described in detail as follows, however, in addition to these detailed descriptions, the application can have other implementation manners.

[0066] Referring to Figure 1 An offshore project demolition operation whole-process monitoring system, the system comprises a data acquisition module, a work ticket generation module, a demolition operation path planning module, a demolition data collection and case filing module, a violation behavior identification module, a risk analysis module, a visualization module;

[0067] The data acquisition module uses a drone to perform three-dimensional scanning on the demolition area, to obtain high-precision three-dimensional data, uses a combination of a laser radar and a high-definition camera to obtain video image data and create a three-dimensional model of the area, and collects demolition team personnel information, main machinery and ship information;

[0068] The work ticket generation module obtains demolition area, demolition team personnel information, demolition project, demolition object, demolition sequence, and demolition expected completion time, and generates a work operation ticket;

[0069] The demolition operation path planning module constructs a high-precision three-dimensional visualization model based on the obtained three-dimensional data, allows a user to intuitively view the structure and potential risks of the demolition area in a virtual environment, and generates an operation path planning graph based on a path planning algorithm of reinforcement learning in combination with the spatial position of the object and the demolition sequence;

[0070] The demolition data collection and case filing module is used to collect all data related to the demolition operation, including work operation ticket content, video images, to form a complete operation file;

[0071] The violation behavior identification module analyzes the real-time video stream of the acquired video image data, and automatically identifies the violation behavior of the work personnel by combining a multi-channel time convolution network with a self-attention mechanism.

[0072] The risk analysis module is used for feedback on the risk type that the violation behavior may correspond to, and simultaneously analyzes and identifies potential risks in the demolition operation process.

[0073] The visualization module is used for providing a visual interface, supporting real-time updating of the mobile terminal and the computer terminal, real-time monitoring and tracking of the demolition progress, whether the object is demolished, and whether the demolition sequence is correct, and displaying the same on the interface.

[0074] For example, the offshore project demolition operation whole-process monitoring system provided by the present application can monitor the whole process of the demolition operation under consideration of a complex offshore environment, and the working principle is as follows: first, the data acquisition module collects information of the demolition team personnel, dynamic information of the main machinery and the ship, and focuses on monitoring the working state and dynamics of the ship, and cooperates with the ship network to view the working condition of the ship, so as to determine whether it is working normally; second, the work ticket generation module obtains the demolition area, demolition team personnel information, demolition project, demolition object, demolition sequence, and demolition expected completion time, and generates a work operation ticket; third, the demolition operation path planning module establishes a three-dimensional view, and generates an operation path planning diagram according to the object position and the demolition sequence; fourth, the demolition data collection and filing module collects all data related to the demolition operation, including the content of the work operation ticket, video images, etc., forms a complete operation archive, and facilitates subsequent analysis and reference; fifth, the violation behavior identification module accurately identifies the violation behavior in the demolition operation process; sixth, the risk analysis module identifies and feeds back the risk type that the violation behavior may correspond to, and simultaneously analyzes and identifies potential risks in the demolition operation process; seventh, the visualization module provides a visual interface, supports real-time updating of the mobile terminal and the computer terminal, real-time monitoring and tracking of the demolition progress, whether the object is demolished, and whether the demolition sequence is correct, and displays the same on the interface, and real-time understanding of the operation condition. Through effective monitoring, the safety and efficiency of the operation can be improved.

[0075] Specifically, high-precision three-dimensional data is acquired for data cleaning, noise and repeated points are removed, the processed three-dimensional data is used to generate a three-dimensional mesh model, a triangular subdivision algorithm is used for smoothing processing to form a continuous and smooth surface, a texture mapping algorithm is used to add material properties to the model, and the positions of structural features and potential hazards are marked on the model.

[0076] Specifically, in the process of generating the three-dimensional mesh model, a triangular subdivision algorithm is used for smoothing processing to form a continuous and smooth surface, and the specific process is as follows:

[0077] For each vertex, identify its neighborhood triangle and calculate the average position of the vertex, neighborhood triangle set to a fixed radius or neighborhood point number, to ensure that the processed results are smooth without losing structure;

[0078] Update the coordinate position of the current point by averaging the neighborhood point coordinates, and maintain the consistency of the normal vector during the smoothing process to avoid causing surface inversion.

[0079] For example, the smoothing process is performed to ensure that the generated three-dimensional mesh model not only has good visual effect, but also effectively preserves the structural characteristics.

[0080] Specifically, a texture mapping algorithm is used to add material properties to the three-dimensional network model, and the location of the structure feature and potential hazard is marked on the model, as follows:

[0081] Insert the calibrated and processed three-dimensional data into the three-dimensional network model, and use normal mapping and displacement mapping to enhance the three-dimensional effect and material texture. Normal mapping can generate detailed surfaces under lighting conditions.

[0082] For example, material mapping and detail enhancement are important steps in three-dimensional modeling, especially when practical applications such as demolition work are required, as this step is directly related to the authenticity and operability of the model.

[0083] Identify and mark key structural features and potential hazard areas in the three-dimensional network model, and extract key structures from the model to ensure that the key points are linked to actual demolition work.

[0084] Color code the potential hazard areas: red represents heavy object hanging areas to alert potential safety risks; yellow represents damaged areas to prompt special attention; blue represents the height of the hazard area, including the gap between the upper and lower layers, to facilitate workers' understanding of the working space.

[0085] Specifically, the work ticket generation module generates a work operation ticket, which includes the following steps:

[0086] In the operation ticket, the name, skill, health status, and responsibility division information of each member in the operation are listed in detail;

[0087] Automatically identify the object to be demolished, including the type, material characteristics, and potential hazards of the object, and display the historical information of the demolition project and the precise marking of the object location;

[0088] In the generated operation ticket, embed the real-time updated three-dimensional model, clearly mark the location of the demolished object, and use color coding to represent different levels of danger;

[0089] Based on previous data analysis and site investigation, the system automatically generates recommended demolition steps and provides operation specifications, including necessary safety measures and emergency plans.

[0090] According to historical cases and real-time data, the duration of each demolition step is predicted, and the estimated completion time is generated in the operation ticket.

[0091] For example, the work operation ticket content is as follows:

[0092] Project Name: Offshore Wind Farm Foundation Demolition

[0093] Creation Time: September 19, 2024

[0094] Project Number: OCEAN-2024-018

[0095] 1. Demolition Team Information

[0096] Team Leader: Li Yi

[0097] Member List:

[0098] Liu Er - Skills: Offshore Construction, Mechanical Operation; Health Status: Good Chen San - Skills: Electrician, Equipment Maintenance; Health Status: Good Zhang Si - Skills: Safety Management; Health Status: Good

[0099] Zhao Wu - Skills: Diving Operations; Health Status: Good

[0100] 2. Demolition Project Description

[0101] Target Object: Offshore Wind Turbine Foundation Demolition Reason: Equipment update, construction of new type of wind turbine

[0102] Object Type: Reinforced Concrete Foundation

[0103] Address: North Sea Wind Farm

[0104] Coordinates: (23.55°N, 118.92°E)

[0105] 3. Object Location

[0106] Marked Location:

[0107] 3D view generated using 3D modeling software (such as Revit or Navisworks): 3D model

[0108] Key areas marked in the diagram:

[0109] Area A: Demolish the first wind turbine foundation

[0110] Area B: Demolish the second wind turbine foundation

[0111] C Zone: Clean up work area

[0112] 4. Dismantling sequence

[0113] Step breakdown:

[0114] Ensure all safety measures are in place (life jackets, diving equipment, etc.).

[0115] Check marine weather conditions and tide forecasts to ensure a safe working window.

[0116] Conduct site reconnaissance to confirm the work area.

[0117] Use cutting equipment to dismantle the first wind turbine foundation in Zone A.

[0118] Use cutting equipment to dismantle the second wind turbine foundation in Zone B.

[0119] Clean up the debris from the dismantling and record it.

[0120] Conduct safety checks and site evaluations.

[0121] 5. Estimated completion time

[0122] Dismantling time estimate:

[0123] Based on historical data analysis and site conditions, the entire dismantling process is estimated to take 10 hours.

[0124] Time allocation:

[0125] Preparation: 1 hour

[0126] Dismantling operation: 7 hours

[0127] Cleaning and safety checks: 2 hours

[0128] Specifically, the specific process of generating a work path planning graph in the demolition operation path planning module based on the reinforcement learning path planning algorithm combined with the spatial position of the object and the demolition sequence is as follows:

[0129] Reinforcement learning update formula:

[0130] Q(s t , a t )←Q(s t , a t )+α((R(s t , a t )+γmax Q(s t+1 , a)-Q(s t , a t ))

[0131] Where Q is the optimal action selected to achieve the maximum total return, Q(st , a t ) represents the expected long-term cumulative reward of performing action a t in state s t , max Q(s t+1 , a) represents the maximum value of Q for all possible actions a in state s t+1 , i.e., the best expected return that can be achieved in the next state s t is the state at time t, a t is the action at time t, α is the learning rate, R is the reward function, γ is the discount factor, and s t+1 is the state at time t+1.

[0132] To maximize the expected reward of the final path P, we have:

[0133] P = arg max Q(s t , a t )

[0134] To introduce the trust level of path selection, define the reputation value C(P) to evaluate the reliability of different paths P, the formula is as follows:

[0135] C(P) = ω c ·C t (P) + ω r ·R(P)

[0136] Where C t (P) is the path feasibility record based on historical data, ω c and ω r are reputation evaluation weights, and R(P) is the safety score of path P, reflecting the potential risk of the path.

[0137] The final optimization objective function is updated as:

[0138]

[0139] Where Γ(P) is the path optimization objective function, ω1 is the time-dependent weight affecting the speed of path selection, T(P) is the total time of path P, measuring the efficiency of the path, ω2 is the risk-dependent weight affecting the speed of path selection, R(P) is the safety score of path P, reflecting the potential risk of the path, and ω3 is the reputation value-dependent weight.

[0140] For example, a path planning algorithm based on reinforcement learning is used, and multi-objective optimization is performed on the path. Finally, the optimized path and related risk markers are displayed on a three-dimensional visualization model, which helps to improve the efficiency and safety of demolition operations.

[0141] Specifically, the process of automatically identifying the violation behavior of the operation personnel by the violation behavior identification module combining the multi-channel time convolution network and the self-attention mechanism specifically comprises:

[0142] First, set the time window as T, and extract the feature vector x t ∈R d from each video frame, where t represents the time index, and d is the feature dimension; the video frame sequence is represented as X=[x1,x2,x3,…,x T ]∈R T×d

[0143] Second, perform convolution operation on the input feature sequence, and the parameterized convolution kernel is W∈R K×d , where K is the size of the convolution kernel, and the convolution output is calculated as:

[0144] Z t =σ(W*x t +b)

[0145] where σ is the activation function, b is the bias, and Z t is the output value of the convolution layer;

[0146] Third, obtain the global feature G by the max-pooling operation:

[0147] G=MaxPool(Z t )

[0148] Fourth, after obtaining the global feature, the self-attention mechanism is introduced to enhance the representation ability of the feature. For any g i ∈G, the attention weight is calculated:

[0149]

[0150] where A ij is the attention weight of the global feature g i to the global feature g j , g i is the i-th global feature, g j is the j-th global feature, n is the number of global features, and g i T represents the transpose of the global feature g i ;

[0151] Fifth, apply the attention weight to the global feature to form the weighted feature G att :

[0152]

[0153] Sixth, map the weighted feature through the fully connected layer to output the probability of the violation behavior:

[0154] y = softmax(W fc G att +b fc )

[0155] wherein W fc and b fc are parameters of the full connection layer, and softmax represents normalization.

[0156] Illustratively, by combining the multi-channel time convolution network with the self-attention mechanism, the accuracy and robustness of the violation behavior recognition are improved, not only the recognition efficiency is improved, but also the adaptability to complex operation environment is enhanced, which provides strong guarantee for on-site operation safety.

[0157] Specifically, the risk analysis module specifically includes calculating the risk value R h of each violation behavior h.

[0158] R h = f(P h , S, E)

[0159] wherein P h is the probability of the violation behavior h, S is the environmental factor, and E is the external event influence factor.

[0160] Illustratively, by analyzing and identifying the violation behavior and potential risk, the occurrence of accidents can be effectively prevented, the safety of personnel and equipment is guaranteed, and the project is promoted to be completed smoothly.

[0161] The above only describes the preferred embodiments of the present application and is not intended to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A system for monitoring the entire process of a demolition work of an offshore project, characterized in that, The system comprises a data acquisition module, a work ticket generation module, a demolition operation path planning module, a demolition data collection and case filing module, a violation behavior identification module, a risk analysis module, and a visualization module; the data acquisition module uses a UAV to perform three-dimensional scanning on a demolition area to obtain high-precision three-dimensional data, uses a laser radar and a high-definition camera in combination to obtain video image data and create a three-dimensional model of the area, and collects demolition team personnel information, main machinery and ship information; the work ticket generation module obtains demolition area, demolition team personnel information, demolition project, demolition object, demolition sequence, and demolition estimated completion time, and generates a work operation ticket; The process of generating the work operation ticket by the work ticket generation module specifically comprises: listing the name, skill, health status, and responsibility division information of each member in the operation ticket; Automatically identifying objects to be demolished, specifically including the type, material properties, and potential hazards of the objects, displaying historical information of the demolition project, and accurately marking the location of the object; In the generated operation ticket, a real-time updated three-dimensional model is embedded, the location of the demolition object is explicitly marked, and different danger levels are represented by color coding; In combination with the data analysis and on-site investigation in the early stage, the system automatically generates recommended demolition steps and provides operation specifications, including necessary safety measures and emergency plans; According to historical cases and real-time data, the duration of each demolition step is predicted, and the estimated completion time is generated in the operation ticket; the demolition operation path planning module constructs a high-precision three-dimensional visualization model based on the obtained three-dimensional data, allows users to intuitively view the structure and potential risks of the demolition area in a virtual environment, and generates a work path planning graph based on a path planning algorithm of reinforcement learning in combination with the spatial location of the object and the demolition sequence; According to the update formula of reinforcement learning, the expected reward of the final path P is maximized, that is: wherein, representing in performing an action a long-term cumulative reward, representing a state at time t, representing an action at time t; The trust degree of path selection is introduced, and the reputation value C(P) is defined to evaluate the reliability of different paths P, and the formula is as follows: wherein, is a path feasibility record based on historical data, and are both reputation evaluation weights, is a security score of the path reflecting potential risks of the path; and the final optimization objective function is: wherein, is a path optimization objective function, is a time-dependent weight affecting the speed of path selection, is a total time of the path, measures the efficiency of the path, is a risk-dependent weight affecting the speed of path selection, is a safety score of the path, reflects the potential risk of the path, is a reputation value-dependent weight; The demolition data collection and case filing module is used to collect all data related to the demolition work, including the content of the work operation ticket, video images, to form a complete job file; The violation behavior identification module analyzes the real-time video stream of the obtained video image data, and uses a combination of a multi-channel time convolution network and a self-attention mechanism to automatically identify the violation behavior of the work personnel; The risk analysis module is used to feedback the risk type that may correspond to the violation behavior, and to analyze and identify the potential risks in the demolition operation process; The visualization module is used to provide a visualization interface, supports real-time updating of mobile and computer terminals, and monitors and tracks the demolition progress, whether the object is demolished, and whether the demolition sequence is correct in real time and displays them on the interface.

2. A system for monitoring the whole process of a marine project demolition operation according to claim 1, characterized in that, The high-precision three-dimensional data obtained is cleaned to remove noise and duplicate points, the processed three-dimensional data is generated into a three-dimensional mesh model, the triangular subdivision algorithm is used for smoothing processing to form a continuous and smooth surface, the texture mapping algorithm is used to add material properties to the model, and the positions of the structural features and potential hazards are marked on the model.

3. A system for monitoring the whole process of a marine project demolition operation according to claim 2, characterized in that, In the process of generating a three-dimensional mesh model, a triangulation algorithm is used for smoothing to form a continuous and smooth surface, which is as follows: for each vertex, identify the neighborhood triangles around it and calculate the average position of the vertex, the neighborhood triangles are set to a fixed radius or a number of neighborhood points, to ensure that the processed result is smooth without losing structure; The coordinate position of the current point is updated by averaging the coordinates of the neighborhood points, and the normal vector consistency is maintained during the smoothing process to prevent surface inversion.

4. A system for monitoring the whole process of a marine project demolition operation according to claim 3, characterized in that, A texture mapping algorithm is used to add material attributes to the three-dimensional mesh model, and the positions of structural features and potential hazards are labeled on the model, which is as follows: insert the calibrated and processed three-dimensional data into the three-dimensional mesh model, use normal mapping and displacement mapping to improve the three-dimensional effect and material texture, normal mapping can generate a detailed surface under lighting conditions; identify and label key structural features and potential hazard areas in the three-dimensional mesh model, extract the key structures in the model to ensure that the key points are linked to the actual demolition operation; Color division is performed on the potential hazard areas: red represents heavy object hanging areas to alert potential safety risks; yellow represents damaged areas to prompt special attention; Blue represents the height of the hazard area, including the gap between the upper and lower layers, to facilitate workers to understand the space of the work.

5. The system according to claim 1, wherein, The update formula of the reinforcement learning is as follows: wherein, is the optimal action selected to obtain the maximum total reward, represents the expected long-term cumulative reward, of performing action in state represents the maximum of the Q-values for all possible actions a in state , i.e. the best expected reward that can be obtained in the next state, is the state at time t, is the action at time t, is the learning rate, is the reward function, is the discount factor, is the state at time t+1.

6. A system for monitoring the whole process of a marine project demolition operation according to claim 5, characterized in that, The process of automatically identifying the violation behavior of the work personnel by the violation behavior identification module combining a multi-channel time convolution network with a self-attention mechanism includes: First, set the time window as T, extract the feature vector for each video frame , where t represents the time index, d is the feature dimension; the video frame sequence is represented as: Second, a convolution operation is performed on the input feature sequence, and the parameterized convolution kernel is , where K is the size of the convolution kernel, and the convolution output is calculated as: wherein, is an activation function, b is a bias, is an output value of the convolution layer; Third, global features are obtained by a max-pooling operation : Fourth, after obtaining the global features, a self-attention mechanism is introduced to enhance the representation ability of the features. For any G, calculate attention weights: wherein, is a global feature attention weight for global feature , is the th global feature, is the th global feature, is the number of global features, denotes the transpose of global feature ; Fifth, attention weights are applied to the global features to form weighted features : Sixth, the weighted features are mapped through a fully connected layer to output the probability of the violation behavior: wherein, and are parameters of the fully connected layer, denotes a normalization thereof.

7. A system for monitoring the whole process of a marine project demolition operation according to claim 6, characterized in that, The risk analysis module specifically includes calculating a risk value h of each violation : wherein, is the probability of violation h, is the environmental factor, is the external event influencing factor.

Images