Safety monitoring device and system for bollard structure in large coastal ports

Abstract

The present invention provides a device and system for monitoring the structural safety of mooring posts in large coastal ports, and relates to the field of data processing technology, wherein the device includes: a design information interaction module, an operation and maintenance information interaction module, a monitoring circle analysis module, a sample data acquisition module, a model building training module, a data acquisition execution module, and a structural safety analysis module. Based on the module scheduling of the device for monitoring the structural safety of mooring posts in large coastal ports, the technical problem that the monitoring of the structural safety status of port mooring posts in the prior art depends on the manual experience of port operation and maintenance staff, and there are defects such as low efficiency and low recognition accuracy in identifying risks of mooring post structures. The technical effect of reducing the difficulty of identifying the structural safety status of port mooring posts during use and improving the recognition accuracy and recognition efficiency of the structural safety status of port mooring posts during use is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a device and system for monitoring the safety of bollard structures in large coastal ports. Background Art

[0002] At present, the safety status monitoring of port bollard structures mainly relies on the manual experience of port operation and maintenance staff. However, manual experience may be limited by individual subjective judgment and experience accumulation, which often makes it difficult to fully and accurately identify potential risks, thereby increasing safety hazards.

[0003] In summary, the safety status monitoring of port bollard structures in the prior art relies on the manual experience of port operation and maintenance personnel, and there are technical problems such as low efficiency and low accuracy in risk identification of bollard structures. Summary of the invention

[0004] The present application provides a device and system for monitoring the safety of mooring bollard structures in large coastal ports, which is used to solve the technical problem that the safety status monitoring of port mooring bollard structures in the prior art relies on the manual experience of port operation and maintenance personnel, and there are defects such as low efficiency and low identification accuracy of mooring bollard structure risks.

[0005] In view of the above problems, the present application provides a device and system for monitoring the safety of mooring bollard structures in large coastal ports.

[0006] The present application provides a device for monitoring the safety of mooring bollard structures in large coastal ports, the device comprising: a design information interaction module, for interactively obtaining target design information, and positioning and calling bollard data based on the target design information, so as to obtain multiple bollard design information of multiple bollards in the target port; an operation and maintenance information interaction module, for interactively obtaining target operation and maintenance information, and positioning and calling bollard data based on the target operation and maintenance information, so as to obtain multiple bollard operation and maintenance information of the multiple bollards; a monitoring circle analysis module, for classifying the multiple bollards according to the multiple bollard design information and the multiple bollard operation and maintenance information, so as to obtain K bollard monitoring circles, wherein each bollard monitoring circle has a bollard monitoring sequence identifier; a sample data acquisition module, for classifying the multiple bollards according to the K bollard monitoring circles The sub-call obtains K sample defect data sets, wherein each sample defect data set includes a sample apparent defect data set and a sample flaw detection data set; a model construction training module is used to construct a model based on a convolutional neural network, and use the K sample defect data sets as model training data to generate K bollard safety monitors; a data acquisition execution module is used to pre-configure K integrated monitoring devices for the K bollard monitoring circles, and schedule the K integrated monitoring devices to perform data acquisition on the multiple bollards according to the K bollard monitoring sequence identifiers of the K bollard monitoring circles, so as to obtain K safety monitoring analysis arrays; a structural safety analysis module is used to synchronize the K safety monitoring analysis arrays to the K bollard safety monitors to perform structural safety analysis, so as to obtain bollard structural risk information.

[0007] The technical solution provided in this application has at least the following technical effects or advantages:

[0008] The device provided in the embodiment of the present application uses a design information interaction module for interactively obtaining target design information, and based on the target design information, locates and calls the mooring bollard data to obtain multiple bollard design information of multiple bollards in the target port; an operation and maintenance information interaction module for interactively obtaining target operation and maintenance information, and based on the target operation and maintenance information, locates and calls the mooring bollard data to obtain multiple bollard operation and maintenance information of the multiple bollards; a monitoring circle analysis module for classifying the multiple bollards according to the multiple bollard design information and the multiple bollard operation and maintenance information to obtain K bollard monitoring circles, wherein each bollard monitoring circle has a bollard monitoring sequence identifier; a sample data acquisition module for obtaining K samples according to the K bollard monitoring circle calls A defect data set, wherein each sample defect data set includes a sample apparent defect data set and a sample flaw detection data set; a model building training module, used to build a model based on a convolutional neural network, and use the K sample defect data sets as model training data to generate K bollard safety monitors; a data acquisition execution module, used to pre-configure K integrated monitoring devices for the K bollard monitoring circles, and schedule the K integrated monitoring devices to perform data acquisition on the multiple bollards according to the K bollard monitoring sequence identifiers of the K bollard monitoring circles, and obtain K safety monitoring analysis arrays; a structural safety analysis module, used to synchronize the K safety monitoring analysis arrays to the K bollard safety monitors to perform structural safety analysis and obtain bollard structural risk information. The technical effect of reducing the difficulty of identifying the structural safety status of port bollards during use and improving the accuracy and efficiency of identifying the structural safety status of port bollards during use is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 Schematic diagram of the structure safety monitoring device for bollard structures in large coastal ports provided for this application;

[0010] Figure 2 A structural schematic diagram of a mooring bollard monitoring circle obtained in the large coastal port mooring bollard structure safety monitoring device provided in this application. DETAILED DESCRIPTION

[0011] The present application provides a device and system for monitoring the safety of mooring bollard structures in large coastal ports, which is used to solve the technical problem that the monitoring of the safety status of port mooring bollard structures in the prior art relies on the manual experience of port operation and maintenance personnel, and there are defects such as low efficiency and low recognition accuracy in identifying the risks of the mooring bollard structures. The technical effect of reducing the difficulty of identifying the structural safety status of port mooring bollards during use and improving the recognition accuracy and recognition efficiency of the structural safety status of port mooring bollards during use is achieved.

[0012] Below, the technical solutions in the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. 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. It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, rather than all of them.

[0013] Embodiment 1

[0014] like Figure 1 As shown, the present application provides a safety monitoring device for mooring bollard structures in large coastal ports, comprising:

[0015] The design information interaction module 100 is used to interactively obtain target design information, and locate and call the bollard data based on the target design information to obtain multiple bollard design information of multiple bollards in the target port;

[0016] Specifically, the design information interaction module in this embodiment is used to call the mooring bollard design data. Specifically, the design information interaction module obtains the target design information that records the port construction design data by interacting with the data storage center of the target port, and then locates and calls the mooring bollard data based on the target design information to obtain multiple mooring bollard design information of multiple mooring bollards in the target port. The design information of each mooring bollard specifically includes the geometric parameters, material performance parameters, thickness of the mooring bollard surface coating material, thickness of the mooring bollard metal material, and location information of the mooring bollard in the target port.

[0017] The operation and maintenance information interaction module 200 is used to interactively obtain target operation and maintenance information, and perform positioning call of bollard data based on the target operation and maintenance information to obtain multiple bollard operation and maintenance information of the multiple bollards;

[0018] Specifically, in this embodiment, the operation and maintenance information interaction module is used to call the historical operation and maintenance time data of the mooring bollard. Specifically, the operation and maintenance information interaction module interacts with the data storage center of the target port to obtain the target operation and maintenance information that records the historical operation and maintenance time of all mooring bollards in the target port, and locates and calls the mooring bollard data based on the target operation and maintenance information to obtain multiple bollard operation and maintenance information of the multiple mooring bollards, and each bollard operation and maintenance information is essentially an operation and maintenance time series.

[0019] A monitoring circle analysis module 300 is used to classify the multiple mooring posts according to the multiple mooring post design information and the multiple mooring post operation and maintenance information to obtain K mooring post monitoring circles, wherein each mooring post monitoring circle has a mooring post monitoring sequence identifier;

[0020] In one embodiment, Figure 2 As shown, the monitoring circle analysis module 300 also includes:

[0021] An analysis network construction module is used to pre-construct a structural design analysis network, wherein the structural design analysis network includes a standard data call layer, a similarity constraint judgment layer and a structural feature analysis layer, and the standard data call layer, the similarity constraint judgment layer and the structural feature analysis layer are cascaded;

[0022] a primary classification execution module, configured to use the design information of the plurality of mooring posts as a classification reference, perform primary classification processing of the plurality of mooring posts based on the structural design analysis network, and obtain a plurality of mooring post classification sets;

[0023] The secondary classification execution module is used to pre-build an operation and maintenance synchronization threshold, and perform secondary classification processing on the multiple mooring bollard classification sets according to the operation and maintenance synchronization threshold and the multiple mooring bollard operation and maintenance information to obtain the K mooring bollard monitoring circles.

[0024] In one embodiment, the analysis network construction module further includes:

[0025] A sample data scheduling module, for interactively obtaining a plurality of standard structural parameter sets of a plurality of standard bollards, and performing data division on the plurality of standard structural parameter sets based on preset structural parameter types, to obtain a plurality of sample geometric parameter sets, a plurality of sample material parameter sets and a plurality of sample coating parameter sets, wherein the preset structural parameter types include geometric parameter types, material parameter types and coating types;

[0026] A data association storage module, used for associating and storing the plurality of standard bollards, the plurality of sample geometric parameter sets, the plurality of sample material parameter sets and the plurality of sample coating parameter sets based on the knowledge graph, and generating the standard data call layer;

[0027] An engine scheduling execution module, used for interactively obtaining a substrate feature calling engine, and synchronizing the substrate feature calling engine to the similarity constraint judgment layer;

[0028] The function construction execution module is used to pre-construct the structure similarity function and synchronize the structure similarity function to the structure feature analysis layer to complete the construction of the structure design analysis network.

[0029] In one embodiment, the function construction execution module further includes:

[0030] A calculation module, the calculation module is used to call the structural similarity function;

[0031] The structural similarity function is as follows:

[0032]

[0033] Where S1 is the geometric similarity coefficient, A is the geometric parameter set of the bollard in the bollard design information, A=a1,a2,...,a i , B is the sample geometric parameter set, B = b1, b2, ..., b i , a j and b j is the parameter value of the jth indicator in the two data sets A and B;

[0034]

[0035] Where S2 is the material similarity coefficient, C is the bollard material parameter set in the bollard design information, C = c1, c2, ..., c i , D is the sample material parameter set, D=d1,d2,...,d i , c j and d j is the parameter value of the jth indicator in the two parameter sets C and D;

[0036]

[0037] Where S3 is the coating similarity coefficient, E is the bollard coating parameter set in the bollard design information, and E = e1, e2, ..., e i , F is the sample coating parameter set, F = f1, f2, ..., f i , e j and f j is the parameter value of the jth indicator in the two data sets E and F;

[0038] S = w1S1+w2S2+w3S3;

[0039] Among them, S is the structural similarity, w1 is the geometric weight value, w2 is the material weight value, and w3 is the coating weight value.

[0040] Specifically, in this embodiment, the function of the monitoring circle analysis module is to group the operation and maintenance of multiple mooring bollards based on the proximity of their operation and maintenance time and the similarity of their types, so as to improve the convenience of safe operation and maintenance of port mooring bollard structures.

[0041] Specifically, the monitoring circle analysis module performs bollard grouping through sub-modules such as the analysis network construction module and the primary classification execution module contained in the module.

[0042] In this embodiment, the analysis network construction module is used to pre-construct the structural design analysis network for automatically classifying mooring posts based on mooring post design. The structural design analysis network includes a standard data call layer, a similarity constraint judgment layer and a structural feature analysis layer. The standard data call layer, the similarity constraint judgment layer and the structural feature analysis layer are cascaded.

[0043] The analysis network construction module specifically includes submodules such as a sample data scheduling module and a data association storage module. The specific process of the analysis network construction module constructing the structural design analysis network is as follows:

[0044] Based on the sample data scheduling module, the interactive bollard manufacturer obtains multiple standard structural parameter sets of multiple standard bollards, and divides the multiple standard structural parameter sets based on preset structural parameter types to obtain multiple sample geometric parameter sets, multiple sample material parameter sets and multiple sample coating parameter sets, wherein the preset structural parameter types include geometric parameter types, material parameter types and coating types.

[0045] It should be understood that the coating is a material applied to the surface of the mooring post to improve the anti-corrosion performance of the mooring post, the material is a metal material of the tensile structure of the mooring post, the material parameters are the properties of the metal material, including but not limited to strength, density, and elastic modulus, and the geometric parameters include geometric properties such as the height, diameter, and side length of the mooring post.

[0046] The data association storage module performs automatic association storage of the multiple standard bollards, the multiple sample geometric parameter sets, the multiple sample material parameter sets and the multiple sample coating parameter sets based on the knowledge graph to generate the standard data call layer.

[0047] The engine scheduling execution module interactively obtains the substrate feature calling engine that can identify and extract the material name of the coating material and the metal material name from the standard structure parameter set, and synchronizes the substrate feature calling engine to the similarity constraint judgment layer.

[0048] The function construction execution module is used to pre-construct a structural similarity function and synchronize the structural similarity function to the structural feature analysis layer to complete the construction of the structural design analysis network.

[0049] The function construction execution module constructs the structural similarity function based on the calculation module therein. Specifically, the structural similarity function constructed by the calculation module is as follows:

[0050]

[0051] Where S1 is the geometric similarity coefficient, A is the geometric parameter set of the bollard in the bollard design information, A=a1,a2,...,a i , B is the sample geometric parameter set, B = b1, b2, ..., b i , a j and b j is the parameter value of the jth indicator in the two data sets A and B;

[0052]

[0053] Where S2 is the material similarity coefficient, C is the bollard material parameter set in the bollard design information, C = c1, c2, ..., c i , D is the sample material parameter set, d=D1,d2,...,d i , c j and d j is the parameter value of the jth indicator in the two parameter sets C and D;

[0054]

[0055] Where S3 is the coating similarity coefficient, E is the bollard coating parameter set in the bollard design information, and E = e1, e2, ..., e i , F is the sample coating parameter set, F = f1, f2, ..., f i , e j and f j is the parameter value of the jth indicator in the two data sets E and F;

[0056] S = w1S1+w2S2+w3S3;

[0057] Among them, S is the structural similarity, w1 is the geometric weight value, w2 is the material weight value, and w3 is the coating weight value.

[0058] The first-level classification execution module starts running after the analysis network construction module completes the construction of the structural design analysis network. Specifically, the process of the first-level classification execution module running the structural design analysis network is as follows:

[0059] The first-level classification execution module randomly calls the first bollard design information of the first bollard from the multiple bollard design information, and divides the first bollard design information based on the preset structural parameter type to obtain a first bollard geometric parameter set, a first bollard material parameter set and a first bollard coating parameter set.

[0060] The first-level classification execution module synchronizes the first bollard geometry parameter set, the first bollard material parameter set and the first bollard coating parameter set to the structural design and analysis network, and obtains the first sample geometry parameter set, the first sample material parameter set and the first sample coating parameter set of the first standard bollard in the standard data calling layer of the structural design and analysis network.

[0061] The first bollard geometry parameter set, the first bollard material parameter set and the first bollard coating parameter set, the first sample geometry parameter set, the first sample material parameter set and the first sample coating parameter set are synchronized to the similarity constraint judgment layer to identify material information and judge whether the material composition of the sample bollard is consistent with the material composition of the first bollard; if they are consistent, the first bollard geometry parameter set, the first bollard material parameter set and the first bollard coating parameter set, the first sample geometry parameter set, the first sample material parameter set and the first sample coating parameter set are mapped and synchronized to the structural feature analysis layer; based on the three function fractions of the structural feature analysis layer, multi-dimensional similarity calculation of geometry structure, material property and metal material-coating material thickness is performed to obtain a first similarity index.

[0062] Then, the primary classification execution module adopts the same method to compare the similarities between the first bollard and multiple standard bollards one by one, and serializes the obtained multiple similarity indexes, and classifies the type of the standard bollard with the highest similarity as the type of the first bollard.

[0063] Then, the primary classification execution module adopts the same method to classify the multiple mooring posts one by one, so as to perform primary classification processing on the multiple mooring posts and obtain multiple mooring post classification sets, which correspond to multiple standard mooring posts.

[0064] The secondary classification execution module is used to perform secondary classification of the mooring posts in each mooring post classification set based on the historical operation and maintenance time of the mooring posts in each mooring post classification set. The specific data processing process of the secondary classification execution module is as follows:

[0065] Based on a pre-built operation and maintenance synchronization threshold (for example, the operation and maintenance time deviation does not exceed 2 hours), after the multiple bollard operation and maintenance information is grouped based on the multiple bollard classification sets to obtain multiple bollard operation and maintenance information groups.

[0066] A first bollard operation and maintenance information group is obtained by calling from multiple bollard operation and maintenance information groups, wherein the first bollard operation and maintenance information group includes H operation and maintenance time series of H bollards, the H operation and maintenance time series are imaged to obtain an operation and maintenance time scatter plot, and multiple connection time bars are generated based on an operation and maintenance synchronization threshold to divide the operation and maintenance time scatter plot, so as to intuitively obtain multiple bollard combinations whose operation and maintenance times are within the same operation and maintenance synchronization threshold, and the multiple bollard combinations are a part of the bollard monitoring circle of the K bollard monitoring circles.

[0067] By analogy, the secondary classification execution module performs operation and maintenance time analysis on each mooring post classification set, obtains multiple groups of mooring post monitoring circles of multiple mooring post classification sets, and obtains the K mooring post monitoring circles after group elimination.

[0068] Furthermore, the port position parameters of the M mooring posts in each mooring post monitoring circle are called, and then the monitoring patrol route is fitted based on the position parameters to generate the mooring post monitoring sequence identifier with the shortest monitoring patrol distance.

[0069] This embodiment is based on the division of mooring bollard monitoring circles to achieve relatively synchronous safety monitoring of several mooring bollards of the same type and with similar historical operation and maintenance times, so as to improve the convenience of safety monitoring and reduce the complexity of safety monitoring.

[0070] The sample data acquisition module 400 is used to obtain K sample defect data sets according to the K bollard monitoring circles, wherein each sample defect data set includes a sample apparent defect data set and a sample flaw detection data set;

[0071] Specifically, in this embodiment, the function of the sample data collection module is to collect data on the apparent defects and flaw detection (internal) defects of historical bollards.

[0072] Specifically, the sample data acquisition module performs data call on apparent corrosion and scratch defects and internal deformation defects of actual mooring posts of various standard mooring posts in port applications according to the correspondence between the K mooring post monitoring circles and the types of various standard mooring posts, so as to obtain K sample defect data sets, wherein each sample defect data set includes a sample apparent defect data set and a sample flaw detection data set.

[0073] A model building training module 500 is used to build a model based on a convolutional neural network and use the K sample defect data sets as model training data to generate K bollard safety monitors;

[0074] In one embodiment, the model building training module 500 further includes:

[0075] an identification branch positioning module, for each of the K bollard safety monitors, including an apparent defect identification branch and a flaw detection defect identification branch;

[0076] A first sample calling module is used to obtain K sample apparent defect data sets based on the K sample defect data sets, wherein the sample apparent defect image in each sample apparent defect data set has a defect size identifier;

[0077] A model training control module, used to construct a standard apparent defect identifier based on a convolutional neural network, and use the K sample apparent defect data sets to train the identifier to obtain K target apparent defect identifiers;

[0078] a first branch optimization module, for synchronizing the K target apparent defect identifiers to the K apparent defect identification branches of the K bollard safety monitors;

[0079] A second sample calling module is used to obtain K sample flaw detection data sets based on the K sample flaw data sets, wherein each sample flaw detection data set includes a plurality of sample flaw detection features, a plurality of sample echo features and a plurality of flaw detection size parameters;

[0080] A sample training control module is used to construct a standard flaw detection defect identifier based on a convolutional neural network, and use the K sample flaw detection data sets to respectively train the standard flaw detection defect identifier to obtain K target flaw detection defect identifiers;

[0081] The second branch optimization module is used to synchronize the K target flaw detection defect identifiers to the K flaw detection defect identification branches of the K bollard safety monitors.

[0082] In one embodiment, the model training control module further includes:

[0083] A standard model optimization module, used to use the K sample apparent defect data sets to perform model training on the standard apparent defect identifier to obtain K optimized apparent defect identifiers;

[0084] A parameter calling execution module, used for calling the parameters of the K optimized apparent defect identifiers to obtain K identifier parameter sets;

[0085] An aggregation algorithm building module, used for pre-building an identifier parameter aggregation algorithm, and performing parameter aggregation on the K identifier parameter sets based on the identifier parameter aggregation algorithm to obtain a target identifier parameter set;

[0086] A parameter adjustment execution module, used for performing parameter adjustment on the standard apparent defect identifier based on the target identifier parameter set to obtain a target apparent defect identifier;

[0087] The model replication execution module is used to perform K replications on the target apparent defect identifier and synchronize the K target apparent defect identifiers obtained by replication to the K apparent defect recognition branches.

[0088] Specifically, in this embodiment, the model building training module is used to develop a bollard safety monitor that can detect surface defects and internal deformation of the bollard to determine whether the bollard has a structural safety risk.

[0089] The model building and training module also includes functional modules such as an identification branch positioning module and a first sample calling module. The specific functional module calling process of the model building and training module to build and generate a bollard safety monitor is as follows:

[0090] The identification branch positioning module is used to construct an identification branch. Finally, each of the K bollard safety monitors includes a pre-constructed apparent defect identification branch and a flaw detection defect identification branch. The two branches respectively identify and judge the defect conditions in two dimensions: apparent defects and internal defects.

[0091] The first sample calling module obtains K sample apparent defect data sets based on the K sample defect data sets, and the sample apparent defect image in each sample apparent defect data set has a defect size mark, and the defect size mark specifically includes the defect area and defect depth (corrosion depth).

[0092] The model training control module is used to build a model and to control the training of the model to ensure that the model output accuracy is higher than a preset threshold (98%). The model training control module includes multiple functional modules such as a standard model optimization module and a parameter call execution module.

[0093] The standard model optimization module in the model training control module specifically controls the model training process. Specifically, the standard model optimization module uses the K sample apparent defect data sets to perform model training of the standard apparent defect identifier using a conventional convolutional neural network training method to obtain K optimized apparent defect identifiers. It should be understood that the input of each optimized apparent defect identifier is the collected bollard apparent image, and the output result is the defect area and defect depth of the identified defect in the bollard apparent image.

[0094] The parameter call execution module is used to call the model parameters of the K optimized apparent defect identifiers to obtain K identifier parameter sets. The aggregation algorithm construction module performs parameter aggregation on the K identifier parameter sets based on the identifier parameter aggregation algorithm (for example, the mean calculation of the same model indicator parameters) to obtain the target identifier parameter set.

[0095] The parameter adjustment execution module performs model parameter adjustment processing on the standard apparent defect identifier based on the target identifier parameter set to obtain a target apparent defect identifier.

[0096] The model replication execution module is used to perform K replications on the target apparent defect identifier, and synchronize the K target apparent defect identifiers obtained by replication to the K apparent defect recognition branches. This embodiment performs model distributed training based on multi-sample data of the standard model and comprehensive model parameter adjustment of the training results, thereby improving the accuracy of defect recognition by the apparent defect branch.

[0097] The first branch optimization module synchronizes the K target apparent defect identifiers by the model replication execution module to the K apparent defect identification branches of the K bollard safety monitors.

[0098] The second sample calling module has the same function as the first sample calling module, and is used to obtain K sample flaw detection data sets from the K sample defect data sets. Each sample flaw detection data set includes multiple sample flaw detection features, multiple sample echo features and multiple flaw detection size parameters. It should be understood that the sample flaw detection features are flaw detection types, including but not limited to protrusions, depressions, and deformations, and the echo features are ultrasonic echo data of flaws of different sizes and flaw detection types.

[0099] The sample training control module constructs a standard flaw detection defect identifier based on a convolutional neural network, and uses the K sample flaw detection data sets to train the standard flaw detection defect identifier respectively using a conventional convolutional neural network training method to obtain K target flaw detection defect identifiers. It should be understood that the input data of each target flaw detection defect identifier is the flaw detection echo feature, and the output result is the flaw detection type and flaw detection size information.

[0100] The second branch optimization module synchronizes the K target flaw detection defect identifiers to the K flaw detection defect identification branches of the K bollard safety monitors.

[0101] This embodiment constructs and trains K bollard safety monitors including multiple defect recognition branches, thereby realizing rapid identification and marking of the bollard surface defects and internal defects after obtaining the bollard surface image and bollard ultrasonic echo data.

[0102] The data collection execution module 600 is used to pre-configure K integrated monitoring devices for the K mooring bollard monitoring circles, and schedule the K integrated monitoring devices to perform data collection on the multiple mooring bollards according to the K mooring bollard monitoring sequence identifiers of the K mooring bollard monitoring circles, so as to obtain K safety monitoring analysis arrays;

[0103] In one embodiment, the data acquisition execution module 600 further includes:

[0104] An information division execution module, configured to divide the plurality of bollard operation and maintenance information according to the K bollard monitoring circles to obtain K groups of bollard operation and maintenance information;

[0105] A monitoring cycle alignment module, used to align the monitoring cycles according to the K groups of bollard operation and maintenance information to obtain K safety monitoring cycles;

[0106] The data acquisition execution module is used to schedule the K integrated monitoring devices to perform data acquisition on the multiple mooring bollards based on the K mooring bollard monitoring sequence identifiers and to obtain the K safety monitoring analysis arrays.

[0107] Specifically, in this embodiment, the K mooring bollard monitoring circles are configured with K integrated monitoring devices for collecting the surface image of the mooring bollard and performing ultrasonic flaw detection of the mooring bollard, and each integrated monitoring device is a movable device that can collect the surface image of the mooring bollard and perform ultrasonic flaw detection of the mooring bollard when it arrives near the mooring bollard.

[0108] The data acquisition execution module controls K integrated monitoring devices to perform defect monitoring on the K mooring post monitoring circles at stable intervals to obtain an analysis of whether the mooring post has structural safety defects as follows:

[0109] The data acquisition execution module divides the multiple mooring bollard operation and maintenance information according to the K mooring bollard monitoring circles based on the information division execution module in the module to obtain K groups of mooring bollard operation and maintenance information.

[0110] The data acquisition execution module performs monitoring cycle alignment based on the monitoring cycle alignment module in the module according to the K groups of bollard operation and maintenance information to obtain K safety monitoring cycles, and the safety monitoring cycle is the time span for controlling the integrated monitoring equipment to perform defect monitoring on the K bollard monitoring circles at stable intervals.

[0111] The data acquisition execution module is based on the data acquisition execution module in the module, and is used to schedule the K integrated monitoring devices to perform data acquisition on the multiple mooring posts according to the K mooring post monitoring sequence identifiers, taking the K safety monitoring cycles as constraints, and obtaining the K safety monitoring analysis arrays, each of which includes the apparent images and ultrasonic echoes of all the mooring posts in the corresponding mooring post monitoring circle. This embodiment achieves the technical effect of reducing the frequency of mooring post fault defect monitoring.

[0112] The structural safety analysis module 700 is used to synchronize the K safety monitoring analysis arrays to the K bollard safety monitors to perform structural safety analysis and obtain bollard structural risk information.

[0113] Specifically, in this embodiment, the structural safety analysis module synchronizes the K safety monitoring and analysis arrays to the K mooring post safety monitors, performs structural safety analysis based on the apparent defect recognition branch and the flaw detection defect recognition branch in each mooring post safety monitor, and obtains K groups of mooring post structural risk information. It should be understood that each mooring post structural risk information is the apparent defect condition and internal defect condition of a mooring post single body.

[0114] This embodiment achieves the technical effect of reducing the difficulty of identifying the structural safety status of port mooring bollards during use, and improving the accuracy and efficiency of identifying the structural safety status of port mooring bollards during use.

[0115] Embodiment 2

[0116] Based on the same inventive concept as the large coastal port mooring bollard structure safety monitoring device in the aforementioned embodiment, the present application also provides a large coastal port mooring bollard structure safety monitoring system, including: a processor, the processor is coupled to a memory, the memory is used to store a program, when the program is executed by the processor, the system executes the steps of the method described in Example 1.

[0117] Any of the above-mentioned devices or steps can be stored as computer instructions or programs in various unlimited types of computer memories, and the computer instructions or programs can be recognized by various unlimited types of computer processors to implement any of the above-mentioned devices or steps.

[0118] Based on the above specific embodiments of the present invention, any improvements and modifications made to the present invention by those skilled in the art without departing from the principles of the present invention shall fall within the patent protection scope of the present invention.

Claims

1. A safety monitoring device for bollard structures in large coastal ports, characterized in that: The device comprises: A design information interaction module, used for interactively obtaining target design information, and positioning and calling the mooring bollard data based on the target design information, to obtain multiple mooring bollard design information of multiple mooring bollards in the target port; An operation and maintenance information interaction module, used for interactively obtaining target operation and maintenance information, and performing positioning and calling of bollard data based on the target operation and maintenance information, to obtain operation and maintenance information of multiple bollards of the multiple bollards; a monitoring circle analysis module, configured to classify the plurality of mooring posts according to the plurality of mooring post design information and the plurality of mooring post operation and maintenance information, and obtain K mooring post monitoring circles, wherein each mooring post monitoring circle has a mooring post monitoring sequence identifier; A sample data acquisition module, used for obtaining K sample defect data sets according to the K bollard monitoring circles, wherein each sample defect data set includes a sample apparent defect data set and a sample flaw detection data set; A model building training module, used for building a model based on a convolutional neural network, and using the K sample defect data sets as model training data to generate K bollard safety monitors; a data collection execution module, configured to pre-configure K integrated monitoring devices for the K mooring bollard monitoring circles, and schedule the K integrated monitoring devices to perform data collection on the multiple mooring bollards according to the K mooring bollard monitoring sequence identifiers of the K mooring bollard monitoring circles, so as to obtain K safety monitoring analysis arrays; A structural safety analysis module, used for synchronizing the K safety monitoring analysis arrays to the K bollard safety monitors to perform structural safety analysis and obtain bollard structural risk information; The plurality of mooring posts are classified according to the plurality of mooring post design information and the plurality of mooring post operation and maintenance information to obtain K mooring post monitoring circles, wherein each mooring post monitoring circle has a mooring post monitoring sequence identifier, including: An analysis network construction module is used to pre-construct a structural design analysis network, wherein the structural design analysis network includes a standard data call layer, a similarity constraint judgment layer and a structural feature analysis layer, and the standard data call layer, the similarity constraint judgment layer and the structural feature analysis layer are cascaded; a primary classification execution module, configured to use the design information of the plurality of mooring posts as a classification reference, perform primary classification processing of the plurality of mooring posts based on the structural design analysis network, and obtain a plurality of mooring post classification sets; a secondary classification execution module, for pre-building an operation and maintenance synchronization threshold, and performing secondary classification processing on the multiple mooring bollard classification sets according to the operation and maintenance synchronization threshold and the multiple mooring bollard operation and maintenance information, to obtain the K mooring bollard monitoring circles; Pre-constructing a structural design and analysis network, wherein the structural design and analysis network includes a standard data call layer, a similarity constraint judgment layer and a structural feature analysis layer, and the standard data call layer, the similarity constraint judgment layer and the structural feature analysis layer are cascaded, including: A sample data scheduling module, for interactively obtaining a plurality of standard structural parameter sets of a plurality of standard bollards, and performing data division on the plurality of standard structural parameter sets based on preset structural parameter types, to obtain a plurality of sample geometric parameter sets, a plurality of sample material parameter sets and a plurality of sample coating parameter sets, wherein the preset structural parameter types include geometric parameter types, material parameter types and coating types; A data association storage module, used for associating and storing the plurality of standard bollards, the plurality of sample geometric parameter sets, the plurality of sample material parameter sets and the plurality of sample coating parameter sets based on the knowledge graph, and generating the standard data call layer; An engine scheduling execution module, used for interactively obtaining a substrate feature calling engine, and synchronizing the substrate feature calling engine to the similarity constraint judgment layer; A function construction execution module, used to pre-construct a structural similarity function, and synchronize the structural similarity function to the structural feature analysis layer to complete the construction of the structural design analysis network; The function construction execution module also includes: A calculation module, the calculation module is used to call the structural similarity function; The structural similarity function is as follows: Where S1 is the geometric similarity coefficient, A is the geometric parameter set of the bollard in the bollard design information, A=a1,a2,...,a i , B is the sample geometric parameter set, B = b1, b2, ..., b i , a j and b j is the parameter value of the jth indicator in the two parameter sets A and B; Where S2 is the material similarity coefficient, C is the bollard material parameter set in the bollard design information, C = c1, c2, ..., c m , D is the sample material parameter set, D=d1,d2,...,d m , c n and d n is the parameter value of the nth indicator in the two parameter sets C and D; Where S3 is the coating similarity coefficient, E is the bollard coating parameter set in the bollard design information, and E = e1, e2, ..., e v , F is the sample coating parameter set, F = f1, f2, ..., f v , e p and f p is the parameter value of the pth indicator in the two parameter sets E and F; S = w1S1+w2S2+w3S3; Among them, S is the structural similarity, w1 is the geometric weight value, w2 is the material weight value, and w3 is the coating weight value.

2. The device according to claim 1, characterized in that The model is constructed based on a convolutional neural network, and the K sample defect data sets are used as model training data to generate K bollard safety monitors. The device also includes: an identification branch positioning module, for each of the K bollard safety monitors, including an apparent defect identification branch and a flaw detection defect identification branch; A first sample calling module is used to obtain K sample apparent defect data sets by calling the K sample defect data sets, wherein the sample apparent defect image in each sample apparent defect data set has a defect size identifier; A model training control module, used to construct a standard apparent defect identifier based on a convolutional neural network, and use the K sample apparent defect data sets to train the identifier to obtain K target apparent defect identifiers; a first branch optimization module, for synchronizing the K target apparent defect identifiers to the K apparent defect identification branches of the K bollard safety monitors; A second sample calling module is used to obtain K sample flaw detection data sets based on the K sample flaw data sets, wherein each sample flaw detection data set includes a plurality of sample flaw detection features, a plurality of sample echo features and a plurality of flaw detection size parameters; A sample training control module is used to construct a standard flaw detection defect identifier based on a convolutional neural network, and use the K sample flaw detection data sets to respectively train the standard flaw detection defect identifier to obtain K target flaw detection defect identifiers; The second branch optimization module is used to synchronize the K target flaw detection defect identifiers to the K flaw detection defect identification branches of the K bollard safety monitors.

3. The device according to claim 2, characterized in that A standard apparent defect identifier is constructed based on a convolutional neural network, and the K sample apparent defect data sets are used to train the identifier to obtain K target apparent defect identifiers, including: A standard model optimization module, used to use the K sample apparent defect data sets to perform model training on the standard apparent defect identifier to obtain K optimized apparent defect identifiers; A parameter calling execution module, used for calling the parameters of the K optimized apparent defect identifiers to obtain K identifier parameter sets; An aggregation algorithm building module, used for pre-building an identifier parameter aggregation algorithm, and performing parameter aggregation on the K identifier parameter sets based on the identifier parameter aggregation algorithm to obtain a target identifier parameter set; A parameter adjustment execution module, used for performing parameter adjustment on the standard apparent defect identifier based on the target identifier parameter set to obtain a target apparent defect identifier; The model replication execution module is used to perform K replications on the target apparent defect identifier and synchronize the K target apparent defect identifiers obtained by replication to the K apparent defect recognition branches.

4. The device according to claim 1, characterized in that The K safety monitoring and analysis arrays are synchronized to the K bollard safety monitors to perform structural safety analysis to obtain bollard structural risk information, wherein the device also includes: An information division execution module, configured to divide the plurality of bollard operation and maintenance information according to the K bollard monitoring circles to obtain K groups of bollard operation and maintenance information; A monitoring cycle alignment module, used to align the monitoring cycles according to the K groups of bollard operation and maintenance information to obtain K safety monitoring cycles; The data acquisition execution module is used to schedule K integrated monitoring devices to perform data acquisition on the multiple mooring bollards based on the K safety monitoring cycles as constraints and according to the K mooring bollard monitoring sequence identifiers to obtain K safety monitoring analysis arrays.

5. Safety monitoring system for bollard structures in large coastal ports, characterized by: include: A processor, the processor is coupled to a memory, the memory is used to store a program, when the program is executed by the processor, the system is enabled to execute the steps of the device according to any one of claims 1 to 4.

Images