Internet of Things distributed monitoring data synchronization method based on port machinery equipment state

Abstract

The invention discloses an Internet of Things distributed monitoring data synchronization method based on port machinery equipment state, which comprises the following steps: taking a fault sign as a node, and if the co-occurrence frequency between two nodes is greater than a preset value, constructing an undirected edge; constructing a fault group according to each connected component; acquiring image data through a plurality of information acquisition terminals, and analyzing whether a fault sign exists or not; obtaining residual fault signs in the corresponding fault group, generating a to-be-detected task, and broadcasting the to-be-detected task in the information acquisition terminal group; the information acquisition terminal calculates the spatial distance between the information acquisition terminal and each to-be-detected task, and the information acquisition terminal corresponding to the minimum value is used as a task execution terminal to confirm residual fault signs; summarizing residual fault sign confirmation conditions, generating a fault confidence coefficient, and forming a recheck message; according to the method, isolated alarm is converted into group early warning, the fault advanced sensing capability is remarkably improved, and the calculation amount of a data center is reduced through co-occurrence relation mining and group division of fault signs.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to an Internet of Things distributed monitoring data synchronization method based on port machinery equipment status. Background Art

[0002] In the power supply management of the port area, the stable supply of electricity is the basis for the normal operation of various businesses in the port area. The substations in the port area are characterized by many points, long lines and wide areas, which require daily inspections by operation and maintenance personnel and real-time monitoring of important equipment. However, due to the need for manual inspections by maintenance personnel, it is limited by manpower and difficult to achieve uninterrupted inspections around the clock. At present, with the popularization of unmanned technology, the Chinese invention patent with publication number CN120185207A discloses an integrated monitoring system suitable for port power supply management. Through the information management module, power dispatching and centralized control module, inspection module and information management module, it obtains information such as substation equipment operation, robots, and maps, inspection information, and uses AI recognition, perception analysis and Internet of Things technology to realize intelligent inspection functions such as intelligent linkage of fault signals of main and auxiliary equipment, remote inspection perception, and intelligent analysis of equipment status, effectively improving the work efficiency of operation and maintenance personnel, reducing labor intensity and improving safety assurance.

[0003] However, the aforementioned patents, or existing port machinery equipment condition monitoring systems, generally adopt a "centralized aggregation-single-machine identification" architecture: each information collection terminal uploads collected images and sensor data to a cloud or local server, which then identifies, records, and issues alarms for fault signs. This architecture, when faced with high-concurrency, widely distributed port operations, exposes the following core technical flaws: 1. Static fault correlation: The cloud or local server at the central node only performs threshold judgments on single-point fault signs. It lacks dynamic modeling of the co-occurrence relationship of multiple fault signs in time and space, making it impossible to detect potential cascading failures in advance. For example, when a port machinery is processing a batch of cargo, all the equipment is under high pressure and may fail at the same time. At the same time, the failure of the previous equipment causes the subsequent equipment to be subjected to stress exceeding the original load, resulting in a cascading failure. 2. Lack of task coordination: When a terminal detects a fault in a port machinery or a port machinery, only that terminal can process the generated pending inspection tasks in sequence. Other terminals cannot receive and assist the terminal in completing pending inspection tasks in real time based on spatial distance and resource status. When faced with inspection tasks requiring rapid confirmation of fault signs, the inspection method of having a single terminal process pending inspection tasks in sequence is inefficient. Therefore, there is an urgent need for a technical solution that can complete fault sign association mining, distributed task collaboration and local data synchronization on the terminal side. Summary of the Invention

[0004] The technical problem solved by the present application is that the prior art is difficult to complete real-time fault sign spatiotemporal co-occurrence relation mining, distributed task cooperation and local data synchronization on the terminal side in the high-concurrency and wide-area distributed operation scene of the port, leading to a bandwidth and computing resource bottleneck of the cloud or local server, and a large safety response delay due to low inspection efficiency.

[0005] To solve the above technical problems, the present application provides the following technical solutions: The Internet of Things distributed monitoring data synchronization method based on the state of port machinery equipment comprises the following steps: S100, record the co-occurrence times of two different fault signs; if the co-occurrence times between two nodes are greater than a preset value, a non-directed edge is constructed; according to each connected component, a fault group is constructed; S200, collect image data of each port machinery equipment through multiple information collection terminals, and analyze whether there is a fault sign; S300, when the information collection terminal detects a fault sign, obtain the remaining fault signs in the corresponding fault group, generate a to-be-detected task, and integrate and generate a task package for broadcasting in the information collection terminal group; S400, after the information collection terminal receives the task package, calculate the spatial distance between itself and each to-be-detected task, take the information collection terminal corresponding to the minimum value as a task execution terminal, and confirm the remaining fault signs; S500, summarize the fault sign confirmation conditions in the corresponding fault group, generate a fault confidence, and form a review message.

[0006] As a preferred scheme of the Internet of Things distributed monitoring data synchronization method based on the state of port machinery equipment, the S1 comprises the following sub-steps: S110, call a historical fault database to obtain the record of the fault sign, and the record field of the fault sign at least contains a fault sign description, a positioning number and a timestamp; the positioning number is composed of an equipment number and a component number; the fault sign description includes structural cracks, overheating and abnormal vibration; S120, based on a preset fault dictionary, unify fault signs with synonymous descriptions into the same standard name; S130, set a first period sliding window, if two fault signs appear successively and the time interval is within the first period sliding window, record it as one co-occurrence; S140, adopt a hash table structure, take a pair of fault sign groups as a key, and accumulate the co-occurrence times; S150, set a node for the fault sign, and traverse the hash table, when the co-occurrence times corresponding to any key are greater than a threshold value, add a non-directed edge between the corresponding two nodes; S160 uses a depth-first search algorithm, starting from any unvisited node, traversing along the undirected edges until it can no longer move forward, marking all traversed nodes as the same fault group, and repeating until all nodes are visited, ultimately obtaining several interrelated fault groups.

[0007] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, S2 includes the following sub-steps: S210, acquiring image data of key parts of the port machinery equipment using a camera device and an infrared thermal imaging device carried by the information collection terminal; the image data includes a real-scene image and an infrared image; S220, preprocessing the image data; wherein the preprocessing includes grayscale conversion, noise reduction, and contrast enhancement; S230, extracting contour information from the real scene image, calculating the ratio of the contour length to the area, and marking it as a suspected crack if the ratio is greater than a preset threshold; S240, identifying the temperature of each area in the infrared image, searching for areas where the temperature is higher than a preset range of the ambient average, and marking the area as suspected overheating if the area is larger than the preset area value; S250: Output the suspected crack and suspected overheating in the form of a structured message, where the message fields include a location number, a description of the fault sign, and a timestamp.

[0008] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, in S210, before the task execution terminal captures the image data, the following steps are also included: S211, while on the way to the task coordinates, the task execution terminal collects the overall image of the port machinery equipment in real time, extracts the outline of the port machinery equipment through grayscale and edge detection, and calculates the centroid of the outline; S212, establishing a virtual horizontal line and a virtual vertical line with the contour centroid as the origin, detecting the angle between the key structure edge and the virtual horizontal line, and detecting the angle between the key structure edge and the virtual vertical line, respectively, and generating a deviation vector when the angle exceeds a preset threshold; wherein the horizontal component of the deviation vector is represented by the horizontal offset between the key structure edge in the image and the virtual vertical line, and the vertical component thereof is represented by the vertical offset between the key structure edge in the image and the virtual horizontal line; S213, based on the position change trajectory of the contour centroid in the continuous multi-frame image, a sliding average-based algorithm is used to extrapolate the predicted centroid position at the next moment, and a circular search zone is delineated in the ground coordinate system with the predicted position as the center; S214: The task execution terminal drives along the circular search zone to the candidate shooting points in sequence. At each point, the vehicle-mounted gimbal camera performs iterative fine-tuning of the vertical pitch and horizontal rotation until the vertical pixel distance and horizontal offset fall within a preset range. The current vehicle coordinates and gimbal attitude angle are recorded as the optimal shooting position. S215 , while maintaining the optimal shooting posture, the task execution terminal executes S210 to S250 to complete the fault sign detection, and writes the image data together with the optimal shooting posture information into the review message.

[0009] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, S210 further includes the following steps: a. When the displacement of the center of mass of the port machinery equipment outline compared with the previous best shooting pose exceeds the set threshold, the task execution terminal immediately triggers a new round of S211 to S215 to form a closed-loop tracking to keep the shooting angle always aligned with the key structure; b. If the confidence level of the mark in the review message is high, the task execution terminal shall perform millimeter-level stepping movement perpendicular to the direction of the crack while maintaining the optimal shooting posture, continuously collect no less than three local images, and attach the local images to the review message.

[0010] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, S3 includes the following sub-steps: S310, taking the currently detected fault sign as a starting point, searching the fault group list for a fault group corresponding to the current fault sign to form a marked fault group; S320, listing the remaining fault signs in the marked fault group to form a list of tasks to be detected; S330: Encapsulate the task list to be detected, fault group identifier, task priority, and timestamp into a task package, and send the package to the information collection terminal group using UDP broadcast from the current information collection terminal. The current information collection terminal is recorded as the broadcast initiator.

[0011] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, S310 further includes the following steps: S311, if the currently monitored fault sign does not retrieve the corresponding fault group in the fault group list, the image data corresponding to the fault sign is sent to the management cloud to confirm the fault sign description, enter the corresponding fault sign into the historical fault database, and record the fault sign description, location number and timestamp corresponding to the fault sign.

[0012] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, the step S4 includes the following sub-steps: S410, each information collection terminal obtains its own real-time coordinates through a built-in positioning module, parses the positioning number corresponding to the task to be detected in the task package, and retrieves the task coordinates corresponding to the task to be detected from a locally pre-stored coordinate search table; S420, each information collection terminal calculates the straight-line distance between its own coordinates and the task coordinates using the Euclidean distance algorithm; S430: Each information collection terminal transmits the calculation result back to the broadcast initiator. The broadcast initiator compares all the returned distances and selects the information collection terminal with the minimum distance to each task to be detected as the task execution terminal. S440, after receiving the message, each task execution terminal immediately goes to the task coordinates to detect signs of faults.

[0013] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to the present invention, S5 includes the following sub-steps: S510, after the task execution terminal arrives at the task coordinates, S201 to S205 are repeated to detect fault signs on the image data captured corresponding to the task coordinates; S520, if there is a fault sign, the corresponding detected fault sign confirmation result is transmitted back to the broadcast initiator in the form of a structured message; S530: The broadcast initiator collects statistics on the confirmation results of all fault signs in the same fault group and registers and confirms the confidence level. If the same sign is confirmed by at least two independent terminals, the confidence level is marked as high; if it is confirmed by only one terminal, the confidence level is marked as medium; if it is not confirmed, the confidence level is marked as low. S540 , the fault group identifier, all fault sign confirmation results, confidence level, on-site image, and timestamp are integrated into a review message and transmitted to the remote monitoring center.

[0014] As a preferred solution of the method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment described in the present invention, the method further includes the following steps: S600 obtains the operating parameters of the port machinery equipment within the current first cycle sliding window through the cloud, compares the operating parameters when multiple sets of fault signs appear with the operating parameters when the fault signs disappear, and analyzes and generates the threshold range of the operating parameters when the port machinery equipment is in safe operation.

[0015] Beneficial effects of the present invention: The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment provided by the present invention realizes the instant construction of fault groups on the terminal side and the dynamic allocation of optimal detection resources by integrating the co-occurrence relationship mining of fault signs, distributed task broadcasting and the minimum distance allocation mechanism model, thereby significantly improving the timeliness of fault chain warning and the system response sensitivity, playing a role in reducing cloud load, reducing redundant data transmission, and suppressing missed detections and false alarms; at the same time, with the help of closed-loop adaptive shooting posture control and operating parameter coupling analysis, the image acquisition accuracy and the reliability of crack and overheating identification are further improved, playing a role in ensuring image detection accuracy in high concurrency and wide-area distribution environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flowchart of the steps of the Internet of Things distributed monitoring data synchronization method based on the port machinery equipment status provided by the present invention. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0018] In order to solve the problem that existing technologies are difficult to complete the spatiotemporal co-occurrence relationship mining of fault signs, distributed task collaboration and local data synchronization in real time on the terminal side in high-concurrency and wide-area distribution operation scenarios in ports, resulting in bottlenecks in cloud or local server bandwidth and computing resources, it is easy to cause large delays in security response due to the low efficiency of inspection methods.

[0019] In this embodiment, the distributed monitoring data synchronization method for port machinery in the Internet of Things operates in a mixed operation scenario consisting of bridge cranes, rail cranes, and tire cranes. Thirty mobile information collection terminals are deployed on-site. Each terminal can be configured as an intelligent inspection vehicle or intelligent inspection drone, or even a fixed-point collection terminal for preliminary data collection. Each terminal integrates a visible light camera, an infrared thermal imager, a positioning module, a 5G dual-mode communication module, and an on-board three-axis gimbal. The terminals are operated by an on-board industrial computer and deploy all the algorithms described in this invention. Optionally, an acoustic wave detector can be added as needed.

[0020] Reference Figure 1 , this application provides an IoT distributed monitoring data synchronization method based on port machinery equipment status, comprising the following steps: S100, record the co-occurrence times of two different failure signs; if the co-occurrence times between two nodes are greater than a preset value (specifically, 0 times can be calculated, or the number of co-occurrences when the actual failure probability is calculated according to the port operation log data), an undirected edge is constructed; according to each connected component, a failure group is constructed.

[0021] S110, in the local memory of each information collection terminal, call the historical failure database to obtain the record of the failure sign, and the record field of the failure sign at least includes the failure sign description, the positioning number and the time stamp; the positioning number is composed of the equipment number and the component number, and the component number can be additionally provided with a region serial number, for example, "QC01-BOOM03" points to the third partition of No. 1 bridge crane boom; the failure sign description includes structural cracks, overheating and abnormal vibration.

[0022] S120, the terminal calls the database to extract the failure sign description, and based on a preset failure dictionary, the dictionary is embedded in the form of a YAML file, and words such as "crack", "crack" or "crack" are uniformly mapped to the standard term "structural crack", and words such as "temperature rise", "overheating" or "hot spot" are uniformly mapped to "overheating", and the dictionary is updated by cloud operation personnel every quarter. The failure signs with synonymous descriptions are uniformly mapped to the same standard name. The standard name is limited to "structural crack, overheating" or "abnormal vibration" can also be added.

[0023] S130, set a first period sliding window (which can be set to 180 seconds), if two kinds of failure signs appear in succession and the time interval is within the first period sliding window, it is recorded as one co-occurrence. Among them, the window sliding step is 30 seconds, which ensures real-time performance.

[0024] S140, adopt a hash table structure, and accumulate the co-occurrence times of the paired failure sign groups; specifically, use a linked hash table to store the co-occurrence count, the key is an ordered tuple (f1, f2) to ensure undirectedness, and the hash load factor is maintained below 0.7 to avoid conflicts.

[0025] S150, set the node for the failure sign, and traverse the hash table, when the co-occurrence times corresponding to any key are greater than a threshold value (specifically, 0 times can be calculated, or the number of co-occurrences when the actual failure probability is calculated according to the port operation log data), an undirected edge is added between the corresponding two nodes; wherein the graph structure adopts an adjacency list form and resides in memory, and the node object additionally saves the latest time stamp for subsequent incremental update.

[0026] S160 uses a depth-first search algorithm, starting from any unvisited node, traversing along the undirected edges until it can no longer move forward, marking all traversed nodes as the same fault group, and repeating until all nodes are visited, ultimately obtaining several interrelated fault groups.

[0027] Through these steps, the terminal can locally detect which fault signs tend to occur together, thereby converting isolated alarms into group-level warnings, significantly improving the ability to proactively detect cascading faults. In other words, the terminal performs local co-occurrence relationship mining and grouping of fault signs, eliminating the need to upload raw data to the data center, reducing data center computing workload and network load. Furthermore, fault grouping provides a precise task list for subsequent distributed task allocation, reducing the probability of missed detections.

[0028] S200, through multiple information collection terminals, collects image data of each port machinery equipment and analyzes whether there are signs of failure.

[0029] S210 uses a camera device (which can use a visible light camera, 2 / 3-inch CMOS, and a focal length of 25mm) and an infrared thermal imaging acquisition device (which can use a 640×512 vanadium oxide non-cooled type) on the information collection terminal. The two are exposed synchronously to ensure that the real-scene image and the infrared image are spatially aligned pixel by pixel, thereby obtaining image data of key parts of the port machinery equipment; the image data includes real-scene images and infrared images.

[0030] In S210, before the task execution terminal captures the image data, the following steps are also included: S211, the task execution terminal collects the overall image of the port machinery equipment in real time on the way to the task coordinates, extracts the outline of the port machinery equipment through grayscale and edge detection, and calculates the outline centroid.

[0031] S212, establishing a virtual horizontal line and a virtual vertical line with the contour centroid as the origin, detecting the angle between the edge of the key structure and the virtual horizontal line, and detecting the angle between the edge of the key structure and the virtual vertical line, respectively, and generating a deviation vector when the angle exceeds a preset threshold (based on the maximum allowable offset angle of the edge of the structural component under port wind conditions of level 5 / 6 / 7, for example, selected within the range of 5 to 10 degrees); wherein the horizontal component of the deviation vector is represented by the degree of horizontal offset between the edge of the key structure in the image and the virtual vertical line, and the vertical component thereof is represented by the degree of vertical offset between the edge of the key structure in the image and the virtual horizontal line.

[0032] S213, based on the position change trajectory of the contour centroid in the continuous multi-frame image, a sliding average basic algorithm is used to extrapolate the predicted centroid position at the next moment, and a circular search zone is drawn in the ground coordinate system with the predicted position as the center.

[0033] S214, the task execution terminal drives along the annular search belt to the candidate shooting points in turn, and performs iterative fine adjustment of vertical pitching and horizontal rotation of the vehicle-mounted gimbal camera at each point until the vertical pixel distance and the horizontal offset fall within the preset interval (±5 pixels and ±10 cm can be selected to meet the crack detection accuracy requirement, and under the 1080p resolution of the camera, ±5 pixels corresponds to actual 2 cm, which meets the crack detection accuracy requirement), and records the current vehicle body coordinates and the gimbal attitude angle as the best shooting pose.

[0034] S215, in the state of maintaining the best shooting pose, the task execution terminal performs S210 to S250 to complete the fault sign detection, and writes the image data and the best shooting pose information into the review message.

[0035] Specifically, through closed-loop pose optimization, the terminal always locks the key structure in the center area of the picture, improves the spatial resolution of crack and overheating detection, and reduces the misjudgment caused by the angle deviation.

[0036] In S210, the following steps are further included: a. When the displacement of the port machinery equipment profile centroid compared with the last best shooting pose port machinery equipment profile centroid exceeds the set threshold (30-50 cm can be specifically used), the task execution terminal immediately triggers a new round of S211 to S215 to form a closed-loop tracking and keep the shooting angle always aligned with the key structure.

[0037] b. If the confidence level in the review message is high, the task execution terminal performs millimeter-level step-by-step movement in the direction perpendicular to the crack direction while maintaining the best shooting pose, continuously collects not less than three local images, and appends the local images to the review message.

[0038] S220, the image data is preprocessed; wherein the preprocessing includes grayscale, noise reduction and contrast enhancement. Specifically, first, weighted grayscale is performed to retain the red channel crack contrast, then 3*3 Gaussian kernel noise reduction is used, and finally CLAHE is used to limit the contrast adaptive histogram equalization to enhance the texture.

[0039] S230, for structural cracks, the contour information in the real scene image is extracted, the contour length and area ratio is counted, and if the ratio is greater than the preset threshold (based on the ROC analysis of the public data set of port crane steel plate cracks (60,000 images), the F1-score maximum point, and 0.25-0.35 mm-1 can be specifically selected), it is marked as a suspected crack. Specifically, the calculation steps can first obtain the binary contour by calling Canny edge detection, then simplify the polyline through Douglas-Peucker algorithm, and calculate the ratio of the length of each simplified polyline to the corresponding convex hull area.

[0040] S240 , identifying the temperature of each area in the infrared image, searching for areas where the temperature is higher than a preset range of the ambient average (specifically, higher than 8 to 15 degrees Celsius). If the area of ​​the area is larger than a preset area value (specifically, higher than 20×20 to 40×40 pixels), it is marked as suspected overheating.

[0041] S250 outputs suspected cracks and suspected overheating in a structured message format (JSON message). The message fields include a location number, a description of the fault signs, and a timestamp.

[0042] Specifically, the terminal designed in this application independently completes image acquisition and preliminary recognition, outputs standardized messages, and provides structured input for subsequent fault group task package generation, ensuring data format consistency and real-time performance.

[0043] S300, when the information collection terminal detects signs of a fault, it obtains the remaining fault signs in the corresponding fault group, generates tasks to be detected, integrates and generates task packages, and broadcasts them within the information collection terminal group. The task package is serialized, and the fields include task list, group ID, priority, broadcast timestamp, and broadcast terminal ID in sequence. The terminal then broadcasts on the local subnet via the UDP multicast address, with the TTL set to 2 to ensure that only adjacent work areas are covered. The broadcast is retransmitted three times at an interval of 300ms to improve the arrival rate. This design allows task information to be received by all reachable terminals within milliseconds, providing a synchronized clock for the nearest selection of S400.

[0044] S310 , starting from the currently detected fault sign, searching the fault group list for a fault group corresponding to the current fault sign to form a marked fault group.

[0045] S311, if the currently monitored fault sign does not retrieve the corresponding fault group in the fault group list, the image data corresponding to the fault sign is sent to the management cloud to confirm the fault sign description, enter the corresponding fault sign into the historical fault database, and record the fault sign description, location number and timestamp corresponding to the fault sign.

[0046] S320: List the remaining fault signs in the marked fault group to form a task list to be detected.

[0047] S330: Encapsulate the task list to be detected, fault group identifier, task priority, and timestamp into a task package, and send the package to the information collection terminal group using UDP broadcast from the current information collection terminal. The current information collection terminal is recorded as the broadcast initiator.

[0048] Specifically, the task package broadcast mechanism designed in this application realizes the instant sharing of group-level tasks among multiple terminals. Unmatched group situations are confirmed through the cloud and written back to the database, ensuring the system's ability to continuously learn new fault signs.

[0049] S400: After receiving the task packet, the information collection terminal calculates the spatial distance between itself and each task to be detected, and uses the information collection terminal corresponding to the minimum value as the task execution terminal to confirm the remaining fault signs.

[0050] In step S410, each information collection terminal obtains its own real-time coordinates through the built-in positioning module, parses the positioning number corresponding to the task to be detected in the task package, and retrieves the task coordinates corresponding to the task to be detected from the locally stored coordinate retrieval table. The coordinate retrieval table is issued daily by the cloud.

[0051] S420, each information collection terminal calculates the straight-line distance between its own coordinates and the task coordinates using the Euclidean distance algorithm.

[0052] S430, each information collection terminal transmits the calculation result back to the broadcast initiator, and the broadcast initiator compares all the returned distances and selects the information collection terminal corresponding to the minimum distance to each task to be detected as the task execution terminal.

[0053] S440, after receiving the message, each task execution terminal immediately goes to the task coordinates to detect signs of faults.

[0054] Specifically, this mechanism physically assigns the nearest terminal to each task, significantly shortening the driving path and reducing the overall detection delay. That is, this application uses a terminal assignment strategy based on minimum distance to assign tasks to the spatially nearest terminal, reducing driving time, achieving load balancing, and reducing overall detection delay.

[0055] S500: Summarize the fault sign confirmation status in the corresponding fault group, generate fault confidence, and form a review message.

[0056] S510: After the task execution terminal arrives at the task coordinates, S201 to S205 are repeated to detect fault signs on the image data captured corresponding to the task coordinates.

[0057] S520: If there are signs of fault, the corresponding detected fault sign confirmation result is transmitted back to the broadcast initiator in the form of a structured message.

[0058] S530, the broadcast initiator counts the confirmation results of all fault signs in the same fault group and performs confidence registration confirmation. If the same sign is confirmed by at least two independent terminals, the confidence is marked as high; if it is confirmed by only one terminal, the confidence is marked as medium; if it is not confirmed, the confidence is marked as low.

[0059] S540 , the fault group identifier, all fault sign confirmation results, confidence level, on-site image, and timestamp are integrated into a review message and transmitted to the remote monitoring center.

[0060] Specifically, the confidence mechanism of the multi-terminal cross-validation provided by the application effectively suppresses single-point false positives, and the review message provides traceable and quantifiable basis for operation and maintenance decisions.

[0061] S600, obtain the running parameters of the port machine equipment in the current first period sliding window through the cloud, compare the running parameters when the fault signs appear with the running parameters when the fault signs disappear, and analyze to generate a threshold range of the running parameters when the port machine equipment is in safe operation.

[0062] Specifically, the application provides a specific use scenario: terminal No. T07 detects a fault sign "structural crack" at position a of the to-be-detected port machine at time t, and finds fault group G05 according to the fault sign; T07 performs a broadcast task package, T12, T19, and T25 return distances, T12 is the smallest and is assigned as the task execution terminal; T12 performs S211-S215 to obtain the best shooting pose, completes detection, and returns the confirmation result; T07 aggregates the results and marks the confidence as high, generates a review message, and uploads it to the cloud; and the cloud completes a complete distributed monitoring data synchronization closed loop.

[0063] Those skilled in the art should understand that the embodiments of the application can be provided as a method, a system or a computer program product. Therefore, the application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the application can take the form of a computer program product implemented on one or more computer-usable storage media having computer-usable program code embodied therein. The storage media can be realized by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk. These computer program instructions can also be stored in a computer readable storage medium that can guide the computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer readable storage medium produce a manufactured product including instruction devices, which realize the processes Figure 1 one flow or multiple flows and / or blocksFigure 1 The function specified in one or more boxes.

[0064] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment, characterized in that: The steps include: S100, recording the co-occurrence counts of two different fault signs; using the fault signs as nodes, if the co-occurrence count between the two nodes is greater than a preset value, constructing an undirected edge; constructing a fault group based on each connected component; S200, through multiple information collection terminals, collects image data of each port machinery equipment and analyzes whether there are signs of failure; S300, when the information collection terminal detects a fault sign, it obtains the remaining fault signs in the corresponding fault group, generates tasks to be detected, integrates the generated task package and broadcasts it within the information collection terminal group; S400, after receiving the task packet, the information collection terminal calculates the spatial distance between itself and each task to be detected, and the information collection terminal corresponding to the minimum value is used as the task execution terminal to confirm the remaining fault signs; S500: Summarize the fault sign confirmation status in the corresponding fault group, generate fault confidence, and form a review message.

2. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 1, characterized in that: The S1 includes the following sub-steps: S110, calling a historical fault database to obtain a record of fault signs, wherein the record field of the fault sign includes at least a fault sign description, a location number, and a timestamp; the location number is a combination of the device number and the component number; the fault sign description includes structural cracks, overheating, and abnormal vibration; S120, based on a preset fault dictionary, unifies fault signs with synonymous descriptions into the same standard name; S130, setting a first period sliding window, if two fault signs appear successively and the time interval is within the first period sliding window, it is recorded as a co-occurrence; S140, using a hash table structure, using paired fault sign groups as keys, and accumulating the number of co-occurrences; S150, setting nodes for fault signs and traversing the hash table. When the number of co-occurrences corresponding to any key is greater than a threshold, adding an undirected edge between the corresponding two nodes; S160 uses a depth-first search algorithm, starting from any unvisited node, traversing along the undirected edges until it can no longer move forward, marking all traversed nodes as the same fault group, and repeating until all nodes are visited, ultimately obtaining several interrelated fault groups.

3. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 1, characterized in that: The S2 includes the following sub-steps: S210, acquiring image data of key parts of the port machinery equipment using a camera device and an infrared thermal imaging device carried by the information collection terminal; the image data includes a real-scene image and an infrared image; S220, preprocessing the image data; wherein the preprocessing includes grayscale conversion, noise reduction, and contrast enhancement; S230, extracting contour information from the real scene image, calculating the ratio of the contour length to the area, and marking it as a suspected crack if the ratio is greater than a preset threshold; S240, identifying the temperature of each area in the infrared image, searching for areas where the temperature is higher than a preset range of the ambient average, and marking the area as suspected overheating if the area is larger than the preset area value; S250: Output the suspected crack and suspected overheating in the form of a structured message, where the message fields include a location number, a description of the fault sign, and a timestamp.

4. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 3, characterized in that: In S210, before the task execution terminal captures the image data, the following steps are also included: S211, while on the way to the task coordinates, the task execution terminal collects the overall image of the port machinery equipment in real time, extracts the outline of the port machinery equipment through grayscale and edge detection, and calculates the centroid of the outline; S212, establishing a virtual horizontal line and a virtual vertical line with the contour centroid as the origin, detecting the angle between the key structure edge and the virtual horizontal line, and detecting the angle between the key structure edge and the virtual vertical line, respectively, and generating a deviation vector when the angle exceeds a preset threshold; wherein the horizontal component of the deviation vector is represented by the horizontal offset between the key structure edge in the image and the virtual vertical line, and the vertical component thereof is represented by the vertical offset between the key structure edge in the image and the virtual horizontal line; S213, based on the position change trajectory of the contour centroid in the continuous multi-frame image, a sliding average-based algorithm is used to extrapolate the predicted centroid position at the next moment, and a circular search zone is delineated in the ground coordinate system with the predicted position as the center; S214: The task execution terminal drives along the circular search zone to the candidate shooting points in sequence. At each point, the vehicle-mounted gimbal camera performs iterative fine-tuning of the vertical pitch and horizontal rotation until the vertical pixel distance and horizontal offset fall within a preset range. The current vehicle coordinates and gimbal attitude angle are recorded as the optimal shooting position. S215 , while maintaining the optimal shooting posture, the task execution terminal executes S210 to S250 to complete the fault sign detection, and writes the image data together with the optimal shooting posture information into the review message.

5. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 4, characterized in that: In S210 The following steps are involved: a. When the displacement of the center of mass of the port machinery equipment outline compared with the previous best shooting pose exceeds the set threshold, the task execution terminal immediately triggers a new round of S211 to S215 to form a closed-loop tracking to keep the shooting angle always aligned with the key structure; b. If the confidence level of the mark in the review message is high, the task execution terminal shall perform millimeter-level stepping movement perpendicular to the direction of the crack while maintaining the optimal shooting posture, continuously collect no less than three local images, and attach the local images to the review message.

6. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 1, characterized in that: The S3 includes the following sub-steps: S310, taking the currently detected fault sign as a starting point, searching the fault group list for a fault group corresponding to the current fault sign to form a marked fault group; S320, listing the remaining fault signs in the marked fault group to form a list of tasks to be detected; S330: Encapsulate the task list to be detected, fault group identifier, task priority, and timestamp into a task package, and send the package to the information collection terminal group using UDP broadcast from the current information collection terminal. The current information collection terminal is recorded as the broadcast initiator.

7. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 6, characterized in that: S310 further includes the following steps: S311, if the currently monitored fault sign does not retrieve the corresponding fault group in the fault group list, the image data corresponding to the fault sign is sent to the management cloud to confirm the fault sign description, enter the corresponding fault sign into the historical fault database, and record the fault sign description, location number and timestamp corresponding to the fault sign.

8. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 1, characterized in that: The S4 includes the following sub-steps: S410, each information collection terminal obtains its own real-time coordinates through a built-in positioning module, parses the positioning number corresponding to the task to be detected in the task package, and retrieves the task coordinates corresponding to the task to be detected from a locally pre-stored coordinate search table; S420, each information collection terminal calculates the straight-line distance between its own coordinates and the task coordinates using the Euclidean distance algorithm; S430: Each information collection terminal transmits the calculation result back to the broadcast initiator. The broadcast initiator compares all the returned distances and selects the information collection terminal with the minimum distance to each task to be detected as the task execution terminal. S440, after receiving the message, each task execution terminal immediately goes to the task coordinates to detect signs of faults.

9. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 1, characterized in that: The S5 comprises the following sub-steps: S510, after the task execution terminal arrives at the task coordinates, S201 to S205 are repeated to detect fault signs on the image data captured corresponding to the task coordinates; S520, if there is a fault sign, the corresponding detected fault sign confirmation result is transmitted back to the broadcast initiator in the form of a structured message; S530: The broadcast initiator counts the confirmation results of all fault signs in the same fault group and registers and confirms the confidence level. If the same sign is confirmed to exist by at least two independent terminals, the confidence level is marked as high. If it is confirmed by only one terminal, the confidence level is marked as medium; if it is not confirmed, the confidence level is marked as low; S540 , the fault group identifier, all fault sign confirmation results, confidence level, on-site image, and timestamp are integrated into a review message and transmitted to the remote monitoring center.

10. The method for synchronizing distributed monitoring data of the Internet of Things based on the status of port machinery equipment according to claim 1, characterized in that: The following steps are also included: S600 obtains the operating parameters of the port machinery equipment within the current first cycle sliding window through the cloud, compares the operating parameters when multiple sets of fault signs appear with the operating parameters when the fault signs disappear, and analyzes and generates the threshold range of the operating parameters when the port machinery equipment is in safe operation.

Images