External damage prevention early warning system for three-dimensional operation crane of transformer substation

Abstract

The invention relates to the technical field of power equipment monitoring and early warning, and particularly discloses a substation three-dimensional operation crane external damage prevention early warning system which comprises an environment sensing module, a dynamic early warning and collision prevention module, a data fusion and decision module, a man-machine interaction operation module and a system self-diagnosis and maintenance module. The environment sensing module is responsible for sensing the surrounding environment of the crane and collecting and analyzing surrounding obstacle information. According to the invention, a mode of fusing various sensor data is adopted, the system can accurately sense the surrounding environment and make a response in a three-dimensional space, especially in a complex and narrow environment such as a transformer substation, the sensor fusion can significantly improve the comprehensiveness and precision of environment sensing, the system can carry out danger prediction and intelligent decision making, and the system has a wide application prospect. And the system not only can make a response in real time, but also can carry out self-learning and optimization according to historical data and environmental changes, so that an optimal early warning strategy is provided under different operation conditions.

Description

Technical Field

[0001] This invention relates to the field of power equipment monitoring and early warning technology, and in particular to a three-dimensional operation crane anti-external damage early warning system for substations. Background Technology

[0002] Substations are an important component of the power system. Their main function is to convert high-voltage electrical energy into low-voltage energy for users. They use equipment such as transformers, circuit breakers, and switching equipment to complete the transformation and distribution of electrical energy. Substations are typically used at different stages of the power transmission process to ensure that electrical energy can be delivered to various areas efficiently and stably.

[0003] Because substations are typically densely packed with facilities and have limited space, with many devices located at high altitudes or underground, three-dimensional cranes can perform precise lifting within limited spaces. They are suitable for the installation, inspection, and maintenance of power equipment. They are widely used in the installation, inspection, and maintenance of power equipment. With their intelligent control system, high load-bearing capacity, and flexible operation, they improve the efficiency of substation operations while ensuring safety.

[0004] The working environment within substations is complex, with densely packed equipment. Crane operation requires precision. Currently, cranes rely solely on monitoring and radar detection during the installation, inspection, or maintenance of power equipment. This approach not only has a low fault tolerance rate but also heavily relies on the operator's technical experience. Even minor operational errors can easily cause external damage to surrounding facilities, equipment, or the environment, potentially leading to power accidents. Therefore, this invention proposes a three-dimensional anti-external damage early warning system for cranes operating in substations to address the problems mentioned in the background. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a three-dimensional operation crane early warning system for substations to prevent external damage.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A three-dimensional operation crane anti-external damage early warning system for substations includes an environmental perception module, a dynamic early warning and anti-collision module, a data fusion and decision-making module, a human-computer interaction operation module, and a system self-diagnosis and maintenance module. The environmental perception module is responsible for sensing the environment around the crane and collecting and analyzing information about surrounding obstacles; the dynamic early warning and collision avoidance module is responsible for monitoring the distance and position of the crane and surrounding obstacles in real time; the data fusion and decision-making module fuses data from various sensors to generate environmental data; the human-machine interaction module provides the operator with an interactive interface with the system; and the system self-diagnosis and maintenance module is responsible for monitoring the working status of each component of the system. The operation of the dynamic early warning and collision avoidance module includes the following steps: Based on the various environmental sensors in the environmental perception module, the surrounding environmental data is collected in real time and the collected data is transmitted to the data fusion and decision-making module. Based on the information analyzed and processed by the data fusion and decision-making module, the system generates a real-time three-dimensional environment map around the crane, marking the position, shape and distance of each obstacle. The system uses computer vision, image processing or deep learning algorithms to identify and classify obstacles, and transmits the generated information to the dynamic early warning and anti-collision module. Based on the 3D environment map and obstacle information generated by the system, the dynamic early warning and collision avoidance module performs a collision risk assessment based on the crane's current position, speed, trajectory and relative position with surrounding obstacles, determines whether there is a potential collision risk, and assigns a risk level to each potential collision event based on the assessment results, and decides on the corresponding response strategy.

[0007] Preferably, the dynamic early warning and collision avoidance module generates a real-time risk assessment curve based on the crane's current position, speed, trajectory, and relative position to surrounding obstacles, and assesses the risk level according to the following collision risk function: in This is the crane's maximum speed, to prevent infinite risk; one It is the maximum acceleration of the crane, used to regulate the effect of acceleration; This indicates that the smaller the distance, the higher the risk; This indicates that the greater the relative speed, the higher the risk, and it is standardized to the ratio of the maximum speed. This indicates the effect of acceleration. A larger acceleration means the crane is making rapid adjustments, increasing the risk of collisions or braking.

[0008] Preferably, the environmental perception module includes a lidar, an ultrasonic sensor, a high-definition camera and vision processing system, and an infrared sensor; wherein the lidar can provide high-precision point cloud data in three-dimensional space for sensing the distance and position of surrounding obstacles; the ultrasonic sensor is used to measure the distance between the crane and surrounding objects; the high-definition camera acquires real-time images of the surroundings, and combined with computer vision technology, can identify the position information of obstacles, personnel, and equipment; the infrared sensor is used to help the crane perceive the temperature differences of invisible obstacles or objects in low-light environments or at night.

[0009] Preferably, the dynamic early warning and collision avoidance system combines LiDAR and visual sensor data to calculate the crane's movement path and dynamically adjust it. If the system detects that an obstacle is about to approach the crane's working path, it automatically adjusts the crane's travel route to avoid the obstacle. Based on sensor data, the system will analyze in real time whether the crane will collide with the obstacle.

[0010] Preferably, the human-machine interaction module, through a visual interface and real-time data display, helps the operator to view the environmental data around the crane in real time through the vehicle display screen or mobile terminal. The system will display a three-dimensional map, alarm information and suggested operations according to environmental changes. When the crane approaches a danger zone, the system will alert the operator through an audible alarm or vibration feedback.

[0011] Preferably, the system self-diagnosis and maintenance module can automatically detect the working status of each sensor and control system, and provide real-time feedback on whether the system has malfunctioned or failed. 7. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that the maintenance module supports remote diagnosis and technical support, enabling the technical team to remotely connect and view the problem in real time when a system problem occurs, and to troubleshoot and repair it.

[0012] Preferably, the risk rating and corresponding decision are transmitted to the human-computer interaction module. The module will trigger different levels of warnings according to the risk level. The operator receives the warning information through the display screen, sound alarm or vibration feedback and takes immediate measures to avoid collision.

[0013] Preferably, when encountering high-risk situations, the anti-collision module will activate an automatic obstacle avoidance mechanism, automatically adjusting the crane's travel path or position to ensure a safe distance from obstacles.

[0014] Preferably, based on the system self-diagnosis and maintenance module, all the aforementioned early warning events and system decisions are recorded for subsequent analysis and optimization. Based on actual operation data, the system will continuously learn and optimize the collision prediction model to improve the accuracy and response speed of the early warning. The self-diagnosis module will periodically check the working status of its sensors and execution systems to ensure the healthy operation of the system. In the event of a fault or sensor accuracy deviation, the system will automatically notify maintenance personnel to check and repair, avoiding system instability.

[0015] Preferably, the human-machine interaction module, through a visual interface and real-time data display, helps the operator to view the environmental data around the crane in real time through the vehicle display screen or mobile terminal. The system will display a three-dimensional map, alarm information and suggested operations according to environmental changes. Through the visual interface, the operator can better understand the current operating status of the crane and make timely adjustments. When the crane approaches a danger zone, the system will alert the operator through sound alarms or vibration feedback. In high-risk situations, the system can also trigger an automated safety shutdown to ensure that the crane stops moving and prevent accidents.

[0016] Preferably, the system self-diagnosis and maintenance module can automatically detect the working status of each sensor and control system, and provide real-time feedback on whether the system has malfunctioned or failed. If a malfunction occurs, the system will issue a warning to the operator and indicate the location of the malfunction for quick repair. The maintenance module supports remote diagnostics and technical support, enabling the technical team to view problems in real time via remote connection when system issues arise, conduct troubleshooting and repairs, and ensure the continuity and safety of crane operations.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention employs a multi-sensor data fusion method to ensure that the system can accurately perceive and react to the surrounding environment in three-dimensional space. Especially in complex and confined environments such as substations, sensor fusion can significantly improve the comprehensiveness and accuracy of environmental perception. The system can perform hazard prediction and intelligent decision-making. Moreover, the system can not only react in real time, but also learn and optimize itself based on historical data and environmental changes, thereby providing the optimal early warning strategy under different operating conditions. 2. The system of the present invention not only relies on automated operation, such as automatic path planning and braking mechanism, but also allows for manual intervention, so that operators can make flexible adjustments according to specific needs in special circumstances. Through intuitive 3D map display, operators can understand the working environment and crane status at any time, reducing the risk of misoperation due to unclear environment. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the operation logic of a three-dimensional operation crane anti-external damage early warning system for substations proposed in this invention; Figure 2 This is a system framework diagram of a three-dimensional operation crane anti-external damage early warning system for substations proposed in this invention; Figure 3 This is a schematic diagram of a three-dimensional operation crane anti-external damage early warning system for substations proposed in this invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Reference Figure 1-3A three-dimensional operation crane anti-external damage early warning system for substations includes an environmental perception module, a dynamic early warning and anti-collision module, a data fusion and decision-making module, a human-computer interaction operation module, and a system self-diagnosis and maintenance module.

[0021] Furthermore, the environmental perception module is responsible for sensing the environment around the crane, collecting and analyzing information about surrounding obstacles to ensure that the trolley can operate safely in complex environments. The environmental perception module includes lidar, ultrasonic sensors, high-definition cameras and vision processing systems, and infrared sensors. Among them, LiDAR can provide high-precision point cloud data in three-dimensional space to perceive the distance and position of surrounding obstacles. LiDAR can be installed in key positions of the crane, such as the end of the boom and around the vehicle body, to scan the surrounding environment in real time; ultrasonic sensors are used to measure the distance between the crane and surrounding objects; high-definition cameras acquire real-time images of the surroundings, and combined with computer vision technology, can identify the position information of obstacles, personnel and equipment. Through image processing, the accuracy of environmental perception is further improved; infrared sensors are used in low-light environments or nighttime operations to help the crane perceive invisible obstacles or temperature differences of objects around it.

[0022] At this point, the environmental perception system achieves comprehensive environmental monitoring through the careful selection and deployment of various sensors. The system employs high-precision 3D LiDAR, installed at key locations such as the end of the boom, the roof, and the sides of the vehicle, acquiring rich point cloud data through 360° panoramic scanning. Simultaneously, it is equipped with high-sensitivity ultrasonic sensors, primarily positioned below the boom and around the vehicle, focusing on detecting nearby obstacles within a 2-5 meter range. For the vision system, high-resolution cameras combined with advanced image recognition algorithms accurately identify and classify different types of obstacles, including electrical equipment, personnel, and vehicles. Furthermore, infrared sensors provide additional safety for operations in nighttime or low-light environments.

[0023] After each sensor operates independently, the system uses fusion algorithms such as Kalman filtering to integrate and process multi-source data, effectively eliminating data redundancy and errors between different sensors, and ultimately generating an accurate 3D environment model. The entire environmental perception process starts from sensor activation, goes through data preprocessing, multi-source fusion, obstacle identification and classification, and finally transmits reliable environmental information to the decision-making module, providing a solid technical foundation for the safe operation of the crane. This multi-sensor fusion scheme not only improves the accuracy and reliability of environmental perception, but also adapts to various complex operating environments, ensuring the safe operation of the crane during construction.

[0024] Furthermore, the dynamic warning and collision avoidance module is responsible for monitoring the distance and position of the crane and surrounding obstacles in real time, predicting collision risks, and issuing warnings when a collision is about to occur to avoid safety accidents. The dynamic warning and collision avoidance system combines LiDAR and visual sensor data to calculate the crane's movement path and dynamically adjust it. If the system detects that an obstacle is about to approach the crane's working path, it automatically adjusts the crane's travel route to avoid the obstacle.

[0025] Based on sensor data, the system will analyze in real time whether the crane will collide with obstacles. Once a danger is detected, the system will issue an alarm signal to remind the operator to adjust the crane's position in time. If a serious threat occurs, the system can suspend the crane's operation through the automatic control system. If the system detects an impending collision and the operator does not respond in time, the automatic braking mechanism will be triggered, and the crane will stop operating immediately to avoid an accident.

[0026] At this point, the dynamic early warning and collision avoidance module is the core system for ensuring the safe operation of the crane. Its workflow begins with the real-time acquisition of environmental data. The system works in concert with lidar, high-definition cameras and ultrasonic sensors to acquire three-dimensional point clouds, images and precise distance information of the surrounding environment. This multi-source data is then sent to the data fusion and decision-making module to be integrated into a real-time three-dimensional map containing key information such as the location, shape, size and distance of obstacles. On this basis, the system uses computer vision and deep learning algorithms to accurately identify and classify obstacles in the map, distinguishing them into fixed obstacles such as buildings and dynamic obstacles such as people and vehicles.

[0027] The collision risk assessment is the decision-making center of the module. Based on the crane's current position, speed, trajectory, and relative relationship with obstacles, it dynamically calculates the collision risk level using a collision risk function that comprehensively considers maximum speed, maximum acceleration, real-time distance, relative speed, and acceleration. This level is divided into three grades: low, medium, and high, triggering corresponding early warning and response mechanisms.

[0028] In low-risk situations, only prompts are displayed on the operating interface; in medium-risk situations, the operator is alerted via sound or vibration alarms; and once a high-risk state is entered, the system will immediately initiate automatic obstacle avoidance or safety shutdown procedures. The automatic obstacle avoidance function uses path planning algorithms such as Dijkstra to calculate the optimal avoidance path and precisely adjusts the crane's movement through PID or fuzzy control algorithms. In extreme cases, if a collision is imminent and the operator fails to respond, the automatic safety shutdown mechanism will force the crane to brake urgently. Through this series of closed-loop processes from perception, assessment to decision-making and execution, this module achieves comprehensive dynamic protection for crane operations, greatly improving operational efficiency and safety in complex environments within substations.

[0029] Specifically, when the collision risk level reaches its highest level and the system determines that the operator has failed to respond in time, or the sensors directly detect that a collision is unavoidable, the automatic safety shutdown mechanism will be triggered. This is a millisecond-level response process: the system will immediately issue the highest priority shutdown command to the crane control system, cut off the power, and activate the emergency braking system, forcing the crane to decelerate to a stop in the shortest possible time. At the same time, audible and visual alarms will be activated to alert all personnel on site. To achieve such an efficient and reliable mechanism, three major technical challenges must be overcome: real-time performance, robustness, and safety. This ensures that the system can make both rapid and safe decisions under any complex working conditions, thereby providing a solid and reliable safety guarantee for high-risk hoisting operations within substations.

[0030] Furthermore, the data fusion and decision-making module integrates data from various sensors to generate accurate and comprehensive environmental data, providing support for system decision-making. At the same time, the core task of the data fusion and decision-making module is to efficiently integrate heterogeneous data from multiple sources such as lidar, visual sensors, ultrasonic and infrared sensors, thereby constructing an accurate, comprehensive and real-time environmental situation map.

[0031] To achieve this goal, the system will employ advanced data fusion algorithms, such as the classic Kalman filter or particle filter suitable for nonlinear systems, to handle the differences in accuracy, format, and update frequency between different sensor data, effectively filter out noise and compensate for uncertainties, and ultimately generate a highly consistent and reliable environmental model.

[0032] This integrated and accurate environmental model is the cornerstone of all subsequent decision support. Based on this, the module can accurately assess the collision risk between the crane and surrounding obstacles, and dynamically classify the risk level by calculating key parameters such as distance, relative speed, trajectory and time window. At the same time, the model also provides the necessary environmental information for the path planning algorithm, enabling it to calculate the optimal path to avoid obstacles.

[0033] The decision-making module integrates the results of risk assessment and path planning to generate specific control commands, directly driving the crane to perform obstacle avoidance or safe shutdown operations. Faced with key technical challenges such as data inconsistency, computational efficiency requirements, and algorithm selection, a well-designed data fusion and decision-making module is the fundamental guarantee that the system can make fast, accurate, and safe decisions under various complex working conditions.

[0034] Furthermore, the human-machine interaction module provides operators with an interactive interface with the system, displaying real-time monitoring data and alarm information, and supporting manual intervention by operators. Through a visual interface and real-time data display, the human-machine interaction module helps operators view environmental data around the crane in real time via the vehicle display screen or mobile terminal. The system will display a 3D map, alarm information, and suggested operations based on environmental changes. Through the visual interface, operators can better understand the current operating status of the crane and make timely adjustments.

[0035] When the crane approaches a danger zone, the system will alert the operator through sound alarms or vibration feedback. In high-risk situations, the system can also trigger an automated safety shutdown to ensure that the crane stops moving and prevent accidents.

[0036] At this point, the human-computer interaction module acts as a bridge connecting the intelligent system and the human operator. Its core design is to transform complex environmental data and system decisions into intuitive, easy-to-understand, and operable information, while always retaining the final decision-making power for human intervention. To achieve this, the interface adopts a sophisticated hierarchical design to avoid information overload. The main view presents a real-time 3D environmental map driven by the data fusion module, which not only accurately displays the crane model and surrounding obstacles, but also classifies and marks personnel, vehicles, and fixed equipment through different colors and icons, and dynamically displays the crane's planned path and the predicted trajectory of obstacles, making potential risks clear at a glance. Next to the main view, the information panel displays key data such as crane speed, boom angle, and distance to the nearest obstacle in real time in the form of an instrument panel, while alarm information will pop up in different eye-catching ways, from yellow banners to red borders, according to the risk level.

[0037] The system also incorporates a multimodal interaction and early warning mechanism to ensure effective information delivery in any environment. In low-risk situations, only dangerous areas are highlighted on the map; in medium-risk situations, an audible alarm or seat vibration is triggered; and in high-risk situations, a full-screen red warning appears accompanied by a rapid alarm sound, and the system immediately enters standby mode. More importantly, the operator always retains ultimate control, able to interrupt automatic operation and switch back to manual mode at any time via interfaces such as canceling alarms and taking over control, to handle special conditions that the algorithm struggles to process. This perfectly embodies the design philosophy of combining automation with human intervention. Furthermore, the system supports compatibility with in-vehicle screens and mobile terminals, facilitating monitoring from different locations.

[0038] When all warnings and alerts fail to prevent a crisis, automated safety shutdown becomes the last line of defense. Upon detecting a high risk and the operator's failure to respond promptly, or if their operational error exacerbates the risk, the system immediately cuts off power and initiates emergency braking, clearly displaying on the screen that an emergency shutdown has been triggered and recording relevant data. Through this closed-loop process—from visualization and multimodal warnings to manual intervention and final assurance—this module successfully transforms a complex collision avoidance system into an efficient, intuitive, and reliable intelligent co-pilot, significantly improving the safety and operational efficiency of crane operations.

[0039] Furthermore, the system self-diagnosis and maintenance module is responsible for monitoring the working status of each component of the system to ensure the normal operation of the system. It also provides diagnostic information and maintenance suggestions when a fault occurs. The system self-diagnosis and maintenance module can automatically detect the working status of each sensor and control system, and provide real-time feedback on whether the system has a fault or failure. If a fault occurs, the system will issue a warning to the operator and mark the fault location for quick repair.

[0040] At this point, the core mission of the system self-diagnosis and maintenance module is to ensure that the system itself is always in the best working state and has the ability to continuously evolve. Through a comprehensive real-time status monitoring mechanism, this module acts like a tireless doctor, monitoring all states from the sensor layer, control system layer to the actuator layer 24 hours a day. It uses diagnostic mechanisms such as threshold judgment and logic verification to instantly detect various potential problems such as interruption of LiDAR data stream, abnormal camera frame rate, or processor overload.

[0041] Once a fault is detected, the system immediately initiates a standardized tiered alarm process. For non-critical issues such as a slight decrease in sensor accuracy, the system will issue a warning with a yellow indicator light on the human-machine interface. For serious faults such as the failure of the main LiDAR, a red alarm will be triggered, and the system may automatically switch to a backup sensor or restrict some functions to ensure basic safety. All fault events, including the time, type, and system status at the time, will be recorded in detail in the fault log, providing crucial information for subsequent analysis. In addition, the system integrates powerful remote diagnostic capabilities. When complex problems cannot be resolved on-site, the technical team can view the system status and analyze the logs in real time through a secure network, and even perform remote repairs and configurations with authorization, greatly reducing maintenance time.

[0042] More forward-lookingly, this module endows the system with the ability to learn and optimize itself. It records every warning event, system decision, and operator intervention as valuable data, and uses machine learning algorithms to continuously iterate and optimize its internal collision prediction model in the background. For example, if the system finds that the visual recognition rate of a specific device is low in light rain, leading to false alarms, it will automatically adjust the algorithm weights for that scenario. Through this data-driven continuous evolution, the system's warning accuracy and response speed in various complex environments will be significantly improved. This is achieved through a complete system that integrates real-time monitoring, intelligent alarms, remote support, and self-learning.

[0043] The maintenance module supports remote diagnostics and technical support, enabling the technical team to view problems in real time via remote connection when system issues arise, conduct troubleshooting and repairs, and ensure the continuity and safety of crane operations.

[0044] At this point, remote diagnostics and technical support are key to ensuring the high availability of the early warning system and enabling rapid maintenance. It completely breaks down geographical and time limitations, seamlessly extending the capabilities of the backend expert team to the work site. To achieve this goal, the system first establishes a stable and reliable communication link through built-in industrial-grade 4G / 5G modules or substation fiber optic networks, and adopts strict security mechanisms such as VPN encrypted tunnels and two-factor authentication to ensure that all remote access is conducted in an absolutely secure environment. At the same time, a refined access control system ensures that technicians at different levels can only perform the operations they are authorized to perform, from reading-only access to reading-write level parameter adjustments and remote control.

[0045] Once a secure connection is established, the remote technical team gains real-time access to everything the crane operator sees on the vehicle's screen, including 3D environment maps, alarm information, and sensor data streams. More importantly, they have deep access to system logs, fault records, and underlying performance data such as CPU and memory. They can also collaborate with on-site operators through built-in VoIP or text chat functions to quickly pinpoint the root cause of problems.

[0046] With explicit authorization, experts can even intervene remotely, adjusting sensor sensitivity, restarting unresponsive software services, or pushing the latest firmware and software updates, greatly shortening fault repair time. The ultimate value of this function lies in maximizing the continuity and safety of operations. Software or configuration problems that used to take hours or even days to resolve can now be handled within minutes, significantly improving crane operating efficiency. The deep involvement of experts at the back end also ensures the professionalism and thoroughness of maintenance. More importantly, when potential safety risks arise in the system that are not detected by on-site personnel, remote experts can guide the implementation of safety measures immediately, and even trigger remote safety shutdowns in emergencies, minimizing the risk of accidents. Through this powerful back-end support system, the early warning system is no longer an isolated standalone machine, but a network node connected to the experts' brains, ensuring that it maintains the highest level of operational continuity and safety when facing various complex challenges.

[0047] The operation of the dynamic early warning and collision avoidance module includes the following steps: Based on various environmental sensors in the environmental perception module, the surrounding environmental data is collected in real time and transmitted to the data fusion and decision module for further processing and analysis. Based on the information obtained from the data fusion and decision-making module, the system generates a real-time 3D environment map around the crane, marking the position, shape and distance of each obstacle. The system uses computer vision, image processing or deep learning algorithms to identify and classify obstacles (such as fixed obstacles, dynamic obstacles, etc.) and transmits the generated information to the dynamic early warning and collision avoidance module to ensure real-time understanding of the dynamic changes in the surrounding environment. Based on the 3D environment map and obstacle information generated by the system, the dynamic early warning and collision avoidance module assesses the collision risk by taking into account the crane's current position, speed, trajectory and relative position with surrounding obstacles. It calculates factors such as the shortest distance between the crane and obstacles, direction of movement, and time window to determine whether there is a potential collision risk. Based on the assessment results, the system assigns a risk level (low risk, medium risk, high risk) to each potential collision event and determines the corresponding response strategy. Risk rating and corresponding decisions are transmitted to the human-machine interaction module. The module will trigger different levels of warnings based on the risk level. The operator receives the warning information through the display screen, sound alarm or vibration feedback and takes immediate measures to avoid collisions. When encountering high-risk situations, the anti-collision module will activate the automatic obstacle avoidance mechanism and automatically adjust the crane's travel path or position to ensure a safe distance from obstacles. If adjusting the path cannot effectively avoid collisions, the system may directly and automatically stop the crane to prevent physical contact or damage. Based on the system self-diagnosis and maintenance module, all the above-mentioned early warning events and system decisions are recorded for subsequent analysis and optimization, including obstacle location, early warning trigger time, decision execution status, etc. Based on actual operation data, the system will continuously learn and optimize the collision prediction model to improve the accuracy and response speed of early warnings. The self-diagnosis module will regularly check the working status of its sensors and execution system to ensure the healthy operation of the system. If a sensor failure or false alarm is found, the system will perform self-diagnosis and notify the maintenance personnel. When a fault occurs or the sensor accuracy deviates, the system will automatically notify the maintenance personnel to check and repair, so as to avoid system instability. The backend system cleans, structures, and automatically labels massive amounts of data. For example, it identifies correct warnings or false alarm scenarios by comparing system warnings with actual results. Based on this high-quality data, the system regularly iterates and retrains the core collision prediction model or fine-tunes key parameters in the risk function. For example, when the system detects a decrease in visual recognition rate in rainy or foggy weather, it learns how to more effectively fuse LiDAR and infrared data to compensate for this deficiency. Before full deployment, the new model undergoes rigorous testing to verify its superiority in real-world operating environments, ensuring the robustness of the upgrade.

[0048] By analyzing long-term trends in sensor performance, the system can predict the remaining lifespan of a component and issue replacement recommendations in advance, such as predicting that an ultrasonic sensor's accuracy will exceed the standard in a month. Through correlation analysis, the system can even discern deep connections between faults. For example, from the concurrency of high CPU usage and LiDAR data packet loss, it can be inferred that heat dissipation may be the root cause. Through this mechanism that integrates comprehensive recording, closed-loop learning, and predictive maintenance, the early warning system truly possesses the ability to self-evolve, learning from each operation and continuously optimizing its early warning strategies under different operating conditions.

[0049] The dynamic early warning and collision avoidance module generates a real-time risk assessment curve based on the crane's current position, speed, trajectory, and relative position to surrounding obstacles, and evaluates the risk level according to the following collision risk function: in This is the crane's maximum speed, to prevent infinite risk; one It is the maximum acceleration of the crane, used to regulate the effect of acceleration; This indicates that the smaller the distance, the higher the risk; This indicates that the greater the relative speed, the higher the risk, and it is standardized to the ratio of the maximum speed. This indicates the effect of acceleration. A larger acceleration means the crane is making rapid adjustments, increasing the risk of collisions or braking.

[0050] Based on the calculated collision risk R(t), the risk can be divided into the following different levels: Low risk: R(t) < 0.2 (indicating a low probability of collision, operation can continue); at the same time, when the risk value is in the low risk range (R(t) < 0.2), the system determines that the crane maintains a sufficient safe distance from the obstacle and the relative motion state is stable; at this time, the system allows the crane to continue normal operation, and the human-machine interface only displays obstacle information in the conventional way to avoid unnecessary alarms from interfering with the operator's attention.

[0051] Medium risk: 0.2≤R(t)<0.5 (Attention is required and preventive measures may be taken); Once the risk value enters the medium risk range (0.2 ≤ R(t)<0.5), the system considers a potential threat to have emerged, possibly due to a decrease in distance or an increase in relative speed. At this time, the system will immediately activate the human-machine collaborative early warning mechanism, reminding the operator to be vigilant through sound alarms or vibration of the operating seat, highlighting the hazard source on the interface, and possibly providing specific operational prompts such as suggestions to slow down, guiding the operator to proactively mitigate the risk.

[0052] High risk: R(t) ≥ 0.5 (immediate stop or take emergency avoidance action); When the risk value exceeds the threshold and reaches the high risk level (R(t) ≥ 0.5), the system judges that a collision is imminent and decisive measures must be taken; In this emergency state, the system will trigger the highest level of multimodal alarm and bypass manual intervention to directly activate the automatic collision avoidance mechanism, such as executing an automatic obstacle avoidance path or immediately triggering a safety shutdown procedure, to minimize the risk of the accident with the highest machine efficiency; This three-level risk classification system enables the system to intelligently allocate different response strategies from silent monitoring, human-machine collaboration to automatic intervention according to the urgency of the threat, achieving the best balance between safety and efficiency.

[0053] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A three-dimensional operation crane anti-external damage early warning system for substations, characterized in that, It includes an environmental perception module, a dynamic early warning and collision avoidance module, a data fusion and decision-making module, a human-computer interaction operation module, and a system self-diagnosis and maintenance module; The environmental perception module is responsible for sensing the environment around the crane and collecting and analyzing information about surrounding obstacles; the dynamic early warning and collision avoidance module is responsible for monitoring the distance and position of the crane and surrounding obstacles in real time; the data fusion and decision-making module fuses data from various sensors to generate environmental data; the human-machine interaction module provides the operator with an interactive interface with the system; and the system self-diagnosis and maintenance module is responsible for monitoring the working status of each component of the system. The operation of the dynamic early warning and collision avoidance module includes the following steps: Based on the various environmental sensors in the environmental perception module, the surrounding environmental data is collected in real time and the collected data is transmitted to the data fusion and decision-making module. Based on the information analyzed and processed by the data fusion and decision-making module, the system generates a real-time three-dimensional environment map around the crane, marking the position, shape and distance of each obstacle. The system uses computer vision, image processing or deep learning algorithms to identify and classify obstacles, and transmits the generated information to the dynamic early warning and anti-collision module. Based on the 3D environment map and obstacle information generated by the system, the dynamic early warning and collision avoidance module performs a collision risk assessment based on the crane's current position, speed, trajectory and relative position with surrounding obstacles, determines whether there is a potential collision risk, and assigns a risk level to each potential collision event based on the assessment results, and decides on the corresponding response strategy.

2. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, The dynamic early warning and collision avoidance module generates a real-time risk assessment curve based on the crane's current position, speed, trajectory, and relative position to surrounding obstacles, and assesses the risk level according to the following collision risk function: in This is the crane's maximum speed, to prevent infinite risk; one It is the maximum acceleration of the crane, used to regulate the effect of acceleration; This indicates that the smaller the distance, the higher the risk; This indicates that the greater the relative speed, the higher the risk, and it is standardized to the ratio of the maximum speed. This indicates the effect of acceleration. A larger acceleration means the crane is making rapid adjustments, increasing the risk of collisions or braking.

3. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, The environmental perception module includes a lidar, an ultrasonic sensor, a high-definition camera and vision processing system, and an infrared sensor. The lidar provides high-precision point cloud data in three-dimensional space to perceive the distance and position of surrounding obstacles. The ultrasonic sensor measures the distance between the crane and surrounding objects. The high-definition camera acquires real-time images of the surroundings, and combined with computer vision technology, it can identify the location information of obstacles, personnel, and equipment. The infrared sensor helps the crane perceive invisible obstacles or temperature differences in objects during low-light environments or nighttime operations.

4. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, The dynamic early warning and collision avoidance system combines LiDAR and visual sensor data to calculate the crane's movement path and dynamically adjust it. If the system detects an obstacle approaching the crane's working path, it automatically adjusts the crane's route to avoid the obstacle. Based on sensor data, the system will analyze in real time whether the crane will collide with the obstacle.

5. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, The human-machine interaction module helps operators view environmental data around the crane in real time through a visual interface and real-time data display. The system will display a 3D map, alarm information and suggested operations according to environmental changes. When the crane approaches a danger zone, the system will alert the operator through an audible alarm or vibration feedback.

6. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, The system self-diagnosis and maintenance module can automatically detect the working status of each sensor and control system, and provide real-time feedback on whether the system has malfunctioned or failed.

7. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, The maintenance module supports remote diagnostics and technical support, enabling the technical team to view problems in real time via remote connection when system issues arise, and to troubleshoot and repair them.

8. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, Risk rating and corresponding decisions are transmitted to the human-machine interaction module. The module will trigger different levels of warnings based on the risk level. The operator receives the warning information through the display screen, sound alarm or vibration feedback and takes immediate measures to avoid collision.

9. The substation three-dimensional operation crane anti-external damage early warning system according to claim 8, characterized in that, When encountering high-risk situations, the anti-collision module will activate an automatic obstacle avoidance mechanism, automatically adjusting the crane's travel path or position to ensure a safe distance from obstacles.

10. The substation three-dimensional operation crane anti-external damage early warning system according to claim 1, characterized in that, Based on the system self-diagnosis and maintenance module, all the aforementioned early warning events and system decisions are recorded for subsequent analysis and optimization. Based on actual operation data, the system will continuously learn and optimize the collision prediction model to improve the accuracy and response speed of early warnings. The self-diagnosis module will periodically check the working status of its sensors and execution systems to ensure the healthy operation of the system. In the event of a fault or sensor accuracy deviation, the system will automatically notify maintenance personnel for inspection and repair to avoid system instability.

Images