Crane unhooking early warning system

Abstract

The invention relates to the technical field of data processing, and particularly provides a crane unhooking early warning system, which comprises: a data acquisition end for acquiring working state information of a crane hook, the working state information comprising a load state of the hook, a position of an anti-falling baffle plate and a state of a locking spring; the data acquisition end comprises a plurality of sensors, and the plurality of sensors are used for measuring the pressure of the locking spring; and the data edge calculation end obtains the pressure of the locking spring, a reference pressure range is set according to the working state of the lifting hook anti-falling device, and if the pressure value of the locking spring exceeds the pressure range, alarm information is generated. The breaking risk caused by the fact that the anti-unhooking device is in a working state for a long time can be effectively avoided.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing, and in particular relates to a crane unhooking early warning system. Background Art

[0002] As a representative of heavy machinery and equipment, cranes are widely used in various large-scale engineering scenarios. Their core function is to complete the lifting of heavy objects efficiently and safely. However, during the lifting process, once the heavy objects on the sling are accidentally unhooked, it may not only directly injure the on-site workers, but also cause immeasurable damage to the surrounding environment and facilities. Therefore, in order to ensure the safety of lifting operations, cranes must be equipped with special hook anti-drop devices.

[0003] The core component of the hook anti-drop device is the anti-drop baffle, which can quickly take effect at the moment of the heavy object being unhooked, effectively preventing the heavy object from falling. However, it is worth noting that the anti-drop baffle will gradually wear out during long-term use and frequent stress, resulting in a decrease in its anti-drop performance. Therefore, it is very necessary to avoid the anti-drop baffle being in a stress state for a long time.

[0004] In view of the above situation, real-time monitoring and early warning of crane unhooking risks are particularly important. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, the present invention provides a crane unhooking warning system to solve the above-mentioned technical problems.

[0006] In a first aspect, the present invention provides a crane unhooking warning system, comprising: A data collection terminal is used to collect the working status information of the crane hook, wherein the working status information includes the load status of the hook, the position of the anti-drop baffle plate, and the status of the locking spring; The data acquisition end includes a plurality of sensors, and the plurality of sensors are used to measure the pressure of the locking spring; The data edge computing end obtains the pressure of the locking spring and sets the reference pressure range according to the working status of the hook anti-dropping device. If the pressure value of the locking spring exceeds the pressure range, an alarm message is generated.

[0007] In an optional embodiment, the anti-unhooking device includes: A spring, a protective pin, an anti-slip baffle, a data acquisition terminal, and a protective cover. The fixed end of the anti-slip baffle is rotatably connected to the hook body through the protective pin, and the suspended end of the anti-slip baffle is clamped on the inner side of the end of the hook; the spring is arranged on the protective pin to push the suspended end of the anti-slip baffle to the inner side of the end of the hook; the data acquisition terminal is installed on the side of the anti-slip baffle facing the hook, and the protective cover is laid on the surface of the data acquisition terminal.

[0008] In an optional embodiment, the data acquisition end includes a patch pressure sensor, a controller and a wireless communication chip, and the detection patch of the patch pressure sensor is attached to the surface of the protective cover and the force-bearing surface of the anti-detachment baffle; the patch pressure sensor and the wireless communication chip are both electrically connected to the controller.

[0009] In an optional implementation, the data acquisition end uses the modbus standard protocol to communicate wirelessly with the data edge computing end, and supports a semi-open ad hoc networking mode.

[0010] In an optional implementation, the data edge computing end includes: The first acquisition module is used to continuously acquire pressure values ​​at the base of the crane hook, and arrange the acquired pressure values ​​into a first data sequence according to the acquisition time; A second acquisition module is used to continuously acquire pressure values ​​of the force-bearing surface of the anti-unhooking device of the crane hook, compare the acquired pressure values ​​with a preset pressure threshold, and arrange the continuous pressure values ​​of the force-bearing surface exceeding the pressure threshold in chronological order of acquisition time into a second data sequence; A risk identification module, used for inputting the first data sequence into a pre-trained long short-term memory neural network model to identify the decoupling risk probability; A length setting module, used to generate a length threshold of a second data sequence based on the decoupling risk probability; The warning generation module is used to confirm that the length of the second data sequence generated synchronously reaches the length threshold and generate a decoupling warning prompt message.

[0011] In an optional embodiment, the pressure values ​​of the force-bearing surface of the anti-unhooking device of the crane hook are continuously collected, the collected pressure values ​​are compared with a preset pressure threshold, and the continuous pressure values ​​of the force-bearing surface exceeding the pressure threshold are arranged in chronological order of collection time into a second data sequence, including: Determine whether the current pressure value exceeding the pressure threshold and the last pressure value of the current second data sequence are continuous pressure values: If so, adding the current pressure value exceeding the pressure threshold to the current second data sequence; If not, the current second data sequence is deleted, and a new second data sequence is created, and the current pressure value exceeding the pressure threshold is used as the starting value of the new second data sequence.

[0012] In an optional implementation, generating a length threshold of the second data sequence based on the decoupling risk probability includes: Pre-set multiple decoupling risk probability levels, and set the length threshold corresponding to each decoupling risk probability level; A target decoupling risk probability level is matched to the decoupling risk probability, and a length threshold corresponding to the target decoupling risk probability level is set as a length threshold of the second data sequence.

[0013] In an optional implementation, confirming that the length of the synchronously generated second data sequence reaches the length threshold and generating decoupling warning prompt information includes: Obtaining a time range of the first data sequence, and verifying whether the second data sequence is within the time range, and if the second data sequence is not within the time range, determining that the second data sequence is invalid; Confirm that the second data sequence is within the time range, and determine whether the length of the second data sequence reaches the length threshold: If yes, a decoupling warning message is generated, and a lifting height restriction instruction is sent to the execution agency; If not, the decoupling risk probability is written into the log file.

[0014] In an optional embodiment, it also includes: Obtain remaining computing resources and communication quality parameters with edge devices; If the remaining computing resources do not exceed the set resource threshold and the communication quality parameter reaches the set quality threshold, the first data sequence is sent to the edge device so that the edge device processes the first data sequence using a pre-deployed pre-trained long short-term memory neural network model.

[0015] The beneficial effect of the present invention is that the crane unhooking warning system provided by the present invention can accurately detect the working status of the anti-unhooking device and then generate warning information, thereby effectively avoiding the risk of breakage caused by the anti-unhooking device being in a working state for a long time.

[0016] In addition, the invention has a reliable design principle, a simple structure and a very broad application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 It is a schematic structural diagram of an anti-unhooking device according to an embodiment of the present invention.

[0019] Figure 2 The figure is a diagram showing the internal structure of an anti-unhooking device according to an embodiment of the present invention.

[0020] Figure 3 The figure is a bottom schematic diagram of an anti-unhooking device according to an embodiment of the present invention.

[0021] Figure 4 It is a schematic diagram of an application scenario of an anti-unhooking device according to an embodiment of the present invention.

[0022] In the figure, 1. spring, 2. protection pin, 3. anti-slip baffle, 4. data acquisition terminal, 5. protective cover, 6. hook, 401. signal generator, 402. circuit board, 403. lithium battery pack, 404. charging interface, 405. device switch, 406. pressure acquisition terminal, 407. strong magnetic adhesive block, 7. portable alarm, 8. data edge computing terminal. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0025] Please refer to Figure 1 , this embodiment provides a crane unhooking warning system, including: A data collection terminal is used to collect the working status information of the crane hook, wherein the working status information includes the load status of the hook, the position of the anti-drop baffle plate, and the status of the locking spring; The data acquisition end includes a plurality of sensors, and the plurality of sensors are used to measure the pressure of the locking spring; The data edge computing end obtains the pressure of the locking spring and sets the reference pressure range according to the working status of the hook anti-dropping device. If the pressure value of the locking spring exceeds the pressure range, an alarm message is generated.

[0026] Crane hook monitoring device Figure 2-Figure 3As shown, it is composed of a spring 1, a protective pin 2, an anti-slip baffle 3, a data acquisition terminal 4, a protective cover 5, a signal generator 401, a circuit board 402, a lithium battery pack 403, a charging interface 404, a device switch 405, a pressure acquisition terminal 406, and a strong magnetic adhesive block 407, and is characterized in that: the anti-slip baffle 3 and the spring 1 are fixed on the hook 6 by the protective pin 2, the lithium battery pack 403 and the circuit board 402 are fixed inside the data acquisition terminal 4, the pressure acquisition terminal 406 is fixed on the bottom front side of the data acquisition terminal 4, the data acquisition terminal 4 is fixed inside the anti-slip baffle 3 by the strong magnetic adhesive block 407, the signal generator 401 is connected to the circuit board 402 and fixed at the front of the data acquisition terminal 4, the device switch 405 is connected to the circuit board 402 and fixed at the right rear side of the data acquisition terminal 4, the charging interface 404 is connected to the lithium battery pack 403 and fixed at the left rear side of the data acquisition terminal 4, and the protective cover 5 is fixed on the upper part of the data acquisition terminal 4.

[0027] When installing, the device is fixed inside the anti-drop baffle by a slide-in installation method using a strong magnetic adhesive block, without changing the original operating procedures of the hook.

[0028] The monitor is divided into two parts: the acquisition end and the data calculation and processing end. The data calculation and processing end is further divided into two types: the data edge computing end and the portable alarm end to adapt to different application scenarios. The data edge computing end has the functions of initializing the acquisition end, returning data calculation and processing, alarm information processing, acquisition end battery power monitoring, sound and light alarm, etc., and has a serial interface and a digital switch interface to facilitate the integration of the upper system. The portable alarm end has the functions of initializing the acquisition end, alarm information processing, and sound and light alarm.

[0029] The acquisition end and the data calculation and processing end are designed separately, and the two types of data calculation and processing ends are more applicable; the acquisition end and the anti-drop baffle are fixed by a strong magnetic adhesive block, which takes into account the convenience of replacement and charging and the reliability during operation; the acquisition end is battery-powered and communicates with the calculation end by wireless communication, avoiding the influence of wiring on the operation of the hook; by measuring the spring force to determine whether the anti-drop baffle is in the correct position, reducing the dependence of the acquisition end on the hook structure; it has the function of autonomously judging whether the hook is working; it has the function of adaptively adjusting the position of the anti-drop baffle, and can autonomously correct the judgment basis according to the actual working data, improving the judgment accuracy The data transmission frequency is controlled according to the working status of the hook to achieve the best balance between power consumption and monitoring timeliness; the data collection end adopts a sliding installation method based on the structure of the anti-slip baffle, which is completely embedded in the anti-slip baffle without changing the original operating procedures of the hook; the wireless communication adopts the modbus standard protocol, which is convenient for docking with other systems; the wireless communication adopts a semi-open self-organizing network mode, and the portable alarm terminal can receive alarms at the same time as the data edge computing end, or it can be used alone to receive alarm prompts.

[0030] In another embodiment of the present invention, the device structure comprises: Spring 1, protective pin 2, anti-slip baffle 3, data acquisition terminal 4, protective cover 5, the fixed end of the anti-slip baffle 3 is rotatably connected to the hook 6 through the protective pin 2, and the suspended end of the anti-slip baffle 3 is stuck on the inner side of the end of the hook; the spring 1 is arranged on the protective pin 2, and is used to push the suspended end of the anti-slip baffle 3 to the inner side of the end of the hook; the data acquisition terminal 4 is installed on the side of the anti-slip baffle 3 facing the hook, and the protective cover 5 is laid on the surface of the data acquisition terminal 4; the data acquisition terminal 4 includes a patch-type pressure sensor, a controller and a wireless communication chip, and the detection patch of the patch-type pressure sensor is attached to the surface of the protective cover 5 and the force-bearing surface of the anti-slip baffle 3; the patch-type pressure sensor and the wireless communication chip are both electrically connected to the controller. Its state during application is as follows Figure 4 As shown, when installing, the device is fixed inside the anti-dropping baffle by a slide-in installation method using a strong magnetic adhesive block, without changing the original operating procedures of the hook.

[0031] The data edge computing end includes: The first acquisition module is used to continuously acquire pressure values ​​at the base of the crane hook, and arrange the acquired pressure values ​​into a first data sequence according to the acquisition time; A second acquisition module is used to continuously acquire pressure values ​​of the force-bearing surface of the anti-unhooking device of the crane hook, compare the acquired pressure values ​​with a preset pressure threshold, and arrange the continuous pressure values ​​of the force-bearing surface exceeding the pressure threshold in chronological order of acquisition time into a second data sequence; A risk identification module, used for inputting the first data sequence into a pre-trained long short-term memory neural network model to identify the decoupling risk probability; A length setting module, used to generate a length threshold of a second data sequence based on the decoupling risk probability; The warning generation module is used to confirm that the length of the second data sequence generated synchronously reaches the length threshold and generate a decoupling warning prompt message.

[0032] In an embodiment of the present invention, based on the first acquisition module, a possible embodiment is given below to illustrate its specific implementation scheme in a non-limiting manner.

[0033] The pressure values ​​at the base of the crane hook are continuously collected. These pressure values ​​can reflect the force applied to the hook during the lifting process. These pressure values ​​are arranged into the first data sequence according to the order of collection time. This step is the basis for understanding the working status of the hook and helps to grasp the dynamic changes of the hook during the lifting operation.

[0034] In an embodiment of the present invention, based on the second acquisition module, a possible embodiment is given below to illustrate its specific implementation scheme in a non-limiting manner.

[0035] The following is a specific example of generating the second data sequence: Assume that there is an anti-unhooking device for a crane hook, and the pressure value of its force-bearing surface is continuously collected. The preset pressure threshold is set to 100 units of pressure. Assume that the time is sampled in seconds, and a pressure value is collected every second.

[0036] Initial state: Before the second data sequence is generated, it is assumed that no pressure value exceeding the pressure threshold is recorded, and thus the second data sequence is empty.

[0037] Pressure value collection and comparison: Time t = 1 second: The collected pressure value is 90 units, which does not exceed the pressure threshold, and no operation is performed.

[0038] Time t = 2 seconds: The collected pressure value is 110 units, which exceeds the pressure threshold. At this time, check whether there is a previous pressure value that exceeds the threshold (that is, check whether the current second data sequence is empty). Since the current second data sequence is empty, a new second data sequence is created, and 110 units is used as the starting value.

[0039] Current second data sequence: {110} Time t = 3 seconds: The collected pressure value is 120 units, which exceeds the pressure threshold. Check whether the current pressure value exceeding the threshold (120 units) and the last pressure value of the current second data sequence (110 units) are consecutive pressure values. Since they are consecutive (i.e., 110 units are followed by 120 units, and both exceed the threshold), 120 units are added to the current second data sequence.

[0040] Current second data sequence: {110, 120} Time t=4 seconds: The collected pressure value is 95 units, which does not exceed the pressure threshold, and no operation is performed.

[0041] Time t = 5 seconds: The collected pressure value is 130 units, which exceeds the pressure threshold. Check again whether the current pressure value exceeding the threshold (130 units) and the last pressure value of the current second data sequence (120 units) are continuous pressure values. Since they are continuous, 130 units are added to the current second data sequence.

[0042] Current second data sequence: {110, 120, 130} Time t=6 seconds: The collected pressure value is 80 units, which does not exceed the pressure threshold, and no operation is performed.

[0043] Time t=7 seconds: The collected pressure value is 20 units, which is far below the pressure threshold and is not continuous with the previous pressure value exceeding the threshold (130 units). At this time, according to the rules, the current second data sequence should be deleted and a new second data sequence should be created. However, since there is no new pressure value exceeding the threshold at this time, the second data sequence will not be created immediately, but will wait for the next pressure value exceeding the threshold to appear.

[0044] In practical applications, if there is no pressure value exceeding the threshold for a long time, a timeout mechanism is set to clear or reset the second data sequence to avoid unnecessary resource occupation.

[0045] Time t=8 seconds: The collected pressure value is 140 units, which exceeds the pressure threshold again. Since the current second data sequence is empty at this time (due to the deletion operation in the previous step), a new second data sequence is created and 140 units is used as the starting value.

[0046] Current second data sequence: {140}.

[0047] In one embodiment of the present invention, based on the risk identification module, a possible embodiment will be given below to illustrate its specific implementation scheme in a non-limiting manner.

[0048] The first data sequence is input into the pre-trained long short-term memory neural network model. This model has been trained with a large amount of data and can accurately identify the probability of the hook being unhooked during the lifting process. Through the calculation of the model, a probability value of the unhooking risk can be obtained.

[0049] Before discussing the process of training the LSTM (Long Short-Term Memory) model for crane unhooking risk, it is necessary to train it first. Collect the pressure value sequence of the hook root before the crane unhooks and the pressure value sequence of the hook root under normal conditions. These data are the basis for model training, and they represent two completely different states: one is the abnormal state of the hook about to unhook, and the other is the normal operating state (although there are natural factors such as wind resistance).

[0050] The training process of the LSTM model can be roughly divided into the following steps: Data preprocessing: Data cleaning: Remove noisy data, such as outliers or invalid data due to equipment failure.

[0051] Feature extraction: Extract features useful for decoupling risk prediction from the original pressure value series. This may include statistics such as the mean, maximum, minimum, and standard deviation of the pressure values, as well as certain transformations of the time series (such as difference, moving average, etc.).

[0052] Labeling: Assign a label to each piece of data to indicate whether a decoupling event has occurred. This is usually a binary classification problem, where 1 indicates decoupling and 0 indicates normal.

[0053] Data partitioning: Divide the dataset into training, validation, and test sets to facilitate model evaluation and tuning during the training process.

[0054] Model construction: LSTM was chosen as the architecture for the model because of its excellent performance in processing time series data.

[0055] Depending on the size and complexity of the data, determine the number of LSTM layers, the number of neurons in each layer, and whether to use other types of layers (such as fully connected layers, Dropout layers, etc.) to enhance the performance of the model.

[0056] Set the loss function (such as cross entropy loss) and optimizer (such as Adam optimizer) to guide the model training process.

[0057] Model training: The training set is input into the LSTM model, and the difference (i.e., loss) between the predicted value and the actual value is calculated through forward propagation.

[0058] The gradients are calculated using the back-propagation algorithm and the model’s weights and biases are updated via the optimizer to minimize the loss.

[0059] This process is repeated until the model's performance on the validation set reaches an acceptable level or a preset number of iterations is reached.

[0060] Model Evaluation: Use the test set to evaluate the performance of the model, including indicators such as accuracy, recall, and F1 score.

[0061] The performance of the model can be further analyzed through visualization tools such as confusion matrix, ROC curve, etc.

[0062] Model tuning: Based on the evaluation results, adjust the model architecture, hyperparameters, or data preprocessing steps to improve the model's performance.

[0063] The process of training, evaluating, and tuning is repeated until the model reaches optimal performance.

[0064] Deployment and application: The trained model is deployed in practical applications to monitor the risk of crane hook unhooking in real time.

[0065] According to the prediction results of the model, corresponding preventive measures are taken to avoid the occurrence of decoupling events.

[0066] In one embodiment of the present invention, based on the length setting module, a possible embodiment will be given below to illustrate its specific implementation scheme in a non-limiting manner.

[0067] Based on this unhooking risk probability value, the system will generate a length threshold of the second data sequence corresponding to the risk probability. This length threshold represents how long the pressure state of the anti-unhooking device's force-bearing surface exceeding the pressure threshold must last before it is considered to be a unhooking risk.

[0068] Risk level classification: Based on historical data, expert experience and safety standards, the probability of decoupling risk can be divided into multiple levels, such as low risk, medium risk, high risk and extremely high risk.

[0069] Each level corresponds to a specific risk probability range. For example, low risk may correspond to a probability of 0%-20%, medium risk corresponds to 21%-50%, high risk corresponds to 51%-80%, and extremely high risk corresponds to 81%-100%.

[0070] Length threshold setting: For each risk level, a corresponding length threshold needs to be set. This threshold indicates how long the pressure state of the anti-unhooking device's force-bearing surface exceeding the pressure threshold needs to last before it is considered a risk of this level by the system.

[0071] The length threshold setting should be appropriately adjusted based on the risk level. For example, for low risk levels, the length threshold can be set relatively long, because even if the stress state continues for a period of time, the risk of decoupling is relatively low. For high or very high risk levels, the length threshold should be set short so that the system can respond quickly and issue an early warning.

[0072] Risk probability calculation and matching: During real-time monitoring, the system continuously collects the pressure values ​​of the crane hook and its anti-unhooking device, and calculates the current unhooking risk probability based on these data.

[0073] The calculated risk probability is matched with the pre-set risk level to determine the current risk level.

[0074] Length threshold setting: Once the current risk level is determined, the system will set the length threshold of the second data sequence according to the length threshold corresponding to the level.

[0075] This length threshold will be used for subsequent risk monitoring and early warning. If the length of the second data sequence reaches this threshold, the system will issue a corresponding early warning message to remind the operator to pay attention and take necessary preventive measures.

[0076] Implementation effects and advantages: Improved warning accuracy: By pre-setting multiple risk levels and corresponding length thresholds, the system can more accurately assess the risk of decoupling and issue warnings in a timely manner when necessary. This helps reduce false positives and missed positives and improves the accuracy and reliability of warnings.

[0077] Enhanced flexibility: Different risk levels and length threshold settings allow the system to respond more flexibly to different levels of risk situations. This helps to achieve more accurate risk management and control in different scenarios.

[0078] Improve safety: Through real-time monitoring and early warning, the system can take timely measures to intervene and prevent uncoupling before it occurs. This helps reduce the incidence of uncoupling accidents and protect the lives of workers and the safety of facilities in the surrounding environment.

[0079] In summary, presetting multiple unhooking risk probability levels and setting corresponding length thresholds for each level is an effective risk management strategy. It can improve the accuracy of early warning, enhance the flexibility of the system, and significantly improve the safety of lifting operations.

[0080] In an embodiment of the present invention, based on the early warning generation module, a possible embodiment will be given below to illustrate its specific implementation scheme in a non-limiting manner.

[0081] Before subsequent processing, it is necessary to first obtain the time range of the first data sequence, which represents the time period during which the pressure value at the base of the hook is continuously collected.

[0082] Next, check whether the second data sequence is within this time range. If the second data sequence is not within the time range of the first data sequence, then the second data sequence can be considered invalid, which may be caused by data collection errors or equipment failures. At this time, the second data sequence should be ignored and no subsequent processing is performed.

[0083] If the second data sequence is within the time range of the first data sequence, then it is necessary to determine whether the length of the second data sequence has reached the length threshold previously generated based on the decoupling risk probability.

[0084] If the length of the second data sequence reaches the length threshold, it means that the anti-unhooking device has been under a force exceeding the safety range for a long time, and there is a high risk of unhooking. At this time, an unhooking warning prompt message should be generated, and a lifting height restriction instruction should be sent to the actuator to reduce the possibility of unhooking accidents.

[0085] If the length of the second data sequence does not reach the length threshold, it means that although the anti-unhooking device is stressed beyond the safety range, the duration is short and the unhooking risk is relatively low. At this time, the current unhooking risk probability can be written into the log file for subsequent data analysis and evaluation.

[0086] In addition, in order to increase the speed of data processing, edge collaborative computing is introduced, including: Get system status information: First, the system needs to obtain the current remaining computing resources, including CPU usage, memory usage, GPU usage (if available), etc. These indicators can reflect the current processing power of the system.

[0087] At the same time, the system also needs to obtain communication quality parameters with edge devices. These parameters may include network delay, packet loss rate, bandwidth, etc., which can reflect the stability and reliability of data transmission.

[0088] Judging computing resources and communication quality: Next, the system needs to determine whether the remaining computing resources do not exceed the set resource threshold. This resource threshold is set based on factors such as the system's processing capacity, business requirements, and expected response time.

[0089] At the same time, the system also needs to determine whether the communication quality parameters have reached the set quality threshold. This quality threshold is set based on factors such as data transmission stability, reliability, and business requirements.

[0090] Data sending decision: If the remaining computing resources do not exceed the set resource threshold and the communication quality parameter reaches the set quality threshold, the system can send the first data sequence (i.e., the pressure value sequence at the root of the crane hook) to the edge device.

[0091] Edge devices are usually deployed close to data sources, with lower network latency and higher data transmission rates. On edge devices, a pre-trained long short-term memory neural network model (LSTM) is pre-deployed, which can efficiently process time series data and predict decoupling risks.

[0092] Data processing and result return: Once the first data sequence is sent to the edge device, it will be processed by the LSTM model on the edge device. This processing may include steps such as data preprocessing, feature extraction, and model inference.

[0093] After processing is completed, the edge device will return the results (such as decoupling risk probability, warning information, etc.) to the original system. The original system can take corresponding measures based on these results, such as generating warning prompt information, sending lifting height restriction instructions, etc.

[0094] Advantages after expansion Improve processing efficiency: By sending data to edge devices with powerful computing capabilities for processing, the computing burden of the original system can be significantly reduced and data processing efficiency can be improved.

[0095] Optimize resource utilization: Making decisions based on the system's current computing resources and communication quality can ensure that resources are fully utilized and avoid unnecessary waste of resources.

[0096] Enhanced system flexibility: The system can flexibly select the data processing method (local processing or sending to edge device processing) according to actual conditions to adapt to different business needs and scenario changes.

[0097] Improve data transmission reliability: By ensuring that the communication quality reaches the set quality threshold before sending data, the packet loss rate and delay during data transmission can be reduced, and the reliability and stability of data transmission can be improved.

[0098] In summary, by expanding and optimizing the original process, the safety monitoring system for crane hoisting operations can be made more efficient, reliable and flexible.

[0099] Although the present invention has been described in detail with reference to the accompanying drawings and in combination with the preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the present invention, a person of ordinary skill in the art may make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions shall be within the scope of the present invention. Any person of ordinary skill in the art may easily think of changes or substitutions within the technical scope disclosed by the present invention, and these shall be within the scope of protection of the present invention.

Claims

1. A crane unhooking warning system, characterized in that: include: A data collection terminal is used to collect the working status information of the crane hook, wherein the working status information includes the load status of the hook, the position of the anti-drop baffle plate, and the status of the locking spring; The data acquisition end includes a plurality of sensors, and the plurality of sensors are used to measure the pressure of the locking spring; The data edge computing end obtains the pressure of the locking spring and sets the reference pressure range according to the working status of the hook anti-dropping device. If the pressure value of the locking spring exceeds the pressure range, an alarm message is generated.

2. The system according to claim 1, characterized in that The anti-unhooking device comprises: A spring, a protective pin, an anti-slip baffle, a data acquisition terminal, and a protective cover. The fixed end of the anti-slip baffle is rotatably connected to the hook body through the protective pin, and the suspended end of the anti-slip baffle is clamped on the inner side of the end of the hook; the spring is arranged on the protective pin to push the suspended end of the anti-slip baffle to the inner side of the end of the hook; the data acquisition terminal is installed on the side of the anti-slip baffle facing the hook, and the protective cover is laid on the surface of the data acquisition terminal.

3. The system according to claim 2, characterized in that The data acquisition end includes a patch pressure sensor, a controller and a wireless communication chip. The detection patch of the patch pressure sensor is attached to the surface of the protective cover and the force-bearing surface of the anti-detachment baffle. The patch pressure sensor and the wireless communication chip are both electrically connected to the controller.

4. The system according to claim 1, characterized in that The data acquisition end uses the modbus standard protocol to communicate wirelessly with the data edge computing end, and supports a semi-open ad hoc networking mode.

5. The system according to claim 3, characterized in that The data edge computing end includes: The first acquisition module is used to continuously acquire pressure values ​​at the base of the crane hook, and arrange the acquired pressure values ​​into a first data sequence according to the acquisition time; A second acquisition module is used to continuously acquire pressure values ​​of the force-bearing surface of the anti-unhooking device of the crane hook, compare the acquired pressure values ​​with a preset pressure threshold, and arrange the continuous pressure values ​​of the force-bearing surface exceeding the pressure threshold in chronological order of acquisition time into a second data sequence; A risk identification module, used for inputting the first data sequence into a pre-trained long short-term memory neural network model to identify the decoupling risk probability; A length setting module, used for generating a length threshold of a second data sequence based on the decoupling risk probability; The warning generation module is used to confirm that the length of the second data sequence generated synchronously reaches the length threshold and generate a decoupling warning prompt message.

6. The system according to claim 5, characterized in that Continuously collecting pressure values ​​of the force-bearing surface of the anti-unhooking device of the crane hook, comparing the collected pressure values ​​with a preset pressure threshold, and arranging the continuous pressure values ​​of the force-bearing surface exceeding the pressure threshold in chronological order of collection time into a second data sequence, including: Determine whether the current pressure value exceeding the pressure threshold and the last pressure value of the current second data sequence are continuous pressure values: If so, adding the current pressure value exceeding the pressure threshold to the current second data sequence; If not, the current second data sequence is deleted, and a new second data sequence is created, and the current pressure value exceeding the pressure threshold is used as the starting value of the new second data sequence.

7. The system according to claim 5, characterized in that Generating a length threshold of a second data sequence based on the decoupling risk probability includes: Pre-set multiple decoupling risk probability levels, and set the length threshold corresponding to each decoupling risk probability level; A target decoupling risk probability level is matched to the decoupling risk probability, and a length threshold corresponding to the target decoupling risk probability level is set as a length threshold of the second data sequence.

8. The system according to claim 5, characterized in that Confirming that the length of the synchronously generated second data sequence reaches the length threshold, and generating a decoupling warning prompt message, including: Obtaining a time range of the first data sequence, and verifying whether the second data sequence is within the time range, and if the second data sequence is not within the time range, determining that the second data sequence is invalid; Confirm that the second data sequence is within the time range, and determine whether the length of the second data sequence reaches the length threshold: If yes, a decoupling warning message is generated, and a lifting height restriction instruction is sent to the execution agency; If not, the decoupling risk probability is written into the log file.

9. The system according to claim 8, characterized in that Also includes: Obtain remaining computing resources and communication quality parameters with edge devices; If the remaining computing resources do not exceed the set resource threshold and the communication quality parameter reaches the set quality threshold, the first data sequence is sent to the edge device so that the edge device processes the first data sequence using a pre-deployed pre-trained long short-term memory neural network model.

Images