A crane hook - detachment warning system

Abstract

The present invention relates to the technical field of data processing, and specifically provides a crane hook detachment warning system, including: a data acquisition end for acquiring the working state information of the crane hook, where the working state information includes the load state of the hook, the position of the anti-detachment baffle, and the state of the locking spring; the data acquisition end includes a plurality of sensors for measuring the pressure of the locking spring; the data edge computing end obtains the pressure of the locking spring, sets a reference pressure range according to the working state of the hook anti-detachment device, and generates an alarm message if the pressure value of the locking spring exceeds the pressure range. The present invention can effectively avoid the fracture risk caused by the long-term working state of the anti-hook detachment device.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and particularly relates to a crane hook - off warning system. Background Art

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

[0003] The core component of the hook anti - detachment device is the anti - detachment baffle, which can quickly play a role at the moment when the heavy object gets unhooked and effectively prevent the heavy object from falling. However, it should be noted that during long - term use and frequent stress, the anti - detachment baffle will gradually wear, resulting in a decline in its anti - detachment performance. Therefore, it is very necessary to avoid the anti - detachment baffle being in a stressed state for a long time.

[0004] In view of the above situation, the real - time monitoring and warning of the crane hook - off risk are particularly important. Summary of the Invention

[0005] Aiming at the above - mentioned deficiencies of the prior art, the present invention provides a crane hook - off warning system to solve the above - mentioned technical problems.

[0006] In a first aspect, the present invention provides a crane hook - off warning system, including:

[0007] A data acquisition end, which is used to acquire the working state information of the crane hook. The working state information includes the load state of the hook, the position of the anti - detachment baffle, and the state of the locking spring;

[0008] The data acquisition end includes multiple sensors, and the multiple sensors are used to measure the pressure of the locking spring;

[0009] The data edge computing end obtains the pressure of the locking spring, sets a reference pressure range according to the working state of the hook anti - detachment device, and generates an alarm message if the pressure value of the locking spring exceeds the pressure range.

[0010] In an optional embodiment, the anti - detachment hook device includes:

[0011] A spring, a protection pin, a anti - detachment baffle, a data acquisition end, and a protective cover. The fixed end of the anti - detachment baffle is rotatably connected to the hook body through the protection pin, and the suspended end of the anti - detachment baffle is stuck inside the end of the hook; the spring is arranged on the protection pin and is used to push the suspended end of the anti - detachment baffle towards the inside of the end of the hook; the data acquisition end is installed on the side of the anti - detachment baffle facing the hook, and the protective cover is laid on the surface of the data acquisition end.

[0012] In an alternative embodiment, the data acquisition end includes a patch - type pressure sensor, a controller, and a wireless communication chip. The detection patch of the patch - type pressure sensor is attached to the surface of the protective cover and the force - receiving surface of the anti - detachment baffle; both the patch - type pressure sensor and the wireless communication chip are electrically connected to the controller.

[0013] In an alternative embodiment, the data acquisition end communicates wirelessly with the data edge computing end using the modbus standard protocol and supports a semi - open self - organizing network mode.

[0014] In an alternative embodiment, the data edge computing end includes:

[0015] A first acquisition module, which is used to continuously acquire the pressure values at the root of the crane hook and arrange the acquired pressure values in a first data sequence in the order of acquisition time;

[0016] A second acquisition module, which is used to continuously acquire the pressure values on the force - receiving surface of the anti - detachment device of the crane hook, compare the acquired pressure values with a preset pressure threshold, and arrange the consecutive pressure values on the force - receiving surface that exceed the pressure threshold in a second data sequence in the order of acquisition time;

[0017] A risk identification module, which is used to input the first data sequence into a pre - trained long - short - term memory neural network model to identify the probability of unhooking risk;

[0018] A length setting module, which is used to generate a length threshold for the second data sequence based on the unhooking risk probability;

[0019] An early warning generation module, which is used to confirm that the length of the synchronously generated second data sequence reaches the length threshold and generate an unhooking early warning prompt message.

[0020] In an alternative embodiment, continuously acquiring the pressure values on the force - receiving surface of the anti - detachment device of the crane hook, comparing the acquired pressure values with a preset pressure threshold, and arranging the consecutive pressure values on the force - receiving surface that exceed the pressure threshold in a second data sequence in the order of acquisition time includes:

[0021] Judging whether the current pressure value exceeding the pressure threshold and the last pressure value of the current second data sequence are consecutive pressure values:

[0022] If so, add the current pressure value exceeding the pressure threshold to the current second data sequence;

[0023] If not, delete the current second data sequence, create a new second data sequence, and use the current pressure value exceeding the pressure threshold as the starting value of the new second data sequence.

[0024] In an alternative embodiment, generating a length threshold for the second data sequence based on the decoupling risk probability includes:

[0025] Preset multiple decoupling risk probability levels and set the corresponding length threshold for each decoupling risk probability level;

[0026] Match the decoupling risk probability to a target decoupling risk probability level and set the length threshold corresponding to the target decoupling risk probability level as the length threshold of the second data sequence.

[0027] In an alternative embodiment, when it is confirmed that the length of the synchronously generated second data sequence reaches the length threshold, generating a decoupling warning prompt message includes:

[0028] Obtain the time range of the first data sequence and verify whether the second data sequence is within the time range. If the second data sequence is not within the time range, it is determined that the second data sequence is invalid;

[0029] 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:

[0030] If so, generate a decoupling warning prompt message and send a lifting height limit instruction to the actuator;

[0031] If not, write the decoupling risk probability into the log file.

[0032] In an alternative embodiment, it further includes:

[0033] Obtain the remaining computing resources and the communication quality parameters with the edge device;

[0034] If the remaining computing resources do not exceed the set resource threshold and the communication quality parameters reach the set quality threshold, send the first data sequence 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.

[0035] The beneficial effect of the present invention is that the crane decoupling warning system provided by the present invention can accurately detect the working state of the anti-decoupling device, and then generate a warning message, thereby effectively avoiding the fracture risk caused by the anti-decoupling device being in a working state for a long time.

[0036] In addition, the design principle of the present invention is reliable, the structure is simple, and it has a very broad application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0038] Figure 1 It is a schematic structural diagram of an anti-disengagement device according to an embodiment of the present invention.

[0039] Figure 2 It is an internal structural diagram of an anti-disengagement device according to an embodiment of the present invention.

[0040] Figure 3 It is a bottom schematic diagram of an anti-disengagement device according to an embodiment of the present invention.

[0041] Figure 4 It is a schematic diagram of an application scenario of an anti-disengagement device according to an embodiment of the present invention.

[0042] In the figure, 1. Spring, 2. Protection pin, 3. Anti-disengagement baffle, 4. Data acquisition end, 5. Protective cover, 6. Hook, 401. Signal generator, 402. Circuit board, 403. Lithium battery pack, 404. Charging interface, 405. Device switch, 406. Pressure acquisition end, 407. Strong magnetic adhesion block, 7. Portable alarm, 8. Data edge computing end. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of the present invention. The terms used in the description of the present invention herein are only for the purpose of describing specific embodiments, and are not intended to limit the present invention.

[0045] Please refer to Figure 1 , this embodiment provides a crane hook disengagement warning system, including:

[0046] A data acquisition terminal, which is used to acquire the working state information of the crane hook, and the working state information includes the load state of the hook, the position of the anti-drop baffle, and the state of the locking spring;

[0047] The data acquisition terminal includes a plurality of sensors, and the plurality of sensors are used to measure the pressure of the locking spring;

[0048] The data edge computing terminal obtains the pressure of the locking spring, sets a reference pressure range according to the working state of the hook anti-drop device, and generates an alarm message if the pressure value of the locking spring exceeds the pressure range.

[0049] The crane hook monitoring device is as Figures 2 - 3 shown, and is composed of a spring 1, a protection pin 2, an anti-drop 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, and a pressure acquisition terminal 406, a strong magnetic adhesive block 407. It is characterized in that: the anti-drop baffle 3 and the spring 1 are fixed on the hook 6 by the protection 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 front side of the bottom of the data acquisition terminal 4, the data acquisition terminal 4 is fixed inside the anti-drop baffle 3 by the strong magnetic adhesive block 407, the signal generator 401 is connected to the circuit board 402 and fixed on the front part of the data acquisition terminal 4, the device switch 405 is connected to the circuit board 402 and fixed on the right rear side of the data acquisition terminal 4, the charging interface 404 is connected to the lithium battery pack 403 and fixed on 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.

[0050] When the device is installed, it is fixed inside the anti-drop baffle by a sliding installation method using a strong magnetic adhesive block, without changing the original operating procedures of the hook.

[0051] The monitor is divided into two parts: an acquisition terminal and a data calculation and processing terminal. The data calculation and processing terminal is further divided into two types: a data edge computing terminal and a portable alarm terminal to adapt to different application scenarios. The data edge computing terminal has functions such as initializing the acquisition terminal, calculating and processing the backhaul data, processing alarm information, monitoring the battery power of the acquisition terminal, and sound and light alarm, and has a serial interface and a digital switch interface for convenient integration with the upper-level system. The portable alarm terminal has functions such as initializing the acquisition terminal, processing alarm information, and sound and light alarm.

[0052] The design of the acquisition end and the data calculation and processing end is separated, and the two types of the data calculation and processing end have stronger applicability; the acquisition end and the anti-drop baffle are fixed by strong magnetic blocks, taking into account the convenience during replacement and charging and the reliability during operation; the acquisition end is powered by a battery and communicates with the calculation end through a wireless communication method, avoiding the influence of wire routing on the operation of the lifting hook; by measuring the spring force, it is judged whether the anti-drop baffle is in the correct position, reducing the dependence of the acquisition end on the structure of the lifting hook; it has the function of autonomously judging whether the lifting hook is working; it has the function of adapting to the position of the anti-drop baffle, and can autonomously correct the judgment basis according to the actual working data, improving the accuracy of judgment and the adaptability to different operators; the acquisition end adopts a low-power design throughout the process, optimizing the effective energy consumption of the battery; according to the working state of the lifting hook, the data transmission frequency is controlled to achieve the best balance between power consumption and monitoring timeliness; the acquisition end adopts a sliding-in installation method according to the structure of the anti-drop baffle and is completely embedded in the anti-drop baffle without changing the original operating procedures of the lifting 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 simultaneously with the data edge calculation end, or the portable alarm terminal can be used alone to receive alarm prompts.

[0053] In another embodiment of the present invention, the device structure includes:

[0054] Spring 1, protection pin 2, anti-drop baffle 3, data acquisition end 4, cover 5. The fixed end of the anti-drop baffle 3 is rotatably connected to the lifting hook 6 through the protection pin 2, and the suspended end of the anti-drop baffle 3 is stuck inside the end of the lifting hook; the spring 1 is arranged on the protection pin 2 and is used to push the suspended end of the anti-drop baffle 3 towards the inside of the end of the lifting hook; the data acquisition end 4 is installed on the side of the anti-drop baffle 3 facing the lifting hook, and the cover 5 is laid on the surface of the data acquisition end 4; the data acquisition end 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 cover 5 and the force-receiving surface of the anti-drop baffle 3; both the patch type pressure sensor and the wireless communication chip are electrically connected to the controller. Its application state is as Figure 4 shown. When the device is installed, it is fixed inside the anti-drop baffle by a sliding-in installation method using strong magnetic blocks without changing the original operating procedures of the lifting hook.

[0055] The data edge calculation end includes:

[0056] The first acquisition module is used to continuously acquire the pressure values at the root of the crane hook and arrange the acquired pressure values in a first data sequence according to the acquisition time sequence;

[0057] The second acquisition module is used to continuously acquire the pressure value of the force-bearing surface of the anti-unhooking device of the crane hook, compare the acquired pressure value with a preset pressure threshold, and arrange the consecutive pressure values of the force-bearing surface that exceed the pressure threshold in the order of acquisition time as the second data sequence;

[0058] The risk identification module is used to input the first data sequence into a pre-trained long short-term memory neural network model to identify the probability of unhooking risk;

[0059] The length setting module is used to generate a length threshold for the second data sequence based on the unhooking risk probability;

[0060] The warning generation module is used to confirm that the length of the synchronously generated second data sequence reaches the length threshold and generate an unhooking warning prompt message.

[0061] In an embodiment of the present invention, based on the first acquisition module, a possible embodiment will be given below to non-restrictively elaborate on its specific implementation scheme.

[0062] Continuously acquire the pressure value at the root of the crane hook. These pressure values can reflect the force condition of the hook during the hoisting process. According to the order of acquisition time, these pressure values are arranged into the first data sequence. This step is the basis for understanding the working state of the hook and helps to master the dynamic changes of the hook during the hoisting operation.

[0063] In an embodiment of the present invention, based on the second acquisition module, a possible embodiment will be given below to non-restrictively elaborate on its specific implementation scheme.

[0064] The following is a specific example of generating the second data sequence:

[0065] Suppose there is an anti-unhooking device for a crane hook, and the pressure value of its force-bearing surface is continuously acquired. The preset pressure threshold is set to 100 unit pressures. Assume that time is sampled in seconds and a pressure value is acquired every second.

[0066] Initial state:

[0067] Before generating the second data sequence, assume that no pressure value exceeding the pressure threshold is recorded, so the second data sequence is empty.

[0068] Pressure value acquisition and comparison:

[0069] Time t = 1 second: The acquired pressure value is 90 units, which does not exceed the pressure threshold, and no operation is performed.

[0070] Time t = 2 seconds: The collected pressure value is 110 units, exceeding the pressure threshold. At this time, check if there is a previous pressure value that exceeded the threshold (i.e., check if 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.

[0071] Current second data sequence: {110}

[0072] Time t = 3 seconds: The collected pressure value is 120 units, exceeding the pressure threshold. Check if the currently exceeded pressure value (120 units) and the last pressure value in the current second data sequence (110 units) are consecutive pressure values. Since they are consecutive (i.e., 120 units follows 110 units immediately, and both exceed the threshold), 120 units is added to the current second data sequence.

[0073] Current second data sequence: {110, 120}

[0074] Time t = 4 seconds: The collected pressure value is 95 units, not exceeding the pressure threshold, and no operation is performed.

[0075] Time t = 5 seconds: The collected pressure value is 130 units, exceeding the pressure threshold. Check again if the currently exceeded pressure value (130 units) and the last pressure value in the current second data sequence (120 units) are consecutive pressure values. Since they are consecutive, 130 units is added to the current second data sequence.

[0076] Current second data sequence: {110, 120, 130}

[0077] Time t = 6 seconds: The collected pressure value is 80 units, not exceeding the pressure threshold, and no operation is performed.

[0078] Time t = 7 seconds: The collected pressure value is 20 units, far lower than the pressure threshold and not consecutive with the previous pressure value that exceeded the threshold (130 units). At this time, according to the rule, 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 that exceeds the threshold at this time, a new second data sequence will not actually be created immediately, but will wait for the next pressure value that exceeds the threshold to appear.

[0079] In practical applications, if there is no pressure value that exceeds the threshold for a long time, set a timeout mechanism to empty or reset the second data sequence to avoid unnecessary resource occupation.

[0080] Time t = 8 seconds: The collected pressure value is 140 units, exceeding 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 second data sequence is newly created, and 140 units is used as the starting value.

[0081] Current second data sequence: {140}.

[0082] In an embodiment of the present invention, based on the risk identification module, a possible embodiment will be given below to non - restrictively elaborate on its specific implementation scheme.

[0083] The first data sequence is input into a 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 becoming unhooked during the hoisting process. Through the calculation of the model, a probability value regarding the unhooking risk can be obtained.

[0084] Before discussing the training process of the LSTM (long short - term memory) model for the crane unhooking risk, it is first necessary to train it. Collect the pressure value sequence at the root of the hook before the crane unhooks and the pressure value sequence at the root of the hook under normal conditions. These data are the basis for model training, representing two completely different states: one is the abnormal state where the hook is about to become unhooked, and the other is the normal operation state (although affected by natural factors such as wind resistance).

[0085] The training process of the LSTM model can be roughly divided into the following steps:

[0086] Data pre - processing:

[0087] Data cleaning: Remove noise data, such as outliers or invalid data generated due to equipment failures.

[0088] Feature extraction: Extract features useful for predicting the unhooking risk from the original pressure value sequence. This may include statistics such as the average value, maximum value, minimum value, standard deviation of the pressure value, and certain transformations of the time series (such as differencing, moving average, etc.).

[0089] Labeling: Assign a label to each piece of data indicating whether an unhooking event has occurred. This is usually a binary classification problem, where 1 represents unhooking and 0 represents normal.

[0090] Data partitioning: Divide the data set into a training set, a validation set, and a test set for model evaluation and tuning during the training process.

[0091] Model construction:

[0092] Select LSTM as the model architecture because it performs well in processing time - series data.

[0093] 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 according to the scale and complexity of the data.

[0094] Set the loss function (such as cross-entropy loss) and the optimizer (such as Adam optimizer) to guide the training process of the model.

[0095] Model training:

[0096] Input the training set into the LSTM model, and calculate the difference (i.e., loss) between the predicted value and the actual value through forward propagation.

[0097] Use the backpropagation algorithm to calculate the gradients, and update the weights and bias terms of the model through the optimizer to minimize the loss.

[0098] Repeat this process until the performance of the model on the validation set reaches an acceptable level or the preset number of iterations is reached.

[0099] Model evaluation:

[0100] Evaluate the performance of the model using the test set, including metrics such as accuracy, recall, and F1-score.

[0101] Use visualization tools (such as confusion matrices, ROC curves, etc.) to further analyze the performance of the model.

[0102] Model tuning:

[0103] According to the evaluation results, adjust the architecture, hyperparameters, or data preprocessing steps of the model to improve the performance of the model.

[0104] Repeat the processes of training, evaluation, and tuning until the model reaches the best performance.

[0105] Deployment and application:

[0106] Deploy the trained model to actual applications for real-time monitoring of the risk of hook detachment of the crane.

[0107] According to the prediction results of the model, take corresponding preventive measures to avoid the occurrence of hook detachment events.

[0108] In an embodiment of the present invention, based on the length setting module, a possible embodiment will be given below to non-restrictively elaborate on its specific implementation scheme.

[0109] Based on this decoupling risk probability value, the system generates a length threshold for the second data sequence corresponding to the risk probability. This length threshold represents how long the pressure state where the force-bearing surface of the anti-decoupling device exceeds the pressure threshold needs to last before it is considered to have a decoupling risk.

[0110] Risk level classification:

[0111] According to historical data, expert experience, and safety standards, the decoupling risk probability can be divided into multiple levels, such as low risk, medium risk, high risk, and extremely high risk, etc.

[0112] Each level corresponds to a specific risk probability range. For example, low risk may correspond to a probability range of 0% - 20%, medium risk corresponds to 21% - 50%, high risk corresponds to 51% - 80%, and extremely high risk corresponds to 81% - 100%.

[0113] Length threshold setting:

[0114] For each risk level, a corresponding length threshold needs to be set. This threshold indicates how long the pressure state where the force-bearing surface of the anti-decoupling device exceeds the pressure threshold needs to last before the system considers it to be a risk of this level.

[0115] The setting of the length threshold should be appropriately adjusted based on the level of the risk. For example, for the low-risk level, the length threshold can be set relatively long because even if the pressure state lasts for a while, the risk of decoupling is relatively low. For high-risk or extremely high-risk levels, the length threshold should be set shorter so that the system can respond quickly and issue a warning.

[0116] Risk probability calculation and matching:

[0117] During the real-time monitoring process, the system continuously collects the pressure values of the crane hook and its anti-decoupling device, and calculates the current decoupling risk probability based on this data.

[0118] The calculated risk probability is matched with the pre-set risk levels to determine the current risk level.

[0119] Length threshold setting:

[0120] Once the current risk level is determined, the system sets the length threshold of the second data sequence according to the length threshold corresponding to this level.

[0121] This length threshold will be used for subsequent risk monitoring and warning. If the length of the second data sequence reaches this threshold, the system will issue a corresponding warning message to remind the operator to pay attention and take necessary preventive measures.

[0122] Implementation effects and advantages:

[0123] Improve warning accuracy: By pre-setting multiple risk levels and corresponding length thresholds, the system can more accurately assess the decoupling risk and issue warnings in a timely manner when necessary. This helps to reduce false alarms and missed alarms, improving the accuracy and reliability of the warnings.

[0124] Enhance flexibility: Different risk level and length threshold settings enable the system to more flexibly respond to different levels of risk situations. This helps to achieve more precise risk management and control in different scenarios.

[0125] Enhance safety: Through real-time monitoring and warning, the system can take measures to intervene and prevent in a timely manner before a decoupling event occurs. This helps to reduce the incidence of decoupling accidents and protect the lives of staff and the safety of facilities in the surrounding environment.

[0126] In summary, pre-setting multiple decoupling risk probability levels and setting corresponding length thresholds for each level is an effective risk management strategy. It can improve the accuracy of warnings, enhance the flexibility of the system, and significantly enhance the safety of lifting operations.

[0127] In an embodiment of the present invention, based on the warning generation module, a possible embodiment will be given below to non-restrictively elaborate on its specific implementation scheme.

[0128] Before proceeding with subsequent processing, it is first necessary to obtain the time range of the first data sequence. This time range represents the period during which the hook root pressure value is continuously collected.

[0129] Next, verify 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 it can be considered that this second data sequence is invalid, possibly due to data collection errors or equipment failures. In this case, this second data sequence should be ignored and no subsequent processing should be performed.

[0130] 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 this second data sequence has reached the length threshold generated based on the decoupling risk probability.

[0131] If the length of the second data sequence reaches the length threshold, it means that the anti-decoupling device has been in a state where the force exceeds the safe range for a long time, and there is a high risk of decoupling. At this time, a decoupling warning prompt message should be generated and a height limit instruction should be sent to the actuator to reduce the likelihood of a decoupling accident.

[0132] If the length of the second data sequence does not reach the length threshold, it indicates that although the force on the anti-disengagement device exceeds the safe range, the duration is short and the risk of disengagement is relatively low. At this time, the current risk probability of disengagement can be written into the log file for subsequent data analysis and evaluation.

[0133] In addition, to improve the speed of data processing, edge collaborative computing is introduced, including:

[0134] Obtain system status information:

[0135] First of all, the system needs to obtain the current remaining computing resources. This includes CPU usage, memory occupancy, the usage of GPU (if available), etc. These metrics can reflect the current processing capacity of the system.

[0136] At the same time, the system also needs to obtain the communication quality parameters with the edge device. These parameters may include network latency, packet loss rate, bandwidth, etc. They can reflect the stability and reliability of data transmission.

[0137] Judge computing resources and communication quality:

[0138] Next, the system needs to judge whether the remaining computing resources do not exceed the set resource threshold. This resource threshold is set according to factors such as the system's processing capacity, business requirements, and expected response time.

[0139] At the same time, the system also needs to judge whether the communication quality parameters reach the set quality threshold. This quality threshold is set according to factors such as the stability and reliability of data transmission and business requirements.

[0140] Data sending decision:

[0141] If the remaining computing resources do not exceed the set resource threshold and the communication quality parameters reach the set quality threshold, then 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.

[0142] Edge devices are usually deployed close to the data source, with low network latency and high data transmission rates. On the edge device, a pre-trained long short-term memory neural network model (LSTM) is pre-deployed, which can efficiently process time series data and predict the risk of disengagement.

[0143] Data processing and result return:

[0144] Once the first data sequence is sent to the edge device, the LSTM model on the edge device will process it. This processing process may include steps such as data preprocessing, feature extraction, and model inference.

[0145] After the processing is completed, the edge device will return the results (such as the probability of decoupling risk, 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 instructions to limit the lifting height, etc.

[0146] Expanded advantages

[0147] Improve processing efficiency:

[0148] By sending data to an edge device with powerful computing capabilities for processing, the computing burden on the original system can be significantly reduced, and the data processing efficiency can be improved.

[0149] Optimize resource utilization:

[0150] Making decisions based on the current computing resources and communication quality of the system can ensure that resources are fully utilized and avoid unnecessary resource waste.

[0151] Enhance system flexibility:

[0152] The system can flexibly select data processing methods (local processing or sending to the edge device for processing) according to the actual situation to adapt to different business requirements and scenario changes.

[0153] Improve data transmission reliability:

[0154] 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.

[0155] In summary, by expanding and optimizing the original process, the safety monitoring system for crane lifting operations can be made more efficient, reliable, and flexible.

[0156] Although the present invention has been described in detail by referring 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, those of ordinary skill in the art can make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions should all be within the scope of the present invention. / Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention.

Claims

1. A crane unhooking warning system, characterized in that: include: A data collection terminal is used to collect working status information of the anti-unhooking device, wherein the working status information includes the load status of the hook, the position of the anti-unhooking 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, sets a reference pressure range according to the working status information of the anti-unhooking device, and generates an alarm message if the pressure value of the locking spring exceeds the pressure range; The anti-unhooking device comprises: a locking spring, a protection pin, an anti-unhooking baffle, a data acquisition terminal, and a protective cover. The fixed end of the anti-unhooking baffle is rotatably connected to the hook body through the protection pin, and the suspended end of the anti-unhooking baffle is clamped on the inner side of the end of the hook; the locking spring is arranged on the protection pin, and is used to push the suspended end of the anti-unhooking baffle to the inner side of the end of the hook; the data acquisition terminal is installed on the side of the anti-unhooking baffle facing the hook, and the protective cover is laid on the surface of the data acquisition terminal; The data acquisition end also 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-slip baffle. The patch pressure sensor and the wireless communication chip are both electrically connected to the controller. The data edge computing end also 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; An early 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 early warning prompt information; 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; 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.

2. 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.

3. The system according to claim 1, 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.

4. The system according to claim 3, 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