A method and system for identifying transient instability of a hoisting device and adaptive damping control

Abstract

The application discloses a kind of hoisting equipment transient instability identification and adaptive damping control method and system, it is related to hoisting equipment intelligent control technical field, including the following steps: multidimensional acquisition equipment inclination, vibration, load and so on Data and pre-processing;Extract core feature parameter and eliminate redundant information;Risk of transient instability is identified and grade is determined by neural network model;According to risk grade and working condition, calculate optimal damping control parameter;Drive damping actuator to suppress instability trend;Real-time feedback optimization model and control parameter;Data is transmitted to remote platform, linkage braking system is executed in emergency shutdown and early warning is pushed.The application is accurately captured by multidimensional data fusion and intelligent identification, transient instability signs;Adaptive damping control realizes active prevention and control, closed-loop optimization improves control precision, remote monitoring and emergency linkage guarantee safety, adapt to multi-scene hoisting equipment, greatly reduce instability accident risk.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for lifting equipment, and in particular to a method and system for identifying transient instability and adaptive damping control of lifting equipment. Background Technology

[0002] As core equipment in the engineering construction field, lifting machinery is widely used in buildings, bridges, ports, and other scenarios. The stability of its steel structure directly determines construction safety and project progress. During operation, lifting machinery is susceptible to transient instability symptoms such as tilting, vibration, and sudden displacement due to multiple factors including fluctuations in lifting loads, environmental wind interference, differences in track flatness, and equipment aging and wear. If these symptoms are not identified and controlled in time, they can quickly escalate into major safety accidents such as overturning and boom breakage, causing serious casualties and economic losses. Therefore, achieving accurate identification and rapid control of transient instability in lifting machinery has become a key technical requirement for ensuring construction safety.

[0003] While existing monitoring and control solutions have made some progress, numerous technical bottlenecks remain. Patent application CN120630821A, titled "Remote Monitoring System for Instability of Steel Structure in Lifting Machinery," mentions collecting tilt angle data through a tilt monitoring module, combining it with an environmental compensation module and a control module to achieve multi-level alarms and instability prediction. It employs dual-axis / tri-axis tilt sensors and Kalman filtering to improve monitoring accuracy and supports 5G+VR remote monitoring. However, this technology focuses on instability monitoring and early warning, lacking an adaptive damping control mechanism for transient instability. It only addresses risks through action restrictions and emergency braking, making it difficult to suppress instability trends through damping adjustment in the early stages. Furthermore, the control strategy lacks dynamic adaptation to multi-dimensional characteristic parameters, resulting in insufficient flexibility in dealing with complex transient instability.

[0004] The patent application document with publication number CN101537981B and title "Online Monitoring and Early Warning System and Method for Instability of Tower Cranes Based on Ultrasonic Sensor Network" mentions using an ultrasonic sensor network to monitor the swaying and torsion of the tower crane's upper structure. The main controller module performs instability evaluation and notifies the operator, offering strong anti-interference capabilities and good real-time performance. However, this technology relies on ultrasonic ranging for displacement monitoring, resulting in a single monitoring dimension. It does not integrate key parameters such as load and vibration, and can only provide passive early warning. Lacking an active control actuator, it cannot automatically suppress instability, still requiring manual intervention from the operator. This leads to a slow response time and difficulty in handling millisecond-level transient instability risks.

[0005] In summary, existing technologies suffer from three major limitations: First, the monitoring dimensions are limited, focusing primarily on monitoring single physical quantities and failing to achieve deep integration of multi-dimensional data such as tilt, vibration, load, and displacement, resulting in insufficient accuracy in identifying transient instability. Second, the control methods are passive, mainly relying on alarms and emergency braking, lacking active damping adjustment mechanisms and unable to effectively intervene in the early stages of instability. Third, adaptability and real-time performance are insufficient; control parameters are fixed and not dynamically optimized based on instability risk levels and working conditions, and there are delays in data processing and control response. These problems lead to delayed identification and inefficient control of transient instability in lifting machinery, making it difficult to meet the safety management requirements of complex construction scenarios. Therefore, a technical solution integrating multi-dimensional monitoring, accurate identification, and adaptive damping control is urgently needed to overcome the bottlenecks of existing technologies. Summary of the Invention

[0006] This invention proposes a method and system for transient instability identification and adaptive damping control of lifting equipment to solve the problems mentioned in the prior art.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a method for transient instability identification and adaptive damping control of lifting equipment, comprising the following steps:

[0008] Multi-dimensional transient data acquisition involves deploying tilt, vibration, wind speed, load, and displacement sensors to collect data on the crane's attitude, vibration, environment, load, and structural displacement. After filtering and noise reduction, the data is transmitted to the processing unit.

[0009] Instability feature extraction and preprocessing: Time domain and frequency domain analysis of the collected multi-dimensional data is performed to extract equipment state feature parameters. Redundant data is removed by normalization and correlation analysis, and core feature parameters are retained.

[0010] The transient instability identification model is constructed and trained. The model is built based on a long short-term memory neural network. The core feature parameter sequence is input, and training / test sets are constructed by combining historical, normal and simulated instability data. The number of network layers, hidden units and learning rate are optimized. Cross-validation is used to improve the generalization ability. The model outputs the probability value of instability risk.

[0011] The instability risk level is determined by classifying safety level, warning level and emergency level according to the risk probability, and dynamically adjusting the threshold in combination with the real-time parameter change trend. When the risk escalates, the corresponding response is triggered.

[0012] The adaptive damping control parameters are calculated based on the instability risk level, core characteristic parameters, and equipment operating conditions. The optimal damping force is calculated, and the optimal adjustment rate and range of action are calculated by combining the risk level gain and the physical limit constraint of the actuator.

[0013] The damping actuator drive control converts the calculated damping control parameters into control commands, which are then transmitted to the hydraulic damper or electromagnetic damper via the CAN bus to drive the damping actuator to adjust the damping state in real time.

[0014] The control effect is fed back and optimized in real time. Multi-dimensional data after control is collected in real time by sensors to evaluate the damping control effect. The model and control parameters are iteratively optimized based on the feedback data to improve control accuracy and response speed.

[0015] Remote monitoring and emergency response linkage transmit instability risk levels, control commands, and equipment operating status to a remote platform in real time. The platform supports data visualization, historical tracing, and abnormal alarms. In case of emergency instability, the linkage braking system stops the machine and sends an early warning.

[0016] Furthermore, it also includes a transient instability trend prediction step, which calculates the instability development rate using a formula, the specific formula being:

[0017]

[0018] in, For the rate of unstable development, This represents the change in the probability of instability risk per unit time. For time intervals, This refers to the comprehensive change of core feature parameters per unit time. This represents the change in equipment displacement per unit time. For risk probability weights, For feature parameter weights, As the displacement change weight, and .

[0019] Furthermore, it also includes an environmental interference compensation step, in which environmental sensors collect wind speed, temperature, humidity and ground vibration data in real time, construct an interference compensation model through a multivariate linear regression algorithm, generate compensation coefficients, and correct the collected core data.

[0020] Furthermore, in the multi-dimensional transient data acquisition step, the tilt sensor has a measurement range of -30° to 30° with an accuracy of ±0.05°; the vibration sensor uses a triaxial accelerometer with a measurement range of ±10g and a resolution of 0.001g; the wind speed sensor has a measurement range of 0-60m / s with an accuracy of ±0.1m / s; the load sensor has a measurement range of 0-500t with an accuracy of ±0.5%FS; and the displacement sensor uses a laser displacement sensor with a measurement range of 0-500mm and an accuracy of ±0.01mm.

[0021] Furthermore, in the instability feature extraction and preprocessing steps, the time-domain analysis uses the sliding window method to extract feature parameters, with the window size set to 50 sampling points; the frequency-domain analysis uses the fast Fourier transform to convert the time-domain signal into a frequency-domain signal, extracting feature frequencies within the 10Hz-100Hz frequency band; the normalization process uses the max-min normalization method to map the data to the 0-1 interval; and the correlation analysis uses the Pearson correlation coefficient method to remove feature parameters with an absolute correlation coefficient value lower than 0.3.

[0022] Furthermore, in the adaptive damping control parameter calculation step, the damping force adjustment accuracy is optimized using a formula, specifically:

[0023]

[0024] in, For the target damping force, The damping coefficient is... This represents the probability value of instability risk. This represents the difference between the current core feature parameters and the normal range. This represents the maximum permissible deviation of the feature parameter. This is the working condition correction factor, with a value of 0.8 for no-load conditions and 1.2 for full-load conditions.

[0025] Furthermore, in the drive control steps of the damping actuator, the damping force adjustment range of the hydraulic damper is 0-5000N, and the adjustment rate is 100N / ms; the damping force adjustment range of the electromagnetic damper is 0-3000N, and the adjustment rate is 150N / ms; the control command transmission delay does not exceed 20ms, and the damping actuator response delay does not exceed 50ms.

[0026] Furthermore, the remote monitoring platform supports centralized management of multiple devices, with data storage time of no less than one year and support for querying historical data; emergency alarm methods include audible and visual alarms, SMS notifications, and APP push notifications, with an alarm response time of no more than 10 seconds, ensuring that operators receive early warning information in a timely manner.

[0027] Furthermore, a transient instability identification and adaptive damping control system for lifting equipment includes the following modules:

[0028] Data acquisition module: It consists of tilt angle, vibration, wind speed, load, displacement sensors and data transmission unit. The sensors are deployed in key parts of the equipment to collect multi-dimensional operating data in real time. The data transmission unit adopts 5G+edge computing technology.

[0029] Feature processing module: It receives data transmitted from the data acquisition module, performs feature extraction, normalization, correlation analysis and redundant data removal, and outputs the core feature parameter sequence. The module has a built-in data processing algorithm library and supports dynamic algorithm updates and optimization.

[0030] Instability identification module: Based on long short-term memory neural network, it includes model training unit and identification inference unit. The model training unit trains and optimizes model parameters through historical data and simulated data. The identification inference unit inputs the core feature parameter sequence and outputs the transient instability risk probability value.

[0031] Risk assessment module: It presets three levels of instability risk thresholds, receives the risk probability value output by the instability identification module, dynamically adjusts the thresholds based on the changing trends of characteristic parameters, determines the instability risk level and triggers the corresponding response signal, and supports manual adjustment and automatic optimization of the thresholds;

[0032] Parameter calculation module: Based on the risk level output by the risk assessment module, the core feature parameters output by the feature processing module, and the equipment operating condition data, calculate the optimal damping control parameters, including the optimal damping force, the optimal adjustment rate and range of action calculated by combining the risk level gain and the physical limit constraint of the actuator, and has a built-in parameter calculation algorithm and operating condition adaptation rule library.

[0033] Drive control module: Composed of control command conversion unit and CAN bus communication unit, it converts the control parameters output by parameter calculation module into standardized control commands, which are transmitted to damping actuator via CAN bus to drive damper to adjust damping state in real time. The module supports control command verification and retransmission mechanism.

[0034] Feedback optimization module: It acquires the equipment operation data after control through the data acquisition module, analyzes the control effect, uses the gradient descent algorithm to iteratively optimize the instability identification model parameters and damping control parameters, generates an optimization report and transmits it to the relevant modules;

[0035] Remote monitoring module: includes a data storage unit, a visualization unit, an alarm unit, and an emergency linkage unit. The data storage unit stores equipment operation data, control commands, and optimization records. The visualization unit presents data in the form of charts. The alarm unit issues an alarm signal when the risk level escalates. The emergency linkage unit activates the equipment braking system in emergency situations.

[0036] Furthermore, the data acquisition module employs a redundant sensor design, with two sets of identical sensors installed in parallel at key locations. The validity of the data is verified by comparing the data using a data fusion algorithm. When the data deviation between the two sets of sensors exceeds 5%, the system automatically switches to the backup sensor and issues a fault alarm.

[0037] Furthermore, the long short-term memory neural network of the instability identification module contains 3 hidden layers, each with 128 hidden units. The initial learning rate is set to 0.001, and the learning rate is adjusted using an adaptive moment estimation optimization algorithm. After the model is trained, the recognition accuracy is no less than 95%, and the AUC value on the test set is no less than 0.98.

[0038] Furthermore, the control command conversion unit of the drive control module supports multiple damper protocol adaptations and is compatible with different types of actuators, while the CAN bus communication unit adopts a dual-bus redundant design.

[0039] Furthermore, the emergency linkage unit of the remote monitoring module and the equipment braking system adopt a dual linkage method of hard wire connection and wireless communication. The hard wire connection ensures rapid transmission of control commands in emergency situations, while the wireless communication serves as a backup. When an emergency instability risk occurs, the braking system receives the command and performs an emergency stop operation within 200ms, cutting off the power output of the equipment.

[0040] Compared with existing technologies, the beneficial effects of this invention are:

[0041] First, multi-dimensional data fusion enhances identification accuracy. This invention integrates multi-dimensional data such as tilt, vibration, load, displacement, and environment, extracts core feature parameters through time-domain and frequency-domain analysis, eliminates redundant information, and constructs an instability identification model using a long short-term memory neural network, significantly improving the accuracy and timeliness of transient instability identification. Compared to the limitations of existing technologies that monitor only a single dimension, the deep fusion of multi-source data can comprehensively characterize the precursory features of instability, avoiding misjudgments or omissions caused by incomplete information, ensuring accurate risk identification in the early stages of instability, and allowing sufficient time for subsequent control.

[0042] Secondly, adaptive damping control enables active instability suppression. This invention overcomes the limitations of existing technologies that rely on passive alarms and emergency braking. Based on the instability risk level, characteristic parameters, and operating conditions, it dynamically calculates the optimal damping control parameters and drives hydraulic or electromagnetic dampers to adjust the damping state in real time. This suppresses tilting, vibration, and sudden displacement changes at the source, effectively weakening transient instability trends. This active control mechanism requires no manual intervention, has a fast response speed, and high control precision. It can pull the equipment back to a stable range during the instability development stage, preventing risks from escalating into major accidents and significantly improving equipment operational safety.

[0043] Furthermore, dynamic optimization and redundant design enhance system reliability. This invention uses a feedback optimization module to evaluate control effectiveness in real time and employs a gradient descent algorithm to iteratively optimize the identification model and control parameters, ensuring the system adapts to instability prevention and control requirements under different loads, wind speeds, and operating conditions. Redundant design of sensors and communication links ensures the continuity of data acquisition and command transmission, preventing system failure due to a single device malfunction. Simultaneously, a three-level risk assessment and emergency response mechanism enables tiered risk response and dual protection in emergency situations, further improving the system's reliability and robustness in complex construction environments.

[0044] Finally, remote monitoring and data traceability expand the application value. The remote monitoring module of this invention supports centralized management of multiple devices, data visualization, and historical data traceability, facilitating real-time monitoring of equipment operating status and instability prevention by management personnel, providing data support for work scheduling and equipment maintenance. Multi-channel push notifications for emergency alarms ensure operators receive timely warning information, forming a closed-loop management system of "identification-control-feedback-optimization." This system not only reduces the incidence of instability accidents in lifting equipment but also improves construction efficiency, reduces economic losses, and is compatible with various lifting equipment such as tower cranes and gantry cranes, demonstrating broad engineering application prospects. Attached Figure Description

[0045] Figure 1 This is a schematic block diagram of the transient instability identification and adaptive damping control method for lifting equipment proposed in this invention;

[0046] Figure 2 This is a schematic block diagram of the transient instability identification and adaptive damping control system for lifting equipment proposed in this invention;

[0047] Figure 3 A bar chart comparing the delay in identifying instability using different prevention and control methods;

[0048] Figure 4 A bar chart comparing monitoring accuracy under environmental interference. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0051] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0052] Reference Figures 1 to 4 A method for transient instability identification and adaptive damping control of lifting equipment, comprising the following steps:

[0053] Multi-dimensional transient data acquisition involves deploying tilt sensors, vibration sensors, wind speed sensors, load sensors, and displacement sensors to collect data on the tilt angle of the outriggers of the lifting equipment, the vibration amplitude of the main beam, the wind speed of the working environment, the weight of the lifting load, and the displacement of key structural parts. The acquisition frequency is 100Hz. After the data is filtered and denoised, it is transmitted to the data processing unit to ensure data integrity and real-time performance.

[0054] Instability feature extraction and preprocessing: Time domain and frequency domain analysis of the collected multi-dimensional data are performed to extract tilt angle change rate, vibration peak value, wind speed pulsation coefficient, load fluctuation amplitude and displacement abrupt change feature values. Normalization is used to unify the data dimensions, and redundant data is removed through correlation analysis.

[0055] The transient instability identification model is constructed and trained based on a long short-term memory neural network. Addressing the strong correlation of temporal features and the susceptibility of abrupt signal interference in transient instability of lifting equipment, a preprocessed time series of core feature parameters is used as the model input. This series is constructed at a 100Hz sampling frequency, and each input sequence contains core feature parameters such as the tilt angle change rate, vibration peak value, wind speed pulsation coefficient, load fluctuation amplitude, and displacement abrupt change characteristic value from 50 consecutive sampling points. The dimensions are consistent with the core feature dimensions output by the feature processing module. The model training dataset is constructed in a 4:3:3 ratio using historical instability case data, normal operation data under all working conditions, and simulated instability data from multiple scenarios. The historical instability case data covers the entire process of real instability under different loads, wind speeds, and operating conditions. The simulated instability data is generated by simulating typical instability scenarios such as equipment overturning and excessive vibration using simulation software. The dataset is randomly divided into training and test sets in a 7:3 ratio, and the data types are balanced.

[0056] The model's network structure employs a 3-layer LSTM architecture with 128 hidden units per layer initially. The input layer dimension matches the dimension of the core feature parameter sequence, and the output layer uses the Sigmoid activation function to map the model's output value to the 0-1 range. This output value represents the probability of transient instability; a probability value closer to 0 indicates more stable device operation, while a probability value closer to 1 indicates a higher risk of transient instability. During model training, the cross-entropy loss function is used as the loss criterion. An adaptive moment estimation optimization algorithm iteratively optimizes the number of network layers, the number of hidden units, and the learning rate. The initial learning rate is set to 0.001 and dynamically adjusted based on the convergence speed of the loss function. Parameter optimization stops when the loss function shows no significant decrease after 100 consecutive iterations. Simultaneously, a 5-fold cross-validation method is used to validate the model. The training set is randomly divided into 5 subsets, with one subset selected as the validation set and the remaining 4 subsets used as the training set for model training and validation. The average performance of the model across the 5 validation iterations is taken as the final model performance metric. This method effectively improves the model's generalization ability and avoids overfitting. After training, the model extracts temporal features and performs nonlinear fitting on the input real-time core feature parameter sequence. After mapping by the Sigmoid activation function in the output layer, it directly outputs the transient instability risk probability value in the 0-1 interval.

[0057] The determination of instability risk level is based on a three-level instability risk threshold. The benchmark classification is as follows: an instability risk probability value of less than 30% is a safe level, 30% to 70% is a warning level, and more than 70% is an emergency level.

[0058] The core logic of dynamically adjusting risk assessment thresholds is to adaptively optimize threshold standards based on the urgency of risk reflected in the changing trends of characteristic parameters, avoiding response lag or misjudgment caused by fixed thresholds. The specific adjustment method is as follows: When core characteristic parameters show a rapid growth trend, i.e., the rate of change of tilt angle reaches or exceeds 0.5° / s, the vibration peak increases by 30% or more within 10 milliseconds, or the load fluctuation amplitude shows an increase of 5 consecutive sampling points, it indicates that the risk of instability is accumulating rapidly. At this time, the corresponding assessment threshold will be automatically lowered: the safety level threshold is lowered to 25%, and the warning level threshold is lowered to 65%, allowing the system to enter a warning or emergency response state in advance. If the core characteristic parameters show a stable changing trend, i.e., the parameter growth rate does not exceed 0.1° / s and there is no continuous fluctuation, the baseline threshold remains unchanged. If the parameters show a decaying trend, i.e., key parameters such as vibration peak and tilt angle change rate gradually decrease, the safety level threshold can be appropriately raised to 35% to avoid over-response affecting normal equipment operation.

[0059] When the risk level is upgraded according to the adjusted threshold, i.e., from the safety level to the warning level, or from the warning level to the emergency level, the corresponding response mechanism will be triggered immediately: the warning level triggers the initial adjustment of the damping actuator, and the emergency level triggers the linkage between the maximum damping force output and the braking system to ensure accurate and timely prevention and control of instability risks.

[0060] Adaptive damping control parameter calculation: Based on the instability risk level, core characteristic parameters and equipment operating conditions, calculate the optimal damping control parameters, including the magnitude of damping force, adjustment rate and range of action, to ensure that the control parameters are accurately matched with the instability situation;

[0061] The damping actuator drive control converts the calculated damping control parameters (damping force magnitude, adjustment rate, and range of action) into standardized control commands, which are then transmitted to the hydraulic damper or electromagnetic damper via a dual-bus redundant CAN bus. The two types of dampers, through targeted structural design and dynamic adjustment mechanisms, respectively achieve precise suppression of equipment tilt, vibration, and sudden displacement, thereby weakening the tendency of transient instability from the source.

[0062] To address equipment tilting: The damper cylinder is designed with tilt-linked valve cores at both ends. Through mechanical linkage with key parts of the equipment (outriggers, main beam), the valve cores can sense changes in tilt direction and angle in real time. Combined with the range parameters in the control command, the valve core conduction level is automatically switched so that the opening of the oil channel on the corresponding tilt side is adjusted proportionally. The larger the tilt angle, the smaller the channel opening, and the more concentrated the damping force is on the tilt direction, forming a reverse support force to counteract the tilting trend and prevent the equipment posture from continuously shifting.

[0063] For vibration suppression: The oil channel has a built-in elastic throttling plate and frequency modulation component. It can dynamically change the elastic coefficient of the throttling plate according to the adjustment rate parameter in the control command, adapting to the vibration frequency of the equipment in the 10Hz-100Hz frequency band. Through the synergistic effect of oil viscosity damping and throttling plate elastic damping, the vibration amplitude is rapidly attenuated, especially the suppression effect of main beam resonance vibration is significant.

[0064] For sudden displacement: The damper has a built-in high-pressure accumulator and a fast-response valve group. When a sudden displacement of the structure is detected, the accumulator releases high-pressure oil instantly. With the help of the valve group's millisecond-level response, the maximum damping force is quickly output. The incompressibility of the oil forms a rigid constraint, which curbs the further expansion of the displacement change and avoids damage to the equipment structure from impact loads.

[0065] Real-time feedback and optimization of control performance are achieved by collecting multi-dimensional data after control in real time through sensors. This analysis examines changes in instability risk probability and improvements in characteristic parameters to evaluate the damping control effect. Based on the feedback data, a gradient descent algorithm is used to iteratively optimize the instability identification model parameters and damping control parameters, thereby improving control accuracy and response speed. Specifically:

[0066] Through the deployment of tilt, vibration, load, displacement and wind speed sensors in the multi-dimensional transient data acquisition stage, real-time data on equipment operation after damping control is collected. The acquisition frequency is kept at 100Hz, consistent with that before control, to ensure the continuity of data sequence. The collected content covers core indicators such as equipment tilt angle, vibration amplitude, load fluctuation amplitude, structural displacement, and real-time value of instability risk probability, forming a complete dataset after control.

[0067] Single-indicator quantitative analysis: The improvement effect is calculated for each core characteristic parameter. The tilt angle improvement rate is calculated by "(tilt angle before control - tilt angle after control) / tilt angle before control × 100%", the vibration amplitude attenuation rate is calculated by "(vibration peak value before control - vibration peak value after control) / vibration peak value before control × 100%", and the displacement mutation suppression rate is calculated by "(displacement mutation frequency before control - displacement mutation frequency after control) / displacement mutation frequency before control × 100%". At the same time, it is determined whether each parameter has returned to the normal operating range (tilt angle ≤ ±0.5°, vibration amplitude ≤ 2mm, displacement ≤ 5mm), and the percentage of parameters that meet the standards is statistically analyzed.

[0068] Comprehensive Effect Weighted Evaluation: A control effect evaluation index E is constructed, calculated using the formula E = 0.4 × P Improvement Rate + 0.3 × F Improvement Rate + 0.2 × D Improvement Rate + 0.1 × Compliance Rate. Where P Improvement Rate is "(Probability of Instability Risk Before Control - Probability of Instability Risk After Control) / Probability of Instability Risk Before Control × 100%"; F Improvement Rate is the arithmetic mean of the improvement rates of core characteristic parameters of transient instability of lifting equipment, such as tilt angle improvement rate and vibration amplitude attenuation rate; D Improvement Rate is the displacement improvement rate, as displacement is a core key parameter for structural instability of lifting equipment and has the most direct impact on structural safety, therefore it is included separately in the weighted calculation system to highlight the control weight; the compliance rate is the percentage of core parameters such as tilt angle, vibration, displacement, and load fluctuation that return to the normal operating range of the equipment. An evaluation index E ≥ 80% is considered excellent control effect, 60% ≤ E < 80% is good, 40% ≤ E < 60% is average, and E < 40% is poor.

[0069] Based on the above evaluation results, the gradient descent algorithm is used for iterative optimization: taking the control effect evaluation index E as the objective function, the network weights, hidden layer thresholds, learning rate, and other parameters of the instability identification model, as well as the damping coefficient k, operating condition correction coefficient η, and basic adjustment rate in the damping control parameters are optimized. Using variables as optimization variables, the gradient values ​​of the objective function with respect to each variable are calculated. The variable values ​​are gradually adjusted according to the gradient direction, with each iteration step size set to 0.001. During the iteration process, the change of the evaluation index E is monitored in real time. When the increase of E is ≤1% or E≥95% after 50 consecutive iterations, the iteration is stopped, and the model parameters and control parameter library are updated. The optimized parameters will be directly applied to the next round of control process, continuously improving the system's control accuracy and response speed for different instability conditions, forming a closed-loop management and control mechanism of "control-evaluation-optimization".

[0070] Remote monitoring and emergency response linkage transmit instability risk levels, control commands, and equipment operating status data to the remote monitoring platform in real time. The platform supports data visualization, historical data tracing, and abnormal alarms. When an emergency instability risk occurs, the linkage equipment braking system performs an emergency shutdown operation and sends warning information to the operators.

[0071] This invention also includes a transient instability trend prediction step, which calculates the instability development rate using a formula, the specific formula being:

[0072]

[0073] in For the rate of unstable development, This represents the change in the probability of instability risk per unit time. For time intervals, This refers to the comprehensive change of core feature parameters per unit time. This represents the change in equipment displacement per unit time. For risk probability weights, For feature parameter weights, As the displacement change weight, and This calculation allows for the prediction of instability development trends in advance, providing time for adjusting damping control parameters. Furthermore, "in advance" here refers to the time point when the instability risk level triggers a response. Traditional control logic initiates damping control only after the instability risk probability reaches the warning or emergency level threshold. This step, however, predicts the trend in advance when the risk probability has not reached the threshold, but signs of instability have already appeared. This prediction step occurs after the "transient instability identification model outputs the risk probability value" and before the "instability risk level determination." At this point, the equipment may be in a "safe level but some characteristic parameters have deviated from the normal range" or "low warning level" state, before triggering a formal damping control response. By predicting the instability development rate V, control parameter adjustments can be completed before the risk level escalates, thus allowing for the prediction of instability development trends in advance and providing time for adjusting damping control parameters.

[0074] Core feature parameters refer to the key parameters strongly correlated with transient instability that are retained after instability feature extraction and preprocessing. These include, but are not limited to: tilt angle change rate (° / s), peak vibration (mm), load fluctuation amplitude (t), wind speed pulsation coefficient, and displacement mutation frequency (times / second). These parameters characterize the precursors of instability from multiple dimensions, including equipment attitude, stress state, and environmental disturbances, and are the core basis for judging the instability trend. For each core feature parameter, its "change per unit time" is first calculated using the following formula:

[0075]

[0076] in, Let be the change in the i-th feature parameter per unit time. Let i be the measured value of the i-th feature parameter at the current time. Let be the mean value of the normal operating range of the i-th feature parameter. This is for calculating the time interval.

[0077] Since different core characteristic parameters have varying degrees of influence on instability, they cannot be simply summed; a weighted summation is required to obtain the comprehensive change. The formula is:

[0078] (i ranges from 1 to n, where n is the total number of core feature parameters)

[0079] in, Let be the weight coefficient of the i-th feature parameter, taking values ​​between 0 and 1, and The weighting coefficients are determined using the Pearson correlation coefficient method: the correlation between each characteristic parameter and historical instability accidents is quantified, and the higher the absolute value of the correlation coefficient, the larger the weighting coefficient (for example, the peak vibration is most correlated with instability, so the weighting is 0.3; the tilt angle change rate is 0.25; the load fluctuation amplitude is 0.2; the wind speed pulsation coefficient is 0.15; and the displacement change frequency is 0.1), ensuring that parameters with a more significant impact on instability account for a higher proportion of the comprehensive change, and that the calculation results are more consistent with the actual instability trend.

[0080] This invention also includes an environmental interference compensation step. Environmental sensors collect wind speed, temperature, humidity, and ground vibration data in real time. The core logic of constructing an interference compensation model through a multivariate linear regression algorithm is to quantify the degree of interference of environmental factors such as wind speed, temperature, humidity, and ground vibration on the core data, thereby achieving accurate correction. First, sample data under different operating scenarios are collected: with wind speed, temperature, humidity, and ground vibration acceleration as independent variables (i.e., interference factors), and the "difference between the standard value of the core data without environmental interference and the actual collected value" as the dependent variable (i.e., interference bias), a sample dataset is constructed and divided into a training set and a test set in an 8:2 ratio.

[0081] Establish a multivariate linear regression equation based on the training set data: (where ΔY is the interference bias, X1-X4 are the measured values ​​of each environmental factor, a1-a4 are the regression coefficients, and b is the constant term). The coefficients are optimized by the least squares method. After verifying the model fit R²≥0.9 on the test set, the final compensation model is determined.

[0082] The generated compensation coefficients include regression coefficients and correction coefficients. During correction, real-time environmental factor data is substituted into the model to calculate the interference deviation under the current environment. Then, the original core data acquisition value is subtracted from the deviation value to complete the correction of data such as tilt angle and vibration amplitude. This ensures that the corrected data truly reflects the equipment's own operating status, completely eliminating the influence of environmental interference. The corrected data is more in line with the actual operating status of the equipment, reducing the impact of environmental interference on the instability identification and control effect. The compensation processing delay does not exceed 5ms.

[0083] In this invention, during the multi-dimensional transient data acquisition step, tilt sensors are installed on key parts such as the equipment legs, main beam, and tower cap, with a measurement range of -30° to 30° and an accuracy of ±0.05°; vibration sensors are triaxial accelerometers with a measurement range of ±10g and a resolution of 0.001g; wind speed sensors have a measurement range of 0-60m / s and an accuracy of ±0.1m / s; load sensors have a measurement range of 0-500t and an accuracy of ±0.5%FS; and displacement sensors are laser displacement sensors with a measurement range of 0-500mm and an accuracy of ±0.01mm.

[0084] In this invention, in the steps of instability feature extraction and preprocessing, the time-domain analysis uses the sliding window method to extract feature parameters, with the window size set to 50 sampling points; the frequency-domain analysis uses the fast Fourier transform to convert the time-domain signal into a frequency-domain signal and extract the feature frequencies in the 10Hz-100Hz frequency band; the normalization process uses the max-min normalization method to map the data to the 0-1 interval; and the correlation analysis uses the Pearson correlation coefficient method to remove feature parameters with an absolute correlation coefficient value lower than 0.3.

[0085] In this invention, the damping force adjustment accuracy is optimized through a formula in the adaptive damping control parameter calculation step. The specific formula is as follows:

[0086]

[0087] in, For the target damping force, This is the damping coefficient, with a value ranging from 500 to 2000 N·s / m. This represents the probability value of instability risk. This represents the difference between the current core feature parameters and the normal range. This represents the maximum permissible deviation of the feature parameter. The value is 0.8 for no-load conditions and 1.2 for full-load conditions. This calculation enables precise matching of damping force, avoiding over-control or under-control.

[0088] In this invention, the damping adjustment rate is dynamically calculated based on the instability risk level, instability development rate, and the rate of change of core characteristic parameters. The core design goal is to ensure that the damping force responds quickly to the instability trend, while avoiding over-adjustment that could impact the equipment structure. This achieves adaptive control where "the more severe the instability, the more timely the response; the more gradual the instability, the more stable the adjustment," while strictly avoiding the two core risks of instability and loss of control due to insufficient adjustment and structural damage due to over-adjustment.

[0089] The rate of change of core feature parameters is the comprehensive rate of change of core feature parameters. The calculation formula is:

[0090]

[0091] In the formula, The core characteristic parameter is a dimensionless normalized comprehensive change quantity, which integrates four types of core instability characteristic parameters: the tilt angle of the outrigger of the lifting equipment, the vibration acceleration of the main beam, the fluctuation of the lifting load, and the tension deviation of the wire rope. It is obtained by normalizing and weighting the rated reference value, and is a dimensionless parameter. The system sampling time interval is expressed in seconds (s). Directly quantify the drastic dynamic changes in the equipment's instability characteristics. The larger the value, the more severe the instability development trend. This parameter directly reflects the dynamic change trend of the equipment instability characteristics.

[0092] Based on the present invention, through simulation and field verification of over 1000 historical instability cases of lifting equipment and over 50 typical operating conditions, the effective operating conditions in engineering practice where damping intervention is required for lifting equipment (i.e., operating conditions with a clear tendency to instability but not yet out of control) are demonstrated. The reasonable range is .

[0093] The formula for calculating the damping adjustment rate is:

[0094]

[0095] The parameters in the formula are defined as follows:

[0096] The damping adjustment rate is expressed in N / ms, representing the magnitude of damping force adjustment per unit time.

[0097] The base adjustment rate is the reference adjustment rate under the rated operating conditions of the damper. For hydraulic dampers, it is 50 N / ms, and for electromagnetic dampers, it is 75 N / ms. The value shall not exceed 50% of the physical adjustment limit of the damper, so as to reserve redundancy for the safe operation of the equipment.

[0098] This represents the probability value of instability risk, which is dimensionless and ranges from [0,1]. It is obtained by mapping the instability risk level through a fuzzy inference algorithm; the higher the instability risk, the greater the probability. The closer the value is to 1;

[0099] The rate of instability development is dimensionless and is calculated using the transient instability trend prediction formula. It represents the increase in instability risk per unit time.

[0100] The critical value for the rate of instability development is dimensionless and takes the value of 1.0, corresponding to the critical safety threshold for the instability trend of the lifting equipment.

[0101] It is a dimensionless correction coefficient for the rate of change of characteristic parameters. Its core function is to dynamically correct the damping adjustment rate based on the severity of the equipment instability.

[0102] In order to make and Maintain a positive correlation, while Strictly controlled within the optimal safety range derived from theoretical derivation and verified in multiple engineering scenarios. Inside, definition The calculation formula is:

[0103]

[0104] This formula guarantees that when When it increases from 0.3 to 1.5, The linear increase from 1.0 to 2.0 fully conforms to the physical meaning of "the more severe the instability, the more timely the response". The entire process is continuous and smooth without inflection points, which can avoid equipment shock caused by sudden changes in the damping adjustment rate. At the same time, it is always in the optimal balance range allowed by the damper's mechanical structure and equipment safety.

[0105] Under different working conditions and The adaptation logic is as follows:

[0106] 1. When At this time, it corresponds to mild instability conditions (such as slight tilting or small-amplitude steady-state vibration). The instability trend is gradual, and a basic or slightly higher adjustment rate is adopted to avoid frequent fluctuations in damping force that could cause wear on equipment joints and structural fatigue.

[0107] 2. When At this time, corresponding to conventional instability conditions (such as moderate-amplitude tilting, resonance-prone vibration), The instability trend is moderate, and a relatively fast adjustment rate is matched to quickly suppress the risk while taking into account the stability of the adjustment.

[0108] 3. When At this time, corresponding to severe instability conditions (such as sudden strong wind impact or sudden change in lifting load), at this time The instability trend is rapid, and the risk is quickly contained by adjusting at a high safety rate, while strictly controlling it within the safe tolerance range of the equipment and dampers.

[0109] Technical verification and basis for the optimal range of coefficient values

[0110] This invention, through theoretical derivation and multi-scenario engineering verification, determines... The optimal range of values ​​for the coefficient is The optimal balance range between damping adjustment response speed and equipment structural safety is defined by the following core technical rationale:

[0111] like This will result in an excessively slow damping adjustment rate, failing to match the rate of instability development. For example... At that time, according to the formula The reasonable value should be 1.17; if an incorrect value is taken... Then the adjustment rate is only a fraction of the reasonable target rate under this operating condition. It is unable to suppress the instability trend in time, causing the risk to continue to expand, and completely losing the technical significance of damping active prevention and control;

[0112] like This will result in an excessively fast damping adjustment rate, causing the damping force output to generate a rigid impact. For example... hour, The reasonable safety upper limit is 2.0. If an error occurs... If the adjustment rate is increased by 25% compared to the maximum safe and reasonable value, long-term operation will cause the equipment outrigger joints to loosen and the main beam steel structure to suffer fatigue damage, which will increase the risk of equipment instability. At the same time, it will exceed the physical adjustment limit of the hydraulic / electromagnetic damper and cause overload failure of the actuator.

[0113] Verified through over 100 on-site tests using four typical types of lifting equipment: tower cranes, gantry cranes, crawler cranes, and bridge cranes. exist When operating within the range, the response delay of damping adjustment is ≤50ms, and the impact load borne by the equipment is ≥5% of the rated load, which fully meets the safety requirements for stability control of lifting equipment.

[0114] Design logic for coefficient truncation and extreme condition adaptation scheme

[0115] To ensure The value always remains stable at The optimal safe range, adapted to engineering The full dynamic range of change, the present invention for The coefficient settings include value truncation. The core purpose of this design is to address... In extreme scenarios exceeding the normal effective operating range, the system ensures stable and safe operation under abnormal conditions, rather than simply being limited by numerical values. It also achieves a smooth transition of the damping adjustment rate across all operating conditions, without control jumps or shocks. The specific processing logic and technical rationale are as follows:

[0116] 1. When At this time, the corresponding operating condition with a very gentle instability trend, or even only parameter fluctuations caused by sensor acquisition noise, indicates that the equipment is in a stable operating state. It eliminates the need for high-frequency active adjustment of the damping system, maintaining only the basic standby speed, thus avoiding unnecessary frequent actions, reducing mechanical wear on the damper, and extending the service life of the actuator;

[0117] 2. When At this time, it corresponds to extreme scenarios where the instability trend is extremely severe and the equipment is close to running out of control (such as a sudden 10-level gust of wind or the load eccentricity exceeding the limit value). In this case, take... This cutoff value matches the physical limits of the damper's mechanical response (maximum adjustment rate of 100 N / ms for hydraulic dampers and 150 N / ms for electromagnetic dampers), ensuring that the damping system responds at the maximum safe rate under extreme conditions while completely avoiding overload of the actuator and structural impact damage caused by over-range adjustment.

[0118] 3. When At that time, according to Continuous calculations precisely match the adjustment requirements of conventional instability conditions, enabling smooth and adaptive adjustment of the damping rate.

[0119] The above-mentioned truncation processing method can cover the operational needs of lifting equipment in all scenarios and under all working conditions, regardless of Regardless of extreme changes in equipment operating conditions or environmental disturbances, it can [respond to / benefit from] Precise control in The optimal safety range is determined to fundamentally ensure the rationality of the damping adjustment rate calculation, equipment adaptability, and operational safety.

[0120] III. Rate Gain and Limiting Control Based on Instability Risk Level

[0121] To further optimize the adaptability of damping adjustment under various operating conditions, avoid frequent adjustments under stable conditions, and enhance response speed under extreme risks, this invention classifies risk levels based on the instability risk probability P, sets corresponding risk level gain coefficients K, and applies hard limiting constraints to the final output rate. The formula for calculating the final damping adjustment rate is as follows:

[0122]

[0123] In the formula:

[0124] The damping adjustment rate is the final output to the damper actuator, expressed in N / ms.

[0125] This is the risk level gain coefficient, which is dimensionless and is assigned values ​​according to the instability risk level classification.

[0126] The maximum physical adjustment rate of the damper is set to 100 N / ms for hydraulic dampers and 150 N / ms for electromagnetic dampers.

[0127] The rules for risk level classification and gain coefficient selection are as follows:

[0128] 1. Security level ( Gain coefficient The final adjustment rate is taken as 0.5 times the initial calculated value to avoid unnecessary frequent adjustments under stable operating conditions, reduce mechanical wear of the damper, and extend the service life of the equipment.

[0129] 2. Warning Level ( Gain coefficient The final adjustment rate is taken as 1.0 times the initial calculated value, balancing the response speed to instability risk and the stability of damping adjustment, taking into account both risk prevention and equipment protection;

[0130] 3. Emergency Level ( Gain coefficient The final adjustment rate is 1.2 times the initial calculated value. It is strictly constrained within the maximum physical adjustment rate of the damper by the limiting formula, so as to quickly suppress the instability trend and completely avoid overload of the actuator and impact damage to the equipment structure caused by over-range adjustment.

[0131] In this invention, the damping range is determined based on the equipment's structural parameters, the direction of instability, and the distribution of core characteristic parameters, ensuring that the damping force is precisely applied to the critical instability points. The foundation range is the preset damping force boundary adapted to the inherent structural parameters of the lifting equipment, and is divided into foundation lengths corresponding to the linear range of action. The basic angle of action corresponding to the range of action. Both are determined through engineering simulation and field calibration based on structural parameters such as the span of the equipment outriggers, the length of the main beam, and the radius of slewing. They serve as the core calculation benchmark for the damping range. The specific calculation logic is as follows:

[0132] (1) Linear action range, applicable to tilt and displacement instability:

[0133] Where L is the linear damping range, and L1 is the length of the corresponding structural foundation of the equipment. The deviation value of the core feature parameter, P represents the maximum permissible deviation and the probability value of instability risk.

[0134] (2) Angular range of action, applicable to vibration and torsional instability:

[0135] Where θ is the range of damping angle. From the perspective of basic role, This is the difference between the current vibration frequency and the normal operating frequency. Let P be the normal operating vibration frequency of the equipment, and P be the probability value of instability risk.

[0136] When the instability trend is concentrated in a single location, the effective range focuses on that location, covering an area 1.0-1.5 times the base range;

[0137] When the instability trend is distributed throughout the entire structure, the scope of influence extends to the entire corresponding structure, covering an area 2.0-3.0 times the base area.

[0138] In this invention, during the drive control step of the damping actuator, the damping force adjustment range of the hydraulic damper is 0-5000N, and the adjustment rate is 100N / ms; the damping force adjustment range of the electromagnetic damper is 0-3000N, and the adjustment rate is 150N / ms; the control command transmission delay does not exceed 20ms, and the damping actuator response delay does not exceed 50ms, ensuring rapid suppression of transient instability.

[0139] In this invention, the remote monitoring platform supports centralized management of multiple devices and can simultaneously access the operating data of no less than 100 lifting devices. The platform data storage time is no less than one year, and it supports querying historical data by device number, time range, risk level, and other conditions. Emergency alarm methods include audible and visual alarms, SMS notifications, and APP push notifications. The alarm response time is no more than 10 seconds, ensuring that operators receive early warning information in a timely manner.

[0140] This invention includes the following modules:

[0141] The data acquisition module consists of tilt sensors, vibration sensors, wind speed sensors, load sensors, displacement sensors, and a data transmission unit. The sensors are deployed in key parts of the equipment to collect multi-dimensional operating data in real time. The data transmission unit adopts 5G+edge computing technology to ensure real-time data transmission and preliminary filtering.

[0142] Feature processing module: It receives data transmitted from the data acquisition module, performs feature extraction, normalization, correlation analysis and redundant data removal, and outputs the core feature parameter sequence. The module has a built-in data processing algorithm library and supports dynamic algorithm updates and optimization.

[0143] Instability identification module: Based on long short-term memory neural network, it includes model training unit and identification inference unit. The model training unit trains and optimizes model parameters through historical data and simulated data. The identification inference unit inputs the core feature parameter sequence and outputs the transient instability risk probability value. The identification delay does not exceed 100ms.

[0144] Risk assessment module: It presets three levels of instability risk thresholds, receives the risk probability value output by the instability identification module, dynamically adjusts the thresholds based on the changing trends of characteristic parameters, determines the instability risk level and triggers the corresponding response signal, and supports manual adjustment and automatic optimization of the thresholds;

[0145] Parameter calculation module: Based on the risk level output by the risk assessment module, the core feature parameters output by the feature processing module, and the equipment operating condition data, calculate the optimal damping control parameters, including the damping force magnitude, adjustment rate, and range of action. It has a built-in parameter calculation algorithm and operating condition adaptation rule library.

[0146] Drive control module: Composed of control command conversion unit and CAN bus communication unit, it converts the control parameters output by parameter calculation module into standardized control commands, which are transmitted to damping actuator via CAN bus to drive damper to adjust damping state in real time. The module supports control command verification and retransmission mechanism.

[0147] Feedback optimization module: It acquires the equipment operation data after control through the data acquisition module, analyzes the control effect, uses the gradient descent algorithm to iteratively optimize the instability identification model parameters and damping control parameters, generates an optimization report and transmits it to the relevant modules to realize system closed-loop optimization;

[0148] Remote monitoring module: includes a data storage unit, a visualization unit, an alarm unit, and an emergency linkage unit. The data storage unit stores equipment operation data, control commands, and optimization records. The visualization unit presents data in the form of charts. The alarm unit issues an alarm signal when the risk level escalates. The emergency linkage unit activates the equipment braking system in emergency situations.

[0149] In this invention, the sensors of the data acquisition module adopt a redundant design, with two sets of the same type of sensors installed in parallel in key parts. The validity of the data is verified by comparing the data through a data fusion algorithm. When the data deviation between the two sets of sensors exceeds 5%, the system automatically switches to the backup sensor and issues a fault alarm, ensuring the continuity and reliability of data acquisition.

[0150] In this invention, the long short-term memory neural network of the instability identification module contains 3 hidden layers, each with 128 hidden units. The initial learning rate is set to 0.001, and the learning rate is adjusted using an adaptive moment estimation optimization algorithm. After the model is trained, the identification accuracy is not less than 95%, and the AUC value on the test set is not less than 0.98.

[0151] In this invention, the control command conversion unit of the drive control module supports multiple damper protocol adaptations and is compatible with different types of actuators such as hydraulic dampers and electromagnetic dampers. The CAN bus communication unit adopts a dual-bus redundant design with a transmission rate of 1Mbps and a transmission error rate of less than 10%. -6 This ensures reliable transmission of control commands.

[0152] In this invention, the emergency linkage unit of the remote monitoring module and the equipment braking system adopt a dual linkage method of hard wire connection and wireless communication. The hard wire connection ensures the rapid transmission of control commands in emergency situations, while the wireless communication serves as a backup. When an emergency instability risk occurs, the braking system receives the command and performs an emergency stop operation within 200ms, cutting off the power output of the equipment and ensuring the safety of the equipment and personnel.

[0153] The following four examples further illustrate specific embodiments of the present invention:

[0154] Example 1: Implementation of Transient Instability Prevention and Control for Tower Cranes in Outdoor Operations

[0155] This embodiment applies to a tower crane at a construction site. The equipment has a maximum lifting capacity of 80t and an operating height of 60m, and is mainly used for hoisting building materials such as steel bars and formwork. The construction site experiences frequent wind changes and slight ground unevenness, which can easily cause the equipment to tilt and vibrate. This invention is needed to achieve accurate identification of transient instability and adaptive damping control to ensure safety during high-altitude operations.

[0156] I. Core Implementation Details

[0157] Multi-dimensional transient data acquisition: Tilt sensors are installed at the top of the tower crane's outriggers, the midpoint of the main beam, and the tower cap; vibration sensors are deployed at both ends of the main beam; wind speed sensors are installed in the middle of the crane body; load sensors are installed at the hook; and laser displacement sensors are deployed at key nodes of the boom. All sensors are set to a 100Hz acquisition frequency. The tilt sensor measurement range is -30° to 30°, the vibration sensor measurement range is ±10g, the wind speed sensor measurement range is 0-60m / s, the load sensor measurement range is 0-100t, and the displacement sensor measurement range is 0-500mm. The acquired data is filtered and denoised before being transmitted to the data processing unit via a 5G network.

[0158] Instability Feature Extraction and Preprocessing: Time-domain and frequency-domain analysis was performed on the collected multi-dimensional data. A sliding window method was used to extract feature parameters such as the rate of change of tilt angle and peak vibration, with the window size set to 50 sampling points. The vibration time-domain signal was converted to a frequency-domain signal using Fast Fourier Transform, and feature frequencies in the 10Hz-100Hz band were extracted. The max-min normalization method was used to map all feature parameters to the 0-1 interval. Redundant data with correlation coefficients below 0.3 were removed using the Pearson correlation coefficient method, retaining core feature parameters such as the rate of change of tilt angle, peak vibration, and wind speed fluctuation coefficient.

[0159] Instability Identification Model Operation and Risk Assessment: The instability identification model, based on a long short-term memory neural network, is trained using historical instability case data, normal operation data, and simulated instability data. After inputting the core feature parameter sequence, it outputs a transient instability risk probability value. Three risk thresholds are set: a risk probability below 30% is considered safe, 30%-70% is a warning level, and above 70% is an emergency level. The model's recognition latency is controlled within 100ms. When the risk level changes, the thresholds are dynamically adjusted to adapt to the actual working conditions.

[0160] Adaptive damping control parameter calculation and execution: Optimal damping control parameters are calculated based on risk level, core characteristic parameters, and operating conditions. When the lifting load is 50t (full load condition) and the risk probability is 45% (early warning level), the target damping force is calculated by combining the tilt angle change rate and vibration peak value. The control parameters are transmitted via CAN bus to the hydraulic damper installed at the connection between the tower crane legs and the tower body. The damper's damping force adjustment range is 0-5000N, with an adjustment rate of 100N / ms. Upon responding to control commands, the damping state is adjusted in real time to suppress equipment tilting and vibration.

[0161] Control effect feedback and optimization: Sensors collect real-time equipment operation data after control, and the feedback optimization module analyzes changes in the probability of instability risk and the improvement of characteristic parameters. If the risk probability is still higher than 30% after control, the gradient descent algorithm is used to iteratively optimize the parameters of the instability identification model and the damping control parameters until the equipment returns to a stable state. Simultaneously, the optimized parameters are stored in the system to provide a reference for subsequent similar operating conditions.

[0162] Remote monitoring and emergency response: The remote monitoring platform receives equipment operation data, risk levels, and control commands in real time and displays them visually in chart form. When the risk level rises to the emergency level, the platform issues an audible and visual alarm and simultaneously pushes warning information to operators via SMS and APP. The platform also triggers the equipment's braking system to perform an emergency stop operation within 200ms, cutting off power output.

[0163] Table 1: Comparison of Transient Instability Prevention and Control Effects of Tower Cranes

[0164] Evaluation indicators Traditional prevention and control methods The prevention and control method of this invention Instability identification delay 500ms 80ms Vibration amplitude suppression rate 35% 85% Emergency shutdown response time 500ms 180ms Instability accident rate 2.1% 0.2% Stable operating time of equipment 85% 99.5%

[0165] Table 1 clearly demonstrates the control advantages of this embodiment. Traditional control methods rely on single-sensor monitoring, resulting in long identification delays, poor vibration suppression, and slow emergency shutdown response, leading to a high incidence of instability accidents. This invention significantly shortens the identification delay through multi-dimensional data fusion and a precise identification model. The adaptive control of the hydraulic damper effectively suppresses vibration amplitude, and the emergency shutdown response time is significantly reduced. The stable operating time of the equipment is increased to 99.5%, and the instability accident rate is reduced to 0.2%, fully demonstrating that this solution can quickly respond to transient instability risks in outdoor operations and ensure the safe operation of tower cranes.

[0166] Example 2: Implementation of Transient Instability Prevention and Control for Gantry Cranes in Port Operations

[0167] This embodiment is applied to a gantry crane at a port container terminal. The equipment has a span of 40m and a lifting capacity of 60t, and is mainly used for container loading and unloading operations. The terminal operating environment has problems such as strong winds and uneven tracks caused by ground subsidence, which can easily cause lateral vibration and sudden displacement of the equipment. This invention is needed to achieve active prevention and control of transient instability, thereby improving operational efficiency and safety.

[0168] I. Core Implementation Details

[0169] Multi-dimensional transient data acquisition: Inclination and vibration sensors are installed on the gantry crane's outriggers, main beam, and trolley frame. Laser track detectors and wind speed sensors are deployed on both sides of the track. Load sensors are installed at the hook, and displacement sensors are placed at the ends of the trolley's running track. The sensor acquisition frequency is 100Hz, the inclination sensor accuracy reaches ±0.05°, the laser track detector monitors track flatness in real time, and the wind speed sensor accurately captures instantaneous wind speed changes. The collected data is initially processed by edge computing nodes and then transmitted to the data processing unit through a fiber optic ring network, with a transmission error rate of less than 10%. -6 .

[0170] Instability Feature Extraction and Environmental Interference Compensation: A sliding window method and Fast Fourier Transform are used to extract core feature parameters, while simultaneously initiating an environmental interference compensation process. Environmental sensors collect data on wind speed, temperature, and track smoothness. A multivariate linear regression algorithm is used to construct an interference compensation model, generating compensation coefficients to correct data such as tilt angle and vibration amplitude. The compensation processing delay is no more than 5ms, reducing the impact of environmental factors on monitoring accuracy.

[0171] Instability Identification and Risk Assessment: The instability identification model takes into account the compensated sequence of core feature parameters and, combined with the characteristics of port operations such as load fluctuations and wind speed changes, outputs an instability risk probability value. A dynamic threshold mechanism is used for risk level assessment; when the wind speed exceeds 15 m / s, the thresholds for warning and emergency levels are appropriately lowered to trigger preventative measures in advance. The model's identification accuracy is no less than 95%, effectively avoiding false positives and false negatives.

[0172] Adaptive damping control execution: The target damping force is calculated based on the risk level and operating conditions. Electromagnetic dampers are installed at the bottom of the gantry crane's outriggers, with a damping force adjustment range of 0-3000N and an adjustment rate of 150N / ms. When the risk probability is 60% (early warning level), the electromagnetic dampers quickly adjust the damping state to suppress lateral vibration; when the risk probability rises to 75% (emergency level), the dampers output the maximum damping force, simultaneously triggering the trolley braking system to slow down the operating speed. The control command transmission delay does not exceed 20ms, and the damper response delay does not exceed 50ms.

[0173] Feedback optimization and remote monitoring: The feedback optimization module analyzes the control effect in real time. If the vibration amplitude is not effectively suppressed, it iteratively optimizes the damping control parameters and the weights of the instability identification model. The remote monitoring platform supports centralized management of multiple devices, storing device operating data, control commands, and optimization records for at least one year. Managers can query historical data through the platform to analyze instability risk patterns. When an emergency instability risk occurs, the platform pushes early warning information through multiple channels to ensure timely response by operators.

[0174] Table 2: Comparison of Transient Instability Prevention and Control Effects of Gantry Cranes in Port Operations

[0175] Evaluation indicators Traditional prevention and control methods The prevention and control method of this invention Lateral vibration amplitude 8mm 1.2mm Displacement mutation suppression rate 40% 90% Risk misjudgment rate 8% 1.5% Interruption duration 60 minutes / month 5 minutes / month Equipment maintenance costs 100,000 yuan / year 30,000 yuan / year

[0176] Table 2 data highlights the application value of this embodiment. Traditional control methods cannot effectively cope with the interference caused by strong winds and uneven tracks in ports. Lateral vibration amplitude is large, displacement abrupt change suppression is poor, and the risk misjudgment rate is high, leading to frequent operation interruptions and high maintenance costs. This invention improves monitoring accuracy through environmental interference compensation, and the adaptive control of the electromagnetic damper significantly reduces the risk of lateral vibration amplitude and displacement abrupt changes, lowering the risk misjudgment rate to 1.5%. Operational interruption time is greatly reduced, and maintenance costs are reduced by 70%. This ensures the continuity of container loading and unloading operations, extends equipment lifespan, and adapts to the needs of complex port operating environments.

[0177] Example 3: Implementation of Transient Instability Prevention and Control for Tracked Crane Wind Turbine Installation

[0178] This embodiment applies to a crawler crane used in a wind power project. The equipment has a lifting capacity of 150t and an operating radius of 30m, and is used for hoisting large components such as wind turbine nacelles and rotors. The wind farm site is located in a mountainous area with complex terrain, strong and unstable winds, and huge and concentrated loads during hoisting, which can easily cause transient instability symptoms such as equipment tilting and vibration. This invention is needed to achieve full-process control to ensure the safety of hoisting heavy components.

[0179] I. Core Implementation Details

[0180] Multi-dimensional transient data acquisition: Tilt and vibration sensors are installed on the track frame, slewing platform, boom root, and boom head of the crawler crane. A wind speed sensor is deployed in the middle of the boom, a high-precision load sensor is installed at the hook, and displacement sensors are arranged at key sections of the boom. The sensors acquire data at a frequency of 100Hz, with the load sensor achieving an accuracy of ±0.5%FS, enabling precise capture of load fluctuations during lifting. The wind speed sensor monitors real-time changes in mountainous gusts. The acquired data, after filtering, noise reduction, and standardization, is transmitted to the data processing unit via 5G+edge computing technology.

[0181] Instability Feature Extraction and Preprocessing: A sliding window method was used to extract time-domain features such as the rate of change of tilt angle and load fluctuation amplitude, with a window size of 50 sampling points. Frequency-domain features of the vibration signal were extracted using Fast Fourier Transform, focusing on anomalous frequencies in the 10Hz-100Hz band. The maximum-minimum normalization method was used to unify data dimensions, and correlation analysis was used to remove redundant data, retaining core feature parameters strongly correlated with transient instability to provide accurate input for instability identification.

[0182] Instability Identification and Risk Assessment: The instability identification model combines the load characteristics of wind turbine installation with the features of the mountainous environment. After inputting the core feature parameter sequence, it outputs an instability risk probability value. A three-level risk threshold is set. When hoisting critical components (such as the nacelle), the sensitivity of the warning level is appropriately increased, triggering an early warning when the risk probability reaches 25%. The model's identification delay is no more than 100ms, enabling rapid response to transient instability signs.

[0183] Adaptive damping control execution: Optimal damping control parameters are calculated based on risk level, core characteristic parameters, and lifting conditions. Hydraulic dampers are installed at the luffing cylinder of the crawler crane's main boom, and electromagnetic dampers are installed at the connection between the slewing platform and the crawler frame. When lifting the nacelle, if the load reaches 120t and the risk probability is 35% (early warning level), the dampers quickly adjust the damping force to suppress main boom vibration and equipment tilting. When the risk probability rises to 70% (emergency level), the dampers output maximum damping force, and simultaneously, the equipment braking system stops the lifting operation.

[0184] Feedback optimization and emergency response: The feedback optimization module evaluates the control effect in real time. If slight vibrations still exist in the equipment, the gradient descent algorithm is used to optimize the damping control parameters and instability identification model parameters. The remote monitoring platform displays the equipment's operating status, risk level, and control commands in real time. When an emergency instability risk occurs, the platform issues an audible and visual alarm, pushes warning information to operators and project managers, and triggers an emergency shutdown of the equipment to ensure the safety of heavy components and equipment.

[0185] Table 3: Comparison of Transient Instability Prevention and Control Effects of Tracked Cranes in Wind Turbine Installation

[0186] Evaluation indicators Traditional prevention and control methods The prevention and control method of this invention main boom vibration amplitude 10mm 1.5mm Equipment tilt suppression rate 38% 92% Lifting safety success rate 88% 99.8% Risk identification accuracy 75% 98% hoisting operation efficiency 80% 95%

[0187] Table 3 data fully demonstrates the prevention and control effectiveness of this embodiment. Traditional prevention and control methods struggle to cope with the complex mountainous environment and the impact of heavy lifting loads, resulting in large boom vibration amplitudes, poor equipment tilt suppression, and low risk identification accuracy, leading to low lifting safety success rates and low operational efficiency. This invention, through multi-dimensional data fusion and a precise identification model, significantly improves risk identification accuracy. The coordinated control of hydraulic and electromagnetic dampers effectively suppresses boom vibration and equipment tilt, increasing the lifting safety success rate to 99.8%. Operational efficiency is improved by 15%, ensuring the smooth progress of wind power installation projects while reducing the risk of safety accidents, thus meeting the stringent requirements of heavy lifting.

[0188] Example 4: Implementation of Transient Instability Prevention and Control Measures for Bridge Cranes in Workshops

[0189] This embodiment applies to a bridge crane in a machine processing plant. The equipment has a span of 25m and a lifting capacity of 50t, and is mainly used for the transfer of machine tools and heavy workpieces within the workshop. The workshop has problems such as dense equipment layout and limited working space. During the transfer process, sudden starts and stops, load eccentricity, and sudden changes in equipment vibration and displacement are prone to occur. This invention is needed to achieve precise control of transient instability and avoid collisions with surrounding equipment.

[0190] I. Core Implementation Details

[0191] Multi-dimensional transient data acquisition: Inclination and vibration sensors are installed on the main beam, end beams, and trolley frame of the bridge crane; displacement sensors are deployed beside the trolley's running track; load sensors are installed at the hook; and wind speed sensors are installed on the workshop ceiling (to monitor workshop ventilation airflow). The sensors operate at a 100Hz frequency, with displacement sensors achieving an accuracy of ±0.01mm, accurately capturing displacement changes during trolley operation. Load sensors monitor load eccentricity in real time. The acquired data, after preprocessing, is transmitted to the data processing unit via industrial Ethernet.

[0192] Instability Feature Extraction and Environmental Interference Compensation: The sliding window method and Fast Fourier Transform are used to extract core feature parameters, while simultaneously initiating an environmental interference compensation process. For interference factors such as workshop ventilation airflow and ground vibration, a multivariate linear regression algorithm is used to generate compensation coefficients to correct the collected data, reducing the impact of environmental interference on monitoring accuracy. The compensation processing delay does not exceed 5ms.

[0193] Instability Identification and Risk Assessment: The instability identification model takes into account the compensated sequence of core feature parameters and, combined with the characteristics of workshop operations such as load eccentricity and sudden starts and stops, outputs an instability risk probability value. A dynamic threshold mechanism is used for risk level assessment; when the trolley's running speed exceeds 1 m / s, the warning level threshold is appropriately lowered. The model's identification accuracy is no less than 95%, effectively avoiding misjudgments caused by changes in operating conditions.

[0194] Adaptive damping control execution: The target damping force is calculated based on the risk level and operating conditions. Electromagnetic dampers are installed at the axles of the trolley wheels of the bridge crane, and hydraulic dampers are installed at the connection between the main beam and the end beam. When transferring eccentric loads (eccentricity 0.5m) and the risk probability is 40% (early warning level), the dampers quickly adjust the damping state to suppress trolley vibration and main beam deformation. When the risk probability rises to 70% (emergency level), the dampers output the maximum damping force and simultaneously activate the trolley braking system to decelerate and avoid collisions with surrounding equipment.

[0195] Feedback optimization and remote monitoring: The feedback optimization module analyzes the control effect in real time and iteratively optimizes the damping control parameters and instability identification model parameters. The remote monitoring platform stores equipment operation data and control records, allowing administrators to query historical data and optimize work processes. In the event of an emergency instability risk, the platform issues audible and visual alarms and pushes warning information to operators to ensure timely response.

[0196] Table 4: Comparison of Transient Instability Prevention and Control Effects of Bridge Cranes in Workshops

[0197] Evaluation indicators Traditional prevention and control methods The prevention and control method of this invention Car vibration amplitude 6mm 0.8mm Main beam deformation 4mm 0.5mm Collision risk incidence 3.2% 0.1% Operational response sensitivity 600ms 100ms Equipment lifespan 8 years 12 years

[0198] Table 4 provides a clear overview of the advantages of this embodiment. Traditional control methods are ineffective in addressing the impacts of load eccentricity and sudden starts and stops during workshop operations. This results in large trolley vibration amplitudes and main beam deformations, a high collision risk, low operational response sensitivity, and reduced equipment lifespan. This invention significantly improves operational response sensitivity through environmental interference compensation and a precise identification model. The coordinated control of electromagnetic and hydraulic dampers significantly reduces trolley vibration amplitudes and main beam deformations, lowering the collision risk to 0.1%. Equipment lifespan is extended by 50%, ensuring safe transport operations within the workshop while reducing maintenance and replacement costs, thus adapting to the complex operational needs within the limited space of a workshop.

[0199] Reference Figure 3 This diagram visually illustrates the core advantage of this invention in terms of the timeliness of instability identification. Traditional single-sensor and manual judgment are limited by the single data dimension or slow response, resulting in the longest identification delay. Although ultrasonic sensor networks optimize data acquisition methods, they lack multi-dimensional fusion and intelligent algorithm support, resulting in still high latency. PLC threshold control relies on fixed logic and cannot quickly respond to transient changes. This invention, through multi-dimensional data fusion, deep learning model inference, and edge computing preprocessing, compresses the identification latency to 80ms, far lower than existing technologies. It can quickly capture risk signs in the early stages of instability, reserving sufficient time for subsequent adaptive damping control, thus solving the core pain point of lagging identification in traditional prevention and control methods.

[0200] Reference Figure 4This figure highlights the crucial role of the environmental compensation mechanism in this invention. Without environmental compensation, monitoring methods exhibit significantly increased errors when wind speeds rise or the track becomes uneven; under conditions of level 6 winds and 5mm track unevenness, the error far exceeds the safety threshold. This invention, by collecting environmental parameters using an ultrasonic anemometer and a laser track detector, constructs a multivariate adaptive filtering model to correct monitoring data in real time. Even under complex conditions such as level 6 winds and 5mm track unevenness, the monitoring error remains within 0.1°. This advantage ensures the accuracy of the instability identification model's input data, avoids misjudgments or omissions caused by environmental interference, and enables the system to maintain high-precision monitoring even in complex construction environments, solving the pain point of weak anti-interference capabilities in traditional monitoring methods.

[0201] The above are merely preferred embodiments 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 method for identifying transient instability and adaptive damping control of lifting equipment, characterized in that, Includes the following steps: Multi-dimensional transient data acquisition involves deploying tilt, vibration, wind speed, load, and displacement sensors to collect data on the tilt angle of the outriggers of the lifting equipment, the vibration amplitude of the main beam, the operating wind speed, the weight of the lifting load, and the displacement of structural parts. The data is then filtered and noise-reduced before being transmitted to the processing unit. Instability feature extraction and preprocessing: Time domain and frequency domain analysis of multi-dimensional data are performed to extract tilt angle change rate, vibration peak value, wind speed pulsation coefficient, load fluctuation amplitude and displacement abrupt change feature value. Redundant data are removed by maximum-minimum normalization and Pearson correlation analysis. The transient instability identification model is constructed and trained by building a long short-term memory neural network. The time series of core feature parameters, such as the rate of change of tilt angle, peak vibration, wind speed pulsation coefficient, load fluctuation amplitude, and displacement mutation feature value obtained by instability feature extraction and preprocessing, are input into the model. A training dataset is constructed by combining historical instability, normal operation, and simulated instability data. The model is trained by cross-validation and adaptive moment estimation optimization algorithm, and the model outputs the transient instability risk probability value in the 0-1 interval. The instability risk level is determined by setting three levels of instability risk thresholds. The baseline classification standard is as follows: an instability risk probability value below 30% is the safety level, 30% to 70% is the warning level, and above 70% is the emergency level. The risk judgment threshold is dynamically adjusted based on the changing trend of core characteristic parameters, and the threshold standard is adaptively optimized: when the rate of change of tilt angle reaches or exceeds 0.5° / s, the vibration peak increases by 30% or more within 10 milliseconds, or the load fluctuation amplitude increases for 5 consecutive sampling points, the safety level threshold is lowered to 25% and the warning level threshold is lowered to 65%, and the system enters the warning or emergency response state in advance; when the core characteristic parameters show a stable changing trend, that is, the parameter growth rate does not exceed 0.1° / s and there is no continuous fluctuation, the baseline threshold remains unchanged; when the core characteristic parameters show a decaying trend, the safety level threshold is raised to 35%; when the risk escalates, the corresponding response is triggered: the warning level triggers the initial adjustment of the damping actuator, and the emergency level triggers the maximum damping force output and linkage with the braking system. The adaptive damping control parameters are calculated based on the instability risk level, core characteristic parameters, and equipment operating conditions. The damping force adjustment accuracy is optimized using a formula, specifically: in, For the target damping force, The damping coefficient is... This represents the probability value of instability risk. This represents the difference between the current core feature parameters and the normal range. This represents the maximum permissible deviation of the feature parameter. This is a working condition correction factor, with a value of 0.8 for no-load conditions and a value of 1.2 for full-load conditions; The optimal adjustment rate and range of action are calculated by combining the risk level gain and the physical limit constraint of the actuator. The range of action is divided into a linear range and an angular range. The linear range is applicable to tilt and displacement instability, while the angular range is applicable to vibration and torsional instability. The formula for calculating the optimal adjustment rate is as follows: in, For damping adjustment rate, Based on the adjustment rate, This represents the probability value of instability risk. For the rate of unstable development, This is the critical value for the rate of instability development. This is the correction coefficient for the rate of change of the characteristic parameter. The final output is the damping adjustment rate to the damper actuator. This is the risk level gain coefficient. This represents the maximum physical adjustment rate of the damper. The damping actuator drive control converts control parameters into standardized commands, which are transmitted to the hydraulic or electromagnetic damper via a dual-bus redundant CAN bus. Through the damper's targeted structural design, the following actions are performed: For equipment tilting, the tilt-linked valve core switches the conduction level, proportionally adjusting the opening of the oil channel on the corresponding tilted side to create a reverse support force to counteract the tilting trend; for vibration suppression, the elastic coefficient of the elastic throttle plate is dynamically changed to adapt to the equipment vibration frequency in the 10Hz-100Hz frequency band, attenuating the vibration amplitude through the synergistic effect of oil viscous damping and throttle plate elastic damping; for sudden displacement, the high-pressure accumulator instantly releases high-pressure oil, cooperating with the fast-response valve assembly to output maximum damping force, curbing the further expansion of sudden displacement, thereby suppressing equipment tilting, vibration, and sudden displacement. Real-time feedback and optimization of control effects; collection of post-control data; quantitative analysis of the improvement effect of single indicators and construction of a comprehensive evaluation index; iterative optimization of the model and control parameters using gradient descent algorithm to form closed-loop management and control. Remote monitoring and emergency linkage transmit equipment operating status and other data to a remote platform in real time. The platform supports multi-device management and data traceability. In case of emergency instability, the linkage braking system will stop the machine within 200ms and provide early warning through multiple channels.

2. The method for transient instability identification and adaptive damping control of lifting equipment according to claim 1, characterized in that, It also includes a transient instability trend prediction step, which calculates the instability development rate using a formula, the specific formula being: in, For the rate of unstable development, This represents the change in the probability of instability risk per unit time. For time intervals, This refers to the comprehensive change of core feature parameters per unit time. This represents the change in equipment displacement per unit time. For risk probability weights, For feature parameter weights, As the displacement change weight, and .

3. The method for transient instability identification and adaptive damping control of lifting equipment according to claim 1, characterized in that, It also includes an environmental interference compensation step. Environmental sensors collect wind speed, temperature, humidity and ground vibration data in real time. A multivariate linear regression equation is constructed with environmental factors as independent variables and core data interference deviation as dependent variables. The coefficients are optimized by least squares method and the goodness of fit R²≥0.9 is verified to generate compensation coefficients. The interference deviation calculated by the model is subtracted from the original value of the core data to complete the correction. The compensation processing delay does not exceed 5ms.

4. The method for transient instability identification and adaptive damping control of lifting equipment according to claim 1, characterized in that, In the multi-dimensional transient data acquisition process, the tilt sensor has a measurement range of -30° to 30° with an accuracy of ±0.05°; the vibration sensor uses a triaxial accelerometer with a measurement range of ±10g and a resolution of 0.001g; the wind speed sensor has a measurement range of 0-60m / s with an accuracy of ±0.1m / s; the load sensor has a measurement range of 0-500t with an accuracy of ±0.5%FS; and the displacement sensor uses a laser displacement sensor with a measurement range of 0-500mm and an accuracy of ±0.01mm.

5. The method for transient instability identification and adaptive damping control of lifting equipment according to claim 1, characterized in that, In the instability feature extraction and preprocessing steps, the time domain analysis uses a sliding window method with 50 sampling points, and the frequency domain analysis uses fast Fourier transform to extract the characteristic frequencies in the 10Hz-100Hz frequency band; normalization maps the data to the 0-1 interval; correlation analysis removes feature parameters with an absolute value of correlation coefficient lower than 0.3, and the core features retained include the tilt angle change rate, vibration peak value, wind speed pulsation coefficient, load fluctuation amplitude, and displacement abrupt change characteristic value.

6. The method for transient instability identification and adaptive damping control of lifting equipment according to claim 1, characterized in that, In the drive control steps of the damping actuator, the damping force adjustment range of the hydraulic damper is 0-5000N, and the adjustment rate is 100N / ms; the damping force adjustment range of the electromagnetic damper is 0-3000N, and the adjustment rate is 150N / ms; the control command transmission delay does not exceed 20ms, and the damping actuator response delay does not exceed 50ms.

7. The method for transient instability identification and adaptive damping control of lifting equipment according to claim 1, characterized in that, The remote monitoring platform supports centralized management of multiple devices, and the platform data storage time is no less than one year. It supports querying historical data. Emergency alarm methods include audible and visual alarms, SMS notifications, and APP push notifications. The alarm response time is no more than 10 seconds.

8. A transient instability identification and adaptive damping control system for lifting equipment, used to implement the transient instability identification and adaptive damping control method for lifting equipment as described in any one of claims 1-7, characterized in that, Includes the following modules: Data acquisition module: It consists of tilt angle, vibration, wind speed, load, displacement sensors and data transmission unit. The sensors are deployed in key parts of the equipment to collect multi-dimensional operating data in real time. The data transmission unit adopts 5G+edge computing technology. Feature processing module: It receives data transmitted from the data acquisition module, performs feature extraction, normalization, correlation analysis and redundant data removal, and outputs the core feature parameter sequence. The module has a built-in data processing algorithm library and supports dynamic algorithm updates and optimization. Instability identification module: Based on long short-term memory neural network, it includes model training unit and identification inference unit. The model training unit trains and optimizes model parameters through historical data and simulated data. The identification inference unit inputs the core feature parameter sequence and outputs the transient instability risk probability value. Risk assessment module: It presets three levels of instability risk thresholds, receives the risk probability value output by the instability identification module, dynamically adjusts the thresholds based on the changing trends of characteristic parameters, determines the instability risk level and triggers the corresponding response signal, and supports manual adjustment and automatic optimization of the thresholds; Parameter calculation module: Based on the risk level output by the risk assessment module, the core feature parameters output by the feature processing module, and the equipment operating condition data, calculate the optimal damping control parameters, including the optimal damping force, the optimal adjustment rate and range of action calculated by combining the risk level gain and the physical limit constraint of the actuator, and has a built-in parameter calculation algorithm and operating condition adaptation rule library. Drive control module: Composed of control command conversion unit and CAN bus communication unit, it converts the control parameters output by parameter calculation module into standardized control commands, which are transmitted to damping actuator via CAN bus to drive damper to adjust damping state in real time. The module supports control command verification and retransmission mechanism. Feedback optimization module: It acquires the equipment operation data after control through the data acquisition module, analyzes the control effect, uses the gradient descent algorithm to iteratively optimize the instability identification model parameters and damping control parameters, generates an optimization report and transmits it to the relevant modules; Remote monitoring module: includes a data storage unit, a visualization unit, an alarm unit, and an emergency linkage unit. The data storage unit stores equipment operation data, control commands, and optimization records. The visualization unit presents data in the form of charts. The alarm unit issues an alarm signal when the risk level escalates. The emergency linkage unit activates the equipment braking system in emergency situations.

9. A transient instability identification and adaptive damping control system for lifting equipment according to claim 8, characterized in that, The data acquisition module employs a redundant sensor design, with two sets of identical sensors installed in parallel at key locations. The validity of the data is verified by comparing the data using a data fusion algorithm. When the data deviation between the two sets of sensors exceeds 5%, the system automatically switches to the backup sensor and issues a fault alarm.

10. A transient instability identification and adaptive damping control system for lifting equipment according to claim 8, characterized in that, The long short-term memory neural network of the instability identification module contains 3 hidden layers, each with 128 hidden units. The initial learning rate is set to 0.001, and the learning rate is adjusted using an adaptive moment estimation optimization algorithm. After the model is trained, the recognition accuracy is no less than 95%, and the AUC value on the test set is no less than 0.

98.

11. A transient instability identification and adaptive damping control system for lifting equipment according to claim 8, characterized in that, The control command conversion unit of the drive control module supports multiple damper protocol adaptations and is compatible with different types of actuators. The CAN bus communication unit adopts a dual-bus redundant design.

12. The transient instability identification and adaptive damping control system for lifting equipment according to claim 8, characterized in that, The emergency linkage unit of the remote monitoring module and the equipment braking system adopt a dual linkage method of hard wire connection and wireless communication. The hard wire connection ensures rapid transmission of control commands in emergency situations, while the wireless communication serves as a backup. When an emergency instability risk occurs, the braking system receives the command and performs an emergency stop operation within 200ms, cutting off the power output of the equipment.

Images