Intelligent wind speed detection and operation control system for tower crane

Abstract

The invention relates to the field of crane intelligent control, in particular to a tower crane wind speed intelligent detection and operation control system which comprises a data collection module, an intelligent control analysis module and an energy recovery module which are in communication connection, and the data collection module is used for obtaining wind data and tower crane state data in real time through a sensor; the intelligent control analysis module is used for realizing wind power prediction and guaranteeing the operation safety of the crane through risk early warning and safety control; the intelligent control analysis module comprises a wind power prediction unit, an early warning unit and a control unit; the energy recovery module is used for conducting electric energy conversion on natural wind power through a wind driven generator so as to meet the requirement for power supply of low-energy-consumption electric equipment in the tower crane. The wind power data can be monitored in real time, the future short-term and long-term wind power conditions can be predicted, and reliable judgment can be conveniently provided for wind power dangerous case early warning.

Description

Technical Field

[0001] The invention relates to the field of intelligent control of cranes, in particular to an intelligent wind speed detection and operation control system for tower cranes. Background Art

[0002] With the continuous development of the construction industry, tower cranes play a vital role in building construction. However, strong winds pose a serious threat to the safe operation of tower cranes. Traditional tower cranes often rely on the operator's experience and limited meteorological information when dealing with wind disaster risks, lacking accurate wind forecasts and effective safety control measures.

[0003] At present, although some tower cranes are equipped with wind speed and direction sensors, they can only provide real-time wind data and cannot accurately predict future wind conditions. At the same time, the judgment and warning of wind force are relatively simple and cannot comprehensively consider multiple factors, resulting in limited safety protection capabilities of tower cranes in strong wind weather. In addition, the electrical equipment of tower cranes usually relies on external power supply, and in some cases, the power supply may be unstable or interrupted, affecting the normal operation of the crane. Summary of the invention

[0004] To solve the above problems, the present invention provides a tower crane wind speed intelligent detection and operation control system, which can monitor wind data in real time and predict future short-term and long-term wind conditions, so as to provide reliable judgment for wind hazard warning.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] A tower crane wind speed intelligent detection and operation control system includes a communication-connected data collection module, an intelligent control analysis module, and an energy recovery module, wherein:

[0007] The data collection module is used to obtain wind data and tower crane status data in real time through sensors;

[0008] The intelligent control analysis module is used to realize wind force prediction and ensure the safety of crane operation through risk warning and safety control; the intelligent control analysis module includes: wind force prediction unit, warning unit and control unit, among which:

[0009] The wind prediction unit is used to predict the upcoming wind conditions through a prediction model based on the wind data collected by the data collection module in a distributed architecture environment;

[0010] The early warning unit is used to classify the wind danger and issue an early warning of the corresponding level according to the wind data, weather forecast information and predicted wind conditions;

[0011] The control unit is used to control the tower crane to adjust and maintain a safe posture according to the warning issued by the warning unit;

[0012] The energy recovery module is used to convert natural wind power into electrical energy through a wind turbine to meet the power supply needs of low-energy-consuming electrical equipment in tower cranes.

[0013] The sensor comprises:

[0014] Wind direction sensor and wind speed sensor installed on the top of the tower crane, height sensor installed next to the lifting drum or wire rope guide device, angle sensor installed at the connection between the boom and the tower body or on the rotating mechanism of the boom, tilt sensor installed on the tower body, and tension sensor installed on the lifting wire rope.

[0015] Furthermore, in the distributed architecture environment, the wind conditions to be generated are predicted by a prediction model based on the wind data collected by the data collection module; the following steps are included:

[0016] A1: Use time series algorithms to analyze wind data trends, seasonality, and periodicity characteristics on historical wind speed and direction data to predict short-term wind trends;

[0017] A2: Use machine learning algorithms to combine wind data and the location of tower cranes to predict wind conditions and determine long-term wind trends;

[0018] A3: The results of short-term wind trend and long-term wind trend are integrated through weighted fusion method to obtain the final judgment.

[0019] Further, the method of analyzing the trend, seasonality and periodicity characteristics of wind data by a time series algorithm and predicting the short-term wind trend according to a mathematical model comprises the following steps:

[0020] A11: Organize the historical wind speed and direction data into a time series format, perform a stationary test on the data through the ADF test, and determine the order of the autoregressive integrated moving average model based on the characteristics of the data and experience;

[0021] A12: Fit the autoregressive integrated moving average model with the training data, diagnose and verify the fitted data, judge the fitting effect of the model by analyzing the residuals of the model, and perform Ljung-Box test on the independence of the residuals. Divide the historical data into training set and test set according to the cross-validation method, fit the model with the training set, and evaluate the model performance with the test set. Judge the generalization ability of the model by comparing the performance of the model on the training set and the test set.

[0022] A13: The wind speed and direction in the short term in the future are predicted based on the fitted autoregressive integrated moving average model. The prediction results are given in the form of numerical values ​​or probability distribution.

[0023] Furthermore, the order of the autoregressive integrated moving average model includes: the value of the autoregressive term, the value of the difference order and the value of the moving average term, specifically: the value range of the autoregressive term and the difference order is preliminarily determined by observing the autocorrelation function and the partial autocorrelation function, the difference order is determined according to the number of times the data is stabilized by the difference processing, and the optimal parameter combination is determined by the AIC information criterion.

[0024] Furthermore, the method of predicting wind conditions by combining wind data and the geographical location of the tower crane with a machine learning algorithm comprises the following steps:

[0025] A21: Collect historical wind speed and direction data and related feature data, organize these data into matrices or data tables, and perform feature engineering on the data, including feature extraction, feature selection, and data normalization;

[0026] A22: The data set is divided into a training set and a test set. The random forest algorithm constructs multiple decision trees by randomly selecting samples and features. In the process of constructing the decision tree, each node is split according to a randomly selected feature until the predetermined stop condition is reached. The random forest algorithm combines the prediction results of the decision trees and obtains the final prediction result by majority voting.

[0027] A23: The model performance is improved by adjusting the parameters of the random forest model, and the optimal parameter combination is determined by grid search. The long-term wind trend in the future is judged by the trained random forest model. The random forest model outputs the category label of the wind trend or the probability value of the wind trend.

[0028] Furthermore, the results of the short-term wind trend and the long-term wind trend are fused by a weighted fusion method, including: standardizing the short-term wind trend and the long-term wind trend judgment results according to Z-score standardization, calculating the variance of the short-term wind trend and the long-term wind trend judgment results, dynamically adjusting the weights of the short-term wind trend and the long-term wind trend according to the statistical characteristics of the data, fusing them by weighted summation, evaluating the fused results by backtesting historical data, and adjusting the weight allocation principle and weight value according to the evaluation results.

[0029] Furthermore, the step of issuing an early warning based on the wind danger classification according to the judgment of the wind prediction unit, the wind data and the weather forecast includes:

[0030] When the real-time wind speed is close to the normal working limit wind speed of the tower crane, and the wind prediction unit determines that the wind speed will further increase in the short term in the future, and the weather forecast also shows that strong winds may occur in the future, a level 1 warning is issued;

[0031] When the real-time wind speed reaches the normal working limit of the tower crane, and the wind prediction unit and weather forecast both indicate that the wind speed may continue or further increase, a Level 2 warning is issued;

[0032] When the real-time wind speed exceeds the normal operating limit wind speed of the tower crane, or the wind prediction unit determines that the future wind speed will cause serious danger to the tower crane, and the weather forecast shows that extremely severe wind conditions are about to occur, a level 3 warning will be issued.

[0033] Furthermore, according to the warning issued by the early warning unit, the tower crane is controlled to adjust and maintain a safe posture, including: when the early warning unit issues a warning that requires stopping the operation, the control system immediately sends a stop operation instruction to each operating component of the crane, clears the load on the hook, and ensures that the lifting, slewing and luffing mechanisms stop running, specifically: by controlling the lifting motor to make the hook reach the maximum limit and fix the hook, controlling the slewing mechanism, adjusting the boom to the downwind direction and locking the slewing mechanism, and monitoring the tower inclination, hook position and boom angle at the same time.

[0034] Furthermore, the intelligent control analysis module further includes a wind imaging unit, which is used to perform the following steps:

[0035] B1: Continuously receive data from the data collection module, and simultaneously obtain wind prediction data from the wind prediction unit, and establish a three-dimensional rectangular coordinate system with the tower crane as the center; in the three-dimensional rectangular coordinate system, the X-axis and the Y-axis represent the horizontal direction, and the Z-axis represents the height direction;

[0036] B2: a solid arrow is placed in the space to represent the data of the real-time receiving data collection module, and a dotted arrow is placed in the space to represent the wind prediction data of the wind prediction unit; wherein the direction of the arrow represents the wind direction, and the length of the arrow represents the wind speed;

[0037] B3: Generate dynamic images through OpenGL graphics drawing tools, generate wind images through visualization models, determine the image update frequency according to the time interval, and realize interactive functions by scaling, rotating and translating the image.

[0038] The beneficial effects of the present invention are:

[0039] 1. By installing wind speed and direction sensors, and using the autoregressive integral moving average model and random forest algorithm to predict and judge wind conditions, it is possible to predict the upcoming wind conditions in advance, providing operators with enough time to take appropriate safety measures. For example, before strong winds come, the hook is raised to the highest position and close to the tower body in time, and the boom is parked downwind and parallel to the wind direction, effectively reducing the risk of damage or accidents to the tower crane in strong winds. A three-level warning for wind hazards enables operators and relevant managers to take targeted measures according to different warning levels, further improving the safety of tower cranes in various wind conditions.

[0040] 2. The distributed architecture and two complementary algorithms are used for progressive data processing, which improves the accuracy and reliability of wind prediction and judgment. At the same time, comprehensive consideration of real-time wind data, weather forecasts and wind prediction results makes the warning more scientific and reasonable, reducing the possibility of false alarms and missed alarms. Small wind turbines power low-energy electrical equipment in tower cranes, increasing the reliability of power supply. In the event of failure or instability of the external power supply, the normal operation of some key equipment can still be guaranteed, improving the overall reliability of the tower crane.

[0041] 3. The entire system has achieved intelligent management, from data collection, wind prediction, risk warning to safety control and power supply management, all of which are carried out in an automated manner, reducing the interference of human factors and improving management efficiency and accuracy. Operators can intuitively understand the real-time and predicted conditions of wind power, as well as the safety posture and power supply status of tower cranes through the display screen, making it easier to make correct decisions in a timely manner and improving the convenience and safety of operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a module diagram of the tower crane wind speed intelligent detection and operation control system.

[0043] Figure 2 It is a flow chart of the wind prediction unit.

[0044] Figure 3 is a flow chart of the wind imaging unit. DETAILED DESCRIPTION

[0045] See also Figure 1 , Figure 2 and Figure 3 As shown, the present invention relates to a tower crane wind speed intelligent detection and operation control system, including a communication-connected data collection module, an intelligent control analysis module and an energy recovery module, wherein:

[0046] The data collection module is used to obtain wind data and tower crane status data in real time through sensors;

[0047] The intelligent control analysis module is used to realize wind force prediction and ensure the safety of crane operation through risk warning and safety control; the intelligent control analysis module includes: wind force prediction unit, warning unit and control unit, among which:

[0048] The wind prediction unit is used to predict the upcoming wind conditions through a prediction model based on the wind data collected by the data collection module in a distributed architecture environment;

[0049] The early warning unit is used to classify the wind danger and issue an early warning of the corresponding level according to the wind data, weather forecast information and predicted wind conditions;

[0050] The control unit is used to control the tower crane to adjust and maintain a safe posture according to the warning issued by the warning unit;

[0051] The energy recovery module is used to convert natural wind power into electrical energy through a wind turbine to meet the power supply needs of low-energy-consuming electrical equipment in tower cranes.

[0052] Specifically, in the energy recovery module, the low-energy-consuming electrical equipment includes: a lighting system, a control system display screen or some sensors. Due to the working environment and installation conditions of the tower crane, a vertical axis wind turbine or a horizontal axis wind turbine is selected; according to the power requirements of the low-energy-consuming electrical equipment, a wind turbine with appropriate output power is determined, and the size, weight, reliability and maintenance requirements of the wind turbine are determined according to the actual situation of the tower crane. The wind turbine is fixed at a higher position on the top of the tower crane, the boom or the tower body to ensure that the installation position does not affect the structural stability and operational safety of the tower crane, and the distance between other equipment is determined to avoid mutual interference. The wind turbine is connected to a rectifier and an inverter. The rectifier converts the AC power generated by the wind turbine into DC power, and the inverter can convert the DC power into AC power suitable for low-energy-consuming electrical equipment, ensuring that it can efficiently convert the output power of the wind turbine into the required voltage and frequency. In order to provide a stable power supply when the wind power is insufficient or the demand of the electrical equipment exceeds the output of the wind turbine, a battery pack is configured. Ensure the safety and reliability of power distribution through distribution boxes, circuit breakers, wires and other equipment, and set up power distribution according to the power requirements and priorities of different equipment. Install a control system to monitor and control wind turbines, power conversion equipment and power distribution systems. When the wind turbine output power is insufficient, it automatically switches to battery power supply; when the battery power is too low, an alarm can be issued and corresponding energy-saving measures can be taken.

[0053] Furthermore, the sensor comprises:

[0054] Wind direction sensor and wind speed sensor installed on the top of tower crane, height sensor installed beside hoisting drum or wire rope guide, angle sensor installed on the connection between boom and tower or on the rotating mechanism of boom, tilt sensor installed on tower, tension sensor installed on hoisting wire rope. Pre-process the data acquired by the sensors.

[0055] Furthermore, in the distributed architecture environment, the wind conditions to be generated are predicted by a prediction model based on the wind data collected by the data collection module; the following steps are included:

[0056] A1: Use time series algorithms to analyze wind data trends, seasonality, and periodicity characteristics on historical wind speed and direction data to predict short-term wind trends;

[0057] A2: Use machine learning algorithms to combine wind data and the location of tower cranes to predict wind conditions and determine long-term wind trends;

[0058] A3: The results of short-term wind trend and long-term wind trend are integrated through weighted fusion method to obtain the final judgment.

[0059] Specifically, the distributed architecture environment configures the master node and slave nodes of Spark, stores the collected data in HDFS, continuously writes real-time data to HDFS, and submits the compiled Spark application to the Spark cluster for execution. Monitor the running status of Spark jobs, including task progress, resource usage, and performance indicators. Optimize based on the monitoring results, including adjusting the parallelism, memory allocation, and data partitioning of the job to improve the efficiency and performance of the job.

[0060] Further, the method of analyzing the trend, seasonality and periodicity characteristics of wind data by a time series algorithm and predicting the short-term wind trend according to a mathematical model comprises the following steps:

[0061] A11: Organize the historical wind speed and direction data into a time series format, perform a stationary test on the data through the ADF test, and determine the order of the autoregressive integrated moving average model based on the characteristics of the data and experience;

[0062] A12: Fit the autoregressive integrated moving average model with the training data, diagnose and verify the fitted data, judge the fitting effect of the model by analyzing the residuals of the model, and perform Ljung-Box test on the independence of the residuals. Divide the historical data into training set and test set according to the cross-validation method, fit the model with the training set, and evaluate the model performance with the test set. Judge the generalization ability of the model by comparing the performance of the model on the training set and the test set.

[0063] A13: The wind speed and direction in the short term in the future are predicted based on the fitted autoregressive integrated moving average model. The prediction results are given in the form of numerical values ​​or probability distribution.

[0064] Specifically, in step A11, the time series data is arranged in chronological order for subsequent analysis and modeling. Stationarity is an important prerequisite for time series analysis. If the data is not stationary, the model will be inaccurate. If the data is not stationary, differential processing is required to make it stationary.

[0065] In step A12, the training data is a part of the historical data used to estimate the parameters of the model, and the model is fitted using a time series analysis library in a statistical software or programming language. During the fitting process, attention is paid to the convergence and stability of the model to ensure that the model can accurately describe the characteristics of the data. The fitted model is diagnosed and verified to ensure the accuracy and reliability of the model. The residual is the difference between the actual value and the model predicted value. If the residual follows a normal distribution with a mean of zero and a constant variance, the model fit is good. In the test of the independence of the residuals, if the residuals are independent, it means that the model does not miss important information.

[0066] Furthermore, the order of the autoregressive integrated moving average model includes: the value of the autoregressive term, the value of the difference order and the value of the moving average term, specifically: the value range of the autoregressive term and the difference order is preliminarily determined by observing the autocorrelation function and the partial autocorrelation function, the difference order is determined according to the number of times the data is stabilized by the difference processing, and the optimal parameter combination is determined by the AIC information criterion.

[0067] Specifically, the orders p, d, and q of the autoregressive integrated moving average model are determined, specifically: the p value representing the number of autoregressive terms, the d value representing the number of differences, and the q value representing the number of moving average terms in the autoregressive integrated moving average model are determined. A more detailed explanation is: the autoregressive integrated moving average model is a commonly used model for time series forecasting and analysis, in which the accurate determination of the three orders p, d, and q is crucial to the accuracy and effectiveness of the model. p determines the degree of dependence of the model on past values, d reflects the number of times the data is differentiated to make it stable, and q represents the number of terms in the moving average part, which together determine the structure and performance of the model. The autocorrelation function represents the correlation between the time series and its own lagged values, and the partial autocorrelation function represents the correlation between the time series and its own lagged values ​​after removing the influence of the intermediate lagged values.

[0068] Furthermore, the method of predicting wind conditions by combining wind data and the geographical location of the tower crane with a machine learning algorithm comprises the following steps:

[0069] A21: Collect historical wind speed and direction data and related feature data, organize these data into matrices or data tables, and perform feature engineering on the data, including feature extraction, feature selection, and data normalization;

[0070] A22: The data set is divided into a training set and a test set. The random forest algorithm constructs multiple decision trees by randomly selecting samples and features. In the process of constructing the decision tree, each node is split according to a randomly selected feature until the predetermined stop condition is reached. The random forest algorithm combines the prediction results of the decision trees and obtains the final prediction result by majority voting.

[0071] A23: The model performance is improved by adjusting the parameters of the random forest model, and the optimal parameter combination is determined by grid search. The long-term wind trend in the future is judged by the trained random forest model. The random forest model outputs the category label of the wind trend or the probability value of the wind trend.

[0072] Specifically, random forest is an ensemble learning algorithm based on decision trees. It improves the accuracy and stability of prediction by constructing multiple decision trees and combining their prediction results. In step A31, the relevant feature data includes: time, geographical location, temperature, air pressure, etc. In the data table, each row represents a sample and each column represents a feature. Feature extraction is to extract useful features from the original data, such as calculating the statistical features such as the mean, standard deviation, maximum, minimum and other statistical features of wind speed, or extracting the angular features of wind direction. Feature selection is to select the features that are most helpful for wind prediction, remove irrelevant or redundant features, and select important features through the feature importance evaluation of random forest. Data normalization is to scale the feature values ​​of the data to a specific range to improve the performance of the algorithm, such as data normalization through methods such as maximum and minimum normalization and mean variance normalization.

[0073] In step A23, the parameters of the random forest model are adjusted, and the main parameters include the number of decision trees, the maximum depth, the minimum number of sample splits, etc. In addition to grid search, random search can also be used to find the optimal parameter combination. In the process of adjusting the parameters, overfitting needs to be avoided to ensure that the model has good generalization ability. The long-term wind trend in the future is judged by the trained random forest model, such as whether the wind will increase or decrease in the future. The random forest model can output the category label of the wind trend, such as "increase", "weaken" or "stable", and can also output the probability value of the wind trend to better evaluate the uncertainty of the prediction.

[0074] Furthermore, the results of the short-term wind trend and the long-term wind trend are fused by a weighted fusion method, including: standardizing the short-term wind trend and the long-term wind trend judgment results according to Z-score standardization, calculating the variance of the short-term wind trend and the long-term wind trend judgment results, dynamically adjusting the weights of the short-term wind trend and the long-term wind trend according to the statistical characteristics of the data, fusing them by weighted summation, evaluating the fused results by backtesting historical data, and adjusting the weight allocation principle and weight value according to the evaluation results.

[0075] Specifically, the weights are dynamically adjusted according to the statistical characteristics of the data. For example, the variance or standard deviation of the short-term prediction and long-term trend judgment results is calculated. Results with larger variances are given lower weights, and results with smaller variances are given higher weights. This can make the weight distribution more reasonable and improve the accuracy of the fusion results. In order to facilitate weighted summation, the short-term wind forecast results and long-term wind trend judgment results need to be standardized. Z-score standardization converts the data into a standard normal distribution with a mean of 0 and a standard deviation of 1.

[0076] Assume that the short-term wind force forecast result is S, the long-term wind force trend judgment result is L, and the weight of the short-term forecast result is ω s , the weight of the long-term trend judgment result is ω L , then the fused result F can be calculated by the following formula: F = w S ×S+w L ×L; the sum of the weight of the short-term prediction result and the weight of the long-term trend judgment result is 1. Evaluate the fused results, judge the accuracy and reliability, backtest through historical data, compare the difference between the fusion results and the actual wind conditions, calculate error indicators such as mean absolute error, root mean square error, etc., analyze the stability and trend of the fusion results, and observe whether they are consistent with the actual wind changes. If the fusion results fluctuate greatly or do not match the actual situation, it is necessary to further adjust the weight distribution or improve the algorithm. According to the evaluation results, adjust the weight distribution principles and specific weight values, try different weight combinations, observe the changes in the fusion results, and find the weight distribution scheme that best suits the current data characteristics and actual needs. Or improve and optimize the algorithm, such as adjusting the parameters of the autoregressive integrated moving average model or the random forest algorithm to improve the accuracy of short-term prediction and long-term trend judgment.

[0077] Furthermore, the step of issuing an early warning based on the wind danger classification according to the judgment of the wind prediction unit, the wind data and the weather forecast includes:

[0078] When the real-time wind speed is close to the normal working limit wind speed of the tower crane, and the wind prediction unit determines that the wind speed will further increase in the short term in the future, and the weather forecast also shows that strong winds may occur in the future, a level 1 warning is issued;

[0079] When the real-time wind speed reaches the normal working limit of the tower crane, and the wind prediction unit and weather forecast both indicate that the wind speed may continue or further increase, a Level 2 warning is issued;

[0080] When the real-time wind speed exceeds the normal operating limit wind speed of the tower crane, or the wind prediction unit determines that the future wind speed will cause serious danger to the tower crane, and the weather forecast shows that extremely severe wind conditions are about to occur, a level 3 warning will be issued.

[0081] Specifically, by continuously receiving real-time wind data from wind speed and wind direction sensors, including wind speed and wind direction information, accessing weather forecast data, obtaining wind forecast information for a period of time in the future, including maximum wind speed forecast, wind direction change trend, etc., and wind forecast results obtained by the wind forecast unit based on the autoregressive integral sliding average model and the random forest algorithm. The first-level warning, for example, when the real-time wind speed reaches 80% of the normal working limit wind speed of the tower crane, combined with the forecast and weather forecast, it is judged that the wind speed may reach the limit wind speed within the next hour, and then a first-level warning is issued. The second-level warning, for example, when the real-time wind speed is equal to the normal working limit wind speed, and it is predicted that the wind speed may continue to increase in the next few hours, and the weather forecast also confirms that strong winds continue, a second-level warning is issued at this time. The third-level warning, for example, when the real-time wind speed exceeds the normal working limit wind speed by more than 10%, or it is predicted that the wind speed will increase significantly in the short term in the future to the extent that it poses a major safety threat to the crane, and the weather forecast warns of extremely strong wind weather, a third-level warning is issued.

[0082] When the data meets the early warning conditions, the control system of the tower crane should immediately trigger the early warning signal of the corresponding level. The early warning signal is issued in a variety of ways, such as displaying the early warning level and specific information on the display screen in the tower crane operation room, and issuing an audible and visual alarm to remind the operator to pay attention. The early warning information is also sent to relevant managers and safety supervision departments through wireless communication so that timely response measures can be taken. In the first-level early warning, the operator should pay close attention to the changes in wind force, be prepared to take further measures at any time, suspend some non-critical operations, and inspect and reinforce the tower crane. In the second-level early warning, the operation of the tower crane should be stopped immediately, the hook should be raised to the highest position and close to the tower body, the boom should be parked downwind and parallel to the wind direction, and the crane should be fully inspected to ensure that all components are in a safe state. In the third-level early warning, in addition to taking the measures of the second-level early warning, the on-site personnel should be evacuated quickly to ensure the safety of the personnel, and the relevant departments should be notified to conduct emergency assessment and treatment of the tower crane to prevent major safety accidents.

[0083] Furthermore, according to the warning issued by the early warning unit, the tower crane is controlled to adjust and maintain a safe posture, including: when the early warning unit issues a warning that requires stopping the operation, the control system immediately sends a stop operation instruction to each operating component of the crane, clears the load on the hook, and ensures that the lifting, slewing and luffing mechanisms stop running, specifically: by controlling the lifting motor to make the hook reach the maximum limit and fix the hook, controlling the slewing mechanism, adjusting the boom to the downwind direction and locking the slewing mechanism, and monitoring the tower inclination, hook position and boom angle at the same time.

[0084] Specifically, the control system of the tower crane continuously monitors the signal sent by the early warning unit, and immediately initiates the corresponding emergency response procedure when receiving the second-level early warning signal. Confirm the accuracy and reliability of the early warning information, eliminate the possibility of false alarms, and ensure the authenticity of the early warning by comparing multiple data sources, such as wind speed and direction sensors, weather forecasts, and the results of wind prediction units. Ensure that the lifting, slewing, and luffing mechanisms stop running to prevent danger caused by continued operation when the wind is strong. Inform the operator to stop the current work task and be prepared to deal with emergencies. The operator should remain calm and operate according to the established emergency plan. Raise the hook to the highest position, which can reduce the swing amplitude of the hook in the wind, reduce the risk of collision with surrounding objects, and ensure that the hook is firmly fixed in the highest position to prevent accidental descent due to wind force. Determine the wind direction by reading the data of the wind direction sensor, and operate the slewing motor to make the boom parallel to the wind direction. This can reduce the force of the wind on the boom and prevent the boom from twisting or bending due to excessive wind load. Fix the slewing position and lock the slewing mechanism through a mechanical lock or an electrical lock. Check the foundation and attachment device of the tower crane to ensure that they are firm and reliable. If the foundation is found to be loose or the attachment device is abnormal, reinforcement measures should be taken immediately, or the counterweight of the tower body should be appropriately increased to improve the stability of the crane. However, the increase in counterweight should be carried out within the design allowable range of the crane to avoid overloading. The wind speed and direction sensor is used to obtain wind condition information in real time, and the various key parts of the crane, such as the inclination of the tower body, the position of the hook, the angle of the boom, etc., are monitored at the same time. If the wind continues to increase or other abnormal conditions occur, further measures should be taken in time, such as issuing a level 3 warning and initiating a higher level of emergency response procedures. Maintain a safe posture to ensure that the crane remains stable before the wind weakens. The operator should be ready to make adjustments according to the situation at any time to ensure the safety of the crane. When the wind weakens to a safe range and the early warning unit confirms that the danger has been lifted, the normal operation of the tower crane can be gradually resumed. Before resuming operation, the crane should be fully inspected to ensure that all components are operating normally and there is no damage or failure due to the influence of wind.

[0085] Furthermore, the intelligent control analysis module further includes a wind imaging unit, which is used to perform the following steps:

[0086] B1: Continuously receive data from the data collection module, and simultaneously obtain wind prediction data from the wind prediction unit, and establish a three-dimensional rectangular coordinate system with the tower crane as the center; in the three-dimensional rectangular coordinate system, the X-axis and the Y-axis represent the horizontal direction, and the Z-axis represents the height direction;

[0087] B2: a solid arrow is placed in the space to represent the data of the real-time receiving data collection module, and a dotted arrow is placed in the space to represent the wind prediction data of the wind prediction unit; wherein the direction of the arrow represents the wind direction, and the length of the arrow represents the wind speed;

[0088] B3: Generate dynamic images through OpenGL graphics drawing tools, generate wind images through visualization models, determine the image update frequency according to the time interval, and realize interactive functions by scaling, rotating and translating the image.

[0089] Specifically, the unit scale of the coordinate system is determined according to the actual situation so as to convert the wind speed and wind direction data into the position and direction information in the image. You can also choose streamlines to represent wind force, draw streamlines according to the wind direction, the direction of the streamlines represents the wind direction, and the thickness or color of the streamlines can represent the size of the wind speed. For example, the greater the wind speed, the thicker the streamlines can be or the brighter the color. The color coding uses different colors to represent different wind speed ranges, such as blue for low wind speed, yellow for medium wind speed, and red for high wind speed, and the continuous change of wind speed is represented by color gradient. The dotted arrow represents the wind force prediction data of the wind force prediction unit, specifically placing a dotted arrow in space, the direction of the dotted arrow represents the wind direction, the length of the dotted arrow can represent the size of the wind speed, and the color of the dotted arrow can also be encoded according to the wind speed to enhance the visualization effect. For the wind force prediction data, according to the predicted time and wind speed size and wind direction angle, the predicted streamlines, arrows or color coding are drawn at the corresponding position in the image, and the change of wind force over time is displayed through animation effects. The interactive functions, such as zooming, rotating, and panning, are used to observe wind images from different angles, add annotations and descriptions, and explain the meanings of various elements and color codes in the image. The generated wind dynamic images are displayed on a suitable platform, such as the display screen in the tower crane's operating room, the large screen in the monitoring center, etc., to ensure that operators and related personnel can understand the real-time and predicted conditions of wind power in a timely and intuitive manner, so as to take corresponding safety measures.

[0090] The above implementation modes are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary engineering and technical personnel in the field shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A tower crane wind speed intelligent detection and operation control system, characterized in that: It includes a data collection module, an intelligent control analysis module and an energy recovery module with communication connection, wherein: The data collection module is used to obtain wind data and tower crane status data in real time through sensors; The intelligent control analysis module is used to realize wind force prediction and ensure the safety of crane operation through risk warning and safety control; the intelligent control analysis module includes: wind force prediction unit, warning unit and control unit, among which: The wind prediction unit is used to predict the upcoming wind conditions through a prediction model based on the wind data collected by the data collection module in a distributed architecture environment; The early warning unit is used to classify the wind danger and issue an early warning of the corresponding level according to the wind data, weather forecast information and predicted wind conditions; The control unit is used to control the tower crane to adjust and maintain a safe posture according to the warning issued by the warning unit; The energy recovery module is used to convert natural wind power into electrical energy through a wind turbine to meet the power supply needs of low-energy-consuming electrical equipment in tower cranes.

2. The tower crane wind speed intelligent detection and operation control system according to claim 1 is characterized in that: The sensor comprises: Wind direction sensor and wind speed sensor installed on the top of the tower crane, height sensor installed next to the lifting drum or wire rope guide device, angle sensor installed at the connection between the boom and the tower body or on the rotating mechanism of the boom, tilt sensor installed on the tower body, and tension sensor installed on the lifting wire rope.

3. The tower crane wind speed intelligent detection and operation control system according to claim 1 is characterized in that: The method of predicting the upcoming wind conditions through a prediction model based on the wind data collected by the data collection module in a distributed architecture environment comprises the following steps: A1: Use time series algorithms to analyze wind data trends, seasonality, and periodicity characteristics on historical wind speed and direction data to predict short-term wind trends; A2: Use machine learning algorithms to combine wind data and the location of tower cranes to predict wind conditions and determine long-term wind trends; A3: The results of short-term wind trend and long-term wind trend are integrated through weighted fusion method to obtain the final judgment.

4. The tower crane wind speed intelligent detection and operation control system according to claim 3 is characterized in that: The method of analyzing the trend, seasonality and periodicity characteristics of wind data by using a time series algorithm to analyze the historical wind speed and wind direction data to predict the short-term wind trend includes the following steps: A11: Organize the historical wind speed and direction data into a time series format, perform a stationary test on the data through the ADF test, and determine the order of the autoregressive integrated moving average model based on the characteristics of the data and experience; A12: Fit the autoregressive integrated moving average model with the training data, diagnose and verify the fitted data, judge the fitting effect of the model by analyzing the residuals of the model, and perform Ljung-Box test on the independence of the residuals. Divide the historical data into training set and test set according to the cross-validation method, fit the model with the training set, and evaluate the model performance with the test set. Judge the generalization ability of the model by comparing the performance of the model on the training set and the test set. A13: The wind speed and direction in the short term in the future are predicted based on the fitted autoregressive integrated moving average model. The prediction results are given in the form of numerical values ​​or probability distribution.

5. The tower crane wind speed intelligent detection and operation control system according to claim 4 is characterized in that: The order of the autoregressive integrated moving average model includes: the value of the autoregressive term, the difference order value and the moving average term value, specifically: the value range of the autoregressive term value and the difference order value is preliminarily determined by observing the autocorrelation function and the partial autocorrelation function, the difference order value is determined according to the number of times the data is stabilized by the difference processing, and the optimal parameter combination is determined by the AIC information criterion.

6. The tower crane wind speed intelligent detection and operation control system according to claim 3 is characterized in that: The method of predicting wind conditions by combining wind data and the geographical location of the tower crane with a machine learning algorithm comprises the following steps: A21: Collect historical wind speed and direction data and related feature data, organize these data into matrices or data tables, and perform feature engineering on the data, including feature extraction, feature selection, and data normalization; A22: The data set is divided into a training set and a test set. The random forest algorithm constructs multiple decision trees by randomly selecting samples and features. In the process of constructing the decision tree, each node is split according to a randomly selected feature until the predetermined stop condition is reached. The random forest algorithm combines the prediction results of the decision trees and obtains the final prediction result by majority voting. A23: The model performance is improved by adjusting the parameters of the random forest model, and the optimal parameter combination is determined by grid search. The long-term wind trend in the future is judged by the trained random forest model. The random forest model outputs the category label of the wind trend or the probability value of the wind trend.

7. The tower crane wind speed intelligent detection and operation control system according to claim 3 is characterized in that: The results of the short-term wind trend and the long-term wind trend are fused by the weighted fusion method, including: standardizing the judgment results of the short-term wind trend and the long-term wind trend according to the Z-score standardization, calculating the variance of the judgment results of the short-term wind trend and the long-term wind trend, dynamically adjusting the weights of the short-term wind trend and the long-term wind trend according to the statistical characteristics of the data, fusing them by weighted summation, evaluating the fused results by backtesting historical data, and adjusting the weight allocation principle and weight value according to the evaluation results.

8. The tower crane wind speed intelligent detection and operation control system according to claim 1 is characterized in that: The step of issuing an early warning based on the wind danger classification according to the judgment of the wind prediction unit, wind data and weather forecast includes: When the real-time wind speed is close to the normal working limit wind speed of the tower crane, and the wind prediction unit determines that the wind speed will further increase in the short term in the future, and the weather forecast also shows that strong winds may occur in the future, a level 1 warning is issued; When the real-time wind speed reaches the normal working limit of the tower crane, and the wind prediction unit and weather forecast both indicate that the wind speed may continue or further increase, a Level 2 warning is issued; When the real-time wind speed exceeds the normal operating limit wind speed of the tower crane, or the wind prediction unit determines that the future wind speed will cause serious danger to the tower crane, and the weather forecast shows that extremely severe wind conditions are about to occur, a level 3 warning will be issued.

9. The tower crane wind speed intelligent detection and operation control system according to claim 1 is characterized in that: The control of the tower crane to adjust and maintain a safe posture according to the warning issued by the warning unit includes: when the warning issued by the warning unit requires stopping the operation, the control system immediately sends a stop operation instruction to each operating component of the crane, clears the load on the hook, and ensures that the lifting, slewing and luffing mechanisms stop running, specifically: by controlling the lifting motor to make the hook reach the maximum limit and fix the hook, controlling the slewing mechanism, adjusting the boom to the downwind direction and locking the slewing mechanism, and monitoring the inclination of the tower body, the hook position and the boom angle at the same time.

10. The tower crane wind speed intelligent detection and operation control system according to claim 1, characterized in that: The intelligent control analysis module further includes a wind imaging unit for performing the following steps: B1: Continuously receive data from the data collection module, and simultaneously obtain wind prediction data from the wind prediction unit, and establish a three-dimensional rectangular coordinate system with the tower crane as the center; in the three-dimensional rectangular coordinate system, the X-axis and the Y-axis represent the horizontal direction, and the Z-axis represents the height direction; B2: a solid arrow is placed in the space to represent the data of the real-time receiving data collection module, and a dotted arrow is placed in the space to represent the wind prediction data of the wind prediction unit; wherein the direction of the arrow represents the wind direction, and the length of the arrow represents the wind speed; B3: Generate dynamic images through OpenGL graphics drawing tools, generate wind images through visualization models, determine the image update frequency according to the time interval, and realize interactive functions by scaling, rotating and translating the image.

Images