Tower crane safety dynamic early warning method and device based on time series analysis

Abstract

The present invention provides a dynamic early warning method and device for tower crane safety based on time series analysis, including establishing a finite element model of the tower crane, obtaining the structural vibration of the tower crane under different typical working conditions, obtaining multi-source time series data corresponding to the dangerous factors of the tower crane under different typical working conditions, establishing deep learning models for different typical working conditions, training each deep learning model, inputting the structural vibration under the corresponding typical working conditions and the multi-source time series data corresponding to the dangerous factors, outputting a joint early warning index, obtaining the typical working conditions corresponding to the real-time working conditions of the tower crane, using the deep learning model of the corresponding typical working conditions for prediction, and obtaining the joint early warning index corresponding to the real-time working conditions of the tower crane. The present invention extracts key feature information based on the multi-source time series data corresponding to the structural vibration under typical working conditions and the dangerous factors, performs feature fusion, establishes a deep learning model, outputs a joint early warning index, and performs real-time dynamic monitoring of the safety of the tower crane.

Description

Technical Field

[0001] The present invention relates to the technical field of tower cranes, and in particular to a tower crane safety dynamic early warning method and device based on time series analysis. Background Art

[0002] The tower crane (i.e. tower crane) safety warning system plays a vital role in the construction site. It can monitor the operating status of the tower crane in real time, discover potential safety hazards in time, and take corresponding early warning measures to ensure the safety of construction workers and the smooth progress of the project.

[0003] In the related technology, a large number of sensors are usually set up to capture the key parameters of the tower crane during operation, such as lifting weight, height, working radius, wind conditions, boom angular displacement and support firmness, etc. These sensors can sense the operating status of the tower crane in real time and provide accurate data support for the early warning system.

[0004] However, in the actual early warning process, the staff will set corresponding thresholds for each key parameter of the tower crane construction for early warning. That is to say, in real-time monitoring, the parameters measured by each sensor are warned separately, and the various parameters cannot be integrated, resulting in the existing early warning means being relatively single, unable to achieve joint early warning, and the early warning results being inaccurate. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a dynamic early warning method and device for tower crane safety based on time series analysis, extract key feature information based on multi-source time series data corresponding to the structural vibration conditions and risk factors under typical working conditions, perform feature fusion, establish a deep learning model, output a joint early warning index, and conduct real-time dynamic monitoring of tower crane safety.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A tower crane safety dynamic early warning method based on time series analysis includes the following steps:

[0008] Establish a finite element model of the tower crane to obtain the structural vibration of the tower crane under different typical working conditions;

[0009] Obtain multi-source time series data corresponding to the risk factors of tower cranes under different typical working conditions;

[0010] Establish deep learning models for different typical working conditions, train each deep learning model, input the structural vibration conditions under the corresponding typical working conditions and multi-source time series data corresponding to the risk factors, and output the joint warning index;

[0011] The typical working condition corresponding to the real-time working condition of the tower crane is obtained, and the deep learning model of the corresponding typical working condition is used for prediction to obtain the joint early warning index corresponding to the real-time working condition of the tower crane.

[0012] Furthermore, typical working conditions of a tower crane include lifting the load as soon as it leaves the ground, restarting the lifting when the load is suspended in the air, braking to lift, braking to lower, and suddenly performing unloading.

[0013] Furthermore, the risk factors for lifting the load off the ground are the lifting weight, ground flatness, lifting moment, inclination angle X and inclination angle Y;

[0014] The risk factors for restarting the lifting when the load is suspended in the air are wind load, wind direction, lifting weight, tower crane verticality, lifting moment, inclination angle X, inclination angle Y and lifting height;

[0015] The risk factors of brake lifting are wind load, wind direction, lifting weight, tower crane verticality, lifting moment, inclination angle X, inclination angle Y and lifting height;

[0016] The risk factors of braking descent are the verticality of the tower crane, the lifting moment, the inclination angle X, and the inclination angle Y;

[0017] The risk factors for sudden unloading are wind load, wind direction, lifting weight, tower crane verticality, lifting moment, inclination angle X

[0018] and inclination angle Y.

[0019] Furthermore, the steps of establishing a finite element model of the tower crane and obtaining the structural vibration conditions of the tower crane under different typical working conditions include:

[0020] Obtain the geometric and physical parameters of the tower crane;

[0021] Establish a three-dimensional model of the tower crane based on its geometric parameters;

[0022] Perform meshing based on the three-dimensional model of the tower crane and generate a finite element model;

[0023] Assign material properties to various parts in the finite element model based on physical parameters;

[0024] The structural vibration of the tower crane under different typical working conditions is solved based on the finite element model.

[0025] Furthermore, deep learning models are established for different typical working conditions, and each deep learning model is trained. The structural vibration conditions under the corresponding typical working conditions and multi-source time series data corresponding to the risk factors are input, and the method for outputting the joint warning index is as follows:

[0026] For each typical working condition, the multi-source time series data corresponding to the structural vibration conditions and risk factors under the corresponding typical working condition are used as the data set, and the data set is preprocessed;

[0027] A deep learning model is constructed, and a fully connected layer is established to extract high-level features of the data set after data preprocessing from the hidden layer output and prepare the final output. The output layer is established, and a linear activation function is used to output the joint warning index. The sigmoid function is applied to limit the joint warning index to the range of [0, 1], indicating the warning level.

[0028] Train the constructed deep learning model.

[0029] Furthermore, data preprocessing includes missing value processing, noise removal and normalization.

[0030] Furthermore, during the training of the constructed deep learning model, mean square error or mean absolute error was selected as the loss function, and Adam and RMSprop were used as optimizers to optimize the network weights;

[0031] The dataset is divided into a training set and a validation set, the deep learning model is trained on the training set, and the hyperparameters are adjusted using the validation set.

[0032] A tower crane safety dynamic early warning device based on time series analysis, comprising:

[0033] The structural vibration condition acquisition module is used to establish the finite element model of the tower crane and obtain the structural vibration conditions of the tower crane under different typical working conditions;

[0034] A multi-source time series data acquisition module is used to obtain multi-source time series data corresponding to the risk factors of the tower crane under different typical working conditions;

[0035] The deep learning model building module is used to build deep learning models for different typical working conditions, train each deep learning model, input the structural vibration conditions under the corresponding typical working conditions and the multi-source time series data corresponding to the risk factors, and output the joint warning index;

[0036] The joint early warning module is used to obtain the typical working condition corresponding to the real-time working condition of the tower crane, and use the deep learning model of the corresponding typical working condition to make predictions, so as to obtain the joint early warning index corresponding to the real-time working condition of the tower crane.

[0037] A computer device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the above-mentioned dynamic early warning method for tower crane safety based on time series analysis is implemented.

[0038] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned dynamic early warning method for tower crane safety based on timing analysis.

[0039] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0040] The tower crane safety dynamic early warning method based on time series analysis provided by the present invention establishes different deep learning models for different typical working conditions, trains each deep learning model, inputs the multi-source time series data corresponding to the structural vibration conditions and risk factors under the corresponding typical working conditions, outputs a joint early warning index, and obtains a trained deep learning model, which can realize real-time dynamic safety early warning of the tower crane and monitor the construction process of the tower crane efficiently and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0042] Figure 1 The present invention is a flow chart of a tower crane safety dynamic early warning method based on time series analysis.

[0043] Figure 2 A schematic diagram of configuring various sensors for the tower crane of the present invention.

[0044] Figure 3 It is a schematic diagram of the tower crane safety dynamic early warning device based on time series analysis of the present invention. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0046] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0047] It should be understood that the present invention can be implemented in different forms and should not be interpreted as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to make the disclosure thorough and complete and to fully convey the scope of the present invention to those skilled in the art.

[0048] Example 1

[0049] Example 1 provides a tower crane safety dynamic early warning method based on time series analysis, such as Figure 1 As shown, the following steps are included:

[0050] Step S1: Establish a finite element model of the tower crane to obtain the structural vibration of the tower crane under different typical working conditions;

[0051] Step S2: Obtain multi-source time series data corresponding to the risk factors of the tower crane under different typical working conditions;

[0052] Step S3: establishing deep learning models for different typical working conditions, training each deep learning model, inputting the structural vibration conditions under the corresponding typical working conditions and multi-source time series data corresponding to the risk factors, and outputting a joint warning index;

[0053] Step S4: Obtain the typical working condition corresponding to the real-time working condition of the tower crane, use the deep learning model of the corresponding typical working condition to make a prediction, and obtain the joint early warning index corresponding to the real-time working condition of the tower crane.

[0054] For the construction of tower cranes, under different typical working conditions, the stress conditions of the tower cranes are different, and the risk factors are different. In the relevant technology, each key parameter in the whole process of tower crane operation is warned separately, and no warning is given separately for different typical working conditions, which leads to inaccurate warning. In actual construction, there is a large amount of valid related information between multi-source data, that is, the occurrence of safety accidents in the construction of tower cranes has multi-source data correlation; at the same time, the occurrence of safety accidents in the construction of tower cranes is a gradual process, which is closely related to the operation status of the tower crane in the previous time period, that is, the safety accidents in the construction of tower cranes have a very strong time correlation.

[0055] The tower crane safety dynamic warning method based on time series analysis provided in this embodiment establishes different deep learning models for different typical working conditions, trains each deep learning model, inputs the multi-source time series data corresponding to the structural vibration conditions and risk factors under the corresponding typical working conditions, outputs a joint warning index, and obtains a trained deep learning model, which can realize real-time dynamic safety warning of the tower crane and monitor the construction process of the tower crane efficiently and accurately.

[0056] In step S1 and step S2 of this embodiment, Figure 2 As shown, a schematic diagram of the tower crane of the present invention is provided with various sensors, and the hoisting conditions of the tower crane are obtained through the various sensors.

[0057] like Figure 2 As shown, a gravity sensor 1, a torque sensor 2, an angle sensor 3, a wind speed sensor 4, a wind direction sensor 5, a distance sensor 6, etc. are arranged on the tower crane to obtain the hoisting condition of the tower crane.

[0058] When the hoisting mechanism operates independently, the typical working conditions of the tower crane can be roughly divided into the following five types:

[0059] Lifting the load as soon as it leaves the ground: This is the first stage of the tower crane's lifting operation. At this time, the tower crane needs to overcome the weight of the material and lift it from the ground. The dynamic load generated by this working condition is relatively large, requiring the tower crane to have sufficient stability and load-bearing capacity.

[0060] As shown in Table 1, these are the risk factors under typical working conditions when the load is lifted off the ground.

[0061] Table 1 Hazard factors under typical working conditions when the load is lifted off the ground

[0062]

[0063] Restart the lifting when the load is suspended in the air: When the material is already suspended in the air, the lifting operation is restarted. In this working condition, the tower crane needs to overcome dynamic factors such as the inertia of the material itself and wind load.

[0064] As shown in Table 2, these are the risk factors under typical working conditions when the load is suspended in the air and then restarted to rise.

[0065] Table 2 Hazard factors under typical working conditions when the load is suspended and then restarted

[0066]

[0067] Braking and rising: Braking is performed during the rising process to stop the material from rising. In this condition, the tower crane needs to have good braking performance and stability to prevent the material from continuing to rise or shaking due to inertia.

[0068] As shown in Table 3, these are the risk factors under typical braking conditions.

[0069] Table 3 Hazard factors under typical braking conditions

[0070]

[0071] Braking descent: Braking is performed during the descent process to ensure smooth descent of materials. With the development of modern tower crane technology, many tower crane hoisting mechanisms have been set with a lower micro-motion descent speed gear for direct braking, so the dynamic load generated under this condition is relatively low.

[0072] As shown in Table 4, these are the risk factors under typical braking descent conditions.

[0073] Table 4 Hazard factors under typical braking conditions

[0074]

[0075] Sudden unloading: Sudden unloading when the material is suspended in the air. In this condition, the tower crane needs to quickly adjust its stability and load-bearing capacity to cope with various dynamic changes that may occur due to the sudden reduction in material weight.

[0076] As shown in Table 5, these are the risk factors under typical working conditions of sudden unloading.

[0077] Table 5 Hazard factors under typical conditions of sudden unloading

[0078]

[0079] In other words, the risk factors of tower cranes are different under different typical working conditions.

[0080] By analyzing historical tower crane inspection data and historical tower crane accident data, common accident causes and risk factors can be identified.

[0081] In step S1, the risk factors under different typical working conditions can be further determined by combining the safety accident data of the tower crane during the hoisting process.

[0082] Step S1 of this embodiment specifically includes the following steps:

[0083] Step S101: Acquire geometric parameters and physical parameters of the tower crane.

[0084] The geometric parameters include the height, arm length, size and shape of each component of the tower crane, among which the components of the tower crane include the tower body, base frame, sleeve frame, tower top, lifting arm and balance arm.

[0085] The physical parameters include the material properties of the tower crane (elastic modulus, density, Poisson's ratio, etc.), cross-sectional properties (moment of inertia, area, etc.), and the type and characteristics of the connectors.

[0086] Step S102: establishing a three-dimensional model of the tower crane based on the geometric parameters of the tower crane.

[0087] Step S103: Meshing is performed based on the three-dimensional model of the tower crane to generate a finite element model.

[0088] The geometric model is meshed to generate a finite element model. The density and type of the mesh (such as quadrilaterals, hexahedrons, triangles, etc.) should be determined according to the analysis accuracy and computing resources. Finer meshes are required for key parts (such as connections and stress concentration areas). Specifically, the connections of the tower crane mainly include the connection of the tower standard section, the connection between the boom and the tower, and the connection between the balance arm and the tower. These connections are connected by fasteners such as pins and bolts to ensure the stability and reliability between the various components of the tower crane. The stress concentration areas of the tower crane include the tower standard section joints, the roots of the boom and the balance arm, and the slewing bearing. The tower standard section joints are prone to stress concentration due to the change in cross-sectional size and the presence of bolt holes. The roots of the boom and the balance arm are the key parts connecting the tower body, which are subject to large bending moments and shear forces. The slewing bearing is a key component in the tower crane to achieve the rotation of the boom, and is subject to complex forces and moments.

[0089] Step S104: assigning material properties to various parts in the finite element model based on the physical parameters.

[0090] Define boundary conditions, including fixed constraints (such as tower base fixation), support conditions (such as the connection between the arm and the tower body), etc.

[0091] Step S105: solving the structural vibration of the tower crane under different typical working conditions based on the finite element model.

[0092] In step S2 of this embodiment, the multi-source time series data includes measurement data and environmental data.

[0093] Multi-source time series data include measurement data corresponding to risk factors, which are obtained through sensors.

[0094] Multi-source time series data also includes environmental data from the construction site, such as temperature, humidity, ground subsidence, etc. These data may affect the safe operation of the tower crane.

[0095] In this embodiment, step S3 specifically includes the following steps:

[0096] Step S301: For each typical working condition, the multi-source time series data corresponding to the structural vibration conditions and risk factors under the corresponding typical working condition are used as a data set, and data preprocessing is performed on the data set.

[0097] Data preprocessing includes missing value processing, noise removal and normalization. Specifically, since the multi-source time series data in this embodiment are all acquired by sensors, the data has a certain degree of continuity and smoothness, so interpolation techniques (such as linear interpolation and polynomial interpolation) are used to process missing values, and moving average or weighted moving average is used to smooth the noise of multi-source time series data. And divide the data set: divide the data set into a training set, a validation set, and a test set.

[0098] Step S302: Build a deep learning model, establish a fully connected layer to extract high-level features of the data set after data preprocessing from the hidden layer output, and prepare the final output, establish an output layer, use a linear activation function to output the joint warning index, and apply the sigmoid function to limit the joint warning index to the range of [0, 1] to indicate the warning level.

[0099] Step S303: training the constructed deep learning model.

[0100] Specifically, the mean square error (MSE) or mean absolute error (MAE) is selected as the loss function. Adam (Adaptive Moment Estimation), RMSprop (Root Mean Square Prop) and other optimizers are used to optimize the network weights.

[0101] Divide the dataset into a training set and a validation set, train the deep learning model on the training set, use the validation set to adjust hyperparameters (such as learning rate, number of hidden layer units, number of layers, etc.) to avoid overfitting, and monitor the loss and validation loss during training to ensure the generalization ability of the model.

[0102] In step S5 of this embodiment, the trained deep learning model is deployed in practical applications, multi-source time series data corresponding to the structural vibration conditions and risk factors under real-time working conditions are received in real time, and a joint warning index is output.

[0103] In this embodiment, in step S302, the method for extracting high-level features is specifically as follows:

[0104] Process the structural vibration to obtain the time domain characteristic signal and frequency domain characteristic signal of the structural vibration, perform wavelet transform on the time domain characteristic signal and frequency domain characteristic signal of the structural vibration, and extract the high-level characteristics of the structural vibration;

[0105] The multi-source time series data corresponding to the risk factors are processed to obtain the time domain characteristic signals and frequency domain characteristic signals corresponding to each risk factor, and the time domain characteristic signals and frequency domain characteristic signals corresponding to each factor are subjected to wavelet transform to extract the high-level features of each risk factor.

[0106] Since the occurrence of each safety accident will cause changes in multiple factors, by coupling the advanced features of each risk factor and conducting in-depth training, the output joint warning index can better reflect the degree of danger.

[0107] Example 2

[0108] Embodiment 2 provides a tower crane safety dynamic warning device based on time series analysis, such as Figure 3 As shown, including:

[0109] The structural vibration condition acquisition module is used to establish the finite element model of the tower crane and obtain the structural vibration conditions of the tower crane under different typical working conditions;

[0110] A multi-source time series data acquisition module is used to obtain multi-source time series data corresponding to the risk factors of the tower crane under different typical working conditions;

[0111] The deep learning model building module is used to build deep learning models for different typical working conditions, train each deep learning model, input the multi-source time series data corresponding to the structural vibration conditions and risk factors under the corresponding typical working conditions, and output the joint warning index;

[0112] The joint early warning module is used to obtain the typical working condition corresponding to the real-time working condition of the tower crane, and use the deep learning model of the corresponding typical working condition to make predictions, so as to obtain the joint early warning index corresponding to the real-time working condition of the tower crane.

[0113] Example 3

[0114] Embodiment 3 provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the above-mentioned tower crane safety dynamic warning method based on timing analysis is implemented.

[0115] Example 4

[0116] Embodiment 4 provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned tower crane safety dynamic warning method based on timing analysis is implemented.

[0117] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0118] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A tower crane safety dynamic early warning method based on time series analysis, characterized in that: The following steps are involved: Establish a finite element model of the tower crane to obtain the structural vibration of the tower crane under different typical working conditions. The typical working conditions of the tower crane include lifting the load off the ground, starting to lift the load when it is suspended in the air, lifting under braking, lowering under braking, and suddenly unloading. Obtain multi-source time series data corresponding to the risk factors of tower cranes under different typical working conditions, including: The risk factors for lifting the load off the ground are the lifting weight, ground flatness, lifting moment, inclination angle X and inclination angle Y; The risk factors for restarting the lifting when the load is suspended in the air are wind load, wind direction, lifting weight, tower crane verticality, lifting moment, inclination angle X, inclination angle Y and lifting height; The risk factors of brake lifting are wind load, wind direction, lifting weight, tower crane verticality, lifting moment, inclination angle X, inclination angle Y and lifting height; The risk factors of brake descent are tower crane verticality, lifting moment, inclination angle X, and inclination angle Y; The risk factors for sudden unloading are wind load, wind direction, lifting weight, tower crane verticality, lifting moment, inclination angle X and inclination angle Y; Establish deep learning models for different typical working conditions, train each deep learning model, input the structural vibration conditions under the corresponding typical working conditions and multi-source time series data corresponding to the risk factors, and output the joint warning index; The typical working condition corresponding to the real-time working condition of the tower crane is obtained, and the deep learning model of the corresponding typical working condition is used for prediction to obtain the joint early warning index corresponding to the real-time working condition of the tower crane.

2. The tower crane safety dynamic early warning method based on time series analysis according to claim 1 is characterized in that: The steps of establishing a finite element model of a tower crane and obtaining the structural vibration of the tower crane under different typical working conditions include: Obtain the geometric and physical parameters of the tower crane; Establish a three-dimensional model of the tower crane based on its geometric parameters; Perform meshing based on the three-dimensional model of the tower crane and generate a finite element model; Assign material properties to various parts in the finite element model based on physical parameters; The structural vibration of the tower crane under different typical working conditions is solved based on the finite element model.

3. The tower crane safety dynamic early warning method based on time series analysis according to claim 1 is characterized in that: Deep learning models are established for different typical working conditions, and each deep learning model is trained. The structural vibration conditions under the corresponding typical working conditions and multi-source time series data corresponding to the risk factors are input, and the method for outputting the joint warning index is as follows: For each typical working condition, the multi-source time series data corresponding to the structural vibration conditions and risk factors under the corresponding typical working condition are used as the data set, and the data set is preprocessed; Build a deep learning model, establish a fully connected layer to extract high-level features of the data set after data preprocessing from the hidden layer output, and prepare the final output, establish an output layer, use a linear activation function to output the joint warning index, and apply the sigmoid function to limit the joint warning index to the range of [0, 1] to indicate the warning level; Train the constructed deep learning model.

4. The tower crane safety dynamic early warning method based on time series analysis according to claim 3 is characterized in that: Data preprocessing includes missing value processing, noise removal and normalization.

5. The tower crane safety dynamic early warning method based on time series analysis according to claim 3 is characterized in that: During the training of the constructed deep learning model, the mean square error or mean absolute error is selected as the loss function, and Adam and RMSprop are used as optimizers to optimize the network weights; The dataset is divided into a training set and a validation set, the deep learning model is trained on the training set, and the hyperparameters are adjusted using the validation set.

6. A tower crane safety dynamic warning device based on time series analysis, characterized in that: The tower crane safety dynamic early warning method based on time series analysis applied to any one of claims 1 to 5 comprises: The structural vibration condition acquisition module is used to establish the finite element model of the tower crane and obtain the structural vibration conditions of the tower crane under different typical working conditions; A multi-source time series data acquisition module is used to obtain multi-source time series data corresponding to the risk factors of the tower crane under different typical working conditions; The deep learning model building module is used to build deep learning models for different typical working conditions, train each deep learning model, input the structural vibration conditions under the corresponding typical working conditions and the multi-source time series data corresponding to the risk factors, and output the joint warning index; The joint early warning module is used to obtain the typical working condition corresponding to the real-time working condition of the tower crane, and use the deep learning model of the corresponding typical working condition to make predictions, so as to obtain the joint early warning index corresponding to the real-time working condition of the tower crane.

7. A computer device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the tower crane safety dynamic early warning method based on time series analysis as described in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the tower crane safety dynamic early warning method based on timing analysis as described in any one of claims 1 to 5 is implemented.

Images