Tower crane collision warning method and system in building construction

Abstract

The present application relates to the technical field of electric digital data processing, in particular to a tower crane mis-collision early warning method and system in building construction, which acquires working state data of all tower cranes in current building construction; data fitting is performed on a tower crane rotating angular velocity sequence to obtain initial tower crane trajectory prediction results of each tower crane; based on a height value of a tower corresponding to each tower crane, a distance value between a hoisted cargo of the tower crane and a tower platform, a tower arm length of the tower crane and the initial tower crane trajectory prediction results, trajectory segments of each tower crane at a future time are determined; whether there is a mis-collision risk between any two adjacent tower cranes is determined according to the trajectory segments of each tower crane; when there is a collision risk, early warning is performed. That is, the scheme of the present application can timely and in advance predict the collision risk between tower cranes, so as to take effective measures to reduce the collision risk.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing, specifically to a method and system for early warning of accidental collisions with tower cranes during building construction. Background Technology

[0002] Tower cranes are an important tool in existing construction projects. When the building area is too large, multiple tower cranes are needed to work together. The more tower cranes there are, the higher the construction efficiency. However, there is a risk of collision between the tower arms of adjacent tower cranes, which can lead to construction safety risks.

[0003] Existing tower crane collision warning systems typically detect the risk of collision by monitoring the rotation position of the tower boom. When an impending collision is detected, if the distance between two tower cranes is less than a set warning distance, the operation of some tower cranes needs to be stopped immediately to avoid the collision risk. However, because the tower cranes are stopped, effective rotation adjustments cannot be made, resulting in low construction efficiency. Therefore, existing tower crane collision warning systems do not consider the possibility of subsequent work scheduling between tower cranes. In other words, although they can avoid the risk of tower crane collisions, they cannot predict the risk of collisions in a timely manner, and thus cannot take corresponding measures to reduce the risk of collisions in advance, affecting the efficiency of construction.

[0004] Therefore, it is particularly important to predict collision risks in a timely and advance manner and to take corresponding measures to reduce collision risks. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for early warning of accidental collisions of tower cranes in building construction, which solves the problem in the prior art that it is impossible to predict collision risks in a timely manner and in advance, and to take corresponding measures in advance to reduce collision risks.

[0006] This invention provides a method for early warning of accidental collisions with tower cranes during building construction, comprising the following steps:

[0007] Obtain the working status data of all tower cranes in the current construction project; the working status data includes the tower crane rotation angular velocity sequence, the tower crane's corresponding tower ratio height value, the distance between the goods lifted by the tower crane and the tower platform, and the tower crane's boom length;

[0008] Data fitting is performed on the rotational angular velocity sequences of any two adjacent tower cranes to obtain the initial tower crane trajectory prediction results for each tower crane; based on the tower ratio height value corresponding to each tower crane, the distance between the goods lifted by the tower crane and the tower platform, the tower crane boom length, and the initial tower crane trajectory prediction results, the trajectory segment of each tower crane at future moments is determined.

[0009] Based on the trajectory segments of each tower crane at future moments, determine whether there is a risk of accidental collision between any two adjacent tower cranes; when there is a risk of collision, issue an early warning.

[0010] The specific process for determining whether there is a risk of accidental collision between two adjacent tower cranes is as follows:

[0011] Randomly select trajectory segments of any two adjacent tower cranes at future times; determine whether the two trajectory segments intersect. If they intersect, then the two adjacent tower cranes are considered to be at risk of collision.

[0012] Optionally, the specific process for determining the trajectory segment of each tower crane at a future moment is as follows:

[0013] If the tower ratio height values ​​of two adjacent tower cranes are the same, then based on the current tower boom length of the two tower cranes as the radius, and on the basis of the initial tower crane trajectory prediction results, determine the arc segment when the tower cranes perform circular motion in the future, and obtain the trajectory segment of the two adjacent tower cranes.

[0014] If the tower ratio height values ​​of two adjacent tower cranes are not the same, then based on the obtained future angular velocity sequence, the lower tower crane uses the tower arm length E as the radius to obtain the arc segment when the tower crane is in circular motion, as the trajectory segment; the higher tower crane uses the distance between the goods lifted by the tower crane and the tower platform as the radius to obtain the arc segment when the goods are in circular motion, as the trajectory segment.

[0015] Optionally, the method also includes a step of calculating the collision avoidance cost of each tower crane based on its future trajectory segments, in order to schedule two adjacent tower cranes with a risk of accidental collision, specifically:

[0016] Construct an objective function for the collision avoidance cost between two adjacent tower cranes among all tower cranes;

[0017] Solve for each objective function, obtain the minimum value of all objective functions, and use the minimum value as the final collision avoidance cost for all tower cranes;

[0018] When the minimum value is less than the first set threshold and greater than or equal to the second set threshold, a false collision warning is issued, and manual dispatch is performed. When the minimum value is less than the second set threshold, the tower crane is automatically dispatched, and the tower crane operator is given an intelligent dispatch reminder. The first set threshold is greater than the second set threshold.

[0019] Optionally, the collision avoidance cost is:

[0020] G i,j =K(min({F i,j})-Fr)×H+max(F' i , F' j )

[0021] Where K represents a 0-1 step function, when min({F i,j When})-Fr≤0, K(min({F i,j When min({F) - Fr) is 0, i,j When})-Fr>0, K(min({F i,j})-Fr) is 1; Fr is a hyperparameter, Fr=2; H is a large value; F' i For the original trajectory segment F i The trajectory point at the last moment and F after adding the adjustment variable i The Euclidean distance F' between the trajectory points at the last moment. j For the original trajectory segment F j The trajectory point at the last moment and F after adding the adjustment variable j The Euclidean distance between the trajectory points at the last moment is calculated using the max() function to find the maximum value and the min() function to find the minimum value.

[0022] Optionally, the solution to each objective function is obtained by using a swarm optimization intelligent algorithm to calculate the changes in the angular velocity of all tower cranes at corresponding future times, thereby determining the value of the objective function.

[0023] Optionally, it also includes a step of obtaining an initial solution for the swarm optimization intelligent algorithm, specifically:

[0024] Obtain the historical working parameter sequences of all tower cranes and the current working parameter sequences, and construct a graph structure dataset of all historical tower cranes and graph structure data of all current tower cranes; wherein, the graph structure data in the graph structure dataset of all historical tower cranes corresponds one-to-one with the historical working state parameters; the working parameter sequences are five-dimensional data; perform one-to-one matching between the graph structure data of all current tower cranes and the graph structure dataset of all historical tower cranes, and obtain the maximum matching subgraph of the graph structure data of all current tower cranes;

[0025] Use the change in angular velocity in the maximum matching subgraph as the initial solution for the change in angular velocity of the current tower crane;

[0026] Simultaneously, the maximum matching subgraph is removed, and the graph structure data of all current tower cranes is re-matched with the graph structure dataset of all historical tower cranes after removal, until the subgraph structure matched by all tower cranes is obtained; then the adjustment variables corresponding to all the obtained matching subgraph structures are used as the initial solution of the objective function of the corresponding two adjacent tower cranes.

[0027] Optionally, during the maximum matching subgraph algorithm, when calculating the distance between two corresponding nodes in the two graph structures, the working state parameter sequences of the two tower cranes at the set time corresponding to the two nodes are merged into a set of working state parameter sequences; the set of working state parameter sequences is dimensionality reduced to obtain the eigenvalues ​​corresponding to the principal component directions of each working state parameter; the eigenvector corresponding to the largest eigenvalue is selected, and the projection value of the five-dimensional data corresponding to each working state on the eigenvector corresponding to the largest eigenvalue is obtained, thereby obtaining the projection value sequences corresponding to the two nodes respectively; and the correlation value between the two projection sequences is obtained using the Pearson correlation coefficient method, which is used as the distance between the two corresponding nodes in the two graph structures.

[0028] This invention also provides a tower crane collision warning system for building construction, comprising:

[0029] processor;

[0030] The memory stores computer instructions for early warning of accidental collisions with tower cranes during construction. When the computer instructions are executed by the processor, the system performs the aforementioned method for early warning of accidental collisions with tower cranes during construction.

[0031] The beneficial effects of this invention are as follows:

[0032] The solution of this invention can predict the working paths of multiple tower cranes in advance, provide early warning of collision risks, and reserve sufficient time for tower crane scheduling, thereby further improving the efficiency of construction.

[0033] Furthermore, by obtaining the collision avoidance cost, the solution of this invention can further determine the scheduling of tower cranes, thereby better saving costs. Attached Figure Description

[0034] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating the steps of a tower crane collision warning method during building construction according to the present invention.

[0036] Figure 2 This is a structural block diagram of a tower crane collision warning system for building construction according to the present invention. Detailed Implementation

[0037] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, describes the specific implementation, structure, features, and effects of the deep learning-based material detection method proposed according to the present invention.

[0038] This invention addresses the application scenario of multiple tower cranes operating simultaneously during building construction, where there is a risk of collision due to the simultaneous operation of multiple tower cranes. Therefore, a specific solution for a tower crane collision early warning method and system in building construction is proposed.

[0039] Figure 1 This is a flowchart illustrating the steps of a tower crane collision warning method in building construction according to the present invention.

[0040] like Figure 1 As shown, a method for early warning of accidental collisions of tower cranes during building construction according to the present invention includes the following steps:

[0041] In step S1, obtain the working status data of all tower cranes in the current building construction.

[0042] In this embodiment, during the current construction, the working status data of each tower crane includes the tower crane's rotational angular velocity value A, the corresponding tower height value C, the distance D between the goods lifted by the tower crane and the tower platform, and the tower crane's boom length E, that is, the working status data A of the i-th tower crane at time t. it C it D it E i ).

[0043] In step S2, the rotational angular velocity sequences of any two adjacent tower cranes are fitted to obtain the initial tower crane trajectory prediction results for each tower crane. Based on the tower height value of each tower crane, the distance between the goods lifted by the tower crane and the tower platform, the tower boom length of the tower crane, and the initial tower crane trajectory prediction results, the trajectory segment of each tower crane at future moments is determined.

[0044] It should be noted that the center coordinates of the tower cranes in this embodiment are two-dimensional coordinates, so all tower cranes can be analyzed in the same plane coordinate system.

[0045] Since tower cranes can only perform circular motion, it is necessary to determine the rotational angular velocity A of the current i-th tower crane. i Predicting the sequence, where A is selected from the first n time intervals. iThe value sequence is used for data prediction to obtain the initial tower crane trajectory prediction result. The time interval value is 2 seconds, but this time interval value can be adjusted in other implementation methods. The value of n can be 100, or it can be adjusted by the implementer according to the specific implementation scenario.

[0046] In obtaining A corresponding to the current time relative to the previous n times... i After the sequence, for A corresponding to the first n time steps i The sequence is fitted using the least squares method to obtain the corresponding fitting function. Substituting future times into the fitting function yields A for those future times. i The corresponding A can be obtained from the value fitting result. i The trajectory results.

[0047] For example, in this embodiment, prediction is performed at m future times to obtain the prediction results corresponding to the m future times; where m is a hyperparameter, which can be adjusted by the implementer according to the specific implementation scenario; of course, the data fitting method mentioned above can also be selected by the implementer according to the specific implementation scenario, such as curve fitting, polynomial fitting method, etc.

[0048] It should be noted that, when making predictions, this invention requires the distance D between the goods lifted by the tower crane and the tower platform to remain constant. This is because, although the distance D can affect tower crane collisions (since the D value of the relatively higher tower crane determines whether its lifted goods can collide with the boom of the relatively lower tower crane), the path that the tower crane can take for lifting changes under different construction progresses, and there is no actual movement pattern. Therefore, if the distance D changes, data prediction cannot be performed. Thus, this embodiment only addresses the case where the distance D between the goods lifted by the tower crane and the tower platform remains constant.

[0049] Meanwhile, if two adjacent tower cranes are at the same height, after obtaining the future angular velocity sequence, it is also necessary to combine the circular motion of the tower cranes, that is, to combine the circular motion formula to determine the trajectory segments of the two tower cranes at the time of the accidental collision in the future; that is, based on the obtained future angular velocity sequence, the trajectory segments (circular arc segments when the tower cranes are performing circular motion) of the future time are obtained with the current tower arm length E of the two tower cranes as the radius.

[0050] If the two tower cranes are not at the same height, after obtaining the future angular velocity sequence, it is necessary to combine the circumference of the tower cranes, i.e., to combine the circular motion formula, to determine the trajectory segments of the two tower cranes at the time of the accidental collision. Specifically, based on the obtained future angular velocity sequence, the lower tower crane uses its boom length E as the radius to obtain the trajectory segment at the future time (the arc segment of the tower crane during circular motion); while the higher tower crane uses the distance D between the hoisted goods and the tower platform as the radius to obtain the trajectory segment at the future time (the arc segment of the goods during circular motion). For example, taking the next m moments as an example, the j-th tower crane adjacent to the i-th tower crane is obtained. The premise for the i-th tower crane and the j-th tower crane to collide is that the trajectory segment F corresponding to the boom position of the i-th tower crane at the next m moments is... i The trajectory segment F corresponding to the tower arm position of the j-th tower crane at m future times. j .

[0051] In step S3, based on the trajectory segments of each tower crane at future moments, it is determined whether there is a risk of accidental collision between any two adjacent tower cranes; when there is a risk of collision, an early warning is issued.

[0052] In one embodiment, when assessing collision risk, it can randomly select trajectory segments from the future times of any two adjacent tower cranes; determine if the two trajectory segments intersect; if an intersection exists, it considers the two adjacent tower cranes to be at risk of collision. Specifically, through trajectory segment F... i and trajectory segment F j Determine if an intersection point exists. If an intersection point exists, issue a risk warning; otherwise, do not issue a risk warning. The trajectory segment F... i and trajectory segment F j The method for determining the existence of intersection points is as follows: obtain trajectory point F i All data in the dataset, if a trajectory point F exists. i The data in the data belongs to trajectory point F. j The data in the data indicates that there is an intersection between the two.

[0053] Of course, as another implementation method, trajectory segment F can also be... i and trajectory segment F j After aligning according to time, calculate the Euclidean distance between two trajectory points at the same time. If there is an Euclidean distance value less than a distance threshold, it indicates that trajectory segment F will be on the future trajectory. i and trajectory segment F j The distances between them will intersect, meaning there will be intersection points, which pose a risk of collision with the tower cranes. The smaller the distance, the higher the risk of collision.

[0054] For example, obtain the sequence of Euclidean distance values ​​{F} corresponding to the next m time steps. i,j}If min({F i,j})<Fr, it is determined that F i and F j have an intersection between them. Wherein Fr is a hyperparameter, Fr=2; of course, the value can also be adjusted by the implementer according to the specific implementation scenario.

[0055] Wherein, the position of the trajectory segment corresponding to min({F i,j}) is exactly the position of the intersection between trajectory segment F i and trajectory segment F j .

[0056] Further, when a collision may occur, in order to reduce the collision risk, scheduling can be performed between the two tower cranes to reduce the risk of false collision of tower cranes, then it is necessary to determine the scheduling degree, that is, a single tower crane may need operations such as stopping, decelerating or retreating, which further affects the operation efficiency of the tower crane. Therefore, it is necessary to calculate the anti-collision cost generated by the two tower cranes while reducing the probability of false collision. Therefore, the early warning method of the present invention further comprises the step of calculating the anti-collision cost of each tower crane according to the initial tower crane trajectory prediction result of each tower crane, so as to schedule two adjacent tower cranes with false collision risk.

[0057] Specifically, in order to reduce the probability of false collision, scheduling is performed between the two tower cranes, wherein the subsequent optimal scheduling is determined by calculating the anti-collision cost, and the specific process is as follows:

[0058] First, construct an objective function of anti-collision cost between two adjacent tower cranes among all tower cranes.

[0059] It should be noted that when constructing the objective function, it is also necessary to obtain the coordinate values of the trajectory points corresponding to the intersection, and then add m adjustment variables x to the angular velocity values corresponding to m moments of the two tower cranes respectively, which is used for iteratively solving the optimal scheduling scheme; m can be 20, 30 or more.

[0060] Wherein, taking the i-th tower crane as an example, after adjusting m angular velocity adjustment variables x, it is required that the first derivative of the angular velocity curve of m times meets the requirement of the variation of the angular velocity of the equipment.

[0061] And then calculate the value of min({F i,j}) corresponding to the i-th tower crane and the j-th tower crane after adding m initial random variables x respectively, wherein the smaller the value of min({F i,j}), the higher the risk, and obtain the maximum value of the respective distance between two trajectory points corresponding to the m-th moment before and after adjustment, which represents the time cost. The larger the value, the higher the time cost.

[0062] Specifically, if a collision is possible between the i-th tower crane and the j-th tower crane, in order to reduce the probability of a false collision, this embodiment can calculate the anti-collision cost between the i-th tower crane and the j-th tower crane as follows:

[0063] G i,j =K(min({F i,j})-Fr)×H+max(F' i , F' j )

[0064] Where K represents a 0-1 step function, when min({F i,j When})-Fr≤0, K(min({F i,j When min({F) - Fr) is 0, i,j When})-Fr>0, K(min({F i,j})-Fr) is 1; Fr is the threshold corresponding to the distance F, which is a hyperparameter; H is a large value, indicating that a large collision avoidance cost value will be obtained when there is a risk. When there is no risk, since K is a step function, H will become 0. Then, by H and K(min({F)) i,j The product of})-Fr) can be used to represent the collision avoidance cost when a collision risk occurs and the collision avoidance cost when no risk occurs; in F' i For the original trajectory segment F i The trajectory point at the last moment and F after adding the adjustment variable i The Euclidean distance between the trajectory points at the last moment is a factor. A larger value indicates a greater impact on the operating efficiency of the i-th tower crane during the collision avoidance scheduling process; a smaller value indicates a smaller impact. Where F' j For the original trajectory segment F j The trajectory point at the last moment and F after adding the adjustment variable j The Euclidean distance between the trajectory points at the last moment is a factor. The larger the value, the greater the impact on the operating efficiency of the j-th tower crane during the collision avoidance scheduling process; the smaller the value, the smaller the impact on the operating efficiency of the j-th tower crane during the collision avoidance scheduling process.

[0065] Since the two tower cranes are dispatched simultaneously, F' is taken. i and F' j The maximum value between max(F') i , F' j ), representing the maximum time cost.

[0066] G i,j The larger the value of G, the higher the cost of collision prevention between the two tower cranes. Therefore, when solving for the adjustment variables, G should be made more stable. i,j It should be small.

[0067] Then, obtain the objective function for the collision avoidance cost between any two adjacent tower cranes (or between any two tower cranes).

[0068] Secondly, each objective function is solved to obtain the minimum value of all objective functions, and the minimum value is used as the final anti-collision cost for all tower cranes.

[0069] In one embodiment, the objective function is solved by taking the current collision avoidance cost as the objective function and finding its minimum value. Specifically, this is achieved by using a swarm optimization algorithm (such as PSO or a genetic algorithm) to solve for the adjustment variables corresponding to the future angular velocities of all tower cranes. Since the specific optimization process in the swarm optimization algorithm is existing technology, it will not be elaborated further here. It should be noted that when using the swarm optimization algorithm, the initial solution for its adjustment variables can be set based on experience.

[0070] Because safety incidents can still occur if the computation time is too long during each scheduling process, a solution time cap is set to determine the number of iterations and calculate the collision avoidance cost. In this scheme, the solution time is set to 2 seconds. The specific number of iterations depends on the computing power of the device within the solution time cap. Within the 2-second solution time, the swarm optimization intelligent algorithm continuously performs iterative calculations.

[0071] Furthermore, in order to quickly obtain the optimal solution and achieve the lowest possible collision avoidance cost within the upper limit of the solution time, this scheme chooses to use the adjustment variable corresponding to the historical tower crane working state as the initial solution for the current actual tower crane working state when solving for the optimal angular velocity change. By setting an initial variable that is closer to the optimal solution, the solution efficiency is improved and the solution is closer to the global optimal solution.

[0072] Get the working status parameters of all tower cranes within n time intervals {(A i B i C i D i E i The sequence is then used; then, a single tower crane is used as a node, and the nodes corresponding to adjacent tower cranes are connected, where the edge weight after two nodes are connected is 1; thus, the graph structure data is constructed, where the method for judging adjacent tower cranes is to be determined by experienced personnel based on the position of the tower cranes.

[0073] The working status data of each tower crane includes the tower crane's rotational angular velocity value A, maximum / minimum angular velocity value B, the corresponding tower height value C, the distance between the goods lifted by the tower crane and the tower platform D, and the tower crane's jib length E, that is, the working status data of the i-th tower crane at time t {(A iB i C i D i E i )}.

[0074] This allows us to obtain the graph structure data P' at the current moment, and similarly, obtain the graph structure data sequence {P1,P2,...} corresponding to each state of the historical data.

[0075] P' is matched one-to-one with {P1,P2,...}. Using a subgraph matching degree algorithm, the maximum matching subgraph between the current working state P' of all tower cranes and all historical tower crane states in {P1,P2,...} is obtained. After obtaining the maximum matching subgraph, the solution for the optimal angular velocity change of the historical tower cranes in the subgraph corresponding to the maximum matching subgraph between all historical tower crane states in {P1,P2,...} and the current working state P' of all tower cranes is used as the initial solution for the angular velocity change of the current tower crane in the subgraph. Then, the tower cranes in the current subgraph are removed from the two graph structure data. Subgraph matching is then performed again using the two graph structure data after removing the subgraphs until all tower cranes have a matched subgraph structure. Finally, the adjustment variables corresponding to all the obtained matched subgraph structures are used as the initial solution for the objective function of the corresponding two adjacent tower cranes.

[0076] In the maximum matching subgraph algorithm, when performing maximum subgraph matching and determining the distance between two corresponding nodes in two graph structures, the working state parameters {(A) of the two tower cranes corresponding to the two nodes within n time intervals are used. i B i C i D i E i The sequence is merged into a single working state parameter {(A)}. i B i C i D i E i The sequence is then used to determine the 2n working state parameters {(A)}. i B i C i D i E iThe sequence is reduced using PCA dimensionality reduction algorithm. After obtaining the five principal component directions, the eigenvector corresponding to the largest eigenvalue is selected. The projection value of the five-dimensional data corresponding to each working state on the eigenvector corresponding to the largest eigenvalue is obtained. Then, the projection sequence corresponding to each of the two nodes is obtained. The length of the projection sequence corresponding to a single node is 100. Then, the correlation value between the two projection sequences corresponding to the two nodes is obtained by using the Pearson correlation coefficient method, which is used as the distance value between the two corresponding nodes in the two graph structures.

[0077] This allows for a fast and optimal solution to the objective function, improving efficiency and reducing the probability of false collisions.

[0078] Then, when the minimum value is less than the first set threshold and greater than or equal to the second set threshold, a false collision warning is issued, and manual scheduling is performed at this time; when the minimum value is less than the second set threshold, the tower crane is automatically scheduled, and the tower crane operator is given an intelligent scheduling reminder; wherein the first set threshold is greater than the second set threshold.

[0079] The first threshold value mentioned above is H, and the second threshold value is 0.9H. For example, if the final value of M is greater than or equal to 0.9*H, it indicates that there is still a risk of accidental collision. Here, 0.9 is a hyperparameter. Although there is no risk of accidental collision when M is less than H but greater than or equal to 0.9*H, the scheduling cost is too high and the scheduling scenario is complex, so an accidental collision warning is still issued, requiring manual scheduling. When the value of M is less than 0.9*H, no accidental collision warning is issued, but since the tower crane is scheduled automatically, intelligent scheduling reminders are sent to the tower crane operators to reduce the risk of accidental collisions caused by operator error.

[0080] It should be noted that there may be multiple tower cranes in a construction project, which requires adjusting the angular velocity of all tower cranes to achieve a lower overall collision avoidance cost. If the collision avoidance cost is too high, it means that the current tower crane scheduling is too complicated, so a false collision warning is issued to allow for manual scheduling and prevent collisions between tower cranes.

[0081] Figure 2 This is a structural block diagram of a tower crane accidental collision early warning system in building construction according to the present invention.

[0082] like Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the above-described method for early warning of tower crane collisions during building construction according to the present invention.

[0083] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0084] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions that can be stored or otherwise maintained by such a computer-readable medium.

[0085] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0086] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early warning of accidental collisions with tower cranes during building construction, characterized in that, Includes the following steps: Obtain the working status data of all tower cranes in the current construction project; the working status data includes the tower crane rotation angular velocity sequence, the tower crane's corresponding tower ratio height value, the distance between the goods lifted by the tower crane and the tower platform, and the tower crane's boom length; Data fitting is performed on the rotational angular velocity sequences of any two adjacent tower cranes to obtain the initial tower crane trajectory prediction results for each tower crane; based on the tower ratio height value corresponding to each tower crane, the distance between the goods lifted by the tower crane and the tower platform, the tower crane boom length, and the initial tower crane trajectory prediction results, the trajectory segment of each tower crane at future moments is determined. Based on the future trajectory segments of each tower crane, determine whether there is a risk of accidental collision between any two adjacent tower cranes; when a collision risk exists, issue an early warning. The specific process for determining whether there is a risk of accidental collision between two adjacent tower cranes is as follows: Randomly select trajectory segments of any two adjacent tower cranes at future times; determine whether the two trajectory segments intersect; if they intersect, then the two adjacent tower cranes are considered to be at risk of collision. It also includes the step of calculating the collision avoidance cost of each tower crane based on its future trajectory segments, in order to schedule two adjacent tower cranes with a risk of accidental collision. Specifically: Construct an objective function for the collision avoidance cost between two adjacent tower cranes among all tower cranes; Solve for each objective function, obtain the minimum value of all objective functions, and use the minimum value as the final collision avoidance cost for all tower cranes; When the minimum value is less than the first set threshold and greater than or equal to the second set threshold, a false collision warning is issued, and manual dispatch is performed. When the minimum value is less than the second set threshold, the tower crane is automatically dispatched, and the tower crane operator is given an intelligent dispatch reminder. The first set threshold is greater than the second set threshold. The cost of collision avoidance Gi,j for: Gi,j =K (my({ Fi,j })- Fr ) × H +max( F'i,F'j ) in, K Represents a 0-1 step function, { Fi,j } represents the trajectory segment after adding adjustment variables. Fi Trajectory segments after adding adjustment variables Fj The sequence of Euclidean distances formed by the Euclidean distances between any two trajectory points, when min({ Fi,j })- Fr When ≤0, K (min({ Fi,j })- Fr When min({ Fi,j })- Fr When >0, K (min({ Fi,j })- Fr K is 1; Fr is a hyperparameter, Fr=2; H is a large value, indicating that a larger collision avoidance cost value will be obtained when there is a risk, and H will become 0 when there is no risk, since K is a step function; F'i To add trajectory segments before and after adjusting variables Fi The Euclidean distance between the trajectory points at the last moment in the middle. F'j To add trajectory segments before and after adjusting variables Fj The Euclidean distance between the trajectory points at the last moment is given by the max() function, which takes the maximum value, and the min() function, which takes the minimum value.

2. The method for early warning of tower crane collisions during building construction according to claim 1, characterized in that, The specific process for determining the trajectory segments of each tower crane at future moments is as follows: If the tower ratio height values ​​of two adjacent tower cranes are the same, then based on the current tower boom length of the two tower cranes as the radius, and on the basis of the initial tower crane trajectory prediction results, determine the arc segment when the tower cranes perform circular motion in the future, and obtain the trajectory segment of the two adjacent tower cranes. If the height values ​​of the tower ratios corresponding to two adjacent tower cranes are not the same, then based on the obtained future angular velocity sequence, the lower tower crane uses the tower arm length E of the tower crane as the radius to obtain the arc segment when the tower crane is in circular motion, as the trajectory segment; the higher tower crane uses the distance between the goods lifted by the tower crane and the tower platform as the radius to obtain the arc segment when the goods are in circular motion, as the trajectory segment.

3. The method for early warning of tower crane collisions during building construction according to claim 1, characterized in that, The objective function is solved by using a swarm optimization intelligent algorithm to calculate the changes in the angular velocity of all tower cranes at corresponding future times, thereby determining the value of the objective function.

4. The method for early warning of tower crane collisions during building construction according to claim 3, characterized in that, It also includes the step of obtaining the initial solution of the swarm optimization intelligent algorithm, specifically: Obtain the historical working parameter sequences of all tower cranes and the current working parameter sequences, and construct a graph structure dataset of all historical tower cranes and graph structure data of all current tower cranes; wherein, the graph structure data in the graph structure dataset of all historical tower cranes corresponds one-to-one with the historical working state parameters; the working parameter sequences are five-dimensional data; perform one-to-one matching between the graph structure data of all current tower cranes and the graph structure dataset of all historical tower cranes, and obtain the maximum matching subgraph of the graph structure data of all current tower cranes; Use the change in angular velocity in the maximum matching subgraph as the initial solution for the change in angular velocity of the current tower crane; Simultaneously, the maximum matching subgraph is removed, and the graph structure data of all current tower cranes is re-matched with the graph structure dataset of all historical tower cranes after removal, until the subgraph structure matched by all tower cranes is obtained; then the adjustment variables corresponding to all the obtained matching subgraph structures are used as the initial solution of the objective function of the corresponding two adjacent tower cranes.

5. The method for early warning of tower crane collisions during building construction according to claim 4, characterized in that, In the process of performing the maximum matching subgraph algorithm, when calculating the distance between two corresponding nodes in two graph structures, the working state parameter sequences of the two tower cranes corresponding to the two nodes at the set time are merged into a set of working state parameter sequences; the dimensionality of the set of working state parameter sequences is reduced to obtain the eigenvalues ​​corresponding to the principal component directions of each working state parameter. The eigenvector corresponding to the largest eigenvalue is selected, and the projection value of the five-dimensional data corresponding to each working state onto the eigenvector corresponding to the largest eigenvalue is obtained. Then, the projection value sequence corresponding to the two nodes is obtained. The correlation value between the two projection sequences is obtained by using the Pearson correlation coefficient method, which is used as the distance value between the two corresponding nodes in the two graph structures.

6. A tower crane collision warning system for building construction, characterized in that, include: processor; A memory storing computer instructions for early warning of accidental collisions with tower cranes during construction, wherein when the computer instructions are executed by the processor, the system performs an early warning method for accidental collisions with tower cranes during construction according to any one of claims 1-5.

Images