A free node coordinate prediction method and system for cable net structure

Abstract

The application provides a free node coordinate prediction method and system for a cable net structure, and the method comprises the following steps: obtaining prestress distribution of the cable net structure, each boundary node coordinate and an optimized direction value of each free node; inputting each boundary node coordinate into a deep learning shape-finding model to output a predicted value of each free node coordinate; calculating a loss function value of the deep learning shape-finding model, adjusting parameters of the deep learning shape-finding model to update the deep learning shape-finding model until the loss function value is smaller than a set error, and obtaining the predicted value of each free node coordinate of the cable net structure. In the application, the boundary node coordinates of the cable net structure are taken as inputs of the deep learning shape-finding model, and the optimized direction value of the free node is given, the deep learning shape-finding model is used to predict the free node coordinates of the cable net structure, and the shape-finding of the cable net structure is achieved. Compared with the existing manual force finding, the application has high calculation efficiency and is fast and accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of architectural design, and more particularly, to a free node coordinate prediction method and system for a cable net structure. BACKGROUND

[0002] Cable net structures have been widely used in large public buildings, presenting novel, beautiful and light effects, and realizing the reasonable unification of force and shape. Due to the presence of prestress in the steel cable, the geometric shape of the cable net structure will change greatly from the zero stress state to the initial prestress state or the load state, which has a great influence on its structural stress performance. Therefore, it is crucial to determine the initial prestress state of the cable net structure, which is the basis for subsequent design and calculation analysis of the structure.

[0003] At present, the shape finding methods of cable net structures include force density method, dynamic relaxation method, nonlinear finite element method, analytical method, etc. These methods have achieved good results and have been applied in many practical engineering projects.

[0004] However, in the above methods, either the mathematical theory is complex and not easy for structural engineers to master, or the stiffness matrix is prone to singular convergence difficulty or needs to be repeatedly adjusted to determine the appropriate initial value. The adjustment process is tedious and time-consuming, and the shape finding efficiency of the cable net structure is low. Therefore, it is necessary to propose a fast and easy-to-use shape finding method for cable net structures. SUMMARY

[0005] The present application provides a free node coordinate prediction method and system for a cable net structure, which overcomes the problems of tedious and time-consuming adjustment process and low shape finding efficiency of the cable net structure.

[0006] According to a first aspect of the present application, a free node coordinate prediction method for a cable net structure is provided, comprising:

[0007] Step 1, obtaining the prestress distribution and the coordinate of each boundary node of the cable net structure, and giving the optimized direction value of each free node of the cable net structure according to the force principle of the cable net structure;

[0008] Step 2, setting the initial parameters of the deep learning shape finding model;

[0009] Step 3, inputting each boundary node coordinate into the deep learning shape finding model, and outputting the predicted value of each free node coordinate;

[0010] Step 4, calculating the loss function value of the deep learning shape finding model according to all boundary node coordinates, all free node optimized direction values and all free node coordinate predicted values;

[0011] Step 5, adjusting parameters of the deep learning form-finding model according to the loss function value to update the deep learning form-finding model, repeating steps 2-5 until the loss function value is less than a set error, and obtaining each free node coordinate prediction value of the cable net structure.

[0012] According to a second aspect of the present application, a free node coordinate prediction system of a cable net structure is provided, comprising:

[0013] An obtaining module is configured to obtain prestress distribution and each boundary node coordinate of the cable net structure, and give each free node optimization direction value of the cable net structure according to force principle of the cable net structure;

[0014] A setting module is configured to set initial parameters of a deep learning form-finding model;

[0015] A prediction module is configured to input each boundary node coordinate into the deep learning form-finding model, and output each free node coordinate prediction value;

[0016] A calculation module is configured to calculate a loss function value of the deep learning form-finding model according to all boundary node coordinates, all free node optimization direction values and all free node coordinate prediction values;

[0017] An adjusting module is configured to adjust parameters of the deep learning form-finding model according to the loss function value to update the deep learning form-finding model, repeatedly call the prediction module and the calculation module until the loss function value is less than a set error, and obtain each free node coordinate prediction value of the cable net structure.

[0018] According to a third aspect of the present application, an electronic device is provided, comprising a memory and a processor, the processor being configured to implement steps of a free node coordinate prediction method of a cable net structure when executing a computer management program stored in the memory.

[0019] According to a fourth aspect of the present application, a computer readable storage medium is provided, having a computer management program stored thereon, the computer management program being configured to implement steps of a free node coordinate prediction method of a cable net structure when executed by a processor.

[0020] The application provides a free node coordinate prediction method and system of a cable net structure, obtains prestress distribution and coordinates of each boundary node of the cable net structure, and gives an optimization direction value of each free node according to a stress principle of the cable net structure; the coordinates of each boundary node are input into a deep learning shape finding model, and a predicted value of each free node coordinate is output; a loss function value of the deep learning shape finding model is calculated, parameters of the deep learning shape finding model are adjusted, and the deep learning shape finding model is updated until the loss function value is less than a set error, and a predicted value of each free node coordinate of the cable net structure is obtained. The application takes the coordinates of the boundary nodes of the cable net structure as input of the deep learning shape finding model, gives the optimization direction value of the free node, uses the deep learning shape finding model to predict the free node coordinates of the cable net structure, and achieves the purpose of shape finding of the cable net structure, and compared with the existing manual force finding, the application has high calculation efficiency and is fast and accurate. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 A free node coordinate prediction method flowchart of a cable net structure is provided in the application;

[0022] Figure 2 A working principle schematic diagram of a deep learning shape finding model is provided in the application;

[0023] Figure 3 A schematic diagram of a chain rule is provided in the application;

[0024] Figure 4 A schematic diagram of an orthogonal cable net structure is provided in the application;

[0025] Figure 5 A shape finding result schematic diagram of the orthogonal cable net structure is provided in the application;

[0026] Figure 6 A schematic diagram of a diagonal cable net structure is provided in the application;

[0027] Figure 7 A shape finding result schematic diagram of the diagonal cable net structure is provided in the application;

[0028] Figure 8 A schematic diagram of an irregular cable net structure is provided in the application;

[0029] Figure 9 A shape finding result schematic diagram of the irregular cable net structure is provided in the application;

[0030] Figure 10-1 A finite element calculation result schematic diagram of cable net stress of the irregular cable net structure is provided in the application;

[0031] Figure 10-2 A finite element calculation result schematic diagram of cable net displacement of the irregular cable net structure is provided in the application;

[0032] Figure 11A structural schematic diagram of a free node coordinate prediction system of a cable net structure provided by the present application is provided.

[0033] Figure 12 A possible hardware structural schematic diagram of an electronic device provided by the present application is provided.

[0034] Figure 13 A possible hardware structural schematic diagram of a computer readable storage medium provided by the present application is provided. DETAILED DESCRIPTION

[0035] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application. In addition, the technical features in each embodiment or in a single embodiment provided by the present application can be combined with each other at will to form a feasible technical solution, and such combination is not subject to the order of steps and / or structure composition mode, but must be based on the implementation by those of ordinary skill in the art. When the combination of technical solutions contradicts each other or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection required by the present application.

[0036] Based on the background art, in view of this, the present application proposes a physical law driven cable net structure deep learning form finding model. The model solves the initial state of the cable net structure through the known prestress distribution of the cable net structure: taking the boundary node coordinates as the input layer and the free node coordinates as the output layer, adding the physical balance equation and the physical stress deformation principle of the known cable prestress into the loss function in the deep learning model, driving the deep learning model to reduce the loss function value through unsupervised training, and gradually approaching to zero, so as to achieve the purpose of form finding of the cable net structure.

[0037] Figure 1 A free node coordinate prediction method flowchart of a cable net structure provided by the present application is provided, which is written in Python language and developed based on the Pytorch library, as shown in Figure 1 The specific steps are as follows:

[0038] Step 1, obtain the prestress distribution of the cable net structure and the coordinates of each boundary node, and according to the stress principle of the cable net structure, give the optimization direction value of each free node of the cable net structure.

[0039] It can be understood that the purpose of the present application is to solve the free node coordinates of the cable net structure when the prestress distribution and each boundary node coordinate of the cable net structure are known, and the shape finding of the cable net structure is performed. Wherein, according to the prestress of the cable net structure and the stress principle of the cable net structure, the optimization direction value of each free node can be given to control the moving direction of the free node.

[0040] Step 2, setting the initial parameters of the deep learning shape finding model.

[0041] Wherein, referring to Figure 2 , the structural diagram of the deep learning shape finding model, the deep learning shape finding model includes an input layer, a plurality of hidden layers and an output layer, then, first, setting the initial parameters of the deep learning shape finding model mainly includes setting the initial values of the weight coefficients and the bias coefficients of each layer of the deep learning shape finding model.

[0042] Step 3, inputting each boundary node coordinate into the deep learning shape finding model, and outputting each free node coordinate prediction value.

[0043] It can be understood that the working principle of the deep learning shape finding model is: inputting the boundary node coordinates (x f ,y f ,z f ), passing through the input layer to the hidden layer (N layers) and the output layer, each neuron of each layer receives the data transmitted by the neuron of the previous layer, and transmits the data to the next layer after weighted summation and activation processing, until the output end outputs the data, that is, each free node coordinate prediction value.

[0044] The weighted summation and activation processing in the above deep learning shape finding model are realized by the following formula:

[0045] y i =σ(w i .y i+1 +b i ), 1≤i≤n

[0046] z=w n+1 ·y n +b n+1

[0047] In the formula, i is the i-th hidden layer, n+1 is the output layer, y i is the output value of the i-th hidden layer, z is the output value of the deep learning shape finding model, σ() is the activation function, w i and b i are the weight coefficient and the bias coefficient of the i-th hidden layer, and w n+1 is the weight coefficient of the output layer, and b n+1 is the bias coefficient of the output layer.

[0048] Step 4, according to all boundary node coordinates, all free node optimization direction values and all free node coordinate prediction values, the loss function value of the deep learning shape finding model is calculated.

[0049] It can be understood that after the deep learning shape finding model makes a prediction, the loss function of the deep learning shape finding model is calculated. In the present application, the loss function of the deep learning shape finding model mainly consists of two parts, one part is the mean square residual error of the physical equilibrium equation of the free node, and the other part is the mean square error of the moving direction of the free node.

[0050] Among them, according to all boundary node coordinates and all free node coordinate prediction values, the mean square residual error MSE x , MSE y and MSE z of the physical equilibrium equation in x, y and z directions of all free nodes are calculated respectively, and according to all free node optimization direction values and all free node coordinate prediction values, the mean square error MSE m of the moving direction of the free node is calculated; according to the mean square residual error MSE x , MSE y and MSE z and the mean square error MSE m , the loss function value of the deep learning shape finding model is calculated.

[0051] Among them, the loss function of the deep learning shape finding model can be expressed as L total =MSE x +MSE y +MSE z +f(n)·MSE m . MSE x , MSE y and MSE z are the mean square residual error of the physical equilibrium equation in x, y and z directions of the free node, which controls the optimization target of the model. After the steel cable is prestressed, according to the principle of physical stress deformation, the intermediate node will gradually move in a straight line direction towards the steel cable until it reaches a balanced stable state. Based on this, the deep learning shape finding model takes f(n) and MSE m as the exponential decay function of the training step and the mean square error of the moving direction of the free node respectively, which controls the optimization direction of the model. With the increase of the training step, the proportion of the optimization direction term in the loss function value gradually decreases, and the optimization target term gradually dominates until it completely dominates, and finally a reasonable shape that satisfies the equilibrium equation is obtained. The specific expression is as follows:

[0052]

[0053] L total =MSE x+MSE y +MSE z +f(n)MSE m

[0054] f(n)=e -αn

[0055] wherein, N s is the number of free nodes, m is the number of cable elements connected to the free node j, L jk and F jk are the length of cable element and cable force between the free node j and the node k, respectively, (x j , y j , z j ) are the coordinates of the free node j, (x k , y k , z k ) are the coordinates of the node k, (x j , y j ) pred and (x j , y j ) opdi are the predicted value of the coordinates of the free node j in x, y direction and the optimization direction value, respectively, f(n) is an exponential decay function of the training step, n is the number of training steps of the deep learning shape finding model, and a is the decay coefficient.

[0056] Step 5: According to the loss function value, the parameters of the deep learning shape finding model are adjusted to update the deep learning shape finding model, and steps 2-5 are repeatedly executed until the loss function value is less than the set error, and the predicted value of the coordinates of each free node of the cable net structure is obtained.

[0057] It can be understood that after the loss function value of the deep learning shape finding model is calculated in step 4, if the loss function value is greater than the set error, the gradients of the loss function with respect to the weight coefficients and bias coefficients of each layer in the back propagation of the deep learning shape finding model are calculated by using the chain rule of differentiation, wherein the chain rule of differentiation can be referred to in Figure 3 The weight coefficients and bias coefficients of the deep learning shape finding model are updated by using an optimization algorithm, steps 3-5 are re-executed, and the loss function value of the deep learning shape finding model is continuously recalculated until the loss function is less than the set error, and the data obtained from the output end is the coordinates (x s , y s , z s ) of the free nodes.

[0058] The method for predicting the coordinates of the free nodes of the cable net structure provided by the present application is described below with reference to several specific embodiments.

[0059] Embodiment 1: Orthogonal cable net structure, node number is as follows Figure 4The nodes numbered 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12 are boundary node coordinates, the nodes numbered 13, 14, 15, 16, 17, 18, 19, 20, and 21 are free nodes, and the ratio of the cable forces in the horizontal and vertical directions is F1:F2 = 1:1. The boundary nodes and the free nodes are fixed on the saddle surface, and the equation of the saddle surface is Z = (X 2 -Y 2 ) / 50, where Figure 3 The boundary node coordinates (x f ,y f ,z f ) and the free node optimization direction values (x j ,y j ) opdi are determined. For this cable net structure, the calculation steps of the deep learning shape finding model are as follows:

[0060] (1) Input the boundary node coordinates (x f ,y f ,z f ), select the optimizer Adam, and set the learning rate to 0.01 and a = 0.01. The data is passed to the hidden layer (6 layers) and the output layer through the input layer, and each layer of neurons (32) receives the data transmitted by the neurons of the previous layer, and after weighted summation and activation (Tanh function) processing, the data is transmitted to the next layer until the output data is obtained.

[0061] (2) After the data is output, the loss function is calculated. If the loss function value is greater than the set error, the gradients of the loss function with respect to the weights and biases of each layer in the deep learning model are calculated through the chain rule of differentiation, the weights and biases in the model are updated through the optimization algorithm, and the calculation is restarted.

[0062] (3) The deep learning model automatically repeats steps (1) and (2), and after n = 1000 training steps, the loss function is less than the set error, the training stops, and the free node coordinates (x s ,y s ,z s ) are output.

[0063] The shape finding results are shown in Figure 5 , and compared with the theoretical results, as shown in Table 1. As can be seen from Table 1, the calculation results of the two methods are in good agreement.

[0064] Table 1 Comparison of free node Z coordinate shape finding results of orthogonal cable net structure

[0065] Number 13 14 15 16 17 18 19 20 21 Theoretical value (m) 0.000 -0.428 0.000 0.428 0.000 0.428 0.000 -0.428 0.000 Predicted value (m) 0.000 -0.438 0.000 0.438 0.000 0.438 0.000 -0.438 0.000 Error (%) 0.0 2.3 0.0 2.3 0.0 2.3 0.0 2.3 0.0

[0066] Example Two: Diagonal Cable Net Structure, Node Numbering as Figure 6As shown, nodes numbered 1, 2, 3, 4, and 5 are boundary nodes, and nodes numbered 6, 7, 8, 9, and 10 are free nodes. The cable force ratio F1:F2:F3:F4:F5 = 1:1:1:1:1. The boundary nodes are fixed on the saddle surface, and their equation is Z = (X... 2 -Y 2 ) / 50, by Figure 6 Determine the boundary node coordinates (x f ,y f ,z f ) and free node optimization direction value (x j ,y j ) opdi The calculation steps are the same as in Example 1.

[0067] The shape finding results are as follows Figure 7 As shown in Table 2, the results obtained by the method of force density are compared with those obtained by the form-finding method. Table 2 shows that the calculation results of the two methods agree well.

[0068] Table 2 Comparison of Form-Finding Results for Oblique Cable Net Structures

[0069]

[0070]

[0071] Example 3: Irregular cable net structure, node numbering as follows Figure 8 As shown, nodes numbered 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 are boundary nodes, and nodes numbered 11, 12, 13, 14, 15, and 16 are free nodes. The cable force ratio F1:F2:F3:F4:F5 = 1:1:1:1:1. (From...) Figure 8 Determine the boundary node coordinates (x f ,y f ,z f ) and free node optimization direction value (x j ,y j ) opdi The calculation steps are the same as in Example 1.

[0072] The shape finding results are as follows Figure 9 As shown, the preset elastic modulus of the cable net is 200 GPa, and the cross-section is 1000 mm². 2 The cable prestress of 200MPa was calculated using finite element method (FEM). The FEM results can be found in [link to FEM calculation]. Figure 10-1 and Figure 10-2 ,in, Figure 10-1 The results are from the finite element analysis of the stress in the cable net. Figure 10-2The finite element calculation results of the cable net displacement are shown in Table 3. It can be seen from Table 3 that the calculation results of the two methods are in good agreement.

[0073] Table 3 Comparison of shape finding results of irregular cable net structure

[0074]

[0075]

[0076] The three cases above show that the shape finding results of the model of the application are in good agreement with the theoretical solution, the calculation results of the linear density method and the finite element method, have high precision, and the time used is not more than 1 min, and the calculation efficiency is high.

[0077] Referring to Figure 11 , a free node coordinate prediction system for a cable net structure of the application is provided, and the system comprises:

[0078] The acquisition module 1101 is configured to acquire prestress distribution and each boundary node coordinate of the cable net structure, and give each free node optimization direction value of the cable net structure according to the force principle of the cable net structure;

[0079] The setting module 1102 is configured to set initial parameters of the deep learning shape finding model;

[0080] The prediction module is configured to input each boundary node coordinate into the deep learning shape finding model, and output each free node coordinate prediction value;

[0081] The calculation module 1103 is configured to calculate a loss function value of the deep learning shape finding model according to all boundary node coordinates, all free node optimization direction values and all free node coordinate prediction values;

[0082] The adjustment module 1104 is configured to adjust parameters of the deep learning shape finding model according to the loss function value, to update the deep learning shape finding model, repeatedly call the prediction module and the calculation module, until the loss function value is less than a set error, and acquire each free node coordinate prediction value of the cable net structure.

[0083] It can be understood that the free node coordinate prediction system for a cable net structure provided by the application corresponds to the free node coordinate prediction method for a cable net structure provided by the above-mentioned embodiments, and the related technical features of the free node coordinate prediction system for a cable net structure can refer to the related technical features of the free node coordinate prediction method for a cable net structure, which will not be repeated here.

[0084] Please refer to Figure 12 , Figure 12An embodiment of the electronic device provided by the present application is shown in the figure. Figure 12 As shown in the figure, the embodiment of the present application provides an electronic device 1200, which comprises a memory 1210, a processor 1220, and a computer program 1211 stored in the memory 1210 and executable on the processor 1220, and the processor 1220 implements the steps of the free node coordinate prediction method of the cable net structure when executing the computer program 1211.

[0085] Please refer to Figure 13 , Figure 13 An embodiment of the computer readable storage medium provided by the present application is shown in the figure. Figure 13 As shown in the figure, the embodiment provides a computer readable storage medium 1300, which stores a computer program 1311, and the computer program 1311 implements the steps of the free node coordinate prediction method of the cable net structure when executed by a processor.

[0086] The embodiment of the present application provides a free node coordinate prediction method and system of a cable net structure, obtains prestress distribution and coordinates of each boundary node of the cable net structure, and gives an optimization direction value of each free node according to the stress principle of the cable net structure; inputs the coordinates of each boundary node into a deep learning shape finding model, and outputs a predicted value of the coordinates of each free node; calculates a loss function value of the deep learning shape finding model, adjusts parameters of the deep learning shape finding model, and updates the deep learning shape finding model until the loss function value is less than a set error, and obtains the predicted value of the coordinates of each free node of the cable net structure. The embodiment of the present application inputs the coordinates of the boundary nodes of the cable net structure into the deep learning shape finding model, and gives the optimization direction value of the free node, and uses the deep learning shape finding model to predict the coordinates of the free node of the cable net structure, so as to achieve the purpose of shape finding of the cable net structure, and compared with the existing manual force finding, the calculation efficiency is high, and the method is fast and accurate.

[0087] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0088] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0089] The present application is described in reference to the flowchart illustrations and / or block diagrams according to the methods, apparatus (systems) and computer program products of embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart Figure 1 one or more functions specified in the flowchart Figure 1 one or more blocks in the flowchart

[0090] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart Figure 1 one or more functions specified in the flowchart Figure 1 one or more blocks in the flowchart

[0091] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart Figure 1 one or more functions specified in the flowchart Figure 1 one or more blocks in the flowchart

[0092] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments by those of skill in the art once they have the benefit of the present disclosure without departing from the spirit and scope of the application. Accordingly, the attached claims are intended to cover all such additional variations and modifications as falling within the scope of the application.

[0093] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A method for predicting coordinates of free nodes of a cable net structure, characterized by, Comprising: Step 1, obtaining the prestress distribution of the cable net structure and the coordinates of each boundary node, and according to the force principle of the cable net structure, giving the optimization direction value of each free node of the cable net structure; Step 2, setting the initial parameters of the deep learning shape finding model; Step 3, inputting each of the boundary node coordinates into the deep learning shape finding model, and outputting the predicted value of each free node coordinate; Step 4, calculating the loss function value of the deep learning shape finding model according to all the boundary node coordinates, all the free node optimization direction values and all the free node coordinate predicted values; Step 5, adjusting the parameters of the deep learning shape finding model according to the loss function value to update the deep learning shape finding model, repeating steps 2-5 until the loss function value is less than the set error, and obtaining the predicted value of each free node coordinate of the cable net structure; In step 4, the loss function value of the deep learning shape finding model is calculated according to all the boundary node coordinates, all the free node optimization direction values and all the free node coordinate predicted values, comprising: Based on all boundary node coordinates and all free node coordinate prediction values, the mean square residual of the physical equilibrium equation in the x, y, and z directions on all free nodes is calculated respectively MSE x , MSE y and MSE z comprising: According to the optimization direction value of all free nodes and the coordinate prediction value of all free nodes, the mean square error of the moving direction of the free nodes is calculated MSE m comprising: According to the mean square residual error MSE x , MSE y And MSE z And the mean square error MSE m , calculating the loss function value of the deep learning shape finding model comprises: L total = MSE x + MSE y + MSE z + MSE m ； wherein, N s is the number of free nodes, m is the number of element connected to the free node j , L jk and F jk are the element length and cable force between the free node j and the node k , x j , y j , z j are the coordinates of the free node j , x k , y k , z k are the coordinates of the node k , x j , y j pred and x j , y j opdi are the coordinate prediction value and optimization direction value of the free node j in x, y direction, f (n) is an exponential decay function of the training step, n is the training step number of the deep learning shape finding model, is the decay coefficient.​​ 2. The free node coordinate prediction method of a cable net structure according to claim 1, wherein, f(n)MSE The deep learning shape finding model comprises an input layer, a plurality of hidden layers and an output layer, comprising:

3. The method of claim 1, wherein, Each of the input boundary node coordinates is transmitted to the plurality of hidden layers through the input layer, the neurons of each hidden layer receive the data transmitted by the neurons of the previous hidden layer, and after weighted summation and activation processing, the data is transmitted to the neurons of the next hidden layer. After passing through a plurality of hidden layers, the predicted value of each free node coordinate is output through the output layer. In step 2, the initial parameters of the deep learning shape finding model are set, comprising:

4. The method of claim 2, wherein, Setting the initial weight coefficients and initial bias coefficients of each layer of the deep learning shape finding model. ， In the formula, i For the first i There are n+1 hidden layers and n+1 output layers. y i For the first i The output value of the hidden layer, z To find the output value of the force model for deep learning, For activation function, w i and b i For the first i The weight and bias coefficients of the hidden layers are parameters of the deep learning shape-finding model. These are the weight coefficients of the output layer. This represents the bias coefficient of the output layer.

5. The method of claim 1, wherein, The weighted summation and activation processing of each of the hidden layers comprises: In step 5, the parameters of the deep learning shape finding model are adjusted according to the loss function value to update the deep learning shape finding model, comprising:

6. A free node coordinate prediction system of a cable net structure, applied to the free node coordinate prediction method of claim 1, characterized in that, If the loss function value is greater than the set error range, the gradients of the loss function to the weight coefficients and bias coefficients of each layer in the back propagation of the deep learning shape finding model are calculated by the chain rule, and the weight coefficients and bias coefficients of the deep learning shape finding model are updated by the optimization algorithm. Comprising: An acquisition module for acquiring the prestress distribution of the cable net structure and the coordinates of each boundary node, and according to the force principle of the cable net structure, giving the optimization direction value of each free node of the cable net structure; A setting module for setting the initial parameters of the deep learning shape finding model; A prediction module for inputting each of the boundary node coordinates into the deep learning shape finding model, and outputting the predicted value of each free node coordinate; A calculation module for calculating the loss function value of the deep learning shape finding model according to all the boundary node coordinates, all the free node optimization direction values and all the free node coordinate predicted values; An adjusting module is configured to adjust parameters of the deep learning shape-finding model according to the loss function value, to update the deep learning shape-finding model, and to repeatedly call the predicting module and the calculating module until the loss function value is less than a set error, and to obtain each free node coordinate prediction value of the cable net structure.

7. An electronic device, comprising: The computer management program is executed by the processor to implement the free node coordinate prediction method of the cable net structure according to any one of claims 1-5.

8. A computer readable storage medium having stored thereon a computer management class program, the program comprising: The computer management program is executed by the processor to implement the free node coordinate prediction method of the cable net structure according to any one of claims 1-5.

Images