Heavy vehicle walking route intelligent arrangement method and equipment based on machine learning

Abstract

According to the machine learning-based heavy vehicle walking route intelligent arrangement method and equipment provided by the invention, the heavy vehicle walking route can be quickly obtained through simple parameter input based on the machine learning method; the problems that a traditional finite element analysis method is complex in modeling process and long in calculation time and needs multiple times of calculation under different working conditions are solved, and the calculation efficiency is improved. According to the invention, a basement roof layout diagram, a structural design general description and heavy vehicle information are input into the information acquisition and numbering model, and the information acquisition and numbering model transmits the identified structural design information and heavy vehicle information, namely load information, to the roof checking calculation module for checking calculation to obtain a checking calculation result; and the checking calculation result is marked through a structure marking module, and a heavy vehicle route arrangement diagram is output. According to the invention, the walking route map of the heavy vehicle can be quickly obtained by automatically identifying the drawing.

Description

Technical Field

[0001] The present invention relates to a method and device for intelligently arranging a heavy vehicle travel route based on machine learning. Background Art

[0002] During the construction process, heavy vehicles are often used to transport various materials. On site, the routes of these heavy vehicles often pass through the basement roof. At this time, it is necessary to verify whether their load exceeds the design requirements of the basement roof, so as to determine the routes that heavy vehicles can travel.

[0003] Existing methods for planning heavy vehicle routes often require the use of finite element software to build a structural model for verification, calculate the response of the basement ceiling when different vehicles move to different positions, and repeatedly modify the load position. The modeling operation process is cumbersome and time-consuming. Summary of the invention

[0004] The object of the present invention is to provide a method and device for intelligently arranging the travel routes of heavy vehicles based on machine learning.

[0005] In order to solve the above problems, the present invention provides a method for intelligently arranging a heavy vehicle travel route based on machine learning, comprising:

[0006] The training information collection and processing module needs to collect the floor plan of the sample basement top plate, the general description of the structural design, and the parameters of the heavy vehicle, and pre-process the structural design information and heavy vehicle information of the sample basement top plate;

[0007] The model training module establishes an information collection and numbering model based on the floor plan, structural design general description, heavy vehicle parameters, and pre-processed structural design information and heavy vehicle information obtained by the training information collection and processing module, and uses machine learning methods;

[0008] Inputting the floor plan, structural design general description and heavy vehicle parameters of the basement top plate to be identified into the information collection and numbering model to obtain the structural design information and heavy vehicle information of the basement top plate to be identified;

[0009] The top plate verification module verifies each floor plate of the basement top plate to be identified based on the beam and plate information collection and the structural design information and heavy vehicle information input by the numbering model, and marks the floor plates that pass the verification and those that fail the verification;

[0010] The structural marking module draws a heavy vehicle route layout diagram in the floor slab that has passed the verification based on the heavy vehicle information.

[0011] Furthermore, in the above method, the parameters of the structural design information include: the standard value of the floor design live load Q, the slab span length L, the slab span width B and the slab thickness hb ;

[0012] The parameters of the heavy vehicle information include: total vehicle weight T, wheel pressure acting distribution area S, vehicle length l, vehicle width b, and vehicle turning radius r.

[0013] Further, in the above method, the training information collection and processing module preprocesses the structural design information and heavy vehicle information of the sample basement roof slab, including:

[0014] Tag the parameters of the structural design information and the parameters of the heavy vehicle information; extract the numerical values of each parameter for unit conversion; and number the floor slab components, with the numbering sequence from the upper left of the drawing from left to right and from top to bottom.

[0015] Further, in the above method, input the floor plan layout, general structural design description, and heavy vehicle parameter input information of the basement roof slab to be recognized into the information collection and numbering model to obtain the structural design information and heavy vehicle information of the basement roof slab to be recognized,

[0016] The input of the beam and slab information collection and numbering model is: the floor plan layout, general structural design description, and heavy vehicle parameters of the basement roof slab to be recognized;

[0017] The output of the beam and slab information collection and numbering model is: the standard value of the designed live load Q of the floor slab, the span length L of the slab, the span width B of the slab, the slab thickness h b , total vehicle weight T, wheel pressure acting distribution area S, vehicle length l, vehicle width b, and vehicle turning radius r.

[0018] Further, in the above method, the roof slab checking module checks each floor slab of the basement roof slab to be recognized according to the structural design information and heavy vehicle information passed in by the beam and slab information collection and numbering model, and marks the floor slabs that pass the check and those that do not pass the check, including:

[0019] According to the preset specifications, based on the floor slab span length L, floor slab span width B, floor slab thickness h b , total vehicle weight T, and wheel pressure acting distribution area S, obtain the equivalent uniformly distributed live load F of the vehicle;

[0020] When Q < F, the floor slab check fails; when Q ≥ F, the floor slab check passes; pass the check result and the corresponding floor slab number to the structure marking module;

[0021] The structure marking module marks the floor slabs that pass the check as green and those that do not pass as red.

[0022] Further, in the above method, the structure marking module draws the layout diagram of the heavy vehicle route on the floor slabs that pass the check according to the heavy vehicle information, including:

[0023] The structure marking module draws the heavy vehicle route layout in the green marked floor slab based on the vehicle length l, vehicle width b and vehicle turning radius r.

[0024] Furthermore, in the above method, the information collection and numbering model and the top plate verification module are trained in the form of DNN, RNN or Transformer model.

[0025] Furthermore, in the above method, the training process of the information collection and numbering model and the top plate verification module adopts the SGD or Adam optimization algorithm until the target loss function reaches a preset expected value, and then the training is stopped.

[0026] According to another aspect of the present invention, there is further provided a computer-readable storage medium having computer-executable instructions stored thereon, wherein when the computer-executable instructions are executed by a processor, the processor is caused to: execute any of the methods described above.

[0027] According to another aspect of the present invention, there is also provided a computer device, comprising:

[0028] Processor; and

[0029] A memory arranged to store computer executable instructions, which when executed cause the processor to: perform any of the methods described above.

[0030] In summary, in order to solve the above-mentioned problem that the heavy vehicle route layout process is cumbersome and time-consuming, the present invention is based on a machine learning method to quickly obtain the heavy vehicle route through simple parameter input, solve the problems of the traditional finite element analysis method that the modeling process is complicated, the calculation is time-consuming, and multiple calculations are required for different working conditions, thereby improving the calculation efficiency. The present invention inputs the basement roof layout diagram, the general structural design description and the heavy vehicle information into the information collection and numbering model. The information collection and numbering model will pass the identified structural design information and heavy vehicle information, i.e., the load information, to the roof verification module for verification to obtain the verification result, mark the verification result through the structure marking module, and output the heavy vehicle route layout diagram. The present invention can quickly obtain the heavy vehicle route map by automatically identifying the drawings.

[0031] Compared with the current heavy vehicle route arrangement method, the beneficial effects of the present invention are mainly as follows:

[0032] (1) High efficiency. The intelligent layout method of heavy vehicle travel routes based on machine learning in the present invention has a layout speed much higher than traditional methods such as finite element calculation;

[0033] (2) Simple operation. Compared with the existing methods, the present invention only requires the construction site personnel to input a few simple parameters or drawings such as the basement roof plan layout, structural design general description, and heavy vehicle parameters to obtain the heavy vehicle route layout;

[0034] (3) Strong adaptability. The present invention can be applied to basement top plates with different spans under any load length and width, and can also be applied to heavy vehicles of any specifications, sizes or tonnage. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of a method for intelligently arranging a heavy vehicle travel route based on machine learning according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] like Figure 1 As shown, the present invention provides a method for intelligently arranging a heavy vehicle travel route based on machine learning, comprising:

[0038] Step S1, the training information collection and processing module needs to collect the floor plan of the sample basement roof, the general description of the structural design, and the parameters of the heavy vehicle, and pre-process the structural design information and heavy vehicle information of the sample basement roof;

[0039] Preferably, the parameters of the structural design information include: standard value of floor design live load Q, slab span length L, slab span width B and slab thickness h b ;

[0040] The parameters of the heavy vehicle information include: vehicle gross weight T, wheel pressure distribution area S, vehicle length l, vehicle width b and vehicle turning radius r;

[0041] The heavy vehicle may be a truck having a maximum gross mass of more than 12,000 kg.

[0042] Preferably, the training information collection and processing module pre-processes the structural design information of the sample basement top plate and the heavy vehicle information, including:

[0043] Label the parameters of the structural design information and the parameters of the heavy vehicle information; extract the values ​​of each parameter and convert the units; and number the floor components in the order from left to right and from top to bottom from the upper left corner of the drawing;

[0044] More preferably, the unit conversion includes:

[0045] Floor span length L, floor span width B, floor thickness h b , vehicle length l, vehicle width b and vehicle turning radius r are in meters;

[0046] The unit of gross vehicle weight T is kN;

[0047] The unit of wheel pressure distribution area S is m 2 ;

[0048] The unit of standard value Q of floor design live load is kN / m 2 .

[0049] Step S2, the model training module establishes an information collection and numbering model based on the floor plan, structural design general description, heavy vehicle parameters, and pre-processed structural design information and heavy vehicle information obtained by the training information collection and processing module, and using a machine learning method;

[0050] Preferably, the training form of the information collection and numbering model is not limited, and models such as DNN, RNN and Transformer can be used to establish the beam and slab information collection and numbering model and the top slab verification module.

[0051] To ensure the training effect, the training set needs more than 10,000 sets of data.

[0052] Preferably, the model training process can use optimization algorithms such as SGD and Adam, and the training can be stopped when the target loss function reaches a preset expected value.

[0053] Step S3, inputting the floor plan of the basement top plate to be identified, the general structural design description and the heavy vehicle parameters into the information collection and numbering model to obtain the structural design information and the heavy vehicle information of the basement top plate to be identified;

[0054] Preferably, the beam and slab information collection and numbering model inputs are: the floor plan of the basement top slab to be identified, the general structural design description and the heavy vehicle parameters;

[0055] The output of the beam and slab information collection and numbering model is: the standard value of the floor slab design live load Q, the slab span length L, the slab span width B, and the slab thickness h of the basement top slab to be identified b , vehicle gross weight T, wheel pressure distribution area S, vehicle length l, vehicle width b and vehicle turning radius r;

[0056] Step S4, the top plate verification module verifies each floor plate of the basement top plate to be identified according to the beam and plate information collection and the structural design information and heavy vehicle information input by the numbering model, and marks the floor plates that pass the verification and those that fail the verification;

[0057] Preferably, according to the preset specifications, based on the floor slab span length L, floor slab span width B, and floor slab thickness h b , the total vehicle weight T, and the wheel pressure acting distribution area S, the equivalent uniformly distributed live load F of the vehicle can be obtained;

[0058] When Q < F, the floor slab check fails; when Q ≥ F, the floor slab check passes; the check result and the corresponding floor slab number are passed to the structure marking module;

[0059] The structure marking module marks the floor slabs that pass the check as green and those that fail as red;

[0060] Step S5, the structure marking module draws the heavy vehicle route layout diagram on the floor slabs that pass the check according to the heavy vehicle information;

[0061] Preferably, the structure marking module draws the heavy vehicle route layout diagram on the green-marked floor slabs according to the vehicle length l, vehicle width b, and vehicle turning radius r.

[0062] According to another aspect of the present invention, there is also provided a computer-readable storage medium on which computer-executable instructions are stored, wherein when the computer-executable instructions are executed by a processor, the processor is caused to: execute the method described in any one of the above.

[0063] According to another aspect of the present invention, there is also provided a calculator device, which includes:

[0064] A processor; and

[0065] A memory arranged to store computer-executable instructions, and when the executable instructions are executed, the processor is caused to: execute the method described in any one of the above.

[0066] In summary, to solve the problems of the cumbersome and time-consuming process of arranging the heavy vehicle walking route as mentioned above, the present invention is based on a machine learning method to achieve quickly obtaining the heavy vehicle walking route through simple parameter input, solving the problems of the complex modeling process, long calculation time, and the need for multiple calculations for different working conditions in the traditional finite element analysis method, and improving the calculation efficiency. The present invention inputs the basement roof layout diagram, the general structure design description, and the heavy vehicle information into the information collection and numbering model, and this information collection and numbering model will transfer the recognized structural design information and heavy vehicle information, that is, load information, to the roof slab check module for checking to obtain the check result, and the structure marking module marks the check result and outputs the heavy vehicle route layout diagram. The present invention can achieve quickly obtaining the heavy vehicle walking route diagram by automatically recognizing the drawings.

[0067] Compared with the current heavy vehicle walking route arrangement method, the beneficial effects of the present invention are mainly as follows:

[0068] (1) High efficiency. The intelligent layout method of heavy vehicle travel routes based on machine learning in the present invention has a layout speed much higher than traditional methods such as finite element calculation;

[0069] (2) Simple operation. Compared with the existing methods, the present invention only requires the construction site personnel to input a few simple parameters or drawings such as the basement roof plan layout, structural design general description, and heavy vehicle parameters to obtain the heavy vehicle route layout;

[0070] (3) Strong adaptability. The present invention can be applied to basement top plates with different spans under any load length and width, and can also be applied to heavy vehicles of any specifications, sizes or tonnage.

[0071] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0072] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0073] Obviously, those skilled in the art can make various changes and modifications to the invention without departing from the spirit and scope of the invention. Thus, if these modifications and variations of the invention fall within the scope of the claims of the invention and their equivalents, the invention is also intended to include these modifications and variations.

Claims

1. A method for intelligently arranging heavy vehicle travel routes based on machine learning, characterized in that: Including: The training information collection and processing module needs to collect the floor plan of the sample basement roof slab, the general structural design specifications, and the heavy vehicle parameters, and preprocess the structural design information and heavy vehicle information of the sample basement roof slab; The model training module, based on the floor plan, general structural design specifications, and heavy vehicle parameters obtained by the training information collection and processing module, as well as the preprocessed structural design information and heavy vehicle information, uses machine learning methods to establish an information collection and numbering model; Input the floor plan, general structural design specifications, and heavy vehicle parameters of the basement roof slab to be recognized into the information collection and numbering model to obtain the structural design information and heavy vehicle information of the basement roof slab to be recognized; The roof slab checking module checks each floor slab of the basement roof slab to be recognized according to the structural design information and heavy vehicle information passed in by the beam-slab information collection and numbering model, and marks the floor slabs that pass the check and those that do not pass the check; The structure marking module draws the heavy vehicle route layout diagram on the floor slabs that pass the check according to the heavy vehicle information.

2. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 1, characterized in that: The parameters of the structural design information include: the standard value of the floor design live load Q, the slab span length L, the slab span width B and the slab thickness h b ; The parameters of the heavy vehicle information include: total vehicle weight T, wheel pressure acting distribution area S, vehicle length l, vehicle width b, and vehicle turning radius r.

3. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 2, characterized in that: The training information collection and processing module preprocesses the structural design information and heavy vehicle information of the sample basement roof slab, including: Tagging the parameters of the structural design information and the parameters of the heavy vehicle information; extracting the numerical values of each parameter for unit conversion; and numbering the floor slab components, with the numbering sequence from the upper left corner of the drawing from left to right and from top to bottom.

4. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 1, characterized in that: Input the floor plan, general structural design specifications, and heavy vehicle parameters of the basement roof slab to be recognized into the information collection and numbering model to obtain the structural design information and heavy vehicle information of the basement roof slab to be recognized, The input of the beam-slab information collection and numbering model is: the floor plan, general structural design specifications, and heavy vehicle parameters of the basement roof slab to be recognized; The output of the beam and slab information collection and numbering model is: the standard value of the floor slab design live load Q, the slab span length L, the slab span width B, and the slab thickness h of the basement top slab to be identified b , vehicle gross weight T, wheel pressure distribution area S, vehicle length l, vehicle width b and vehicle turning radius r.

5. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 1, characterized in that: The roof slab checking module checks each floor slab of the basement roof slab to be recognized according to the structural design information and heavy vehicle information passed in by the beam-slab information collection and numbering model, and marks the floor slabs that pass the check and those that do not pass the check, including: According to the preset specifications, based on the floor span length L, floor span width B, floor thickness h b , vehicle gross weight T and wheel pressure distribution area S, and obtain the vehicle equivalent uniform live load F; When Q < F, the floor slab check fails; when Q ≥ F, the floor slab check passes; transmit the check result and the corresponding floor slab number to the structure marking module; The structure marking module marks the floor slabs that pass the check as green and those that do not pass as red.

6. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 1, characterized in that: The structure marking module draws the heavy vehicle route layout diagram on the floor slabs that pass the check according to the heavy vehicle information, including: The structure marking module draws the heavy vehicle route layout diagram on the green-marked floor slabs according to the vehicle length l, vehicle width b, and vehicle turning radius r.

7. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 1, characterized in that: The training forms of the information collection and numbering model and the roof slab checking module adopt DNN, RNN, or Transformer models.

8. The method for intelligently arranging heavy vehicle travel routes based on machine learning as claimed in claim 1, characterized in that: The optimization algorithms of SGD or Adam are adopted during the training process of the information collection and numbering model and the roof slab checking module. Training stops when the target loss function reaches the pre-set expected value.

9. A computer-readable storage medium having computer-executable instructions stored thereon, wherein: When the computer executable instructions are executed by a processor, the processor is caused to: perform the method according to any one of claims 1 to 8.

10. A computer device, wherein: include: processor; as well as A memory arranged to store computer executable instructions which, when executed, cause the processor to: perform a method as claimed in any one of claims 1 to 8.

Images