Method, processor, apparatus, and storage medium for determining arm parameters
By training a three-level neural network model, the design parameters of the truck crane boom can be quickly determined, solving the problems of long cycle and low efficiency in traditional design methods, and achieving efficient boom parameter optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZOOMLION HEAVY INDUSTRY SCIENCE AND TECHNOLOGY CO LTD
- Filing Date
- 2022-12-30
- Publication Date
- 2026-05-19
AI Technical Summary
Existing truck crane boom designs are characterized by long design cycles, high costs, and low efficiency. Traditional design methods rely on experience and iteration, making it difficult to quickly obtain optimized design parameters.
A three-level neural network model is adopted. By acquiring known boom geometry and load parameters, supervised training is performed using neural networks of the main boom, superlift, and auxiliary boom to output optimal design parameters, including the key dimensions of the main boom, superlift, and auxiliary boom.
It enables the rapid and accurate determination of boom design parameters, reducing design cycle and cost, and improving design efficiency.
Smart Images
Figure CN116244845B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of mechanical engineering, and more specifically, to a method, processor, apparatus, and storage medium for determining boom parameters. Background Technology
[0002] The boom structure of a truck crane consists of a main boom and a jib. The boom structure is complex, composed of multiple main and jib sections connected together. A complete boom design includes modeling parameters for each section as well as assembly parameters. The modeling parameters for the entire boom are numerous, the data volume is large, and the design cycle is long.
[0003] Existing product design methods mainly rely on benchmarking or upgrading. Traditional design methods are based on experience or involve modifying and upgrading existing boom structures. This traditional approach often requires repeated manual iterations, resulting in long development cycles, wasted resources, low efficiency, and the final boom design parameters may not be satisfactory. Summary of the Invention
[0004] The purpose of this application is to provide a method, processor, device, and storage medium for rapidly determining boom design parameters.
[0005] To achieve the above objectives, this application provides a method for determining boom parameters, wherein the boom includes multiple boom sections, and the boom parameters include design parameters for the main boom, design parameters for the auxiliary boom, and design parameters for the superlift. The method includes:
[0006] Obtain the first geometric parameters of a known boom segment and the first load parameters of a known boom under multiple working conditions;
[0007] The first geometric parameters and the first load parameters are input into the main arm neural network so that the main arm neural network can output the optimal design parameters of the main arm to be designed.
[0008] The optimal design parameters of the main arm are input into the super-giant neural network, so that the optimal design parameters of the super-giant to be designed are output through the super-giant neural network.
[0009] The optimal design parameters of the super-start are input into the secondary arm neural network, so that the optimal design parameters of the secondary arm to be designed are output through the secondary arm neural network.
[0010] In the embodiments of this application, the first geometric parameters include the number of known boom sections Z_X, the boom section length Z_L of each known boom section, the maximum height Z_H of the basic boom section, and the maximum width Z_B of the basic boom section. The first load parameters include the working radius Z of the known boom under each working condition. i _, Load size Z i _、 Elevation angle Z i _.
[0011] In the embodiments of this application, the optimal design parameters of the main boom include the straight edge outward opening Z1r of the upper cover plate of the first main boom section, the bending angle opening Z1b of the upper cover plate, the height of the upper cover plate Z1h, the included angle Z1θ1 of the upper and lower cover plates, and the parameters of the reinforcing plate; the optimal design parameters of the superlift include the length of the superlift pull plate C_LL, the length of the superlift mast C_L, the hoisting radius C_R, and the opening angle of the superlift mast C_α; the optimal design parameters of the auxiliary boom include the length of the auxiliary boom F_L, the width of the auxiliary boom section F_B, the length of the web member F_FL, the width of the web member spacing F_FX, the diameter of the chord member F_XD, and the diameter of the web member F_FD.
[0012] In the embodiments of this application, vehicle parameters of multiple existing construction machinery are obtained, including the second geometric parameters of the boom segment and the second load parameters of the boom. The second geometric parameters and the second load parameters are input into the vehicle neural network to train the vehicle neural network. After the vehicle neural network is trained, the second geometric parameters and the second load parameters are input into the trained vehicle neural network again. The predicted design parameters output by the trained vehicle neural network are obtained, including the predicted design parameters of the main boom, the predicted design parameters of the superlift, and the predicted design parameters of the auxiliary boom. The main boom neural network, the superlift neural network, and the auxiliary boom neural network are trained in a supervised manner using the predicted design parameters.
[0013] In the embodiments of this application, training the main boom neural network, the super-lift neural network, and the auxiliary boom neural network by predicting design parameters includes: acquiring sample geometric parameters and sample load parameters of multiple existing construction machinery. The sample geometric parameters include the number of boom sections Z_X, the boom section length Z_L, the maximum height Z_H of the basic boom cross-section, and the maximum width Z_B of the basic boom cross-section. The sample load parameters include the boom's working radius Z under each working condition. i _D, Load size Z i _F, Elevation Angle Z i _θ; Set the first error parameter and first step length of the main arm neural network; After normalizing the sample geometric parameters and sample load parameters, input the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network in sequence; Obtain multiple main arm design parameters output by the main arm neural network under supervised training, wherein the number of multiple main arm design parameters is determined according to the first error parameter and the first step length; Repeat the step of inputting the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network until the main arm neural network is trained.
[0014] In the embodiments of this application, training the main arm neural network, the super-start neural network, and the secondary arm neural network using predicted design parameters includes: setting a second error parameter and a second step size for the super-start neural network; inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-start in the predicted design parameters into the super-start neural network; obtaining multiple super-start design parameters output by the super-start neural network under supervised training, wherein the number of multiple super-start design parameters is determined according to the second error parameter and the second step size; and repeatedly executing the step of inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-start in the predicted design parameters into the super-start neural network until the super-start neural network is trained.
[0015] In the embodiments of this application, training the main arm neural network, the super-start neural network, and the secondary arm neural network using predicted design parameters includes: setting a third error parameter and a third step size for the secondary arm neural network; inputting the super-start predicted design parameters output by the trained super-start neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network; obtaining multiple secondary arm design parameters output by the secondary arm neural network under supervised training, wherein the number of multiple secondary arm design parameters is determined based on the third error parameter and the third step size; and repeatedly executing the step of inputting the super-start design parameters output by the trained super-start neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network until the secondary arm neural network is trained.
[0016] A second aspect of this application provides a processor configured to perform any of the above-described methods for determining boom parameters.
[0017] A third aspect of this application provides an apparatus for determining boom parameters, including the processor described above.
[0018] A fourth aspect of this application provides a machine-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform a method for determining boom parameters according to any one of the preceding claims.
[0019] The above technical solution includes a three-level neural network, with each level interconnected. Ultimately, the design parameters of the entire boom system can be obtained through initial design parameters and working conditions.
[0020] Other features and advantages of this application will be described in detail in the following detailed description section. Attached Figure Description
[0021] The accompanying drawings are provided to further illustrate the present application and form part of the specification. They are used together with the following detailed description to explain the present application, but do not constitute a limitation thereof. In the drawings:
[0022] Figure 1 A schematic flowchart of a method for determining boom parameters according to an embodiment of this application is shown.
[0023] Figure 2 A schematic flowchart of a method for determining boom parameters according to yet another embodiment of this application is shown.
[0024] Figure 3 The diagram illustrates the internal structure of a computer device according to an embodiment of this application. Detailed Implementation
[0025] The specific embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit this application.
[0026] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.
[0027] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.
[0028] Figure 1 A schematic flowchart illustrating a method for determining boom parameters according to an embodiment of this application is shown. Figure 1 As shown in one embodiment of this application, a method for determining boom parameters is provided, comprising the following steps:
[0029] Step 101: Obtain the first geometric parameters of the known boom segment and the first load parameters of the known boom under multiple working conditions;
[0030] Step 102: Input the first geometric parameters and the first load parameters into the main arm neural network so that the main arm neural network can output the optimal design parameters of the main arm to be designed.
[0031] Step 103: Input the optimal design parameters of the main arm into the super-giant neural network, so as to output the optimal design parameters of the super-giant to be designed through the super-giant neural network;
[0032] Step 104: Input the optimal design parameters of the super-start into the secondary arm neural network, so that the optimal design parameters of the secondary arm to be designed can be output through the secondary arm neural network.
[0033] The boom can include multiple boom segments, and boom parameters can include design parameters for the main boom, auxiliary boom, and superlift. The processor can acquire the first geometric parameters of a known boom segment and the first load parameters of the known boom under multiple operating conditions. The acquired first geometric parameters and first load parameters are input into the main boom neural network, which outputs the optimal design parameters for the main boom. These optimal design parameters are then input into the superlift neural network, which outputs the optimal design parameters for the superlift to be designed. The processor can further input the optimal superlift design parameters into the auxiliary boom neural network, which outputs the optimal design parameters for the auxiliary boom to be designed.
[0034] In one embodiment, the first geometric parameters include the number of known boom sections Z_X, the length of each known boom section Z_L, the maximum height of the basic boom section Z_H, and the maximum width of the basic boom section Z_B. The first load parameters include the working radius Z of the known boom under each working condition. i _D, Load size Z i _F, Elevation Angle Z i _θ.
[0035] The processor can acquire the first geometric parameters of a known boom segment and the first load parameters of a known boom, which are then input into the main boom neural network. The first geometric parameters may include the number of known boom segments Z_X, the length of each known boom segment Z_L, the maximum height of the basic boom cross-section Z_H, and the maximum width of the basic boom cross-section Z_B, etc. The first load parameters include the working radius Z of the known boom under each working condition. i _D, Load size Z i _F, Elevation Angle Z i _θ etc.
[0036] In one embodiment, the optimal design parameters for the main boom include the straight edge outward opening Z1r of the upper cover plate of the first main boom section, the bending angle opening Z1b of the upper cover plate, the height of the upper cover plate Z1h, the included angle Z1θ1 between the upper and lower cover plates, and the parameters of the reinforcing plate; the optimal design parameters for the superlift include the length of the superlift pull plate C_LL, the length of the superlift mast C_L, the hoisting radius C_R, and the opening angle of the superlift mast C_α; the optimal design parameters for the auxiliary boom include the length of the auxiliary boom F_L, the width of the auxiliary boom section F_B, the length of the web member F_FL, the width of the web member spacing F_FX, the diameter of the chord member F_XD, and the diameter of the web member F_FD.
[0037] The main boom neural network can output the optimal design parameters of the main boom to be designed based on the input first geometric parameters and first load parameters. The optimal design parameters output by the main boom neural network can include basic boom section parameters such as the straight edge outward opening Z1r of the first main boom cover plate, the bending angle opening Z1b of the cover plate, the height of the cover plate Z1h, and the included angle Z1θ1 between the upper and lower cover plates, as well as stiffening plate parameters. The stiffening plate is a thin plate attached to the inner and outer surfaces of the main boom, serving a local reinforcement function. The stiffening plate parameters include the length Z_JL, width Z_JB, and thickness Z_JH of the stiffening plate. After inputting the optimal design parameters of the main boom into the super neural network, the super neural network can output the optimal design parameters of the superlift to be designed. The super optimal design parameters output by the super neural network can include the length of the superlift pull plate C_LL, the length of the superlift mast C_L, the hoisting radius C_R, and the superlift mast opening angle C_α, etc. After obtaining the super optimal design parameters, the processor can input the super optimal design parameters into the secondary arm neural network. The secondary arm neural network outputs the optimal design parameters of the secondary arm to be designed. The optimal design parameters of the secondary arm output by the secondary arm neural network may include the secondary arm length F_L, the secondary arm cross-sectional width F_B, the web member length F_FL, the web member spacing width F_FX, the chord diameter F_XD, the web member diameter F_FD, etc.
[0038] In one embodiment, vehicle parameters of multiple existing construction machinery are obtained, including the second geometric parameters of the boom segment and the second load parameters of the boom; the second geometric parameters and the second load parameters are input into the vehicle neural network to train the vehicle neural network; after the vehicle neural network is trained, the second geometric parameters and the second load parameters are input into the trained vehicle neural network again; the predicted design parameters output by the trained vehicle neural network are obtained, including the predicted design parameters of the main boom, the predicted design parameters of the superlift, and the predicted design parameters of the auxiliary boom; the main boom neural network, the superlift neural network, and the auxiliary boom neural network are trained in a supervised manner using the predicted design parameters.
[0039] The processor can establish and train the main boom neural network, superlift neural network, and auxiliary boom neural network. The processor can acquire vehicle parameters from multiple existing construction machinery, which may include the second geometric parameters of the entire boom segment and the second load parameters of the entire boom. The processor can establish a whole-vehicle neural network and train it by inputting the second geometric parameters and second load parameters. After the whole-vehicle neural network is trained, the processor can again input the second geometric parameters and second load parameters into the trained whole-vehicle neural network, thereby outputting predicted design parameters, including predicted design parameters for the main boom, superlift, and auxiliary boom. The predicted design parameters output by the trained whole-vehicle neural network are used to perform supervised training of the main boom neural network, superlift neural network, and auxiliary boom neural network.
[0040] In one embodiment, training the main boom neural network, the super-lift neural network, and the auxiliary boom neural network using predictive design parameters includes: acquiring sample geometric parameters and sample load parameters of multiple existing construction machinery. The sample geometric parameters include the number of boom sections Z_X, the length of each boom section Z_L, the maximum height of the basic boom cross-section Z_H, and the maximum width of the basic boom cross-section Z_B. The sample load parameters include the working radius Z of the boom under each working condition. i _D, Load size Z i _F, Elevation Angle Z i _θ; Set the first error parameter and first step length of the main arm neural network; After normalizing the sample geometric parameters and sample load parameters, input the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network in sequence; Obtain multiple main arm design parameters output by the main arm neural network under supervised training, wherein the number of multiple main arm design parameters is determined according to the first error parameter and the first step length; Repeat the step of inputting the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network until the main arm neural network is trained.
[0041] The processor can acquire sample set parameters and sample load parameters from multiple existing construction machinery. The sample set parameters include the number of boom sections Z_X, the length of each boom section Z_L, the maximum height of the basic boom cross-section Z_H, and the maximum width of the basic boom cross-section Z_B. The sample load parameters include the boom's working radius Z under each working condition. i _D, Load size Z i _F, Elevation Angle Z iThe processor normalizes the sample set parameters and sample payload parameters, then sequentially inputs the processed sample geometric parameters, sample payload parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network. This yields multiple main arm design parameters output by the main arm neural network under supervised training. The processor then obtains the actual main arm design parameters output by the main arm neural network and compares these actual parameters with the predicted design parameters to determine the error between them. The processor can set the first error parameter and the first step length of the main arm neural network. The first error parameter is used to judge the error between the actual main arm design parameters and the predicted design parameters. For example, the processor can set error parameters Z-maxloss and Z-minloss for the main arm neural network. The main arm neural network is trained by repeatedly executing the steps of inputting the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters. This continues until the error between the actual and predicted main arm design parameters reaches Z-maxloss. Then, the processor can start recording the design parameters output by the main arm neural network according to the set first step length, until the error between the main arm design parameters output by the main arm neural network and the predicted design parameters reaches Z-minloss. The processor can compare and calculate the recorded data to select the optimal data as the final output of the main arm neural network. Here, the processor can determine the optimal data from the recorded data according to the user's selection instructions and determine it as the design parameter corresponding to the main arm structure.
[0042] In one embodiment, training the main arm neural network, the super-start neural network, and the secondary arm neural network using predicted design parameters includes: setting a second error parameter and a second step size for the super-start neural network; inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-start in the predicted design parameters to the super-start neural network; obtaining multiple super-start design parameters output by the super-start neural network under supervised training, wherein the number of multiple super-start design parameters is determined according to the second error parameter and the second step size; and repeatedly performing the step of inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-start in the predicted design parameters to the super-start neural network until the super-start neural network is trained.
[0043] The processor can set a second error parameter and a second step size for the super-giant neural network. The processor can input the main arm design parameters output by the trained main arm neural network and the super-giant prediction design parameters from the preset design parameters into the super-giant neural network. Multiple super-giant design parameters output by the super-giant neural network under supervised training are obtained, where the number of multiple super-giant design parameters is determined by the second error parameter and the second step size. After inputting the main arm design parameters output by the main arm neural network and the super-giant prediction design parameters from the preset design parameters into the super-giant neural network, the processor can output the actual super-giant design parameters through the super-giant neural network. The processor can compare the obtained actual super-giant design parameters with the predicted super-giant design parameters to obtain the error between the two, and train the super-giant neural network using the second error parameter and the second step size set by the processor. For example, for a super-giant neural network, the processor can set the second error parameters as C-maxloss and C-minloss. By repeatedly executing the step of inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the supergiant from the predicted design parameters into the supergiant neural network, the process continues until the error between the supergiant design parameters output by the super-neural network and the predicted design parameters of the supergiant is between C-maxloss and C-minloss. The processor can select several supergiant design parameters output by the supergiant neural network when the error between the supergiant design parameters output by the supergiant neural network and the predicted supergiant design parameters is between C-maxloss and C-minloss, based on a second step size. The processor can perform comparative calculations on the recorded data to select the optimal data as the final output of the supergiant neural network. Here, the processor can determine the optimal data from the recorded data according to the user's selection instructions and determine it as the supergiant design parameter.
[0044] In one embodiment, training the main arm neural network, the super-start neural network, and the secondary arm neural network using predicted design parameters includes: setting a third error parameter and a third step size for the secondary arm neural network; inputting the super-start predicted design parameters output by the trained super-start neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network; obtaining multiple secondary arm design parameters output by the secondary arm neural network under supervised training, wherein the number of multiple secondary arm design parameters is determined based on the third error parameter and the third step size; and repeatedly performing the step of inputting the super-start design parameters output by the trained super-start neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network until the secondary arm neural network is trained.
[0045] The processor can set the third error parameter and the third step size of the secondary arm neural network. The processor can input the hyper-engine design parameters output by the trained hyper-engine neural network and the predicted secondary arm design parameters from the preset design parameters into the secondary arm neural network. Multiple secondary arm design parameters output by the secondary arm neural network under supervised training are obtained; the number of these parameters is determined by the third error parameter and the third step size. After inputting the hyper-engine design parameters output by the hyper-engine neural network and the predicted secondary arm design parameters from the preset design parameters into the secondary arm neural network, the processor can output the actual secondary arm design parameters from the secondary arm neural network. The processor can compare the obtained actual secondary arm design parameters with the predicted secondary arm design parameters to obtain the error between them, and then train the secondary arm neural network using the third error parameter and the third step size set by the processor. For example, for a secondary arm neural network, the processor can set a third error parameter as F-maxloss and F-minloss. By repeatedly executing the step of inputting the hyper-engine design parameters output by the trained hyper-engine neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network, the process continues until the error between the secondary arm design parameters output by the secondary arm neural network and the predicted design parameters of the secondary arm is between F-maxloss and F-minloss. The processor can select several secondary arm design parameters output by the secondary arm neural network based on the third step size when the error between the output and predicted secondary arm design parameters is between F-maxloss and F-minloss. The processor can perform comparative calculations on the recorded data to select the optimal data as the final output of the secondary arm neural network. Here, the processor can determine the optimal data from the recorded data according to the user's selection instructions and determine it as the design parameters for the secondary arm.
[0046] In one embodiment, a processor is provided, configured to perform a method for determining boom parameters according to any one of the foregoing.
[0047] like Figure 2 The diagram illustrates a schematic flowchart of a method for determining boom parameters according to an embodiment of this application. It includes the following steps:
[0048] Step 201: Design the main arm neural network, the super-branch neural network, and the secondary arm neural network;
[0049] Step 202: Perform theoretical or optimization processing on the initial samples;
[0050] Step 203: Organize existing crane-related implementation data as network training samples;
[0051] Step 204: Train the vehicle neural network using training samples to obtain the corresponding vehicle neural network model;
[0052] Step 205, Vehicle geometric parameters and operating load parameters;
[0053] Step 206: Input the main arm neural network and iterate the main arm neural network by setting the error Z-maxloss and Z-minloss;
[0054] Step 207: Select multiple main arm design parameters according to the first preset step size;
[0055] Step 208: Select the optimal main boom design parameters;
[0056] Step 209: Input the supercomputer neural network and iterate the supercomputer neural network by setting the error C-maxloss and C-minloss;
[0057] Step 210: Select multiple super-start design parameters according to the second preset step size;
[0058] Step 211: Select the optimal overstart design parameters;
[0059] Step 212: Input the secondary arm neural network and iterate the secondary arm neural network by setting the error F-maxloss and F-minloss;
[0060] Step 213: Select multiple auxiliary arm design parameters according to the third preset step size;
[0061] Step 214: Select the optimal auxiliary boom design parameters.
[0062] The processor can design neural networks for the main boom, super boom, and auxiliary boom. It can organize existing crane design example data as training samples for the neural networks and perform theoretical or optimization processing on the initial training samples. The processor can then use the training samples to train the neural networks, thereby obtaining the corresponding neural network model. The processor can acquire the overall vehicle's geometric parameters and operating load parameters. The geometric parameters may include the number of known boom sections Z_X, the length of each known boom section Z_L, the maximum height of the basic boom section Z_H, and the maximum width of the basic boom section Z_B, etc. The load parameters include the known boom's working radius Z under each operating condition. i _D, Load size Z i _F, Elevation Angle Z i _θ etc.
[0063] The processor can build a vehicle-wide neural network and train it by inputting training samples. After the vehicle-wide neural network is trained, a corresponding neural network model is obtained. The processor can then input the sample training parameters back into the trained vehicle-wide neural network, thereby outputting predicted design parameters. These predicted design parameters include those for the main arm, super-lift, and auxiliary arm. The predicted design parameters output by the trained vehicle-wide neural network are used to perform supervised training on the subsequent main arm neural network, super-lift neural network, and auxiliary arm neural network.
[0064] The processor can input the vehicle's geometric parameters and load parameters into the main arm neural network to obtain multiple main arm design parameters output by the main arm neural network under supervised training. For the main arm neural network, the processor can set error parameters Z-maxloss and Z-minloss to iterate the network. The main arm neural network can output main arm design parameters based on the input vehicle geometric parameters and load parameters. The processor can compare the main arm design parameters output by the main arm neural network with the predicted design parameters corresponding to the vehicle geometric parameters and load parameters, repeatedly executing the step of inputting the vehicle geometric parameters and load parameters into the main arm neural network until the error between the two is between Z-maxloss and Z-minloss. At this point, the processor can determine that the main arm neural network training is complete. The processor can select multiple main arm design parameters output by the main arm neural network according to a first preset step size and choose the optimal main arm design parameter from among them.
[0065] The processor inputs the selected optimal main boom design parameters into the super-lift neural network. For the super-lift neural network, the processor can set error parameters C-maxloss and C-minloss to iterate the super-lift neural network. The super-lift neural network can output super-lift design parameters based on the input optimal main boom design parameters. The processor can compare the super-lift design parameters output by the super-lift neural network with the super-lift predicted design parameters corresponding to the vehicle's geometric parameters and working load parameters. The step of inputting the main boom design parameters output by the trained main boom neural network into the super-lift neural network is repeated until the error between the two is between C-maxloss and C-minloss. At this point, the processor can determine that the super-lift neural network training is complete. The processor can select multiple super-lift design parameters output by the super-lift neural network according to a second preset step size, and select the optimal super-lift design parameter from the multiple super-lift design parameters.
[0066] The processor inputs the selected optimal hyperlift design parameters into the secondary arm neural network. For the secondary arm neural network, the processor can set error parameters F-maxloss and F-minloss to iterate the secondary arm neural network. The secondary arm neural network can output secondary arm design parameters based on the input optimal hyperlift design parameters. The processor can compare the secondary arm design parameters output by the secondary arm neural network with the predicted secondary arm design parameters corresponding to the vehicle's geometric parameters and working load parameters. The step of inputting the hyperlift design parameters output by the trained hyperlift neural network into the secondary arm neural network is repeated until the error between the two is between F-maxloss and F-minloss. The processor can then determine that the secondary arm neural network training is complete. The processor can select multiple secondary arm design parameters output by the secondary arm neural networks according to a third preset step size, and select the optimal secondary arm design parameters from the multiple secondary arm design parameters.
[0067] After obtaining the optimal parameters for the boom design, the processor can store these parameters in a database and use them to train the vehicle's neural network, resulting in a neural network model corresponding to the boom design. When the user operates the system, they can input the vehicle's geometric parameters and operating load parameters into the neural network model to obtain the corresponding design parameters for the entire vehicle.
[0068] The aforementioned technical solution includes a three-level neural network, with each level interconnected. Ultimately, the design parameters of the entire boom system can be derived using initial design parameters and operating conditions. The processor can train each neural network using sample parameters and set corresponding errors and step sizes for each network to select the optimal design parameters. After obtaining the optimal design parameters, the entire vehicle's neural network is trained using these parameters, enabling it to quickly output the boom's design parameters based on geometric parameters and operating load parameters.
[0069] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.
[0070] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3As shown. The computer device includes a processor A01, a network interface A02, a memory (not shown), and a database (not shown) connected via a system bus. The processor A01 provides computing and control capabilities. The memory includes internal memory A03 and a non-volatile storage medium A04. The non-volatile storage medium A04 stores an operating system B01, a computer program B02, and a database (not shown). The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 in the non-volatile storage medium A04. The database stores relevant data about the construction machinery and data input by the operators. The network interface A02 communicates with external terminals via a network connection. When executed by the processor A01, the computer program B02 implements a method for determining boom parameters.
[0071] Figure 1 This is a flowchart illustrating a method for determining boom parameters in one embodiment. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
[0072] This application provides a device including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: obtaining the first geometric parameters of a known boom segment and the first load parameters of a known boom under multiple working conditions; inputting the first geometric parameters and the first load parameters into a main boom neural network to output the optimal design parameters of the main boom to be designed through the main boom neural network; inputting the optimal design parameters of the main boom into a superlift neural network to output the optimal design parameters of the superlift to be designed through the superlift neural network; and inputting the optimal design parameters of the superlift into a secondary boom neural network to output the optimal design parameters of the secondary boom to be designed through the secondary boom neural network.
[0073] In one embodiment, the first geometric parameters include the number of known boom sections Z_X, the length of each known boom section Z_L, the maximum height of the basic boom section Z_H, and the maximum width of the basic boom section Z_B. The first load parameters include the working radius Z of the known boom under each working condition. i _D, Load size Z i _F, Elevation Angle Z i _θ.
[0074] In one embodiment, the optimal design parameters for the main boom include the straight edge outward opening Z1r of the upper cover plate of the first main boom section, the bending angle opening Z1b of the upper cover plate, the height of the upper cover plate Z1h, the included angle Z1θ1 between the upper and lower cover plates, and the parameters of the reinforcing plate; the optimal design parameters for the superlift include the length of the superlift pull plate C_LL, the length of the superlift mast C_L, the hoisting radius C_R, and the opening angle of the superlift mast C_α; the optimal design parameters for the auxiliary boom include the length of the auxiliary boom F_L, the width of the auxiliary boom section F_B, the length of the web member F_FL, the width of the web member spacing F_FX, the diameter of the chord member F_XD, and the diameter of the web member F_FD.
[0075] In one embodiment, vehicle parameters of multiple existing construction machinery are obtained, including the second geometric parameters of the boom segment and the second load parameters of the boom; the second geometric parameters and the second load parameters are input into the vehicle neural network to train the vehicle neural network; after the vehicle neural network is trained, the second geometric parameters and the second load parameters are input into the trained vehicle neural network again; the predicted design parameters output by the trained vehicle neural network are obtained, including the predicted design parameters of the main boom, the predicted design parameters of the superlift, and the predicted design parameters of the auxiliary boom; the main boom neural network, the superlift neural network, and the auxiliary boom neural network are trained in a supervised manner using the predicted design parameters.
[0076] In one embodiment, training the main boom neural network, the super-lift neural network, and the auxiliary boom neural network by predicting design parameters includes: acquiring sample geometric parameters and sample load parameters of multiple existing construction machinery. The sample geometric parameters include the number of boom sections Z_X, the length of each boom section Z_L, the maximum height of the basic boom cross-section Z_H, and the maximum width of the basic boom cross-section Z_B. The sample load parameters include the working radius Z of the boom under each working condition. i _D, Load size Z i _F, Elevation Angle Z i_θ; Set the first error parameter and first step length of the main arm neural network; After normalizing the sample geometric parameters and sample load parameters, input the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network in sequence; Obtain multiple main arm design parameters output by the main arm neural network under supervised training, wherein the number of multiple main arm design parameters is determined according to the first error parameter and the first step length; Repeat the step of inputting the processed sample geometric parameters, sample load parameters, and the predicted design parameters of the main arm from the predicted design parameters into the main arm neural network until the main arm neural network is trained.
[0077] In one embodiment, training the main arm neural network, the super-start neural network, and the secondary arm neural network using predicted design parameters includes: setting a second error parameter and a second step size for the super-start neural network; inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-start in the predicted design parameters to the super-start neural network; obtaining multiple super-start design parameters output by the super-start neural network under supervised training, wherein the number of multiple super-start design parameters is determined according to the second error parameter and the second step size; and repeatedly performing the step of inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-start in the predicted design parameters to the super-start neural network until the super-start neural network is trained.
[0078] In one embodiment, training the main arm neural network, the super-start neural network, and the secondary arm neural network using predicted design parameters includes: setting a third error parameter and a third step size for the secondary arm neural network; inputting the super-start predicted design parameters output by the trained super-start neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network; obtaining multiple secondary arm design parameters output by the secondary arm neural network under supervised training, wherein the number of multiple secondary arm design parameters is determined based on the third error parameter and the third step size; and repeatedly performing the step of inputting the super-start design parameters output by the trained super-start neural network and the predicted design parameters of the secondary arm from the predicted design parameters into the secondary arm neural network until the secondary arm neural network is trained.
[0079] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0080] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this 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 processor, 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 a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0081] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0083] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0084] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.
[0085] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.
[0086] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0087] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for determining boom parameters, characterized in that, The boom comprises multiple boom sections, and the boom parameters include design parameters for the main boom, the auxiliary boom, and the superlift. The method includes: The first geometric parameters of a known boom segment and the first load parameters of a known boom under multiple working conditions are obtained, as well as the vehicle parameters of multiple existing construction machinery. The vehicle parameters include the second geometric parameters of the entire boom segment and the second load parameters of the entire boom. The second geometric parameters and the second load parameters are input into the vehicle neural network to train the vehicle neural network; After the vehicle neural network is trained, the second geometric parameters and the second load parameters are input into the trained vehicle neural network again. Obtain the predicted design parameters output by the trained vehicle neural network, including the predicted design parameters of the main boom, the predicted design parameters of the super-lift, and the predicted design parameters of the auxiliary boom. The main arm neural network, super-branch neural network, and secondary arm neural network are trained in a supervised manner using the aforementioned predictive design parameters. The first geometric parameters and the first load parameters are input into the trained main arm neural network so that the main arm neural network can output the optimal design parameters of the main arm to be designed. The optimal design parameters of the main arm are input into the trained super-start neural network, so that the optimal design parameters of the super-start to be designed are output through the super-start neural network. The optimal design parameters of the super-start are input into the trained secondary arm neural network, so that the optimal design parameters of the secondary arm to be designed are output through the secondary arm neural network.
2. The method according to claim 1, characterized in that, The first geometric parameters include the number of known boom sections Z_X, the length of each known boom section Z_L, the maximum height of the basic boom section Z_H, and the maximum width of the basic boom section Z_B. The first load parameters include the working radius of the known boom under each working condition. Load size Angle of elevation .
3. The method according to claim 1, characterized in that, The optimal design parameters for the main boom include the outward opening of the straight edge of the first section of the main boom's upper cover plate. , Upper cover plate bend angle opening Height of the top cover plate Angle between the upper and lower cover plates And the parameters of the reinforcing plate; The optimal design parameters for the superlift include the superlift pull plate length C_LL, the superlift mast length C_L, the hoist radius C_R, and the superlift mast opening angle C_α; The optimal design parameters for the auxiliary arm include the auxiliary arm length F_L, the auxiliary arm cross-sectional width F_B, the web member length F_FL, the web member spacing width F_FX, the chord diameter F_XD, and the web member diameter F_FD.
4. The method according to claim 1, characterized in that, The step of training the main arm neural network, the super-branch neural network, and the secondary arm neural network using the predicted design parameters includes: Obtain sample geometric parameters and sample load parameters for multiple existing construction machinery. The sample geometric parameters include the number of boom sections Z_X, the boom section length Z_L, the maximum height of the basic boom section Z_H, and the maximum width of the basic boom section Z_B. The sample load parameters include the boom's working radius under each working condition. Load size Angle of elevation ; Set the first error parameter and first step length of the main arm neural network; After normalizing the sample geometric parameters and the sample load parameters, the processed sample geometric parameters, sample load parameters, and the prediction design parameters of the main arm in the prediction design parameters are sequentially input into the main arm neural network. Obtain multiple main arm design parameters output by the main arm neural network under supervised training, wherein the number of the multiple main arm design parameters is determined based on the first error parameter and the first step length; Repeat the step of inputting the processed sample geometric parameters, sample load parameters, and the main arm's predicted design parameters from the predicted design parameters into the main arm neural network until the main arm neural network is trained.
5. The method according to claim 1, characterized in that, The step of training the main arm neural network, the super-branch neural network, and the secondary arm neural network using the predicted design parameters includes: Set the second error parameter and the second step size for the hyper-neural network; The main arm design parameters output by the trained main arm neural network and the predicted design parameters of the hyperbola in the predicted design parameters are input into the hyperbola neural network. Obtain multiple hyper-start design parameters output by the hyper-start neural network under supervised training, wherein the number of the multiple hyper-start design parameters is determined based on the second error parameter and the second step size; Repeat the step of inputting the main arm design parameters output by the trained main arm neural network and the predicted design parameters of the super-giant in the predicted design parameters into the super-giant neural network until the super-giant neural network is trained.
6. The method according to claim 1, characterized in that, The step of training the main arm neural network, the super-branch neural network, and the secondary arm neural network using the predicted design parameters includes: Set the third error parameter and the third step size for the secondary arm neural network; The super-start prediction design parameters output by the super-start neural network after training, as well as the prediction design parameters of the secondary arm in the prediction design parameters, are input into the secondary arm neural network. Obtain multiple secondary arm design parameters output by the secondary arm neural network under supervised training, wherein the number of the multiple secondary arm design parameters is determined based on the third error parameter and the third step size; Repeat the step of inputting the super-engine design parameters output by the trained super-engine neural network and the predicted design parameters of the secondary arm in the predicted design parameters into the secondary arm neural network until the secondary arm neural network is trained.
7. A processor, characterized in that, It is configured to perform the method for determining boom parameters as described in any one of claims 1 to 6.
8. A device for determining boom parameters, characterized in that, Includes the processor according to claim 7.
9. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, this instruction causes the processor to be configured to perform the method for determining boom parameters according to any one of claims 1 to 6.