CN-121982249-A - Storehouse three-dimensional model componentization construction method and system based on digital twin

CN121982249ACN 121982249 ACN121982249 ACN 121982249ACN-121982249-A

Abstract

The application relates to the technical field of three-dimensional modeling, in particular to a method and a system for constructing a warehouse three-dimensional model assembly based on digital twinning, wherein the method comprises the steps of dividing a three-dimensional space into a plurality of grid units according to the distribution density of material point location data; setting assembly priority for each three-dimensional component according to the support dependency relationship, initializing the wave function state of each grid unit, calculating shannon entropy of each grid unit, carrying out weighted correction on shannon entropy by utilizing the assembly priority to obtain a weighted entropy value, selecting the grid unit with the smallest weighted entropy value for carrying out state collapse to determine the grid unit as a specific three-dimensional component, updating the feasible state set of the adjacent grid units according to the adjacent constraint rule until all the grid unit states are determined, and generating a storehouse three-dimensional model. According to the technical scheme, the storehouse three-dimensional model which accords with the physical rule and has high precision can be quickly generated.

Inventors

ZHOU QIAN
ZHONG TONGQING

Assignees

湖北华中电力科技开发有限责任公司

Dates

Publication Date: 20260505
Application Date: 20251211

Claims (10)

1. The method for constructing the digital twin-based storehouse three-dimensional model assembly is characterized by comprising the steps of obtaining material point position data of a storehouse, dividing a three-dimensional space into a plurality of grid units according to the distribution density of the material point position data, wherein the sizes of the grid units are inversely related to the distribution density; Establishing a component model library comprising a plurality of three-dimensional components, defining an adjacency constraint rule among the three-dimensional components, and setting assembly priority for the three-dimensional components according to the support dependency relationship; Initializing the wave function state of the grid units according to the wave function collapse algorithm, calculating shannon entropy of each grid unit, and carrying out weighting correction on the shannon entropy by utilizing the assembly priority to obtain a weighted entropy value; And updating a feasible state set of the adjacent grid units according to the adjacent constraint rule, and repeatedly executing the steps of state collapse and updating until all the states of the grid units are determined, and then instantiating each three-dimensional component to generate a storehouse three-dimensional model.
2. The method for componentized construction of a digital twin-based three-dimensional model of a warehouse of claim 1, wherein the dividing the three-dimensional space into a plurality of grid cells according to the distribution density of the material point data comprises: Dividing the three-dimensional space into a plurality of macro areas, counting the number of material point location data in each macro area to construct a material density field, carrying out normalization processing on the material density field to obtain the distribution density of each position in the three-dimensional space, calculating corresponding grid step length according to the distribution density of each position, and dividing each macro area according to the grid step length to generate non-uniform grid units.
3. The method for constructing the digital twin-based storehouse three-dimensional model assembly according to claim 2, wherein the method for calculating the grid step length comprises the following steps: Setting a basic grid step length and a density sensitivity coefficient, taking the basic grid step length as a numerator, taking the sum of products of 1 and the distribution density and the density sensitivity coefficient as a denominator, taking the ratio of the two as a first step length, and taking the maximum value between the first step length and the minimum value of a preset step length as the grid step length.
4. The method for constructing the digital twin-based storehouse three-dimensional model assembly according to claim 1 is characterized in that the setting of the assembly priority for each three-dimensional assembly according to the supporting dependency relationship comprises a basic assembly, a skeleton assembly, a structural assembly and a material assembly, wherein the assembly priority of the basic assembly is 0 and comprises a ground and a wall, the assembly priority of the skeleton assembly is 1 and comprises a shelf upright post and a stacker track, the assembly priority of the structural assembly is 2 and comprises a shelf cross beam and a laminate, and the assembly priority of the material assembly is 3 and comprises a tray and goods.
5. The method of constructing a digital twin based three-dimensional model of a warehouse of claim 1, wherein the performing a weighted correction on shannon entropy using assembly priorities to obtain weighted entropy values comprises: Calculating a priority attenuation coefficient which is inversely related to the assembly priority of the three-dimensional component with the highest probability in the candidate component set of the grid unit; And multiplying the shannon entropy of the grid unit by the priority attenuation coefficient to obtain a weighted entropy value.
6. The method of digital twinning-based warehouse three-dimensional model componentization construction of claim 1, wherein initializing the wave function state of the grid cells in accordance with a wave function collapse algorithm comprises: judging whether material point position data exist in the grid unit or not; Setting the initial existence probability of the corresponding material component in the grid unit to be a first probability value in response to the material point data, wherein the first probability value is larger than the sum of probability values of the rest feasible components; And in response to the absence of the material point data, maintaining the grid unit in a full stack state, wherein the probabilities of all three-dimensional components except the material components in the full stack state are the same.
7. The method for componentized construction of a digital twin based library three-dimensional model according to claim 1, wherein during the repeated execution of the state collapse and update steps, the construction method further comprises a dynamic subdivision step: the method comprises the steps of calculating the spatial change rate of shannon entropy between a current grid cell and an adjacent grid cell to obtain a conflict gradient, splitting the current grid cell into a plurality of sub-grid cells in response to the conflict gradient meeting preset subdivision conditions, and re-executing the state collapse and updating steps within the range of the sub-grid cells.
8. The method for constructing the digital twin-based storehouse three-dimensional model assembly according to claim 7, wherein the method for calculating the collision gradient comprises the following steps: Obtaining the maximum shannon entropy of all grid units in the primary iteration, and carrying out normalization processing on the shannon entropy of the current grid unit by utilizing the maximum shannon entropy; calculating the difference value of the normalized entropy value of the current grid unit and the adjacent grid unit in each direction of the three-dimensional space, calculating the ratio of the difference value to the size of the grid unit, and taking the average value of all the ratios as a collision gradient.
9. The method for componentized construction of a digital twin-based library three-dimensional model of claim 7, wherein the predetermined subdivision condition comprises a collision gradient of the current grid cell being greater than a predetermined gradient threshold, a normalized entropy value of the current grid cell being greater than a predetermined uncertainty threshold, and a number of candidate components in the current grid cell being greater than or equal to 2.
10. A digital twin based three-dimensional model building system comprising a processor and a memory, the memory storing computer program instructions which, when executed by the processor, implement a digital twin based three-dimensional model building method according to any one of claims 1 to 9.

Description

Storehouse three-dimensional model componentization construction method and system based on digital twin Technical Field The application relates to the technical field of three-dimensional modeling, in particular to a warehouse three-dimensional model componentization construction method and system based on digital twinning. Background Along with the development of smart power grids and digital twin technologies, management and control of electric power materials in a warehouse also tend to be intelligent. As a key link for realizing visual management and control of materials, constructing a high-fidelity three-dimensional model of a warehouse, and has important significance for improving warehouse management efficiency, optimizing space utilization and guaranteeing operation safety. Therefore, how to realize the automatic construction of the three-dimensional model of a large-scale storehouse scene on the premise of ensuring the model precision is a problem to be solved urgently. In the existing three-dimensional model construction process, picture data in a storehouse are often required to be acquired, or a laser radar is adopted to scan the storehouse to acquire point cloud data, so that a three-dimensional model of the storehouse is reconstructed. However, laser scanning and image data acquisition depend on expensive acquisition equipment, and acquired mass point clouds or image data need to be subjected to complex processing, so that the requirement of rapid model updating under frequent change of warehouse materials is difficult to meet, and a three-dimensional warehouse model cannot be quickly and accurately constructed. Disclosure of Invention In order to solve the technical problem that a three-dimensional model of a storehouse cannot be quickly and accurately built, the application provides a digital twinning-based storehouse three-dimensional model componentization building method and system, which can quickly generate a storehouse three-dimensional model which accords with a physical rule and has high precision. The application provides a digital twinning-based storehouse three-dimensional model assembly construction method, which comprises the steps of obtaining material point position data of a storehouse, dividing a three-dimensional space into a plurality of grid units according to distribution density of the material point position data, wherein the size and the distribution density of the grid units are inversely related, establishing an assembly model library comprising a plurality of three-dimensional assemblies, defining an adjacent constraint rule among the three-dimensional assemblies, setting assembly priorities for the three-dimensional assemblies according to supporting dependency relations, initializing wave function states of the grid units according to a wave function collapse algorithm, calculating shannon entropy of each grid unit, carrying out weighting correction on the shannon entropy by utilizing the assembly priorities to obtain a weighted entropy value, selecting the grid unit with the smallest weighted entropy value for carrying out state collapse, determining the grid unit as a specific three-dimensional assembly according to the material point position data, updating a feasible state set of an adjacent grid unit according to the adjacent constraint rule, and repeatedly executing the state collapse and updating steps until all three-dimensional assemblies are determined to instantiate the three-dimensional assemblies to generate a storehouse three-dimensional model. The method has the advantages that the grid units with the size inversely related to the density are generated according to the distribution density of the material point data, the self-adaptive discretization of the three-dimensional space is realized, the total number of grids is greatly reduced on the premise of guaranteeing the assembly accuracy, meanwhile, the shannon entropy is corrected by defining the adjacency constraint rule and the assembly priority and utilizing the assembly priority weighting, the physical support logic corresponding to the assembly priority is followed in the wave function collapse process, the topological errors such as component suspension are fundamentally eliminated, and the physical reality of the generated model is guaranteed. Preferably, dividing the three-dimensional space into a plurality of grid units according to the distribution density of the material point location data comprises dividing the three-dimensional space into a plurality of macroscopic regions, counting the number of the material point location data in each macroscopic region to construct a material density field, carrying out normalization processing on the material density field to obtain the distribution density of each position in the three-dimensional space, calculating corresponding grid step lengths according to the distribution density of each position, and dividing each macroscopic region acc