Performance optimization method for power digital twin model

Abstract

The invention discloses an electric power digital twin model performance optimization method, belongs to the technical field of three-dimensional model processing, aims to solve the technical problem of electric power digital twin model performance optimization, and realizes model lightweight processing by combining 3dsMax, Maya and other modeling software. The model loading speed is improved, the problems of frame loss, excessive resource occupation and collapse after model loading are solved, and the method comprises the following steps: step 1, determining a lagging phenomenon and a range; step 2, troubleshooting problem reasons; 3, carrying out model optimization implementation; the model loading speed is increased in a mode of surface reduction, material adjustment and node combination of the model; through model hierarchical structure optimization, surface number optimization and material optimization, the operability of the model after loading is improved, so that frame loss is avoided after the model is loaded; after the model is optimized, memory occupied by the model is reduced, resources occupied during loading are reduced, and the requirement for used hardware is lowered.

Description

Technical Field

[0001] A method for optimizing the performance of a power digital twin model belongs to the technical field of three-dimensional model processing, and specifically relates to the technical field of methods for optimizing the performance of a power digital twin model. Background Art

[0002] Main relevant contents of existing technology: Manual modeling: Models are created manually using specialized software (3dsMax, Maya, etc.). This approach relies on the designer's spatial imagination and operational skills, starting with basic geometric shapes (cubes, spheres, etc.). Details are then shaped through operations like "extrusion," "cutting," and "subdivision," and then given materials and textures. Scanning and reconstruction: Data is collected by scanning real-world objects using physical equipment (such as laser scanners). The resulting point cloud data is converted into a 3D model, achieving a high-precision "real-to-virtual" reproduction. Procedural modeling: Complex models are automatically generated through algorithms, rules, or node logic, eliminating the need for manual editing of details. This approach is suitable for large-scale, repetitive scenes.

[0003] Analysis of problems and causes of existing technologies: Slow model loading: In the process of using modeling software to create models, in order to achieve the original geometric features and visual fidelity, there will be too many faces, too many materials, and too many nodes, resulting in slow model loading. Loading jams lead to model frame drops: After the model is loaded, the model will jam when operating the model, resulting in unclear model loading and frame drops. Large resource usage: When a twin scene is composed of many models, each model has too many faces, too many materials, and too many redundant nodes, the computer resources occupied during loading will increase, resulting in loading crashes. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the existing defects and provide a method for optimizing the performance of electric power digital twin models, thereby solving the technical method for optimizing the performance of electric power digital twin models. By combining modeling software such as 3dsMax and Maya, the model lightweight processing is realized, the model loading speed is improved, and the problems of frame loss after model loading and crash caused by excessive resource occupation are solved.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for optimizing the performance of a power digital twin model includes the following steps: Step 1: Confirm the jamming phenomenon and scope ①Record the frequency of freezes (continuous / intermittent) and triggering conditions (when operating the device, loading scenes); ② Identify lag symptoms, including frame rate drops, operation delays, data refresh lags, or system crashes; Step 2: Problem Cause Investigation ① Hardware Resource Investigation a Real-time monitoring of CPU utilization during lag periods, marked as abnormal when usage exceeds 80%; b Using GPU performance analysis tools to detect shader load during three-dimensional rendering, if texture sampling time consumption ratio > 50%, determine that the model precision is too high; c Monitor memory usage, including whether there is memory leakage (available memory continuously decreases over time) and insufficient memory caused by excessive model loading; ② Model Design Reasonability Investigation a Check the number of polygons, including redundant surfaces of individual device models and whether the total number of scene polygons exceeds the rendering threshold; b Detect whether the texture resolution exceeds the actual display requirements, and the number of transparent and reflective materials (such materials have 3-5 times the rendering calculation of ordinary materials); c Check whether there are redundant transformation nodes in a single model; Step 3: Model Optimization Implementation ① Model Face Number Optimization: Reduce the number of triangular facets by grid simplification and edge folding algorithm while maintaining the original geometric characteristics and visual fidelity; ② Model Material Optimization: Reduce the number of material balls and lower the texture resolution through material merging and texture compression techniques; integrate multiple maps using texture set and adjust model map UV; ③ Model Level Node Optimization: Merge redundant transformation nodes; merge sub-meshes with the same material and ensure that the UV and normal data after merging are correctly baked.

[0006] As a preferred technical solution of the present application, the grid simplification algorithm in step 3 needs to satisfy the condition that the error rate of key geometric features of the model is ≤5% while reducing the number of triangular facets.

[0007] As a preferred technical solution of the present application, the texture compression technique in step 3 uses ETC, ASTC, or BC series compression formats to reduce the map resolution from 4K to 1K while maintaining a visual quality loss rate ≤10%.

[0008] As a preferred technical solution of the present application, the level node optimization in step 3 also includes restructuring multiple components of a single device, reducing the number of objects by merging sub-meshes with the same material, and ensuring the correctness of UV and normal data after merging through baking.

[0009] As a preferred technical solution of the present application, the GPU performance analysis tool in step 2 is GPU-Z, which obtains shader load and texture sampling time consumption ratio data through this tool.

[0010] As a preferred technical solution of the present application, in step 3, the UV remapping technology is used to adjust the UV coordinates of the model map when the texture set integrates multiple maps, so as to realize the ordered arrangement of multiple maps in a single texture set.

[0011] As a preferred technical solution of the present application, in step 2, the memory monitoring tool is used to record the available memory change curve in real time, and when the available memory decreases by ≥20% within 30 consecutive minutes, it is determined that there is a memory leak.

[0012] As a preferred technical solution of the present application, in step 3, during the material merging process, the material balls with the same basic attributes are subjected to parameter unification processing, so as to retain the key visual feature parameters while reducing the number of material balls.

[0013] Compared with the prior art, the beneficial effects of the present application are: by using the three-dimensional model lag optimization method of the present application, after optimizing the number of faces, the material and the node of the model, significant effects are achieved: the loading speed of the model is improved by about 40%, the CPU usage is reduced by about 30%; the loaded model is not lagging, and the overall resource occupation is reduced by about 30%, effectively solving the lag problem in the loading and running process of the three-dimensional model, and improving the user experience.

[0014] By reducing the number of faces, adjusting the material, and merging the nodes, the model loading speed is improved. By optimizing the model hierarchy, the number of faces, and the material, the operability of the loaded model is improved, so that the model does not lose frames after loading. After model optimization, the model occupies less memory, and the resources occupied during loading will be reduced, and the hardware requirements will be reduced. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0016] The present application provides a technical solution: A power digital twin model performance optimization method, comprising the following steps: Lag phenomenon and range confirmation ① Record the frequency (continuous / intermittent) and triggering conditions (operating equipment, loading scene) of the lag occurrence; ② Identify the lag performance, including frame rate drop, operation delay, data refresh lag, or system crash. Through this step, the specific situation of the card can be accurately located, providing a clear direction for subsequent problem troubleshooting and optimization. Problem Cause Troubleshooting ① Hardware Resource Troubleshooting a Monitor the CPU utilization rate during the card period in real time, and mark it as abnormal when the utilization rate exceeds 80%. CPU is the core of processing data and instructions, and its high utilization rate will cause data processing to be not timely, thus causing card. b Use GPU performance analysis tools (such as GPU-Z) to detect the shader load during three-dimensional rendering. If the texture sampling time consumption ratio is >50%, it is determined that the model precision is too high. GPU is responsible for generating and drawing graphics in three-dimensional rendering, and excessive shader load will affect rendering speed. c Monitor the memory occupation state, including whether there is memory leakage (available memory continuously decreases over time) and insufficient memory caused by excessive model loading. Memory is a temporary storage space for data, and memory problems will affect model loading and running. ② Model Design Reasonability Troubleshooting a Check the number of polygons, including whether the redundant surfaces of a single device model and the total number of polygons in the scene exceed the rendering threshold. Excessive polygons will increase the rendering calculation amount, causing card. b Detect whether the texture resolution exceeds the actual display requirements, and the number of transparent and reflective materials (such materials have 3-5 times the rendering calculation amount of ordinary materials). High-resolution textures and complex materials will increase the computational burden of GPU. c Check whether there are redundant transformation nodes in a single model. Redundant nodes will increase the calculation amount of model loading and running. Model Optimization Implementation ① Model Face Number Optimization: Under the premise of maintaining the original geometric characteristics and visual fidelity, reduce the number of triangular patches through mesh simplification and edge folding algorithm. Among them, the mesh simplification algorithm needs to meet the error rate ≤5% of the key geometric characteristics of the model while reducing the number of triangular patches, to ensure that the optimized model can still meet the visual requirements. ② Model Material Optimization: Reduce the number of material balls and reduce the map resolution through material merging and texture compression technology. Texture compression technology uses ETC, ASTC or BC series compression format to reduce the map resolution from 4K to 1K while maintaining a visual quality loss rate ≤10%. At the same time, multiple maps are integrated into a single texture set using UV remapping technology to adjust the UV coordinates of the model map, achieving the orderly arrangement of multiple maps in a single texture set. In addition, during the material merging process, material balls with the same basic attributes are parameterized to reduce the number of material balls while retaining key visual feature parameters. ③Model level node optimization: merge redundant transformation nodes; merge submeshes with the same material and ensure that the UV and normal data after merging remain correct baking. In addition, it also includes restructuring multiple parts of a single device, reducing the number of objects by merging submeshes with the same material, and ensuring that the UV and normal data after merging remain correct through baking.

[0017] (I) Stutter phenomenon and range confirmation When running a three-dimensional model in a certain scene, users reported intermittent stuttering, mainly triggered when loading the scene, with the frame rate dropping from the normal 60 frames per second to below 30 frames per second, and the operation was significantly delayed. (II) Problem cause investigation Hardware resource investigation ① Real-time monitoring found that the CPU usage rate reached 85% during the stuttering period, exceeding the 80% threshold, which was an abnormal state. ② Use GPU-Z to detect the shader load during three-dimensional rendering, find that the texture sampling time consumption accounts for 55%, and determine that the model precision is too high. ③ Monitor memory usage, find that available memory continues to decrease over time, there is a memory leak problem, and at the same time, due to the large number of models loaded, there is also a shortage of memory. Model design rationality investigation ① Check the number of polygons, find that there are too many redundant surfaces in a single device model, high-precision modeling is used for models that should be simplified, and the total number of polygons in the scene exceeds the rendering threshold. ② Detect textures and materials, find that the device map uses 4K resolution, while the actual display only needs 1K, and there are a large number of transparent and reflective materials. ③ Check the scene level, find that there are redundant transformation nodes in a single model. (III) Model optimization implementation Model face number optimization: use mesh simplification, edge folding and other algorithms to reduce the number of high-polygon models, reduce the number of triangular patches while maintaining the error rate of key geometric features of the model ≤5%, and achieve model lightweight. Model material optimization: use material merging technology to unify the parameters of material balls with the same basic attributes, retain key visual feature parameters, reduce the number of material balls; use ETC compression format to compress textures, reduce 4K resolution maps to 1K, and maintain a visual quality loss rate ≤10%; use texture set integration to integrate multiple maps, use UV remapping technology to adjust the UV coordinates of the model map, and realize the ordered arrangement of multiple maps in a single texture set. Model level node optimization: merge redundant transformation nodes, restructure multiple parts of a single device, merge submeshes of the same material, and ensure that the UV and normal data after merging remain correct through baking. (Four) optimization results After the above optimization, the loading speed of the model is improved by about 40%, and the CPU usage is reduced by about 30%; the running model after loading is smooth and does not block, and the overall resource occupation is reduced by about 30%.

[0018] Although embodiments of the present application have been described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the performance of a power digital twin model, characterized by: The following steps are involved: Step 1: Confirm the jamming phenomenon and scope ①Record the frequency of freezes (continuous / intermittent) and triggering conditions (when operating the device, loading scenes); ② Identify lag symptoms, including frame rate drops, operation delays, data refresh lags, or system crashes; Step 2: Troubleshooting ①Hardware resource troubleshooting a) Monitor the CPU utilization during the lag period in real time and mark it as abnormal when the utilization exceeds 80%; b. Use GPU performance analysis tools to detect shader load during 3D rendering. If the texture sampling time accounts for >50%, the model accuracy is judged to be too high. c. Monitor memory usage, including whether there are memory leaks (available memory continues to decrease over time) and insufficient memory caused by excessive model loading; ② Check the rationality of model design a Check the number of polygons, including the redundant faces of a single device model and whether the total number of polygons in the scene exceeds the rendering threshold; b. Check whether the texture resolution exceeds the actual display requirements, and the number of transparent and reflective materials (the rendering calculation of such materials is 3-5 times that of ordinary materials); c. Check whether there are redundant transformation nodes in a single model; Step 3: Model optimization implementation ① Model face optimization: While maintaining the original geometric features and visual fidelity, the number of triangles is reduced through mesh simplification and edge folding algorithms; ②Model material optimization: reduce the number of material balls and lower the texture resolution through material merging and texture compression technology; use texture sets to integrate multiple textures and adjust the model texture UV; ③ Model-level node optimization: merge redundant transformation nodes; merge sub-meshes of the same material, and ensure that UV and normal data remain baked correctly after merging.

2. The method for optimizing the performance of a power digital twin model according to claim 1, characterized in that: The mesh simplification algorithm described in step 3 must reduce the number of triangles while keeping the error rate of key geometric features of the model ≤5%.

3. The method for optimizing the performance of a power digital twin model according to claim 1, characterized in that: The texture compression technology described in step 3 uses ETC, ASTC or BC series compression formats to maintain a visual quality loss rate of ≤10% when reducing the map resolution from 4K to 1K.

4. The method for optimizing the performance of a power digital twin model according to claim 1, wherein: The hierarchical node optimization described in step 3 also includes structural reorganization of multiple components of a single device, reducing the number of objects by merging sub-meshes of the same material, and maintaining the correctness of the merged UV and normal data through baking.

5. The method for optimizing the performance of a power digital twin model according to claim 1, wherein: The GPU performance analysis tool mentioned in step 2 is GPU-Z, which is used to obtain shader load and texture sampling time ratio data.

6. The method for optimizing the performance of a power digital twin model according to claim 1, characterized in that: When the texture set described in step 3 integrates multiple maps, the UV remapping technology is used to adjust the UV coordinates of the model map to achieve an orderly arrangement of multiple maps in a single texture set.

7. The method for optimizing the performance of a power digital twin model according to claim 1, characterized in that: The memory leak detection described in step 2 uses a memory monitoring tool to record the available memory change curve in real time. When the available memory decreases by ≥20% within 30 consecutive minutes, a memory leak is determined to exist.

8. The method for optimizing the performance of a power digital twin model according to claim 1, characterized in that: During the material merging process described in step 3, the parameters of the material balls with the same basic attributes are unified, retaining the key visual feature parameters while reducing the number of material balls.