A method and system for fatigue damage analysis of industrial speed reducers based on finite element analysis.

Abstract

This invention provides a method and system for fatigue damage analysis of industrial speed reducers based on finite element analysis (FEM), belonging to the field of FEM technology. The method includes: for each expert, generating their current proposed analysis strategy based on collaborative interaction behavior data parsing; calculating the weight coefficient of the analysis strategy by combining pre-set expert characteristic data and the correlation between the analysis strategy and the collaborative task objective; aggregating the analysis strategies and their weight coefficients of all experts to construct a multi-dimensional decision matrix; performing strategy fusion processing based on the multi-dimensional decision matrix to generate a target simulation scheme; controlling the finite element analysis model to execute the fatigue damage analysis of the target simulation scheme, and dynamically adjusting the execution pace during execution to adapt to the real-time pace of collaborative discussion and decision-making among experts. By quantifying expert opinions and optimizing the simulation scheme through strategy fusion, the real-time nature of expert collaboration and the effectiveness of decision-making are enhanced, making the analysis results closer to actual needs.

Description

Technical Field

[0001] This invention relates to the field of finite element analysis technology, and in particular to a method and system for fatigue damage analysis of industrial speed reducers based on finite element analysis. Background Technology

[0002] Currently, with the rapid development of industrial technology, the fatigue damage problem of industrial speed reducers, as core components of mechanical equipment, is receiving increasing attention. Traditional fatigue damage analysis methods often rely on the experience and judgment of a single expert, lacking real-time support and multi-party collaboration, resulting in insufficient scientific rigor and applicability of the analysis results. In recent years, finite element analysis technology has been widely used in speed reducer fatigue damage analysis; however, how to effectively integrate the collaborative wisdom of multiple experts with finite element analysis technology to construct a dynamic and efficient analysis process remains a pressing technical challenge. Summary of the Invention

[0003] One of the objectives of this invention is to provide a fatigue damage analysis method and system for industrial speed reducers based on finite element analysis, in order to solve the problems pointed out in the background art.

[0004] In a first aspect, the fatigue damage analysis method for industrial speed reducers based on finite element analysis provided in the embodiments of the present invention includes:

[0005] 101. On the expert collaboration platform, collect in real time the collaborative interaction data generated by each expert on the finite element analysis model of the industrial reducer;

[0006] 102. For each expert, the analysis strategy currently advocated by the expert is generated based on the analysis of collaborative interaction behavior data. The weight coefficient of the analysis strategy is calculated by combining the pre-set expert feature data and the correlation between the analysis strategy and the collaborative task objective.

[0007] 103. Aggregate the analytical strategies and weighting coefficients of all experts to construct a multidimensional decision matrix;

[0008] 104. Based on the multi-dimensional decision matrix, perform strategy fusion processing to generate the target simulation scheme for the current round of finite element simulation;

[0009] 105. Control the finite element analysis model to perform fatigue damage analysis of the target simulation scheme, and dynamically adjust the execution pace during the execution process to adapt to the real-time pace of collaborative discussion and decision-making among experts;

[0010] 106. After the current round of finite element simulation is completed, return to step 101 to execute the next round of operation until the analysis termination condition is met.

[0011] Optionally, in step 102, generating the analysis strategy currently advocated by the expert based on the analysis of collaborative interaction behavior data includes: extracting features from the target data related to the expert in the collaborative interaction behavior data to obtain multiple first feature values;

[0012] Based on each first feature value, construct the first feature description vector;

[0013] Based on a pre-built analysis strategy library, the analysis strategy currently advocated by the expert is matched according to the first feature description vector.

[0014] Optionally, in step 102, the weighting coefficient of the analysis strategy is calculated by combining preset expert feature data and the correlation between the analysis strategy and the collaborative task objective, including:

[0015] Based on a pre-defined quantification system, the authority and relevance of the analytical strategy represented by expert feature data are quantified respectively.

[0016] Based on preset weights, the authority and relevance are weighted and calculated to obtain the weight coefficients of this analysis strategy.

[0017] Optionally, step 104, performing strategy fusion processing based on a multi-dimensional decision matrix to generate the target simulation scheme for the current round of finite element simulation, includes:

[0018] Based on the local matrix truncation constraint, multiple local matrices are extracted from the multidimensional decision matrix;

[0019] For each local matrix, feature extraction is performed on the local matrix to obtain multiple second feature values. Based on each second feature value, a second feature description vector is constructed. Based on the pre-set strategy collaborative execution value library, the value degree of collaborative execution of each analysis strategy in the local matrix is ​​matched according to the second feature description vector.

[0020] By aggregating the analysis strategies in the local matrix with the highest aggregation value, the target simulation scheme for the current round of finite element simulation is obtained;

[0021] The local matrix truncation constraints include:

[0022] The proportion of analysis strategies whose weight coefficients exceed a preset coefficient threshold in the local matrix exceeds a preset proportion threshold;

[0023] The analysis objectives differ between pairwise analysis strategies within the same local matrix; the overall difference between the analysis strategies in each pairwise local matrix exceeds the preset difference threshold.

[0024] Optionally, in step 105, the execution pace is dynamically adjusted during the execution process to adapt to the real-time pace of collaborative discussion and decision-making among experts, including:

[0025] Obtain the remaining fatigue damage analysis tasks for the finite element analysis model;

[0026] Predict future trends in the pace of collaborative discussions and decision-making among experts;

[0027] Based on the remaining fatigue damage analysis tasks, future pace changes, and real-time pace, planning experts will collaboratively discuss and decide on the best next pace.

[0028] The remaining fatigue damage analysis tasks are split into task units and serialized to obtain a sequence of sub-task units.

[0029] Select the first N subtask units from the subtask unit sequence; where N is a positive integer, and the value of N is such that when the finite element analysis model executes the selected subtask units sequentially, the expected actual rhythm of guiding the experts to collaborate and make decisions is closest to the optimal next rhythm.

[0030] Dynamically adjust the execution rhythm, including controlling the finite element analysis model to execute the first N selected sub-task units sequentially in the next unit rhythm time.

[0031] Optionally, the prediction of future trends in the pace of collaborative discussion and decision-making among experts includes:

[0032] Acquire the historical rhythm of collaborative discussions and decision-making among various experts;

[0033] Based on real-time and historical rhythms, predict future trends in the rhythm of collaborative discussions and decision-making among experts.

[0034] Optionally, the step of planning the optimal next pace based on the remaining fatigue damage analysis tasks, future pace trends, and real-time pace, through expert collaborative discussion and decision-making, includes:

[0035] Based on the remaining fatigue damage analysis tasks, future rhythm change trends, and real-time rhythm, a third feature description vector is constructed.

[0036] Based on the optimal rhythm library, the best next rhythm for expert collaborative discussion and decision-making is matched according to the third feature description vector.

[0037] Optionally, the step of obtaining the real-time rhythm of collaborative discussion and decision-making among experts includes: determining the real-time rhythm of collaborative discussion and decision-making among experts based on updated real-time collected collaborative interaction behavior data.

[0038] Optionally, the analysis termination conditions include:

[0039] Historically, the simulation results from multiple rounds of finite element simulations have aligned with the objectives of the collaborative mission.

[0040] Secondly, the fatigue damage analysis system for industrial speed reducers based on finite element analysis provided in this embodiment of the invention includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the steps of the fatigue damage analysis method for industrial speed reducers based on finite element analysis described above.

[0041] The present invention has achieved the following beneficial effects:

[0042] By collecting real-time collaborative interaction data from experts through an expert collaboration platform, analyzing and generating expert analysis strategies, and combining expert characteristic data with the correlation between strategies and task objectives to calculate weight coefficients, a multi-dimensional decision matrix is ​​constructed for strategy fusion to generate a target simulation scheme. Simultaneously, the execution pace is dynamically adjusted during execution to adapt to the real-time discussion and decision-making rhythm of experts. Finally, through multiple iterations, high-precision and high-efficiency fatigue damage analysis is achieved. This significantly improves the scientific rigor of the analysis, optimizes the simulation scheme by quantifying expert opinions and strategy fusion, and enhances the real-time nature of expert collaboration and the effectiveness of decision-making, making the analysis results more closely aligned with actual needs.

[0043] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0044] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0045] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0046] Figure 1 This is a flowchart of the fatigue damage analysis method for industrial speed reducers based on finite element analysis in an embodiment of the present invention;

[0047] Figure 2 This is another flowchart of the fatigue damage analysis method for industrial speed reducers based on finite element analysis in this embodiment of the invention;

[0048] Figure 3 This is another flowchart of the fatigue damage analysis method for industrial speed reducers based on finite element analysis in this embodiment of the invention;

[0049] Figure 4 This is another flowchart of the fatigue damage analysis method for industrial speed reducers based on finite element analysis in this embodiment of the invention. Detailed Implementation

[0050] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0051] The research and development approach of this application combines expert collaboration with finite element analysis (FEM) technology to construct a dynamic and iterative analysis process for fatigue damage problems in industrial speed reducers. The core of this approach lies in using real-time interactive data from experts to generate analysis strategies, forming optimized simulation schemes through weight quantification and strategy fusion, and dynamically adjusting the pace during execution. Ultimately, through multiple iterations, high-precision and high-efficiency fatigue damage analysis is achieved. This method not only enhances the scientific rigor of the analysis but also strengthens the real-time nature of expert collaboration and the effectiveness of decision-making.

[0052] Figure 1 A flowchart of a fatigue damage analysis method for industrial speed reducers based on finite element analysis is provided for embodiments of this application, such as... Figure 1 As shown, the method includes:

[0053] 101. On the expert collaboration platform, collect in real time the collaborative interaction data generated by each expert on the finite element analysis model of the industrial reducer.

[0054] In this step, the collaborative interaction data generated by various experts on the finite element analysis model of the industrial reducer is collected in real time on the expert collaboration platform. This data reflects the experts' operational, decision-making, and communication behaviors during the analysis process, forming the basis for subsequent analysis of expert strategies. The real-time nature of the data ensures that the analysis keeps pace with the latest expert ideas, while the collaboration reflects the collective wisdom of multiple experts.

[0055] Here's an implementation example: In a fatigue analysis project for an industrial speed reducer, three experts worked on the same finite element model through a collaborative platform. Expert A adjusted the mesh density of the gear region, increasing it from 50 elements to 100 elements, and ran a stress analysis, recording the maximum stress value. Expert B modified the material properties of the bearing section, adjusting the elastic modulus from 200 GPa to 210 GPa, and submitted the thermal stress simulation results. Expert C, during the discussion, suggested paying attention to fatigue crack propagation at the gear meshing point and marked the relevant areas. This behavioral data (including parameter adjustments, simulation runs, and discussions) was collected in real time and stored in the expert collaboration platform's database as collaborative interaction behavioral data, with each data point bearing a timestamp and expert identifier.

[0056] 102. For each expert, generate the analysis strategy currently advocated by the expert based on the collaborative interaction behavior data analysis, and calculate the weight coefficient of the analysis strategy by combining the preset expert feature data and the correlation between the analysis strategy and the collaborative task objective.

[0057] In this step, for each expert, their currently advocated analytical strategy is analyzed based on collaborative interaction behavior data. This strategy's weight coefficient is then calculated by combining expert characteristic data (such as experience and domain expertise), the relevance of the analytical strategy to the task objective, and the weight coefficient reflects the importance of the analytical strategy in the overall decision-making process and serves as a crucial basis for subsequent fusion.

[0058] like Figure 2 As shown, step 102 specifically includes the following sub-steps:

[0059] 201. Extract features from the target data related to the expert in the collaborative interaction behavior data to obtain multiple first feature values.

[0060] In this step, several primary feature values ​​are extracted from the expert-related target data in the collaborative interaction behavior data. These primary feature values ​​include quantitative indicators such as operation frequency, parameter adjustment range, and simulation type.

[0061] Continuing with the above implementation example: taking expert A as an example, analyzing the collaborative interaction behavior data, several primary characteristic values ​​were determined, namely, the grid density adjustment frequency is 2 times / hour, the simulation type is stress analysis, and the parameter adjustment magnitude is an increase of 100% in the number of grid cells. These primary characteristic values ​​quantify the behavior pattern of expert A.

[0062] 202. Construct the first feature description vector based on each first feature value.

[0063] In this step, a first feature description vector is constructed based on the first feature value, serving as a structured representation of expert behavior.

[0064] Continuing with the above implementation example: Expert A's feature vector is [2, stress analysis, 100%], which corresponds to frequency, simulation type, and adjustment range, respectively.

[0065] 203. Based on the pre-set analysis strategy library, match the analysis strategy currently advocated by the expert according to the first feature description vector.

[0066] Based on a pre-built analysis strategy library, the analysis strategy currently advocated by experts is matched using the first feature description vector. The strategy library contains analysis strategies corresponding to different first feature description vectors, such as high-density mesh stress analysis and material property optimization analysis.

[0067] Continuing with the above implementation example: Expert A's feature vector [2, stress analysis, 100%] matches the high-density mesh stress analysis in the analysis strategy library, indicating that the analysis strategy currently advocated by this expert is to focus on the refined analysis of stress concentration areas.

[0068] 204. Based on the pre-set quantification system, quantify the authority and relevance of the analysis strategy represented by the expert feature data.

[0069] In this step, based on a pre-defined quantification system, the authority and relevance of the analytical strategies represented by expert feature data (such as historical fatigue analysis experience of industrial reducers) are quantified separately. When quantifying authority, the proportion of relevant experience in the expert's historical fatigue analysis experience of industrial reducers can be used as the level of authority. When quantifying relevance, a mapping relationship between different relevance levels and the relevance they represent is pre-defined, and the quantification is performed by directly querying this mapping relationship based on the relevance.

[0070] Continuing with the above implementation example: Expert A's analysis strategy focuses on detailed analysis of stress concentration areas. In their historical experience analyzing fatigue in industrial reducers, this strategy accounts for 90% of their relevant experience, thus their authority level is 0.9. The collaborative task objective is gear fatigue damage analysis, and Expert A's analysis strategy is highly correlated with it, with a quantitative correlation level of 0.95.

[0071] 205. Based on the preset weights, the authority and relevance are weighted and calculated to obtain the weight coefficients of the analysis strategy.

[0072] In this step, based on preset weights (such as authority accounting for 60% and relevance accounting for 40%), the authority and relevance are weighted and calculated to obtain the weight coefficient of the analysis strategy. The weight coefficient represents the importance of the analysis strategy.

[0073] Continuing with the above implementation example: The weight coefficient of expert A's analysis strategy = 0.9 × 0.6 + 0.95 × 0.4 = 0.54 + 0.38 = 0.92, indicating that the analysis strategy is of great importance.

[0074] 103. Aggregate the analytical strategies and weighting coefficients of all experts to construct a multidimensional decision matrix.

[0075] In this step, the analytical strategies and their weighting coefficients of all experts are aggregated to construct a multidimensional decision matrix. This matrix integrates scattered expert opinions into a structured framework, providing a data foundation for strategy fusion.

[0076] Here is an implementation example: Expert A's analysis strategy is high-density mesh stress analysis with a weighting factor of 0.92. Expert B's analysis strategy is material property optimization analysis with a weighting factor of 0.85. Expert C's analysis strategy is crack propagation analysis with a weighting factor of 0.78. The resulting multidimensional decision matrix is ​​shown in Table 1 below.

[0077] Table 1: Example of a multidimensional decision matrix

[0078]

[0079] 104. Based on the multidimensional decision matrix, perform strategy fusion processing to generate the target simulation scheme for the current round of finite element simulation.

[0080] In this step, based on the multidimensional decision matrix, the analytical strategies proposed by multiple experts are fused using methods such as local matrix truncation and collaborative value assessment to generate the target simulation scheme for the current round of finite element simulation. The multidimensional decision matrix is ​​a structured data framework that aggregates the analytical strategies of all experts and their corresponding weight coefficients; each row represents an expert's strategy and its importance assessment result. By performing strategy fusion processing on the matrix, the system can select the optimal strategy combination from numerous expert opinions, ensuring that the simulation scheme is both comprehensive and efficient. This process consists of three specific sub-steps: local matrix truncation, feature extraction and value assessment, and target scheme generation.

[0081] like Figure 3 As shown, step 104 specifically includes the following sub-steps:

[0082] 401. Based on local matrix truncation constraints, extract multiple local matrices from a multidimensional decision matrix. The local matrix truncation constraints include:

[0083] Constraint 1: The proportion of analysis strategies whose weight coefficients in the local matrix exceed the preset coefficient threshold exceeds the preset proportion threshold.

[0084] This constraint ensures that the strategies in the local matrix have sufficient importance. A preset coefficient threshold (e.g., 0.8) is used to filter high-weight strategies, while a preset proportion threshold (e.g., 0.5) requires that most strategies in the local matrix meet this importance criterion.

[0085] Constraint 2: The analysis objectives differ between pairwise analysis strategies within the same local matrix.

[0086] This constraint ensures the diversity of strategies within the local matrix. The analysis objective refers to the goal of the strategy (such as stress analysis or material optimization), requiring that each strategy has a distinct objective to avoid redundancy.

[0087] Constraint 3: The overall difference between the analysis strategies in each pairwise local matrix exceeds the preset difference threshold.

[0088] This constraint ensures that the strategy combinations between different local matrices are sufficiently distinctive. The overall dissimilarity can be quantified by the difference in strategy type or weight coefficients, with a preset dissimilarity threshold used to measure this degree of distinctiveness.

[0089] Through these three constraints, the extracted local matrix not only contains strategies of high importance, but also possesses internal diversity and external differences, providing high-quality candidate combinations for subsequent fusion.

[0090] Continuing with the above implementation example:

[0091] The multidimensional decision matrix contains the following data:

[0092] Expert D: Load spectrum optimization analysis, weighting coefficient 0.88;

[0093] Expert E: Thermal stress coupling analysis, weighting coefficient 0.90.

[0094] The preset coefficient threshold is set to 0.8, the preset proportion threshold to 0.5, and the preset difference threshold to 0.6. The process of extracting a local matrix is ​​as follows:

[0095] Local matrix 1:

[0096] Includes expert A, expert B, expert C, and their respective analysis strategies and weighting coefficients;

[0097] The proportion of analysis strategies with a weight greater than 0.8 (the analysis strategies of expert A and expert B) is 2 / 3 ≈ 0.67 > 0.5, which satisfies constraint one;

[0098] Analysis objectives: stress analysis, material optimization, and crack propagation are all different and satisfy constraint two.

[0099] Local matrix 2:

[0100] Includes expert D, expert E, and their respective analytical strategies and weighting coefficients;

[0101] The proportion of strategies with a weight exceeding 0.8 (both the analysis strategies of expert D and expert E exceed this weight) is 2 / 2 = 1 > 0.5, satisfying constraint one;

[0102] Analysis objectives: Load optimization and thermal stress analysis, which are different from each other and satisfy constraint two;

[0103] In addition, the strategies of local matrix 1 and local matrix 2 are completely different (non-overlapping type), with a difference of 1>0.6, satisfying constraint 3.

[0104] 402. For each local matrix, feature extraction is performed on the local matrix to obtain multiple second feature values. Based on each second feature value, a second feature description vector is constructed. Based on the pre-set strategy collaborative execution value library, the value degree of collaborative execution of each analysis strategy in the local matrix is ​​matched according to the second feature description vector.

[0105] In this step, features are extracted from each local matrix to obtain multiple second eigenvalues, which are used to quantify the characteristics of the strategy combination. Based on these eigenvalues, a second feature description vector is constructed, and the value of strategy collaboration in the local matrix is ​​evaluated through a pre-defined strategy collaboration execution value library. The second eigenvalues ​​include: the type of analysis strategy (e.g., stress analysis, crack propagation analysis), the average weight coefficient (reflecting overall importance), and the correlation between pairs of analysis strategies (e.g., whether there is execution dependency, such as stress analysis providing data for crack propagation). The second feature description vector is a structured representation of these eigenvalues, used to characterize the strategy combination characteristics of the local matrix. The pre-defined strategy collaboration execution value library stores the value levels corresponding to different eigenvectors based on historical experimental data. The value level measures the positive impact of strategy collaboration on fatigue analysis of industrial reducers, such as improving the comprehensiveness of analysis or the efficiency of expert collaboration.

[0106] Continuing with the above implementation example: process local matrix 1 and local matrix 2:

[0107] Local matrix 1 (Experts A, B, C):

[0108] Second eigenvalue:

[0109] Analysis strategy types: high-density mesh stress analysis, material property optimization analysis, crack propagation analysis;

[0110] Average weighting coefficient: (0.92 + 0.85 + 0.78) / 3 ≈ 0.85;

[0111] Correlation: High-density mesh stress analysis provides stress data for crack propagation analysis, indicating a dependency relationship;

[0112] Second feature description vector: [Type diversity = 3, average weight = 0.85, dependency relationship = yes]

[0113] Value assessment: Based on the strategy collaborative execution value library, the value of the second feature descriptor vector feature vector is 0.9 (high value, due to the enhanced comprehensiveness of the analysis due to diversity and dependencies).

[0114] Local matrix 2 (experts D, E):

[0115] Second eigenvalue:

[0116] Analysis strategy types: Load spectrum optimization analysis, thermal stress coupling analysis;

[0117] Average weighting coefficient: (0.88 + 0.90) / 2 = 0.89;

[0118] Relationship: No direct dependency

[0119] Second feature description vector: [Type diversity = 2, average weight = 0.89, dependency relationship = no]

[0120] Value assessment: Based on the strategy collaborative execution value library, the value of the second feature description vector is 0.7 (medium value, low synergistic effect due to lack of dependency).

[0121] 403. By aggregating the analysis strategies in the local matrix with the highest aggregation value, the target simulation scheme for the current round of finite element simulation is obtained.

[0122] In this step, the most valuable local matrix is ​​selected, and the analytical strategies within it are aggregated to form the target simulation scheme for the current round of finite element simulation. This scheme represents the optimal combination of expert strategies, maximizing the value of collaborative execution.

[0123] Continuing with the above implementation example: Select local matrix 1 and aggregate its strategies: high-density mesh stress analysis, material property optimization analysis, and crack propagation analysis, as the target simulation scheme for the current round.

[0124] 105. Control the finite element analysis model to perform fatigue damage analysis of the target simulation scheme, and dynamically adjust the execution pace during the execution process to adapt to the real-time pace of collaborative discussion and decision-making among experts. The steps for obtaining the real-time pace of collaborative discussion and decision-making among experts include: determining the real-time pace of collaborative discussion and decision-making among experts based on updated real-time collected collaborative interaction behavior data.

[0125] In this step, the finite element analysis model is controlled to perform fatigue damage analysis of the industrial reducer according to the target simulation scheme generated in step 104. Simultaneously, the execution pace of the simulation is dynamically adjusted using real-time collected collaborative interaction data to keep it synchronized with the pace of expert discussion and decision-making. This dynamic adjustment ensures that the simulation results can promptly reflect the latest expert opinions, improving the real-time nature of the analysis and collaborative efficiency.

[0126] Here's an implementation example: The target simulation includes high-density mesh stress analysis, material property optimization analysis, and crack propagation analysis. During execution: The system monitors the frequency of expert discussions. For example, if expert C frequently mentions crack propagation (5 times / minute), the system speeds up the crack propagation analysis (e.g., increasing from 10 iterations / second to 15). If expert discussions slow down (e.g., the frequency drops to 1 time / minute), the system slows down the stress analysis, waiting for more input. The real-time pace is determined through collaborative interaction data (e.g., discussion frequency, number of parameter adjustments), ensuring seamless integration between simulation and expert collaboration.

[0127] 106. After the current round of finite element simulation is completed, return to step 101 to execute the next round of operation until the analysis termination condition is met. The analysis termination condition includes: the simulation results of multiple rounds of finite element simulation in the past meet the collaborative task objective.

[0128] In this step, after completing the current round of simulation, the system returns to step 101, re-collects expert interaction data, generates a new simulation scheme, and enters the next iteration. This process continues until the simulation results meet the analysis termination condition, that is, the results of multiple rounds of simulation are consistent with the collaborative task objective (such as fatigue life prediction error being less than 5%).

[0129] Here's an implementation example: After the first round of simulation, the results show that the gear fatigue life prediction error is 8%, which fails to meet the target. The system returns to step 101, collects new expert suggestions on the crack propagation region (such as increasing mesh density), generates a new scheme, and executes the second round of simulation. After multiple iterations, the error drops to 4%, meeting the termination condition, and the analysis ends.

[0130] Overall, the above method collects real-time collaborative interaction data from experts through an expert collaboration platform, analyzes and generates expert analysis strategies, and calculates weight coefficients by combining expert characteristic data and the correlation between strategies and task objectives. A multi-dimensional decision matrix is ​​then constructed for strategy fusion to generate a target simulation scheme. Simultaneously, the execution pace is dynamically adjusted during execution to adapt to the real-time discussion and decision-making rhythm of experts. Finally, through multiple iterations, high-precision and high-efficiency fatigue damage analysis is achieved. This significantly improves the scientific rigor of the analysis, optimizes the simulation scheme by quantifying expert opinions and strategy fusion, and enhances the real-time nature of expert collaboration and the effectiveness of decision-making, making the analysis results more closely aligned with actual needs.

[0131] Specifically, step 102 involves real-time collection of expert interaction data on the collaborative platform, parsing and generating analysis strategies for each expert, and calculating the weight coefficient of each strategy by combining expert characteristic data (such as experience and expertise) and the relevance of the strategy to the task objective. This transforms the experts' subjective opinions into quantitative, structured analysis strategies and weights, achieving a scientific expression of expert knowledge. This not only provides an objective and reliable data foundation for subsequent strategy fusion but also reflects the importance of each expert in decision-making through weight calculation, ensuring the authority and relevance of the analysis plan, thereby significantly improving the scientific rigor of fatigue damage analysis for industrial reducers.

[0132] Step 104, based on a multidimensional decision matrix, integrates the analytical strategies of multiple experts through local matrix extraction and collaborative value assessment to generate the target simulation scheme for the current round of finite element simulation. Its technical advantage lies in selecting highly important and complementary strategy combinations by setting constraints (such as weight thresholds, diversity, and variability), ensuring that the integrated simulation scheme is both comprehensive and efficient. This method effectively integrates the collective wisdom of experts, optimizes the design of the simulation scheme, and significantly improves the scientific rigor and accuracy of fatigue damage analysis of industrial reducers, while providing a high-quality execution basis for finite element simulation.

[0133] After completing the current round of finite element simulation, step 106 returns to step 101 to re-collect collaborative interaction data from experts and generate a new simulation scheme, entering the next iteration until the simulation results meet the preset analysis termination conditions (such as fatigue life prediction error being less than a certain threshold). Through multiple rounds of dynamic iteration, the analysis process is continuously optimized, allowing the simulation results to gradually approach the collaborative task objectives. This cyclical mechanism ensures that the analysis can adapt to real-time feedback and adjustments from experts, thereby significantly improving the accuracy and efficiency of fatigue damage analysis for industrial reducers and achieving continuous improvement in analysis optimization.

[0134] In some embodiments, in step 105, the execution pace is dynamically adjusted during execution to adapt to the real-time pace of collaborative discussion and decision-making among experts, such as... Figure 4 As shown, it includes:

[0135] 501. Obtain the remaining fatigue damage analysis tasks for the finite element analysis model.

[0136] The purpose of this step is to identify the unfinished tasks in the finite element analysis model during the current round, providing a task basis for subsequent pace adjustments. The remaining tasks are the unexecuted parts extracted from the target simulation scheme generated in step 104, reflecting the progress of the analysis. Examples include stress distribution calculations and crack propagation simulations.

[0137] Here is an implementation example:

[0138] The target simulation scheme generated in step 104 includes the following three main tasks:

[0139] High-density mesh stress analysis (for gear meshing regions);

[0140] Material property optimization analysis (for bearing materials);

[0141] Crack propagation analysis (for gear surfaces).

[0142] During execution, the high-density mesh stress analysis is 50% complete, while the remaining two tasks have not yet started. The system retrieves the following remaining tasks by querying the execution log of the finite element analysis model:

[0143] The remaining part of the high-density mesh stress analysis (calculating the stress distribution in the remaining mesh region);

[0144] Material property optimization analysis (complete task, not yet executed);

[0145] Crack propagation analysis (complete task, not yet executed).

[0146] These task information are stored as a task list, with each task accompanied by a computational estimate (such as the number of mesh cells and the number of iterations) and an estimated time (e.g., the remaining part of the stress analysis will take 30 minutes).

[0147] 502. Predict the future trend of changes in the pace of collaborative discussions and decision-making among experts.

[0148] This step analyzes the historical and current pace of expert discussions to predict future trends in the pace of expert collaboration. These trends reflect the potential evolution of the frequency, depth, and focus of expert discussions, serving as a key basis for dynamically adjusting the execution pace.

[0149] Step 502 specifically includes the following sub-steps:

[0150] Obtain the historical rhythm of collaborative discussions and decision-making among various experts.

[0151] Historical rhythm data are quantitative metrics extracted from interaction records on expert collaboration platforms, including discussion frequency (e.g., number of speeches per minute), decision-making speed (e.g., intervals for parameter adjustments), and duration of focus on a particular topic. This data reflects the behavioral patterns of experts in past analyses.

[0152] Here's an implementation example: In the past two hours of analysis, the system recorded the following: Expert A suggested mesh adjustments every 10 minutes (0.1 times / minute); Expert B adjusted material parameters every 15 minutes (0.067 times / minute); Expert C mentioned crack propagation every 5 minutes (0.2 times / minute). Historical data is stored as a time series, with timestamps and content for each interaction.

[0153] Based on real-time and historical rhythms, predict future trends in the rhythm of collaborative discussions and decision-making among experts.

[0154] By combining real-time and historical rhythms, statistical analysis or machine learning models (collecting a large amount of historical data on rhythm changes resulting from discussions and decisions made by experts using finite element analysis models of industrial reducers for collaborative fatigue damage analysis, and using this data to train machine learning; machine learning is a current technology and will not be elaborated upon) can be used to infer future rhythms. For example, if real-time rhythms show an increase in discussion frequency, combining historical trends can predict whether this increase will continue.

[0155] Continuing with the above implementation example: Current real-time data shows that Expert C's discussion frequency has increased to 0.3 times / minute (once every 3 minutes) in the past 10 minutes, focusing on crack propagation; Experts A and B's frequencies remain stable. Combining historical data (Expert C's highest frequency is 0.2 times / minute), the system uses a machine learning model to predict that Expert C's discussion frequency may rise to 0.35 times / minute and stabilize within the next 30 minutes, while Experts A and B will remain at a low frequency. This prediction indicates that crack propagation will become a hot topic of discussion in the short term.

[0156] 503. Based on the remaining fatigue damage analysis tasks, future rhythm change trends, and real-time rhythm, plan the optimal next rhythm through collaborative discussion and decision-making among experts.

[0157] This step, based on the complexity of the remaining tasks, the predicted pace trend, and the current discussion pace, plans a pace target best suited for the next stage of expert collaboration. The optimal next pace aims to balance computational efficiency with the needs of expert decision-making, ensuring that simulation results support discussions in a timely manner, while avoiding both excessive speed and slowness that could lead to wasted resources or a breakdown in collaboration.

[0158] Step 503 specifically includes the following sub-steps:

[0159] Based on the remaining fatigue damage analysis tasks, future rhythm change trends, and real-time rhythm, a third feature description vector is constructed.

[0160] The third feature description vector is a multi-dimensional data structure that integrates task characteristics (such as computational load and priority), prediction rhythm (such as discussion frequency trends), and real-time rhythm (such as current speaking frequency) to characterize the overall state of the current analysis scenario.

[0161] Continuing with the above implementation example: the remaining tasks include high-density mesh stress analysis (medium computational cost), material property optimization analysis (high computational cost), and crack propagation analysis (medium computational cost); the prediction frequency shows that expert C's discussion frequency has increased to 0.35 times / minute (crack propagation focus); the real-time frequency is 0.3 times / minute for expert C, and lower for experts A and B. The third feature description vector is constructed as: [task complexity = medium to high, prediction frequency = 0.35, real-time frequency = 0.3, focus = crack propagation].

[0162] Based on the optimal rhythm library, the best next rhythm for expert collaborative discussion and decision-making is matched according to the third feature description vector.

[0163] The optimal rhythm library is a pre-trained database that stores the mapping relationship between different feature vectors and recommended rhythms. Rhythms include discussion intervals (e.g., every 5 minutes) and focus allocation (e.g., 50% of the time discussing crack propagation). The optimal rhythm scheme is found through vector matching.

[0164] Continuing with the above implementation example: the third feature description vector is input into the optimal rhythm library, and the matching result is that the optimal next rhythm is once every 4 minutes (frequency 0.25 times / minute), of which 60% of the time focuses on crack propagation, 30% discusses stress analysis, and 10% discusses material optimization. This rhythm is slightly lower than the predicted frequency (0.35) to avoid expert fatigue, while prioritizing responses to the crack propagation focus.

[0165] 504. Perform task unit splitting and serialization on the remaining fatigue damage analysis tasks to obtain a sub-task unit sequence.

[0166] This step breaks down the remaining fatigue damage analysis task into smaller subtask units, arranging them in execution order to form a serialized list of subtasks. The granularity of the task decomposition needs to be fine enough to support flexible pacing adjustments while maintaining the coherence of the computational logic. Serialization ensures that subtasks are executed in an orderly manner according to priority and dependencies.

[0167] Continuing with the above implementation example: The remaining fatigue damage analysis task is broken down as follows:

[0168] High-density mesh stress analysis:

[0169] Subtask Unit 1: Refine the mesh of the gear meshing area (1000 elements);

[0170] Subtask Unit 2: Calculate the stress distribution in this region;

[0171] Subtask Unit 3: Generate stress visualization plots.

[0172] Material property optimization analysis:

[0173] Subtask Unit 4: Adjust the elastic modulus of the bearing material (210 GPa).

[0174] Subtask Unit 5: Run a thermal stress simulation;

[0175] Subtask Unit 6: Evaluate material fatigue life.

[0176] Crack propagation analysis:

[0177] Subtask Unit 7: Set initial crack parameters (length 0.1mm);

[0178] Subtask Unit 8: Simulate crack propagation for 10 iterations;

[0179] Subtask Unit 9: Output crack propagation path.

[0180] The serialization result is a sequence of subtask units: [subtask unit 1, subtask unit 2, subtask unit 3, subtask unit 4, subtask unit 5, subtask unit 6, subtask unit 7, subtask unit 8, subtask unit 9], in which each subtask unit is arranged in the logical order of stress analysis, material optimization, and crack propagation.

[0181] 505. Select the first N subtask units from the subtask unit sequence. Where N is a positive integer, and the value of N is such that when the finite element analysis model executes the selected subtask units sequentially, the expected actual rhythm of guiding the experts to collaborate and make decisions is closest to the optimal next rhythm.

[0182] Select the first N subtasks (N is a positive integer) from the subtask sequence, so that the pace at which the finite element analysis model executes these subtasks guides the actual pace of the expert discussion to be close to the optimal next pace planned in step 503. The selection of N is based on the matching degree between the time consumption of the subtasks and the focus of the discussion, ensuring that the simulation progress is synchronized with the needs of the experts.

[0183] Continuing with the above implementation example: the optimal next pace is to discuss every 4 minutes, focusing 60% on crack propagation. Subtask unit time estimates: Subtask Unit 1 (5 minutes), Subtask Unit 2 (10 minutes), Subtask Unit 7 (3 minutes), Subtask Unit 8 (8 minutes). To match the pace, N=3 is selected, and the subtask units are [Subtask Unit 1, Subtask Unit 2, Subtask Unit 7]:

[0184] Subtask units 1 and 2 provide stress data (15 minutes) and support 30% stress discussion;

[0185] Subtask Unit 7 initiates crack propagation (3 minutes), occupying 60% of the focus.

[0186] The total time was 18 minutes, with results output on average every 6 minutes, close to a 4-minute pace. After adjustments, the experts were guided to discuss the topic every 5 minutes.

[0187] 506. Dynamically adjust the execution rhythm, including: controlling the finite element analysis model to execute the selected first N sub-task units sequentially in the next unit rhythm time.

[0188] The finite element analysis model is controlled to execute the selected N sub-tasks sequentially in the next unit of time (e.g., one hour). By adjusting the execution speed and output interval, the discussion rhythm of the experts is dynamically influenced to align with the optimal next rhythm. This step is the execution phase of rhythm adjustment.

[0189] Continuing with the above implementation example: Select [Subtask Unit 1, Subtask Unit 2, Subtask Unit 7], and set the next unit's tempo time to 30 minutes. System Settings:

[0190] Subtask Unit 1: Complete in 5 minutes, output grid data;

[0191] Subtask Unit 2: Complete in 10 minutes, output stress diagram;

[0192] Subtask Unit 7: Completed in 3 minutes, outputting crack parameters.

[0193] The execution pace was adjusted to output results every 6 minutes (slightly slower than the 4-minute pace to accommodate expert reaction time), totaling 18 minutes to complete, with the remaining time for discussion. After receiving the results, the experts discussed them every 5 minutes, approaching the optimal pace.

[0194] Overall, the above method, by acquiring the remaining fatigue damage analysis tasks, predicting the future pace of expert collaborative discussion and decision-making, planning the optimal next pace, and performing unit-based and serialized processing of the tasks, selects the first N sub-task units to guide the actual discussion pace of the experts close to the optimal pace, and finally controls the finite element analysis model to execute the selected sub-tasks in the next unit of time. This ensures that the simulation process can respond promptly to the latest discussions and decisions of the experts, improving the real-time performance and collaborative efficiency of the analysis, while enhancing the synchronization between simulation results and expert opinions through refined pace adjustments.

[0195] Specifically, step 505 selects the first N sub-task units (N is a positive integer) from the sub-task unit sequence. By controlling the pace at which the finite element analysis model executes these sub-tasks, the pace of the experts' actual discussion and decision-making is guided to approach the pre-planned optimal next pace. Through fine-grained task decomposition and dynamic pace adjustment, the simulation process is ensured to remain synchronized with the real-time collaboration needs of the experts. This method not only enables the simulation results to respond promptly to changes in the experts' discussion focus and decisions, but also enhances the real-time nature and collaborative efficiency of the analysis process, thereby improving the overall effectiveness and synergy of fatigue damage analysis for industrial reducers.

[0196] This application provides a fatigue damage analysis system for industrial speed reducers based on finite element analysis, including a processor and a memory. The memory stores a computer program, which, when executed by the processor, implements the steps of the fatigue damage analysis method for industrial speed reducers based on finite element analysis described above.

[0197] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A fatigue damage analysis method for industrial speed reducers based on finite element analysis, characterized in that, include:

101. On the expert collaboration platform, collect in real time the collaborative interaction data generated by each expert on the finite element analysis model of the industrial reducer; 102. For each expert, the analysis strategy currently advocated by the expert is generated based on the analysis of collaborative interaction behavior data. The weight coefficient of the analysis strategy is calculated by combining the pre-set expert feature data and the correlation between the analysis strategy and the collaborative task objective.

103. Aggregate the analytical strategies and weighting coefficients of all experts to construct a multidimensional decision matrix; 104. Based on the multi-dimensional decision matrix, perform strategy fusion processing to generate the target simulation scheme for the current round of finite element simulation; 105. Control the finite element analysis model to perform fatigue damage analysis of the target simulation scheme, and dynamically adjust the execution pace during the execution process to adapt to the real-time pace of collaborative discussion and decision-making among experts; 106. After the current round of finite element simulation is completed, return to step 101 to execute the next round of operation until the analysis termination condition is met; The 104th step involves strategy fusion processing based on a multi-dimensional decision matrix to generate the target simulation scheme for the current round of finite element simulation, including: Based on the local matrix truncation constraint, multiple local matrices are extracted from the multidimensional decision matrix; For each local matrix, feature extraction is performed on the local matrix to obtain multiple second feature values. Based on each second feature value, a second feature description vector is constructed. Based on the pre-set strategy collaborative execution value library, the value degree of collaborative execution of each analysis strategy in the local matrix is ​​matched according to the second feature description vector. By aggregating the analysis strategies in the local matrix with the highest aggregation value, the target simulation scheme for the current round of finite element simulation is obtained; The local matrix truncation constraints include: The proportion of analysis strategies whose weight coefficients exceed a preset coefficient threshold in the local matrix exceeds a preset proportion threshold; The analysis objectives differ between pairwise analysis strategies within the same local matrix; the overall difference between the analysis strategies in each pairwise local matrix exceeds the preset difference threshold.

2. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 1, characterized in that, In step 102, generating the analysis strategy currently advocated by the expert based on the analysis of collaborative interaction behavior data includes: extracting features from the target data related to the expert in the collaborative interaction behavior data to obtain multiple first feature values; Based on each first feature value, construct the first feature description vector; Based on a pre-built analysis strategy library, the analysis strategy currently advocated by the expert is matched according to the first feature description vector.

3. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 1, characterized in that, In step 102, the weighting coefficient of the analysis strategy is calculated by combining the preset expert feature data and the correlation between the analysis strategy and the collaborative task objective, including: Based on a pre-defined quantification system, the authority and relevance of the analytical strategy represented by expert feature data are quantified respectively. Based on preset weights, the authority and relevance are weighted and calculated to obtain the weight coefficients of this analysis strategy.

4. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 1, characterized in that, In step 105, the execution pace is dynamically adjusted during the execution process to adapt to the real-time pace of collaborative discussion and decision-making among experts, including: Obtain the remaining fatigue damage analysis tasks for the finite element analysis model; Predict future trends in the pace of collaborative discussions and decision-making among experts; Based on the remaining fatigue damage analysis tasks, future pace changes, and real-time pace, planning experts will collaboratively discuss and decide on the best next pace. The remaining fatigue damage analysis tasks are split into task units and serialized to obtain a sequence of sub-task units. Select the first N subtask units from the subtask unit sequence; where N is a positive integer, and the value of N is such that when the finite element analysis model executes the selected subtask units sequentially, the expected actual rhythm of guiding the experts to collaborate and make decisions is closest to the optimal next rhythm. Dynamically adjust the execution rhythm, including controlling the finite element analysis model to execute the first N selected sub-task units sequentially in the next unit rhythm time.

5. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 4, characterized in that, The predicted future trends in the pace of collaborative discussion and decision-making among experts include: Acquire the historical rhythm of collaborative discussions and decision-making among various experts; Based on real-time and historical rhythms, predict future trends in the rhythm of collaborative discussions and decision-making among experts.

6. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 4, characterized in that, The optimal next pace, based on the remaining fatigue damage analysis tasks, future pace trends, and real-time pace, is planned through expert collaborative discussion and decision-making, including: Based on the remaining fatigue damage analysis tasks, future rhythm change trends, and real-time rhythm, a third feature description vector is constructed. Based on the optimal rhythm library, the best next rhythm for expert collaborative discussion and decision-making is matched according to the third feature description vector.

7. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 1, characterized in that, The steps for obtaining the real-time rhythm of collaborative discussion and decision-making among experts include: determining the real-time rhythm of collaborative discussion and decision-making among experts based on updated real-time collected collaborative interaction behavior data.

8. The fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in claim 1, characterized in that, The analysis termination conditions include: Historically, the simulation results from multiple rounds of finite element simulations have aligned with the objectives of the collaborative mission.

9. A fatigue damage analysis system for industrial speed reducers based on finite element analysis, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, which, when executed by the processor, implements the steps of the fatigue damage analysis method for industrial speed reducers based on finite element analysis as described in any one of claims 1 to 8.

Images