Method and system for constructing digital twin model of speed reducer for optimizing machining precision

Abstract

The invention provides a speed reducer digital twinning model construction method and system for processing precision optimization, and relates to the technical field of digital twinning, and the method comprises the steps: predicting a plurality of optimization thinking paths which may be generated when a user carries out the processing precision optimization through employing a digital twinning body of a speed reducer; building an active auxiliary module of each optimization thinking path; wherein when the operation behavior of the user is matched to any optimization thinking path, the corresponding active auxiliary module is activated in real time, and an optimization decision data stream is generated through a simulation engine of the digital twin; and embedding the active auxiliary module of each optimized thinking path into an interactive logic layer of a digital twinborn body, and constructing a speed reducer digital twinborn model. According to the invention, dynamic and all-around active assistance of the user processing precision optimization thinking path is realized. The processing precision optimization efficiency is obviously improved, and the humanization and intelligence level of the digital twin model of the speed reducer is greatly enhanced at the same time.

Description

Technical Field

[0001] The present invention relates to the field of digital twin technology, and in particular to a method and system for constructing a digital twin model of a reducer for optimizing machining accuracy. Background Art

[0002] The machining accuracy of the reducer directly determines its working performance, so it is very important to optimize its machining accuracy.

[0003] At present, digital twin technology is becoming increasingly popular due to its real-time data synchronization and precise simulation features. The use of this technology to build a digital twin model of the reducer and provide users with processing accuracy optimization has emerged.

[0004] However, traditional reducer digital twin models typically only offer basic functions such as data display and performance simulation, lacking dynamic, comprehensive, and proactive support for users' thinking on how to optimize machining accuracy. This results in low efficiency when users use the model to optimize machining accuracy, and the model's humanization and intelligence levels are insufficient.

[0005] Therefore, an effective solution is urgently needed to make up for this deficiency. Summary of the Invention

[0006] One of the purposes of the present invention is to provide a method for constructing a digital twin model of a reducer for optimizing machining accuracy, so as to solve the problems in the background technology.

[0007] The method for constructing a digital twin model of a reducer for optimizing machining accuracy provided by an embodiment of the present invention includes: Predict the various optimization paths that users may have when optimizing machining accuracy using the reducer’s digital twin; Build an active auxiliary module for each optimization thinking path. When the user's operation behavior matches any optimization thinking path, the corresponding active auxiliary module is activated in real time, and the optimization decision data flow is generated through the simulation engine of the digital twin. The active auxiliary module of each optimized thinking path is embedded in the interactive logic layer of the digital twin to build a digital twin model of the reducer.

[0008] Optionally, the optimization thinking path includes: a dynamic technical path for generating a reducer processing accuracy optimization solution based on a decision-making reasoning process.

[0009] Optionally, the prediction steps of the multiple optimization thinking paths include: Based on the historical records of users optimizing the machining accuracy of reducers, the first probability size of each standard path in the standard optimization thinking path library is predicted when users use digital twins to optimize the machining accuracy; All standard paths corresponding to the first possibility values exceeding the first possibility threshold are screened out as the final multiple optimization thinking paths.

[0010] Optionally, the steps of building the active auxiliary module for each optimized thinking path include: Traverse each optimization thinking path in turn; When traversing to any optimization thinking path, a judgment rule is generated to match the user's operation behavior with the traversed optimization thinking path, and a control rule is generated to control the simulation engine of the digital twin to generate the optimization decision data flow of the traversed optimization thinking path; Integrate judgment rules and control rules to obtain an active auxiliary module for traversing the optimized thinking path.

[0011] Optionally, the judgment rule includes: Analyze the thinking path reflected by the user's operation behavior; Eliminate the first partial path where the user may regret from the thinking path to obtain the path to be matched; Match the path to be matched with the optimized thinking path traversed; When, in the traversed optimized thinking path, there is a second partial path whose matching degree with the path to be matched exceeds the matching degree threshold, and the second partial path indicates that the second probability of the user generating other third partial paths in the future exceeds the second probability threshold, then it is determined that the user's operation behavior matches the traversed optimized thinking path.

[0012] Optionally, the step of obtaining the first partial path that the user may regret includes: Determine multiple path points to be verified from the thinking path; wherein the path points to be verified must satisfy the following conditions: the path feature sets of the preset path segments before and after the path points match the standard path feature set representing the user's possible irrational thinking; Verify the user's rational thinking at each path point to be verified; Based on all the to-be-verified path points whose rationality of thinking is lower than the rationality threshold, and according to their distribution in the thinking path, a template for delineating possible regret paths is matched; Based on the template for defining possible regret paths, the first partial path that the user may regret is defined in the thinking path.

[0013] Optionally, the step of verifying the user's rational thinking when at each path point to be verified includes: Traverse each path point to be verified in turn; When traversing to any path point to be verified, determine the fourth partial path before the traversed path point to be verified from the thinking path. Based on the fourth partial path, match the path rationality degree and the user's rational thinking ability degree when the traversed path point is to be verified, and perform weighted calculation on the path rationality degree and the rational thinking ability degree, and use the weighted calculation result as the thinking rationality degree of the user when the traversed path point is to be verified.

[0014] Optionally, the step of obtaining the second probability of the second partial path indicating that the user will generate other third partial paths in the future includes: A weighted calculation is performed on the beginning and end boundary values of the path position ratio interval of the second partial path in the thinking path, the interval length, and the correlation between other third partial paths and the second partial path. The weighted calculation result is used as the second partial path to indicate the second possibility of the user generating other third partial paths in the future.

[0015] Optionally, the step of generating control rules for the optimization decision data flow traversed to the optimization thinking path by the simulation engine controlling the digital twin includes: Based on the optimized thinking path traversed and the user's new operation behavior, the control rules are matched to generate templates; Based on the control rule generation template, the simulation engine that controls the digital twin generates control rules for the optimization decision data flow that traverses the optimization thinking path.

[0016] An embodiment of the present invention provides a system for constructing a digital twin model of a reducer for optimizing machining accuracy, comprising: The prediction module is used to predict the various optimization paths that users may generate when optimizing machining accuracy using the reducer's digital twin. A building module is used to build an active auxiliary module for each optimization thinking path. When the user's operation behavior matches any optimization thinking path, the corresponding active auxiliary module is activated in real time, and the optimization decision data flow is generated through the simulation engine of the digital twin; A building module is used to embed the active auxiliary module of each optimized thinking path into the interactive logic layer of the digital twin to build a digital twin model of the reducer.

[0017] The present invention has achieved the following beneficial effects: The system predicts the various optimization thinking paths that may be generated when users use the reducer digital twin to optimize processing accuracy, builds a dedicated active assistance module for each path, and pre-embeds it in the interactive logic layer of the digital twin, together forming the core intelligent assistance capability of the reducer digital twin model. When the user's operating behavior matches a preset optimization thinking path, the corresponding active assistance module will be activated in real time and drive the simulation engine of the digital twin to dynamically generate an optimization decision data stream that is highly adapted to the path, guiding the user to efficiently advance the optimization process along the matching path, and realizing dynamic and comprehensive active assistance for the user's processing accuracy optimization thinking path. It significantly improves the efficiency of processing accuracy optimization and greatly enhances the humanization and intelligence level of the reducer digital twin model.

[0018] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0019] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 Schematic diagram of a method for constructing a digital twin model of a reducer for optimizing machining accuracy in an embodiment of the present invention; Figure 2 Schematic diagram of an implementation method of a reducer digital twin model construction method for optimizing machining accuracy in an embodiment of the present invention; Figure 3 Schematic diagram of a reducer digital twin model construction system for optimizing machining accuracy in an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0022] Example 1: The embodiment of the present invention provides a method for constructing a digital twin model of a reducer for optimizing machining accuracy, such as Figure 1 Shown, including: S1. Predicting multiple optimization thinking paths that a user may generate when optimizing the machining accuracy of a reducer using a digital twin of the reducer. The optimization thinking paths include: generating a dynamic technical path for optimizing the machining accuracy of the reducer based on a decision-making reasoning process.

[0023] like Figure 2 As shown, sensors at the physical layer collect real-time physical data about the reducer (such as model, geometry, performance parameters, and real-time operating status). This data is transmitted to the digital twin layer via a data synchronization mechanism. Within the digital twin layer, the data layer receives and integrates the physical data. The simulation engine, combined with the model layer (including geometric and physical models), performs 3D dynamic simulation of the reducer, ultimately constructing a digital twin corresponding to its physical entity. When users optimize machining accuracy based on this digital twin, they typically follow this process: first, they set a clear optimization goal (such as gear system optimization, housing / shaft system optimization, or dynamic performance compensation). Next, they engage in decision-making and reasoning around this goal, exploring and exploring specific methods to achieve it. Throughout this process, they continuously evaluate different options, weigh parameter adjustments, and predict potential impacts, ultimately determining a feasible solution for optimizing reducer machining accuracy that meets this goal. The series of steps, thought nodes, and their logical connections throughout this decision-making and reasoning process constitute the optimization thinking path.

[0024] S2. Build an active auxiliary module for each optimization thinking path; when the user's operation behavior matches any optimization thinking path, the corresponding active auxiliary module is activated in real time, and the optimization decision data stream is generated through the simulation engine of the digital twin.

[0025] like Figure 2As shown in the figure, users input operational actions through the user interface layer, including viewing the digital twin's status, adjusting parameters, and selecting analysis functions. When the system identifies that the user's current operational actions match a specific optimization path, indicating the user's intention to optimize machining accuracy along that path, the active assistance module corresponding to that path is activated in real time. The activated active assistance module drives the simulation engine in the digital twin layer to perform specialized simulation calculations for that optimization path, generating an optimization decision data stream. This data stream is then fed back to the user interface layer, providing real-time decision support. The core function of the optimization decision data stream is to assist users in efficiently advancing accuracy optimization along the matching optimization path. It contains key information that guides subsequent operations, such as parameter optimization suggestions (e.g., for the gear system optimization path, recommended adjustment ranges for gear module, pressure angle, and displacement coefficient, optimal value predictions, and their impact curves on transmission error); process adjustment plans (e.g., for the housing / shaft optimization path, optimal bearing preload / clearance settings, assembly sequence improvements, or adjustment ranges for critical fit tolerances based on thermal deformation simulation results); and potential problem warnings.

[0026] S3. Embed the active auxiliary module of each optimized thinking path into the interactive logic layer of the digital twin to build a digital twin model of the reducer.

[0027] like Figure 2 As shown in the figure, the active auxiliary module of each optimized thinking path is embedded in the interactive logic layer of the digital twin, which continuously monitors the user's operation behavior and waits for activation. After activation, it executes the simulation drive and decision support functions.

[0028] In an embodiment of the present invention, the system predicts the various optimization thinking paths that may be generated when the user uses the reducer digital twin to optimize the processing accuracy, builds a dedicated active assistance module for each path, and pre-embeds it in the interactive logic layer of the digital twin, which together constitute the core intelligent assistance capabilities of the reducer digital twin model. When the user's operating behavior matches a preset optimization thinking path, the corresponding active assistance module will be activated in real time, and drive the simulation engine of the digital twin to dynamically generate an optimization decision data stream that is highly adapted to the path, guiding the user to efficiently advance the optimization process along the matching path, and realizing dynamic and comprehensive active assistance for the user's processing accuracy optimization thinking path. It significantly improves the efficiency of processing accuracy optimization and greatly enhances the humanization and intelligence level of the reducer digital twin model.

[0029] Example 2: In the embodiment of the present invention, in S1, the step of predicting multiple optimization thinking paths includes: S11. Based on the historical records of users optimizing the machining accuracy of reducers, predict the first probability size of each standard path in the standard optimization thinking path library generated by the user when using the digital twin to optimize the machining accuracy.

[0030] By collecting a large number of historical optimization thinking paths generated by different technicians when using the reducer digital twin to optimize processing accuracy, a representative standard optimization thinking path library is formed. The user's historical records of reducer processing accuracy optimization (such as historical optimization results, the most recent optimization target, the commonly used parameter adjustment range, the types of successful / failed optimization cases, etc.) contain their optimization strategy preferences and habits. Therefore, based on the user's historical records, the system can predict the first probability of the user generating each standard optimization thinking path. Specifically, when implementing this prediction process, a large number of historical records marked with the standard paths that the user would choose can be used to train a neural network model to obtain a selection prediction model. The current historical records and each standard path are input into the selection prediction model. The selection prediction model will output the probability value of the user selecting each standard path. This probability is the first probability size.

[0031] S12. All standard paths corresponding to the first likelihood values exceeding the first likelihood threshold are screened out as the final multiple optimization thinking paths.

[0032] The first possibility threshold is set in advance. When the first possibility exceeds the first possibility threshold, it means that the user is very likely to generate a corresponding standard path, which is selected as the final multiple optimization thinking paths.

[0033] An embodiment of the present invention introduces a standard optimization thinking path library. Based on the historical records of users optimizing the processing accuracy of reducers, it predicts the first probability size of each standard path in the standard optimization thinking path library generated by users, and takes all standard paths corresponding to the first probability size exceeding the first probability threshold as the final multiple optimization thinking paths, thereby improving the accuracy, comprehensiveness and efficiency of the prediction of multiple optimization thinking paths, improving the accuracy of using them to build active auxiliary modules, and accurately preparing to provide users with reducer optimization assistance.

[0034] Example 3: In the embodiment of the present invention, in S2, the steps of building the active auxiliary module of each optimized thinking path include: S21. Traverse each optimization thinking path in turn.

[0035] S22. When traversing to any optimization thinking path, generate a judgment rule for matching the user's operation behavior with the traversed optimization thinking path, and generate a control rule for controlling the simulation engine of the digital twin to generate the optimization decision data flow for the traversed optimization thinking path. In S22, the judgment rule includes: S221. Analyze the thought process reflected by the user's operational behavior. The user's operational behavior is the outward manifestation of their decision-making and reasoning process for optimizing the reducer's machining accuracy. It can reflect their reasoning steps, key thought process nodes, and the logical connections between them. For example, when a user adjusts the gear module and pressure angle, this behavior reflects their decision intent: to improve the gear system's accuracy by optimizing the basic gear parameters (module and pressure angle). Therefore, the system can infer the user's current optimization thought process by directly analyzing the user's operational behavior sequence. This analysis process requires pre-definition of: 1. Operational behavior-reasoning element mapping rules: These rules clearly define the specific decision-making reasoning steps or thought process nodes corresponding to different operational behaviors; and 2. A logical association library between reasoning elements: These libraries define the possible logical relationships between different reasoning steps or thought process nodes (e.g., sequential execution, conditional branching, loop iteration, etc.). During the analysis process, the mapping rules are used to identify the decision-making reasoning steps or thought process nodes corresponding to each user operational behavior. These identified reasoning steps or nodes are sequentially connected according to the chronological order of the operational behaviors to form an initial behavioral deduction sequence. Based on a predefined logical association library, the system annotates the logical associations between the identified reasoning steps or nodes in the initial sequence, ultimately constructing a thinking path that reflects the user's current decision-making reasoning process.

[0036] S222. Eliminate the first partial path that the user may regret from the thinking path to obtain the path to be matched. However, given the high complexity of the reducer processing accuracy optimization problem, the user's optimization thinking process is inevitably complex, dynamic, and highly uncertain. In addition, when exploring solutions, users often perform a series of intermittent trial operations. Therefore, the thinking path will inevitably contain partial path segments that the user may intend to abandon or modify, that is, the first partial path that the user may regret. In S222, the steps for obtaining the first partial path that the user may regret include: S2221. Determine multiple path points to be verified from the thinking path; wherein, the path point to be verified must meet the following requirements: the path feature set of each preset path segment extending before and after it matches the standard path feature set representing the possible irrational thinking of the user. The system pre-sets a standard path feature set, which characterizes the typical irrational decision-making mode that may occur when the user optimizes the processing accuracy. When analyzing the thinking path, for each path point, the system will extract a path feature set consisting of path segments with preset lengths extending forward and backward (such as the first M thought nodes and the last N thought nodes, where M and N are positive integers) with the point as the center. Subsequently, the path feature set is matched and analyzed with the standard path feature set. If the two are completely matched, it indicates that the user has a high tendency to make irrational decisions at this path point, and it is used as a path point to be verified. Specifically, when pre-setting the standard path feature set, a series of path features that can reflect the potential irrationality of decisions are defined based on a large number of historical optimization records (especially cases containing known irrational decisions or leading to optimization failures), such as: parameter mutation features (such as drastic and unrelated changes in the adjustment amplitude or direction of key parameters such as modulus, pressure angle, preload, etc. in a short period of time), target drift features (the optimization target switches frequently and without logical connection, such as suddenly jumping from reducing transmission error to minimizing temperature rise and then jumping back), etc., and a standard path feature set is constructed based on these path features.

[0037] S2222. Verify the user's rational thinking level when at each path point to be verified.

[0038] S2223. Based on all the to-be-verified path points whose rationality of thinking is lower than the rationality threshold, and according to their distribution in the thinking path, a template for delineating possible paths of regret is matched. The system presets a rationality threshold. When the rationality of thinking of the user at a specific to-be-verified path point obtained by analysis and calculation is lower than the threshold, it is determined that the user has a significant lack of rationality in thinking at that point. In the entire thinking path, the spatial distribution of all to-be-verified path points that are determined to be insufficiently rational comprehensively indicates the overall irrational tendency intensity and pattern of the user in the process of generating the path. In order to predict the local path segment (i.e., the first local path) that the user is most likely to abandon based on this distribution, a template library for delineating the path of regret is pre-set, in which each template is associated with a distribution situation and a specific strategy for delineating the first local path under this distribution situation is set. Specifically, during pre-setting, the distribution of irrational thinking path points is determined from a large number of historical thinking paths for optimizing reducer machining accuracy with irrational thinking. Based on this distribution, other affected path points are determined (e.g., path points that have a causal logical relationship with the irrational thinking path points). The local path formed by the irrational thinking path points and the affected path points is then determined. The delineation logic of this local path is then analyzed (i.e., how this local path was delineated in the historical thinking path). The delineation logic is then paired with the distribution to form a delineation template for a reversal path. After multiple templates are formed, they are stored in a database to obtain a delineation template library for a reversal path.

[0039] S2224. Based on the template for defining possible regret paths, define the first partial path in the thinking path where the user may regret.

[0040] S223. Match the path to be matched with the traversed optimization thinking path.

[0041] S224. When, in the optimized thinking path that has been traversed, there is a second partial path whose matching degree with the path to be matched exceeds the matching degree threshold, and the second partial path indicates that the user will generate other third partial paths in the future. The user's operation behavior is judged to match the optimized thinking path that has been traversed. The matching degree threshold is set in advance. When the matching degree of the second partial path and the path to be matched exceeds the matching degree threshold, it indicates that the matching degree between the two is high. The second possibility threshold is set in advance. When the second partial path indicates that the user will generate other third partial paths in the future. The second possibility threshold is set in advance. When the matching degree exceeds the matching degree threshold and the second possibility exceeds the second possibility threshold, it indicates that the second possibility is large. When the matching degree exceeds the matching degree threshold and the second possibility exceeds the second possibility threshold, it indicates that the user is very likely to make a processing accuracy optimization decision according to the corresponding optimized thinking path. The user's operation behavior is judged to match the optimized thinking path that has been traversed.

[0042] The step of generating control rules for the optimization decision data flow traversed by the simulation engine of the control digital twin includes: S225. Based on the traversed optimization thinking path and the user's new operation behavior, a control rule generation template is matched. The user's new operation behavior refers to the behavior of the digital twin of the reducer generated by the user after the user's operation behavior is judged to match the traversed optimization thinking path. Pre-set control rule generation templates that match different optimization thinking paths and the user's new operation behavior. The control rule generation template is used to generate control rules for the simulation engine of the control digital twin to generate the optimization decision data flow for the optimization thinking path traversed. Specifically, when pre-setting the control rule generation template, the user's current auxiliary needs are determined according to different optimization thinking paths and the user's new operation behavior, and then which optimization decision data flow can meet the auxiliary needs is determined. Finally, the rules for the control simulation engine to generate the optimization decision data flow are set as the control rule generation template that matches the optimization thinking path and the user's new operation behavior. For example: the auxiliary demand is that the system needs to provide a recommended adjustment range of the gear module, pressure angle, and displacement coefficient. The control rule generation template is to control the simulation engine to generate the optimization data flow of the reducer state under different recommended gear modules, pressure angles, and displacement coefficients.

[0043] S226. Based on the control rule generation template, generate the control rules for the simulation engine of the control digital twin to generate the optimization decision data flow of the optimization thinking path traversed.

[0044] S23, integrating the judgment rule and the control rule to obtain an active auxiliary module for the optimized thinking path traversed to. Finally, the judgment rule and the control rule are integrated into the same rule execution module to obtain an active auxiliary module for the optimized thinking path traversed to.

[0045] In this embodiment of the present invention, when generating judgment rules, the system proactively removes the first identified partial path (the path segment where the user might change their mind) from the user's optimization thinking path. Matching judgments based on these removed paths significantly improve the accuracy, comprehensiveness, and applicability of the match between user behavior and the target optimization thinking path. This directly optimizes the accuracy of the judgment rules' decision support within the proactive assistance module. Without requiring the user to actively input their thinking path, the system can seamlessly and accurately identify and focus on the user's most determined core optimization path, significantly enhancing the system's applicability in practical applications.

[0046] When identifying the first partial path, the system first quickly screens potentially irrational path points for verification based on pre-set criteria. It then performs centralized "rationality of thinking" verification only on these selected points, significantly reducing verification computing resource consumption and significantly improving verification efficiency. Based on the distribution of all path points that verify below the rationality threshold, the system intelligently matches a pre-set reversal path delineation template. Using this matching template, the first partial path is efficiently and accurately delineated within the thinking path. This significantly improves the accuracy and efficiency of first partial path identification and ensures the timely generation of judgment rules.

[0047] When generating control rules, the system intelligently matches predefined control rule generation templates based on the currently matched optimization thinking path and the user's most recent operational behavior. Using this template, control rules are dynamically generated, ensuring that after rule execution, the simulation engine generates an optimization decision data stream that is highly adapted to the current optimization path and the user's most recent intent. This significantly improves the accuracy and contextual awareness of real-time user assistance.

[0048] Overall, the system's ability to build efficient proactive assistance modules for each optimized thinking path has been significantly enhanced, significantly improving the overall effectiveness of the proactive assistance modules in providing users with dynamic, comprehensive, and highly adaptable proactive assistance.

[0049] Example 4: In the embodiment of the present invention, in S2222, the step of verifying the user's rational thinking level at each path point to be verified includes: Traverse each path point to be verified in turn; When traversing to any path point to be verified, determine the fourth partial path before the traversed path point to be verified from the thinking path. Based on the fourth partial path, match the path rationality degree and the user's rational thinking ability degree when the traversed path point is to be verified, and perform weighted calculation on the path rationality degree and the rational thinking ability degree, and use the weighted calculation result as the thinking rationality degree of the user when the traversed path point is to be verified.

[0050] Different thinking paths are pre-set with a matching path rationality level (different thinking paths are collected and their path rationality is comprehensively assessed by multiple experts). This matching path rationality level represents the overall rationality of the user's thinking process as reflected by the thinking path. The user's rational thinking ability after generating different thinking paths is pre-determined through experiments (multiple experts comprehensively assess the rational thinking ability of different users after generating their thinking paths). This rational thinking ability level represents the user's ability to continue rational thinking after generating a thinking path. To verify the rationality level of thinking, a fourth partial path is determined from the thinking path before the point to be verified. Based on this fourth partial path, the path rationality level is matched with the user's rational thinking ability level at the point to be verified. Based on pre-determined weights based on the degree to which the path rationality level and rational thinking ability level each influence the rationality level of thinking, a weighted calculation is performed on the path rationality level and the rational thinking ability level to obtain the user's rationality level of thinking at the point to be verified. It improves the accuracy, comprehensiveness and efficiency of verifying the rationality of users' thinking when they are at each path point to be verified, and improves the accuracy of their subsequent use in determining distribution to match the template for delineating possible paths of regret.

[0051] Example 5: In the embodiment of the present invention, in S224, the step of obtaining the second probability of the second partial path indicating that the user will generate other third partial paths in the future includes: A weighted calculation is performed on the beginning and end boundary values of the path position ratio interval of the second partial path in the thinking path, the interval length, and the correlation between other third partial paths and the second partial path. The weighted calculation result is used as the second partial path to indicate the second possibility of the user generating other third partial paths in the future.

[0052] The path position ratio interval means, for example: there are 20 path points on the thinking path, and the second local path is from the 7th path point to the 11th path point on the thinking path, then the path position ratio interval is [7 / 20, 11 / 20], and the first and last boundary values are 7 / 20 and 11 / 20 respectively.

[0053] The larger the head-tail boundary value is, the further back the second partial path is in the thinking path, and the more it represents the user entering the part of the thinking path that reflects the subsequent thinking process, indicating that the user is more likely to generate other third partial paths in the future. Therefore, the head-tail boundary value is positively correlated with the size of the second possibility.

[0054] The longer the interval length is, the greater the proportion of the second partial path in the thinking path is, which means that the more partial paths in the thinking path have been generated, the more likely the user will generate other third partial paths in the future. Therefore, the interval length is positively correlated with the size of the second possibility.

[0055] The degree of correlation between other third partial paths and the second partial path refers to the degree of logical correlation between the two paths (the correlation degrees corresponding to different logical correlation relationships can be pre-set. When determining the logical correlation degree, the corresponding correlation degree is directly determined based on the correlation relationship between the two paths). The greater the correlation degree, the more likely the user will generate other third partial paths in the future. Therefore, the correlation degree is positively correlated with the size of the second possibility.

[0056] Based on the pre-set weights for the initial and final boundary values, interval length, and correlation, each factor influences the second likelihood. A weighted calculation is performed on these four factors to determine the second likelihood. This improves the accuracy of the second likelihood and its applicability for matching user actions with the resulting optimized thinking paths.

[0057] Example 6: The embodiment of the present invention provides a reducer digital twin model construction system for optimizing machining accuracy, such as Figure 3 Shown, including: Prediction module 1 is used to predict the various optimization paths that users may generate when optimizing machining accuracy using the reducer's digital twin. Module 2 is used to build an active auxiliary module for each optimization thinking path. When the user's operation behavior matches any optimization thinking path, the corresponding active auxiliary module is activated in real time, and the optimization decision data flow is generated through the simulation engine of the digital twin. Construction module 3 is used to embed the active auxiliary module of each optimized thinking path into the interactive logic layer of the digital twin to build a digital twin model of the reducer.

[0058] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A method for constructing a digital twin model of a reducer for optimizing machining accuracy, characterized in that: include: Predict the various optimization paths that users may have when optimizing machining accuracy using the reducer’s digital twin; Build an active auxiliary module for each optimization thinking path. When the user's operation behavior matches any optimization thinking path, the corresponding active auxiliary module is activated in real time, and the optimization decision data flow is generated through the simulation engine of the digital twin. The active auxiliary module of each optimized thinking path is embedded in the interactive logic layer of the digital twin to build a digital twin model of the reducer.

2. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 1, wherein: The optimization thinking path includes: a dynamic technical path for generating a reducer processing accuracy optimization solution based on a decision-making reasoning process.

3. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 1, wherein: The prediction steps of the multiple optimization thinking paths include: Based on the historical records of users optimizing the machining accuracy of reducers, the first probability size of each standard path in the standard optimization thinking path library is predicted when users use digital twins to optimize the machining accuracy; All standard paths corresponding to the first possibility values exceeding the first possibility threshold are screened out as the final multiple optimization thinking paths.

4. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 1, wherein: The steps for building the active auxiliary module of each optimized thinking path include: Traverse each optimization thinking path in turn; When traversing to any optimization thinking path, a judgment rule is generated to match the user's operation behavior with the traversed optimization thinking path, and a control rule is generated to control the simulation engine of the digital twin to generate the optimization decision data flow of the traversed optimization thinking path; Integrate judgment rules and control rules to obtain an active auxiliary module for traversing the optimized thinking path.

5. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 4, wherein: The judgment rules include: Analyze the thinking path reflected by the user's operation behavior; Eliminate the first partial path where the user may regret from the thinking path to obtain the path to be matched; Match the path to be matched with the optimized thinking path traversed; When, in the traversed optimized thinking path, there is a second partial path whose matching degree with the path to be matched exceeds the matching degree threshold, and the second partial path indicates that the second probability of the user generating other third partial paths in the future exceeds the second probability threshold, then it is determined that the user's operation behavior matches the traversed optimized thinking path.

6. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 5, wherein: The step of obtaining the first partial path that the user may regret comprises: Determine multiple path points to be verified from the thinking path; wherein the path points to be verified must satisfy the following conditions: the path feature sets of the preset path segments before and after the path points match the standard path feature set representing the user's possible irrational thinking; Verify the user's rational thinking at each path point to be verified; Based on all the to-be-verified path points whose rationality of thinking is lower than the rationality threshold, and according to their distribution in the thinking path, a template for delineating possible regret paths is matched; Based on the template for defining possible regret paths, the first partial path that the user may regret is defined in the thinking path.

7. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 6, wherein: The steps for verifying the user's rational thinking when at each path point to be verified include: Traverse each path point to be verified in turn; When traversing to any path point to be verified, determine the fourth partial path before the traversed path point to be verified from the thinking path. Based on the fourth partial path, match the path rationality degree and the user's rational thinking ability degree when the traversed path point is to be verified, and perform weighted calculation on the path rationality degree and the rational thinking ability degree, and use the weighted calculation result as the thinking rationality degree of the user when the traversed path point is to be verified.

8. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 5, wherein: The step of obtaining the second probability of the second partial path indicating that the user will generate other third partial paths in the future includes: A weighted calculation is performed on the beginning and end boundary values of the path position ratio interval of the second partial path in the thinking path, the interval length, and the correlation between other third partial paths and the second partial path. The weighted calculation result is used as the second partial path to indicate the second possibility of the user generating other third partial paths in the future.

9. The method for constructing a digital twin model of a reducer for optimizing machining accuracy according to claim 4, wherein: The step of generating control rules for the optimization decision data flow traversed by the simulation engine of the control digital twin includes: Based on the optimized thinking path traversed and the user's new operation behavior, the control rules are matched to generate templates; Based on the control rule generation template, the simulation engine that controls the digital twin generates control rules for the optimization decision data flow that traverses the optimization thinking path.

10. A reducer digital twin model construction system for machining accuracy optimization, characterized in that: include: The prediction module is used to predict the various optimization paths that users may generate when optimizing machining accuracy using the reducer's digital twin. A building module is used to build an active auxiliary module for each optimization thinking path. When the user's operation behavior matches any optimization thinking path, the corresponding active auxiliary module is activated in real time, and the optimization decision data flow is generated through the simulation engine of the digital twin; A building module is used to embed the active auxiliary module of each optimized thinking path into the interactive logic layer of the digital twin to build a digital twin model of the reducer.

Images