Parameterized instruction prediction method and system for CAD modeling

Abstract

The invention provides a parameterization instruction prediction method and system for CAD modeling. The parameterization instruction prediction method comprises the following steps that S1, a parameterization sequence of an existing CAD model of a user is collected; s2, predicting future finite step number operation of the user; s3, pushing the prediction operation to the user; and S4, according to user selection, executing corresponding functions, such as node selection and parameter input. According to the method, a user clicks a mouse key rightwards, the first n predicted instructions appear around a mouse cursor in a circular area, the user moves a mouse to click the instructions, and the predicted instructions or sequences are pushed to the user in a list mode. The predicted instructions of the first n items appear in the instruction prediction area, the user moves a mouse to click the instructions, the user can directly, intuitively and quickly select and click the instructions, and then the next round is started, so that the instruction click time of each round is greatly shortened, and the completion of the generation of a design task is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of CAD modeling parameterized instruction prediction, and in particular to a CAD modeling parameterized instruction prediction method and system. Background Art

[0002] CAD models are usually generated in the form of parametric instructions. Usually, the corresponding parametric instructions are generated according to the function options clicked by the user, and the parameter values ​​are determined according to the specific values ​​subsequently entered by the user. Commonly used design software, such as Creo, Ug, AM, CATIA, etc., can prompt the user to enter the parameters according to the function selected by the user, but only the parameters required by the selected function itself are limited.

[0003] As users spend more time using design software, designers often need to design models that are similar to previous design work, or continuously use the same series of functional processes. Existing design software is unable to record and reuse designers' existing work records, causing inconvenience to design tasks. Therefore, a parametric instruction prediction method for CAD modeling is involved, which can use the historical work records of designers calling parametric instructions to train a model for predicting future parametric instructions. Summary of the invention

[0004] (I) Purpose of the invention

[0005] The purpose of the present invention is to provide a parametric instruction prediction method for CAD modeling in view of the deficiencies of the prior art. When the user right-clicks the mouse button, the predicted instructions of the first n items appear in a circular area around the mouse cursor. The user moves the mouse to click the instruction, and the predicted instruction or sequence is pushed to the user in the form of a list. The method can also be: setting an instruction prediction area on the work interface, the predicted instructions of the first n items appear in the instruction prediction area, the user moves the mouse to click the instruction, the user can directly and intuitively and quickly select and click, and then enter the next round, which greatly shortens the instruction clicking time of each round and facilitates the completion of the design task.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A parametric instruction prediction method for CAD modeling comprises the following steps:

[0009] S1, collecting the parameterized instruction sequence of the user's existing CAD model;

[0010] S2, a sequence input recurrent neural network model, predicts the user's future parameterized instruction options with a finite number of steps;

[0011] S3, push the prediction instruction to the user;

[0012] S4. Execute corresponding instructions according to user selection;

[0013] S5a. If the user does not select the recommended instruction and executes other instruction operations, the prediction is deemed incorrect, and the existing sequence and the predicted sequence are updated according to the other instructions executed by the user.

[0014] S5b, if the user selects the recommended instruction, it is deemed that the prediction is correct, and the corresponding parameterized instruction is called according to the option selected by the user;

[0015] S6. Provide reference values ​​of nodes, parameters, etc.;

[0016] S7, determining specific parameters of the instruction according to the user's feedback;

[0017] S8. Execute corresponding CAD parameterization instructions.

[0018] The beneficial effects of the present invention are:

[0019] (1) The present invention can collect the parametric instruction sequence of the user's existing CAD model to predict the user's future limited-step parametric instruction options. The instructions are pushed to the user in the form of a list. At the same time, two methods are provided to push instructions to the user. The user can intuitively and quickly select and click the instruction, enter the parameters to execute, and then enter the next round of parametric instruction prediction. The instruction click time for each round is extremely short, which is convenient for the selection and completion of design tasks.

[0020] (2) In the present invention, when the current n predicted instructions do not include options that meet the user's expectations, the user can click the switch instruction in the circular area around the mouse cursor; or move the mouse cursor to the instruction prediction area and turn the middle mouse button to switch the predicted instruction. This helps the user select the instruction that does not appear at the front of the prediction list and wants to click, thereby shortening time and improving efficiency.

[0021] (3) In the present invention, if the predicted instruction is incorrect, the user can directly click on the operation or click on the search bar to search and then click to select the operation he wants; the operation selected by the user after directly clicking on the operation or clicking on the search bar to search will be recorded and used in the recurrent neural network model for regular iterative updates, which provides convenience for the selection of subsequent instructions. The data in the user's recurrent neural network is updated and trained at regular intervals, and the updated neural network model can be shared with other users. New users can directly reuse the existing work records of other designers, which provides great convenience. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0023] Figure 2 This is a schematic diagram of the structure of the parameterized instruction sequence set module of the present invention;

[0024] Figure 3 It is a branch flow chart of a time-ordered instruction sequence set of the present invention;

[0025] Figure 4 This is a schematic diagram of the process of arranging the number of clicks from high to low in the present invention;

[0026] Figure 5 This is a schematic diagram of pushing the first predicted operation instruction of the present invention;

[0027] Figure 6 This is a schematic diagram of pushing the second predicted operation instruction of the present invention;

[0028] Figure 7 This is a schematic diagram of pushing the third predicted operation instruction of the present invention;

[0029] Figure 8 This is a schematic diagram of pushing the fourth predicted operation instruction of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0032] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0033] Embodiment 1: Figure 1-Figure 6As shown, this embodiment provides a parameterized instruction prediction method for CAD modeling, comprising the following steps:

[0034] S1, collecting the parameterized instruction sequence of the user's existing CAD model;

[0035] S2, a sequence input recurrent neural network model, predicts the user's future parameterized instruction options with a finite number of steps;

[0036] S3, push the prediction instruction to the user;

[0037] S4. Execute corresponding instructions according to user selection;

[0038] S5a. If the user does not select the recommended instruction and executes other instruction operations, the prediction is deemed incorrect, and the existing sequence and the predicted sequence are updated according to the other instructions executed by the user.

[0039] S5b, if the user selects the recommended instruction, it is deemed that the prediction is correct, and the corresponding parameterized instruction is called according to the option selected by the user;

[0040] S6. Provide reference values ​​of nodes, parameters, etc.;

[0041] S7, determining specific parameters of the instruction according to the user's feedback;

[0042] S8. Execute corresponding CAD parameterization instructions.

[0043] like Figure 2 As shown, the detailed steps of step S1 are as follows:

[0044] Collecting a set of parameterized instruction sequences (M1, M2, M3, M4...Mn) of the model in the user's current workspace, wherein the instruction sequence M1 includes a set of sub-instruction sequences (M11, M12, M13, M14...M1n);

[0045] The instruction sequence M2 includes a set of sub-instruction sequences (M21, M22, M23, M24...M2n);

[0046] The instruction sequence M3 includes a set of sub-instruction sequences (M31, M32, M33, M34...M3n);

[0047] The instruction sequence M4 includes a set of sub-instruction sequences (M41, M42, M43, M44...M4n); ...

[0049] The instruction sequence Mn includes a set of sub-instruction sequences (Mn1, Mn2, Mn3, Mn4...Mnn);

[0050] It should be noted that: M1 itself is an instruction. Clicking M1 instruction enters the working module of M1 instruction, which includes instructions M11, M12, M13, M14...M1n. The user then clicks, and it can be said that M11, M12, M13, M14...M1n are sub-instructions of M1.

[0051] Collecting individual instructions in the parameterized instruction sequence set clicked by the user, and arranging them in chronological order to obtain a chronological instruction sequence set (P1, P2, P3, P4...Pn), wherein P1, P2, P3, P4...Pn belong to instructions in (M1, M2, M3, M4...Mn) or (M11, M12, M13, M14...M1n) or (M21, M22, M23, M24...M2n) or (M31, M32, M33, M34...M3n) or (M41, M42, M43, M44...M4n),...or (Mn1, Mn2, Mn3, Mn4...Mnn);

[0052] After multiple collections, we can obtain the time sequence instruction sequence set branch process data set (T1, T2, T3, T4...Tn), obtain the time sequence instruction sequence set branch process data set (T1, T2, T3, T4...Tn), and transmit the time sequence instruction sequence set branch process data set (T1, T2, T3, T4...Tn) and the time sequence instruction sequence set branch process flow chart to the recurrent neural network model for reasoning. The time sequence instruction sequence set (P1, P2, P3, P4...Pn) is just a transfer relative to (T1, T2, T3, T4...Tn) and can be omitted;

[0053] The detailed steps of step S2 are as follows:

[0054] The recurrent neural network model infers the branch process data set (T1, T2, T3, T4...Tn) of the time-ordered instruction sequence set to predict the user's future operation intentions for a limited number of steps.

[0055] Define any Ti in the time-order instruction sequence set branch process data set, where Ti∈(T1, T2, T3, T4...Tn); according to the time-order instruction sequence set branch flow chart, such as Figure 3 As shown;

[0056] The next step of Ti operation specifies to arrange them in descending order according to the number of clicks to obtain the sequence Ta1, Ta2, Ta3, Ta4...Tan, such as Figure 4 As shown;

[0057] In step S3, the first n items in the sequence Ta1, Ta2, Ta3, Ta4, ... Tan are positioned as predicted instructions and pushed to the user in a list form.

[0058] The predicted instructions or sequences are pushed to the user via the area around the mouse cursor as follows:

[0059] The user right-clicks the mouse button, and the first n predicted instructions appear in a circular area around the mouse cursor. The user can move the mouse and click to select an instruction. In this embodiment, n is 8. Figure 5 shown.

[0060] The predicted instructions or sequences are pushed to the user through the instruction prediction area in the following way:

[0061] An instruction prediction area is set on the work interface, and the first n predicted instructions appear in the instruction prediction area. The user moves the mouse to click on the instruction. In this embodiment, n is 8, where Ta1, Ta2, Ta3, Ta4, Ta5, Ta6, Ta7, and Ta8 are the instructions with the highest number of clicks after instruction Ti predicted by the user's recurrent neural network model, such as Figure 6 As shown in the figure, users can intuitively and quickly select and click commands, input parameters to execute, and then enter the next round of parameterized command prediction. The time for clicking commands in each round is extremely short, which is convenient for selecting and completing design tasks.

[0062] In step S4, if the predicted instructions of the first n items are incorrect, the user can directly click on the operation or click on the search bar to search and then click to select the operation they want; the operation selected by the user after directly clicking on the operation or clicking on the search bar to search will be recorded and used in the recurrent neural network model for regular iterative updates, which provides convenience for the selection of subsequent instructions. The data in the user's recurrent neural network is updated and trained at regular intervals, and the updated recurrent neural network model can be shared with other users. New users can directly reuse the existing work records of other designers, which provides great convenience.

[0063] Embodiment 2: Figure 7 and Figure 8 As shown, the components identical or corresponding to those in the first embodiment are marked with the corresponding reference numerals in the first embodiment. For the sake of simplicity, only the differences from the first embodiment are described below. The second embodiment differs from the first embodiment in that:

[0064] The instruction selection in this embodiment is more convenient and quick. In the first embodiment, if the predicted instructions of the first n items are incorrect and do not include the instructions that the user expects to use, the user can click the search bar to search for the desired instruction. Although the search can find the instruction that the customer wants to click, it takes longer;

[0065] The predicted instructions of the current n items appear in a circular area around the mouse cursor, and a switch instruction is set in the circular area; after the mouse cursor clicks the switch instruction, the predicted instructions in the circular area are switched to n+1-2n bits; the switch predicted instructions, such as Figure 7 As shown;

[0066] Or the predicted instructions of the current n items appear in the instruction prediction area. Move the mouse cursor to the instruction prediction area and turn the middle mouse button to switch the predicted instructions, such as Figure 8 As shown; it helps users select the command they want to click that does not appear at the top of the prediction list, so as to shorten the time and improve efficiency.

[0067] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A parametric instruction prediction method for CAD modeling, characterized in that: The following steps are involved: S1, collecting the parameterized instruction sequence of the user's existing CAD model; S2, a sequence input recurrent neural network model, predicts the user's future parameterized instruction options with a finite number of steps; S3, push the prediction instruction to the user; S4. Execute corresponding instructions according to user selection; S5a. If the user does not select the recommended instruction and executes other instruction operations, the prediction is deemed incorrect, and the existing sequence and the predicted sequence are updated according to the other instructions executed by the user. S5b, if the user selects the recommended instruction, the prediction is deemed to be correct, and the corresponding parameterized instruction is called according to the option selected by the user; S6. Provide reference values ​​of nodes, parameters, etc.; S7, determining specific parameters of the instruction according to the user's feedback; S8. Execute corresponding CAD parameterization instructions.

2. A parameterized instruction prediction method for CAD modeling according to claim 1, characterized in that: The detailed steps of step S1 are as follows: Collecting a set of parameterized instruction sequences (M1, M2, M3, M4...Mn) of the model in the user's current workspace, wherein the instruction sequence M1 includes a set of sub-instruction sequences (M11, M12, M13, M14...M1n); The instruction sequence M2 includes a set of sub-instruction sequences (M21, M22, M23, M24...M2n); The instruction sequence M3 includes a set of sub-instruction sequences (M31, M32, M33, M34...M3n); The instruction sequence M4 includes a set of sub-instruction sequences (M41, M42, M43, M44...M4n); ... The instruction sequence Mn includes a set of sub-instruction sequences (Mn1, Mn2, Mn3, Mn4...Mnn).

3. A parameterized instruction prediction method for CAD modeling according to claim 2, characterized in that: Collect single instructions in the parameterized instruction sequence set clicked by the user, and arrange them in chronological order to obtain a chronological instruction sequence set (P1, P2, P3, P4...Pn), where P1, P2, P3, P4...Pn belong to the instructions in (M1, M2, M3, M4...Mn) or (M11, M12, M13, M14...M1n) or (M21, M22, M23, M24...M2n) or (M31, M32, M33, M34...M3n) or (M41, M42, M43, M44...M4n),...or (Mn1, Mn2, Mn3, Mn4...Mnn). After multiple collections, a time sequence instruction sequence set branch process data set (T1, T2, T3, T4...Tn) can be obtained, and a time sequence instruction sequence set branch flow chart can be obtained. The time sequence instruction sequence set branch process data set (T1, T2, T3, T4...Tn) and the time sequence instruction sequence set branch flow chart can be transmitted to the recurrent neural network model for reasoning.

4. A parameterized instruction prediction method for CAD modeling according to claim 3, characterized in that: The detailed steps of step S2 are as follows: The recurrent neural network model infers the branch process data set (T1, T2, T3, T4...Tn) of the time-ordered instruction sequence set to predict the user's future operation intention for a limited number of steps.

5. A parameterized instruction prediction method for CAD modeling according to claim 4, characterized in that: Define an arbitrary Ti in the time-ordered instruction sequence set branch process data set, where Ti∈(T1, T2, T3, T4...Tn); A flow chart that branches according to a collection of chronological instruction sequences. The next step of Ti operation specifies to arrange them in descending order according to the number of clicks to obtain the sequence Ta1, Ta2, Ta3, Ta4...Tan; In step S3, the first n items in the sequence Ta1, Ta2, Ta3, Ta4, ... Tan are positioned as predicted instructions and pushed to the user in a list form.

6. A parameterized instruction prediction method for CAD modeling according to claim 5, characterized in that: The predicted instructions or sequences are pushed to the user via the area around the mouse cursor as follows: The user right-clicks the mouse button, and the top n predicted instructions appear in a circular area around the mouse cursor. The user can move the mouse and click to select an instruction. The predicted instructions or sequences are pushed to the user through the instruction prediction area in the following way: An instruction prediction area is set on the work interface, and the predicted instructions of the first n items appear in the instruction prediction area. The user moves the mouse and clicks the instruction.

7. A parameterized instruction prediction method for CAD modeling according to claim 6, characterized in that: In step S4, if the predicted instructions of the first n items are incorrect, the user can directly click on the operation or click on the search bar to search and then click to select the operation he wants.

8. A parameterized instruction prediction method for CAD modeling according to claim 7, characterized in that: The operations selected by the user after directly clicking on the operation or clicking on the search bar to search will be recorded and used in the recurrent neural network model for regular iterative updates.

9. A parameterized instruction prediction method for CAD modeling according to claim 8, characterized in that: The predicted instructions for the current n items appear in a circular area around the mouse cursor, and a switching instruction is provided in the circular area; Or the predicted instructions of the current n items appear in the instruction prediction area. Move the mouse cursor to the instruction prediction area and turn the middle mouse button to switch the predicted instructions.

10. A parametric instruction prediction system for CAD modeling, applied to the parametric instruction prediction method for CAD modeling according to any one of claims 1-9.

Images