Il metodo di estrazione delle caratteristiche per il modello 3D delle parti
La premessa della progettazione del processo è la conversione delle informazioni sulla parte dalla conoscenza del progetto alla conoscenza del processo. Nell’ingegneria reale, la conoscenza del design è inclusa nel modello 3D della parte e nei disegni 2D. Nel processo di conversione alla conoscenza del processo, il progettista del processo deve leggere il modello 3D e i disegni della parte. , Acquisire artificialmente conoscenze di progettazione e progettare processi corrispondenti.
With the emergence of the “multi-variety, small batch” production model, the way to obtain design knowledge by humans has severely restricted work efficiency. People hope that the acquisition of design knowledge can be realized automatically through computers.
Design knowledge is mainly oriented to parts and mainly refers to the relationship between machining features and their feature attributes and machining features. Therefore, the purpose of feature extraction is to obtain the processing feature element matrix and the processing feature relationship matrix.
In the current mainstream 3D modeling process, feature-based description ideas are also adopted. To achieve feature extraction, it is necessary to define features in advance. There are two types of feature definition methods:
- 1) Predefine the generation method of the feature parameters of the product, drive the model generation by controlling the feature parameter value, and realize the feature parameter extraction while designing the model. The two complement each other, which is called feature pre-definition, or feature-based parameterization Modeling.
- 2) Use basic geometric elements such as points, lines, and surfaces to define feature parameters. Model design and feature extraction do not interfere with each other. After the model design is completed, the feature extraction is completed by identifying the organizational form of the geometric elements of the model and matching the feature parameters. It is the post-definition of features, or feature recognition.
Compared with feature recognition methods, parametric modeling methods have the following advantages:
- 1) Parametric modeling can not only complete feature extraction, but also assist designers in rapid modeling.
- 2) The parameterized modeling method extracts features more accurately, and does not cause misjudgments or omissions. The feature recognition method is only suitable for features with large differences, and features with similar geometric shapes are easy to confuse.
Therefore, this article is based on the Siemens NX10.0 secondary development API, and adopts the parametric modeling method to realize the feature extraction of parts processing. Since the parameterized modeling has defined the characteristic parameters in advance, the parameter value can be directly called by the parameter name to construct the processing characteristic element matrix.
On the other hand, it provides users with the function of selecting the reference surface, which can obtain the reference feature by identifying the reference surface; the attachment feature can be obtained through the attachment surface; the same type of feature can be obtained by matching the feature parameters of the same type of processing feature attached to the same surface. This can build a matrix of characteristic elements.
In actual engineering, similar features are generally treated as a composite feature and processed together in one step. Therefore, post-processing is required during feature extraction, that is, after identifying similar features, they are combined into a composite feature. Since the other parameters of the similar features except for the position are the same, the position parameters of the merged similar features in the feature element matrix adopt the position of the feature modeled first in the modeling process. In addition, after merging, the remaining processing features do not have similar features, so the feature relationship matrix only needs to retain the first two symbols, that is, the reference relationship and the dependency relationship.