Knowledge engineering technology for gearbox of agricultural machinery tractor

Knowledge is the decisive factor of new product competitiveness. The product design based on knowledge engineering is to summarize and sort out the mature design experience and knowledge, inherit and reuse them in the new product design, so as to achieve the effect of rapid response design. Knowledge engineering technology mainly includes knowledge acquisition, knowledge representation and knowledge application. Take gear as an example for specific analysis.

1.Gear knowledge acquisition

Knowledge acquisition is to extract the required knowledge from some knowledge sources, including network data, domain experts, text documents, etc. According to the different ways of knowledge acquisition, knowledge acquisition can be divided into manual, semi-automatic and automatic acquisition.

In this paper, gear knowledge acquisition mainly through manual way, from the design manual, scientific literature, case model to sort out the gearbox gear parameters, formulas, features, structure and other design knowledge, after refining into the CAD system, so as to achieve the reuse of gear design knowledge.

2.Gear knowledge representation

Knowledge representation refers to the representation of design knowledge with symbols that can be recognized and processed by computers. Common knowledge representation methods include production representation, framework representation, object-oriented representation, artificial neural network representation and ontology based representation, and different representation methods have their own advantages and disadvantages and application scope.

The knowledge in the process of gear modeling is mainly parameter formula and condition judgment. In this paper, production representation is used. The knowledge of gear is expressed as the language that NX expression can recognize, and the knowledge-based intelligent parametric design effect is realized by intelligent judgment of gear structure through design rules. For example, the knowledge of gear structure division can be expressed as:

Rule1: if e < 2m then gear shaft

Rule2: if e ⩾ 2M and Da ⩽ 160 then solid gear

Rule3:IF e⩾2m AND 160

Rule4:IF e⩾2m AND 500

Where, e — distance from the bottom of keyway to the root circle, m m; m — modulus, mm; Da — diameter of tooth top circle, mm.

The design knowledge of gear has the characteristics of fuzziness, implicitness and diversity. It is difficult to describe all the design knowledge effectively and comprehensively by using a single knowledge representation method. In this paper, ontology representation is applied to the knowledge representation of gear field to further improve the flexibility and accuracy of gear design knowledge representation. Ontology, originated from the concept of philosophy, is a clear formal specification of shared conceptual model. The construction of gear knowledge base based on ontology can better realize the sharing, reuse, expansion and maintenance of gear knowledge.

Prot é g é is an ontology development tool developed by Stanford University. It is fully open-source and has many plug-ins. It has become the most popular ontology editor. In this paper, gear domain knowledge is collected and sorted out, and gear domain ontology is built by using prot é g é – 5.5.0 tools. The link relationship between gear concepts is shown in Figure 2. Firstly, the important concepts of gears are sorted out, and the hierarchical relations of corresponding classes and subclasses are defined in prot é g é. Then, the attributes and instances of classes are defined. Finally, the ontology concept tree is improved and updated.

Ontology can not only describe the semantic network relationships among different concepts, such as kind of (inheritance relationship), part of (part and whole relationship), instance of (concept and instance relationship), attribute of (attribute relationship), etc., but also define the data attributes, object attributes, annotation attributes of different concepts, and explicitly express various concept relationships [23]. The gear knowledge base based on ontology can ensure the consistency of gear concept knowledge, and infer the implicit relationship between concept knowledge, so that the gear domain knowledge can be more effectively expressed and reused.

3. Gear knowledge application

The use of knowledge means that computers can simulate people’s thinking mode, use knowledge to reason and solve problems, and have certain intelligent design ability. For the gearbox virtual design system, knowledge is mainly used in the intelligent judgment and selection of gear structure features, so that a model has the ability to evolve multiple structures. Knowledge engineering technology is integrated into CAD system to improve the intelligence of gear design. As shown in figure a, the simple parametric system can only realize the rapid change of the parameters or dimensions of the gear geometry model, and can not change the structural features of the gear. And Figure 3B The intelligent parametric system shown in this paper integrates the structural design knowledge of gears, defines the structural design rules of gears, automatically selects the structural features of gears according to the input parameters, and determines the number of keyways, key width, key height size, chamfer radius and other specific parameters according to the diameter of gear shaft hole. The designer’s “intellectual activities” refer to the process program of judging the design manual So that CAD system can simulate human thinking for intelligent design.