A Prototype System Realization Of Port-Based Ontology Towards Conceptual Design

With the development of science and technology, the trend of economic globalization is more and more obvious in the increasingly market competition, product requirements are unceasing changes. In order to remain invincible on the market, we must pay attention to product conceptual design stage. People are looking for creative ideas and design methods to promote product development so as to improve product quality and function. If we can achieve the result at this stage, it is not only substantial cost savings but also to meet the diverse needs of people. This paper puts forward the matrix based on port-based ontology and prototype software system. In this research, main work and results are described as follows.1. From the perspective of port, we set up port ontology knowledge base, rule base, component database and formulate the rules and pass the compatibility of the port for a multi-attribute classification, and design ontology prototype system by C++ programs.2. PBO incidence matrix model, through analyzing the properties of connections between ports to build network connectivity model and using graph theory, PBO network model can be transformed into incidence matrix. Using port attributes, we can simplify PBO incidence matrix. The connections of component are turned into coding connections of port attribute.3. Thought of using genetic algorithms operate PBO incidence matrix codes, and choosing roulette wheel selection method, single-point crossover method, the boundary variation method-based algorithms, as well as crossover rate. Mutation rate control parameter conducted a selection, using multi-objective comprehensive evaluation method to evaluate the solutions of the products. Select the optimal solution of the product conceptual design.4. Above the theoretical basis ,to study probe case, using Protégésoftware to establish liquid mixture container ontology, and using MATLAB software programming to operate the PBO incidence matrix code by choose, crossover and mutation, and resulting in a number of solutions which verifies the feasibility of incidence matrix modeling method.