Abstract:When designing complex engineered systems, designers typically encounter different types of complexity in decision-making, such as learning trade-offs among conflicting system goals and concurrently meeting specifications of subsystems while complying with all constraints. Therefore, when managing knowledge of the concurrent design of multi-component systems, one must consider the decision interactions among designers with different interests and the coupling effects of subsystems. Conventional multi-objective optimization methods are inadequate in providing decision support on compromising the achievement of conflicting goals under various design circumstances; they are incompetent to provide knowledge from the supply side. They may also fail to solve the problems endorsed by decision-makers' evolving preferences; that is, they are incompetent in gaining knowledge from the demand side. Enhancing decision support and managing various stakeholders' preferences is essential for engineering design problems. To address this issue, we propose a method to give systematic guidance for knowledge discovery and reuse in learning the trade-offs among conflicting goals with various preferences. Using the proposed method, a designer can (1) select appropriate design scenarios to compromise the achievement of system goals while maintaining an acceptable level of fidelity and (2) explore and refine the subsystems coupling in a structured and computable manner. We made the contributions above through (1) reusable knowledge identification and ontology development for creating and archiving the knowledge associated with the problem-solving process, (2) implementing interactive decision-making by adopting different guiding strategies, (3) working on satisfying solutions, and (4) developing icon-based knowledge representation and execution of decision workflows in the knowledge-based design guidance system. The whole process forms a loop, namely, the formulation-refinement-exploration-improvement loop. We demonstrate the efficacy of our method using a test problem, the design of a small-scale thermal system that is widely applied in freshwater production and agricultural irrigation.

Exploratory Study on Cognitive Information Gain Modeling and Optimization of Personalized Recommendations for Knowledge Reuse

Meta Concept Recommendation Based on Knowledge Graph

Study of the Platform for Customized Knowledge Discovery from Industry Web Resources and Personalized Sharing Service

Knowledge Graph-Based Mapping and Recommendation to Automate Life Cycle Assessment

Research on Personalized Recommendation Algorithms for the Web with Knowledge Representation

Personalized knowledge push system based on design intent and user interest

Research on Personalized Learning Resource Recommendation Based on Knowledge Graph Technology

Personalized Recommendation of Learning Resources Based on Knowledge Graph

Analysis and Design of a Personalized Recommendation System Based on a Dynamic User Interest Model

Knowledge recommendation for workplace learning: a system design and evaluation perspective

Creating Personalized Recommendations in a Smart Community by Performing User Trajectory Analysis through Social Internet of Things Deployment

Empowering recommender systems using automatically generated Knowledge Graphs and Reinforcement Learning

Personalized recommendation system based on knowledge embedding and historical behavior

Research on Personalized Recommendation Algorithm of Internet Platform Goods Based on Knowledge Graph

Personalized Online Learning Resource Recommendation Based on Artificial Intelligence and Educational Psychology

Decision-guidance method for knowledge discovery and reuse in multi-goal engineering design problems

Construction of personalized recommendation model for educational video game resources based on knowledge graph

Give Me Something Unknown: Incorporate Exploration Preference In Cognition Into Recommender System

A Personalized Collaborative Filtering Recommendation System of Network Document Resource Based on Knowledge Graph

Paper Recommendation Based on the Knowledge Gap Between A Researcher'S Background Knowledge and Research Target

A knowledge graph algorithm enabled deep recommendation system