Model Construction for Complex Product Design Change with Improved Dendritic Neural Network

Abstract

Design changes are widespread during the produce development due to many uncertain factors, e.g., the impact of the epidemic on the supply chain, changes in customer demands, and upgrades demand by the produce production itself. It is very important for companies to predict product’s parameter changes and clarify the mechanism of change propagation for saving costs and improving efficiency. However, the current product change methods are often based only on the experience and knowledge of experts, and lack of effective use of the historical data of design changes. Motivated by this, we here propose a data-driven mechanism to construct the design change model with an improved dendritic neural network. A dendritic neural network is first improved to model the design change into a complex network according to the features of the design change data. The method for obtaining the linkage strengthen weights is developed based on the input-output relationship of the units in the trained dendritic neural network. The effectiveness of the proposed algorithm is demonstrated by applying to the design of a 2D-modeled bicycle.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 4 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.