The acquisition of valuable design knowledge from massive fragmentary data is challenging for designers in conceptual product design.This study proposes a novel method for acquiring design knowledge by combining deep ...The acquisition of valuable design knowledge from massive fragmentary data is challenging for designers in conceptual product design.This study proposes a novel method for acquiring design knowledge by combining deep learning with knowledge graph.Specifically,the design knowledge acquisition method utilises the knowledge extraction model to extract design-related entities and relations from fragmentary data,and further constructs the knowledge graph to support design knowledge acquisition for conceptual product design.Moreover,the knowledge extraction model introduces ALBERT to solve memory limitation and communication overhead in the entity extraction module,and uses multi-granularity information to overcome segmentation errors and polysemy ambiguity in the relation extraction module.Experimental comparison verified the effectiveness and accuracy of the proposed knowledge extraction model.The case study demonstrated the feasibility of the knowledge graph construction with real fragmentary porcelain data and showed the capability to provide designers with interconnected and visualised design knowledge.展开更多
During the last two decades,partial discharges(PDs)modelling methods have been used as a complement of insulation diagnosis systems of electrical assets.Finite-element-analysis models for simulating PDs in cavities wi...During the last two decades,partial discharges(PDs)modelling methods have been used as a complement of insulation diagnosis systems of electrical assets.Finite-element-analysis models for simulating PDs in cavities within solid dielectric materials are reviewed and a novel model is presented,which combines the main advantages of electrostatic and electric current models.The theoretical background is presented and some limitations and restrictions are discussed.A case of study was implemented for three different ageing conditions and simulation results exhibit good agreement with reported values by other authors in the literature.An analysis of variations of PD behaviour with ageing as a function of temperature and pressure is briefly presented.It is concluded that more research is needed to include physical and chemical interactions on the void surface and that the cavity surface conductivity plays a fundamental role in the PD simulations at advanced ageing conditions.展开更多
Electrical trees are defects originated and driven by partial discharges(PD)activity and this is the reason why their modelling and simulation are fundamental for the ageing mechanisms comprehension and diagnosis anal...Electrical trees are defects originated and driven by partial discharges(PD)activity and this is the reason why their modelling and simulation are fundamental for the ageing mechanisms comprehension and diagnosis analyses of insulation systems compound by solid dielectrics.This study presents a brief review of the different models implemented to simulate PD in electrical trees inside solid dielectrics.In addition,an improved deterministic model as well as a stochastic model are presented,which allow predicting adequately the PD pulses distribution,in magnitudes and phase,for non‐conducting electrical trees of different shape and length.Two case studies were simulated and their results exhibit good agreement when compared with measurements presented in the literature.It is concluded that the main parameters that state the PD behaviour in trees are the inception and extinction voltage mag-nitudes,including their probability distributions,and the tree geometry.The models can be used for prognosis analysis based on on‐line measurements.展开更多
基金This research is supported by the Chinese Special Projects of the National Key Research and Development Plan(2019YFB1405702).
文摘The acquisition of valuable design knowledge from massive fragmentary data is challenging for designers in conceptual product design.This study proposes a novel method for acquiring design knowledge by combining deep learning with knowledge graph.Specifically,the design knowledge acquisition method utilises the knowledge extraction model to extract design-related entities and relations from fragmentary data,and further constructs the knowledge graph to support design knowledge acquisition for conceptual product design.Moreover,the knowledge extraction model introduces ALBERT to solve memory limitation and communication overhead in the entity extraction module,and uses multi-granularity information to overcome segmentation errors and polysemy ambiguity in the relation extraction module.Experimental comparison verified the effectiveness and accuracy of the proposed knowledge extraction model.The case study demonstrated the feasibility of the knowledge graph construction with real fragmentary porcelain data and showed the capability to provide designers with interconnected and visualised design knowledge.
文摘During the last two decades,partial discharges(PDs)modelling methods have been used as a complement of insulation diagnosis systems of electrical assets.Finite-element-analysis models for simulating PDs in cavities within solid dielectric materials are reviewed and a novel model is presented,which combines the main advantages of electrostatic and electric current models.The theoretical background is presented and some limitations and restrictions are discussed.A case of study was implemented for three different ageing conditions and simulation results exhibit good agreement with reported values by other authors in the literature.An analysis of variations of PD behaviour with ageing as a function of temperature and pressure is briefly presented.It is concluded that more research is needed to include physical and chemical interactions on the void surface and that the cavity surface conductivity plays a fundamental role in the PD simulations at advanced ageing conditions.
基金niversidad de Antioquia‐Departamento de Ingeniería Eléc-trica,Universidad Politécnica de Madrid,Fondo Sapiencia‐Alcaldía de Medellín and Ministry of Science and Innovation of Spain,National Program of Scientific and Technical Research and Innovation(project PID2019‐107126RB‐C21).
文摘Electrical trees are defects originated and driven by partial discharges(PD)activity and this is the reason why their modelling and simulation are fundamental for the ageing mechanisms comprehension and diagnosis analyses of insulation systems compound by solid dielectrics.This study presents a brief review of the different models implemented to simulate PD in electrical trees inside solid dielectrics.In addition,an improved deterministic model as well as a stochastic model are presented,which allow predicting adequately the PD pulses distribution,in magnitudes and phase,for non‐conducting electrical trees of different shape and length.Two case studies were simulated and their results exhibit good agreement when compared with measurements presented in the literature.It is concluded that the main parameters that state the PD behaviour in trees are the inception and extinction voltage mag-nitudes,including their probability distributions,and the tree geometry.The models can be used for prognosis analysis based on on‐line measurements.
