Research On Customer Value-oriented Conceptual Design Method For Smart Product

The development of computer and internet technologies improves the intelligence of product which consists of complex systems rather than simple parts or components.The product is changed greatly.Traditional product is being replaced by smart product.However,smart product employs connected systems to communicate with the working environment timely.It has one or more capacities involving autonomy,self-learning,self-adaptation,self-reasoning,self-coordination,sefl-monitoring,selfdiagnosis,self-repair and self-organization,etc.The interaction between traditional product and the working environment is very weak.Traditional product usually does not have these capacities.Even if it has one or several of the above capacitiess,the intelligent level of traditional product is less than that of smart product.So the conceptual design theories and methods of traditional product are not suitable for smart product.Moreover,the development of smart product is in the initial stage.The relevant theories,methods and techniques need to be studied,and smart product has not been systematically developed.This causes the development of smart product slowly.Therefore,this paper focuses on the key issues of smart product conceptual design based on analysis of existing research methods and industry practice.(1)Conceptual design framework of smart product.On the basis of analyzing technology stack and working principle of smart product,the conceptual design framework of smart product is proposed,and the design processes and key technologies which support the framework are also developed.(2)Identification and analysis of customer requirements for smart product.Customer requirements are classified into must-be quality requirements,one-dimensional requirements and attractive requirements by Kano model.Functional requirements are identified to satisfy customer requirements correspondingly in the axiomatic design.Other stakeholders pay the cost to realize the functions of smart product,and customers use the functions to obtain the value.After the linear equilibrium strategy,the profit is maximized.Finally,customer requirements and functional requirements are determined.(3)Requirement conversion and resolution of technical conflicts.The hierarchical classification method is adopted to identify technical attributes from functional requirements.The relationships between customer requirements and technical attributes are determined in the fuzzy QFD.Fuzzy TOPSIS is improved to evaluate the technical attribute importance.To identify technical conflicts between technical attributes,the line-fitting method is used.Then,the TRIZ theory is improved to solve technical conflicts.(4)Design of monitoring and control functions of smart product.Pairwise comparative algorithm is used to cluster technical attributes into modules which are assembled into smart products.To monitor smart product,wireless sensor network is developed,and K-means algorithm is adopted to deal with monitoring data.Triangular fuzzy numbers are used to evaluate the maturity of monitoring function.Self-tuning fuzzy PID controller employs fuzzy rules,which have been optimized by improved adaptive genetic algorithm,to control smart product remotely.After evaluating the maturity of control function by triangular fuzzy numbers,the design of control function is determined.(5)Design of optimization and autonomy functions of smart product.Iterative learning control system adopts PD learning law to realize optimization function of smart product.Triangular fuzzy numbers are employed to evaluate the maturity of optimization function.To design autonomy function,adaptive neuro fuzzy inference system and multi-agent system are integrated.The former deals with the information from monitoring,control and optimization functions,and the latter gives the optimal strategy by using Q-learning algorithm.The maturity of autonomy function is evaluated by triangular fuzzy numbers.Finally,all systems are integrated to form a closed loop,and the conceptual design of smart products is completed.To verify the theories,methods and technologies of conceptual design of smart product,intelligent cleaning robot and intelligent refrigerator are chosen.The results are good.