Research On Modeling Analysis Theory And Method Of Crowd Intelligence Collaborative Innovation Process For Complex Product

With the continuous improvement of society’s demand for products,complex product design is confronted with difficulties such as high R&D costs,lack of design capabilities,low user satisfaction,and low design effectiveness.Recently,more companies are building Internet community and complete complex product innovation cooperatively with external intelligence resources,which forms a new product innovation model: crowd intelligence collaborative innovation(CICI).At present,the academic research pays little attention to the in-depth mechanism and principles of CICI,such as the theoretical framework,internal relations,and process stability.To solve the practical problems in the application and explore the laws and principles of complex product CICI,this paper stresses the issues of dynamic resource identification,resource converging uncertainty,and process instability during the innovation process,and studies the following content:(1)The theoretical model of complex products CICI is proposed.Faced with the absence of a complete and effective theoretical model of complex product CICI,this paper first analyzes the organizational model of complex product CICI and gives the definition.Then,the features such as openness,diversity,uncertainty,and dynamic evolution are concluded.After that,the three elements of CICI,including innovation projects,crowd intelligence resource(CIR),and crowd intelligence community(CIC)are deeply analyzed.Based on this,the framework model of complex product CICI is established.The research results lay a solid theoretical foundation for the postscript.(2)The success factor model of complex product CICI,complexity measurement model of innovation activity(IA),and knowledge expression method are put forward.Based on the theory of persuasion and diffusion of innovation,a model of key factors affecting the success of complex product CICI is established.The empirical analysis results show that the project itself as well as the project creators and participants are all affecting factors.Starting from the perspectives of process complexity and problem complexity,a complexity measurement model of IA is established,and a complexity measurement method based on information entropy is proposed.This method can accurately measure the complexity of IA.With the goal of discovering and calculating knowledge capabilities of community and the CIR,an ontology-based keyword set knowledge representation method is proposed.This method can extract knowledge quickly and easily from content published in the community.(3)A dynamic CIR identification method based on participation behavior is proposed.The participation behavior is both social behavior and innovative behavior.According to the participation way,the pattern of establishing interaction relationship is defined and the method of constructing interaction relationship matrix is proposed.The network diagram of interaction relationship is generated to analyze network characteristics.The results show that the user presents as “Core-Edge” structure.Then,based on user-relationship characteristics and innovation characteristics,the K-Means clustering algorithm is used to classify users into different types.Through statistical analysis methods,six different role types and the contribution number and quality of each type users to the innovation process are identified.The results show that there are significant differences in the behavior and capabilities of different user types.(4)A dynamic CIR agglomeration model and method based on the demand for IA is put forward.Process organizational requirements,problem-solving needs,and knowledge requirements are measured to represent the IA demand model.Then,the CIR model is proposed,including resource type capability model,resource individual knowledge model,and dynamic change model of resource quantity.Taking meeting the requirements of IA as a constraint,a two-stage resource dynamic agglomeration method is studied and proposed based on the matching relationship between design requirements and resource capabilities.The first stage determines the CIR type and quantity based on genetic algorithms,and the second stage uses the text similarity algorithm and ant colony algorithm to determine and select CIR individuals.Through case analysis,it shows that the aggregation method can get the optimized CIR combination that meets the demand and has the highest degree of interaction.(5)A CIR optimization configuration method for CICI process is put forward.The project process model is established based on hierarchical Petri nets.The upper-level network describes the interrelationship among various IAs and the substitution transition represents the logical relationship of the implementation process of IA.To involve more CIR into the innovation process and increase the participation rate,a resource optimization configuration model and solution algorithm are proposed.Case analysis shows that the model and algorithm are reasonable and effective.To encourage the configured CIR to participate in the corresponding IAs,the participation motivations are studied based on the expectations theory and the nethnography.A numerical analysis of the users’ behavioral quantity and behavioral attributes yields behavioral expectations and behavioral motivations,which provides theoretical methods for formulating incentive measures and management methods.This paper studies the CICI from the three levels of “elements-relation-systems”.The research results improve and expand the theory and method of CICI,and lay the foundation for the in-depth study of CICI.Besides,it provides theoretical guidance for application CICI in practice and has strong theoretical and practical value.