This paper is concerned with analyzing the evolution of green innovation collaborative network in Beijing- Tianjin-Hebei region and finding the unknown cooperative relationship by using the complex network and link prediction theory. Based on the Green Patent Cooperation data of Beijing- Tianjin-Hebei region from 2009 to 2020, this paper constructed cooperation networks and analyzed the evolution from three aspects: structure, centrality and clustering. Then used the link prediction theory to evaluate 14 prediction indexes to find the optimal link prediction algorithm of green innovation collaborative network, and explored the recommended cooperation in the innovation network, so as to put forward constructive suggestions for the coordinated development of green innovation in Beijing-Tianjin-Hebei region. The experimental results showed that the green patent cooperation network shows an overall growth trend, the structure evolves in a scale-free direction, and the cluster effect presents phased characteristics in the evolution process. In recent years, the centrality of core nodes has decreased slightly, and green innovation synergy has a diversified development trend. In the Beijing-Tianjin-Hebei green patent cooperation network, the link prediction accuracy using Katz index is the highest.

This paper proposes a simulation study of the flexibility of production system based on hybrid modeling to analyze the influence of personnel collaboration on the flexibility of production system, and optimize the organization configuration of production system. Firstly, this study constructs the evaluation index of the flexibility of production system from four dimensions and puts forward the specific calculation formula. Secondly, this study establishes the simulation model of production system and gives the experimental principle. Finally, this study defines three production collaboration strategies, which are used in the key parts production system of a motorcycle engine to simulate the flexibility of production system, on the basis of the modeling method of personnel collaboration. This simulation research makes it possible to realize the comparative analysis of multiple collaboration strategies. Results from the simulation show that personnel collaboration strategy is an important factor affecting the flexibility of production system. In the production environment of medium and large fluctuation, the flexibility of the four dimensions of system output, machine utilization, personnel utilization and production efficiency are improved with the change of collaboration strategy. Results from the experiment also show that it is the hybrid modeling method that can make the dynamic allocation and operation evaluation of elements of production system flexibility come true. Meanwhile, it can provide a dynamic and quantitative research method for production organization optimization.

This paper builds industry-university-research collaborative innovation networks based on the patent cooperation data in the Beijing-Tianjin-Hebei region from 2010 to 2019. Based on the SIS propagation model, this paper conducts simulation experiments to make the university nodes in the networks lose randomly or deliberately. Then, the changes in relative size of largest connected subgraph and network efficiency can be obtained from the simulation experiments. From the experimental results, it can be concluded that the loss of university nodes has a significant impact on the network structure. In addition, in the deliberate loss method, the robustness of the betweenness loss is weaker than that of the degree loss. Therefore, we should focus on strengthening the protection of the university nodes which are top ranked in betweenness centrality. At the same time, this paper innovatively proposes a quantitative index system of robustness, and uses the data obtained from simulation experiments to quantitatively analyze robustness of networks. The results show that the robustness of the Beijing-Tianjin-Hebei collaborative innovation networks have been on the rise in the past decade.

With the development of Internet technology and the advent of the era of knowledge economy, people have higher and higher requirements for the personalized design of products. The transformation of product innovation model from traditional model to open mass innovation model has promoted the rapid development of open source product design pattern. Contributors share opinions, insights, experiences and viewpoints in the open source product community, and a series of knowledge contribution behaviors, continuously injecting high-quality content into the promotion and evolution of products. However, willingness of contributors to contribute knowledge in open source product community may be largely affected by their sentiment attitudes. Based on the research on the sentiment state of contributors in the open source product community, this paper adopts the hidden Markov model to study the willingness of contributors to contribute knowledge and its changing laws. First, we use a pretraining model named XLNet to perform sentiment classification of contributors. Then, the sentiment state of contributors is used as the input of the hidden Markov model, and the state transition probability matrix is obtained after training the model. Finally, according to the state transition probability matrix, knowledge contribution willingness of contributors and its change law are analyzed.

In order to identify the opinion leaders in open source design community, a comprehensive evaluation method of node importance based on TOPSIS is proposed in this paper. Firstly, a user collaboration network model of community organization is established by using complex social network technology. Secondly, in order to improve the accuracy of opinion leader identification, the benefit indicators and cost indicators of each node are measured. Degree centrality, median centrality and near centrality are used as the benefit indicators, meanwhile structural hole is used as the cost indicator. Then the comprehensive evaluation index of nodes is taken as the attribute of decision evaluation scheme to calculate synthetically to determine its importance in the network. On this basis, this paper takes the Local Motors community as the research object and establishes a user cooperative directed weighted network. The static topology index of the network is analyzed. According to the results, the opinion leader node of the user cooperative network is identified based on TOPSIS method. Experiments have shown that more accurate evaluation results can be obtained than using a single indicator.

The worker assignment problem is of great importance in high-variety, small-quantity workshop production systems. This paper presents a cooperative worker assignment model by introducing a flexible worker cooperation strategy and multi-objective optimization to minimize the makespan and maximum worker load and total worker load. To solve the model, an improved NSGAII algorithm is designed, which brings about improvement to the method of selection and trim operation. Then, the contrast experiments of the model and algorithm were carried out and the results show that the cooperative worker assignment model is more competent with the makespan and maximum worker load of the workshop and more assignment scheduling plans will be obtained. In addition, the improved algorithm can effectively avoid local optimization during the calculation process.

Aiming at the problem of the characteristics of different participants influencing on the evolution of the community project, a simulation model is established based on the agent-based method. The mechanism of participant selection and performing tasks and the mechanism of cooperation between participants are added in the model. The evolution situation under different composition structures of participants is compared and analyzed, and a sensitivity analysis of the cooperation probability between participants is introduced. The results shows that the joining of innovation leader could impact the evolution of the community project largely, but the impact is weaken with the increase of the number of innovation leaders. The increase of the number of core developers can promote the evolution of the community project, but the impact is limited. The cooperation probability of participants has a greater influence on the evolution community project. Finally, the management recommendations are given based on the results.

Firstly, a complex network model of participants in the open source design community OpenIDEO is built, and the network index such as degree and degree distribution of complex network model are analyzed. By clustering analysis based on the network index, participants are divided into three categories, respectively the core users, information dissemination users and ordinary users. In addition, we assume that the proportion and activeness of information transmission users plays a key role for the evolution of the network. Then, by clustering participants of each project and analyzing the performance of each project, the assumption is verified. The study result is helpful to optimize the mechanism construction of the open source design community and to promote the benign development of the community, which lay the foundation of improving the efficiency and quality of group development.

Prior researches on knowledge spillover have focused on the innovation of enterprise and ignore the product evaluation. We built a mass collaborative product development model with an effective knowledge spillover methodology. Through contrast experiment, we find that collaborative society networks are small-world, and there was a positive correlation between knowledge spillover rate and cooperate effects. In addition, Knowledge spillover rate can decrease the preferential attachment's Influence Power. Finally, high knowledge spillover rate will lead to a more closely society network.

This paper devoted to propose a human-machine integrated simulation method to evaluate the effect of corrective maintenance (CM) strategies on manufacturing cells considering the worker's autonomous and collaborative behaviors. First, a human-machine integrated model is developed to integrate the worker behaviors in the CM process by using agent technology. In this way, worker's dynamic and cooperative behaviors can be explicitly modeled and simulated. Then, the collaboration behaviors in the CM process and the measurement of the simulation model are discussed. Third, the proposed simulation model is applied in a motor-cycle engine cell, and a comparative simulation experiment is designed and conducted for two different maintenance strategies. Simulation results show the strategy designed with collaboration behavior performed better during the CM process than maintenance strategy without collaboration behavior. The proposed simulation method can dynamically and quantitatively evaluate the CM strategy optimization, especially when considering worker's collaborative and autonomous behaviors.

Mass collaborative product development (MCPD) is an emerging product development mode which makes full use of open group intelligence with the advantage of low cost and high innovation level. In this paper, behaviors of individuals in MCPD, especially collaborative behaviors, are modeled and simulated, and the effects of collaborative behaviors on the whole MCPD process are identified based on the simulation study. Firstly, a multi-agent process model of MCPD is built, which includes product model, community model, and three-level agent model. Secondly, collaboration behaviors including collaboration request, collaborator choosing, collaboration judgment, and collaboration response are depicted. Finally, based on the process model and the individual behaviors of MCPD, a simulation system for MPCD is designed and developed, and two sets of experiments are carried out. The results of experiments show that different weights of collaborator choosing indices significantly influence OSC evolution, and individuals' collaboration intensions significantly influence product evolution. The findings of this paper are expected to help the community managers develop more reasonable management and service mechanisms for participants of the MCPD.

In cellular manufacturing, it is essential to focus on human issues such as worker assignment strategies, skill identification, teamwork and communication. In order to analyze the impact of collaborative work in cellular manufacturing, agent model of workers was developed and integrated into the production simulation model. Then, comparative simulation experiments of different collaborative work schemes are designed based on the background of a motorcycle engine box production cell. The results indicate that collaboration of multi-skill workers can effectively shorten the lead time of the orders if the multi-skill level of the workers reaches a reasonable level, and the autonomous and collaborative organization is more robust facing to the unstable production environment.

Resource conflict resolution of the collaborative product development process is a crucial problem for the project management. To describe the arbitration behavior in the agent-based simulation, an arbitration agent was introduced and modeled. Some arbitration strategies were constructed according to the task utility and resource matching degree. Then, the agent and it's strategies were integrated into the agent-based simulation platform. Some comparative experiments were carried out and the results were generated. The results show that the proposed strategies were reasonable and effective in dealing with the resource conflict problems in collaborative product development process.

In order to model the designer's autonomous real-time task scheduling behavior and use it in the agent-based product development process simulation, this paper constructed an agent's benefit function in which task urgency, task importance, agent individual preference, and recovery time are considered. To validate and show applications of the benefit function in agent-based product development process simulation, some comparative simulation experiments are carried out in a case study. The result shows that the simulation analysis based on the benefit function can help an organization identify reasonable scheduling behaviors in the collaborative and complex product development process.

Modern product customization aims at satisfying customers' individual requirements by providing customers various customization approaches. Recently, customers, especially lead users, are proven that they can play an active role in product innovation by using their potential user knowledge and creative skills. However, there is an obvious lack of systematic and deep research on customers' innovation or co-creation in current product customization systems. By customer integration and team collaborative creativity, a conceptual framework for collaborative customer innovation (CCI) is formulated in this paper, which proposes a new customer integrating approach beyond product customization. First, the concept and connotation of CCI are presented by comparing this emergent innovation approach with existing researches. Second, by studying the essential working principle of CCI, technology-driven knowledge spiral collaborative innovation mechanism and an agent-based team working model are formulated and described in detail. To realize customer collaborative innovation, the function structure of the innovation platform of CCI is analyzed and proposed. An innovation platform for household ceramic products is constructed as the application of CCI, and a case study is presented to validate the above principle and working model. The proposed conceptual framework for CCI can be used to build the collaborative customer innovation organization, develop innovation toolkits, and instruct enterprises to realize CCI effectively.