Research On Heterogeneous Multi-skill Project Scheduling With Cooperative Effect

In recent years,enterprises gradually face the drop in productivity and the loss of skilled workers in many industries.In the past,they chose to fill the huge talent gap by importing outsiders simply.However,it becomes more difficult with the increase of labor cost and talent gap.So,the enterprises begin to pay attention to the real disposable employees.The existence of multi-skill is beneficial to reduce hiring costs,reduce the idle time of on-the-job employees and increase the flexibility of scheduling.It has become a source for the enterprises to obtain sustaining competitiveness.In the cruel and competitive market,it is of considerable realistic to ensure the quality of temporary project team while taking into account the long-term development of employees’ skills.In this context,this thesis studies the multi-skill project scheduling to optimize the team effectiveness and individual skills.Compared with other resources,project human share the characteristics:(1)heterogeneity: employees differ in the skill category/level and the learning/forgetting rate;(2)dynamic: the skill level is dynamic under the learning-forgetting effect;(3)social relations: there is a cooperative relationship among individual members.And,the positive and effective cooperation is an important guarantee for team effectiveness.Therefore,if we consider the feature of human resource,the MSPSP is far more complex than the RCPSP.It means the former has unconventional optimization goals and more constraints.Besides,they also put forward new challenges for data collection and quantification.At present,most research about MSPSP focus on the match between skills and tasks to optimize traditional goal-makespan and cost.Further,for the MSPSP considering the cooperative effect,employee heterogeneity and the dynamic of skill level,this research develop nonlinear integer programming model and solve it by the CPLEX and NSGA-II.For the small-scale MSPSP-CE with learning effect,based on the linearization and “ε-constraint method”,the model is solved by the commercial solver – CPLEX.For the medium/large-scale MSPSP-CE with learning-Forgetting effect,based on the classical NSGA-II,the thesis designs a new mechanism of encoding and decoding,adjusts the crossover and mutation operator,proposes the adjusted NSGA-II.At the end,verifies the effectiveness of the proposed model and algorithm by two real projects.The multi-skill project scheduling problem is the integration of human resource allocation and project scheduling.However,the relevant study is less at home and abroad.So,the research of this thesis can enrich the existing researches and has a strong theoretical value.In addition,the MSPSP-CE to optimize team effectiveness and individual skills conforms to the real needs of enterprises,and its results can provide specific proposals for managers directly.It also has a strong practical significance...