GEP Algorithm-based Optimization Research For Precast Components Production Scheduling Under Due Date Variation

As an important tool for the transformation and upgrading of China’s construction industry,prefabricated construction transfers the components needed for construction projects to off-site production through prefabrication technology.The production of precast components(PCs)usually adopts the production-by-order method,and the production arrangement is based on the due date and demand of the order of the construction part.Due to the complexity of the actual construction project,the on-site construction schedule is often uncertain and susceptible to changes in the actual delivery requirements of the precast components produced off-site due to various factors.Failure to properly consider due date variations may result in ineffective production planning,high inventory costs,delayed on-site assembly schedules,customer dissatisfaction and loss of market share.Therefore,it is of practical importance to consider dynamic production scheduling of prefabricated components in terms of delivery date variations.In addition,the gene expression programming(GEP)algorithm is widely used in a wide range of application situations,and in the manufacturing field,it can outperform empirical dispatching rules in constructing dispatching rules and is suitable for solving optimization of dynamic scheduling,but there is no application of the GEP algorithm in the precast components production scenario yet.This paper first describes the PCs production scheduling problem based on production scheduling theory,combined with the production method and process analysis of precast components for assembled buildings.And on this basis,through the method of discrete-time simulation,considering the constraints of working time,mold resources,buffer space and other constraints,the due dates variation generator is introduced to simulate the due date variation,and the specific dispatching rules are used as the judgment criteria of order priority in the production process,and the penalty cost is used as the target value to establish the due date variation response model.Secondly,the Gene Expression Programming algorithm(GEP)is applied to perform the target search.Based on the component production environment and order attribute characteristics,gene elements are defined.Gene elements are combined by K-expressions to generate dispatching rules,and by genetic operations such as selective replication,mutation,IS string insertion,and gene recombination,superior dispatching rules are constructed to achieve model optimization under the corresponding due date change scenarios.Finally,a single-case analysis,multi-scenario simulation analysis and performance trend analysis are performed by due date variation simulation with a case study of PCs production in a precast component plant.The results of the case application verify the effectiveness of the model,while the superiority of the dispatching rules constructed by the GEP algorithm is verified by comparing other dispatching rules under different degrees of due date variation scenarios.The model established in this paper can effectively reduce the penalty cost when the component plant faces the impact of due date change.The innovations in this article are described below: Compared with the existing research on the optimization of PCs production scheduling,the impact of due date variation is considered.The paper proposes a due date variation response model.Besides,by applying the GEP algorithm to achieve model optimization by constructing dynamic dispatching rules,which can cope with different due date variation scenarios.