Rolling Horizon Based Equipment Replacing With Finite Horizon

Businesses require equipments in order to function and deliver their outputs. In the global competitive environment, equipments are critical to success. However, equipments generally degrade with age and usage, and investment is required to maintain the functional performance of equipments. Making good equipment replacement policy is critical to labor productivity and economic benefits of enterprises. There are many fields concerning ERP, i.e. agricultural machinery, medical devices, manufacturing equipments and electric power assets. With the rapid development of modern global engineering economy, equipment replacement problem plays a more and more important role in modern industry. Many problems hard to solve existing in ERP attract more theoretical and practical researchers. ERP is stated as a class of combinatorial optimization in engineering economy.ERP is confronted with the computational complexity of large-scale replacing problems and the uncertainty of dynamic market information. Existing disaggregating methods to solve the computational complexity and uncertainty of ERP have such-and-such defects and malpractices. While rolling horizon procedure (RHP) is a versatile method to settle large scale combinatorial optimization problems. However, RHP has not been applied to equipment replacement problem, and this research direction is blank both at home and abroad.This dissertation investigates RHP in large-scale deterministic replacing problems in the context of job shop equipments replacement. The effectiveness and performance of RHP in equipment replacement problem under different circumstances are extensively verified and analyzed. Summarily, the main research work of this dissertation lies in five aspects as follows.(1) The principles and key techniques of RHP for equipment replacement are analyzed, using the successful applications of RHP in the fields of production schedule and path programming as references. The generic description of RHP is proposed formally. (2) RHP is applied to stationary ERP with multiple alternatives. Two types of rolling mechanisms as well as their respective algorithms are developed. Extensive experiments were conducted. Computational results demonstrate the effectiveness of RHP for ERP. General rules for selecting rolling mechanisms are provided.(3) Improvements are carried out on rolling horizon replacing method proposed. A terminal penalty function was added into the local objective of sub-problems to make it be consistent with the global one. The analytical and computational results demonstrate that this procedure can make a great improvement and the global solution is better than traditional rolling horizon procedure for ERP with moderate computational efforts.(4) A forward rolling procedure Rolling Planning Horizon Procedure (RPHP) for monotonic equipment replacement problem is proposed. The procedure adopts present planning horizon procedure and the frame of RHP to obtain the forecasting and planning horizons of rolling windows. Clench constraint relationship between present decisions and future information can be better solved. The procedure is carried out by Excel, which raises informationization of engineering management and policy.