Research

year 2013 
author Jun Young Moon 
Keyword 생산 일정 계획, 에너지 효율성, 스마트 그리드, 수요 반응, 분산 전원, 에너지 스토리지, 유전 알고리즘, 제약 프로그래밍 
Abstract Currently, manufacturing industries in many advanced countries pay electricity rates stratified by the time of the day (i.e., peak load, mid-load, and off-peak load). These countries are trying to widen the difference between the peak load rate and off-peak load rate to reduce electricity demand at peak load times. Differential electricity pricing affords consumers an opportunity to shift the timing of electricity usage to time periods with a lower electricity price, thereby reducing electricity costs.
Similarly, a production scheduling process, which considers time-dependent and machine-dependent electricity costs, enables industrial electricity consumers to minimize energy expenses.
Additionally, the emerging Smart Grid is supposed to require industries to pay real-time hourly electricity costs. More energy-efficient, intelligent production scheduling is thus possible.
The present thesis deals with minimizing the total cost of production schedule by means of energy scheduling considering both the usage and generation of electricity. The proposed method allows each decision maker in the manufacturing industry to seek a compromise solution with total production cost by considering electricity costs with distributed energy resources and energy storage in the Smart Grid environment of the future.
The method employs hybrid algorithms that include genetic algorithm, constraint programming, and mixed-integer linear programming approaches to solve the problems and compare the proposed models with the classical computational method.
In addition, smart production scheduling models are formulated considering distributed generations (DGs), including energy storage system and renewable energy resources such as solar, wind, and fuel cell. This issue is more complex as manufacturers have to consider an hourly generation schedule of DGs and hourly charging/discharging schedule of energy storage system. In order to solve this problem, an algorithm and solution technique using mixed integer programming and constraint programming is introduced.
Finally, this thesis deals with multi-factory smart production scheduling problems by grouping factories. One model considering separated DGs and another considering centralized DGs are analyzed and compared. This can be a guideline to determine the setup of DGs for manufacturing companies with a number of factories.
The models considering time-dependent and machine-dependent electricity costs were proposed for efficient use of electricity in factories. However, processing times for each job in the suggested models were assumed to be integer-multiples. Therefore, improvements are still needed for more realistic models applicable to the field. 
c PhD 

Download :