یک فرمول مسئله جریان کارگاهی جهت برنامه ریزی عملیات حمل و نقل زیست توده
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|11795||2013||8 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Computers and Electronics in Agriculture, Volume 91, February 2013, Pages 49–56
Currently, the planning of the biomass collection operations is performed relying on the contractor’s experience without the use of any dedicated planning tool. It is expected that more explicitly formulated planning tools would provide benefits on securing the supply chain of biomass feedstock. In this study, the problem of finding a permutation schedule for a number of geographically dispersed fields where biomass handling operations have to be carried out involving a number of sequential tasks, was formulated as a flow shop with sequence depended set up times scheduling problems as known in the industrial domain. By applying the approach to a case study involving experimental recording of the operations executions, the execution of an optimal schedule was found to give a reduction of 9.8% in the total time as compared to a schedule based on the tacit knowledge of the operations manager.
Traditionally, in biomass production systems the developed engineering scheduling approaches deals mainly with the seasonal planning of field operations (e.g. Sørensen, 1999). This type of pure scheduling problems involves assigning labour and machinery recourses to the operations and assigning the operations to periods in time. The first scientific approaches for supporting this decision process in agriculture appeared in early 1980s (e.g. van Elderen, 1980 and Fokkens and Puylaert, 1981), mainly based on linear programming methodologies. Wijngaard (1988) implemented and compared three different models for farm operations management namely, a dynamic programming model, a linear programming model, and a simulation model. Recent approaches for a long term cropping schedule involve other methodologies such as stochastic programming (Darby-Dowman et al., 2000), hybrid petri nets (Guan et al., 2008) and metaheuristics (e.g. simulated annealing, and genetic algorithms) (Guan et al., 2009). The above mentioned scheduling type belongs to the operational planning level of the farm operations management system (Sørensen et al., 2010). Another type of scheduling in biomass production systems that also belongs to the operational planning level is the sequencing of tasks that compete for shared resources. An example of this scheduling type is the planning of large-scale biomass harvesting and handling operations where a series of sequential operations (e.g. cutting, raking, baling, and loading) have to be performed at a number of geographically dispersed fields. This planning task is becoming particularly important as an integral part of the biomass collection for bioenergy production use since securing the supply chain of biomass feedstock will increase the demand for advanced agricultural fleet management tools. In most cases, the planning of the biomass collection operations is performed relying on the contractor’s experience without the use of any dedicated planning tool (Sørensen and Bochtis, 2010). Currently, the research efforts within this planning domain are limited. As an example, Basnet et al. (2006) introduced a scheduling method for harvesting of renewable resources based on a Travelling Salesman Problem (TSP) approach combined with greedy and tabu search heuristics. Bochtis and Sørensen (2010) showed that scheduling and planning problems for agricultural field operations can be cast as vehicle routing problems with time windows (VRPTWs) instances and, consequently, it is possible to apply advanced methods developed specifically for the solution of these instances. However, the development of a system that can be implemented to real-life biomass collection planning has to be combined with tools that can precisely predict the time requirements of all tasks involved in the related operations. This is also important in terms of reliability of the biomass supply chain which takes place in a stochastic environment. However, the difficulty in dealing with stochastic measures like availability and reliability concerns the practical implementation of these measures in terms of quantification and task times prediction (Sørensen, 1999). In this paper, an industrial engineering approach as regard the problem of scheduling of sequential biomass handling operations is presented. The particular objectives are: (1) The formulation of the planning problem of multiple-fields sequential operations as a flow shop with set up times as part of the scheduling problem. (2) The demonstration of the approach implementation using a real-life small scale example. (3) To assess the impact of the uncertainty in the task times prediction on the applicability of the pursued approach.
نتیجه گیری انگلیسی
A novel approach regarding the problem of scheduling sequential biomass handling operations was presented. It was shown that this scheduling problem for biomass handling is equivalent to the flow shop with sequence dependent set up times known from scheduling problems in the industrial domain. By applying the approach on a case study involving experimental recording of the operations executions, there was a reduction of 9.8% in the total operations time when the optimal schedule were executed instead of the one based on the tacit knowledge and experience of the operations manager. The assessment of the impact of the uncertainty in the task times prediction on the applicability of the approach showed that appropriate models that simulate in detail the involved in-field activities have to be developed in order to secure a sufficient stable scheduling solution. As far as it concerns the scheduling itself, a number of future advances could be included: • To incorporate the case of multiple-machinery systems, namely the case where an operation type is carried out by more than one (not necessarily identical) units. • The case where the operations have to be interrupted during the night time where machinery has to return the depot. • To take the transportation time cycle into account. • The temporal constraints that biomass handling operations subject should also be considered. These temporal constraints regard minimum and maximum time lags between starting and completion times of any two sequential operations. The duration of these time lags is driven by the interaction of the physical properties of the cut biomass and the prevailing local weather conditions and consequently, dedicated prediction systems are also needed. The reliability of the biomass supply chain which takes place in a stochastic environment is also an important aspect. However, the difficulty in dealing with stochastic measures like availability and reliability concerns the practical implementation of these measures in terms of quantification and task times prediction. A precise task times prediction can not only be based on off-line estimation using average norms (like the approach presented here) but should preferable include models that are adaptable in order to update the model recorded parameters so they precisely reflect the conditions on the specific field based on automatic data collection from historical operations.