Operations Research - Master's degree
http://hdl.handle.net/1721.1/7720
2015-12-22T11:32:01ZOptimization of yard operations in maritime container terminals
http://hdl.handle.net/1721.1/99589
Optimization of yard operations in maritime container terminals
Borjian Boroujeni, Setareh
With the continuous growth in international container shipping, many container terminals in maritime ports face congestion, particularly during peak hours of service, and when there is limited space in the storage area. Thus, there has been increasing interest in improving operations efficiency in container terminals. An efficient terminal, in general, is one that discharges containers from the ships in a timely manner and delivers containers to customers with a reasonable wait time. Moreover, a key performance measure in the storage area is the number of moves performed by yard cranes. Due to limited space in the storage area, containers are stacked on top of each other, forming a column of containers that can be accessed by yard cranes only from the top. Therefore, in order to retrieve a container that is covered by other containers, the blocking containers must be relocated to other slots. Because such relocation moves are costly for the port operators and result in service delays, one of the main challenges in the storage area is to plan the moves such that the number of relocations is minimized. This problem is referred to as the Container Relocation Problem (CRP). The CRP in its most simplified setting is concerned with finding a sequence of moves that retrieves all containers in a pre-defined order with a minimum number of relocations, assuming that no new containers are stacked during the retrieval process. Also, it is often assumed that the non-blocking containers cannot be relocated (i.e., repositioning moves are not allowed), an assumption that can result in a sub-optimal solution. Other variants of the container relocation problem include the dynamic CRP and the CRP with incomplete information. The former involves minimizing the number of relocations when containers are continuously stacked in and retrieved from the storage area, and the latter refers to the case that the departure times of containers are not fully known in advance. For example, a probabilistic distribution of container departure orders, or approximate departure times (in the form of time windows) might be known. Another important efficiency metric, in addition to the number of relocations, is customer wait times during the retrieval process. In particular, when repositioning moves are allowed in the system, there is a trade-off between the total number of relocations (including repositionings) and wait times, because such repositioning moves make the retrieval process faster for trucks arriving in the future. Also, it might be desired to prioritize some customers so that those prioritized experience shorter wait times. For example, in terminals with appointment systems, shorter waiting time guarantees can be given to customers who book in advance a time slot for picking up their containers. In this thesis, we propose optimization models that capture service-based and cost-based objectives and study different service policies. In the first part of this thesis, we study the CRP with complete information using an optimization model and heuristic approach. In particular, we formulate CRP (with no restrictive assumptions on repositioning moves) as an Integer Program that minimizes the weighted sum of the number of relocations and the total wait time of customers. Our integer program provides the optimal sequence of moves for retrieving containers subject to various service policies. For example, it can be used by port operators to minimize customer wait times, or to give different waiting time guarantees to different customers to reflect relative priorities. Moreover, by assigning different weight factors to the two objectives, one can use our model to plan repositioning moves. We also extend our model to the dynamic CRP and illustrate how the flexibility in the stacking process can be exploited to optimize jointly the sequence of moves and the stacking position of containers. Additionally, we propose a class of flexible retrieval policies. We demonstrate that flexible policies can result in fewer relocations and shorter wait times, thereby benefiting both the port operators and customers. In the second part of this thesis, we study the CRP with incomplete information in a 2-stage setting where the departure times of a subset of containers are initially known and the departure times of other containers are revealed at once at a later time. The contributions are twofold. First, we propose an approximate stochastic optimization algorithm, called ASA*, which is a branch-and-bound framework combined with a sampling technique, and to the best of our knowledge is the first optimization algorithm proposed for this problem. We provide theoretical bounds on the approximation errors and present numerical results showing the computational tractability and efficiency of our algorithm. Second, we use the ASA* algorithm and a myopic heuristic to study the value of information, that is, the effect of the number of containers initially known on the number of relocations. In the last part of this thesis, we introduce a simulator that is capable of integrated simulation of port operations, including the retrieval process, the stacking process, and other aspects such as allocating cranes to containers and allocating trucks to cranes. Our simulator captures the practical details of operations that cannot be modelled in an optimization framework and is capable of simulating long periods (e.g. a week) of realistic-scale operations.
Thesis: S.M. in Transportation, Massachusetts Institute of Technology, Department of Civil and Environmental Engineering, 2015.; Thesis: S.M., Massachusetts Institute of Technology, Sloan School of Management, Operations Research Center, 2015.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 107-109).
2015-01-01T00:00:00ZAnomaly detection methods for unmanned underwater vehicle performance data
http://hdl.handle.net/1721.1/98718
Anomaly detection methods for unmanned underwater vehicle performance data
Harris, William Ray
This thesis considers the problem of detecting anomalies in performance data for unmanned underwater vehicles(UUVs). UUVs collect a tremendous amount of data, which operators are required to analyze between missions to determine if vehicle systems are functioning properly. Operators are typically under heavy time constraints when performing this data analysis. The goal of this research is to provide operators with a post-mission data analysis tool that automatically identifies anomalous features of performance data. Such anomalies are of interest because they are often the result of an abnormal condition that may prevent the vehicle from performing its programmed mission. In this thesis, we consider existing one-class classification anomaly detection techniques since labeled training data from the anomalous class is not readily available. Specifically, we focus on two anomaly detection techniques: (1) Kernel Density Estimation (KDE) Anomaly Detection and (2) Local Outlier Factor. Results are presented for selected UUV systems and data features, and initial findings provide insight into the effectiveness of these algorithms. Lastly, we explore ways to extend our KDE anomaly detection algorithm for various tasks, such as finding anomalies in discrete data and identifying anomalous trends in time-series data.
Thesis: S.M., Massachusetts Institute of Technology, Sloan School of Management, Operations Research Center, 2015.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 101-102).
2015-01-01T00:00:00ZModeling and design of material recovery facilities : genetic algorithm approach
http://hdl.handle.net/1721.1/98715
Modeling and design of material recovery facilities : genetic algorithm approach
Testa, Mariapaola
In the Organisation for Economic Co-operation and Development (OECD) area, the production of numerical solid waste (MI\SW) increased by 32% between 1990 and 2011, exceeding 660 million tonnes in 2011; the world-wide production of waste is estimated to grow further due to increasing GDP in developing economies. Given this scenario, effective treatment and recovery of wastes becomes a priority. In developed countries, MSW is usually sent to materials recovery facilities (MRFs), which use mechanical and manual sorting units to extract valuable components. In this work, we define a network flow model to represent a MRF that sorts wastes using multi-output units with recirculating streams. For each material in the system, we define a matrix to describe the sorting process. We then formulate a genetic algorithm (GA) that generates alternative configurations of a MRF having a given set of sorting units with known separation parameters and selects those with highest profit and efficiency. The GA incorporates a heuristic for personnel allocation to manual units. We code the algorithm in Java and apply it to an existing MRF. The results show a 33.4% improvement in profit and a 1.7% improvement in efficiency with respect to the current configuration without hand sorting; and a 6.7% improvement in profit and a 3.9% improvement il efficiency, with respect to the current configuration with hand sorting.
Thesis: S.M., Massachusetts Institute of Technology, Sloan School of Management, Operations Research Center, 2015.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 187-193).
2015-01-01T00:00:00ZDelay characterization and prediction in major U.S. airline networks
http://hdl.handle.net/1721.1/98567
Delay characterization and prediction in major U.S. airline networks
Hanley, Zebulon James
This thesis expands on models that predict delays within the National Airspace System (NAS) in the United States. We propose a new method to predict the expected behavior of the NAS throughout the course of an entire day after only a few flying hours have elapsed. We do so by using k-means clustering to classify the daily NAS behavior into a small set of most commonly seen snapshots. We then use random forests to map the delay behavior experienced early in a day to the most similar NAS snapshot, from which we make our type-of-day prediction for the NAS. By noon EST, we are able to predict the NAS type-of-day with 85% accuracy. We then incorporate these NAS type-of-day predictions into previously proposed models to predict the delay on specific origin-destination (OD) pairs within the U.S. at a certain number of hours into the future. The predictions use local delay variables, such as the current delay on specific OD pairs and airports, as well network-level variables such as the NAS type-of-day. These OD pair delay prediction models use random forests to make classification and regression predictions. The effects of changes in classification threshold, prediction horizon, NAS type-of-day inclusion, and using wheel off/on, actual, and scheduled gate departure and arrival times are studied. Lastly, we explore how the delay behavior of the NAS has changed over the last ten years and how well the models perform on new data.
Thesis: S.M., Massachusetts Institute of Technology, Sloan School of Management, Operations Research Center, 2015.; This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.; Cataloged from student-submitted PDF version of thesis.; Includes bibliographical references (pages 101-102).
2015-01-01T00:00:00Z