Person:
AVISHAN, Farzad

Profile Picture

Email Address

Birth Date

WoSScopusGoogle ScholarORCID

Name

Job Title

First Name

Farzad

Last Name

AVISHAN
OrgUnit Logo
Organizational Unit

Filters

2022 - 20232
Reset filters

Settings

Publication Search Results

Now showing 1 - 2 of 2
  • Placeholder
    ArticlePublicationMetadata only
    Robust strategic planning of dynamic wireless charging infrastructure for electric buses
    (Elsevier, 2022-02-01) Alwesabi, Y.; Avishan, Farzad; Yanıkoğlu, İhsan; Liu, Z.; Wang, Y.; Industrial Engineering; AVISHAN, Farzad; YANIKOĞLU, Ihsan
    Electromobility in public bus systems is growing rapidly and experiencing a fundamental transformation in their infrastructure and operations. The dilemma of limited driving range and charging time of battery electric buses (BEBs) hinders their adoption. A novel approach to address BEB limitations is to utilize dynamic wireless charging (DWC) technology that allows buses to charge while in motion. This paper aims to analyze robust strategic planning of DWC and BEB fleet scheduling based on a real bus network at Binghamton University. The problem is first formulated as a new deterministic mixed-integer linear programming model to simultaneously optimize both the charging planning problem and fleet scheduling problem in an integrated fashion. To address the uncertainty of energy demand and charging time, a robust counterpart model (RCM) has been derived. To increase RCM flexibility, the battery status variable is formulated in a cumulative form. Dependent and independent budget uncertainty sets have been developed to control the robustness. A sensitivity analysis has been conducted to study the system behavior in response to different charging types, auxiliary energy demand, depth of discharge, charging options at terminals, battery degradation, and electricity cost. The deterministic model shows that eight homogeneous BEBs are required to operate on the selected routes with a battery capacity of 54.01 kWh and a total cost of $3,636,347. The results show that joint planning of charging infrastructure and fleet scheduling can save 19.2% of total cost compared to disjoint planning. The RCM results in 10 BEBs to ensure feasiblility against uncertainty.
  • Placeholder
    ArticlePublicationMetadata only
    Humanitarian relief distribution problem: an adjustable robust optimization approach
    (Informs, 2023-07) Avishan, Farzad; Elyasi, Milad; Yanıkoğlu, İhsan; Ekici, Ali; Özener, Okan Örsan; Industrial Engineering; AVISHAN, Farzad; YANIKOĞLU, Ihsan; EKİCİ, Ali; ÖZENER, Okan Örsan; Elyasi, Milad
    Management of humanitarian logistics operations is one of the most critical planning problems to be addressed immediately after a disaster. The response phase covers the first 12 hours after the disaster and is prone to uncertainties because of debris and gridlock traffic influencing the dispatching operations of relief logistics teams in the areas affected. Moreover, the teams have limited time and resources, and they must provide equitable distribution of supplies to affected people. This paper proposes an adjustable robust optimization approach for the associated humanitarian logistics problem. The approach creates routes for relief logistics teams and decides the service times of the visited sites to distribute relief supplies by taking the uncertainty in travel times into account. The associated model allows relief logistics teams to adjust their service decisions according to the revealed information during the process. Hence, our solutions are robust for the worst-case realization of travel times, but still more flexible and less conservative than those of static robust optimization. We propose novel reformulation techniques to model these adjustable decisions. The resulting models are computationally challenging optimization problems to be solved by exact methods, and, hence, we propose heuristic algorithms. The state-of-the-art heuristic, which is based on clustering and a dedicated decision-rule algorithm, yields near-optimal results for medium-sized instances and is scalable even for large-sized instances. We have also shown the effectiveness of our approach in a case study using a data set obtained from an earthquake that hit the Van province of Turkey in 2011.