Back to search

Technical University of Crete (TUC)

Βέλτιστη διαχείριση ενέργειας σε λιμάνια με περιορισμό των εκπομπών αερίων του θερμοκηπίου

Abstract

dc:description

The amount of electrical energy required by ports is closely associated with adjustable loads such as refrigerated containers (reefers), plug-in electric vehicles (PEVs), and electric shore power supplies (SPS) (cold ironing) for ships at berth. At the same time, ports present significant opportunities for energy generation from renewable energy sources (RES). However, controlling these diverse adaptable loads is challenging with conventional control systems. Multi-agent systems (MAS) have proven to be highly efficient in managing complex systems and as demonstrated in this paper, can also be effective for port management applications. This research proposes a distributed demand-response system using MAS, implemented in the Java Agent DEvelopment Framework (JADE), to manage power demand for adjustable loads in ports. The ASEME Methodology, which has been successfully employed in the smart grid application domain, was followed for the MAS analysis and design. Two independent fuzzy logic-based systems are integrated into the MAS to minimize the operating costs of reefers and PEVs, respectively. In addition, ships at berth, operating in parallel with SPS and acting as smart prosumers, can effectively optimize operational costs while maintaining emissions within the prescribed limits. The efficiency of the proposed method is evaluated by detailed simulations of a port. Results demonstrate, in the specific systems and operating scenarios, a 38.54% reduction in the total energy cost per kWh for reefers and PEVs. A further 17.68% reduction in overall port operational costs is achieved when including ships at berth, with the reduction limited to 15.92% or 9.32% when emission control is applied per ship, or per port, respectively. This thesis also presents a real–time simulation of multi-agent systems for decentralized voltage control in distribution network. The multi-agent approach is used to share the voltage regulation effort and perform the coordination of distributed loads and generators with flexibility and reliability in distribution network. In simulations, the MAS with local voltage support keeps voltages within ±5% while preserving near-optimal cost and runtime suitable for real-time control.Nowadays, power consumption at container terminals is related to flexible loads. Arguably, the most significant flexible loads are the refrigerated containers, termed reefers. On average, reefers are responsible for 47.87% of the total energy consumption of container terminals. To control the reefers in a container terminal, it is hard to succeed with conventional methods. The multi-agent systems (MAS) are well suited to control complex systems. In this thesis, it is shown that MAS can be particularly well-suited to control reefers in a container terminal. It is demonstrated using real-world data, that a MAS system, implemented in JADE, is suitable for active power management of reefers. Additionally, it is used a fuzzy logic system, integrated in the MAS structure, to estimate the coefficient of utilization of the reefers. The results of the simulations with real data of a large terminal proved the effectiveness of the method. Ιn the specific systems and operating scenarios, a reduction of 37.08% to total energy cost of reefers at a container terminal was achieved.Finally, this thesis introduces an innovative, emission-aware optimal operation scheduling method for ports, designed to manage power demand across distributed flexible electric loads. The proposed approach simultaneously addresses multiple objectives: limiting in-port emissions, minimizing operational costs, optimizing port load time-shifting, and satisfying the operational constraints of numerous system components. To date, few studies have treated ships as smart prosumers capable of effectively mitigating all harmful emissions. To solve this highly complex problem, is employed Multi-Agent Systems (MAS), implemented in JADE. MAS was selected due to the complex nature of the power system under study and the need for a fast and reliable solution. Each agent solves a local optimization problem very quickly and robustly, enabling the overall system to reach a globally optimal solution. The proposed method was rigorously evaluated through detailed simulations in a case study involving a port with hundreds of reefers, plug-in electric vehicles (PEVs), and two shore power supplies. The results confirm the method’s effectiveness, achieving optimal outcomes, while fully meeting all optimization goals and operational constraints. The proposed methodology ensures rapid and reliable convergence of the optimization algorithm in one round and exhibits enhanced performance relative to existing approaches.

Degree

thesis:*
Grantor dc:publisher
Technical University of Crete (TUC)
Year dc:date
2026

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Ntountounakis, Emmanouil
  • Ντουντουνάκης, Εμμανουήλ

Subjects

dc:subject × 40

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/60924

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Ntountounakis, Emmanouil; Ντουντουνάκης, Εμμανουήλ. Βέλτιστη διαχείριση ενέργειας σε λιμάνια με περιορισμό των εκπομπών αερίων του θερμοκηπίου. Technical University of Crete (TUC), 2026. http://hdl.handle.net/10442/hedi/60924