Michael Devetsikiotis's research while affiliated with University of New Mexico and other places

Charging station location decisions are a critical element in mainstream adoption of electric vehicles (EVs). The consumer confidence in EVs can be boosted with the deployment of carefully-planned charging infrastructure that can fuel a fair number of trips. The charging station (CS) location problem is complex and differs considerably from the classical facility location literature, as the decision parameters are additionally linked to a relatively longer charging period, battery parameters, and available grid resources. In this study, we propose a three-layered system model of fast charging stations (FCSs). In the first layer, we solve the flow capturing location problem to identify the locations of the charging stations. In the second layer, we use a queuing model and introduce a resource allocation framework to optimally provision the limited grid resources. In the third layer, we consider the battery charging dynamics and develop a station policy to maximize the profit by setting maximum charging levels. The model is evaluated on the Arizona state highway system and North Dakota state network with a gravity data model, and on the City of Raleigh, North Carolina, using real traffic data. The results show that the proposed hierarchical model improves the system performance, as well as the quality of service (QoS), provided to the customers. The proposed model can efficiently assist city planners for CS location selection and system design.

We consider a cluster of buildings within proximity that share a large-capacity battery for peak-shaving purposes, and draw power from the grid at a premium once they reach a certain threshold. Our goal is to identify a resource allocation policy that minimizes the amount of energy the cluster draws at a premium, while also ensuring fair access to all of its members. We introduce an adaptive policy that allows for maximum energy savings when the network load is low, and ensures fairness when the aggregate power level is high. We compare this adaptive policy with two standard resource allocation strategies with complementary advantages, and demonstrate through an extensive performance evaluation, that it combines the benefits of both. It is therefore suitable for a microgrid operator where equal weight is given to both cluster-wide cost minimization and fairness among all customers.

Motivated by the recent explosion of interest around blockchains, we examine whether they make a good fit for the Internet of Things (IoT) sector. Blockchains allow us to have a distributed peer-to-peer network where non-trusting members can interact with each other without a trusted intermediary, in a verifiable manner. We review how this mechanism works and also look into smart contracts-scripts that reside on the blockchain that allow for the automation of multi-step processes. We then move into the IoT domain, and describe how a blockchain-IoT combination: 1) facilitates the sharing of services and resources leading to the creation of a marketplace of services between devices and 2) allows us to automate in a cryptographically verifiable manner several existing, time-consuming workflows. We also point out certain issues that should be considered before the deployment of a blockchain network in an IoT setting: from transactional privacy to the expected value of the digitized assets traded on the network. Wherever applicable, we identify solutions and workarounds. Our conclusion is that the blockchain-IoT combination is powerful and can cause significant transformations across several industries, paving the way for new business models and novel, distributed applications.

The charging power of plug-in electric vehicles (PEVs) decreases significantly when the state of charge (SoC) gets closer to the fully charged state, which leads to a longer charging duration. Each time when the battery is charged at high rates, it incurs a significant degradation cost that shortens the battery life. Furthermore, the differences between demand preferences, battery types, and charging technologies make the operation of the charging stations a complex problem. Even though some of these issues have been addressed in the literature, the charging station modeling with battery models and different customer preferences have been neglected. To that end, this paper proposes two queueing-based optimization frameworks. In the first one, the goal is to maximize the system revenue for single class customers by limiting the requested SoC targets. The PEV cost function is composed of battery degradation cost, the waiting cost in the queue, and the admission fee. Under this framework, the charging station is modeled as a $M/G/S/K$ queue, and the system performance is assessed based on the numerical and simulation results. In the second framework, we describe an optimal revenue model for multi-class PEVs, building upon the approach utilized in the first framework. Two charging strategies are proposed: 1) a dedicated charger model and 2) a shared charger model for the multi-class PEVs. We evaluate and compare these strategies. Results show that the proposed frameworks improve both the station performance and quality of service provided to customers. The results show that the system revenue is more than doubled when compared with the baseline scenario which includes no limitations on the requested SoC.