Scalable Model Predictive Control for Autonomous Mobility-on-Demand Systems

Carron, Andrea; Seccamonte, Francesco; Ruch, Claudio; Zeilinger, Melanie N.

doi:10.1109/tcst.2019.2954520

Cited by 31 publications

(19 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The resulting real-time implementation is formally an ILP, which however reduces to a LP thanks to total unimodularity (93). In network flow models, rebalancing problems are often formulated as (integer) optimization problems (72,94,118), possibly considering competitive behaviors (119). For instance, recall the dynamic model in Equation 1..…”

Section: Review Of Methodsmentioning

confidence: 99%

“…For instance, in network flow models, we lose the identity of single AVs. Thus, the problems of dispatching, routing, and rebalancing AVs are implicitly solved when computing minimum cost flows, which are usually formulated as LPs and quadratic programs (QPs) for non-integer flows (72,94,122), and as ILPs and mixed integer linear programs (MILPs) for integer flows (120), possibly considering charging (123,124) and parking routines (69).…”

Section: Review Of Methodsmentioning

confidence: 99%

“…For instance, a queueing-theoretic model consists of N stations and m vehicles; thus, fleet size is implicitly specified by the model of the transportation network and AVs. Second, fleet size sometimes results from the initial condition of a dynamic model of the transportation system (35,94). For instance, (35) shows that, in the setting of Equation 1., the number of AVs is constant (i.e., d dt ∑ i vi(t) = 0).…”

Section: Operational Constraintsmentioning

confidence: 99%

See 2 more Smart Citations

Analysis and Control of Autonomous Mobility-on-Demand Systems

Zardini,

Lanzetti,

Pavone

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Challenged by urbanization and increasing travel needs, existing transportation systems call for new mobility paradigms. In this article, we present the emerging concept of Autonomous Mobility-on-Demand, whereby centrally orchestrated fleets of autonomous vehicles provide mobility service to customers. We provide a comprehensive review of methods and tools to model and solve problems related to Autonomous Mobility-on-Demand systems. Specifically, we first identify problem settings for their analysis and control, both from the operational and the planning perspective. We then review modeling aspects, including transportation networks, transportation demand, congestion, operational constraints, and interactions with existing infrastructure. Thereafter, we provide a systematic analysis of existing solution methods and performance metrics, highlighting trends and trade-offs. Finally, we present various directions for further research.

show abstract

Section: Review Of Methodsmentioning

confidence: 99%

Section: Review Of Methodsmentioning

confidence: 99%

Section: Operational Constraintsmentioning

confidence: 99%

See 1 more Smart Citation

Analysis and Control of Autonomous Mobility-on-Demand Systems

Zardini,

Lanzetti,

Pavone

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The authors of [20] propose a rebalancing algorithm using queuing Jackson networks. A scalable model predictive control (MPC) solution approach for fleet rebalancing on AMoD systems was provided in [21]. The authors in [22] describe a mathematical program of static user equilibrium for an MoD system, and its solution was used as the basis for a linear program built to rebalance empty vehicles.…”

Section: A Model-based Methodsmentioning

confidence: 99%

A Modular and Transferable Reinforcement Learning Framework for the Fleet Rebalancing Problem

Skordilis,

Hou,

Tripp

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…As we analyze the case of the 20 Italian regions that are heterogeneous in terms of the health care system, size, population, economy, and demography method should focus on how to design a predictive model with a strong capability of learning big data. Therefore we estimate the relationships between real data and the SIRCQTHE model's characteristic parameters employing some of the most known machine learning techniques [44], [45]. In particular, let us adopt an approach based on a least-squares optimization technique combined with constraints to enforce the prior scientific knowledge of the COVID-19 pandemic.…”

Section: Long-term Identification Of the Model Parametersmentioning

confidence: 99%

Stochastic Non-Pharmaceutical Optimal Control Strategies to Mitigate COVID-19

Scarabaggio¹,

Carli²,

Cavone³

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper proposes a stochastic non-linear model predictive controller to support policy-makers in determining robust optimal strategies to tackle the COVID-19 secondary waves. First, a time-varying <i>SIRCQTHE </i>epidemiological model (considering Susceptible, Infected, Removed, Contagious, Quarantined, Threatened, Healed, and Extinct compartments of individuals) is defined to get reliable predictions on the pandemic dynamics on a regional basis. A stochastic Model Predictive Control problem is then formulated to select the necessary control actions to minimize the arising socio-economic costs. <br>In particular, considering the unavoidable uncertainty characterizing this decision-making process, we ensure that the capacity of the network of regional healthcare systems is not violated in accordance with a chance constraint approach.<br>Furthermore, since the infection rate depends on people’s mobility, differently from the related literature, we model the control actions as interventions affecting the mobility levels associated to different socio-economic categories.<br><div>The effectiveness of the presented method in properly supporting the definition of diversified regional strategies for tackling the COVID-19 spread is tested on the network of Italian regions using real data from the Italian Civil Protection Department. However, provided the availability of reliable data, the proposed approach can be easily extended to cope with other countries' characteristics and different levels of the spatial scale.</div><div><br></div><div>Preprint of paper submitted to IEEE Transactions on Automation Science and Engineering (<em>T-ASE</em>)</div>

show abstract

Scalable Model Predictive Control for Autonomous Mobility-on-Demand Systems

Cited by 31 publications

References 22 publications

Analysis and Control of Autonomous Mobility-on-Demand Systems

Analysis and Control of Autonomous Mobility-on-Demand Systems

A Modular and Transferable Reinforcement Learning Framework for the Fleet Rebalancing Problem

Stochastic Non-Pharmaceutical Optimal Control Strategies to Mitigate COVID-19

Contact Info

Product

Resources

About