Stochastic Model Predictive Control for Autonomous Mobility on Demand

TL;DR

This paper introduces a stochastic MPC algorithm for autonomous mobility-on-demand systems that improves dispatching and rebalancing efficiency by leveraging probabilistic forecasts, with demonstrated significant reductions in customer wait times.

Contribution

The paper develops a scalable stochastic MPC framework with performance guarantees, separating dispatch and rebalancing tasks for large-scale autonomous vehicle systems.

Findings

62.3% reduction in customer waiting time in simulations

Outperforms prior state-of-the-art algorithms

Scalable solution using linear programming relaxations

Abstract

This paper presents a stochastic, model predictive control (MPC) algorithm that leverages short-term probabilistic forecasts for dispatching and rebalancing Autonomous Mobility-on-Demand systems (AMoD, i.e. fleets of self-driving vehicles). We first present the core stochastic optimization problem in terms of a time-expanded network flow model. Then, to ameliorate its tractability, we present two key relaxations. First, we replace the original stochastic problem with a Sample Average Approximation (SAA), and characterize the performance guarantees. Second, we separate the controller into two separate parts to address the task of assigning vehicles to the outstanding customers separate from that of rebalancing. This enables the problem to be solved as two totally unimodular linear programs, and thus easily scalable to large problem sizes. Finally, we test the proposed algorithm in two scenarios based on real data and show that it outperforms prior state-of-the-art algorithms. In particular, in a simulation using customer data from DiDi Chuxing, the algorithm presented here exhibits a 62.3 percent reduction in customer waiting time compared to state of the art non-stochastic algorithms.