A significant challenge of adopting battery electric buses into fleets lies in scheduling the charging, which in turn is complicated by considerations such as timing constraints imposed by routes, long charging times, limited numbers of chargers, and utility cost structures. This work builds on previous network-flow-based charge scheduling approaches and includes both consumption and demand time-of-use costs while accounting for uncontrolled loads on the same meter. Additionally, a variable-rate, non-linear partial charging model compatible with the mixed-integer linear program (MILP) is developed for increased charging fidelity. To respond to feedback in an uncertain environment, the resulting MILP is adapted to a hierarchical receding horizon planner that utilizes a static plan for the day as a reference to follow while reacting to stochasticity on a regular basis. This receding horizon planner is analyzed with Monte-Carlo techniques alongside two other possible planning methods. It is found to provide up to 52% cost savings compared to a non-time-of-use aware method and significant robustness benefits compared to an optimal open-loop method.

As transit authorities begin adopting Battery Electric Buses (BEBs), they must address the challenge of extended battery charging sessions while maintaining fixed bus schedules and managing the cost of energy. This paper proposes a novel technique for minimizing the monthly cost of energy for operating electric bus fleets in the presence of time-varying uncontrolled loads. The charging problem is cast as a constrained bin packing problem and expressed as a mixed integer linear program (MILP). Among other things, we show that the proposed method significantly decreases demand on power infrastructure by temporally balancing the loads from buses with uncontrolled loads.

A major challenge to adopting battery electric buses into bus fleets is the scheduling of the battery charging while considering route timing constraints and battery charge. This work develops a scheduling framework to balance the use of slow and fast chargers assuming the bus routes and charger locations are fixed. Slow chargers are utilized when possible to lower the cost of charging and fast chargers are used when needed to meet timing constraints and to ensure a sufficient charge for route execution. A directed graph is used to model the available charge times for buses that periodically return to the station to pick up passengers and to recharge its battery. A constrained network flow Mixed-Integer Linear Program (MILP) problem is formulated to optimize the scheduling of chargers as well as to determine the number of chargers required to meet battery state of charge thresholds. Using a randomly generated route schedule for thirty buses, results are presented that demonstrate the ability of the proposed method to find optimal charging plans while considering time-of-use (TOU) costs and allowing for fixed and variable numbers of chargers. These optimal charging plans reduce costs up to 20.9% compared to a thresholding-based plan and up to 12.3% compared to an optimal strategy that does not consider the TOU costs.

Formation control is a canonical problem in multi-agent systems as many multi-agent problems require agents to travel in coordination at some point during execution. This paper develops a method for coordinated moving formation control by building upon existing virtual structures approaches to define the relative vehicle positions and orientations and building upon clothoid-based motion planning to create the desired motion of the structure. The result is a coordinated formation control method that respects individual curvature constraints of each agent while allowing agents to track their desired positions within the formation with asymptotic convergence.

This paper presents a new method for injecting human inputs into mixed-initiative interactions between humans and robots. The method is based on a model-predictive control (MPC) formulation, which inevitably involves predicting the system (robot dynamics as well as human input) into the future. These predictions are complicated by the fact that the human is interacting with the robot, causing the prediction method itself to have an effect on future human inputs. We investigate and develop different prediction schemes, including fixed and variable horizon MPCs and human input estimators of different orders. Through a search-and-rescue-inspired human operator study, we arrive at the conclusion that the simplest prediction methods outperform the more complex ones, i.e., in this particular case, less is indeed more.

Snake robots are controlled by implementing gaits inspired from their biological counterparts. However, transitioning between these gaits often produces undesired oscillations which cause net movements that are difficult to predict. In this paper we present a framework for implementing gaits which will allow for smooth transitions. We also present a method to determine the optimal time for each module of the snake to switch between gaits in a decentralized fashion. This will allow for each module to participate in minimizing a cost by communicating with a set of modules in a local neighborhood. Both of these developments will help to maintain desired properties of the gaits during transition.

Model predictive control can be computationally intensive as it has to compute an optimal control trajectory at each time instant. As such, we present a method in which parametrized behaviors are introduced as a level of abstraction to give a finite representation to the control trajectory optimization. As these control laws can be designed to accomplish different tasks, the robot is able to use the presented framework to tune the parameters online to achieve desirable results. Moreover, we build on switch-time optimization techniques to allow the model predictive control framework to optimize over a series of given behaviors, allowing for an added level of adaptability. We illustrate the utility of the framework through the control of a nonholonomic mobile robot.

Receding horizon control strategies have proven effective in many control and robotic applications. These methods simulate the state a certain time horizon into the future to choose the optimal trajectory. However, in many cases, such as in mobile robot navigation, the selection of an appropriate time horizon is important as too long of a time horizon can amplify the detrimental effects caused by environmental uncertainties in the prediction of the future state while too of a short horizon will lead to reduced performance. In this paper we strike a balance between these two conflicting objectives by first introducing a receding horizon method for navigation founded on schema-based behaviors. We then suggest a method of adapting the time horizon by minimizing a cost function which balances the performance of the underlying control problem (which prefers longer horizons) with the performance of our state prediction (which prefers shorter time horizons). We illustrate the operation with an example which shows the usefulness of our navigation scheme with an adaptive time horizon.