Filter-Embedded UAV Task Assignment Algorithms for Dynamic Environments

2013

J Intell Robot Syst

Planning for multi-agent systems such as task assignment for teams of limited-fuel unmanned aerial vehicles (UAVs) is challenging due to uncertainties in the assumed models and the very large size of the planning space. Researchers have developed fast cooperative planners based on simple models (e.g., linear and deterministic dynamics), yet inaccuracies in assumed models will impact the resulting performance. Learning techniques are capable of adapting the model and providing better policies asymptotically compared to cooperative planners, yet they often violate the safety conditions of the system due to their exploratory nature. Moreover they frequently require an impractically large number of interactions to perform well. This paper introduces the intelligent Cooperative Control Architecture (iCCA) as a framework for combining cooperative planners and reinforcement learning techniques. iCCA improves the policy of the cooperative planner, while reduces the risk and sample complexity of the learner. Empirical results in gridworld and task assignment for fuel-limited UAV domains with problem sizes up to 9 billion state-action pairs verify the advantage of iCCA over pure learning and planning strategies.

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Intelligent Cooperative Control Architecture: A Framework for Performance Improvement Using Safe Learning

Geramifard

Redding

2013

J Intell Robot Syst

“…Thus, this failure mode increases the rate at which fuel is burned (and thus decreases the time a vehicle can complete tasks). The following section illustrates how the Receding Horizon Task Assignment (RHTA) algorithm (defined in [24]) can be modified to include the vehicle's fuel state and how this can improve the algorithm's performance in situations where vehicles can experience failures, need refueling or require repair. The main idea -to consider the health data as information that constrains the possible list of activities each UAV can perform -can be extended to many different types of health data (engine performance, sensor performance, etc) and tasking algorithms.…”

Section: Task Assignment With Health Constraintsmentioning

confidence: 99%

“…This was accomplished by including an estimate of each vehicle's operational radius [24], which is defined here as…”

Section: Task Assignment With Health Constraintsmentioning

confidence: 99%

“…This information can be used to effectively prune the list of petals that RHTA considers in order to ensure that the vehicle can always safely return to base before its fuel is exhausted. More specifically, the following pruning criterion was added to the RHTA algorithm (Algorithm 2.3.1 in [24]): For every petal under consideration, reject the petal if…”

Section: Task Assignment With Health Constraintsmentioning

confidence: 99%

See 1 more Smart Citation

Embedding Health Management into Mission Tasking for UAV Teams

Valenti

Bethke

2007 American Control Conference

et al. 2007

Abstract-Coordinated multi-vehicle autonomous systems can provide incredible functionality, but off-nominal conditions and degraded system components can render this capability ineffective. This paper presents techniques to improve missionlevel functional reliability through better system self-awareness and adaptive mission planning. In particular, we extend the traditional definition of health management, which has historically referred to the process of actively monitoring and managing vehicle sub-systems (e.g., avionics) in the event of component failures, to the context of multiple vehicle operations and autonomous multi-agent teams. In this case, health management information about each mission system component is used to improve the mission system's self-awareness and adapt vehicle, guidance, task and mission plans. This paper presents the theoretical foundations of our approach and recent experimental results on a new UAV testbed.

A robust approach to the UAV task assignment problem

Alighanbari

Intl J Robust & Nonlinear

2007

SUMMARYThis paper presents a new robust approach to the task assignment of unmanned aerial vehicles (UAVs) operating in uncertain dynamic environments for which the optimization data, such as target cost and target-UAV distances, are time varying and uncertain. The impact of this uncertainty in the data is mitigated by tightly integrating two approaches for improving the robustness of the assignment algorithm. One approach is to design task assignment plans that are robust to the uncertainty in the data, which reduces the sensitivity to errors in the situational awareness (SA), but can be overly conservative for long duration plans. A second approach is to replan as the SA is updated, which results in the best plan given the current information, but can lead to a churning type of instability if the updates are performed too rapidly. The strategy proposed in this paper combines robust planning with the techniques developed to eliminate churning. This combination results in the robust filter-embedded task assignment algorithm that uses both proactive techniques that hedge against the uncertainty, and reactive approaches that limit churning behavior by the vehicles. Numerous simulations are shown to demonstrate the performance benefits of this new algorithm.