Chelsea Sabo

A Formulation and Heuristic Approach to Task Allocation and Routing of UAVs under Limited Communication

Unmanned Air Vehicle (UAV) teams are anticipated to provide surveillance support through algorithms, software, and automation. It is desirable to have algorithms that compute effective and efficient routes for multiple UAVs across a variety of missions. These algorithms must be realizable, practical, and account for uncertainties. In surveillance missions, UAVs act as mobile wireless communication nodes in a larger, underlying network consisting of targets where information is to be collected and base stations where information is to be delivered. The role of UAVs in these networks has primarily been to maintain or improve connectivity while undervaluing routing efficiency. Moreover, many current routing strategies for UAVs ignore communication constraints even though neglecting communication can lead to suboptimal tour designs. Generating algorithms for autonomous vehicles that work effectively despite these communication restrictions is key for the future of UAV surveillance missions. A solution is offered here based on a variation of the traditional vehicle routing problem and a simple communication model. In this work, the new routing formulation is defined, analyzed, and a heuristic approach is motivated and described. Simulation results show that the heuristic algorithm gives near-optimal results in real time, allowing it to be used for large problem sizes and extended to dynamic scenarios.

Fuzzy logic unmanned air vehicle motion planning

There are a variety of scenarios in which the mission objectives rely on an unmanned aerial vehicle (UAV) being capable of maneuvering in an environment containing obstacles in which there is little prior knowledge of the surroundings. With an appropriate dynamic motion planning algorithm, UAVs would be able tomaneuver in any unknown environment towards a target in real time. This paper presents a methodology for two-dimensional motion planning of a UAV using fuzzy logic. The fuzzy inference system takes information in real time about obstacles (if within the agents sensing range) and target location and outputs a change in heading angle and speed. The FL controller was validated, and Monte Carlo testing was completed to evaluate the performance. Not only was the path traversed by the UAV often the exact path computed using an optimal method, the low failure rate makes the fuzzy logic controller (FLC) feasible for exploration. The FLC showed only a total of 3% failure rate, whereas an artificial potential field (APF) solution, a commonly used intelligent control method, had an average of 18% failure rate. These results highlighted one of the advantages of the FLC method: its adaptability to complex scenarios while maintaining low control effort.

SMART Heuristic for Pickup and Delivery Problem (PDP) with Cooperative UAVs

Pickup and Delivery Problems (PDPs) are a subset of Vehicle Routing Problems (VRPs) which require a vehicle to service targets by picking them up at an origin and delivering them to their unique destination. With respect to surveillance functions, this becomes a realistic problem as UAVs are restricted by operating range, data rate, Anti-Jam margins, and cost. Therefore, UAVs must be allocated to “pickup” targets and then “deliver” them, from within a prescribed communication space, back to a command and control HQ. To maximize the speed/amount of information transmitted from this communication region, the objective of allocating the UAVs is such that the total service time (pickup and delivery) of all the targets is minimized. Previous work on PDPs has shown that as the problem gets more complicated (i.e. more targets and more vehicles) the solution space increases exponentially, and the execution time to find an optimal solution is impossible to implement. Additionally, previously related work using a heuristic solution has been applied to this problem showing that good result can be maintained (within ~1520% of the optimal solution for small cases). The focus of this research is to develop an alternate heuristic algorithm, deemed SMART from here on, that can perform near optimally (within ~5%) and scales as the problem gets more complicated. Also, the algorithm is developed such that it is easily extendable to a dynamic scenario as this research progresses. This algorithm is described in detail and has shown that it reaches this performance metric while requiring a significantly reduced computational time when compared to the time needed to obtain the optimal solution.

Minimum Service Time for UAV Cooperative Control Subject to Communication Constraints

The allocation of multiple Unmanned Air Vehicles (UAVs) under communication constraints for a minimum service time is the main focus of this paper. Previous work on task allocation of multiple UAVs has shown that obtaining the optimal policy for relevant problems can be computationally overwhelming as the number of states and controls become very large. Allowing communication only within a limited area adds an operationally realistic and complicated constraint to the problem. It is shown that generating the optimal solution for even a simplified version of the problem becomes computationally overpowering quickly. A solution based on heuristics for target clustering and assignment of vehicles to targets is shown here and is based on study of the optimal policy. A limited look-ahead method is then used to approximate the dynamic programming solution for one vehicle with a particular target assignment. It is shown that good results can be obtained much faster. In addition to this, it is feasible to obtain solutions with much higher complexity (i.e. more targets and vehicles).

Effectiveness of 2D Path Planning in Real Time using Fuzzy Logic

The effectiveness of 2D path planning of a UAV using fuzzy logic for the purpose of decision making in real time is explored in this paper. Previous work has shown that by using a fuzzy inference system, an agent can navigate an unknown environment. It does this by taking information about obstacles (if within the agent’s sensing range) and target location, and outputting a change in heading angle and speed. Often there are scenarios in which it is desirable for an aircraft to redirect its path midflight. These situations can involve threats, changing mission objectives, and/ or be very complex; with multiple and moving obstacles. A system that can handle these varying conditions rapidly and efficiently is imperative to the future of autonomous aircraft. A fuzzy logic approach is used here for its ability to imitate human heuristics and simplicity to implement. The effectiveness of this methodology is analyzed by comparing it to an optimal path planning approach. While the optimal path will give either the shortest path (or time to a target), control algorithms are incapable of being re-tasked. Presented here is a fuzzy inference system (FIS) for path planning with obstacle avoidance. This FIS has been tweaked and tested for robustness by comparing it to other 2D path planning methods on numerous obstacle environments.

Path Planning of a Fire-Fighting Aircraft using Fuzzy Logic

The effectiveness of path planning of a fire-fighting aircraft using fuzzy logic for the purpose of fighting forest fires is explored in this paper. Using previous work on path tracking and obstacle avoidance with fuzzy logic as a starting point, the effort has been expanded to include logic for continuously changing target location and verification of results for moving obstacles. When combating wildfires, situations can change rapidly due to fluctuating environmental conditions (wind, fuel, terrain, etc) that can affect the target location. A system that can handle these varying conditions rapidly and efficiently is imperative to these situations. Presented here is a fuzzy inference system that takes information about obstacles (if within the agent’s sensing range) and target location and outputs a change in heading angle and speed accordingly. The agent’s objective is to take the shortest path to the target area while also avoiding obstacles. These obstacles could be mountains, no-fly zones, areas in which it is too dangerous to fly, or other agents. Presented here is a path planning with obstacle avoidance fuzzy inference system and verification on a simplified fire growth model.

Dynamic Allocation of Unmanned Aerial Vehicles with Communication Constraints

A subset of Vehicle Routing Problems (VRPs) address the problem in which a vehicle is required to service targets by picking them up at an origin and deliver them to their destination. With respect to surveillance functions, this becomes a realistic problem as UAVs are restricted by operating range, data rate, Anti-Jam margins, and cost. Therefore, UAVs must be allocated to pickup targets and then deliver them to a communication range to be able to transmit information back to a command and control HQ. Research on similar problems illustrates the importance of studying dynamic scenarios as most of these problems are dynamic in nature (similar to this one). To maximize the information return to the communication range, the objective of allocating the UAVs is such that the total service time (pickup and delivery) of all the targets is minimized. A heuristic solution has been applied to this problem in previous work showing that good result can be obtained for the static problem. The focus of this research is to extend the work to operate in a dynamic scenario where requests arrive according to a stochastic process. It is shown that a solution to the static problem is also a solution to the dynamic problem. To address how the solution would perform in a dynamic environment, the scalability of the static solution is analyzed. Furthermore, the performance is analyzed under varying degrees of dynamism. Results show that the short execution times of the heuristic solution allow it to be used in a dynamic setup and show good performance under varying arrival rates. Furthermore, the boundaries for the regions of light and heavy load cases were identified and presented.

VRP with Minimum Delivery Latency Using Linear Programming

Vehicle Routing Problems (VRPs) concern allocating multiple vehicles to service requests with routes that start and end at a depot, visit all requests, and minimize some operational cost. Many UAV situations can be modeled in a manner that represents tasks to be accomplished by service requests and UAVs as agents that fulfill these requests. Often, communications are restricted by available power, data rate, noise margins, and cost. These constraints cause information gathered at request locations to be delivered only when the UAV returns to the “depot.” In other words, UAVs must be allocated to services requests and then deliver information gathered at the request location to a communication range to be able to transmit information back to a command and control HQ. Because of limited communication ranges which result from limited power onboard small UAVs, information is not delivered until the UAV returns to the depot. Research on similar problems show some related work on minimizing the latency of the pickup of requests but not the delivery. This work attempts to study the VRP with minimum delivery latency and formulate a Linear Programming (LP) solution. The objective function is shown to be nonlinear, but it can be linearized at a cost: many new variables are added to linearize it. This in turn, makes the problem intractable, and only solutions to small problems are obtained.

Experimental Validation of the Allocation of UAVs Under Communication Restrictions

A subset of Vehicle Routing Problems (VRPs) address the problem in which a vehicle is required to service targets by picking them up at an origin and deliver them to their destination. With respect to surveillance functions, this becomes a realistic problem as Unmanned Aerial Vehicles (UAVs) are restricted by operating range, data rate, Anti-Jam margins, and cost. Therefore, UAVs must be allocated to pickup targets and then deliver them to a communication range to be able to transmit information back to a command and control HQ. Because most VRPs adopt a minimum distance objective function, this can lead to suboptimal results when information is time critical. Related research developed a new VRP formulation and cost function (minimum delivery latency) and a heuristic for this problem with near-optimal performance and almost linear scalability to address this deficiency. To highlight the usability of this approach, experimental testing in a laboratory environment was completed. Both static and dynamic cases were validated in flight testing on commercially available platforms. Furthermore, the heuristic solutions to the minimum delivery latency objective function were compared to optimal solutions for the minimum distance method to demonstrate how much time can be saved in more realistic scenarios using the new approach.