George York

Cooperative Control of UAVs for Localization of Intermittently Emitting Mobile Targets

Compared with a single platform, cooperative autonomous unmanned aerial vehicles (UAVs) offer efficiency and robustness in performing complex tasks. Focusing on ground mobile targets that intermittently emit radio frequency signals, this paper presents a decentralized control architecture for multiple UAVs, equipped only with rudimentary sensors, to search, detect, and locate targets over large areas. The proposed architecture has in its core a decision logic which governs the state of operation for each UAV based on sensor readings and communicated data. To support the findings, extensive simulation results are presented, focusing primarily on two success measures that the UAVs seek to minimize: overall time to search for a group of targets and the final target localization error achieved. The results of the simulations have provided support for hardware flight tests.

Developing a Control Architecture for Multiple Unmanned Aerial Vehicles to Search and Localize RF Time-Varying Mobile Targets: Part I

In this paper, we present a control architecture that allows multiple Unmanned Aerial Vehicles (UAVs) to cooperatively detect mobile RF (Radio Frequency) emitting ground targets. The architecture is developed under the premise that UAVs are controlled as a distributed system. The distributed system-based technique maximizes the search and detection capabilities of multiple UAVs. We use a hybrid approach that combines a set of intentional cooperative rules with emerging properties of a swarm to accomplish the objective. The UAVs are equipped only with low-precision RF direction finding sensors and we assume the targets may emit signals randomly with variable duration. Once a target is detected, each UAV optimizes a cost function to determine whether to participate in a cooperative localization task. The cost function balances between the completion of detecting all targets (global search) in the search space and increasing the precision of cooperatively locating already detected targets. A search function for each UAV determines the collective search patterns of collaborating UAVs. Two functions used by each UAV determine (1) the optimal number of UAVs involved in locating targets, (2) the search pattern to detect all targets, and (3) the UAV flight path for an individual UAV. We show the validity of our algorithm using simulation results. Hardware implementation of the strategies is planned for this coming year.

Design of an anthropomorphic robot head for studying autonomous development and learning

We describe the design of an anthropomorphic robot head intended as a research platform for studying autonomously learning active vision systems. The robot head closely mimics the major degrees of freedom of the human neck/eye apparatus and allows a number of facial expressions. We show that our robot head can shift its direction of gaze at speeds which come close to that of human saccades. Since our design only makes use of low cost consumer grade components, it paves the way for widespread use of anthropomorphic robot heads in science, education, health-care, and entertainment.

Localization of ground targets using a flying sensor network

In this paper we propose a novel cooperation control method for multiple unmanned aerial vehicles (UAVs) to search, detect, and locate ground targets. We assume that our UAVs are only equipped with rudimentary angle of arrival sensors and the ground targets are mobile radio frequency (RF) emitters. In addition to the mobility, we assume our targets emit signals with random duration and frequency. Our cooperative UAVs function as a reconfigurable sensor network whose goal is to minimize target localization time and the target location error. The proposed control method extends our previous work on distributed control architecture for multiple UAVs by introducing localized hierarchical control structures. Such structures are dynamically established by local leaders that autonomously emerge in response to individual sensor readings, allowing efficient coordination of activities among nearby UAVs. The sensor network uses Kalman filtering techniques to minimize the target location error and the target localization time by combining current and past captured sensor values from multiple UAVs. We validate our proposed control method with simulation results that illustrate its performance under varying levels of cooperation among the members of the sensor network

Localizing mobile RF targets using multiple unmanned aerial vehicles with heterogeneous sensing capabilities

In this paper, we consider the problem of locating a mobile radio frequency (RF) target using multiple unmanned aerial vehicles (UAVs) equipped with sensors with varying accuracies. We investigate the localization task performance as we vary (1) the configuration of multiple UAVs (sensor locations), (2) the type of sensors onboard the UAVs, and (3) the sensor sequence. We use the well known optimal recursive estimation techniques (Kalman filtering) to combine captured sensor values from multiple UAVs and to investigate sensor scheduling issues to minimize the target location error. We present our findings in the form of simulation results.

Ground Target Detection Using Cooperative Unmanned Aerial Systems

In this paper, we present a comparative study that evaluates the merit of cooperative unmanned aerial sensor systems against that of a single system with equivalent capabilities. The motivation for our study stems from the current lack of theoretical and empirical work that shows the effectiveness of multiple cooperative unmanned vehicles (UAVs) over the proponents of a single, sophisticated UAV. Using a case study of searching and detecting ground targets with both electro-optical (EO) and infrared (IR) signatures, we quantify the advantage of multiple cooperative UAVs over a single UAV with equivalent sensing capabilities. Simulation results that support the use of cooperative systems over a single system are included.

Comparison of Cooperative Search Algorithms for Mobile RF Targets Using Multiple Unmanned Aerial Vehicles

In this chapter, we compare two cooperative control algorithms for multiple Unmanned Aerial Vehicles (UAVs) to search, detect, and locate multiple mobile RF (Radio Frequency) emitting ground targets. We assume the UAVs are equipped with low-precision RF direction finding sensors with no ranging capability and the targets may emit signals randomly with variable duration. In the first algorithm the UAVs search a large area cooperatively until a target is detected. Once a target is detected, each UAV uses a cost function to determine whether to continue searching to minimize overall search time or to cooperate in localization of the target, joining in a proper orbit for precise triangulation to increase localization accuracy. In the second algorithm the UAVs fly in formations of three for both search and target localization. The first algorithm minimizes the total search time, while the second algorithm minimizes the time to localize targets after detection. Both algorithms combine a set of intentional cooperative rules with individual UAV behaviors optimizing a performance criterion to search a large area. This chapter will compare the total search time and localization accuracy generated by multiple UAVs using the two algorithms simulations as we vary ratios of the numbers of UAVs to the number of targets.

Information-Based Cooperative Control for Multiple Unmanned Aerial Vehicles

In this paper, we present a distributed control algorithm to locate multiple mobile radio frequency (RF) targets using cooperative unmanned aerial vehicles (UAVs) equipped only with Global Positioning System (GPS) sensors and angle of arrival sensors. The proposed control algorithm considers the number of UAVs involved, sensor schedules, and sensor trajectories to increase the performance of cooperative UAVs localizing ground targets. The UAVs use optimal recursive estimation techniques to minimize target localization errors and differential geometry techniques to generate UAV (sensor) trajectories

Trajectory Transcriptions for Potential Autonomy Features in UAV Maneuvers

If a mission operation in a battlefield environment is complex, unmanned aerial vehicles (UAVs) and unmanned combat air vehicles (UCAVs) are expected to assist human pilots and ground personnel so that the threats to life are mitigated. Although these vehicles are controlled remotely by humans through ground commands or through another vehicle, possibilities to engage the onboard computer commands to respond to an environment and control the vehicle are not ruled out. Such systems are referred as autonomous UAVs and UCAVs. Following the traditional control technique adopted in aircraft motion technology, this paper assumes that autonomy and decision based controls for unmanned aircraft maneuvers are the features associated with the trajectory transcription techniques when the linear controllers at the inner loop are reconfigured. Hence in this paper, a tutorial on the switched linear controllers resulting from a non-zero equilibrium state from where an unmanned aircraft maneuver is required are presented. These evolving maneuvers instantaneously demanded in pitch plane are formulated using linear functional controllers of a micro air vehicle model. Linear functional controllers are the multiple input state feedback controllers which retain desired closed loop eigenvalues at fixed locations but the infinite eigenvector options that these controllers offer are transcripted for various ascent and descent angles of unmanned aircraft maneuvers.

An extended time horizon search technique for cooperative unmanned vehicles to locate mobile RF targets

In this paper, we present a behavior-based, distributed, cooperative search algorithm for multiple unmanned aerial vehicles (UAVs) to cooperatively find sub-optimal search patterns to detect moving radio frequency (RF) signal emitting targets. The overall goal of the search algorithm is to compute sub-optimal flight trajectories for participating UAVs to minimize the combined search cost: search coverage, time, fuel usage, and communication overhead. The focus for this paper is to extend our existing search algorithm s ability to incorporate evaluations of flight path options beyond the immediate time horizon. The paper explores the trade-offs over the additional computation cost and the reduction of the total search time. In addition to finding a set of sub-optimal UAV search paths, the search algorithm also generates a priority list of possible, search paths. The list is then used by an individual UAV to adjust its path selection to minimize a global search cost. Collectively, the selected UAV paths produce sub-optimal search patterns for a group of UAVs. The validity of the search algorithm is demonstrated using computer simulation