Showing papers presented at "Mediterranean Conference on Control and Automation in 2020"

Lithium-Ion Battery Monitoring and Observability Analysis with Extended Equivalent Circuit Model

Abstract: Lithium-Ion battery (LIB) on-line monitoring based on the extended equivalent circuit model (ECM) has received considerable attention. However, up to now, only linear observability analysis has been studied. The main purpose of this paper is to derive observability conditions for the extended ECM. Firstly, the battery ECM is decomposed into two sub-models mathematically, because there is no mutual coupling relationship between the battery states. Then, for the nonlinear extended one, an observability analysis is conducted. It shows that the observability necessary conditions depend on the battery current, the initial value of the battery capacity and the square of the derivative of the open circuit voltage (OCV) with respect to the state of charge (SOC). Numerical simulations have been conducted and have successfully assessed these conditions.

Attack Resilient Heterogeneous Vehicle Platooning Using Secure Distributed Nonlinear Model Predictive Control

Abstract: Recently, vehicle platoons have offered significant enhancements in traffic management, energy consumption and safety in intelligent transportation systems. Despite the benefits brought by the platoons, they potentially suffer from insecure networks which provide the connectivity among the vehicles participating in the platoon. This paper deals with the secure control of vehicle platoons under the risk of a common cyber attack, namely Denial of Service (DoS) attack. A DoS intruder can endanger the security of platoon by jamming the communication network among the vehicles which is responsible to transmit inter-vehicular data throughout the platoon. This can potentially result in huge performance degradation or even hazardous collisions. We propose a secure distributed nonlinear model predictive control algorithm consisting of i) detection and ii) mitigation phases. The algorithm is capable of handling DoS attack performed on a platoon equipped by different communication topologies and at the same time it guarantees the desired formation control performance. Stability analysis of the attacked platoon running the given algorithm is also presented. Simulation results on a sample heterogeneous attacked platoon exploiting two-predecessor follower communication environment demonstrates the effectiveness of the method.

A Rendezvous Algorithm for Multi-agent Systems in Disconnected Network Topologies

Abstract: This paper addresses the problem of having a multi-agent system converging to a rendezvous location for networks of agents without any type of localization sensor. A central node or tower is able to determine the noisy position of each agent and transmit it using a directional antenna. Given the asynchronous communication setup, the topology will not be connected in general, which precludes the use of set-consensus algorithms available in the literature. By devising a flocking rule with mechanisms to prevent collisions, nodes explore the mission plane while maintaining the current connectivity. The process is followed by a convergence phase using a modified set-consensus algorithm with collision-free guarantees and asymptotic convergence. Results are also presented that, if there is a sufficient density of nodes, convergence occurs to a single cluster. The performance of the proposed algorithm is assessed through simulations, illustrating the cases where convergence occurs to a single or multiple clusters.

Model-Free Control as a Service in the Industrial Internet of Things: Packet loss and latency issues via preliminary experiments

Abstract: Model-Free Control (MFC), which is easy to implement both from software and hardware viewpoints, permits the introduction of a high level control synthesis for the Industrial Internet of Things (IIoT) and the Industry 4.0. The choice of the User Diagram Protocol (UDP) as the Internet Protocol permits to neglect the latency. In spite of most severe packet losses, convincing computer simulations and laboratory experiments show that MFC exhibits a good Quality of Service (QoS) and behaves better than a classic PI regulator.

Deep Weed Detector/Classifier Network for Precision Agriculture

Abstract: The productivity of crop farming keeps diminishing at an alarming rate due to infestation of weeds and pests. Deep learning is becoming as the approach for identifying weeds on farmlands. However, training weed data sets with deep learning classification alone trains the whole images consisting of the weed and its background (soil) without categorically telling which particular item in the image is a weed. This makes utilising this classification approach for precision agriculture difficult. We present an alternative approach, which involves incorporating a pre-trained network in this case ResNet-50 and YOLO v2 object detector for weed detection/classification on farmlands. Thus, weeds can precisely be located, identified (type), sprayed with the appropriate herbicide or removed with the appropriate mechanism. This sums up weeding process in precision agriculture.

Design and Modeling of Unconventional Quadrotors

Abstract: In this paper, we propose a new design of unconventional quadrotors with simple mechanisms. They are able to change the orientation and the length of their arms independently thanks to their adaptive morphologies, these quadrotors can be used for specific missions and in more congested spaces. Thus, the usefulness of certain configurations that our quadrotors can make will be shown. In addition, a detailed and generic model is proposed for three possible cases: 1) the extension of the arms, 2) the rotation of the arms, 3) the hybridization between the extension and rotation of the arms. We stress that the modeling of these unconventional quadrotors is more challenging compared to the standard quadrotors, because of the geometrical asymmetry of the quadrotor structures and the constant change of the inertia and the center of gravity during the flight. In this work, a preliminary prototype of our quadrotor with rotating arms is presented with the electronic architecture.

Trajectory Reference Generation and Guidance Control for Autonomous Vehicle Lane Change Maneuver

Abstract: One of the expected tasks that autonomous vehicle may do is driving on highway which makes performing lane change maneuvers inevitable. In order to ensure the safety of the lane change, the contingency planning approach must generate smooth and feasible local trajectory adequate for overtaking or collision avoidance applications. This paper presents a method to carry out a lane change maneuver based on a sigmoid function trajectory. The main contribution of this paper is to provide an envelope of the set of the feasible parameterized sigmoid functions based on vehicle dynamics and geometric constraints. The few parameters used in the trajectory adjustment make the lane change maneuver safe, comfortable laterally and applicable in real-time. The proposed lateral controller switches between attenuating the lateral error and the orientation error to better track the desired reference trajectory. CarMaker simulation results show the applicability of the proposed approach.

Vehicle odometry model identification considering dynamic load transfers

Abstract: The paper proposes a parameter identification method for a vehicle model using real measurements of onboard sensors. The motivation of the paper is to improve the localization of the vehicle when the accuracy of the regular methods is poor, e.g. in the case of unavailable GNSS signals, no enough feature for vision, or low acceleration for IMU-based techniques. In these situations the wheel encoder based odometry may be an appropriate choice for pose estimation, however, this method suffers from parameter uncertainty and unmodelled effects. The utilized vehicle model operates with dynamic wheel radius. The proposed identification method combines the Kalman-filter and least square techniques in an iterative loop for estimating the parameters. The estimation process is verified by real test of a compact car. The results are compared with the nominal setting, in which there is no estimation.

Pursuit-evasion Game for Nonholonomic Mobile Robots With Obstacle Avoidance using NMPC

Abstract: In this work, non-cooperative competitive games between two unmanned ground robots using Nonlinear Model Predictive Control (NMPC) while incorporating obstacle avoidance techniques are studied. The objective of the first player (pursuer) is to minimize the relative distance and orientation between itself and the second player (evader) while avoiding obstacles, whereas the evader does the opposite. The Pursuit-Evasion Game (PEG) being a typical class of a differential game is formulated as a zero-sum game with two homogeneous players in five different game scenarios. The objective function of each player is formulated as a double optimization problem and is solved separately using NMPC techniques. The optimal trajectory of each player is computed iteratively by considering the best response of the opponent player. The level of information is assumed to be symmetric. Simulations of various scenarios show the winning possibility of each player.

Interval Observer Design for Uncertain Linear Continuous-time Metzlerian Systems

Abstract: For linear continuous-time positive systems the paper proposes an approach, reflecting structural constraints and positiveness in the problem of Metzlerian and strictly Metzlerian interval observers design. Every interval matrix boundary is represented by a set of linear matrix inequalities representing Metzler matrix parameter constraints and reflecting potential zero elements in a Metzler matrix structure by structural diagonal matrix variables. Combined with couple of Lyapunov inequalities, the observer Metzler matrix parameters are guaranteed and interval stability is attained. A numerical example is included to assess the feasibility technique.

Analysis of a novel autonomous underwater robot for biofouling prevention and inspection in fish farms

Abstract: Biofouling is a challenge for finfish farming as it can impact cage stability and fish health. Amongst others, current strategies against biofouling rely heavily on removal of biofouling using in-situ pressure cleaning of nets. The cleaning waste is released into the water where it can impact the health of the cultured fish. Grooming, the regular cleaning of nets to prevent biofouling communities from establishing, is one novel strategy that is currently explored using tethered underwater robots. In addition, remotely operated vehicles (ROVs) are used for inspection of the net to ensure its integrity and prevent fish escapes, while stationary sensors at the farm barge are employed to extrapolate on environmental conditions in the net pens. In this paper, the requirements for a permanently resident, autonomous and tetherless subsea robot for cleaning and inspection are proposed. As such, the robot aims to combine several application areas and offer a solution to the biofouling challenge while at the same time improving safeguarding of net integrity and monitoring of environmental conditions directly in the pen. Using the SEATONOMY method from an operational viewpoint, the robot's individual operations are analysed. This included i) Environmental condition monitoring, ii) Net and biofouling inspection, iii) Growth prevention and iv) Docking. As a result, the specifications and requirements for the development of a novel robotic system that is able to simultaneously and autonomously perform inspection, growth prevention and monitoring in fish cages were derived. The paper proposes the development of a new technology and biofouling management strategy that will contribute to increase the efficiency and production demands in the aquaculture industry.

LPV MPC Control of an Autonomous Aerial Vehicle

Abstract: This work proposes the use of the LPV MPC approach for the position-heading control problem of a small quadrotor. The use of a LPV representation of the attitude subsystem allows to consider it as independent from the position one allowing the use of a cascade control structure. For the inner attitude control loop, the proposed LPV MPC controller is solved efficiently using quadratic programming. The proposed approach is applied to an AscTec Hummingbird UAV in simulation.

Offset-free Model Predictive Control: A Study of Different Formulations with Further Results

Abstract: This paper presents discussions on offset-free model predictive control (MPC) methods for linear discrete-time systems in the presence of deterministic system disturbances. The general approach is based on the use of a disturbance model and an observer to estimate the disturbance states. The recent development in offset-free MPC has established the equivalence of the velocity form (without output delay) to a specific choice of the disturbance model and observer. In this note, it was shown that this particular disturbance model and observer is not necessarily equivalent to the velocity form with output delay. Nevertheless, it was shown that the velocity form with output delay is equivalent to a different choice of the disturbance model and observer. An import of this result is that the velocity forms (with and without delayed output) belong to the same general approach - disturbance model and observer. Furthermore, areas that may be considered in future researches are also highlighted.

PCA Methods and Evidence Based Filtering for Robust Aircraft Sensor Fault Diagnosis

Abstract: In this paper PCA and D-PCA techniques are applied for the design of a Data Driven diagnostic Fault Isolation (FI) and Fault Estimation (FE) scheme for 18 primary sensors of a semi-autonomous aircraft. Specifically, Contributions-based, and Reconstruction-based Contributions approaches have been considered. To improve FI performance an inference mechanism derived from evidence-based decision making theory has been proposed. A detailed FI and FE study is presented for the True Airspeed sensor based on experimental data. Evidence Based Filtering (EBF) showed to be very effective particularly in reducing false alarms.

MAV Development Towards Navigation in Unknown and Dark Mining Tunnels

Abstract: The Mining industry considers the deployment of Micro Aerial Vehicles (MAVs) for autonomous inspection of tunnels and shafts to increase safety and productivity. However, mines are challenging and harsh environments that have a direct effect on the degradation of high-end and expensive utilized components over time. Inspired by this effect, this article presents a low cost and modular platform for designing a fully autonomous navigating MAVs without requiring any prior information from the surrounding environment. The design of the proposed aerial vehicle can be considered as a consumable platform that can be instantly replaced in case of damage or defect, thus comes into agreement with the vision of mining companies for utilizing stable aerial robots with reasonable cost. In the proposed design, the operator has access to all on-board data, thus increasing the overall customization of the design and the execution of the mine inspection mission. The MAVs platform has a software core based on Robot Operating System (ROS) operating on an Aaeon UP-Board, while it is equipped with a sensor suite to accomplish the autonomous navigation equally reliable when compared to high-end and expensive platforms.

Adaptive Control of a Linear Hyperbolic PDE with Uncertain Transport Speed and a Spatially Varying Coefficient

Abstract: Recently, the first result on backstepping-based adaptive control of a 1-D linear hyperbolic partial differential equation (PDE) with an uncertain transport speed was presented. The system also had an uncertain, constant in-domain coefficient, and the derived controller achieved convergence to zero in the $L_{\infty}$ -sense in finite time. In this paper, we extend that result to systems with a spatially varying in-domain coefficient, achieving asymptotic convergence to zero in the $L_{\infty}$ -sense. Additionally, for the case of having a constant in-domain coefficient, the new method is shown to have a slightly improved finite-time convergence time. The theory is illustrated in simulations.

Implementation of a Petri-net based Digital Twin for the development procedure of an Electric Vehicle

Abstract: In the current work the development procedure (design, manufacture and assembly) of an electric vehicle is considered. Uncertainties make difficult to follow initial time plan, so monitoring of the development procedure is necessary. To handle delays a method using Petri nets to model the tasks of the development procedure and their dependencies is introduced. The model is not used offline as a passive element but is connected and interacts with the physical system (development procedure). Based on the information exchange between physical and digital system, alternative ways to overcome delays are studied and the optimal solution is calculated, tested and applied. Results are provided according to different scenarios, in order to show the efficiency and applicability of the proposed method.

Identification of the Quadcopter Vertical Translation Dynamics

Abstract: The paper deals with the quadcopter dynamics model identification. The vertical translation is separated as the SISO control loop. The transfer functions from the throttle to vertical velocity and to altitude are studied. The finite-frequency identification is exploited to estimate the transfer function coefficients. This approach includes the feature of the closed loop identification, that is, a controller operates to stabilize the system and to provide performance during the identification test. However, the controller formulae and parameters are not needed for identification. It is important for a quadcopter controlled by standard autopilot software. A standard quadcopter is used as the experimental setup. The identification results from the experimental flight data are considered.

Model Predictive Control of a Convertible Tiltrotor Unmanned Aerial Vehicle

Abstract: This paper presents a Model Predictive Control (MPC) based control structure for a convertible Unmanned Aerial Vehicle (UAV) with fixed wings and tiltrotor thrust vectoring. The controller encompasses full flight envelope trajectory tracking, thereby optimally exploiting the aircraft's Vertical Take Off and Landing (VTOL) and cruising-forward flight capabilities. An adaptive control allocation is designed to handle the changing control authorities of the actuators and efficiently distribute the required control actions between propellers, tilt servos and control surfaces. Preliminary simulation results show the feasibility of the proposed control approach.

Gantry Crane Position Control via Parallel Feed-forward Compensator

Abstract: This article is concerned with an output feedback position control of gantry (overhead) cranes applying a parallel feed-forward compensator (PFC) that allows for a reduction of payload swinging without additional sensors or payload swing angle estimations. Performance and stability of the controlled system are achieved by defining a new output as a combination of the original output and the output of an appropriate PFC. The later provides an almost strict positive real (ASPR) condition for the augmented plant. Thus, high gain output feedback control becomes applicable. The proposed feedback control approach is successfully validated in a numerical study and in experiments on a laboratory plant.

Fused-PID Control for Tilt-Rotor VTOL Aircraft

Abstract: This work presents a fused-PID (FPID) control strategy for a tilt-quadrotor VTOL (vertical takeoff and landing) where the FPID generates attitude setpoints based on the pilot's velocity commands. The FPID controller consists of two separate PID controllers where one can handle fixed-wing aircraft and the other is designed for multicopters. The output of both controllers is fused depending on the airspeed to obtain an attitude setpoint and a tilt angle of the rotors. In contrast to the commonly-used binary switch between helicopter flight (propellers pointing upwards) and airplane flight (propellers pointing forwards), the FPID makes a smooth transition between the two flight modes possible. The proposed controller design has been extensively tested in simulation and with a real radio-controlled tilt-quadrotor VTOL aircraft. The results presented in this work clearly show that the FPID approach, despite its simplicity, is suited very well for tilt-rotor VTOL aircraft.

Gain Scheduled Subspace Predictive Control of a Pressurized Water-type Nuclear Reactor

Abstract: This work presents a methodology for designing subspace-based gain scheduled predictive controller for nuclear reactor power control. The main idea is to design a family of predictive controllers directly from measurements and integrate them without employing any explicit process model. The developed controller incorporates the robustness feature of subspace identification with the adaptive capability of gain scheduling in a predictive control set-up. The controller is designed to handle process variations effectively. The efficacy of the proposed controller is demonstrated for load-following transients using a simulated model of a PWR-type nuclear reactor. Simulation results show that the proposed strategy is effective in addressing the load-following control problem of a non-linear parameter-varying PWR nuclear reactor system.

Application of Pareto Optimization in an Economic Model Predictive Controlled Microgrid

Abstract: This paper presents an economic model predictive control approach for a linear microgrid model The microgrid in grid-connected mode represents a medium-sized company building including storage systems, renewable energies and couplings between the electrical and heat energy system Economic model predictive control together with Pareto optimization is applied to find suitable compromises between two competing objectives, ie monetary costs and thermal comfort Using realworld data from 2018 and 2019, the model is simulated with auto-detection of the Pareto solution which is closest to the Utopia point The results show that the Pareto optimization can either be used in real-time control of the microgrid, or to obtain suitable weights from long term simulations Both approaches result in significant cost reductions

Trajectory Tracking Control for a Kinematic Bicycle Model

Abstract: This paper deals with the trajectory tracking control problem in the kinematic bicycle model. To avoid possible singular states in the control system for the model under consideration we apply a sigmoid function (hyperbolic tangent) in the feedback loop. We present an error function between the real and virtual vehicles that follows the required trajectory, and introduce a control law that stabilizes asymptotically the zero error state in the error dynamics. In addition, the proposed algorithm allows us to handle the control problem in cases where actuator saturations and state constraints exist. The paper is concluded with an example that demonstrate the characteristics of the control law and its performance.

Distributed Time-varying Kalman Filter Design and Estimation over Wireless Sensor Networks Using OWA Sensor Fusion Technique

Abstract: In this paper, a novel estimation procedure is proposed, which consists of designing a distributed class of time-varying Kalman filter based on wireless sensor networks topology along with a new sensor fusion method. The proposed technique is employed to estimate the states and outputs of a linear time-varying system with a high level of accuracy. Both the dynamics of the system and the measurements are assumed to be contaminated by external noises. The notion of Orness and Ordered Weighted Averaging (OWA) operator technique are utilized to fuse the estimation of the sensors. O'Hagan method, along with the gradient descent method, is employed to find the optimal weights. In the introduced approach, OWA weights are learned for each observation such that they efficiently minimize the estimation error for that particular observation. This will result in an outstanding high accurate sensor fusion outcome. In addition, two optimistic and pessimistic exponential OWA operators are used and compared together to achieve a pre-specified level of Orness. The simulation results are shown on a given linear time-varying system to verify the effectiveness of the proposed sensor fusion distributed filtering design method.

Indoor Localization and Mapping: Towards Tracking Resilience Through a Multi-SLAM Approach

Abstract: This paper presents a use case for SLAM techniques applied to real time localization and detailed mapping for emergency response personnel in non cooperative environments. Such environments tend to defeat conventional localization approaches, therefore we must ensure continuous operation of our localization and mapping regardless of the difficulties encountered (lack of GPS signals, lighting conditions, smoke, etc.). The proposed system fuses two SLAM algorithms, a LiDAR-based and a camera-based. Since LiDAR-based SLAM uses dense 3D measurements, it is well suited to the construction of a detailed map, while the visual SLAM allows to quickly recognize already visited places in order to apply loop closure corrections, by using a key frames graph. The currently proposed system allows collaboration between these two SLAMs through pose sharing and relocalization.

Event-Trigger based Adaptive-Robust Guidance Strategy with Input Saturation

Abstract: Over the last few decades, the computational capabilities of the state-of-the-art processors have increased extensively and are now capable of performing multiple parallel computations. However, due to constraint on cost, weight, and space, they are not generally installed as an on-board processor of an interceptor. This limitation of the computational capabilities of the missile processor is addressed in this paper, by designing an event-triggered strategy which reduces the number of control updates. In addition to achieving the primary objective of capturing the target, the proposed guidance strategy enables the interceptor to tackle external disturbances that it encounters. Further the conservative assumption of apriori knowledge about the upper bound of the disturbances is eliminated by adaptively tuning the gains of the proposed guidance law. Input saturation has also been considered on the control effort of the missile so as to be consistent with most practical interceptor systems. Lyapunov theory has been used to demonstrate asymptotic stability of the closed-loop system states under the proposed event-triggered control scheme. Numerical simulations for the guidance strategy on the commonly occurring tail-chase and head-on engagement scenarios are performed along with a comparative performance analysis with the periodic time-triggered technique.

Ethical Considerations for a Decision Making System for Autonomous Vehicles During an Inevitable Collision

Abstract: The automotive industry is heading towards the introduction of fully autonomous vehicles. However before these type of vehicles are commercially available at mass scale, some issues need to be solved. A major issue is the ethics involved in the decision-making during an accident; this paper presents an analysis of how to solve this issue. The proposal is a pre-programmed system with different ethical settings based on six formal ethical theories. For the implementation, eight ethical concerns are defined and ordered according to each theory. These concerns are defined in terms of harm to self and harm to others. The ethical concerns are used as a guideline to define the level of importance of each person or object in an accident scenario. With the proposed system, the vehicle will be partially tailored to the preferences of different users while still being bounded by legal requirements to avoid any misuse.

Robust Fault Detection for Switched Systems based on Interval Observers

Abstract: This paper deals with fault detection (FD) for a class of discrete-time switched systems with sensor faults. Under the assumption that disturbances and measurement noise are unknown but bounded, two methods, respectively based on the $H_{\infty}$ and the $L_{\infty}$ formalisms, are introduced to attenuate the effects of the uncertainties and to improve the accuracy of the proposed residual framers. The design conditions of the interval observer are given in terms of Linear Matrix Inequalities (LMIs). Furthermore, a FD decision is developed to indicate the presence of faults. A numerical example is performed to illustrate the effectiveness of the proposed method based on the $L_{\infty}$ performance through a comparison with the results obtained using the $H_{\infty}$ analysis.

On estimation of the parameters in the complex network via the adaptive observer

Abstract: In this paper, the problem of estimation of the unknown parameters in the complex networks composed of identical nodes being generalized Lorenz chaotic systems will be discussed. The estimation is provided by an adaptive observer. Estimation of the parameters is provided on the complex network with ring topology and bidirectional coupling between nodes. The extra node being an adaptive observer is connected with the permanent node of the analyzed complex network. During the relatively short period, this additional node will be identically synchronized with the analyzed complex network without precise knowledge of the unknown parameters in the permanent nodes. The results are illustrated by the numerical simulation as well.