On-board State Estimation Research Articles

Onboard State Estimation Around Didymos With Recurrent Neural Networks and Segmentation Maps

Onboard State Estimation Around Didymos With Recurrent Neural Networks and Segmentation Maps

Battery Charge Curve Prediction via Feature Extraction and Supervised Machine Learning.

Real-time onboard state monitoring and estimation of a battery over its lifetime is indispensable for the safe and durable operation of battery-powered devices. In this study, a methodology to predict the entire constant-current cycling curve with limited input information that can be collected in a short period of time is developed. A total of 10066 charge curves of LiNiO2 -based batteries at a constant C-rate are collected. With the combination of a feature extraction step and a multiple linear regression step, the method can accurately predict an entire battery charge curve with an error of < 2% using only 10% of the charge curve as the input information. The method is further validated across other battery chemistries (LiCoO2 -based) using open-access datasets. The prediction error of the charge curves for the LiCoO2 -based battery is around 2% with only 5% of the charge curve as the input information, indicating the generalization of the developed methodology for predicting battery cycling curves. The developed method paves the way for fast onboard health status monitoring and estimation for batteries during practical applications.

Editorial: Towards Real-World Deployment of Legged Robots

part of its weight up while walking. This novel design provides the advantage of being extremely safe and 18 low-cost. The design of the robot incorporates tendon wires to actuate the articulations attached to their 19 thin linkages. The robot employs a data-driven neural network approach for performing walking behaviors.Employing such approach enables to handle the underactuated nature of the robot since BALLU2 has only 21 2 actuated degrees of freedom per leg. In order to determine the transition function for walking, the authors 22 perform a correlation study to chose significant state variables. Then, they train a neural network to learn 23 transition times on both single and double support phases. Walking is accomplished in the real system at a 24 speed of 0.18m/s using an RGBD camera for onboard state estimation.

Multi-IMU Based Alternate Navigation Frameworks: Performance &amp; Comparison for UAS

On-board state estimation is a persistent challenge to fielding unmanned aerial systems (UAS), particularly when global positioning system (GPS) measurements are not available. The dominant estimation tool remains the extended Kalman filter (EKF) applied to an inertial measurement unit (IMU). The growing availability of low-cost commercial-grade IMUs raises questions about how to best improve sensor readings into an estimate when measurements are available from multiple IMUs. This paper evaluates four different approaches to attitude estimation from multiple IMU measurements and their application in high dynamic motion. The four approaches are fusion of measurements (virtual IMU), fusion of state estimates (Federated KF), feedback fusion state estimate (Feedback Federated KF), and an EKF design incorporating the additional measurements (Augmented KF); these correspond to fusion before, within, or after state estimation. The performance of the approaches are quantified for varying IMU number theoretically and experimentally. The experiments use on-board autopilot hardware implementations of the estimators during maneuvers in a motion capture volume and flights on-board an Unmanned Aerial Systems (UAS) implementation, with the peak and root-mean-square (RMS) errors used to quantify accuracy. The RMS error results indicate that the feedback federated Kalman Filter using five IMUs returns 38% compared to general federated Kalman Filter using 37% accuracy improvement over a single IMU. This improvement compares to a 19% improvement for virtual IMU and 9% improvement for the Augmented KF respectively. These results indicate that both the Federated KF approach achieves the lowest RMS error relative to the virtual IMU and augmented KF approaches and inform the design of multi-IMU UAS pose estimators.

Simulation Processes for Onboard State Estimation in a Small UAV Environment

Simulation Processes for Onboard State Estimation in a Small UAV Environment

On-Board State Estimation in Electrical Vehicles: Achieving Accuracy and Computational Efficiency Through an Electrochemical Model

Success of electric mobility and connected future depends on advanced high capacity Lithium-ion batteries and a tailored battery management system that keeps track of battery health and safety to optimize the performance. The drive for advanced batteries is being pursued on one front via the development of novel chemistry, for example, blended composite cathodes to achieve enhanced performance and life. On the other front, the need for optimal battery performance via operational control viz. a robust Battery Management System (BMS) that accurately predicts state-of-charge (SOC), state-of-power (SOP) and state-of-health (SOH) in a computationally efficient way, has been lacking in many ways due to the reliance on simple and fast-to-compute models such as equivalent circuit models (ECM) that lack the accuracy needed for large battery packs. This paper reports an on-board reduced-order electrochemical thermal model (ROTM) based SOC and voltage estimation for a 12S1P configuration composite-cathode battery pack and demonstrates its practicality by implementing it on four different micro-controller units (MCU) (ATmega:2560&328, Infineon:TC275&TC297). We show that on-board ROTM based SOC and voltage prediction have better accuracy compared to the conventional ECM based methods under both static and dynamic load conditions.

A Closed Form Reduced Order Electrochemical Model for Lithium-Ion Cells

The state of a lithium (Li)-ion cell as a function of measureable quantities like cell voltage and current is not available in closed form without simplifying assumptions like uniform reaction rate, thus limiting the applicability of reduced order electrochemical models for on-board estimation. In the present work, an analytic form for non-uniform reaction rate profile is derived under certain limiting conditions. Using this form and polynomial approximations for the concentration profiles, a direct non-iterative solution scheme is developed for estimation of cell voltage, state of charge (SOC) and other internal variables. The predicted cell voltage is validated against experimental results for commercial cells while the internal variables are verified against the full pseudo 2 dimensional (P2D) model. The results show that, in spite of the order reduction, the model predicts cell voltage as well as the physical processes of the Li-ion cell accurately with less than 1% error for a wide range (0.2C – 5C) of operating conditions. The self-consistent predictive capability along with low computational cost, makes the proposed model an ideal candidate for physics based on-board state estimation.

High-voltage Lithium-ion Batteries — Methods for On-board State Estimation

High-voltage Lithium-ion Batteries — Methods for On-board State Estimation

Design of a Two-Wheel Self-Balancing Robot with the Implementation of a Novel State Feedback for PID Controller using On-Board State Estimation Algorithm

Design of a Two-Wheel Self-Balancing Robot with the Implementation of a Novel State Feedback for PID Controller using On-Board State Estimation Algorithm

Hybrid State Space Modeling of a Spark Ignition Engine for Online Fault Diagnosis

In this work, a nonlinear hybrid state space model of a complete spark ignition (SI) gasoline engine system from throttle to muffler is developed using the mass and energy balance equations. It provides within-cycle dynamics of all the engine variables such as temperature, pressure, and mass of individual gas species in the intake manifold (IM), cylinder, and exhaust manifold (EM). The inputs to the model are the same as that commonly exercised by the engine control unit (ECU), and its outputs correspond to available engine sensors. It uses generally known engine parameters, does not require extensive engine maps found in mean value models (MVMs), and requires minimal experimentation for tuning. It is demonstrated that the model is able to capture a variety of engine faults by suitable parameterization. The state space modeling is parsimonious in having the minimum number of integrators in the model by appropriate choice of state. It leads to great computational efficiency due to the possibility of deriving the Jacobian expressions analytically in applications such as on-board state estimation. The model was validated both with data from an industry standard engine simulation and those from an actual engine after relevant modifications. For the test engine, the engine speed and crank angle were extracted from the crank position sensor signal. The model was seen to match the true values of engine variables both in simulation and experiments.

Asteroid rotation and orbit control via laser ablation

This paper presents an approach to control the rotational motion of an asteroid while a spacecraft is deflecting its trajectory through laser ablation. During the deflection, the proximity motion of the spacecraft is coupled with the orbital and rotational motion of the asteroid. The combination of the deflection acceleration, solar radiation pressure, gravity field and plume impingement will force the spacecraft to drift away from the asteroid. In turn, a variation of the motion of the spacecraft produces a change in the modulus and direction of the deflection action which modifies the rotational and orbital motion of the asteroid. An on-board state estimation and control algorithm is then presented that simultaneously provides an optimal proximity control and a control of the rotational motion of the asteroid. It will be shown that the simultaneous control of the rotational and proximity motions of asteroid and spacecraft has a significant impact on the required deflection time.

A reduced order electrochemical thermal model for lithium ion cells

A reduced order model (ROM) is proposed for accurate prediction of electrochemical and thermal response of lithium ion cells. The order reduction is carried on the coupled partial differential equations (PDE) of the electrochemical thermal model by consistent volume averaging of local heat generation and spatial temperature variation. The model is validated with experimental data for temperatures ranging from 253 K–333 K. It is seen that modification of ROM to account for low electronic conductivity results in accurate voltage estimation of cells with lithium nickel cobalt aluminium oxide (LNCO) cathodes. A detailed parametric sensitivity to operating conditions is provided. The utility of ROM for on-board state estimation is demonstrated by applying to realistic drive cycle protocols such as the Hybrid Pulse Power Characterization (HPPC) and the Urban Dynamometer Driving Schedule (UDDS) data. The electrochemical structure of ROM enables identification of controlling processes, and analysis of HPPC results reveal that Ohmic drop of cathode is controlling at high rates and the electrolyte potential during rest phase. Based on accurate voltage prediction, computational speed and physical insights, it can be concluded that the proposed ROM is an adequate state estimation and a cell design tool.

H∞ Filtering for Cloud-Aided Semi-active Suspension with Delayed Road Information

The paper considers a cloud-aided semi-active vehicle suspension. In this system, road profile information is downloaded from a cloud database to facilitate the onboard state estimation. Time-varying delays are considered during the data transmission. An H∞ filter is designed that exploits both onboard sensor measurements and delayed road profile information from the cloud. Disturbances due to GPS localization error and time delay inaccuracies are treated. The H∞ performance analysis is presented and sufficient conditions for the existence of the filter are derived as linear matrix inequalities (LMIs). Filter design is then developed with the Projection Lemma. Numerical simulations are presented to illustrate the effectiveness of the designed filter in estimating the suspension states.

On-board vision-based spacecraft estimation algorithm for small body exploration

A methodology is summarized for designing on-board state estimators in support of spacecraft exploration of small bodies such as asteroids and comets. This paper focuses on an estimation algorithm that incorporates two basic computer-vision measurement types: a landmark table (LMT) and a paired feature table (PFT). Several innovations are developed to incorporate these measurement types into the on-board state estimation algorithm. Simulations are provided to demonstrate the feasibility of the approach.