PyTorchKR Papers

최신 논문

총 1,050편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2109.14985

The Artificial Intelligence behind the winning entry to the 2019 AI Robotic Racing Competition

Robotics is the next frontier in the progress of Artificial Intelligence (AI), as the real world in which robots operate represents an enormous, complex, continuous state space with inherent real-time requirements. One extreme challenge in robotics is currently formed by autonomous drone racing. Human drone racers can fly through complex tracks at speeds of up to 190 km/h. Achieving similar speeds with autonomous drones signifies tackling fundamental problems in AI under extreme restrictions in terms of resources. In this article, we present the winning solution of the first AI Robotic Racing (AIRR) Circuit, a competition consisting of four races in which all participating teams used the same drone, to which they had limited access. The core of our approach is inspired by how human pilots combine noisy observations of the race gates with their mental model of the drone's dynamics to achieve fast control. Our approach has a large focus on gate detection with an efficient deep neural segmentation network and active vision. Further, we make contributions to robust state estimation and risk-based control. This allowed us to reach speeds of ~9.2m/s in the last race, unrivaled by previous autonomous drone race competitions. Although our solution was the fastest and most robust, it still lost against one of the best human pilots, Gab707. The presented approach indicates a promising direction to close the gap with human drone pilots, forming an important step in bringing AI to the real world.

arXiv:2109.13873

Establishment and placement of a Multi-purpose Phasor measurement unit to improve parallel state estimation in distribution Networks

Today, microgrids in distribution networks are in dire need of improvement to cope with economic challenges, human losses, and equipment placement issues. Today, there is the issue of scattered resources in distribution systems, which has created many problems in the areas of environment, economy, and human and animal losses. The most important challenge in this section is the existence of voltage and frequency fluctuations during the occurrence of possible events such as severe load changes or errors in distribution networks. Having such a big problem can call a distribution network into question and destroy it. Therefore, it is necessary to provide an optimal method that can meet and cover these challenges. For this purpose, the present research deals with the problem of establishing and placing a multifunctional phasor measurement unit to improve the parallel state estimation in distribution networks, which offers a control approach. This approach determines the time of occurrence of internal and external disturbances after using the phasor unit. The approach of this research is to use a neural-fuzzy method because there is uncertainty in the distribution network due to the mentioned challenges, and training in the system is needed to accurately deploy and place possible errors. Do not occur. When setting up and placing the phasor measuring unit, the most important issue is the proper distribution of the load in the distribution network. The simulation results in the MATLAB / Simulink environment show the improvement of the results according to the proposed approach. Keywords: Distribution Network, Neural-Fuzzy Network, Optimal Load Distribution, Parallel State Estimation, Phasor Measurement Unit.

arXiv:2109.12255

Constrained Attack-Resilient Estimation of Stochastic Cyber-Physical Systems

In this paper, a constrained attack-resilient estimation algorithm (CARE) is developed for stochastic cyber-physical systems. The proposed CARE can simultaneously estimate the compromised system states and attack signals. It has improved estimation accuracy and attack detection performance when physical constraints and operational limitations are available. In particular, CARE is designed for simultaneous input and state estimation that provides minimum-variance unbiased estimates, and these estimates are projected onto the constrained space restricted by inequality constraints subsequently. We prove that the estimation errors and their covariances from CARE are less than those from unconstrained algorithms and confirm that this property can further reduce the false negative rate in attack detection. We show that estimation errors of CARE are practically exponentially stable in mean square. Finally, an illustrative example of attacks on a vehicle is given to demonstrate the improved estimation accuracy and detection performance compared to an existing unconstrained algorithm.

arXiv:2109.12269

Integrating Recurrent Neural Networks with Data Assimilation for Scalable Data-Driven State Estimation

Data assimilation (DA) is integrated with machine learning in order to perform entirely data-driven online state estimation. To achieve this, recurrent neural networks (RNNs) are implemented as surrogate models to replace key components of the DA cycle in numerical weather prediction (NWP), including the conventional numerical forecast model, the forecast error covariance matrix, and the tangent linear and adjoint models. It is shown how these RNNs can be initialized using DA methods to directly update the hidden/reservoir state with observations of the target system. The results indicate that these techniques can be applied to estimate the state of a system for the repeated initialization of short-term forecasts, even in the absence of a traditional numerical forecast model. Further, it is demonstrated how these integrated RNN-DA methods can scale to higher dimensions by applying domain localization and parallelization, providing a path for practical applications in NWP.

arXiv:2101.00161

Design of heterogeneous multi-agent system for distributed computation

A group behavior of a heterogeneous multi-agent system is studied which obeys an "average of individual vector fields" under strong couplings among the agents. Under stability of the averaged dynamics (not asking stability of individual agents), the behavior of heterogeneous multi-agent system can be estimated by the solution to the averaged dynamics. A following idea is to "design" individual agent's dynamics such that the averaged dynamics performs the desired task. A few applications are discussed including estimation of the number of agents in a network, distributed least-squares or median solver, distributed optimization, distributed state estimation, and robust synchronization of coupled oscillators. Since stability of the averaged dynamics makes the initial conditions forgotten as time goes on, these algorithms are initialization-free and suitable for plug-and-play operation. At last, nonlinear couplings are also considered, which potentially asserts that enforced synchronization gives rise to an emergent behavior of a heterogeneous multi-agent system.

arXiv:2109.08248

Assessments of epistemic uncertainty using Gaussian stochastic weight averaging for fluid-flow regression

We use Gaussian stochastic weight averaging (SWAG) to assess the model-form uncertainty associated with neural-network-based function approximation relevant to fluid flows. SWAG approximates a posterior Gaussian distribution of each weight, given training data, and a constant learning rate. Having access to this distribution, it is able to create multiple models with various combinations of sampled weights, which can be used to obtain ensemble predictions. The average of such an ensemble can be regarded as the `mean estimation', whereas its standard deviation can be used to construct `confidence intervals', which enable us to perform uncertainty quantification (UQ) with regard to the training process of neural networks. We utilize representative neural-network-based function approximation tasks for the following cases: (i) a two-dimensional circular-cylinder wake; (ii) the DayMET dataset (maximum daily temperature in North America); (iii) a three-dimensional square-cylinder wake; and (iv) urban flow, to assess the generalizability of the present idea for a wide range of complex datasets. SWAG-based UQ can be applied regardless of the network architecture, and therefore, we demonstrate the applicability of the method for two types of neural networks: (i) global field reconstruction from sparse sensors by combining convolutional neural network (CNN) and multi-layer perceptron (MLP); and (ii) far-field state estimation from sectional data with two-dimensional CNN. We find that SWAG can obtain physically-interpretable confidence-interval estimates from the perspective of model-form uncertainty. This capability supports its use for a wide range of problems in science and engineering.

arXiv:2109.07982공식 코드 · Papers with Code

R3LIVE: A Robust, Real-time, RGB-colored, LiDAR-Inertial-Visual tightly-coupled state Estimation and mapping package

In this letter, we propose a novel LiDAR-Inertial-Visual sensor fusion framework termed R3LIVE, which takes advantage of measurement of LiDAR, inertial, and visual sensors to achieve robust and accurate state estimation. R3LIVE is contained of two subsystems, the LiDAR-inertial odometry (LIO) and visual-inertial odometry (VIO). The LIO subsystem (FAST-LIO) takes advantage of the measurement from LiDAR and inertial sensors and builds the geometry structure of (i.e. the position of 3D points) global maps. The VIO subsystem utilizes the data of visual-inertial sensors and renders the map's texture (i.e. the color of 3D points). More specifically, the VIO subsystem fuses the visual data directly and effectively by minimizing the frame-to-map photometric error. The developed system R3LIVE is developed based on our previous work R2LIVE, with careful architecture design and implementation. Experiment results show that the resultant system achieves more robustness and higher accuracy in state estimation than current counterparts (see our attached video). R3LIVE is a versatile and well-engineered system toward various possible applications, which can not only serve as a SLAM system for real-time robotic applications, but can also reconstruct the dense, precise, RGB-colored 3D maps for applications like surveying and mapping. Moreover, to make R3LIVE more extensible, we develop a series of offline utilities for reconstructing and texturing meshes, which further minimizes the gap between R3LIVE and various of 3D applications such as simulators, video games and etc (see our demos video). To share our findings and make contributions to the community, we open source R3LIVE on our Github, including all of our codes, software utilities, and the mechanical design of our device.

arXiv:2109.03962공식 코드 · Papers with Code

Learning-based Moving Horizon Estimation through Differentiable Convex Optimization Layers

To control a dynamical system it is essential to obtain an accurate estimate of the current system state based on uncertain sensor measurements and existing system knowledge. An optimization-based moving horizon estimation (MHE) approach uses a dynamical model of the system, and further allows for integration of physical constraints on system states and uncertainties, to obtain a trajectory of state estimates. In this work, we address the problem of state estimation in the case of constrained linear systems with parametric uncertainty. The proposed approach makes use of differentiable convex optimization layers to formulate an MHE state estimator for systems with uncertain parameters. This formulation allows us to obtain the gradient of a squared and regularized output error, based on sensor measurements and state estimates, with respect to the current belief of the unknown system parameters. The parameters within the MHE problem can then be updated online using stochastic gradient descent (SGD) to improve the performance of the MHE. In a numerical example of estimating temperatures of a group of manufacturing machines, we show the performance of tuning the unknown system parameters and the benefits of integrating physical state constraints in the MHE formulation.

arXiv:2109.04139

Robot Localization and Navigation through Predictive Processing using LiDAR

Knowing the position of the robot in the world is crucial for navigation. Nowadays, Bayesian filters, such as Kalman and particle-based, are standard approaches in mobile robotics. Recently, end-to-end learning has allowed for scaling-up to high-dimensional inputs and improved generalization. However, there are still limitations to providing reliable laser navigation. Here we show a proof-of-concept of the predictive processing-inspired approach to perception applied for localization and navigation using laser sensors, without the need for odometry. We learn the generative model of the laser through self-supervised learning and perform both online state-estimation and navigation through stochastic gradient descent on the variational free-energy bound. We evaluated the algorithm on a mobile robot (TIAGo Base) with a laser sensor (SICK) in Gazebo. Results showed improved state-estimation performance when comparing to a state-of-the-art particle filter in the absence of odometry. Furthermore, conversely to standard Bayesian estimation approaches our method also enables the robot to navigate when providing the desired goal by inferring the actions that minimize the prediction error.

arXiv:2109.04087공식 코드 · Papers with Code

Learning Cross-Scale Visual Representations for Real-Time Image Geo-Localization

Robot localization remains a challenging task in GPS denied environments. State estimation approaches based on local sensors, e.g. cameras or IMUs, are drifting-prone for long-range missions as error accumulates. In this study, we aim to address this problem by localizing image observations in a 2D multi-modal geospatial map. We introduce the cross-scale dataset and a methodology to produce additional data from cross-modality sources. We propose a framework that learns cross-scale visual representations without supervision. Experiments are conducted on data from two different domains, underwater and aerial. In contrast to existing studies in cross-view image geo-localization, our approach a) performs better on smaller-scale multi-modal maps; b) is more computationally efficient for real-time applications; c) can serve directly in concert with state estimation pipelines.

arXiv:2109.03604

Power to the Relational Inductive Bias: Graph Neural Networks in Electrical Power Grids

The application of graph neural networks (GNNs) to the domain of electrical power grids has high potential impact on smart grid monitoring. Even though there is a natural correspondence of power flow to message-passing in GNNs, their performance on power grids is not well-understood. We argue that there is a gap between GNN research driven by benchmarks which contain graphs that differ from power grids in several important aspects. Additionally, inductive learning of GNNs across multiple power grid topologies has not been explored with real-world data. We address this gap by means of (i) defining power grid graph datasets in inductive settings, (ii) an exploratory analysis of graph properties, and (iii) an empirical study of the concrete learning task of state estimation on real-world power grids. Our results show that GNNs are more robust to noise with up to 400% lower error compared to baselines. Furthermore, due to the unique properties of electrical grids, we do not observe the well known over-smoothing phenomenon of GNNs and find the best performing models to be exceptionally deep with up to 13 layers. This is in stark contrast to existing benchmark datasets where the consensus is that 2 to 3 layer GNNs perform best. Our results demonstrate that a key challenge in this domain is to effectively handle long-range dependence.

arXiv:2109.03561

A Computationally Efficient EK-PMBM Filter for Bistatic mmWave Radio SLAM

Millimeter wave (mmWave) signals are useful for simultaneous localization and mapping (SLAM), due to their inherent geometric connection to the propagation environment and the propagation channel. To solve the SLAM problem, existing approaches rely on sigma-point or particle-based approximations, leading to high computational complexity, precluding real-time execution. We propose a novel low-complexity SLAM filter, based on the Poisson multi-Bernoulli mixture (PMBM) filter. It utilizes the extended Kalman (EK) first-order Taylor series based Gaussian approximation of the filtering distribution, and applies the track-oriented marginal multi-Bernoulli/Poisson (TOMB/P) algorithm to approximate the resulting PMBM as a Poisson multi-Bernoulli (PMB). The filter can account for different landmark types in radio SLAM and multiple data association hypotheses. Hence, it has an adjustable complexity/performance trade-off. Simulation results show that the developed SLAM filter can greatly reduce the computational cost, while it keeps the good performance of mapping and user state estimation.

arXiv:2109.02767

A Recursive Delay Estimation Algorithm for Linear Multivariable Systems with Time-varying Delays

Time delay estimation plays a critical role in control, stabilization and state estimation of many practical system with time delay. In this paper, we propose a method to estimate delay for discrete time linear multiple-input multiple-output systems with time-varying input delays. This method is purposefully given for situations where only a limited amount of information is available for the system. Although, this approach is primarily developed in a deterministic framework, it can also be applied to noisy data under special circumstances. In addition, switched linear autoregressive models with exogenous inputs are introduced as possible applications of the presented algorithm provided that the switching frequencies are small. Finally, effectiveness of the algorithm is illustrated by two numerical examples.

arXiv:2109.02016

Guaranteed State Estimation via Indirect Polytopic Set Computation for Nonlinear Discrete-Time Systems

This paper proposes novel set-theoretic approaches for state estimation in bounded-error discrete-time nonlinear systems, subject to nonlinear observations/constraints. By transforming the polytopic sets that are characterized as zonotope bundles (ZB) and/or constrained zonotopes (CZ), from the state space to the space of the generators of ZB/CZ, we leverage a recent result on the remainder-form mixed-monotone decomposition functions to compute the propagated set, i.e., a ZB/CZ that is guaranteed to enclose the set of the state trajectories of the considered system. Further, by applying the remainder-form decomposition functions to the nonlinear observation function, we derive the updated set, i.e., an enclosing ZB/CZ of the intersection of the propagated set and the set of states that are compatible/consistent with the observations/constraints. Finally, we show that the mean value extension result in [1] for computing propagated sets can also be extended to compute the updated set when the observation function is nonlinear.

arXiv:2109.00538

Physics-integrated hybrid framework for model form error identification in nonlinear dynamical systems

For real-life nonlinear systems, the exact form of nonlinearity is often not known and the known governing equations are often based on certain assumptions and approximations. Such representation introduced model-form error into the system. In this paper, we propose a novel gray-box modeling approach that not only identifies the model-form error but also utilizes it to improve the predictive capability of the known but approximate governing equation. The primary idea is to treat the unknown model-form error as a residual force and estimate it using duel Bayesian filter based joint input-state estimation algorithms. For improving the predictive capability of the underlying physics, we first use machine learning algorithm to learn a mapping between the estimated state and the input (model-form error) and then introduce it into the governing equation as an additional term. This helps in improving the predictive capability of the governing physics and allows the model to generalize to unseen environment. Although in theory, any machine learning algorithm can be used within the proposed framework, we use Gaussian process in this work. To test the performance of proposed framework, case studies discussing four different dynamical systems are discussed; results for which indicate that the framework is applicable to a wide variety of systems and can produce reliable estimates of original system's states.

arXiv:2108.11558

Targeted False Data Injection Attacks Against AC State Estimation Without Network Parameters

State estimation is a data processing algorithm for converting redundant meter measurements and other information into an estimate of the state of a power system. Relying heavily on meter measurements, state estimation has proven to be vulnerable to cyber attacks. In this paper, a novel targeted false data injection attack (FDIA) model against AC state estimation is proposed. Leveraging on the intrinsic load dynamics in ambient conditions and important properties of the Ornstein-Uhlenbeck process, we, from the viewpoint of intruders, design an algorithm to extract power network parameters purely from PMU data, which are further used to construct the FDIA vector. Requiring no network parameters and relying only on limited phasor measurement unit (PMU) data, the proposed FDIA model can target specific states and launch large deviation attacks. Sufficient conditions for the proposed FDIA model are also developed. Various attack vectors and attacking regions are studied in the IEEE 39-bus system, showing that the proposed FDIA method can successfully bypass the bad data detection and launch targeted large deviation attacks with very high probabilities.

arXiv:2108.11311

A new state estimation approach-Adaptive Fading Cubature Kalman filter

This paper presents a novel adaptive fading cubature Kalman filter (AFCKF) based on double transitive factors. The developed adaptive algorithm is explained in two stages; stage (i) a single transitive factor is used to update the predicted state error covariance, ${\bf \hat P_{k}}^{-}$ based on innovation or residual vector, whereas, in stage (ii), the measurement noise covariance matrix, ${\bf \hat R_{k}^{*}}$ is scaled by another transitive factor. Furthermore, showing the proof concept for estimation of the process noise, ${\bf \hat Q_{k}^{*}}$ and measurement noise covariance matrices by combining the innovation and residual vector in the AFCKF algorithm. It can provide reliable state estimation in the presence of unknown noise statistics. Bench-marking target tracking example is considered to show the performance improvement of the developed algorithms. As compared with existing adaptive approaches, the proposed fading algorithm can provide better estimation results.

arXiv:2108.07698

An Experimental Urban Case Study with Various Data Sources and a Model for Traffic Estimation

Accurate estimation of the traffic state over a network is essential since it is the starting point for designing and implementing any traffic management strategy. Hence, traffic operators and users of a transportation network can make reliable decisions such as influence/change route or mode choice. However, the problem of traffic state estimation from various sensors within an urban environment is very complex for several different reasons, such as availability of sensors, different noise levels, different output quantities, sensor accuracy, heterogeneous data fusion, and many more. To provide a better understanding of this problem, we organized an experimental campaign with video measurement in an area within the urban network of Zurich, Switzerland. We focus on capturing the traffic state in terms of traffic flow and travel times by ensuring measurements from established thermal cameras by the city's authorities, processed video data, and the Google Distance Matrix. We assess the different data sources, and we propose a simple yet efficient Multiple Linear Regression (MLR) model to estimate travel times with fusion of various data sources. Comparative results with ground-truth data (derived from video measurements) show the efficiency and robustness of the proposed methodology.

arXiv:2108.07134공식 코드 · Papers with Code

Neural Predictive Monitoring under Partial Observability

We consider the problem of predictive monitoring (PM), i.e., predicting at runtime future violations of a system from the current state. We work under the most realistic settings where only partial and noisy observations of the state are available at runtime. Such settings directly affect the accuracy and reliability of the reachability predictions, jeopardizing the safety of the system. In this work, we present a learning-based method for PM that produces accurate and reliable reachability predictions despite partial observability (PO). We build on Neural Predictive Monitoring (NPM), a PM method that uses deep neural networks for approximating hybrid systems reachability, and extend it to the PO case. We propose and compare two solutions, an end-to-end approach, which directly operates on the rough observations, and a two-step approach, which introduces an intermediate state estimation step. Both solutions rely on conformal prediction to provide 1) probabilistic guarantees in the form of prediction regions and 2) sound estimates of predictive uncertainty. We use the latter to identify unreliable (and likely erroneous) predictions and to retrain and improve the monitors on these uncertain inputs (i.e., active learning). Our method results in highly accurate reachability predictions and error detection, as well as tight prediction regions with guaranteed coverage.

arXiv:2108.04591

Event-Triggered State Estimation with Multiple Noisy Sensor Nodes

General nonlinear continuous-time systems are considered for which its state is estimated via a packet-based communication network. We assume that the system has multiple sensor nodes, affected by measurement noise, which can transmit at discrete (non-equidistant) points in time. Moreover, each node can transmit asynchronously. For this setup, we develop a state estimation framework, where the transmission instances of the individual sensor nodes can be generated in either time-triggered or event-triggered fashions. In the latter case, we guarantee the absence of Zeno behavior by construction. It is shown that, under the provided design conditions, an input-to-state stability property is obtained for the estimation error with respect to the measurement noise and process disturbances and that the state is thus reconstructed asymptotically in the absence of noise. A numerical case study shows the strengths of the developed framework.