SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Topological Data Analysis for Adversarial Detection (LANL O4937) - Detects adversarial examples in vision-language models using persistent homology and two-sample testing. Combines TDA features from CLIP embeddings with statistical methods (ME, SCF, SAMMD, C2ST) for robust detection across ImageNet, CIFAR-10/100.
Resilient observer design for Cyber-Physical Systems (CPS) in the presence of adversarial false data injection attacks (FDIA) is an active area of research. The existing state-of-the-art algorithms tend to break down as more and more knowledge of the system is built into the attack model; also as the percentage of attacked nodes increases. From the view of optimization theory, the problem is often cast as a classical error correction problem for which a theoretical limit of has been established as the maximum percentage attacked nodes for which state recovery is guaranteed. Beyond this limit, the performance of -minimization based schemes, for instance, deteriorates rapidly. Similar performance degradation occurs for other types of resilient observers beyond certain percentages of attacked nodes. In order to increase the corresponding percentage of attacked nodes for which state recoveries can be guaranteed, researchers have begun to incorporate prior information into the underlying resilient observer design framework. For the most pragmatic cases, this prior information is often obtained through a data-driven machine learning process. Existing results have shown a strong positive correlation between the maximum attacked percentages that can be tolerated and the accuracy of the data-driven model. Motivated by these results, this chapter examines the case for pruning algorithms designed to improve the Positive Prediction Value (PPV) of the resulting prior information, given stochastic uncertainty characteristics of the underlying machine learning model. Theoretical quantification of the achievable improvement is given. Simulation results show that the pruning algorithm significantly increases the maximum correctable percentage of attacked nodes, even for machine learning model whose prediction power is comparable to the random flip of a coin.
The present volume on applications of artificial intelligence with regard to knowledge-based systems in aerospace and industry discusses machine learning and clustering, expert systems and optimization techniques, monitoring and diagnosis, and automated design and expert systems. Attention is given to the integration of AI reasoning systems and hardware description languages, care-based reasoning, knowledge, retrieval, and training systems, and scheduling and planning. Topics addressed include the preprocessing of remotely sensed data for efficient analysis and classification, autonomous agents as air combat simulation adversaries, intelligent data presentation for real-time spacecraft monitoring, and an integrated reasoner for diagnosis in satellite control. Also discussed are a knowledge-based system for the design of heat exchangers, reuse of design information for model-based diagnosis, automatic compilation of expert systems, and a case-based approach to handling aircraft malfunctions.
High impedance faults (HIFs) have always been significant challenge in the power grids. Researchers have developed some advanced protective methods to detect the HIFs. To test and validate these methods, large amounts of HIF data are required. This paper presents a synthetic HIF data generating method using the deep convolutional generated adversarial network (DCGAN). The DCGAN includes a generator module to create synthetic HIF waveform from random noises; and a discriminator module to identify the flaws of those synthetic data, which ultimately help improve the quality of the synthetic data created by the generator. To test the fidelity of the generated synthetic HIF data, two different HIF-detection methods have been applied. Extensive simulation results have validated the effectiveness of using the DCGAN to create synthetic HIF data.
Various Internet of Things (IoT) devices generate complex, dynamically changed, and infinite data streams. Adversaries can cause harm if they can access the user’s sensitive raw streaming data. For this reason, protecting the privacy of the data streams is crucial. In this paper, we explore local differential privacy techniques for streaming data. We compare the techniques and report the advantages and limitations. We also present the effect on component (e.g., smoother, perturber) variations of distribution-based local differential privacy. We find that combining distribution-based noise during perturbation provides more flexibility to the interested entity.
The boom in single-cell technologies has brought a surge of high dimensional data that come from different sources and represent cellular systems from different views. With advances in these single-cell technologies, integrating single-cell data across modalities arises as a new computational challenge. Here, we present an adversarial approach, sciCAN, to integrate single-cell chromatin accessibility and gene expression data in an unsupervised manner. We benchmarked sciCAN with 5 existing methods in 5 scATAC-seq/scRNA-seq datasets, and we demonstrated that our method dealt with data integration with consistent performance across datasets and better balance of mutual transferring between modalities than the other 5 existing methods. We further applied sciCAN to 10X Multiome data and confirmed that the integrated representation preserves biological relationships within the hematopoietic hierarchy. Finally, we investigated CRISPR-perturbed single-cell K562 ATAC-seq and RNA-seq data to identify cells with related responses to different perturbations in these different modalities.
Presently the training of neural networks relies on data of comparable accuracy but in real applications only a very small set of high-fidelity data is available while inexpensive lower fidelity data may be plentiful. We propose a new composite neural network (NN) that can be trained based on multi-fidelity data. It is comprised of three NNs, with the first NN trained using the low-fidelity data and coupled to two high-fidelity NNs, one with activation functions and another one without, in order to discover and exploit nonlinear and linear correlations, respectively, between the low-fidelity and the high-fidelity data. We first demonstrate the accuracy of the new multi-fidelity NN for approximating some standard benchmark functions but also a 20-dimensional function that is not easy to approximate with other methods, e.g. Gaussian process regression. Subsequently, we extend the recently developed physics-informed neural networks (PINNs) to be trained with multi-fidelity data sets (MPINNs). MPINNs contain four fully-connected neural networks, where the first one approximates the low-fidelity data, while the second and third construct the correlation between the low- and high-fidelity data and produce the multi-fidelity approximation, which is then used in the last NN that encodes the partial differential equations (PDEs). Specifically, by decomposing the correlation into a linear and nonlinear part, the present model is capable of learning both the linear and complex nonlinear correlations between the low- and high-fidelity data adaptively. By training the MPINNs, we can: (1) obtain the correlation between the low- and high-fidelity data, (2) infer the quantities of interest based on a few scattered data, and (3) identify the unknown parameters in the PDEs. In particular, we employ the MPINNs to learn the hydraulic conductivity field for unsaturated flows as well as the reactive models for reactive transport. The results demonstrate that MPINNs can achieve relatively high accuracy based on a very small set of high-fidelity data. Despite the relatively low dimension and limited number of fidelities (two-fidelity levels) for the benchmark problems in the present study, the proposed model can be readily extended to very high-dimensional regression and classification problems involving multi-fidelity data.
The availability of fine grained time series data is a pre-requisite for research in smart-grids. While data for transmission systems is relatively easily obtainable, issues related to data collection, security and privacy hinder the widespread public availability/accessibility of such datasets at the distribution system level. This has prevented the larger research community from effectively applying sophisticated machine learning algorithms to significantly improve the distribution-level accuracy of predictions and increase the efficiency of grid operations. Synthetic dataset generation has proven to be a promising solution for addressing data availability issues in various domains such as computer vision, natural language processing and medicine. However, its exploration in the smart grid context remains unsatisfactory. Previous works have tried to generate synthetic datasets by modeling the underlying system dynamics: an approach which is difficult, time consuming, error prone and often times infeasible in many problems. In this work, we propose a novel data-driven approach to synthetic dataset generation by utilizing deep generative adversarial networks (GAN) to learn the conditional probability distribution of essential features in the real dataset and generate samples based on the learned distribution. To evaluate our synthetically generated dataset, we measure the maximum mean discrepancy (MMD) between real and synthetic datasets as probability distributions, and show that their sampling distance converges. To further validate our synthetic dataset, we perform common smart grid tasks such as k-means clustering and short-term prediction on both datasets. Experimental results show the efficacy of our synthetic dataset approach: the real and synthetic datasets are indistinguishable by solely examining the output of these tasks.
As power delivery systems evolve and become increasingly reliant on accurate forecasts, they will be more and more vulnerable to cybersecurity issues. A coordinated data attack by sophisticated adversaries can render existing data corrupt or outlier detection methods ineffective. This would have a very negative impact on operational decisions. The focus of this paper is to develop descriptive analytics-based methods for anomaly detection to protect the load forecasting process against cyberattacks to essential data. We propose an integrated solution (IS) and a hybrid implementation of IS (HIIS) that can detect and mitigate cyberattack induced long sequence anomalies. HIIS is also capable of improving true positive rates and reducing false positive rates significantly comparing with IS. The proposed HIIS can serve as an online cybersecure load forecasting scheme.
The implications are explored of the specific decision context approach to anticipatory project assessment. More specifically, it is hypothesized that with respect to any given effect of a proposed project or action (mobility, job opportunities, air pollution, population distribution, etc.) such effect will likely differ in probability and/or magnitude from one decisional context to another; that the social desirability or undesirability of a given effect is a function (will differ with) each specific decisional context; that therefore the social impact of such effect will in all likelihood differ with each specific decisional context; and that the social significance of even the dame social impact of a given effect will vary from one decisional context to another when such social impact interacts with (competes with or reinforces) the social impacts of other effects. It also follows from this analysis that the respective roles of scientific method (demonstrable data) and adversarial system will not only differ with each specific decisional context but with each alternative course of action available to the decisional entity in each specific context.
We present NNREPAIR, a constraint-based technique for repairing neural network classifiers by i) fixing the logic of the network at an intermediate layer, and ii) fixing the decision conditions of the network at the last layer. NNREPAIR first uses fault localization to find potentially faulty network parameters (specifically weights) and then performs repair using constraint solving to apply small modifications to remedy the defects. We present novel strategies to enable precise yet efficient repair such as inferring correctness specifications to act as oracles for intermediate layer repair, and generation of experts wrt each class. We demonstrate the technique in the context of three different scenarios: (1) Improving the over all accuracy of a model, (2) Fixing security vulnerabilities caused by poisoning of training data and (3) Improving the robustness of the network against adversarial attacks. Our evaluation on MNIST and CIFAR-10 models shows that NNREPAIR can improve the accuracy by 45.56% on poisoned data and 11.11% on adversarial data. NNREPAIR also provides improvement, albeit modest, in the overall accuracy of models, without requiring new data or re-training.
The accurate prediction of photovoltaic (PV) power generation is significant to ensure the economic and safe operation of power systems. To this end, the paper proposes a new digital twin (DT) empowered PV power prediction framework that is capable of ensuring reliable data transmission and employing the DT to achieve high accuracy of power prediction. With this framework, considering potential data contamination in the collected PV data, a generative adversarial network is employed to restore the historical data set, which offers a prerequisite to ensure accurate mapping from the physical space to the digital space. Further, a new DT empowered PV power prediction method is proposed. Therein, we model a DT that encompasses a digital physical model for reflecting the physical operation mechanism and a neural network model (i.e., a parallel network of convolution and bidirectional long-short-term memory model) for capturing the hidden spatial-temporal features. The proposed method enables the use of the DT to take advantages of the digital physical model and the neural network model, resulting in enhanced prediction accuracy. Finally, a real data set is conducted to access the effectiveness of the proposed method.
Recent year has brought considerable advancements in Electric Vehicles (EVs) and associated infrastructures/ communications. Intrusion Detection Systems (IDS) are widely deployed for anomaly detection in such critical infrastructures. This paper presents an Interpretable Anomaly Detection System (RX-ADS) for intrusion detection in CAN protocol communication in EVs. Contributions include: 1) window based feature extraction method; 2) deep Autoencoder based anomaly detection method; and 3) adversarial machine learning based explanation generation methodology. The presented approach was tested on two benchmark CAN datasets: OTIDS and Car Hacking. The anomaly detection performance of RX-ADS was compared against the state-of-the-art approaches on these datasets: HIDS and GIDS. The RX-ADS approach presented performance comparable to the HIDS approach (OTIDS dataset) and has outperformed HIDS and GIDS approaches (Car Hacking dataset). Further, the proposed approach was able to generate explanations for detected abnormal behaviors arising from various intrusions. Furthermore, these explanations were later validated by information used by domain experts to detect anomalies. Other advantages of RX-ADS include: 1) the method can be trained on unlabeled data; 2) explanations help experts in understanding anomalies and root course analysis, and also help with AI model debugging and diagnostics, ultimately improving user trust in AI systems.
ABSTRACT There are a number of hypotheses underlying the existence of adversarial examples for classification problems. These include the high‐dimensionality of the data, the high codimension in the ambient space of the data manifolds of interest, and that the structure of machine learning models may encourage classifiers to develop decision boundaries close to data points. This article proposes a new framework for studying adversarial examples that does not depend directly on the distance to the decision boundary. Similarly to the smoothed classifier literature, we define a (natural or adversarial) data point to be ( γ , σ)‐stable if the probability of the same classification is at least for points sampled in a Gaussian neighborhood of the point with a given standard deviation . We focus on studying the differences between persistence metrics along interpolants of natural and adversarial points. We show that adversarial examples have significantly lower persistence than natural examples for large neural networks in the context of the MNIST and ImageNet datasets. We connect this lack of persistence with decision boundary geometry by measuring angles of interpolants with respect to decision boundaries. Finally, we connect this approach with robustness by developing a manifold alignment gradient metric and demonstrating the increase in robustness that can be achieved when training with the addition of this metric.
Many approaches at the forefront of structural health monitoring rely on cutting-edge techniques from the field of machine learning. Recently, much interest has been directed towards the study of so-called adversarial examples; deliberate input perturbations that deceive machine learning models while remaining semantically identical. This article demonstrates that data-driven approaches to structural health monitoring are vulnerable to attacks of this kind. In the perfect information or ‘white-box’ scenario, a transformation is found that maps every example in the Los Alamos National Laboratory three-storey structure dataset to an adversarial example. Also presented is an adversarial threat model specific to structural health monitoring. The threat model is proposed with a view to motivate discussion into ways in which structural health monitoring approaches might be made more robust to the threat of adversarial attack.
In overhead image segmentation tasks, including additional spectral bands beyond the traditional RGB channels can improve model performance. However, it is still unclear how incorporating this additional data impacts model robustness to adversarial attacks and natural perturbations. For adversarial robustness, the additional in-formation could improve the model’s ability to distinguish malicious inputs, or simply provide new attack avenues and vulnerabilities. For natural perturbations, the additional information could better inform model decisions and weaken perturbation effects or have no significant influence at all. In this work, we seek to characterize the performance and robustness of a multispectral (RGB and near infrared) image segmentation model subjected to adversarial attacks and natural perturbations. While existing adversarial and natural robustness research has focused primarily on digital perturbations, we prioritize on creating realistic perturbations designed with physical world conditions in mind. For adversarial robustness, we focus on data poisoning attacks whereas for natural robustness, we focus on extending ImageNet-C common corruptions for fog and snow that coherently and self-consistently perturbs the input data. Overall, we find both RGB and multispectral models are vulnerable to data poisoning attacks regardless of input or fusion architectures and that while physically-realizable natural perturbations still degrade model performance, the impact differs based on fusion architecture and input data.
With the ever-increasing reliance on data for data-driven applications in power grids, such as event cause analysis, the authenticity of data streams has become crucially important. The data can be prone to adversarial stealthy attacks aiming to manipulate the data such that residual-based bad data detectors cannot detect them, and the perception of system operators or event classifiers changes about the actual event. This paper investigates the impact of adversarial attacks on convolutional neural network-based event cause analysis frameworks. We have successfully verified the ability of adversaries to maliciously misclassify events through stealthy data manipulations. The vulnerability assessment is studied with respect to the number of compromised measurements. Furthermore, a defense mechanism to robustify the performance of the event cause analysis is proposed. The effectiveness of adversarial attacks on changing the output of the framework is studied using the data generated by real-time digital simulator (RTDS) under different scenarios such as type of attacks and level of access to data.