Data Mining-Based Cyber-Physical Attack Detection Tool for Attack-Resilient Adaptive Protective Relays

Protection systems based on optical fiber communication are considered to be one of the reliable and efficient protection systems. Besides, number of advantages like high reliability, less delay and latency and end to end encryption, optical fibers has certain drawbacks such as highly susceptible to breakage, complex and expensive installation. An ordinary wireless communication is not considered to be an alternate to the existing optical fiber based communication. Most importantly, the existing conventional wireless systems are susceptible to high electromagnetic interferences which also prohibit its large scale application. Therefore, this paper aims to present an efficient integrated protection system based on differential protection principals using wireless communication system which is the latest emerging and cutting edge 5G (URLLC). This system provides 1ms latency with 99.999% reliability. However, the wireless reliability can be compromised in cases such as in presence of jammers or any unusual events in space due to which GPS synchronization gets disturbed and causes delay in communication. Alternately, the system should have a reliable backup protection plan which can perform in case of primary protection failure. The backup protection is based on directional protection using power-line carrier communication. The simulation model is designed in Power System Computer Aided Design (PSCAD) tool, whereas, communication network is simulated using NETSIM 12.2 software. The throughput, delay, packet loss and jitter calculated in NETSIM are promising for differential protection to replace optical fibers for distribution grid protection. 5G based protection system is one of the most potential protection system and it is one of the important direction of protection in the near future. Additionally, high performance of primary protection using wireless communication is balanced by the reliable backup protection using wired PLC communication and this makes the proposed system cost effective, efficient and reliable protection system. Detailed simulation and results are discussed in the paper.

Critical infrastructure protection and national security are enhanced by the security and reliability of networks. Various types of information circulate on these networks, ranging in classification from open to closed. The consequences of cyberattacks on these networks can be severe, including reputational damage, financial loss, operational disruption and data leakage. Traditional security methods, such as firewalls and anti-virus software, are becoming less effective against modern and ever-changing cyber threats. As a result, powerful network intrusion detection systems (IDS) have become indispensable for proactive detection and mitigation of cyber attacks. Machine learning has become a viable method for creating adaptive intrusion detection tools that can detect new and complex attack patterns. By learning from huge labelled network traffic datasets, ML models can understand the subtle patterns and differentiating features of normal and abnormal or malicious traffic flows. This allows them to detect possible cyber threats and intrusions that traditional signature-based IDSs cannot detect. Extracting discriminative features and training appropriate classification models from such data is a challenging task. In the presented study, we analyse the effectiveness of ML algorithms for detecting cyberattacks, in particular distributed denial of service (DDoS) attacks, in network traffic data. In the presented study, a network attack detection system is developed using ML and deep learning (DL) models and experimented on the CICIDS2017 dataset. The main objectives of the study are to develop a strategy for extracting valuable information from raw network streams; to study the impact of data preparation on the false positive rate; and to conduct a comparative analysis of ML models for cyberattack detection. The main goal of the study is to provide an understanding of the development of a reliable adaptive network intrusion detection system using ML approaches that increase cybersecurity capabilities and protect against future cyberattacks.

Most of current products and models are poor at detecting novel attacks without an acceptable level of accuracy or false alarms. In order to figure out this problem, a network based intrusion detection system has been established, and many up-to-date attack tools are used to attack the network. On the basis of the intrusion experiment, 29 variables are chosen as intrusion features to characterize the status of network connection. At the same time, the rough sets theory is exploited as a detector of network connection. The experimental results indicate that the features extracted from network connection are good indicators of the status of network and the rough sets theory is powerful in multi-class classification as well as effective in unknown attack detection.

This paper is concerned with cyber attack detection problem in a platoon-based vehicular networked control system. In such a system, the information among vehicles is transmitted through a shared wireless communication network and also each vehicle has access to its own information measured by local sensors. These kind of systems are highly vulnerable to cyber attacks and therefore, cyber-security issues need to be properly addressed to ensure the safety of the systems. Among various cyber-security aspects, reliable attack detection is of utmost importance as the ability to detect cyber attacks in a timely manner can reduce the damage to the systems. Therefore, we present a cyber attack detection algorithm that is capable of detecting attacks violating both measurements and control command data. This algorithm is based on an ellipsoidal set-membership filtering approach which consists of two sets: prediction ellipsoid set and an estimation ellipsoid set calculated through updating the prediction ellipsoid set with the measurement data. The detection method depends on the existence of intersection between these two sets computed by the filter. Simulation results for some possible cyber attacks are provided to demonstrate the effectiveness of the proposed method.

Comparison of PDEM and MCS: Accuracy and efficiency

Reliable detection and precise volumetric quantification of brain metastases (BM) on MRI are essential for guiding treatment decisions. We evaluate the potential of CNNs for automated detection and quantification of BM by training models on 308 patients with BM (with a 4:1 split for training/testing) for automated volumetric assessment of contrast-enhancing tumors (CE) and non-enhancing FLAIR signal abnormality including edema (NEE). An independent test set of 30 patientswas used for external testing.We assessed performance case-wise for CE and NEE and lesion-wise for CE using the casewise/lesion-wise DICE-coefficient (C/L-DICE), positive predictive value (L-PPV) and sensitivity (C/L-Sensitivity). The performance of detecting CE lesions on the validation dataset was not significantly affected when evaluating different volumetric thresholds (0.001–0.2 cm 3 ; P = .2028). The median L-DICE and median C-DICE for CE lesions were 0.78 (IQR = 0.6–0.91) and 0.90 (IQR = 0.85–0.94) in the institutional as well as 0.79 (IQR = 0.67–0.82) and 0.84 (IQR = 0.76–0.89) in the external test dataset. The corresponding median L-Sensitivity and median L-PPV were 0.81 (IQR = 0.63–0.92) and 0.79 (IQR = 0.63–0.93) in the institutional test dataset, as compared to 0.85 (IQR = 0.76–0.94) and 0.76 (IQR = 0.68–0.88) in the external test dataset. The median C-DICE for NEE was 0.96 (IQR = 0.92–0.97) in the institutional test dataset as compared to 0.85 (IQR = 0.72–0.91) in the external test dataset. The developed ANN-based algorithm (publicly available at www.github.com/NeuroAI-HD/HD-BM) allows reliable detection and precise volumetric quantification of CE and NEE compartments in patients with BM. The full paper has been published at Neuro-Oncology Advances [1].

The realisation of an eventual low-voltage (LV) Smart Grid with a complete communication infrastructure is a gradual process. During this evolution the protection scheme of distribution networks should be continuously adapted and optimised to fit the protection and cost requirements at the time. This paper aims to review practices and research around the design of an effective, adaptive and economical distribution network protection scheme. The background of this topic is introduced and potential problems are defined from conventional protection theories and new Smart Grid technologies. Challenges are identified with possible solutions defined as a pathway to the ultimate flexible and reliable LV protection systems.

The spread of infectious diseases among aquaculture species has a serious impact on the aquaculture industry. Simple, specific and low-cost detection methods are urgently needed for early diagnosis and timely treatment, particularly for on-site identifying and tracking of pathogens. Vibrio splendidus (V. splendidus) is regarded as one of the main pathogenic bacteria causing skin ulcerative syndrome in cultured sea cucumbers, leading to massive mortality and severe economic losses. We herein present a microfluidic-based real-time fluorogenic loop-mediated isothermal amplification (LAMP) system for simple and reliable detection of V. splendidus. A LAMP primer set with six primers (arsB1) specifically targeting the arsB gene of V. splendidus was successfully designed and tested on the portable microfluidic system for the first time. Only a single step of sample loading using a pipette is required to fill an array of reaction wells (with 10 or 18 wells) in a disposable chip for multiplex detection. A dedicated plastic shell is then utilized to tightly seal the openings of the chip by buckling to prevent contamination and evaporation. Up to four chips (one sample per chip) can be held in the stand-alone and inexpensive microdevice simultaneously, enabling on-demand detection of multiple samples in a single run. Reproducible (relatively low intra- and inter-chip variability) and sensitive (as few as ∼20 CFU, Colony-Forming Units, per reaction well) on-chip arsB1-LAMP assay was demonstrated by using diluted lysate of V. splendidus. A linear standard curve (R2 > 0.98) was attained over the template concentration range of 5 × 103 to 5 × 106 CFU mL-1. V. splendidus can be detected in samples containing different bacteria, indicating the feasibility of the portable microfluidic LAMP system for parallel detection of multiple bacterial pathogens. The proposed on-chip LAMP assay is simple to operate, reliable for amplification, flexible in detection and cost-effective in instrumentation and testing, holding great potential for on-site rapid detection and routine monitoring of aquaculture pathogens.

Detection of evolving cyber attacks is a challenging task for conventional network intrusion detection techniques. Various supervised machine learning algorithms have been implemented in network intrusion detection systems. However, traditional algorithms require long training time and have high computational complexity. Therefore, we propose detection of denial of service cyber attacks in communication networks by employing the broad learning system (BLS) that requires shorter training time while achieving comparable performance. Because designing effective detection systems relies on training and test datasets that contain anomalous network traffic data, in this paper we evaluate the performance of various BLS models by using recently generated network intrusion datasets. The best accuracy and F-Score were often achieved using BLS with cascades while BLS with incremental learning usually required shorter training time.

Collapse pressure is an important part of calculating the lower limit of the safe mud density window, which is crucial for optimizing the well trajectory, designing the drilling fluid density, and ensuring drilling safety. However, collapse pressure prediction methods based on physical mechanisms are theoretically complex, computationally intensive, and slow. Data mining-based prediction methods often rely on conventional machine learning models, which suffer from low prediction accuracy, high data demand, and poor interpretability. In this paper, a novel hybrid-driven model combining mechanistic knowledge and machine learning methods is proposed, which has a faster computational speed in collapse pressure prediction compared with the traditional analytical model, and a better performance compared with the existing data-driven models. The model incorporates the stress transformation, rock strength criteria, and other knowledge to ensure the robustness and interpretability of this prediction model. The neural network structure and model hyperparameters are optimized using a Bayesian optimization algorithm. To consider the influence of uncertainty of input parameters on collapse pressure, the Monte Carlo method is used to quantify the influence of uncertainty of input parameters on collapse pressure prediction results based on the hybrid-driven prediction model, and the sensitivity of different input parameters to the outcomes is determined. The proposed model, tested on a test dataset, demonstrated high prediction accuracy and prediction stability with an average absolute error of 0.0037 g/cm3 and a root mean square error of 0.0104 g/cm3 for the collapse pressure equivalent density. Furthermore, a horizontal well was selected for validation, with predicted results exhibiting an average absolute error of only 0.0045 g/cm3 compared to the logging interpretation results, and a computational speed nearly 100 times faster than traditional analytical models. Three points with different stress conditions were selected on this well and their equivalent density of collapse pressure hemispheric projection maps were predicted, and the results were consistent with the actual results, indicating that the model can accurately capture the variation of collapse pressure with well inclination and azimuth. To quantify the effect of input parameter uncertainty on wellbore stability, the influence of input parameter uncertainty on the equivalent density of collapse pressure is simulated based on the above prediction model in combination with the Monte Carlo method, and the corresponding confidence intervals are given. The results found that the effect of uncertainty in ground stress on collapse pressure is relatively significant. In conclusion, the hybrid-driven model effectively integrates physical knowledge, enabling rapid and accurate prediction of collapse pressure in horizontal and inclined wells, offering an innovative approach for intelligent wellbore stability assessment and uncertainty analysis.

Accurate and near real-time estimates of electromechanical modes are of great importance since the modal damping is a key indicator of the stability of the power system. If the estimates of the electromechanical modes are to be useful, knowing the variability in the estimates is critically important. This paper presents a method of directly estimating the variance of each mode estimate in addition to estimating the frequency and damping of each mode in an online setting using a recursive maximum likelihood (RML) estimator. The variance estimates are achieved using two closed-form multidimensional Taylor series approximations, the details of which are fully derived here. The proposed method is validated using a Monte Carlo simulation with a low order model of the Western Electricity Coordinating Council (WECC) power system under both ambient and probing conditions, with multiple modes closely spaced in frequency, and is compared to the regularized robust recursive least squares (R3LS) method. It is also successfully applied to phasor measurement unit (PMU) data collected from the actual WECC system, also under both ambient and probing conditions.

Accurate estimation of lithium-ion battery life is essential to assure the reliable operation of the energy supply system. This study develops regression models for battery prognostics using statistical methods. The resultant regression models can not only monitor a battery’s degradation trend but also accurately predict its remaining useful life (RUL) at an early stage. Three sets of test data are employed in the training stage for regression models. Another set of data is then applied to the regression models for validation. The fully discharged voltage (Vdis) and internal resistance (R) are adopted as aging parameters in two different mathematical models, with polynomial and exponential functions. A particle swarm optimization (PSO) process is applied to search for optimal coefficients of the regression models. Simulations indicate that the regression models using Vdis and R as aging parameters can build a real state of health profile more accurately than those using cycle number, N. The Monte Carlo method is further employed to make the models adaptive. The subsequent results, however, show that this results in an insignificant improvement of the battery life prediction. A reasonable speculation is that the PSO process already yields the major model coefficients.

A highly sensitive and reliable tunable diode laser absorption spectroscopy gas detection system with a temperature-pressure compensation algorithm is demonstrated for detecting C H 4 concentrations in near space. Near space generally refers to the airspace 20-100km away from the ground, where temperature and pressure changes are complex. Since the gas absorption spectrum is easily affected by temperature and pressure, a temperature-pressure compensation algorithm is proposed and used in the C H 4 sensor to improve the detection accuracy of the sensor. First, we measured the basic characteristics of the sensor in the laboratory, such as linearity and long-term stability. Experimental results showed that the linear correlation coefficient R-square can reach 0.999, and the concentration fluctuation of C H 4 is less than 0.17ppm within 3.5h. Then the sensor was applied to a research activity in Qinghai Province, China, in September, and the results show that the sensor can effectively monitor the C H 4 concentration in near space.

A physiological measures-based method for detecting inattention in drivers using machine learning approach

Purpose:Currently, proton and carbon ions are used for cancer treatment. More recently, other light ions including helium ions have shown interesting physical and biological properties. The purpose of this work is to study the biological and physical properties of helium ions (He‐3) in comparison to protons.Methods:Monte Carlo simulations with FLUKA, GEANT4 and MCNPX were used to calculate proton and He‐3 dose distributions in water phantoms. The energy spectra of proton and He‐3 beams were calculated with high resolution for use in biological models. The repair‐misrepairfixation (RMF) model was subsequently used to calculate the RBE.Results:The proton Bragg curve calculations show good agreement between the three general purpose Monte Carlo codes. In contrast, the He‐3 Bragg curve calculations show disagreement (for the magnitude of the Bragg peak) between FLUKA and the other two Monte Carlo codes. The differences in the magnitude of the Bragg peak are mainly due to the discrepancy in the secondary fragmentation cross sections used by the codes. The RBE for V79 cell lines is about 0.96 and 0.98 at the entrance of proton and He‐3 ions depth dose respectively. The RBE increases to 1.06 and 1.59 at the Bragg peak of proton and He‐3 ions. The results demonstrated that LET, microdosimetric parameters (such as dose‐mean lineal energy) and RBE are nearly constant along the plateau region of Bragg curve, while all parameters increase within the Bragg peak and at the distal edge for both proton and He‐3 ions.Conclusion:The Monte Carlo codes should revise the fragmentation cross sections to more accurately simulate the physical properties of He‐3 ions. The increase in RBE for He‐3 ions is higher than for proton beams at the Bragg peak.