Sensor Observations Research Articles

Ultra-Short-Term Wind Power Forecasting Based on CGAN-CNN-LSTM Model Supported by Lidar.

Accurate prediction of wind power is of great significance to the stable operation of the power system and the vigorous development of the wind power industry. In order to further improve the accuracy of ultra-short-term wind power forecasting, an ultra-short-term wind power forecasting method based on the CGAN-CNN-LSTM algorithm is proposed. Firstly, the conditional generative adversarial network (CGAN) is used to fill in the missing segments of the data set. Then, the convolutional neural network (CNN) is used to extract the eigenvalues of the data, combined with the long short-term memory network (LSTM) to jointly construct a feature extraction module, and add an attention mechanism after the LSTM to assign weights to features, accelerate model convergence, and construct an ultra-short-term wind power forecasting model combined with the CGAN-CNN-LSTM. Finally, the position and function of each sensor in the Sole du Moulin Vieux wind farm in France is introduced. Then, using the sensor observation data of the wind farm as a test set, the CGAN-CNN-LSTM model was compared with the CNN-LSTM, LSTM, and SVM to verify the feasibility. At the same time, in order to prove the universality of this model and the ability of the CGAN, the model of the CNN-LSTM combined with the linear interpolation method is used for a controlled experiment with a data set of a wind farm in China. The final test results prove that the CGAN-CNN-LSTM model is not only more accurate in prediction results, but also applicable to a wide range of regions and has good value for the development of wind power.

Cropland Extraction in Southern China from Very High-Resolution Images Based on Deep Learning

Accurate cropland information is crucial for the assessment of food security and the formulation of effective agricultural policies. Extracting cropland from remote sensing imagery is challenging due to spectral diversity and mixed pixels. Recent advances in remote sensing technology have facilitated the availability of very high-resolution (VHR) remote sensing images that provide detailed ground information. However, VHR cropland extraction in southern China is difficult because of the high heterogeneity and fragmentation of cropland and the insufficient observations of VHR sensors. To address these challenges, we proposed a deep learning-based method for automated high-resolution cropland extraction. The method used an improved HRRS-U-Net model to accurately identify the extent of cropland and explicitly locate field boundaries. The HRRS-U-Net maintained high-resolution details throughout the network to generate precise cropland boundaries. Additionally, the residual learning (RL) and the channel attention mechanism (CAM) were introduced to extract deeper discriminative representations. The proposed method was evaluated over four city-wide study areas (Qingyuan, Yangjiang, Guangzhou, and Shantou) with a diverse range of agricultural systems, using GaoFen-2 (GF-2) images. The cropland extraction results for the study areas had an overall accuracy (OA) ranging from 97.00% to 98.33%, with F1 scores (F1) of 0.830–0.940 and Kappa coefficients (Kappa) of 0.814–0.929. The OA was 97.85%, F1 was 0.915, and Kappa was 0.901 over all study areas. Moreover, our proposed method demonstrated advantages compared to machine learning methods (e.g., RF) and previous semantic segmentation models, such as U-Net, U-Net++, U-Net3+, and MPSPNet. The results demonstrated the generalization ability and reliability of the proposed method for cropland extraction in southern China using VHR remote images.

An Earth Observation Task Representation Model Supporting Dynamic Demand for Flood Disaster Monitoring and Management

A comprehensive, accurate, and timely expression of earth observation (EO) tasks is the primary prerequisite for the response to and the emergency monitoring of disasters, especially floods. However, the existing information model does not fully satisfy the demand for a fine-grain observation expression of EO task, which results in the absence of task process management. The current study proposed an EO task representation model based on meta-object facility to address this problem. The model not only describes the static information of a task, but it also defines the dynamics of an observation task by introducing a functional metamodel. This metamodel describes the full life cycle of a task; it comprises five process methods: birth, separation, combination, updating, and extinction. An earth observation task modeling and management prototype system (EO-TMMS) for conducting a remote sensing satellite sensor observation task representation experiment on flooding was developed. In accordance with the results, the proposed model can describe various EO tasks demands and the full life cycle process of an EO task. Compared with other typical observation task information models, the proposed model satisfies the dynamic and fine-grain process representation of EO tasks, which can improve the efficiency of EO sensor utilization.

Anatomy of the Tsunami and Lamb Waves-Induced Ionospheric Signatures Generated by the 2022 Hunga Tonga Volcanic Eruption

As tsunamis propagate across open oceans, they remain largely unseen due to the lack of adequate sensors. To address this fundamental limitation of existing tsunami warnings, we investigate Global Navigation Satellite Systems (GNSS) data to monitor the ionosphere Total Electron Content (TEC) for Traveling Ionospheric Disturbances (TIDs) created by tsunami-induced internal gravity waves (IGWs). The approach has been applied to regular tsunamis generated by earthquakes, while the case of undersea volcanic eruptions injecting energy into both the ocean and the atmosphere remains mostly unexplored. With both a regular tsunami and air-sea waves, the large 2022 Hunga Tonga-Hunga Ha’apai volcanic eruption is a challenge. Here, we show that even in near-field regions (1000–1500 km), despite the complex wavefield, we can isolate the regular tsunami signature. We also highlight that the eruption-generated Lamb wave induces an ionospheric disturbance with a similar waveform and an amplitude spatial pattern consistent with IGW origin but with a quasi-constant propagation speed (~ 315 m/s). These results imply that when GNSS-TEC measurements are registered near an ocean bottom pressure sensor, they can help discriminating the regular tsunami from the initial air-sea waves appearing in the sensor observations.

Cross-calibration method based on an automated observation site.

Cross-calibration methods are widely used in high-precision remote sensor calibrations and ensure observational consistency between sensors. Because two sensors must be observed under the same or similar conditions, the cross-calibration frequency is greatly reduced; performing cross-calibrations on Aqua/Terra MODIS, Sentinel-2A/Sentinel-2B MSI and other similar sensors is difficult due to synchronous-observation limitations. Additionally, few studies have cross-calibrated water-vapor-observation bands sensitive to atmospheric changes. In recent years, standard automated observation sites and unified processing technology networks, such as an Automated Radiative Calibration Network (RadCalNet) and an automated vicarious calibration system (AVCS), have provided automatic observation data and means for independently, continuously monitoring sensors, thus offering new cross-calibration references and bridges. We propose an AVCS-based cross-calibration method. By limiting the observational-condition differences when two remote sensors transit over wide temporal ranges through AVCS observation data, we improve the cross-calibration opportunity. Thereby, cross-calibrations and observation consistency evaluations between the abovementioned instruments are realized. The influence of AVCS-measurement uncertainties on the cross-calibration is analyzed. The consistency between the MODIS cross-calibration and sensor observation is within 3% (5% in SWIR bands); that for the MSI is within 1% (2.2% in the water-vapor-observation band); and for the cross-calibration of Aqua MODIS and the two MSI, the consistency between the cross-calibration-predicted TOA reflectance and the sensor-measured TOA reflectance was within 3.8%. Thus, the absolute AVCS-measurement uncertainty is also reduced, especially in the water-vapor-observation band. This method can be applied to cross-calibrations and measurement consistency evaluations of other remote sensors. Later, the spectral-difference influences on cross-calibrations will be further studied.

A Variational Bayesian Blind Calibration Approach for Air Quality Sensor Deployments

Air pollution has become a global threat to urban environments and public health. Low-cost air quality sensor systems have been deployed to support fine-grained monitoring, and in-field calibration methods are necessary to assure the accuracy of sensor observation. Existing methods use in-field reference to train some specific calibration models. Nevertheless, collecting sufficient reference data after deployment is challenging. First, the synchronized data pair of reference observation is hard to obtain. Thus, many blind calibration approaches rely on some alternative reference like a historical reference before sensor observation. However, since the relationship model between alternative reference and sensor observation becomes more complicated, the amount of collected data is still insufficient to learn the complex model. To address the above challenge, we propose a Variational Bayesian Blind Calibration Algorithm. Our method introduces a supplement set that only involves historical reference and synchronized reference measurement without sensor observation to alleviate the few-reference pressure. Large amounts of the introduced supplement sets have been collected in different cities by accurate stations, which we adopt to conduct prediction tasks and build a Bayesian model to learn how to combine the formulated prediction task and target calibration task in a theoretically better way. Furthermore, we design a variational Bayesian framework to make good use of the easily obtained supplement set to train a better prediction model for improving calibration performance and relieving the pressure of collecting costly reference measurements after deployment. Evaluations of real-world and synthetic datasets show that the proposed approach has a better performance than previous baselines.

Sensitivity tests performed on passive low frequency seismic (LFS) data acquired in the Central North Sea to delineate hydrocarbon deposits

In 2018 TenzorGEO carried out a trial in the North Sea to acquire passive seismic data (LFS – low frequency seismic technology) as part of a blind test with an unnamed operator. The result from the trial is a hydrocarbon probability map showing data informed zones with delineated hydrocarbons. To further test the acquired data and the level of sensitivity of the model used in the analysis, a sensitivity test was performed to assess the level of model misfit and validate the vertical resolution of the technology. The results from the sensitivity tests was able to show how the hydrocarbon probability map will respond to: changes in depth, presence of other hydrocarbon layers within the study area, and the exclusion of sensor observation points should the sensors be affected or damaged during installation. Overall, the tests were able to validate the capability and potential of the LFS technology in identifying and delineating multiple hydrocarbon reservoirs with a predetermined interval spacing between the reservoirs across a field. The sensitivity tests provide an opportunity to test and fine-tune the application of LFS in detecting and delineating deeper structures with potential hydrocarbons and also monitoring CO2 storage reservoirs.

Real-time detection of deception attacks in cyber-physical systems

AbstractDetection of deception attacks is pivotal to ensure the safe and reliable operation of cyber-physical systems (CPS). Detection of such attacks needs to consider time-series sequences and is very challenging especially for autonomous vehicles that rely on high-dimensional observations from camera sensors. The paper presents an approach to detect deception attacks in real-time utilizing sensor observations, with a special focus on high-dimensional observations. The approach is based on inductive conformal anomaly detection (ICAD) and utilizes a novel generative model which consists of a variational autoencoder (VAE) and a recurrent neural network (RNN) that is used to learn both spatial and temporal features of the normal dynamic behavior of the system. The model can be used to predict the observations for multiple time steps, and the predictions are then compared with actual observations to efficiently quantify the nonconformity of a sequence under attack relative to the expected normal behavior, thereby enabling real-time detection of attacks using high-dimensional sequential data. We evaluate the approach empirically using two simulation case studies of an advanced emergency braking system and an autonomous car racing example, as well as a real-world secure water treatment dataset. The experiments show that the proposed method outperforms other detection methods, and in most experiments, both false positive and false negative rates are less than 10%. Furthermore, execution times measured on both powerful cloud machines and embedded devices are relatively short, thereby enabling real-time detection.

Assessment of the impact of waste fires on air quality and atmospheric aerosol optical depth: A case study in Poland

In this study the impact on air quality and on atmospheric aerosol optical depth (AOD) of selected waste fires in Poland in the last few years was assessed using measurement results derived from the public air quality monitoring (AQM) system and satellite observations of the MODIS sensor. Additional meteorological data from nearest ground stations as well as results of HYSPLIT forward and backward modeling and high-resolution satellite imagery were used to initially identify locations potentially exposed to smoke from the fires and hours of exposure. The concentrations of selected air pollutants (PM10, PM2.5, SO2, NO2, NOx, CO, and C6H6) were treated as characteristic substances of waste fires. Depending on the location of a given AQM station in relation to the place of waste fire and the prevailing meteorological conditions, a more or less significant air concentration peaks (lasting at least several hours) of the analyzed substances were observed. The study showed that the fumes from large waste fires spread over a long distance, so that the increase in air pollutant concentrations can be also recorded at AQM stations located further away (approx. 200–300 km) from the fire site. In the layer of the atmosphere where the smoke is transported, its influence on the AOD data from satellite observations was also visible.

Super-resolution and uncertainty estimation from sparse sensors of dynamical physical systems

The goal of this study is to leverage emerging machine learning (ML) techniques to develop a framework for the global reconstruction of system variables from potentially scarce and noisy observations and to explore the epistemic uncertainty of these models. This work demonstrates the utility of exploiting the stochasticity of dropout and batch normalization schemes to infer uncertainty estimates of super-resolved field reconstruction from sparse sensor measurements. A Voronoi tessellation strategy is used to obtain a structured-grid representation from sensor observations, thus enabling the use of fully convolutional neural networks (FCNN) for global field estimation. An ensemble-based approach is developed using Monte-Carlo batch normalization (MCBN) and Monte-Carlo dropout (MCD) methods in order to perform approximate Bayesian inference over the neural network parameters, which facilitates the estimation of the epistemic uncertainty of predicted field values. We demonstrate these capabilities through numerical experiments that include sea-surface temperature, soil moisture, and incompressible near-surface flows over a wide range of parameterized flow configurations.

Pecora 22: Remote Sensing for Freshwater and Marine Environments

Pecora 22: Remote Sensing for Freshwater and Marine Environments

Protocol-Based Fuzzy Control of Networked Systems Under Joint Deception Attacks

The security control issue of nonlinear networked systems is considered in this article on the basis of interval type-2 fuzzy modeling strategy. For avoiding communication congestion, a stochastic scheduling strategy called Markovian communication protocol is introduced to coordinate the sensor transmission order. An asynchronous observer is designed for estimating the unmeasured states via the hidden Markov model. In addition, a more comprehensive scenario on deception attack is considered, in which attacks occur both in the sensor–observer and controller–actuator communication channels with different types of deception signals. In view of slack matrix approach and stochastic analysis technique, some sufficient conditions for ensuring the ultimately boundedness of the resulting closed-loop system are obtained. In the end, simulations show the validity of the proposed protocol-based fuzzy control method.

Multisource information fusion-based environment perception and dynamic model of underwater vehicle in irregular ocean environment

The irregular ocean environment brings difficulties to observation and affects the movement state of underwater vehicles. Multisource information fusion can provide more reliable results on information-rich datasets. This paper proposes a multisource information fusion-based environment perception and dynamics model for underwater vehicles in irregular ocean environments. First, the multisource observation data is decomposed by the latent feature disentanglement learning model into the latent stable and dynamical features. The dynamical feature is predicted by the retrospect regression method and then integrated with the latent stable feature to predict the ocean field. After that, the environment-vehicle coupling effect is explored by thermodynamics and statics simulation. Then, a real-time dynamic model of the underwater vehicle is constructed by combining multisource information from physical principles, environmental perception, and sensor observations. Finally, extensive experiments are performed on a novel underwater glider and a publicly available South China Sea dataset to verify the proposed method.

Efficient iteratively reweighted algorithms for robust hyperbolic localization

Hyperbolic localization utilizes the time-difference-of-arrival (TDOA) measurements obtained from multiple coordinated sensors to locate an emitting source. A challenge in such an area, is that adverse factors like the non-line-of-sight propagation in typical environments may lead to outlying sensor observations and therefore deteriorated positioning performance. To mitigate the negative effects of outliers in the raw TDOA-based range-difference (RD) data on localization accuracy, the conventional ℓ2-norm based RD linear least squares (RD-LLS) location estimator is robustified here using the Geman–McClure (GM) and correntropy-induced loss functions, respectively. Two soft-redescending M-estimators are consequently developed, whereafter two iterative algorithms are presented to handle the corresponding optimization formulations. The first method applies the iteratively reweighted least squares to implement GM loss minimization, whereas the second employs the technique of half-quadratic optimization, through which the correntropy loss is maximized in an alternating manner. Either way the problem comes down to updating a series of weight-like variables and tackling a simplifed formulation that is closely related to RD-LLS at each iteration, thereby leading to fairly computationally efficient iterative solutions. Numerical results confirm the capabilities of the proposed schemes to outperform several existing hyperbolic positioning approaches.

Physical-Assisted Routing for Proactive Avoidance of Nomadic Obstacles in IoT

With the broadening of the radio spectrum to higher frequency bands, wireless links are more prone to blockages by nomadic obstacles. However, existing routing schemes mostly follow the network-oriented design principle, which makes it difficult to react quickly to sudden obstruction. This article proposes PAR, a novel physical-assisted routing scheme for the Internet of Things. PAR takes a physical-oriented viewpoint attempting to mitigate unexpected link blockages by leveraging sensor observations of obstacles at individual nodes. It analyzes the signatures of sensor measurements to directly estimate the positions and sizes of obstacles and to proactively infer remedial routing decisions even before the transmission failure occurs. During the network deployment time, PAR lets each node learn and calibrate the geographic distribution of neighbors with respect to its sensor measurements by exchanging the sensor observations of a mobile guidance obstacle. During the runtime, an obstacle bypassing algorithm is then developed to find the shortest detour routes by comparing the local sensor measurements with the calibrated positions of neighbors to immediately resume data forwarding. We evaluate the efficacy of PAR using simulation and a real-world testbed. It is observed that PAR significantly improves the success rate while reducing redundant hops and routing control overhead.

A Graph-Attention-Based Method for Single-Resident Daily Activity Recognition in Smart Homes.

In ambient-assisted living facilitated by smart home systems, the recognition of daily human activities is of great importance. It aims to infer the household's daily activities from the triggered sensor observation sequences with varying time intervals among successive readouts. This paper introduces a novel deep learning framework based on embedding technology and graph attention networks, namely the time-oriented and location-oriented graph attention (TLGAT) networks. The embedding technology converts sensor observations into corresponding feature vectors. Afterward, TLGAT provides a sensor observation sequence as a fully connected graph to the model's temporal correlation as well as the sensor's location correlation among sensor observations and facilitates the feature representation of each sensor observation through receiving other sensor observations and weighting operations. The experiments were conducted on two public datasets, based on the diverse setups of sensor event sequence length. The experimental results revealed that the proposed method achieved favorable performance under diverse setups.

Receding Horizon-Based Infotaxis With Random Sampling for Source Search and Estimation in Complex Environments

This article proposes a receding horizon-based information-theoretic source search and estimation strategy for a mobile sensor in an urban environment in which an invisible harmful substance is released into the atmosphere. The mobile sensor estimates the source term including its location and release rate by using sensor observations based on Bayesian inference. The sampling-based sequential Monte Carlo method, particle filter, is employed to estimate the source term state in a highly nonlinear and stochastic system. Infotaxis, the information-theoretic gradient-free search strategy is modified to find the optimal search path that maximizes the reduction of the entropy of the source term distribution. In particular, receding horizon Infotaxis (RHI) is introduced to avoid falling into the local optima and to find more successful information gathering paths in obstacle-rich urban environments. Besides, a random sampling method is introduced to reduce the computational load of the RHI for real-time computation. The random sampling method samples the predicted future measurements based on current estimation of the source term and computes the optimal search path using sampled measurements rather than considering all possible future measurements. To demonstrate the benefit of the proposed approach, comprehensive numerical simulations are performed for various conditions. The proposed algorithm increases the success rate by about 30% and reduces the mean search time by about 40% compared with the existing information-theoretic search strategy.

Comprehensive growth index monitoring of desert steppe grassland vegetation based on UAV hyperspectral.

The goal of this study was to establish a comprehensive growth index (CGI) of grassland vegetation for monitor the overall condition of the grassland. Taking the desert grassland in Otuoke Banner, Ordos City, Inner Mongolia as the research object, this study integrates five indicators. First, the optimal band of the unmanned aerial vehicle hyperspectral data is optimized using the correlation analysis, successive projection algorithm (SPA), optimum index factor method, and band combination index method. A dual-band spectral index in good correlation with the CGI is then constructed in the optimal band. Afterwards, a CGI characterization model is established in accordance with the partial least squares regression (PLSR) algorithm and its accuracy is analyzed. Finally, the CGI of the study area is estimated. The experimental results are as follows. 1) The R2 of models built using the training samples of the spectral indices corresponding to the optimal spectra screened by the SPA method was 0.7835, RMSE was 0.0712, and RE was 6.89%, less than 10%. The R2 of the Validation samples was 0.7698, RMSE was 0.0471, and RE was 6.36%, less than 10%, highest precision. 2) Models were built using the spectral indices corresponding to the optimal spectra screened by the SPA method, and the CGI mean values were inverted. A comparison of the mean measured CGI values of the sample quadrat of the test area showed that the mean relative error was 3.82%. The results show that the vegetation growth of desert-steppe grasslands can be adequately monitored, providing technical support for the rapid and accurate diagnosis of grassland conditions. However, there are still shortcomings in this study. 1) The research area for this study was mainly in the desert steppe in Otuoke Banner, Ordos, hence the relevance and universality of the findings need to be verified, and subsequent experiments need to be carried out on desert steppes in other regions or even other types of grasslands to test the universality of the model. 2) In this study, the influence of soil background and litter on the spectral reflectance is not considered in depth. In addition, the influence of sensor observation angle and solar elevation angle on the inversion model demands further investigation efforts.

Cooperative Navigation Algorithm of Extended Kalman Filter Based on Combined Observation for AUVs

The navigation and positioning of multi-autonomous underwater vehicles (AUVs) in the complex and variable marine environment is a significant and much-needed area of attention, especially considering the fact that cooperative navigation technology is the essential method for multiple AUVs to solve positioning problems. When the extended Kalman filter (EKF) is applied for underwater cooperative localization, the outliers in the sensor observations cause unknown errors in the measurement system due to deep-sea environmental factors, which are difficult to calibrate and cause a significant reduction in the co-location accuracy of AUVs, and can even cause problems with a divergence of estimation error. In this paper, we proposed a cooperative navigation method of the EKF algorithm based on the combined observation of multiple AUVs. Firstly, the corresponding cooperative navigation model is established, and the corresponding measurement model is designed. Then, the EKF model based on combined observation is designed and constructed, and the unknown error is eliminated by introducing a previously measured value. Finally, simulation tests and lake experiments are designed to verify the effectiveness of the algorithm. The results indicate that the EKF algorithm based on combined observation can approximately eliminate errors and improve the accuracy of cooperative localization when the unknown measurement error cannot be calibrated by common EKF methods. The effect of state estimation is improved, and the accuracy of co-location can be effectively improved to avoid serious declines in—and divergence of—estimation accuracy.

Data level and decision level fusion of satellite multi-sensor AOD retrievals for improving PM2.5 estimations, a study on Tehran

One of the techniques for estimating the surface particle concentration with a diameter of fewer than 2.5 micrometers (PM2.5) is using aerosol optical depth (AOD) products. Different AOD products are retrieved from various satellite sensors, like MODIS and VIIRS, by various algorithms, such as Deep Blue and Dark Target. Therefore, they do not have the same accuracy and spatial resolution. Additionally, the weakness of algorithms in AOD retrieval reduces the spatial coverage of products, particularly in cloudy or snowy areas. Consequently, for the first time, the present study investigated the possibility of fusing AOD products from observations of MODIS and VIIRS sensors retrieved by Deep Blue and Dark Target algorithms to estimate PM2.5 more accurately. For this purpose, AOD products were fused by machine learning algorithms using different fusion strategies at two levels: the data level and the decision level. First, the performance of various machine learning algorithms for estimating PM2.5 using AOD data was evaluated. After that, the XGBoost algorithm was selected as the base model for the proposed fusion strategies. Then, AOD products were fused. The fusion results showed that the estimated PM2.5 accuracy at the data level in all three metrics, RMSE, MAE, and R2, was improved (R2=0.64). Despite the simplicity and lower computational cost of the data level fusion method, the spatial coverage did not improve considerably due to eliminating poor quality data through the fusion process. Afterward, the fusion of products at the decision level was followed in eleven scenarios. In this way, the best result was obtained by fusing Deep Blue products of MODIS and VIIRS sensors (R2=0.81)