Search

Smart agricultural sensing has enabled great advantages in practical applications recently, making it one of the most important and valuable systems. For outdoor plantation farms, the prediction of climate data, such as temperature, wind speed, and humidity, enables the planning and control of agricultural production to improve the yield and quality of crops. However, it is not easy to accurately predict climate trends because the sensing data are complex, nonlinear, and contain multiple components. This study proposes a hybrid deep learning predictor, in which an empirical mode decomposition (EMD) method is used to decompose the climate data into fixed component groups with different frequency characteristics, then a gated recurrent unit (GRU) network is trained for each group as the sub-predictor, and finally the results from the GRU are added to obtain the prediction result. Experiments based on climate data from an agricultural Internet of Things (IoT) system verify the development of the proposed model. The prediction results show that the proposed predictor can obtain more accurate predictions of temperature, wind speed, and humidity data to meet the needs of precision agricultural production.

The power load prediction is significant in a sustainable power system, which is the key to the energy system’s economic operation. An accurate prediction of the power load can provide a reliable decision for power system planning. However, it is challenging to predict the power load with a single model, especially for multistep prediction, because the time series load data have multiple periods. This paper presents a deep hybrid model with a serial two‐level decomposition structure. First, the power load data are decomposed into components; then, the gated recurrent unit (GRU) network, with the Bayesian optimization parameters, is used as the subpredictor for each component. Last, the predictions of different components are fused to achieve the final predictions. The power load data of American Electric Power (AEP) were used to verify the proposed predictor. The results showed that the proposed prediction method could effectively improve the accuracy of power load prediction.

Time series data exist in various systems and affect the following management and control, in which real time series data sets are often composed of multiple variables. For predicting the future of data, not only the historical value of the variable but also other implicit influence factors should be considered. Therefore, we propose a prediction method based on the convolutional neural network (CNN) and Bi-directional long short term memory (Bi-LSTM) networks with the multidimensional variable. CNN is used to learn the horizontal relationship between variables of multivariate raw data, and Bi-LSTM is used to extract temporal relationships. Experiments are carried out with Beijing meteorological data, and the results show the high prediction accuracy of wind speed and temperature data. It is indicated that the proposed model can explore effectively the features of multivariable non-stationary time series data.

The state estimation method is one of the important techniques for target tracking and those techniques based on Kalman filtering have attracted the attention of many researchers. Among these existing methods, the very basic and necessary premise is that the model of the target is with high precision and good prior knowledge of process noise as well as measurement noise are also available. However, it is quite difficult to achieve in the practical applications. Therefore, in this paper, the Gated Recurrent Unit (GRU)-based estimation method is proposed to estimate the actual moving trajectory via the simulated GPS data with noise. GRU has a good advantage in the processing of time series data, and it has a memory function for data information. In the experiment part, by comparing with the traditional Kalman method, it is found that the GRU can obtain better estimation results.

An inertial measurement unit-based pedestrian navigation system that relies on the intelligent learning algorithm is useful for various applications, especially under some severe conditions, such as the tracking of firefighters and miners. Due to the complexity of the indoor environment, signal occlusion problems could lead to the failure of certain positioning methods. In complex environments, such as those involving fire rescue and emergency rescue, the barometric altimeter fails because of the influence of air pressure and temperature. This paper used an optimal gait recognition algorithm to improve the accuracy of gait detection. Then a learning-based moving direction determination method was proposed. With the Kalman filter and a zero-velocity update algorithm, different gaits could be accurately recognized, such as going upstairs, downstairs, and walking flat. According to the recognition results, the position change in the vertical direction could be reasonably corrected. The obtained 3D trajectory involving both horizontal and vertical movements has shown that the accuracy is significantly improved in practical complex environments.

The monitoring-blind area exists in the industrial park because of private interest and limited administrative power. As the atmospheric quality in the blind area impacts the environment management seriously, the prediction and inference of the blind area is explored in this paper. Firstly, the fusion network framework was designed for the solution of &ldquo, Circumjacent Monitoring-Blind Area Inference&rdquo, In the fusion network, the nonlinear autoregressive network was set up for the time series prediction of circumjacent points, and the full connection layer was built for the nonlinear relation fitting of multiple points. Secondly, the physical structure and learning method was studied for the sub-elements in the fusion network. Thirdly, the spatio-temporal prediction algorithm was proposed based on the network for the blind area monitoring problem. Finally, the experiment was conducted with the practical monitoring data in an industrial park in Hebei Province, China. The results show that the solution is feasible for the blind area analysis in the view of spatial and temporal dimensions.

The simulated experiment of A. parasiticus isolated from the paddy was carried out during the paddy storage for 20 days. The growth and mycotoxin data were collected for constructing kinetic and probability models of moulds. The Baranyi and Gompertz model was employed as the primary model and estimated the lag phase and maximum specific growth rate. Secondary models, such as polynomial, Davey and Gibson model were used and completely evaluated under different conditions. The polynomial equation was highly rated compared with Gibson and Davey model and gave realistic temperatures and aw for mould growth. Logistic model showed promising results on the prediction of growth boundary and AFB1 production. Employed models showed promising predicted results, indicating that it is an effective tool for describing and predicting the growth of moulds under different temperatures and aw. The results can be applied to develop the optimal strategy to prevent fungal spoilage and aflatoxin production during paddy storage.

Time-series data generally exists in many application fields, and the classification of time-series data is one of the important research directions in time-series data mining. In this paper, univariate time-series data are taken as the research object, deep learning and broad learning systems (BLSs) are the basic methods used to explore the classification of multi-modal time-series data features. Long short-term memory (LSTM), gated recurrent unit, and bidirectional LSTM networks are used to learn and test the original time-series data, and a Gramian angular field and recurrence plot are used to encode time-series data to images, and a BLS is employed for image learning and testing. Finally, to obtain the final classification results, Dempster–Shafer evidence theory (D–S evidence theory) is considered to fuse the probability outputs of the two categories. Through the testing of public datasets, the method proposed in this paper obtains competitive results, compensating for the deficiencies of using only time-series data or images for different types of datasets.

Precision farming aims to optimizing the crop production process and managing sustainable supply chain practices as more efficient and reasonable as possible. Recently, various advanced technologies, such as deep-learning and internet of things (IoT), have achieved remarkable intelligence progress in realistic agricultural conditions. However, crops species recognition can be considered as fine-grained visual classification (FGVC) problem, suffering the low inter-class discrepancy and high intra-class variances from the subordinate categories, which is more challenging than common basic-level category classification depended on traditional deep neural networks (DNNs). This paper presents a fine-grained visual recognition model named as MCF-Net to classifying different crop species in practical farmland scenes. Proposed MCF-Net is consisted of cross stage partial network (CSPNet) as backbone module, three parallel sub-networks, and cross-level fusion module. With multi-stream hybrid architecture utilizing massive fine-granulometric information, MCF-Net obtains preferable representation ability for distinguishing interclass discrepancy and tolerating intra-class variances. In addition, the end-to-end implementation of MCF-Net is optimized by cross-level fusion strategy to accurately identify different crop categories. Several experiments on in-field CropDeepv2 datasets demonstrate that our method favorably against the state-of-the-art methods. The recognition accuracy and F1-score of MCF-Net achieves very competitive results up to 90.6% and 0.962 separately, both of which outperforming contrasted models indicate better recognition accuracy and model stability of our method. Moreover, the overall parameters of MCF-Net are only 807 MByte with achieving a good balance between model’s performance and complexity. It is acceptable and suitable to the implementation of IoT platforms in precision agricultural practices.

MEMS (Micro-Electro-Mechanical Systems) gyroscope is the core component in the posture recognition and assistant positioning, of which the complex noise limits its performance. It is essential to filter the noise and obtain the true value of the measurements. Then an adaptive filtering method was proposed. Firstly, noises of MEMS gyroscope were analyzed to build the basic framework of the dynamic noise model. Secondly, the dynamic Allan variance was improved with a novel truncation window based on the entropy features, which referred to the parameters in the noise model. Thirdly, the adaptive Kalman filter was derived from the dynamic noise model. Finally, the simulation and experiment were carried out to verify the method. The results prove that the improved dynamic Allan variance can extract noise feature distinctly, and the filtering precision in the new method is relatively high.

Automation is a major challenge in the application of agricultural robots. Therefore, in order for agricultural robots to automatically classify and detect vegetables, an effective identification system should be established, which will increase production efficiency. In this paper, we propose a real-time vegetables recognition system based on deep learning network to detect and classify vegetables. On the basis of the large vegetable dataset obtained by our pinking robot, our goal is to find the appropriate framework for pinking mission with high accuracy and fast speed. Therefore, we select several one-stage and two-stage networks as alternative detectors including Faster R-CNN, SSD, RFB Net, YOLOv2, and YOLOv3. With the promotion of data augmentation, we demonstrate the recognition network based on YOLOv3 combing the depth meta-framework and feature extractor achieve the well performance in terms of both accuracy and speed by reducing the number of false positives in training phase. Comparative experiments indicate that our recognition system can effectively identify seven different kinds of vegetables (the mean AP can reach 87.89 %, and the detection speed is up to 38FPS), which would handle with real-time detection task and improve pinking capability of our agricultural robot.

In this paper, we propose an end-to-end deep learning framework that could extract the global spatial-temporal features from the videos based on the NetVladframework optimized with data augmentation. Given keyframes extracted from the original videos, we have presented a three-step action recognition framework: the first step of our framework is given by a data augmentation based on central crop, random crop and keyframe scaling. Then the second step is given by a local feature descriptor of each frame with Two-Stream Inflated 3D ConvNet (I3D) that is based on 2D ConvNet inflation, which enables the network to capture spatial and temporal features simultaneously. The third step is given by aggregating global features for action recognition through the new generalized “Vector of Locally Aggregated Descriptors” (NetVLAD) layer optimized with a novel pooling strategy, avoiding the misjudgment caused by local features. The whole framework is trained and fine-tuned via an end-to-end way. It's demonstrated that the performance of our framework outperforms the state-of-the-art algorithms on UCF10l dataset. The competitive results clearly reveal an efficient action recognition of high accuracy (up to 91.25%) in fast time (close to 1.8s), which will significantly improve performance and application of video action recognition.

State estimation is widely used in various automated systems, including IoT systems, unmanned systems, robots, etc. In traditional state estimation, measurement data are instantaneous and processed in real time. With modern systems’ development, sensors can obtain more and more signals and store them. Therefore, how to use these measurement big data to improve the performance of state estimation has become a hot research issue in this field. This paper reviews the development of state estimation and future development trends. First, we review the model-based state estimation methods, including the Kalman filter, such as the extended Kalman filter (EKF), unscented Kalman filter (UKF), cubature Kalman filter (CKF), etc. Particle filters and Gaussian mixture filters that can handle mixed Gaussian noise are discussed, too. These methods have high requirements for models, while it is not easy to obtain accurate system models in practice. The emergence of robust filters, the interacting multiple model (IMM), and adaptive filters are also mentioned here. Secondly, the current research status of data-driven state estimation methods is introduced based on network learning. Finally, the main research results for hybrid filters obtained in recent years are summarized and discussed, which combine model-based methods and data-driven methods. This paper is based on state estimation research results and provides a more detailed overview of model-driven, data-driven, and hybrid-driven approaches. The main algorithm of each method is provided so that beginners can have a clearer understanding. Additionally, it discusses the future development trends for researchers in state estimation.

Trend prediction based on sensor data in a multi-sensor system is an important topic. As the number of sensors increases, we can measure and store more and more data. However, the increase in data has not effectively improved prediction performance. This paper focuses on this problem and presents a distributed predictor that can overcome unrelated data and sensor noise: First, we define the causality entropy to calculate the measurement’s causality. Then, the series causality coefficient (SCC) is proposed to select the high causal measurement as the input data. To overcome the traditional deep learning network’s over-fitting to the sensor noise, the Bayesian method is used to obtain the weight distribution characteristics of the sub-predictor network. A multi-layer perceptron (MLP) is constructed as the fusion layer to fuse the results from different sub-predictors. The experiments were implemented to verify the effectiveness of the proposed method by meteorological data from Beijing. The results show that the proposed predictor can effectively model the multi-sensor system’s big measurement data to improve prediction performance.

Based on the collected weather data from the agricultural Internet of Things (IoT) system, changes in the weather can be obtained in advance, which is an effective way to plan and control sustainable agricultural production. However, it is not easy to accurately predict the future trend because the data always contain complex nonlinear relationship with multiple components. To increase the prediction performance of the weather data in the precision agriculture IoT system, this study used a deep learning predictor with sequential two-level decomposition structure, in which the weather data were decomposed into four components serially, then the gated recurrent unit (GRU) networks were trained as the sub-predictors for each component. Finally, the results from GRUs were combined to obtain the medium- and long-term prediction result. The experiments were verified for the proposed model based on weather data from the IoT system in Ningxia, China, for wolfberry planting, in which the prediction results showed that the proposed predictor can obtain the accurate prediction of temperature and humidity and meet the needs of precision agricultural production.

Air pollution (mainly PM2.5) is one of the main environmental problems about air quality. Air pollution prediction and early warning is a prerequisite for air pollution prevention and control. However, it is not easy to accurately predict the long-term trend because the collected PM2.5 data have complex nonlinearity with multiple components of different frequency characteristics. This study proposes a hybrid deep learning predictor, in which the PM2.5 data are decomposed into components by empirical mode decomposition (EMD) firstly, and a convolutional neural network (CNN) is built to classify all the components into a fixed number of groups based on the frequency characteristics. Then, a gated-recurrent-unit (GRU) network is trained for each group as the sub-predictor, and the results from the three GRUs are fused to obtain the prediction result. Experiments based on the PM2.5 data from Beijing verify the proposed model, and the prediction results show that the decomposition and classification can develop the accuracy of the proposed predictor for air pollution prediction greatly.

In this paper, a novel semi-supervised segmentation framework based on a spot-divergence supervoxelization of multi-sensor fusion data is proposed for autonomous forest machine (AFMs) applications in complex environments. Given the multi-sensor measuring system, our framework addresses three successive steps: firstly, the relationship of multi-sensor coordinates is jointly calibrated to form higher-dimensional fusion data. Then, spot-divergence supervoxels representing the size-change property are given to produce feature vectors covering comprehensive information of multi-sensors at a time. Finally, the Gaussian density peak clustering is proposed to segment supervoxels into sematic objects in the semi-supervised way, which non-requires parameters preset in manual. It is demonstrated that the proposed framework achieves a balancing act both for supervoxel generation and sematic segmentation. Comparative experiments show that the well performance of segmenting various objects in terms of segmentation accuracy (F-score up to 95.6%) and operation time, which would improve intelligent capability of AFMs.

Over the last decade, quantitative investment has been adopted by the international investment community as a new method that combines financial data with mathematics and computer technology. This paper proposes a quantitative investment prediction model that combines multifactor and candlestick models to predict stock prices. Firstly, relevant indicators were identified to construct a stock pool containing stocks with good future trends. Then, the candlestick model was used to decide when to buy or sell stocks in the pool. The method developed here can be used for quantitative investment strategies, and the experimental results show it can have good returns in the investment stock market based on the transaction data.

For high-frequency fluctuations of time-series data, it is necessary to extract and analyze the main trend for many applications. This paper focuses on the main trend analysis by Kalman filter, gives the extraction model and the transform model, and discusses the suitable value for the key parameters to guarantee the system convergence. The high order dimensional dynamics of the main trend are analyzed by the estimate results. The simulations show that the developed method is effective for extracting the main trend of the time-series data and able to explain accurately the characteristics of the main trend.

It is crucial to predict PM2.5 concentration for early warning regarding and the control of air pollution. However, accurate PM2.5 prediction has been challenging, especially in long-term prediction. PM2.5 monitoring data comprise a complex time series that contains multiple components with different characteristics, therefore, it is difficult to obtain an accurate prediction by a single model. In this study, an integrated predictor is proposed, in which the original data are decomposed into three components, that is, trend, period, and residual components, and then different sub-predictors including autoregressive integrated moving average (ARIMA) and two gated recurrent units are used to separately predict the different components. Finally, all the predictions from the sub-predictors are combined in fusion node to obtain the final prediction for the original data. The results of predicting the PM2.5 time series for Beijing, China showed that the proposed predictor can effectively improve prediction accuracy for long-term prediction.

Dealing with noisy time series is an important task in many data-driven real-time applications. In order to improve the veracity of the measured time series data, an effective denoising method is of great significance. For some applications with online requirement, the measurement would need to be processed to get rid of noise as soon as it is obtained. In this paper, a novel method was proposed to process relatively smooth time series data with annoying complex noise based on a second-order adaptive statistics model (SASM). However, in practical process, the nonzero mean measurement noise covariance “R” was unknown, and unfortunately it usually has a huge impact on the denoising effect. Therefore, this paper proposed a self-adjustment algorithm for measurement variance searching, by means of introducing a forgetting factor to estimate “R”. In this way, “R” would be convergent to the real value reasonably fast. The effectiveness of the method was verified by the simulation experiment. The results show that the proposed method can not only make “R” to be convergent to real value but also achieve the favorable denoising effect.

In a multi-sensor environment, it is efficient to record and reflect people's information of activities, using the large amount of data. However, the data cannot directly display the form of activity itself so that it is necessary to do the further job of exploration and processing. Deep Learning (DL) methods have attracted more attention and have shown some superior performance, while they have the problem of structural complexity. Therefore, this paper creatively uses Broad Learning System (BLS) method for human activity recognition. We use sliding window to get the data segmented. The weights involved are fine-tuned by pseudo-inverse and ridge regression algorithms, and we achieve an accurate classification of activities. The method is verified by using OPPORTUNITY data set. The results show that this method can greatly shorten the learning time and improve the accuracy, as well as the performance in comparison with traditional methods.