Explore the vast range of titles available.

Understanding the pattern of financial activities and predicting their development and changes are research hotspots in academic and financial circles. Because financial data contain complex, incomplete and fuzzy information, predicting their development trends is an extremely difficult challenge. Fluctuations in financial data depend on a myriad of correlated constantly changing factors. Therefore, predicting and analysing financial data are a nonlinear, time-dependent problem. Deep neural networks (DNNs) combine the advantages of deep learning (DL) and neural networks and can be used to solve nonlinear problems more satisfactorily compared to conventional machine learning algorithms. In this paper, financial product price data are treated as a one-dimensional series generated by the projection of a chaotic system composed of multiple factors into the time dimension, and the price series is reconstructed using the time series phase-space reconstruction (PSR) method. A DNN-based prediction model is designed based on the PSR method and a long- and short-term memory networks (LSTMs) for DL and used to predict stock prices. The proposed and some other prediction models are used to predict multiple stock indices for different periods. A comparison of the results shows that the proposed prediction model has higher prediction accuracy.

Building extraction from remote sensing (RS) images is a fundamental task for geospatial applications, aiming to obtain morphology, location, and other information about buildings from RS images, which is significant for geographic monitoring and construction of human activity areas. In recent years, deep learning (DL) technology has made remarkable progress and breakthroughs in the field of RS and also become a central and state-of-the-art method for building extraction. This paper provides an overview over the developed DL-based building extraction methods from RS images. Firstly, we describe the DL technologies of this field as well as the loss function over semantic segmentation. Next, a description of important publicly available datasets and evaluation metrics directly related to the problem follows. Then, the main DL methods are reviewed, highlighting contributions and significance in the field. After that, comparative results on several publicly available datasets are given for the described methods, following up with a discussion. Finally, we point out a set of promising future works and draw our conclusions about building extraction based on DL techniques.

Water is the source of human life and water pollution is becoming more and more serious with the development of cities. The supervision and treatment of water resources have become a big problem of urban development. Water quality monitoring is not timely, flood warning is not timely is directly related to the livelihood of the people. And the development of smart water utilities can solve problems timely and accurately. By placing water quality sensors in the urban water supply network, real-time monitoring of water quality can be performed to prevent incidents of drinking water pollution. After an incident of drinking water pollution occurs, reverse locating the pollution source through the information detected by the water quality sensors represents a challenging problem because in the actual water supply network, the direction and speed of the water flow will change with the water demand of the residents, thus leading to uncertainty in this problem. In conventional studies of pollution source location problems, it is often assumed that the water demand is fixed. However, due to the variability of the water demand of residents, this problem is actually a dynamic change problem and thus can be considered as a dynamic optimization problem. In this study, a Poisson distribution model was used to simulate the change of water demand among urban residents. On this basis, we proposed an improved genetic algorithm to solve the pollution source location problem and implemented two different water supply networks to perform the simulation experiments, which could accurately locate the pollution sources. The simulation results were compared with the standard genetic algorithm to verify the accuracy and robustness of the proposed algorithm.

When drinking water flows into the water distribution network from a reservoir, it is exposed to the risk of accidental or deliberate contamination. Serious drinking water pollution events can endanger public health, bring about economic losses, and be detrimental to social stability. Therefore, it is obviously crucial to research the water contamination source identification problem, for which scholars have made considerable efforts and achieved many advances. This paper provides a comprehensive review of this problem. Firstly, some basic theoretical knowledge of the problem is introduced, including the water distribution network, sensor system, and simulation model. Then, this paper puts forward a new classification method to classify water contamination source identification methods into three categories according to the algorithms or methods used: solutions with traditional methods, heuristic methods, and machine learning methods. This paper focuses on the new approaches proposed in the past 5 years and summarizes their main work and technical challenges. Lastly, this paper suggests the future development directions of this problem.

Reducing pollutant detection time based on a reasonable sensor combination is desirable. Clean drinking water is essential to life. However, the water supply network (WSN) is a vulnerable target for accidental or intentional contamination due to its extensive geographic coverage, multiple points of access, backflow, infrastructure aging, and designed sabotage. Contaminants entering WSN are one of the most dangerous events that may cause sickness or even death among people. Using sensors to monitor the water quality in real time is one of the most effective ways to minimize negative consequences on public health. However, it is a challenge to deploy a limited number of sensors in a large-scale WSN. In this study, the sensor placement problem (SPP) is modeled as a sequential decision optimization problem, then an evolutionary reinforcement learning (ERL) algorithm based on domain knowledge is proposed to solve SPP. Extensive experiments have been conducted and the results show that our proposed algorithm outperforms meta-heuristic algorithms and deep reinforcement learning (DRL).

To improve the filtration efficiency of submicron dust by dual-layer granular bed filters, filtration experiments for micro-silica powder were conducted for removing particles smaller than 1 μm that account for more than 96% (by volume) using a dual-layer granular bed filter with an external electric field. Electrostatic enhancement methods, including dust pre-charging, application of an electric field to the lower filter layer, and that to both the upper and lower filter layers, were examined. Results showed that the average filtration efficiency of a dual-layer granular bed filter for micro-silica powder without electric field was 76.52%, the average outlet dust concentration was 263.53 mg/m 3 , and the filtration cycle time was 73 min. With pre-charged dust, the average outlet dust concentration dropped to 82.51 mg/m 3 . A decrease in the thickness of the lower filter layer from 45 to 25 mm with electric field reduced the pressure drop from 2570 to 1770 Pa. Meanwhile, the application of an electric field to the lower/upper filter layer reduced the average outlet dust concentration to 25.98 mg/m 3 . Increasing the initial face velocity from 0.25 to 0.45 m/s increased the average outlet dust concentration from 25.98 to 30.27 mg/m 3 and increased the pressure drop from 2570 to 3500 Pa.

The real-time location of pollution sources is the process of inverting pollution sources based on the dynamic optimization model constructed by the time-varying pollution concentration detected by the water quality sensor. Due to the vast quantities of the water supply networks, the water quality sensors will only be placed on critical nodes, resulting in multiple solutions. However, the increased monitoring data enhances the uniqueness of the solution. Combined with the real-time location of pollution sources, this work proposed a multi-strategy dynamic multi-mode optimization algorithm based on domain knowledge, which could guide the population search and avoid trapped into local optimal. The merging mechanism was used to keep the diversity of the population and prevent sub-population clustering on the same optimal solution. The simulation results showed that the algorithm could effectively solve the real-time location problem of pollution sources in different pipe networks and pollution scenarios.

Urban water supply network is easily affected by intentional or occasional chemical and biological pollution, which threatens the health of consumers. In recent years, drinking water contamination happens occasionally, which seriously harms social stabilization and safety. Placing sensors in water supply pipes can monitor water quality in real time, which may prevent contamination accidents. However, how to reversely locate pollution sources through the detecting information from water quality sensors is a challengeable issue. Its difficulties lie in that limited sensors, massive pipe network nodes and dynamic water demand of users lead to the uncertainty, large-scale and dynamism of this optimization problem. This paper mainly studies the uncertainty problem in contaminant sources identification (CSI). The previous study of CSI supposes that hydraulic output (e.g., water demand) is known. Whereas, the inherent variability of urban water consumption brings an uncertain problem that water demand presents volatility. In this paper, the water demand of water supply network nodes simulated by Gaussian model is stochastic, and then being used to solve the problem of CSI, simulation–optimization method finds the minimum target of CSI and concentration which meet the simulation value and detected value of sensors. This paper proposes an improved genetic algorithm to solve the CSI problem under uncertainty water demand and comparative experiments are placed on two water distribution networks of different sizes.

The utilization of clean energy is a matter of primary importance for sustainable development as well as a vital approach for solving worldwide energy-related issues. If the low utilization rate of nuclear fuel, nuclear proliferation, and insufficient nuclear safety can be solved, nuclear fission energy could be used as a sustainable and low-carbon clean energy form for thousands of years, providing steady and base-load electrical resources. To address these challenges, we propose an accelerator-driven advanced nuclear energy system (ADANES), consisting of a burner system and a fuel recycle system. In ADANES, the ideal utilization rate of nuclear fuel will be >95%, and the final disposal of nuclear waste will be minimized. The design of a high-temperature ceramic reactor makes the burner system safer. Part of fission products (FPs) are removed during the simple reprocessing in the fuel recycle system, significantly reducing the risks of nuclear proliferation of nuclear technology and materials. The ADANES concept integrates nuclear waste transmutation, nuclear fuel breeding, and safety power production, with an ideal closed loop operation of nuclear fission energy, constituting a major innovation of great potential interest for future energy applications.

During the molybdenite mining process, conveyor belts stretching for miles are used to transport ore between the blasting sites, crushing stations, and the concentrator plant. In order to ensure the safety and stability of the industrial production process, this paper introduces a foreign matter detection method based on deep learning for the belt conveyor. Aiming at the problems of insufficient feature extraction capabilities in existing machine vision-based foreign body detection methods and poor detection accuracy due to imbalanced positive and negative samples, an improved foreign body detection method for anchorless frame-type metal mine belt conveyors is proposed. This method introduces atrous convolution in the pooling layer to increase the receptive field of feature extraction and improve the ability of extracting feature details of foreign objects. By optimizing the ratio of positive and negative samples in the training process, the overall loss function value of the algorithm is reduced to ensure the accuracy of foreign body recognition. Finally, the improved model is trained after enhancing and labeling the sample dataset. The experimental results show that the average mean accuracy of foreign body detection (MAP) is 90.9%, better than existing methods. It can be used as an effective new method for detecting foreign objects on molybdenum mine belt conveyors.