Aquaculture plays a crucial part in the coastal region, India. Aquaculture depends on the water quality categorized into physical, chemical, and biological parameters. Over a huge number of a farmer's life relies on aquaculture. This paper shows the monitoring of water quality based on Machine Learning and IoT. Based on the change in parameters like pH, water level, temperature, foul smell, and turbidity, the water quality is monitored and the growth of fisheries in water is determined. This paper presents the remote monitoring of parameters on screen as well as on an Android App after storing them into the cloud and also helps to find the accuracy of the water quality based on the machine learning algorithm.

This paper presents investigations of physical properties of hydrogel (HG) based on the lightly crosslinked polymer of acrylic acid "Carbopol ETD 2020" brand under conditions of electrical and thermal modifications when exposed to alternating frequency-modulated signal (AFMS). Data on change in the hydrogel viscosity depending on the gel-forming component concentration, AFMS treatment, and temperature were obtained. It is established that the comparative heating time of modified hydrogels increases with increased concentration of the gel-forming component and under AFMS exposure for thermally modified hydrogels near the water critical point of 4 °C. The advantages of using modernized HG with respect to the traditional fire-extinguishing agent are shown. Thermophysical and rheological properties of HG obtained under electrical modification and hydraulic calculations of watering systems for above-ground vertical stock tanks (VST) of petroleum product proved the justification for their use and showed improving efficiency of installations for thermal protection of tanks storing petroleum products. The compositions of modified HG and their field of application in the nomenclature of tanks for petroleum products were determined. Practical recommendations with calculation results for the use of modified HG on existing fire extinguishing installations in tank batteries are proposed.

Traction substation is the only power source for subway trains, and its reliable operation is essential to guarantee the safety of the AC metro systems. Therefore, it is significant to accurately assess the reliability and provide guidance on reliability enhancement approaches. In this paper, firstly, by dividing the AC metro traction substation into two series blocks according to the system structure and operation mode, the fault tree analysis (FTA) model is established; then, using sequential Monte Carlo simulation, the reliability indices are obtained and the critical components of the system are identified; finally, the reliability level of the AC metro traction substation is considerably improved by enhancing the robustness of critical components. The proposed method is verified through case study, and it shows that: (1) with consideration of time-varying characteristics, the sequential Monte Carlo method can effectively capture the dynamic variation of system reliability with the accumulation of operation time; (2) the co-phase power supply device (TX1), circuit breaker (L315), 25 kV bus, and cable (L5 and L6) are identified as the critical components of the system; by enhancing the reliability of these pieces of equipment, the system reliability can be effectively improved. The research in this paper can help designers to identify and enhance the critical components to improve system reliability in the designing stage, and they can also enable maintenance personnel to comprehend the system operation condition and pay more attention to critical equipment during the inspection and maintenance period.

Conventional neural networks-based network security situation (NSS) prediction methods have low prediction accuracy and low efficiency. To solve such shortages of NSS prediction, this paper proposes a novel method based on combining 3D convolutional neural networks (3DCNNs) and bidirectional recurrent neural networks (Bi-RNNs). Because the prediction data of NSS includes multi-dimensional time series, the NSS can be better predicted by combining the spatial features and sequential features, and the prediction accuracy and efficiency will be improved by using the combined features. Therefore, the 3DCNNs model is adopted to extract spatial features from different network nodes, and the Bi-RNNs model with gated recurrent units (Bi-GRUs) is adopted to extract the sequential features based on the extracted spatial features. Finally, the NSS prediction results are obtained by using the fused spatial-sequential features. In order to validate the feasibility and effectiveness of the proposed method, comparable experiments are performed on three different datasets. Experimental results have shown that the proposed 3DCNNs-Bi-GRUs model achieves the optimal NSS prediction results among all datasets under different situations. The efficiency of the proposed model meets the requirements for real-world application scenarios. By combining the spatial features and sequential features, the proposed model confirms higher prediction accuracies of NSS, and such a model has good application value for the rapid development of computer networks and intelligent technologies.

It is difficult to detect the faulty feeder when a single-phase-to-ground (SPG) fault occurs in the resonant earthed system due to the low amplitude of current, and different faulty conditions leading to different components of transient zero-sequence current (TZSC) make the detection much harder. A novel faulty feeder detection method based on the parameter-optimization variational mode decomposition (VMD) is proposed in this paper. Firstly, utilize the fruit fly optimization algorithm (FOA) to obtain the optimal parameter combination [K, α] of VMD for the fault components extracted accurately. Second, considering that low-frequency decaying DC components only exist in the faulty feeder, the modal energy of decaying DC components of each feeder are calculated for a preliminary judgment. Then, due to the amplitude and polarity of the high-frequency transient capacitive current are both different between the faulty and sound feeder, the correlation analysis of high-frequency components is utilized to reflect the mismatch of zero-sequence current waveform in high frequency band between faulty and sound feeder. Finally, the dual criterion combining modal energy of decaying DC components in low-frequency band and waveform similarity in high-frequency band is constituted. A large number of MATLAB/Simulink simulation results show the great effectiveness and reliability of this method at different fault conditions.

Accurate medical image segmentation can assist doctors in disease diagnosis. It is very important to segment the liver accurately from medical images in the field of the liver. However, the low contrast of tissues and organs and uneven distribution of CT values in abdominal CT images makes liver segmentation difficult. In this paper, we propose a method of combining the improved U-Net network model and the region growing algorithm. The feature information of the pooling layer is directly extracted after two convolution and ReLU functions. The up-sample layer copies the feature information of the corresponding down-sampling layer. Softmax Layer calculates the amount of information loss to reduce the loss of feature information. Finally, the region growing algorithm is used to optimize the initial results. Five parameters of medical image segmentation are used for evaluation. DICE can reach more than 95.0% accuracy and other parameters have been increased accordingly. Experimental results show this method can accurately segment the liver area, solve the problems of blurred edges and unclear areas, and provide an effective basis for the diagnosis of liver disease.

Defects such as blowholes and cracks are inevitable in the manufacturing of the magnet tile, and online full detection is a necessary process. Image-based surface defect detection is of great significance for improving product quality and production efficiency. This paper introduces a pixel-level surface defect detection method based on a deep full convolutional network, which realizes the detection and classification of defects at the same time. Combining the U-Net architecture and the squeeze-excitation module, an SE-U-Net that adaptively fuses shallow local information and deep semantic information is constructed. With a small amount of additional computation, U-Net's accuracy in detecting small defects from the large background is improved. Data augmentation via data transfer reduces the imbalance between image background and defects and improves the learning speed of the model. The proposed method was compared with SegNet and U-Net and achieved more accurate defect detection, and the average pixel accuracy reached 0.97, which demonstrates the superiority of the improved SE-U-Net for magnetic tile surface defects.

Test case prioritization is an effective way to improve test efficiency and to reduce test costs. To deal with the problem that the application of group intelligent searching algorithm in large-scale test case prioritization causes slow convergence rate and premature convergence, leading to unstable error detection ability, in this paper, we propose a novel approach for large-scale test case prioritization based on the Viterbi algorithm. Specifically, we first use the sequential grid search idea of the Viterbi algorithm to encode the test case set in the priority ranking model of multi-objective test cases to form the state space of sequential grid search. Then, the code coverage and the effective execution time of test cases are taken as the path measurement of state transition. Finally, the path with the largest path metric is selected as the survival path to complete the priority ranking of test cases. Experimental results show that the algorithm has a higher convergence speed and higher defect detection rate.

Traditional collaborative filtering algorithms make recommendations based on user behavior without considering semantic relationships between objects. Based on the normalization of the item similarity matrix according to the maximum value, this paper normalizes the user similarity matrix. That is, while reducing the influence of popular items on the recommended results, it also reduces the impact of active users. This paper uses the knowledge graph between items as auxiliary information, makes recommendations based on the confidence of the multi-path relationship, and merges the recommendation results with the user- and item-based collaborative filtering recommendation results. The processing method in this paper makes up for the defects of traditional collaborative filtering algorithms, which do not sufficiently consider hidden information, and it has higher accuracy, coverage, and recall.

For inaccurate internal resistance measurements of battery packs of under-well water distributor power supplies, a new compensation method for internal resistance measurement is proposed in this paper. On the basis of analyzing numerous factors influencing the detection disorder of internal resistance of batteries, multiple linear regression based on statistical prediction is applied to compensate for the resistance measurement error caused by equipment disturbance and aging, and SPSS is employed for prediction analysis and regression model testing. The test results show that the minimum resolution of resistance is 1mΩ, and the measurement precision can reach 1%, which fully meets the performance requirements of down-well intelligent water distributors for battery packs.

The power grid is located at the end of the power transmission chain. The further away from the power supply center, the worse the power quality. These areas often contain abundant natural energy such as wind, solar and tidal energy. Here, we investigate the influence of distributed power supply on distribution networks. A two-layer optimal configuration model including operational level and planning level is established. Then, the energy storage target model is established for two kinds of power with different power sources in order to adapt to different planning purposes. Furthermore, the two-layer model is solved and analyzed. The research proves that the energy storage system using the double-layer optimization model can optimize the distribution network with a distributed power supply.

In order to achieve better prediction results on emerging trends of keywords, a graph neural network is used to mine the relationship between keywords. This paper first constructs a keyword network based on co-occurrence relationship. Then, a feature template is carefully designed. Finally, the relationships among keywords are captured by a graph neural network, and the method for predicting emerging trends is proposed. The results on the test set show that the average Euclidean distance of the prediction results of this method is 12.33, and the rank correlation is 0.85, which significantly outperforms the baseline methods.

Smart school is the product of a new round of information technology reform and the further development of knowledge economy. It is a manifestation of the deep integration of industrialization, urbanization, and informatization, as well as the progress towards a higher stage. The standardization of smart school construction must be combined with the reality of smart school construction to carry out the standardization work pertinently. Intelligent school construction is accelerating step by step. Internet technology, mobile communication technology, and intelligent terminals are combined to be applied to all areas of social life. This paper makes full use of the concepts and technological achievements of smart school construction, from the aspects of data regulation and results building, conflict detection, and contradiction coordination, results management, and resource sharing. The construction of multi-information information fusion visualization system for sharing, unified, efficient, and dynamic updating is studied. The open big data teaching mode is used to conduct theoretical study and practical exercises online. This approach allows learners to master the necessary knowledge and skills as soon as possible and provides sufficient and excellent big data talents for enterprises in a timely manner.

In recent years, the frequent occurrence of safety accidents in construction projects has brought a lot of adverse effects to the enterprise and the community. In the event of unexpected incidents, how to choose the best contingency plan and how to carry on the scientific appraisal to the project emergency management have been a key research focus. Based on this, the characteristics and emergency management of large-scale construction projects were studied, and the theory of interval valued intuitionistic fuzzy sets and interval numbers were studied in detail. According to the project criteria with incomplete uncertain information on weights of emergency decision problems, the emergency management model of construction project based on interval valued intuitionistic fuzzy information was put forward, which can effectively solve the fuzzy project and uncertainty in emergency management. Taking the collapse of a certain project as an example, the concrete steps of emergency management decision-making evaluation were studied. It is proven that the research on emergency management evaluation of large scale construction projects based on interval intuitionistic fuzzy information is beneficial to ensure the scientific and rationality of emergency management, and it can effectively reduce the damage caused by unexpected events.

Database-assisted dynamic spectrum allocation, a dynamic spectrum allocation that is closest to practical application, is regarded as the core key technology of next-generation radio networks. However, it is easy for malicious users to obtain spectrum information and spectrum access opportunities from the spectrum database due to its openness and shareability. The database-assisted dynamic spectrum allocation will deal with more serious security problems. We discuss a practical and promising dynamic spectrum allocation scenario where some malicious users eavesdrop on legitimate users' transmission to access and imitate them. We propose a database-assisted dynamic spectrum allocation method with spectrum security. In the proposed method, legitimate users are required to transmit their spectrum identity information when transmitting communication signals. The spectrum database must consider the transmission of spectrum identity information when it dynamically allocates spectrum to these legitimate users. We establish the optimization model and its solution of the database-assisted dynamic spectrum allocation with spectrum security when the spectrum watermark of these legitimate users is embedded and extracted by the spread spectrum technology. The alternate optimization and differential convex (D.C.) programming optimization method is used to solve the proposed optimization problem based on the suboptimal equivalent that the channel allocation coefficients with values of is replaced by these continuous values of . The simulation results show that the proposed method can converge quickly and obtain approximate optimal performance with spectrum security.