A novel method based on health index similarity in multiple time scales with autoencoder (AE MTS-HI) was proposed aiming at the shortage of the traditional similarity-based method in extracting health index and similarity matching. Autoencoder was applied to construct the health index based on monitoring data, which can minimize the loss of nonlinear information. The health index in multiple time scales was developed for similarity matching by considering the fluctuation of the length of test degradation trajectories. The method can remove the accuracy limitation caused by fixed time scales and enhance the prediction robustness. Performance of the proposed method was evaluated on public turbofan engines datasets. Results demonstrate that the method can improve the remaining useful life (RUL) prediction accuracy and provide stable support for predictive maintenance.

In order to promote the transformation of industrial cyber security defense mode from static passive defense to active defense, and alleviate the contradiction between the serious shortage of security experts and the sharp increase of cyber security demands, a cyber security active defense system framework of digital twin system was built from the perspective of bionics, and then five kinds of key technologies focusing on active defense were proposed based on the digital twin security brain (DTSB), including security data interaction and systems collaborative defense based on cloud-edge collaboration, cyber security active defense model of parallel digital twin system, situation awareness of parallel digital twin systems based on digital twin security brain, active defense and control technical framework for digital twin system based on immune system, and anti-attack intelligent recognition of digital twin system based on artificial intelligence. A case study of a digital twin workshop was given to demonstrate the successful application of digital twin cyber security in smart manufacturing.

A new deep neural network with attention convolutional gated recurrent unit-based autoencoder (CGRUA-AE) and a process fault detection method based on CGRUA-AE were proposed aiming at the problem that the existing fault detection algorithms were difficult to extract the internal information of data deeply and accurately. First, a convolutional gated recurrent unit (ConvGRU) was effectively extracted the spatial and temporal features of input data. Secondly, an auto-encoder based on ConvGRU was established, using unsupervised learning to extract features from time series data, introducing an attention mechanism to calculate the weight of corresponding features to realize the effective selection of key features. Finally, the process monitoring model based on $ {T}^{2} $ and SPE statistics were established in feature space and residual space respectively to realizes effective feature extraction and fault detection for multivariate data. Numerical case and Tennessee-Eastman process fault detection results show that CGRUA-AE has good feature extraction ability and fault detection ability, and its performance is superior to the common process fault detection methods.

Based on Navier-Stokes equations for incompressible fluids and magnetohydrodynamics (MHD) basic equations, a complete transfer function model for Hartmann flow of metallic fluid in a rectangular annular tube under current and voltage control mode were built up, and the effects of viscous force and boundary layer on the output performance of the momentum wheel were analyzed. By using the finite element simulation software COMSOL, the fluid motion characteristics and velocity distribution were simulated and verified. The influencing factors of output indexs, including current, magnetic field and characteristic parameters of the fluid, were totally analyzed. In current control mode, the angular momentum output scale factor of the momentum wheel is about 9.68×10^-5 N·m·s/A, which can provide the basis for the design and optimization of the momentum wheel.

An underwater image enhancement algorithm was proposed based on generative adversarial networks (GAN) and improved convolutional neural networks (CNN) in order to solve the problems of haze blurring and color distortion of underwater image. Generative adversarial network was used to synthesize underwater images to effectively expand the paired underwater data set. The underwater image was decomposed by multi-scale wavelet transform without losing the feature resolution. Then, combined with CNN, the compact learning method was used to extract features from multi-scale images, and skip connection was used to prevent gradient dispersion. Finally, the fog blur effect of the underwater image was resolved. In order to improve the color correction ability of the model and overcome the problem of color distortion of underwater images, the correlation between different channels of color images was learned by using the style cost function. Experimental results show that, in subjective visual and objective indicators, the proposed algorithm is superior to the contrast algorithm in comprehensive performance and robustness.

The researches on data security management and privacy protection technologies at home and abroad were analyzed and summarized aiming at current problems in blockchain security, such as unreasonable data management mode, unreliable data sharing scheme, smart contract vulnerabilities not easily fixed and incomplete privacy protection of multiple types of data. Various security problems and reasonable solutions in current blockchain systems were outlined from four aspects: data storage security, data privacy security, data access security and data sharing security. The challenges and future research directions of data security in blockchain were discussed. Some reference for the future work of researchers was provided in the field of blockchain security.

In order to achieve the parallel optimization of route planning, and solve the problem of high response time of location-based services (LBS) caused by extensive concurrent queries during peak hours, a dual-level grid index (DLG-index) was firstly introduced, and then, based on DLG-index, a parallel optimization algorithm of route planning (PORP) was introduced. The top layer of DLG-index is a skeleton graph consisting of border nodes of the entire graph, and the bottom layer is composed of all grids partitioned by the entire graph. For a given query, the first step is to compute a global path based on the skeleton graph. Then the route planning task is divided into multiple local optimizations in grids passed by the global path. At the same time, each local optimization is maintained independently by different processors. The algorithm can optimize the planned route in real time based on varying traffic conditions. The entire optimization is implemented in several segments, which can be handled by multi-processors and achieve rapid response to massive concurrent queries. Experiments results showed that compared with CANDS algorithm, the response time of PORP was reduced by an average of 49.6% and the processing time was saved by an average of 28.5%.

An efficient ShuffleNetv2 and YOLOv3 integrated network static gesture real-time recognition method was proposed to reduce the computing power requirements of the model on the hardware aiming at the characteristics of limited computing resources and small storage space under the mobile terminal platform. The computational complexity of the model was reduced by replacing Darknet-53 with the lightweight network ShuffleNetv2 as the backbone network. The CBAM attention mechanism module was introduced to strengthen the network’s attention to space and channels. The K-means clustering algorithm was used to regenerate the aspect ratio and number of Anchors, so that the regenerated Anchors size can accurately locate the target to improve the detection accuracy of the model. The experimental results showed that the average recognition accuracy of the proposed algorithm on gesture recognition was 99.2%, and the recognition speed was 44 frames/s. The inference time of a single 416×416 picture on the GPU was 15 ms, and the inference time on the CPU was 58 ms. The memory occupied by the model was 15.1 MB. The method has the advantages of high recognition accuracy, fast recognition speed, and low memory occupancy rate, which is conducive to the deployment of models on mobile terminals.

The research works of Chinese font style transfer were classified according to different stages of research development. The traditional methods were briefly reviewed and the deep learning-based methods were combed and analyzed. The commonly used open data sets and evaluation criteria were introduced. The future research trends were expected from four aspects, which were to improve the generation quality, enhance personalized differences, reduce the number of training samples, and learn calligraphy font style.

In order to solve the problems of difficult equipment deployment, velocity measurement and river monitoring in flooding environment, a series of image velocimetry techniques from particle image velocimetry (PIV) to deep learning methods were outlined based on non-invasive, low-cost and efficient measurement means in conjunction with nearly ten years of research in the field of river monitoring. The mechanism and issues of river surface velocimetry were discussed in the sections of image acquisition, image analysis, and image post-processing. By comparing and summarizing the differences of each method, the requirement of the existing methods were proposed, aiming to improve the river flow velocity measurement efficiency.

A mixed integer programming model was constructed for parallel disassembly line balancing problem aiming at the problem that the task definition of each disassembly line is unclear and the mathematical models are conceptual models in the existing parallel disassembly line. The difference of operators between workstations was considered. The number of workstations, the number of robots, disassembly cost and idle time balancing index were minimized. An improved brain storm optimization algorithm was proposed. A feasible disassembly sequence was constructed through double-layer coding, and the original operation was discretized. A mutation and crossover mode was designed corresponding to the generation mechanism of a single individual and two individuals. The operation strategy of four-point crossover was designed in order to increase the diversity of population individuals. Pareto solution set and crowding distance were introduced to screen non-inferior solutions of multi-objectives aiming at the multiplicity of optimization objectives. CPLEX and LINGO were used to solve the exact solution of small-scale examples. The correctness of the model and the effectiveness of the algorithm were verified compared with the results of the algorithm. The algorithm was applied to solve P25 classic examples and compared with the results of many existing literatures. The superiority of the algorithm was verified. The proposed model and algorithm were applied to the parallel disassembly line of TV and refrigerator, and the advantages of the proposed algorithm were verified by different comparative experiments.

A novel deep-learning-based model, adaptive graph generation jump network (AG-JNet), was proposed to solve the problem that traffic flow data has complex spatial-temporal correlations. The model consisted of two spatial-temporal modules, each of which was divided into two critical components, i.e., temporal correlation block and spatial correlation block. The temporal correlation block used multi-layer dilated convolution to increase the receptive field in temporal dimension while reducing computational cost. The spatial correlation block used adaptive graph generation convolution, which did not rely on the fixed graph structure to extract spatial correlation. Stacking multiple layers by jumping connections was used in both temporal and spatial modeling in order to improve the ability of extracting deep features of the model. The temporal feature and the spatial feature were fused by gated mechanism to obtain the spatial-temporal features for traffic flow prediction. Extensive experiments were conducted on two public datasets, i.e., PeMSD4 and PeMSD8. The experimental results showed that the AG-JNet achieved excellent performance under different traffic indicators.

The full-coverage film cooling was analyzed for the high temperature area on combustor in order to prolong the service life of the micro gas turbine combustor. The effects of arrangements and outer ring expansion film holes on the film cooling and the overall performance of combustor were compared under the actual conditions based on the test of KJ-66 micro gas turbine. Results showed that the average overall cooling efficiency of the order arrangement was lower than that of the cross arrangement in the actual micro gas turbine model, but the comprehensive cooling effect was higher. The film cooling effect was gradually improved as the outlet diameter of expansion holes increased, but the uniformity of temperature distribution at the combustor outlet was decreased. The secondary flow into the mainstream was deflected due to the influence of the cooling holes at the back of the combustor, which can improve the effects of film cooling. The full-coverage film cooling has a good cooling effect on the combustor wall under actual combustion conditions. Expansion film holes can effectively improve the film cooling effects of the combustor outer ring.

A publicly verifiable secret sharing technology was proposed based on the threshold secret sharing technology, in order to study the security of the user’s private secret in the blockchain. The secret fragments can be verified after the participating nodes receiving them, which can effectively preventing the master splitting node from doing evil when splitting the key. The secret fragments of the nodes participating in the secret splicing are publicly verified through the secret recovery phase, to prevent the nodes in the secret recovery phase from doing evil. Identity IDs are added to the participating nodes during the secret distribution phase, thus malicious nodes can be tracked and the node status can be updated in real time. The dynamic threshold mechanism was designed so that after the node holding the secret fragment is offline, the owner of the secret fragment and the master node can redistribute the secret fragment to the new participating nodes to ensure the integrity of the private secret fragment. Experimental results show that the private secret recovery rate of this scheme can reach 80%, and it has threshold characteristics, traceability, unforgeability and recoverability.

An algorithm of fractional frequency phase locked loop (PLL) output signal phase synchronization was proposed, in order to realize the phase synchronization of PLL on multiple transceiver chips or a single transceiver chip in a multi-channel radio frequency (RF) communication system. A selection algorithm of sampling points for phase accumulation was designed. The sampling points selected by the algorithm were used to accumulate the triangulation results of PLL’s output signal under sampled by reference clock and reference signal generated by NC oscillator (NCO), so as to eliminate the high-order harmonic component and reduce the error of the phase difference calculation result effectively. According to the size of the phase difference, the fractional frequency ratio of the input of delta-sigma modulator (DSM) in PLL was adjusted by feedback, so that the phase of PLL output signal was adjusted linearly, and the phase synchronization of multiple PLL output signals with reference signal was realized. The correctness of the algorithm is verified by simulation, and the phase error is 0.35 ° after the final phase synchronization, and the time required to complete the synchronization is 210 ms.

The definition of virtual obstacles for high temperature heat sources was proposed in order to solve the safety and efficiency problems faced by the global path planning of mobile robots in high temperature scenarios. A hybrid obstacle space model was established. The path planning problem in high temperature scene was transformed into a multi-objective optimization problem in high temperature mixed obstacle space, which considered the cost of path temperature and length. The NSGA-Ⅱ algorithm was improved to expand the population by selecting excellent non-feasible solutions, which improved the population diversity and population evolution efficiency. A new adaptive crossover and mutation probability calculation method was proposed. The process adjustment value realized the balance between the search ability in the early stage of the population and the convergence in the later stage according to the individual cost function value and the overall evolution process of the population. The simulation results of the optimal path show that although the path length cost of the proposed improved algorithm is slightly higher than that of the original algorithm and other improved algorithms, the temperature cost is greatly reduced. The proposed improved algorithm is more effective to avoid falling into the local optimal solution.

Difference between non-liveness attack types is neglected, and adverse impact of category imbalance between liveness and non-liveness samples on model training is not considered in existing studies of silent liveness detection. In this paper, non-liveness attacks were subdivided into two categories, print attack and display attack, which transformed silent liveness detection from traditional two-classification problem into multi-classification problem. And the cross-entropy was used as the loss function to train network model. Thus, the disadvantage of binary classification and category imbalance can be eliminated, common features of the non-liveness face samples were likely to be identified more accurately through model training, and the accuracy of the network model was improved for non-liveness recognition. Moreover, a two-stream feature fusion the network model was constructed to further improve the feature representation capacity of the network model, which adopted the attention mechanism to adaptively fuse the feature vectors extracted from RGB and YCrCb. Abundant comparative experiments were performed on four public datasets, CASIA-FASD, Replay-Attack, MSU-MFSD and OULU-NPU. Experimental results indicate that silent liveness detection model adopting multi-classification strategy and feature fusion can effectively reduce the classification error and improve over-generalization ability.

A new knowledge-enhanced graph convolutional neural network (KEGCN) classification model was proposed aiming at the problem of text classification. In the KEGCN model, firstly a text graph containing word nodes, document nodes, and external entity nodes was constructed on the entire text set. Different similarity calculation methods were used between different types of nodes. After the text graph was constructed, it was input into the two-layer graph convolutional network to learn the representation of the node and classified. The KEGCN model introduced external knowledge to compose the graph, and captured the long-distance discontinuous global semantic information, and was the first work to introduce knowledge information into the graph convolution network for classification tasks. Text classification experiments were conducted on four large-scale real data sets, 20NG, OHSUMED, R52 and R8, and results showed that the classification accuracy of the KEGCN network model was better than that of all baseline models. Results show that integrating knowledge information into the graph convolutional neural network is conducive to learning more accurate text representations and improving the accuracy of text classification.

A fault-tolerant strategy based on cache queue was proposed, aiming at the problem of low efficiency of stream processing job recovery due to global rollback after Flink task fails. In the job, the operator with the longest recovery time is taken as the key operator, the processed data are stored in the buffer queue, and active backup is performed for it. The backup operator will also accept the data from the upstream to reach the the effect that the job can be restored instantaneously after a failure. In order to solve the additional consumption caused by active backup, a data filtering algorithm was proposed. The backup operator will retrieve the current data from the cache component before processing the data each time to determine whether to continue processing. When the Flink operator itself fails, it will use the buffer queue in the strategy and Flink’s JobManager to send the data information at the time of the failure to the backup operator. When the backup operator receives the data, it will realize the effect of instant recovery. The strategy was evaluated on four evaluation indicators. Compared with the failure recovery mode of Flink1.8, the proposed strategy had a significant improvement in Flink task failure recovery. The recovery efficiency was increased by 56.3%, 51.3%, 46.2% and 45.8% under failure times of 1, 2, 3 and 4 separately. And the proposed strategy brings only a very small price in terms of processing delay, CPU utilization and memory usage.

A combined prediction model combining stacked bidirectional gated recurrent unit (SBiGRU), complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and error correction was proposed, aiming at the problem of error accumulation in iterative training process of load forecasting model. The SBiGRU model was established to learn the time series characteristics of load series under the influence of temperature and date type, and the error characteristics were reflected in the error series generated in the prediction process of SBiGRU model. Then CEEMDAN algorithm was used to decompose the error series into several intrinsic mode function (IMF) components and trend components. For each component, SBiGRU model was established again for learning and forecasting, and the predicted values of all components were reconstructed to obtain the error prediction results. Finally, the prediction results were summed to correct the error. Model evaluation results show that the prediction accuracy of the combined model is 98.86%. Compared with SBiGRU, BiRNN, support vector regression, ect., the combined model has better accuracy.