Search

With the construction of the modern power system, power load forecasting is significant to keep the electric Internet of Things in operation. However, it usually needs to collect massive power load data on the server and may face the problem of privacy leakage of raw data. Federated learning can enhance the privacy of the raw power load data of clients by frequently transmitting model updates. Concerning the increasing communication burden of resource-heterogeneous clients resulting from frequent communication with the server, a communication-efficient federated learning algorithm based on Compressed Model Updates and Lazy uploAd (CMULA-FL) was proposed to reduce the communication cost. CMULA-FL also integrates the error compensation strategy to improve the model utility. First, the compression operator is used to compress the transmitted model updates, of which large norms are uploaded to reduce the communication cost of each epoch and transmission frequency. Second, by measuring the error of compression and lazy upload, the error is accumulated to the next epoch to improve the model utility. Finally, based on simulation experiments on the benchmark power load data, the results show that the communication cost decreases at least 60% with controlled loss of model prediction compared with baseline.

Most cluster identification studies regarding consumer electricity load is faced with problems of erroneous clustering method similarity, low clustering quality and poor identification accuracy. To solve these problems, this paper utilizes the elbow method, k- Means ++, entropy weight method and a graph convolutional neural network to provide a means for cluster identification based on electrical appliance power data collected via smart sockets. In this article, elbow and entropy weight methods were used to achieve the adaptive clustering algorithm. To obtain the electrical appliance load curves, Euclidean and dynamic time warping (DTW) distances were integrated and the similarity measurement method was used to improve the k-means ++ algorithm, which was then applied to data collected via smart socket clustering. Next, clustering results were input into the graph convolutional neural network (GCN) for identification purposes and appliance type information was obtained. Finally, experiments were conducted using electrical load data from 20 commercial users. The method used was a combination of the k-means algorithm and long short-term memory network (LSTM). The results show that under optimal K value conditions (as determined by the elbow score), the methods used in this paper have improved clustering quality and recognition accuracy, when compared to LSTM.

Many-objective optimization problems are widely applied and complex to solve. For most many-objective evolutionary algorithms, maintaining the balance of solution convergence and diversity is a challenging problem. Considering the convergence and diversity at the same time, a many-objective artificial bee colony algorithm based on adaptive grid(MOABCAG) is proposed. By searching for the ideal target point in the grid, the solution selection pressure is enhanced, the solution is positioned under the designed grid, and the grid is adaptively divided to improve the convergence and diversity of the solution. The algorithm improves the position sharing mechanism of the leader and follower bees in the artificial bee colony algorithm, and it sets the variable neighborhood search method of the follower bee to improve the precision of the solution vectors. MOABCAG is compared with five well-known evolutionary algorithms on thirteen benchmark test functions. The results show that the proposed MOABCAG algorithm obtains better performance than other related state-of-the-art algorithms in solving such many-objective optimization problems.

Differentially private stochastic gradient descent (DP-SGD) that perturbs the clipped gradients is a popular approach for private machine learning. Gaussian mechanism GM, combined with the moments accountant (MA), has demonstrated a much better privacy-utility tradeoff than using the advanced composition theorem. However, it is unclear whether the tradeoff can be further improved by other mechanisms with different noise distributions. To this end, we extend GM ( $p=2$ ) to the generalized $p$ -power exponential mechanism ( $p$ EM with $p>0$ ) family and show its privacy guarantee. Straightforwardly, we can enhance the privacy-utility tradeoff of GM by searching noise distribution in the wider mechanism space. To implement $p$ EM in practice, we design an effective sampling method and extend MA to $p$ EM for tightly estimating privacy loss. Besides, we formally prove the non-optimality of GM based on the variation method. Numerical experiments validate the properties of $p$ EM and illustrate a comprehensive comparison between $p$ EM and the other two state-of-the-art methods. Experimental results show that $p$ EM is preferred when the noise variance is relatively small to the signal and the dimension is not too high.

This paper presents a thorough evaluation of twenty-one state-of-the-art widely-used crop segmentation algorithms, motived by their significance in vision tasks for further analysis. An ideal crop segmentation algorithm can effectively extract crop information, thus providing an important precondition for the application of intelligent agriculture analytics. In order to enable researchers in this field to fully understand various crop segmentation methods, this paper proposes a new classification strategy of object segmentation by dividing the algorithms into pixel-based and region-based approaches at first, and then systematically evaluating various crop segmentation methods with a unified data benchmark and four common metrics. A new dataset which incorporates crop variety, environment condition and observation distance into consideration is constructed for demonstrating the experiments and comparisons. The effectiveness and robustness of these algorithms were evaluated by three sets of comparative experiments. Based on the quantitative results, we summarize the advantages and disadvantages of the evaluated algorithms from the segmentation performances with four metric indicators. Furthermore, the discussion and evaluation results will provide great support for precision agriculture analysis.

In this work, a novel process monitoring method in a block-wised partitioning manor is proposed for plant-wide processes which can be partitioned into several sub-blocks and monitored parallelly. The focus of this method is to reduce the high complexity of global plant-wide process, while to improve the efficiency of local feature extraction. In this method, considering that not all process knowledge is available in the block division process strategy, a novel community discovery (CD) algorithms, based on the similarity of neighbor node weighted Louvain, is introduced into the framework of the multi-block Bayesian inference and principal component analysis (PCA) based plant-wide process monitoring scheme. Firstly, the complex network (CN) theory is used to establish the network topology structure for the global variables of the plant-wide process. Secondly, by analyzing the graph characteristic structure, considering the connection strength between nodes, a more reasonable sub-block division is conducted according to the improved Louvain algorithm. Then, PCA method is used to establish process monitoring model for each sub-block to obtain sub-block monitoring statistics. Finally, the total joint statistics is obtained through Bayesian inference for fault detection. The feasibility and effectiveness, in terms of the detection performance, of this method are demonstrated in a simulated plant-wide process by compared with other state-of-the-art PCA based monitoring methods.

In order to machine complex box and shell structures, the cutter with large ratio of length to diameter is usually used, whereas the low rigidity of the cutter hinders the increase of the stable machining parameters. This article develops one kind of damping cutter to improve the stable processing condition in this milling operation. The influences of modal characteristic of the cutter on chatter stability are investigated, where it is proven that increasing the dynamic stiffness can efficiently extend the stable zone. On this basis, two damping cutters are designed by inlaying the strips and damping core into the toolholder body. The influences of geometrical parameters and material property of the damping cutters on modal characteristics are theoretically analyzed, and the optimal geometrical dimensions are determined using the finite element simulation. Then, the two damping cutters are fabricated. After modal tests and milling machining experiments, the measured results validate the effectiveness of the developed damping cutters in terms of improvement of the stability frontier.

As more complex curves are used in current productions, curve speed planning has become a key technique to overcome the bottleneck of high-speed and high-precision computerized numerical control (CNC) systems. This paper first establishes the model of particle velocity, acceleration, and jerks in the Cartesian coordinate system, and then improves the uniaxial performance limit algorithm. We set up a real-time speed planning process of spline curves, design an S-model look-ahead algorithm, simplify the S-model speed planning algorithm, and achieve real-time non-uniform rational B-splines (NUBRS) curve S-model speed planning based on uniaxial performance limitation. The simulation results show that all the actual interpolation velocity, acceleration, and jerk under the proposed method meet the preset single-axis limit. The experimental results show that the tracking performance under the proposed method has been significantly improved compared with that under the small line segments method. Compared with the NUBRS curve trapezoidal model speed planning, the vibration spikes during machining can be eliminated.

Effective cryptocurrency key management has become an urgent requirement for modern cryptocurrency. Although a large body of cryptocurrency wallet-management schemes has been proposed, they are mostly constructed for specific application scenarios and often suffer from weak security. In this paper, we propose a more effective, usable, and secure cryptocurrency wallet-management system based on semi-trusted social networks, therein allowing users to collaborate with involved parties to achieve some powerful functions and recovery under certain circumstances. Furthermore, we employ an identity-based hierarchical key-insulated encryption scheme to achieve time-sharing authorization and present a semitrusted portable social-network-based wallet-management scheme that provides the features of securityenhanced storage, portable login on different devices, no-password authentication, flexible key delegation, and so on. The performance analysis shows that our proposed schemes require minimal additional overhead and have low time delays, making them sufficiently efficient for real-world deployment.

Credit scoring models are the cornerstone of the modern financial industry. After years of development, artificial intelligence and machine learning have led to the transformation of traditional credit scoring models based on statistics. In this study, a novel multi-stage ensemble model with a hybrid genetic algorithm is proposed to achieve accurate and stable credit prediction. To alleviate the adverse effects of imbalanced data in credit scoring models, the Instance Hardness Threshold method is extended using a majority voting strategy to deal with data imbalance. To eliminate redundant and irrelevant features in the dataset and select well-performing base classifiers, a new hybrid genetic algorithm is proposed to obtain the optimal feature subset and base classifier subset. To aggregate the predictive power of the base classifiers, a stacking approach is adopted to integrate the optimal base classifiers into the ensemble model. The proposed model is tested on three standard imbalanced credit scoring datasets, compared with similar state-of-the-art approaches, and evaluated using four well-known evaluation indicators. The experimental results prove the effectiveness of the proposed model and demonstrate its superiority.

The rapid growth in the scale of the intelligent connected vehicles (ICVs) brings about the need for novel random access schemes that realize collision avoidance and reliable communication. The coded random access (CRA) scheme utilizes successive interference cancellation (SIC) across slots to enable the extraction of multiple users' packets from the collision slots and proves to be an efficient solution to collision resolution. In this paper, we propose a random backoff-based CRA scheme to achieve a tradeoff between the access success probability and access delay for CRA. Specifically, we partition the traditional CRA frame into subframes, so that the user access delay is reduced to the same scale of the subframe length. In addition, the backoff operation among subframes is introduced so that access failures originated from the subframe length limit can be reduced with the retransmission in subsequent subframes. In particular, we consider practical Nakagami-m fading channel in the CRA performance analysis and take into account the capture effect in the iterative interference cancellation process. Simulation results show that backoff-based CRA can significantly reduce the average access delay without severe negative effect on the access success probability for moderate system load compared with traditional CRA. Comparisons with the state-of-the-art non-orthogonal random access (NORA) scheme proposed for 5G indicates that backoff-based CRA provides a feasible solution for random access of large-scale ICVs.

Radio frequency (RF) relative measurement provides an effective means of communication and autonomous navigation for the spacecraft formation flying. Multipath and Doppler are two main factors that affect the accuracy of RF relative measurement. In this study, a framework that exploits joint multipath-Doppler diversity is proposed to reduce multipath errors and improve the measurement accuracy. The proposed framework first utilizes an Extended Kalman Filter (EKF) estimator to estimate the parameters of Doppler and multipath. Different from existing research, the proposed framework reconstructs the received signal according to the estimated parameters, so as to mitigate multipath signals and enhance the direct signal. Numerical results demonstrate that the framework is suitable for both multipath and multipath-Doppler scenarios, and has a significant performance improvement over existing multipath mitigation methods.

To improve the interpolation accuracy for the non-uniform rational B-spline (NURBS) curves, this paper proposes a novel bidirectional adaptive feedrate scheduling (BAFS) method based on S-shaped acceleration/deceleration (ACC/DEC) algorithm. The NURBS interpolator with the proposed BAFS method is also presented. The BAFS method has two meanings. First, to consider the continuous constraints of feedrate especially in the areas near the endpoints of each NURBS curve segment, an adaptive feedrate scheduling method is designed, which can improve the interpolation accuracy compared with the conventional method only considering the constraints of endpoints. Second, a bidirectional adaptive interpolation strategy including an adaptive task scheduling method and a meeting processing method is proposed to conduct the acceleration and deceleration interpolation stages of each NURBS segment separately in two directions. The two directions interpolation can be realized orderly by the adaptive task scheduling and the smoothness of feedrate profile can be guaranteed by the meeting processing method whose solvability is also proved. Therefore, the feedrate constraints at both the start and end points areas can be considered, which can improve the interpolation accuracy. Meanwhile, the round-off errors caused by cycle sampling are also considered and compensated in the bidirectional feedrate scheduling process to maintain the motion smoothness. In addition, an optimized look-ahead strategy is designed to ensure the correctness of the endpoints feedrate. Finally, a series of simulations and experiments are conducted to validate the good performance and applicability of the proposed method.