20 November 2025, Volume 32 Issue 11
    

  • Select all
    |
  • Stateless Control for Suspension System Combining Optimal Control and Observer
    WU Jianwei, JIANG Qiubo, FU Qidi, SUN Beibei
    Control Engineering of China. 2025, 32(11): 1921-1928.
    Abstract ( )   Knowledge map   Save
    For the limitation of existing control methods for obtaining intermediate states, a practical control method (stateless control) is proposed for suspension system by combining the linear quadratic regulator with the Luneberger observer. It directly establishes the relationship between current control input with historical control inputs and outputs. Using a quarter-car suspension system model as an example, the practical control performance of the stateless method is investigated. It incorporates sensor noise, modeled according to sensor precision, and accounts for suspension system parameter uncertainties. The results demonstrate that under both excitation and random excitation, the stateless control method effectively balances the three ride comfort performance indices of the suspension system, achieving excellent overall control performance. Crucially, the stateless control method only requires historical sensor outputs and control inputs to realize optimal control effectiveness even with limited sensors, thus significantly simplifying the control system design and offering significant practical value.
  • Scheduling Strategy for Unmanned Inspection System in Underground Mines Based on Supervisory Control
    ZHANG Mingzhen, LIU Chao, WANG Xiaodong, MA Weidong, SHEN Yi, TAI Ruochen
    Control Engineering of China. 2025, 32(11): 1929-1936.
    Abstract ( )   Knowledge map   Save
    In recent years, inspection robots have greatly improved the safety of mine operations. However, the unstructured environment of mines poses difficulties in the high cost and maintenance to track inspection robots, and the explosion-proof requirement of mines limits the hardware design of inspection robots. A mine inspection unmanned aerial vehicle (UAV) system that meets intrinsic safety standards is proposed for the first time, and designs low-power UAVs and charging ground cabins for the explosion-proof requirement. In addition, the UAV inspection scheme is adopted to overcome the difficulties caused by the unstructured environment of mines. Furthermore, a scheduling strategy model for unmanned inspection tasks in underground mines based on automata is established, and synthesized a complete scheduling strategy for unmanned inspection tasks in underground mines based on the supervised control theory. Finally, the effectiveness of the scheduling strategy is verified through scheduling experiments with a UAV and a ground handling cabin.
  • PID Design of Quad-rotor Aircraft Based on Multi-objective Genetic Algorithm
    WU Zhenlong, ZHANG Can, LIU Yanhong
    Control Engineering of China. 2025, 32(11): 1937-1946.
    Abstract ( )   Knowledge map   Save
    PID control algorithm is a widely used control strategy with reliable control performance, and it has many advantages such as simple structure, easy tuning and easy implementation. In order to solve the parameter tuning problem of PID controller with robustness constraint, we apply non-dominated sorting genetic algorithms-II (NSGA-II) to tune the PID controller parameters and applies the optimized PID to the control of flight attitude and altitude of quad-rotor aircraft. Firstly, the quad-rotor aircraft model is linearized. For the altitude and attitude channels of full drive control, the linear model of the sub-channel is obtained. Then, in the NSGA-II, PID parameters are selected as decision variables, the rejection ability performance and tracking performance are taken as the two objective functions, and the maximum sensitivity function is taken as the robustness constraint. Finally, by comparing the parameter control performance of the simulations with that of the traditional method, the superiority of multi-objective genetic algorithm tuning PID in tracking and jamming is verified.
  • Backstepping Height Control for Pneumatic Suspension System with Prescribed Performance
    AN Runxing, Pak Kin WONG, TANG Chuanyin, LIANG Zhongchao, ZHAO Jing
    Control Engineering of China. 2025, 32(11): 1947-1954.
    Abstract ( )   Knowledge map   Save
    The electric control air suspension can improve the comfort and handle stability by adjusting the vehicle height. A quarter vehicle air suspension model is established for air suspension system height adjustment in this research, then a vehicle height adjustment controller combining the prescribed performance function and Backstepping method is designed to guarantee that: ① the ride height of the vehicle can converge on a neighborhood of the desired height; ② the height tracking error is inside the boundary of the performance function. In addition, a sliding mode observer is designed to estimate the air pressure in the air spring considering the missing air pressure sensor in practical application. Finally, the effectiveness and application of the proposed control algorithm are verified by AMESim with Simulink co-simulations and hardware-in-the-loop test.
  • Fuzzy Control of Nonlinear Active Suspension System with Finite Frequency Domain Constraint
    JIA Junru, ZHENG Liming, ZHANG Zhen
    Control Engineering of China. 2025, 32(11): 1955-1963.
    Abstract ( )   Knowledge map   Save
    A fuzzy finite frequency output feedback control method is proposed considering the finite frequency characteristics of pavement interference and human sensitive frequency segment for nonlinear active suspension system. Firstly, the Takagi-Sugeno (T-S) fuzzy method was used to characterize the time-delay system of nonlinear active suspension, so as to vividly describe the nonlinear characteristics of spring and damper. Secondly, the fuzzy finite-frequency state feedback controller is designed based on the finite-frequency performance index and the condition of the robust performance. Thirdly, considering that some states of the suspension system are difficult to be measured online in practice, a fuzzy finite-frequency output feedback controller design method on the basis of matrix separation technology is produced. Finally, the experimental results verify the effectiveness of the proposed fuzzy controller.
  • A Lightweight CNN Human Activity Recognition Model
    JIAN Xianzhong, LIU Bingyan, HUANG Hong
    Control Engineering of China. 2025, 32(11): 1964-1971.
    Abstract ( )   Knowledge map   Save
    For the low recognition accuracy and several parameters of the current human activity recognition model, a lightweight convolutional neural network (CNN) human activity recognition model is proposed. Firstly, the sensor data is preprocessed. Then, the processed data is inputted into the CNN model to identify the specificity of the human body activities. Finally, the squeeze-and-excitation (SE) attention module is embedded in the feature extraction backbone network, and different weights are assigned to each convolution channel to strengthen key features and improve model accuracy. The performance of the model is evaluated on three public datasets of UCI-HAR, WISDM and OPPORTUNITY. The 1F of the model on the UCI-HAR dataset is 97.54%, with 17 198 parameters. 1F on the WISDM dataset is 97.66%, with 16 622 parameters; 1F on the OPPORTUNITY dataset is 82.38%, with 27 545 parameters. Compared with the existing advanced human activity recognition models, the recognition accuracy is higher, the model parameters are fewer, and the model generalization ability is enhanced.
  • Bearing Life Prediction Algorithm Based on Improved Multi-scale Temporal Convolutional Network
    WANG Shuai, LIU Yefeng, LIU Jingjing, SUN Hanlin, WEN Xuezhi
    Control Engineering of China. 2025, 32(11): 1972-1978.
    Abstract ( )   Knowledge map   Save
    Rolling bearings are commonly used in rotating machinery, and their remaining useful life (RUL) prediction is crucial to ensure safe system operation and reduce maintenance costs. Current methods usually extract features from a single scale only, ignoring information from other scales, and suffer from the problem of insufficient feature extraction. Therefore, a bearing RUL prediction model based on multi-head attention (MA) with multi-scale temporal convolutional network (MsTCN) is proposed. The model utilizes MsTCN to extract multi-scale features and introduces the MA mechanism to assist MsTCN. MA is located at the head of MsTCN and captures important features by processing multiple attention mechanisms in parallel, thus improving the extraction efficiency of MsTCN. Experiments are conducted on the PHM 2012 dataset, and the experimental results show that the proposed method outperforms the other comparative models in at least 10 values of MAE, RMSE, and Score, indicating that the proposed prediction framework improves the prediction accuracy, has better robustness, and is suitable for RUL prediction of rolling bearings.
  • High Density Intra-oral Point Cloud Registration with Spatial Consistency
    CAI Xianhang, CHEN Sheng
    Control Engineering of China. 2025, 32(11): 1979-1987.
    Abstract ( )   Knowledge map   Save
    Digital intraoral scanning technology has gradually become an important part of modern oral medicine. In order to solve the real-time registration problem caused by dense point clouds with low signal-to-noise ratio in the mouth, an efficient registration model—3D-spatial consistent grouping (3D-SCG) is proposed. The Tooth-Net sub-module is designed based on the attention mechanism and the spatial characteristics of point clouds to extract the high-dimensional geometric spatial features of point pairs. Design the Outlier Removal module in the point cloud to effectively eliminate outlier pairs in the point pair set. Four high-density large and four small-scale point cloud datasets, as well as a real oral tooth point cloud dataset, were fabricated by scanning with an oral scanning instrument, which were respectively used to train and test the feasibility of real-time matching of the 3D-SCG model. Compared with the traditional registration scheme RANSAC+ICP, our model achieves the same registration accuracy while reducing the overall time consumption by 80% and lowering the proportion of matching failures by 33%. The experimental results show that the proposed registration model effectively improves the quality and speed of dense point cloud registration with low signal-to-noise ratio in the oral cavity, ensuring the real-time performance requirements of the oral scanner.
  • Dual Semantic-controlled Conversion of Virtual to Real Images for Maritime Unmanned Boats
    LIU Fan, WANG Qihang, ZHU Yifan, XIANG Lei, JING Shilong, ZHOU Yuan
    Control Engineering of China. 2025, 32(11): 1988-1995.
    Abstract ( )   Knowledge map   Save
    To address the significant differences between virtual unmanned surface vessel images simulated by hardware-in-the-loop systems and real images – which cause models trained on virtual data to perform poorly in real-world scenarios – we propose a virtual-to-real image conversion method based on a dual semantic control diffusion model (DSCDM). The approach utilizes both grayscale and RGB semantic images as conditions to extract semantic and spatial information through class-adaptive normalization and cross-attention. Comparative experiments show that on real datasets, our method reduces the fréchet inception distance (FID) by 6.7 on average and increases the mean intersection over union (mIoU) by nearly 1%. On virtual datasets, FID drops by 12.6 on average with mIoU matching state-of-the-art methods. Ablation experiments demonstrate that our method achieves complementary integration of semantic category information and spatial features with fewer parameters, generating higher-quality real images. These results validate the practical value of our approach for augmenting real datasets.
  • College Students’ Comprehensive Grade Prediction Method Based on Multi-model Fusion
    LI Shipeng, WEI Gaoqi, ZHAO Ziyan
    Control Engineering of China. 2025, 32(11): 1996-2003.
    Abstract ( )   Knowledge map   Save
    The rapid advancement of information and artificial intelligence technology heralds new opportunities for innovation and transformation in higher education. Combining them with higher education teaching has led to innovative applications of data-driven assisted teaching decision-making. Traditional grade analysis and prediction models have the defects of single structure, poor prediction effect and a plethora of required features. By analyzing the characteristics of a variety of machine learning models and taking full advantages of basic models, a Stacking model is constructed and optimized through Bayesian hyperparameter optimization. It realizes the accurate prediction of college students’ comprehensive grades and can achieve ideal prediction results with fewer input features. The academic grades of students majoring in Automation at Northeastern University are used as experimental data to verify the model. The constructed comprehensive grade prediction model has guiding significance in assisting the teaching research department in the process of curriculum and discipline construction.
  • Harmonic Detection Based on Improved VMD and Wavelet Adaptive Threshold Denoising
    ZHAO Guoqiang, CAO Yudong, FU Hua
    Control Engineering of China. 2025, 32(11): 2004-2013.
    Abstract ( )   Knowledge map   Save
    Variational mode decomposition (VMD) has end effect on the natural modes of signal decomposition, decomposition effect is affected by noise and the K value needs to be determined in advance. The waveform extension method combining waveform matching degree and instantaneous frequency is proposed. The improved wavelet threshold function is used to de-noise the signal. After de-noising, the optimal K value is selected using waveform amplitude characteristics. Simulation results show that the algorithm can effectively improve the VMD endpoint effect problem, reduce the noise effect, and achieve the optimal K value VMD decomposition. Compared with the multivariate ensemble empirical mode decomposition (MEEMD) and the complementary ensemble empirical mode decomposition with adaptive noise (CEEMDAN) algorithms, the proposed algorithm is more effective.
  • Important-data-based DoS Attack Design and Control for UDS
    WANG Maodong, TIAN En’gang, WANG Jing, QV Feng
    Control Engineering of China. 2025, 32(11): 2014-2020.
    Abstract ( )   Knowledge map   Save
    A novel attack-defense framework is designed for the unmanned driving system(UDS). Firstly, a novel denial-of-service (DoS) attack mechanism, called important-data-based (IDB) DoS attack strategy, is proposed from the perspective of adversaries to reinforce the destructiveness of the attack. Different from most of the existing DoS attack strategies, the proposed one can intercept the output data, identify the importance degree of the data and only launch attacks for the important data. In this way, the proposed IDB DoS attack strategy will cause larger attack damages. Secondly, in order to resist the negative effect caused by the IDB DoS attack strategy, a novel resilient H∞ control technique is constructed to alleviate the attack damages. Finally, simulation results demonstrate that: ① The proposed IDB DoS attack strategy will result in worse system degradation when compared with some existing DoS attack models. ② The designed resilient H∞ controller can effectively mitigate the attack destructiveness of the proposed IDB DoS attack strategy.
  • Identification of Wiener Model Wind Power System Based on Correlation Analysis
    DING Zhenyu, LI Feng, HE Naibao, CAO Qingfeng
    Control Engineering of China. 2025, 32(11): 2021-2028.
    Abstract ( )   Knowledge map   Save
    Wind power generation systems are characterized by strong nonlinearity and large volatility. Establishing high-precision models and achieving effective predictions under random wind speed conditions is the current research focus. A Wiener model modeling and identification method based on correlation analysis is proposed for 3.5 MW wind turbine units. Firstly, 3σ data cleaning is adopted to remove outliers, and the agent model is trained using the cleaned data. Secondly, a Wiener model is constructed, with the separable signal and the actual wind speed as the input, and the output of the proxy model and the actual power as the output. Finally, correlation analysis and recursive augmented stochastic gradient method with forgetting factor are respectively adopted to identify the parameters of dynamic linear and static nonlinear subsystems. The simulation results show that the average absolute percentage error of the proposed method is 0.55% lower than that of the recursive least squares method, and 2.46% lower than that of the more new interest augmented stochastic gradient method. The method can effectively improve the modeling and prediction accuracy of wind power systems.
  • Closed-loop Optimal Control Strategy of Drilling System Based on Hierarchical Structure
    XU Baochang, LU Yaoyao, MENG Zhuoran, LIU Wei
    Control Engineering of China. 2025, 32(11): 2029-2038.
    Abstract ( )   Knowledge map   Save
    With the improvement of drilling automation, drilling objectives tend to be more unified in economy, safety and efficiency. The control of traditional drilling engineering is carried out only by drilling experience, which is difficult to meet the drilling requirements in terms of control accuracy and drilling efficiency. Therefore, a closed-loop optimal control strategy of drilling system based on hierarchical structure is proposed to achieve efficient drilling in the safety pressure window. The upper layer takes the optimal mechanical specific energy as the optimization objective, and combines the hard constraints such as pressure window to carry out steady-state optimization to obtain the optimal drilling parameters under the current working conditions. The lower layer takes some parameters obtained from the upper layer as the optimal set value, and combines the bottom hole pressure control objective for optimal control. The experimental results show that the control strategy can comprehensively improve the drilling speed and drilling
  • Predictive Control of Rare Earth Extraction Process with Unmodeled Dynamic Compensation
    LU Rongxiu, LIU Xiaoxia, YANG Hui, DAI Wenhao, ZHU Jianyong
    Control Engineering of China. 2025, 32(11): 2039-2047.
    Abstract ( )   Knowledge map   Save
    Rare earth extraction process is a typical multivariable nonlinear system, which often faces uncertain factors such as external unknown interference and internal parameter estimation error. It can easily cause instability in the control system. In view of this, a generalized predictive control method with unmodeled dynamic compensation is proposed. Firstly, according to the rare earth extraction process technology, the extraction process is described as the sum of linear model and unmodeled dynamic. On this basis, a multivariable generalized predictive controller suitable for rare earth extraction process is designed. And the extreme learning machine optimized by slime mold algorithm is used to estimate the unmodeled dynamics and a compensator is designed to eliminate its influence on the system. By comparing with the backpropagation- proportional integral derivative controller, the results show that the proposed method can reduce the amount of extractant and detergent on the whole on the basis of ensuring small overshoot and short regulation time.
  • Performance Assessment Method of Cold Rolling Single Loop Control System Based on Hurst Exponent
    LI Shu, GUO Xiangchen, HU Yunjian, PENG Wen, SUN Jie, ZHANG Dianhua
    Control Engineering of China. 2025, 32(11): 2048-2054.
    Abstract ( )   Knowledge map   Save
    To evaluate the control performance of the cold rolling single-loop control system and determine the controller performance quality under the current operating conditions, an approach for control system performance assessment based on the Hurst exponent is adopted to address the issue that the minimum variance performance evaluation method relies on the mathematical model of the controlled object. This method is based on the regular operation data of the rolling process, which can avoid solving the mathematical model of the controlled object required in the minimum variance method and enable the acquisition of the controller performance quality of the current rolling system loop. Through the operation data of the simulation model, this index is used to evaluate the simulation data results and optimize the controller. The method is also applied to the actual rolling production process to verify its feasibility and effectiveness. Experimental results show that the Hurst exponent can accurately measure the current performance of the controller. Moreover, it can be used to diagnose the characteristics of the current control loop based on the Hurst exponent, and combined with the characteristics of control parameters, provide optimization suggestions for the current controller parameters. After parameter optimization, the performance of the controller is significantly improved compared with the initial performance.
  • Spatial Dynamic Linearization Based Iterative Learning Formation Control
    SUN Shiyong, CHI Ronghu, LIN Na
    Control Engineering of China. 2025, 32(11): 2055-2064.
    Abstract ( )   Knowledge map   Save
    A spatial dynamic linearization based iterative learning formation control (SDL-ILFC) method is proposed for nonlinear nonaffine multi-agent systems with single sequence structure. Firstly, the spatial dynamic linear data model among adjacent agents is established to reveal the dynamic relationship between the parent and the child agents of the multi-agent system. Then, considering the iteration-varying formation control task, an iterative learning formation control law and a parameter updating law are designed based on the quadratic performance index. The proposed SDL-ILFC method utilizes the information of previous iterative operations, and utilizes the information of parent agents, which improves the control performance. Meanwhile, the proposed method does not require the system to satisfy the assumption of the same initial conditions, so it is more suitable for actual systems. The design and analysis of this controller directly utilize system I/O data, without requiring an accurate system dynamics model. Mathematical analysis proves the bounded convergence of the formation tracking error. Simulation studies further verify the effectiveness of the proposed method.
  • Sliding Mode Consensus Control Based on ET-WTOD Multi-node Transmission Hybrid Scheduling for Multi-agent Systems
    DING Meng, CHEN Bei
    Control Engineering of China. 2025, 32(11): 2065-2072.
    Abstract ( )   Knowledge map   Save
    The problem of consensus tracking control for a class of multi-agent systems under constrained communication channels is studied. A hybrid ET-WTOD multi-node transmission protocol combining event triggering (ET) mechanism and weight try-once-discard (WTOD) protocol is proposed, i.e., during information transmission between agents, each state variation of agent is first compared with two thresholds to determine whether each state component satisfies the trigger condition, and the WTOD protocol is introduced to filter the state components between the two thresholds. Therefore, the number of transmitted components may be dynamically regulated to improve the flexibility of transmission under limited bandwidth. Then, a consensus error function under the ET-WTOD scheduling protocol is constructed, a linear sliding mode function and a distributed sliding mode controller are designed, and the corresponding stability conditions are obtained. Finally, the numerical simulation verifies that the designed distributed sliding mode control strategy can guarantee the consensus of multi-agent system under the communication scheduling protocol.
  • Research and Application of an Optimization Method for Power Grid Error Prevention and Control
    MENG Chengzhen, HAO Fangzhou, LI Huanhuan, LUO Qi, WEI Zhijun, MA Yihang
    Control Engineering of China. 2025, 32(11): 2073-2080.
    Abstract ( )   Knowledge map   Save
    The increase in electrical equipment will lead to an increased risk of misoperation. A power grid error prevention regulation control system based on edge computing and improved support vector machine is proposed. Combining deep learning with error prevention analysis theory, a stack sparse autoencoder is introduced to construct an error prevention analysis neural network model. The experimental results show that when the dataset size is 1 000, the accuracy of the kernel ridge regression model, gradient boosting machine model, support vector machine model, and improved support vector machine model are 0.91, 0.93, 0.96, and 0.98, respectively. The improved support vector machine model has a judgment accuracy of 98.1%, 98.6%, 96.7%, and 89.4% for controllable operations, erroneous operations, pending confirmation operations, and all operations, respectively. The research results indicate that the proposed model can effectively achieve optimization of power grid error prevention and control, providing a reliable and feasible method for the field of power grid error prevention.
  • Finite-time Synchronization and Energy Consumption Estimation for Fractional-order Coupled Neural Networks
    WEI Changjiang, MA Ben, TONG Dongbing, CHEN Qiaoyu, ZHOU Wuneng
    Control Engineering of China. 2025, 32(11): 2081-2086.
    Abstract ( )   Knowledge map   Save
    The study concentrates the finite-time synchronization of a class of fractional-order coupled neural networks and the upper bound problem of estimating synchronization time and energy consumption. First, using fractional calculus theory and inequality techniques, sufficient conditions are provided to ensure finite-time synchronization of the considered system. Second, based on the finite-time theory, the upper bounds of synchronization time and energy consumption in the synchronization process of fractional-order coupled neural networks are obtained. The method extends the research results of energy consumption in network control to the fractional-order field, which greatly expands the application scope of the system. In addition, the influence of system order on synchronization time and energy consumption is revealed through MATLAB simulation. Finally, a numerical example is provided to demonstrate the validity of the obtained results.
  • Control of Permanent Magnet Synchronous Motor Based on Hardware in Loop System
    ZHANG Shouhe, BAI Rui
    Control Engineering of China. 2025, 32(11): 2087-2095.
    Abstract ( )   Knowledge map   Save
    PMSM is widely used in the motor speed regulating system because of its small size and high the power density. PI control algorithm is mostly utilized in the PMSM controller. In fact, PI controller has some disadvantages such as the slow response speed, poor anti-disturbance ability. It is difficult to solve the contradiction between stability and speed. Therefore, for the speed control system of PMSM, the excellent control performance cannot be implemented by PI control. In order to solve the above-mentioned problems, linear active disturbance rejection controller is used to replace PI controller in the speed loop of PMSM vector control system. Meanwhile, a hardware-in-loop experiment system of PMSM is established using NI platform. The PMSM vector control system based on PI controller and LADRC controller are studied by the established experiment system. In the experiment results, compared with PI algorithm, the LADRC algorithm has stronger anti-interference ability, so that the motor controller has better control effect. Meanwhile, the hardware platform provides the valuable experience for the rapid design and development of high-performance motor controller.
  • Quality-related Dynamic Slow Feature Analysis for Fault Detection
    SHAO Yuanzhe, ZHAO Zhonggai, LIU Fei
    Control Engineering of China. 2025, 32(11): 2096-2104.
    Abstract ( )   Knowledge map   Save
    Product quality is key-performance-indicator in industrial process monitoring. The previous quality-related monitoring algorithms only focus on the monitoring statistics of the steady-state characteristics of the system, and cannot distinguish between changes in operating conditions caused by changes in operating conditions and real process failures, which is detrimental to the continuity and economy of production. Canonical correlation analysis (CCA) and dynamic slow feature analysis (DSFA) are combined to construct a new optimization objective function to extract the slow variables and their first-order derivatives that are strongly related to quality in the process data. They are the steady-state information and dynamic information that reflect the essential changes of the system, respectively. The method decomposes process variables into quality-related subspaces and quality-unrelated subspaces. Then the corresponding steady-state and dynamic monitoring indicators are established in the two subspaces respectively. The method uses the process data obtained in real time to detect faults and judge whether they affect the quality. It can distinguish between normal operating conditions deviation and real faults and has good dynamic performance. Finally, the effectiveness of the proposed method is verified by comparing the previous methods through simulation experiments.
  • Sensor Network Cloud Intrusion Behavior Detection Based on Improved BP Neural Network
    YUAN Jinming, GENG Xiaofen, NA Chongzheng
    Control Engineering of China. 2025, 32(11): 2105-2112.
    Abstract ( )   Knowledge map   Save
    Sensor network cloud intrusion detection is easily affected by the uneven energy consumption of interactive information nodes, which leads to a decrease in the performance of some nodes and a decrease in the accuracy of intrusion behavior detection. To address this, an improved BP neural network-based sensor network cloud intrusion behavior detection method is proposed. Firstly, the sparse projection data algorithm is used to collect the sparse projection data of the sensor network cloud. Then, the sparse representation based learning method is used to perform sparse representation on the collected data, in order to obtain sensing network cloud data features with spatiotemporal correlation. Finally, by adaptively adjusting the learning rate and summing and accumulating to improve the neural network, the sensor network cloud data feature data is used as the network input to achieve sensor network cloud intrusion detection. Through experiments, it has been proven that the proposed method achieves a recognition rate of over 96.7%, a detection speed of only 34 ms, an average fluctuation coefficient of less than 0.20, and a CPU utilization rate of only 14% at its peak, demonstrating good intrusion detection performance.
Office Online
Journal Online
Download
More>>
京ICP备05021913号-100
Copyright © Control Engineering of China, All Rights Reserved.
Tel: 024-23883498（传真），024-83688973-16/17/18　
E-mail：kzgcbjb@mail.neu.edu.cn
Powered by Beijing Magtech Co. Ltd.