This research provides a novel approach for detecting multi-legged robot actuator *** most significant concept is to design the Fault Diagnosis Generative Adversarial Network(FD-GAN)to fully adapt to the fault diagnos...
详细信息
This research provides a novel approach for detecting multi-legged robot actuator *** most significant concept is to design the Fault Diagnosis Generative Adversarial Network(FD-GAN)to fully adapt to the fault diagnosis problem with insufficient *** found that it is difficult for methods based on classification and prediction to learn failure patterns without enough data.A straightforward solution is to use massive amounts of normal data to drive the diagnostic *** introduce frequency-domain information and fuse multi-sensor data to increase the features and expand the difference between normal data and fault data.A GAN-based framework is designed to calculate the probability that the enhanced data belongs to the normal *** uses a generator network as a feature extractor,and uses a discriminator network as a fault probability evaluator,which creates a new use of GAN in the field of fault *** the many learning strategies of GAN,we find that a key point that can distinguish the two types of data is to use the hidden layer noise with appropriate discrimination as the *** also design a fault location method based on binary search,which greatly improves the search efficiency and engineering value of the entire *** have conducted a lot of experiments to prove the diagnostic effectiveness of our architecture in various road conditions and working *** compared FD-GAN with popular diagnostic *** results show that our method has the highest accuracy and recall rate.
In this paper, an event-triggered time-varying formation tracking control for a class of second-order nonlinear multiagent systems(MAS) operating within a constrained region is investigated. To mitigate the negative e...
详细信息
In this paper, an event-triggered time-varying formation tracking control for a class of second-order nonlinear multiagent systems(MAS) operating within a constrained region is investigated. To mitigate the negative effects of external unknown disturbance, a novel disturbance observer with performance guarantees is proposed, enabling precise disturbance *** the artificial potential field(APF) method, a repulsive potential function is introduced to prevent inter-agent collisions as well as collisions with environmental obstacles. To reduce continuous communication and frequent system updates, a sliding mode technique is incorporated into the formation tracking controller, utilizing an event-triggered mechanism. The controller is also applicable to the formation control of MAS in switching-constrained regions. The achievement of the specified timevarying geometric formation is rigorously demonstrated through the Lyapunov framework. Numerical simulations are presented to validate the effectiveness of the theoretical results.
This paper investigates an interval analysis method for neural networks and applies it to fault detection for systems with unknown but bounded measurement noise. First, a novel interval analysis method is presented, w...
详细信息
This paper investigates an interval analysis method for neural networks and applies it to fault detection for systems with unknown but bounded measurement noise. First, a novel interval analysis method is presented, which can compute the bounds of the output of a feedforward neural network subject to a bounded input. By applying the proposed interval analysis method to a network trained with fault-free system data, adaptive thresholds for fault detection are computed. Finally, one can acquire fault detection results via a fault detection strategy. The proposed method can achieve tight bounds of the network output and employ simple operations, which leads to accurate fault detection results and a low computational burden.A numerical simulation and an experiment on an AC servo motor are given to illustrate the effectiveness and superiority of the proposed method.
This paper develops distributed algorithms for solving Sylvester *** authors transform solving Sylvester equations into a distributed optimization problem,unifying all eight standard distributed matrix *** the authors...
详细信息
This paper develops distributed algorithms for solving Sylvester *** authors transform solving Sylvester equations into a distributed optimization problem,unifying all eight standard distributed matrix *** the authors propose a distributed algorithm to find the least squares solution and achieve an explicit linear convergence *** results are obtained by carefully choosing the step-size of the algorithm,which requires particular information of data and Laplacian *** avoid these centralized quantities,the authors further develop a distributed scaling technique by using local information *** a result,the proposed distributed algorithm along with the distributed scaling design yields a universal method for solving Sylvester equations over a multi-agent network with the constant step-size freely chosen from configurable ***,the authors provide three examples to illustrate the effectiveness of the proposed algorithms.
The damage caused by the earthquake is *** the purpose of post-earthquake rescue is to reduce *** it is necessary to improve the efficiency of rescue as much as *** paper focuses on multi-agent cooperative target sear...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
The damage caused by the earthquake is *** the purpose of post-earthquake rescue is to reduce *** it is necessary to improve the efficiency of rescue as much as *** paper focuses on multi-agent cooperative target search in the post-earthquake rescue and how to exploit the prior information of post-disaster building to improve the searching efficiency.A regional necessity based searching strategy is proposed which utilizes pre-disaster building information and adapts to dynamic ***,the target probability map is estimated according to the distribution information of pre-disaster ***,the regional necessity map is constructed by integrating the target probability map and the perception of the environment during the search ***,the search strategy is proposed based on the regional necessity map and the particle swarm optimization *** the gradient direction of regional necessity and local optimal velocity,the agent dynamically selects the search *** comparing with other search strategies in simulation experiments,the proposed strategy could shorten the searching time and improve the searching efficiency in the small and medium-scale task ***,the search strategy can adapt to the post-earthquake dynamic environment,which is confirmed by dynamic experiments.
When arriving at the airport,flight needs to be served by special *** at the dynamic time window scheduling problem of airport refueling vehicles,this paper establishes a vehicle routing problem model with time window...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
When arriving at the airport,flight needs to be served by special *** at the dynamic time window scheduling problem of airport refueling vehicles,this paper establishes a vehicle routing problem model with time window to minimize the operating ***,a multi-strategy genetic algorithm is designed to gain the solve time window scheduling problem,which employs the crossover based on particle swarm optimization to accelerate the early search capability,and the local search method based on simulated annealing to increase the local optimization *** aiming at dynamically adjusting vehicle routes on the basis of static scheduling,a local replanning strategy based on a dynamic time window is introduced,which uses the original route matching and rescheduling *** results show that the multi-strategy hybrid algorithm can effectively reduce the number of routes and vehicles the airport *** different scales' dynamic changes of time windows,the algorithm could enable vehicles to still meet time constraints and effectively minimize the change of routes.
In order to study to design a photoelectric intelligent platform based on multifunctional edge computing equipment by using photoelectric detection equipment TC505C, RK-Series Development Kits computing platform and N...
详细信息
The localization accuracy of visual SLAM depends on the image ***,in postdisaster rescue missions,the images obtained by the camera often contain considerable noise,which affects the pose estimation based on visual **...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
The localization accuracy of visual SLAM depends on the image ***,in postdisaster rescue missions,the images obtained by the camera often contain considerable noise,which affects the pose estimation based on visual *** this paper,we study the influence of random impulse noise in images on the localization accuracy of visual SLAM,and reduce these influences by denoising and removing ***,the camera image is preprocessed by the traditional image noise reduction *** at the problem of a large number of mismatches in optical flow tracking due to the influence of residual noise,the improved random sample consensus method is adopted to remove *** judge the correct matching probability of optical flow tracking results by normalized cross-correlation matching before random *** use guided sampling to select matching points to estimate the camera motion model,to increase the robustness of the SLAM ***,our method is verified in the open-source solution *** show that after random impulse noise is added to the KITTI dataset,the pose estimation accuracy of the improved SLAM is higher than the pose estimation accuracy after noise reduction only,and it is also higher than the pose estimation results of the original images in multiple sequences of the KITTI dataset.
With the development of deep learning technology,the dehazing method based on convolutional neural network has also developed ***,it still faces some problems such as incomplete dehazing and difficulty in detail *** a...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
With the development of deep learning technology,the dehazing method based on convolutional neural network has also developed ***,it still faces some problems such as incomplete dehazing and difficulty in detail *** at the problems,we propose a dehazing network based on U-Net structure and residual *** the down sampling process,we introduce the residual block of attention mechanism,which effectively improves the feature extraction and expression ability of the *** the middle part of the network,the smooth dilation convolution residual block is used to expand the receptive field of the network and improve the restoration effect of small *** the same time,we introduce a feature fusion mechanism based on attention mechanism,which can reduce the loss of information and weight the down-sampling features into intermediate *** the process of up sampling,the method of adding elements instead of feature stitching is used to reduce the number of parameters and realize the function of jump ***,dense convolution residual blocks are added to the jump connection between input and output to further improve the *** experimental results show that the PSNR of this method can reach 32.893,which is better than the existing methods.
In recent years,the introduction of Siamese network has brought new vitality to the object tracking ***,high-performance Siamese trackers cannot run at a real-time speed on mobile devices due to their complex and huge...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
In recent years,the introduction of Siamese network has brought new vitality to the object tracking ***,high-performance Siamese trackers cannot run at a real-time speed on mobile devices due to their complex and huge *** distillation is a common and effective model compression method,but it is difficult to be applied to the challenging task like object *** find out the fundamental cause is that the imbalance between the foreground and background in the object tracking task,which aggravates the problem of insufficient feature extraction ability of small ***,we propose the attention mask distillation(AMD) to help the student tracker focus on the foreground area faster and more *** attention mask can be easily obtained from the feature maps and brings fine-granularity to the traditional binary *** experimental results on OTB100 and VOT2018 show that our method enables the student tracker perform as well as the teacher *** the same time,it's able to run on the CPU at a hyper-real-time of 66 fps and achieves nearly 9 times model compression *** low computational and storage costs make it possible to deploy high-performance trackers on resource-constrained platforms.
暂无评论