Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis

In machinery fault diagnosis,labeled data are always difficult or even impossible to obtain.Transfer learning can leverage related fault diagnosis knowledge from fully labeled source domain to enhance the fault diagnosis performance in sparsely labeled or unlabeled target domain,which has been widely used for cross domain fault diagnosis.However,existing methods focus on either marginal distribution adaptation(MDA)or conditional distribution adaptation(CDA).In practice,marginal and conditional distributions discrepancies both have significant but different influences on the domain divergence.In this paper,a dynamic distribution adaptation based transfer network(DDATN)is proposed for cross domain bearing fault diagnosis.DDATN utilizes the proposed instance-weighted dynamic maximum mean discrepancy(IDMMD)for dynamic distribution adaptation(DDA),which can dynamically estimate the influences of marginal and conditional distribution and adapt target domain with source domain.The experimental evaluation on cross domain bearing fault diagnosis demonstrates that DDATN can outperformance the state-of-the-art cross domain fault diagnosis methods.

Balanced Discriminative Transfer Feature Learning for Visual Domain Adaptation

Transfer learning aims to transfer source models to a target domain.Leveraging the feature matching can alleviate the domain shift effectively,but this process ignores the relationship of the marginal distribution matching and the conditional distribution matching.Simultaneously,the discriminative information of both domains is also neglected,which is important for improving the performance on the target domain.In this paper,we propose a novel method called Balanced Discriminative Transfer Feature Learning for Visual Domain Adaptation(BDTFL).The proposed method can adaptively balance the relationship of both distribution matchings and capture the category discriminative information of both domains.Therefore,balanced feature matching can achieve more accurate feature matching and adaptively adjust itself to different scenes.At the same time,discriminative information is exploited to alleviate category confusion during feature matching.And with assistance of the category discriminative information captured from both domains,the source classifier can be transferred to the target domain more accurately and boost the performance of target classification.Extensive experiments show the superiority of BDTFL on popular visual cross-domain benchmarks.

Adaptive Time Frequency Distribution Based on Linear Chirp Modulated Gaussian Functions

We propose a method that uses linear chirp modulated Gaussian functions as the elementary functions, by adaptively adjusting variances, time frequency centers and sweep rates, to decompose signals. By taking WVD, an improved adaptive time frequency distribution is developed, which is non negative, free of cross term interference, and of better time frequency resolution. The paper presents an effective numerical algorithm to estimate the optimal parameters of the basis. Simulations indicate that the proposed approach is effective in analyzing signal's time frequency behavior.

Adaptive mechanisms of Ardisia crenata var.bicolor along an elevational gradient on Gaoligong Mountain,Southwest China

Plants overcome environmental stress by generating metabolic pathways.Thus,it is crucial to understand the physiological mechanisms of plant responses to changing environments.Ardisia crenata var.bicolor has an important ornamental and medicinal value.To reveal the impact of elevational gradient on the habitat soil and plant physiological attributes of this species,we collected root topsoil(0–20 cm)and subsoil(20–40 cm)samples and upper leaves at the initial blooming phase,in a survey of six elevations at 1,257 m,1,538 m,1,744 m,1,970 m,2,135 m,and 2,376 m,with 18 block plots,and 5sampling points at each site.Temperature decreases with an increase in elevation,and soil variables,and enzymatic activities fluctuated in both the topsoil and subsoil,with all of them increasing with elevation and decreasing with soil depth.Redundancy analysis was conducted to explore the correlation between the distribution of A.crenata var.bicolor along the elevational gradient and soil nutrients and enzyme activities,the soil properties were mainly affected by p H at low elevations,and governed by total phosphorus(TP)and available nitrogen(AN)at high elevations.The levels of chlorophyll,carbohydrates,and enzymatic activity except for anthocyanin in this species showed significant variation depending on physiological attributes evaluated at the different collection elevations.The decline in chlorophyll a and b may be associated with the adaptive response to avoid environmental stress,while its higher soluble sugar and protein contents play important roles in escaping adverse climatic conditions,and the increases in activities of antioxidant enzymes except peroxidase(POD)reflect this species’higher capacity for reactive oxygen scavenging(ROS)at high elevations.This study provides supporting evidence that elevation significantly affects the physiological attributes of A.crenata var.bicolor on Gaoligong Mountain,which is helpful for understanding plant adaptation strategies and the plasticity of plant physiological traits along the elevational gradients.

Design of Clustering Techniques in Cognitive Radio Sensor Networks

In recent decades,several optimization algorithms have been developed for selecting the most energy efficient clusters in order to save power during trans-mission to a shorter distance while restricting the Primary Users(PUs)interfer-ence.The Cognitive Radio(CR)system is based on the Adaptive Swarm Distributed Intelligent based Clustering algorithm(ASDIC)that shows better spectrum sensing among group of multiusers in terms of sensing error,power sav-ing,and convergence time.In this research paper,the proposed ASDIC algorithm develops better energy efficient distributed cluster based sensing with the optimal number of clusters on their connectivity.In this research,multiple random Sec-ondary Users(SUs),and PUs are considered for implementation.Hence,the pro-posed ASDIC algorithm improved the convergence speed by combining the multi-users clustered communication compared to the existing optimization algo-rithms.Experimental results showed that the proposed ASDIC algorithm reduced the node power of 9.646%compared to the existing algorithms.Similarly,ASDIC algorithm reduced 24.23%of SUs average node power compared to the existing algorithms.Probability of detection is higher by reducing the Signal-to-Noise Ratio(SNR)to 2 dB values.The proposed ASDIC delivers low false alarm rate compared to other existing optimization algorithms in the primary detection.Simulation results showed that the proposed ASDIC algorithm effectively solves the multimodal optimization problems and maximizes the performance of net-work capacity.

Adaptive distributed formation maintenance for multiple UAVs:Exploiting proximity behavior observations

The formation maintenance of multiple unmanned aerial vehicles(UAVs)based on proximity behavior is explored in this study.Individual decision-making is conducted according to the expected UAV formation structure and the position,velocity,and attitude information of other UAVs in the azimuth area.This resolves problems wherein nodes are necessarily strongly connected and communication is strictly consistent under the traditional distributed formation control method.An adaptive distributed formation flight strategy is established for multiple UAVs by exploiting proximity behavior observations,which remedies the poor flexibility in distributed formation.This technique ensures consistent position and attitude among UAVs.In the proposed method,the azimuth area relative to the UAV itself is established to capture the state information of proximal UAVs.The dependency degree factor is introduced to state update equation based on proximity behavior.Finally,the formation position,speed,and attitude errors are used to form an adaptive dynamic adjustment strategy.Simulations are conducted to demonstrate the effectiveness and robustness of the theoretical results,thus validating the effectiveness of the proposed method.

Tracking consensus for nonlinear heterogeneous multi-agent systems subject to unknown disturbances via sliding mode control

We investigate the tracking control for a class of nonlinear heterogeneous leader-follower multi-agent systems（MAS）with unknown external disturbances. Firstly, the neighbor-based distributed finite-time observers are proposed for the followers to estimate the position and velocity of the leader. Then, two novel distributed adaptive control laws are designed by means of linear sliding mode（LSM） as well as nonsingular terminal sliding mode（NTSM）, respectively. One can prove that the tracking consensus can be achieved asymptotically under LSM and the tracking error can converge to a quite small neighborhood of the origin in finite time by NTSM in spite of uncertainties and disturbances. Finally, a simulation example is given to verify the effectiveness of the obtained theoretical results.

Distributed adaptive direct position determination based on diffusion framework

The conventional direct position determination（DPD） algorithm processes all received signals on a single sensor.When sensors have limited computational capabilities or energy storage,it is desirable to distribute the computation among other sensors.A distributed adaptive DPD（DADPD）algorithm based on diffusion framework is proposed for emitter localization.Unlike the corresponding centralized adaptive DPD（CADPD） algorithm,all but one sensor in the proposed algorithm participate in processing the received signals and estimating the common emitter position,respectively.The computational load and energy consumption on a single sensor in the CADPD algorithm is distributed among other computing sensors in a balanced manner.Exactly the same iterative localization algorithm is carried out in each computing sensor,respectively,and the algorithm in each computing sensor exhibits quite similar convergence behavior.The difference of the localization and tracking performance between the proposed distributed algorithm and the corresponding CADPD algorithm is negligible through simulation evaluations.

Adaptive Energy-Efficient Power Allocation for DAS with Imperfect Channel State Information and Antenna Selection

Considering that perfect channel state information(CSI) is difficult to obtain in practice,energy efficiency(EE) for distributed antenna systems(DAS) based on imperfect CSI and antennas selection is investigated in Rayleigh fading channel.A novel EE that is defined as the average transmission rate divided by the total consumed power is introduced.In accordance with this definition,an adaptive power allocation(PA) scheme for DAS is proposed to maximize the EE under the maximum transmit power constraint.The solution of PA in the constrained EE optimization does exist and is unique.A practical iterative algorithm with Newton method is presented to obtain the solution of PA.The proposed scheme includes the one under perfect CSI as a special case,and it only needs large scale and statistical information.As a result,the scheme has low overhead and good robustness.The theoretical EE is also derived for performance evaluation,and simulation result shows the validity of the theoretical analysis.Moreover,EE can be enhanced by decreasing the estimation error and/or path loss exponents.

Research on Transfer Learning of Vision-based Gesture Recognition

Gesture recognition has been widely used for human-robot interaction.At present,a problem in gesture recognition is that the researchers did not use the learned knowledge in existing domains to discover and recognize gestures in new domains.For each new domain,it is required to collect and annotate a large amount of data,and the training of the algorithm does not benefit from prior knowledge,leading to redundant calculation workload and excessive time investment.To address this problem,the paper proposes a method that could transfer gesture data in different domains.We use a red-green-blue(RGB)Camera to collect images of the gestures,and use Leap Motion to collect the coordinates of 21 joint points of the human hand.Then,we extract a set of novel feature descriptors from two different distributions of data for the study of transfer learning.This paper compares the effects of three classification algorithms,i.e.,support vector machine(SVM),broad learning system(BLS)and deep learning(DL).We also compare learning performances with and without using the joint distribution adaptation(JDA)algorithm.The experimental results show that the proposed method could effectively solve the transfer problem between RGB Camera and Leap Motion.In addition,we found that when using DL to classify the data,excessive training on the source domain may reduce the accuracy of recognition in the target domain.

Distributed coordinated adaptive tracking in networked redundant robotic systems with a dynamic leader

Distributed coordinated control of networked robotic systems formulated by Lagrange dynamics has recently been a subject of considerable interest within science and technology communities due to its broad engineering applications involving complex and integrated production processes,where high flexibility,manipulability,and maneuverability are desirable characteristics.In this paper,we investigate the distributed coordinated adaptive tracking problem of networked redundant robotic systems with a dynamic leader.We provide an analysis procedure for the controlled synchronization of such systems with uncertain dynamics.We also find that the proposed control strategy does not require computing positional inverse kinematics and does not impose any restriction on the self-motion of the manipulators;therefore,the extra degrees of freedom are applicable for other sophisticated subtasks.Compared with some existing work,a distinctive feature of the designed distributed control algorithm is that only a subset of followers needs to access the position information of the dynamic leader in the task space,where the underlying directed graph has a spanning tree.Subsequently,we present a simulation example to verify the effectiveness of the proposed algorithms.

Recent advances in distributed adaptive consensus control of uncertain nonlinear multi-agent systems

So far,distributed adaptive consensus problems for uncertain nonlinear multi-agent systems have not yet been extensively studied.Compared with centralised adaptive control,some new challenges need to be well addressed,for examples,(i)how to reach asymptotically consensus tracking with directed topology condition,by using totally distributed adaptive control strategies;(ii)how to ensure globally uniform boundedness of closed-loop systems while achieving leaderless consensus with semi-positive definite Laplacian matrix;(iii)how to maintain system performance while effectively reducing the communication burden among connected agents.This paper is mainly devoted to report some recent advances in distributed adaptive consensus control.Besides,some interesting research topics which are worthy of further investigation will also be discussed.

Distributed event-triggered adaptive control for second-order nonlinear uncertain multi-agent systems

In this paper,the event-triggered consensus control problem for nonlinear uncertain multi-agent systems subject to unknown parameters and external disturbances is considered.The dynamics of subsystems are second-order with similar structures,and the nodes are connected by undirected graphs.The event-triggered mechanisms are not only utilized in the transmission of information from the controllers to the actuators,and from the sensors to the controllers within each agent,but also in the communication between agents.Based on the adaptive backstepping method,extra estimators are introduced to handle the unknown parameters,and the measurement errors that occur during the event-triggered communication are well handled by designing compensating terms for the control signals.The presented distributed event-triggered adaptive control laws can guarantee the boundness of the consensus tracking errors and the Zeno behavior is avoided.Meanwhile,the update frequency of the controllers and the load of communication burden are vastly reduced.The obtained control protocol is further applied to a multi-input multi-output second-order nonlinear multi-agent system,and the simulation results show the effectiveness and advantages of our proposed method.

Adaptive distributed observer design for containment control of heterogeneous discrete-time swarm systems

This paper develops both adaptive distributed dynamic state feedback control law and adaptive distributed measurement output feedback control law for heterogeneous discrete-time swarm systems with multiple leaders.The convex hull formed by the leaders and the system matrix of leaders is estimated via an adaptive distributed containment observer.Such estimations will feed the followers so that every follower can update the system matrix of the corresponding adaptive distributed containment observer and the system state of their neighbors.The followers cooperate with each other to achieve leader-follower consensus and thus solve the containment control problem over the network.Numerical results demonstrate the effectiveness and computational feasibility of the proposed control laws.

Characteristic model-based consensus of networked heterogeneous robotic manipulators with dynamic uncertainties

In this paper, we address the characteristic model-based discrete-time consensus problem of networked robotic manipulators with dynamic uncertainties. The research objective is to achieve joint-position consensus of multiple robotic agents interconnected on directed graphs containing a spanning tree. A novel characteristic model-based distributed adaptive control scenario is proposed with a state-relied projection estimation law and a characteristic model-based distributed controller. The performance analysis is also unfolded where the uniform ultimate boundedness(UUB) of consensus errors is derived by resorting to the discrete-time-domain stability analysis tool and the graph theory. Finally, numerical simulations illustrate the effectiveness of the proposed theoretical strategy.

Model Reference Adaptive Control for Networked Distributed Systems with Strong Interconnections and Communication Delays

In recent years, networked distributed control systems（NDCS） have received research attention. Two of the main challenges that such systems face are possible delays in the communication network and the effect of strong interconnections between agents. This paper considers an NDCS that has delays in the communication network, as well as strong interconnections between its agents. The control objective is to make each agent track efficiently a reference model by attenuating the effect of strong interconnections via feedback based on the delayed information. First, the authors assume that each agent knows its own dynamics, as well as the interconnection parameters, but receives information about the states of its neighbors with some communication delay. The authors propose a distributed control scheme and prove that if the interconnections can be weakened and if the communication delays are small enough, then the proposed scheme guarantees that the tracking error of each agent is bounded with a bound that depends on the size of the weakened interconnections and delays, and reduces to zero as these uncertainties reduce to zero. The authors then consider a more realistic situation where the interconnections between agents are unknown despite the cooperation and sharing of state information. For this case the authors propose a distributed adaptive control scheme and prove that the proposed scheme guarantees that the tracking errors are bounded and small in the mean square sense with respect to the size of the weakened interconnections and delays, provided the weakened interconnections and time delays are small enough. The authors then consider the case that each agent knows neither its dynamics nor the interconnection matrices. For this case the authors propose a distributed adaptive control scheme and prove that the proposed scheme guarantees that the tracking errors are bounded and small in the mean square sense provided the weakened interconnections and time delays are small enough. Finally, the authors present an illustrative example to present the applicability and effectiveness of the proposed schemes.

COMPARING TWO CLUSTERING-BASED TECHNIQUES TO TRACK MOBILE NODES IN INDUSTRIAL WIRELESS SENSOR NETWORKS

Tracking mobile nodes in dynamic and noisy conditions of industrial environments has provided a paradigm for many issues inherent in the area of distributed control systems in general and wireless sensor networks in particular. Due to the dynamic nature of the industrial environments, a practical tracking system is required that is adaptable to the changes in the environment. More specifically, given the limited resources of wireless nodes and the challenges created by harsh industrial environments there is a need for a technique that can modify the configuration of the system on the fly as new wireless nodes are added to the network and obsolete ones are removed. To address these issues, two cluster-based tracking systems, one static and the other dynamic, are proposed to organize the overall network field into a set of tracking zones, each composed of a sink node and a set of corresponding anchor nodes. To manage the wireless nodes activities and inter and intra cluster communications, an agent-based technique is employed. To compare the architectures, we report on a set of experiments performed in JADE （Java Agent Development Environment）. In these experiments, we compare two agent-based approaches （dynamic and static） for managing clusters of wireless sensor nodes in a distributed tracking system. The experimental results corroborate the efficiency of the static clusters versus the robustness and effectiveness of the dynamic clusters.

Adaptive distributed observer and the cooperative control of multi-agent systems

The distributed observer approach has been an effective way to synthesise a distributed control law for a multi-agent system.However,this approach assumes that all the followers know the system matrix of the leader system,and this assumption may not be desirable in some applications.In this note,we will further introduce an adaptive distributed observer,which is able to estimate both the state and the system matrix of the leader and thus does not require that all the followers know the system matrix of the leader system.We will also point out some applications of the adaptive distributed observer to several cooperative control problems of multiagent systems such as the leader-following consensus problem,the cooperative output regulation problem and rendezvous and/or flocking.