Search

In this article, the observability of the Boolean multiplex(multiplevel) networks (BMNs) is studied. First, the dynamical model and structure of BMNs are constructed. By using of Semi-Tensor Product (STP) approach, the logical dynamics of BMNs is converted into an equivalent algebraic representation. Then, the observability of the BMNs is investigated by giving necessary and sufficient conditions. Finally, an example is given to illustrate the efficiency of the obtained theoretical results.

Headbox is the key component in a paper machine which has a significant effect on product quality. In this paper, a hybrid model which combines mixing tank and air-cushioned headbox is proposed. In the hybrid model, the mixing tank is taken into consideration, which makes the consistency of pulp controllable. An internal model controller with both good control effect and strong robustness is designed to control the proposed model. With the generalized decoupled object, computational work is avoided, and the designed controller has low order. Numerical simulations are performed to illustrate the effectiveness of this methodology.

In papermaking process, headbox is a key component of the paper machine, whose output has big influences on the produced paper quality and operation safety. In this paper, the double air-cushioned headboxes structure is modelled in a cascaded way, in which we take the liquid level and air-cushioned pressure into consideration. To achieve stable and fast-responded control performance for the proposed cascaded model, a Kalman Filter (KF)-based model predictive control (MPC) method is designed. The KF is utilized to estimate the state of the system under disturbance near the headbox operating points. Optimal state estimation results obtained from the KF are then applied into the MPC algorithm to predict future liquid level and air-cushioned pressure of the headbox. Simulation results show that the KF-based MPC control strategy achieves satisfactory performance on suppressing system disturbance and reducing output overshoot.

The problem of the stability analysis of genetic regulatory networks (GRNs) with time varying delays are proposed in this paper based on a new delay-decomposing method. Some appropriate Lyapunov-krasovskiis functionals are defined on each of the obtained subintervals, some integral inequalities and the convex combination technique are utilized to deal with the bound of the derivative of the Lyapunov-krasovskiis functional. Then the criteria are derived in terms of linear matrix inequalities (LMIs). Finally, a numerical example is used to illustrate the usefulness and less conservative than relative references.

Modeling and control methods for systems with dynamic hysteresis are proposed in this paper based on a type of Hammerstein models. Experiments are conducted on a typical dynamic hysteresis system, i.e., a giant magnetostrictive actuator. The modeling, real-time tracking, and real-time active vibration control results are presented. These results verify the effectiveness and accuracy of the modeling and control methods based on the Hammerstein model.

This paper gives a Hinf-LPV(Linear-Parameter-Variable) yaw control method to improve vehicle yaw response performance. The controller works in the linear area of the tyre force mainly to improve vehicle response to the steering wheel angle. By means of writing the Dugoff tyre model into two terms (linear term and nonlinear term), Lipschitz condition is used to approximating nonlinear term with linear expression. And the tyre lateral stiffness slight variation caused by tyre pressure is taken as model uncertainty. In order to suppress the influence of model uncertainty and disturbance, a Hinf-LPV controller is designed. At last, simulations are done to check the controller performance. Simulation results show that controller designed in this paper can improve vehicle yaw response performance at both high friction coefficient and medium friction coefficient road.

Object tracking is a fundamental skill of mobile robotics. For a multi-robot system the object tracking accuracy of each robot is usually different because of the dynamic environment and the different sensing ability. The focus of cooperative object tracking is how to fuse different robots' observation to obtain more robust and accurate estimation of the object. This paper investigates this problem using an adaptive consensus based distributed particle filter algorithm. In this algorithm, each robot runs a particle filter to achieve local object tracking. Then a consensus algorithm which can establish an agreement between all robots is utilized to generalize a global posterior about the object state. To fully consider the difference between each robot's object tracking accuracy, Renyi entropy is used to adaptively adjust the weight of the consensus algorithm. The concept of Renyi entropy is utilized to measure the distance between a robot's local posterior and the global posterior and then a weight will be assigned to the robot according to the Renyi entropy. This weight will be used continuously in the next consensus step. By the proposed algorithm in this paper the accuracy and the robust of the cooperative can be improved. Finally the effectiveness of the proposed algorithm is verified by simulation results.

This paper investigates two finite-time attitude tracking controllers for spacecraft with attitude dynamics presented by rotation matrix. With the use of rotation matrix, unwinding problem widely exists in controllers based on quaternion can be dealt with. Based on backstepping method, the first controller can be developed in the presence of external disturbance. However, the upper bound of the disturbance must be exactly known, which makes it inconvenient to implement. The second controller is constructed employing the property of hyperbolic tangent function, with which the virtual command is derived so that a finite-time convergence can be obtained. Furthermore, an adaptive law is utilized to estimate the upper bound of the disturbance online. Mathematical demonstration and simulation results validate the effectiveness of the controllers developed in this paper.

Cloud Control System is proposed based on Cloud Computing, Networked Control System and Internet of Things. It inherits the “3C” (Computing, Control and Communication) from Cyber-physical System and provides powerful ability of data storage and processing to serve control system. It is getting popularity day by day in the modern CS architecture, because of its ground breaking future applications. From the perspective of engineering application, this paper makes an investigation on engineering problems in initial phase of Cloud Control System, addressing the issues regarding communication protocol, gateway and time delay of network, based on some simple experiments. An experiment based on ball-beam and Tencent Cloud is presented with a simple compensator used to compensate the delay of network. It shows that Cloud Control System is feasible. And work in this paper will lay a good foundation for further research on Cloud Control System.

This paper presents a new Lyapunov theorem on almost surely finite-time stability for stochastic nonlinear systems with Markovian switching. Unlike the work in [3] that consider finite-time stability and stabilisation of conventional stochastic differential equation (SDE) systems, this paper aims to propose a weaker finite-time stability theory for a more general class of SDE systems with Markovian switching. A lemma is presented to discuss conditions that ensure the existence of a unique strong solution for such SDE systems with Markovian switching. Extended Comparison Principle and Bihari's inequality are derived, which relaxes some previous conditions and play an important role in the proof of the new Lyapunov theorem. Weaker conditions are proposed to ensure finite-time stability in probability one with supportive examples. Two simulation examples are given to illustrate the theoretical analysis.

This paper investigates the problem of formation control for multi-agent systems with general linear dynamics and input saturation. The time-varying formation tracking this paper deal with is required to be piecewise differentiable. Firstly, a time-varying formation control protocol is presented. Then an algorithm consisting of 3 steps to design the formation control protocol is proposed. As follows, it is proven that under the designed formation control protocol, the multi-agent systems with input saturation can achieve the desired time-varying formation tracking which belongs to the feasible formation set. Finally, numerical simulations are presented to verify the effectiveness of the proposed algorithms.

The two-layer restricted Boltzmann machine (RBM) for rating prediction problems in recommendation systems has become one of the significant researches. A dual conditional restricted Boltzmann machine (dCRBM) model is proposed in this paper. In the dCRBM model, the training process uses rated/unrated information. Meanwhile, we also utilize the dual patterns of users and items, and build two CRBMs based on users and items respectively. Moreover, rating predictions of these two CRBMs via regression models are combined. The experimental results from MovieLens dataset show that the dCRBM model proposed in this paper helps to improve the prediction accuracy of the recommendation system.

In this paper, unrelated parallel machine scheduling problem with job rejection and earliness-tardiness penalties is investigated. The objective is to minimize the total penalty cost by deciding job acceptance, assigning jobs on unrelated machines, and determining the processing sequence of jobs on each machine. To solve this problem, a mixed integer programming (MIP) model is established, and a hybrid genetic algorithm which hybridizes Genetic Algorithm (GA) and tabu search (TS) with a concise encoding method, special genetic operators is proposed. Computational experiments are performed on three different sized sets of test instances which are randomly generated, and the results of comparative experiment demonstrate that the algorithm proposed in this paper can effectively solve the problem.

This paper designs a vibration controller based on back propagation (BP) neural network for jacket-type offshore platforms with time delay. With an exosystem designed, we can describe the irregular wave forces in the context of Morison equation and wave theory. The original vibration control can be seen as an optimal control problem. This linear discrete-time system is affected by wave forces as external disturbances. A BP neural network controller is designed based on a feedforward and feedback optimal control law with memory (FFOCLM). Neural network has been widely applied in controlling with strong capability of approximation and adaption. The tuning rule for updating weights and thresholds of the network is derived to approximate the dynamics of the control system. Data of the FFOCLM are sampled to train the BP neural network. The simulation results prove the feasibility and efficiency of the BP neural network controller designed in paper for attenuating vibrations of offshore platforms.

In this paper, it is discussed the modeling problem of a rotor/fixed-wing hybrid unmanned aerial vehicle (UAV). The vehicle has two main advantage functions, i.e., fixed-wing cruise and vertical take-off and landing (VTOL). This paper focuses on analysis of the transition process for the rotor/fixed-wing hybrid UAV, which consists of four stages. Based on the six degree of freedom model, the modeling algorithm of the rotor/fixed-wing hybrid UAV is proposed.

This paper considers a nonlinear attitude estimator to estimate the attitude and gyro-bias exponentially for rigid body whose kinematics is evolved on the Special orthogonal group SO (3), which can describe the attitude globally and uniquely. Compared to the asymptotic stabilities which are presented in the previous literature, the estimator presented in this paper has exponential convergence. The exponential stability analysis is demonstrated by singular perturbation method which divide the closed loop system into fast system and slow system. The results can be apply to the vehicles which are equipped with the low-cost inertial measurement unit (IMU) such as gyroscope. Finally, the theoretical results are verified through a numerical simulation.

This paper provides a simple approach for group consensus of networked gyroscopes systems from the analytical dynamics point of view. By using the Udwadia-Kalaba fundamental equation theory, a novel robust group control law is presented and a sufficient condition which ensure achieving group consensus of non-identical gyroscopes networks is given in this paper. Finally, numerical simulations conducted on networks composed by five non-identical gyroscopes is given to demonstrate effectiveness of the proposed control methodology.

Traditional convergence theory of self-tuning regulators requires boundedness of the conditional variances of the systems noise processes. However, this requirement cannot be satisfied for many practical models such the well-known ARCH (Autoregressive Conditional Heteroscedasticity) model in economic systems. The aim of this paper is to provide a convergence theory of self-tuning regulators for linear uncertain systems with conditional heteroscedastic noises, where the conditional variance is unbounded. To be specific, we consider weighted least-squares-based self-tuning regulators, and establish both the global stability and the optimality of tracking under some natural conditions on the system models as well as on the conditional heteroscedastic noises. To the best of the authors' knowledge, this is the first paper that investigates this kind of problems with a convergence theory, and makes the self-tuning regulators applicable to systems with noised modeled by ARCH.

This paper is devoted to the semiglobal finite-time stabilization via output-feedback for a class of high-order uncertain nonlinear systems. This problem can hardly be solved by the existing schemes, mainly since the systems under investigation contain unknown control coefficients which are dominated by the known functions of system states, and possess more serious nonlinearities which allow arbitrary-high-order growing unmeasurable states. However, for the global case, the system unknown control coefficients should be dominated by the known functions of output and the system nonlinearities merely allow homogeneous-order growing unmeasurable states, and therefore our semiglobal objective significantly weakens the restrictions on the control coefficients and the nonlinearities of the systems. In this paper, with the assumption that all the system states are available for feedback, a state-feedback controller is first constructed by adding a power integrator, which ensures that the closed-loop system is globally finite-time stable. Then, by certainty equivalence principle and saturation technique, an output-feedback controller is successfully constructed, which makes the closed-loop system semiglobal attractive and local finite-time stable. A simulation example is presented to illustrate the effectiveness of the proposed method.

This paper proposes a robotic grasping system for automatically sorting garbage based on machine vision. This system achieves the identification and positioning of target objects in complex background before using manipulator to automatically grab the sorting objects. The object identification in complex background is the key problem that machine vision algorithm is trying to solve. This paper uses the deep learning method to achieve the authenticity identification of target object in complex background. In order to achieve the accurate grabbing of target object, we apply the Region Proposal Generation (RPN) and the VGG-16 model for object recognition and pose estimation. The machine vision system sends the information of the geometric centre coordinates and the angle of the long side of the target object to the manipulator which completes the classification and grabbing of the target object. The results of sorting experiment of the bottles in the garbage show that the vision algorithm and the manipulator control method of the proposed system can achieve the garbage sorting efficiently.

The common PID controller with fixed parameters is difficult to control the speed of autonomous vehicle under changing operation conditions, thus a indirect single neuron adaptive PID controller based on quadratic criterion is put forward in this paper. The output error and control increment are assigned to be performance indexs in the learning algorithm. In order to further restrain the control increment and reduce the influence of measurement error, the controller proposed in this paper is changed into indirect structure. A method, which is based on identification model of vehicle and simulation experiment, to obtain the initial parameters of the controller, is presented in this paper. Furthermore, this paper uses a miniature autonomous vehicle to test the performance of the controller.

For air-breathing hypersonic vehicle, there exits significant aerodynamic/propulsion/structure couplings due to the integrated design, which requires detailed performance limitation analysis during the concept design phase. Aerodynamic derivatives are important parameters for hypersonic vehicle while open-loop performance affects the achievable closed-loop performance, there is mutual relationship between aerodynamic derivatives and the open-loop performance, the changes of aerodynamic derivatives under different flight conditions will directly affect the performance of the open loop. In this paper, the waverider model like X-43A is established and linearized, then aerodynamic derivatives will extract from the model state and control matrices, nine aerodynamic derivatives are considered in this paper. In addition, three kinds of open-loop performance are introduced including the deflection of elevator, the lag of path-attitude response, and the right half plane zero bandwidth constraints, which are obtained from aerodynamic derivatives. Finally, by analyzing them under different flight conditions, the best flight condition range is obtained.

This paper is concerned with the problem of distributed state estimation for linear discrete system with packet loss. A state prediction consensus mechanism is introduced to promote the estimation performance. By applying matrix theory, a sufficient condition for the convergence of the estimation error system is derived. Then, a numerical example is given to show the algorithm proposed in this paper has better consensus performance compared with a classical Kalman consensus filtering (KCF) by some simulations in real and packet-dropping cases, respectively. Finally, the influence of consensus coefficient and prediction consensus coefficient on the estimation error covariance are presented by simulation.

There are some problems such as low utilization rate, high energy consumption, low productivity and poor quality of cutting surface in the traditional metal bar cutting technology. Aimed at this situation, this paper analyzes that the low-frequency load causes the crack initiation at the tip of the v-groove of the bar, the high-frequency load causes the crack propagation at the tip of the V type groove of the bar, and give the criteria for the initiation of the ν type groove at the tip of the bar. a new precision dual frequency vibration cutting technology is adopted, and a dual frequency excitation device is proposed. On this basis, this paper carries out a dynamic analysis of the experimental device, which can fully guarantee the separation of metal bar in theory. Single frequency and dual frequency load experiments were carried out on 7A09 aviation aluminum, the experimental results show that the dual frequency vibration can improve the quality of the section and shorten the fracture time. Therefore, It gives support to the cutting technology for metal round bar under the dual frequency vibration load both in theory and the experiment.

This paper gives some Razumikhin-type theorems on pth moment boundedness of impulsive stochastic functional differential equations (ISFDEs) by using Razumikhin tenique and comparison principle. Some improved conditions on pth moment stability are also proposed. The main results of this paper allow the estimated upper bound of the diffusion operator associated with the underlying (ISFDEs) of the Lyapunov function to have time-varying coefficients(the coefficients may even be sign-changing functions). Examples are provided to illustrate the effectiveness of the proposed results.

A secrecy transmission method with robust power control is investigated in this paper for a downlink two-tier femtocell network, where an eavesdropper attempts to wiretap the legitimate macrocell users. Considering the imperfect channel gains, a probability constraint robust optimization problem is formulated to satisfy the quality-of-service (QoS) of users. We aim to maximize the secrecy rate with the constrains of power outage probability and limited power at base stations (BSs). Firstly, the problem of power outage probability is solved by a novel conversion method. Unfortunately, the object function is a non-convex. A simple method based on variable substitution and Taylor expansion is presented to solve this puzzle in this paper. Then, the iterative algorithm is provided to get the practical operation for users. Finally, numerical simulation results are given to illustrate the effectiveness of the proposed algorithm.

In distribution network, with the installed of photovoltaic (PV), bought many problems. One of the most concerned problems is the overvoltage in the terminal bus of distribution network. It is caused by the reverse power flow which brought out by the increase of the PV sources penetration. In this paper, an improved PQ-Q (V) control strategy is proposed to alleviate such problem. The economic performance and the efficiency are considered in this prosed scheme. The photovoltaic grid-connected inverter is controlled to compensate the overvoltage or under voltage by compensating the inductive and capacitive reactive power to ensure voltage are within rated range. Its performance is compared with a traditional control method. On this basis, the influence of the method on the network is investigated, which have been found that the value of tune parameter will affect the regulate ability and the system stability. In this paper, the 13-bus IEEE network is used in the simulation of DIgSILENT software.

In this paper, the exponential stochastic synchronization of coupled neural networks via adaptive periodically intermittent control is further investigated. Based on the Lyapunov stability theory combined with the method of the adaptive control, periodically intermittent control and the properties of Weiner process, some simple criteria are derived for the exponential stochastic synchronization of the coupled neural networks with coupling delays under stochastic perturbations. The adaptive periodically intermittent control which we have obtained can cut down control cost. The sufficient conditions of this paper for network synchronization are less conservative and can be applied in a wider area. The numerical example of the stochastically coupled neural networks is given to demonstrate the effectiveness of the control strategies.

In this paper, the finite-time boundedness and stabilization problems of a class of networked control systems (NCSs) with bounded packet dropout are investigated. The main results provided in the paper are sufficient conditions for finite-time boundedness and stability via state feedback. An iterative approach is proposed to model NCSs with bounded packet dropout as jump linear systems (JLSs). Based on Lyapunov stability theory and JLSs theory, the sufficient conditions for finite-time boundedness and stabilization of the underlying systems are derived via linear matrix inequalities (LMIs) formulation. Moreover, both sensor-to-controller and controller-to-actuator packet dropouts are considered simultaneously. Lastly, an illustrative example is given to demonstrate the effectiveness of the proposed results.

With the development of intelligent device and social media, the data bulk on Internet has grown with high speed. As an important aspect of image processing, object detection has become one of the international popular research fields. In recent years, the powerful ability with feature learning and transfer learning of Convolutional Neural Network (CNN) has received growing interest within the computer vision community, thus making a series of important breakthroughs in object detection. So it is a significant survey that how to apply CNN to object detection for better performance. First the paper introduced the basic concept and architecture of CNN. Secondly the methods that how to solve the existing problems of conventional object detection are surveyed, mainly analyzing the detection algorithm based on region proposal and based on regression. Thirdly it mentioned some means which improve the performance of object detection. Then the paper introduced some public datasets of object detection and the concept of evaluation criterion. Finally, it combed the current research achievements and thoughts of object detection, summarizing the important progress and discussing the future directions.

Dynamics model is the basis of controller design for a robot. This paper presents dynamics modeling method for the arm of humanoid robot Nao. The dynamics model is analyzed based on Euler-Lagrange equations. Firstly, this paper analyzes the upper limb topology of Nao model and establishes the dynamics models. Secondly, Fuzzy-PD controller is used in the dynamics model and compared with traditional PD controller. Finally, the proposed solution is tested in Matlab environment to demonstrate the correctness of the robot dynamics model.

In the study of controllability of complex networks, there are two core issues: the minimal number of inputs and the minimal number of controlled nodes needed to achieve full control. In this paper, we pay attention to find the optimal minimum driver nodes set with more stronger control ability by using the methods arising from graph theory and matrix theory. Firstly, we determine the minimum number of driver nodes; Secondly, the optimal minimum driver nodes set is determined by searching the optimal connection added between new control signal u+(t) and state node i with more stronger control ability. Finally, the input signals are added to the optimal minimum driver nodes to complete the controllability of the network. In this paper, the algorithm of the optimal minimum driver nodes set is also designed for the controllability of complex networks.

The PID control technique which is improved by linear extended state observer (LESO) is applied to the design of the flight control system for parafoils with Global Position System(GPS). Firstly, in this paper, the nonlinear dynamics model of the parafoil system is outlined. Then, method of guidance and the theory of LESO is presented in detail. Next, the control system using improved PID control technique is introduced integrally. The results of simulation and flight test are presented and analyzed at the end of the paper. The results of the improved controller show the great superiority to the traditional PID controller. In the flight test, the error of the trajectory tracking is less than twenty meters and the control strategy has great application value on the precision aerial delivery system.

In this paper, a modeling and simulation approach for the boiler coil pipe bending production line is proposed based on the eM-plant software. The aim of this paper is to analyze the real workshop of manufacturing resource and process data based on the object-oriented simulation modeling technology and give the arrange strategy for improving the continuity and reliability of the manufacturing process. Firstly, the mathematical models of production processing and welding combination processes are studied, including the production line configuration, working time for each product station set theory and the production line buffer size. Secondly, the maximum processing capacity and each station working load are analyzed according to the working state of the workshop and bottleneck analysis. Finally, a simulation model is established to determine optimal control strategy for the material management and storage, human configuration optimization and production scheduling. Simulation results show that the designed boiler coil pipe bending production line can effectively improve the production efficiency.

In order to improve the reliability and viability, advanced aircraft is equipped with abundant multiple control surfaces. Control allocation is utilized to assign the virtual control torque to these redundant control surfaces. Due to physical and aerodynamic factors, there are some constraints impacting on each control surface, which makes the control allocation problem become more complex. In this paper, a weighted pseudo-inverse based control allocation algorithm is presented. Although the weighted pseudo-inverse method is simple and efficient, this control allocation algorithm cannot reach the optimal allocation achievement. For obtaining the maximum attainable moment set, genetic algorithm is employed to train the weighted matrices in this paper. In order to improve performances of the method, the space of the expected moment is divided into multiple parts, and the genetic algorithm is used to find the optimal weighted matrices for each part. Compared with a single weighted matrix, multiple weighted matrices can ensure that the proposed algorithm performs better in each part. In order to verify the validity of this algorithm, the direct allocation method is employed as a comparison. Simulations demonstrate and verify that performances of the new method is better than those of the conventional weighted pseudo-inverse method.

When the Buck converter is under the chaotic state, its working performance deteriorates. This paper proposes a novel nonlinear controller in the form of the cubic difference to eliminate the chaotic behavior in the Buck converter. This nonlinear controller is designed on basis of an idea that the cubic difference between the output voltage and the input voltage is multiplied by a feedback coefficient. Based on the state equations and the discrete mapping equations of the original Buck converter, it is confirmed that the chaotic phenomenon exists in the Buck converter. In terms of the controlled Buck converter model which can be obtained by introducing the nonlinear controller into the original Buck converter, the effectiveness of this controller is verified via analyzing the bifurcation diagrams with the feedback coefficient increasing, the inductor current waveform, the output voltage waveform, the Poincare section, the phase portrait and so on. Results show that without knowing the targeting orbit beforehand, this controller can transform the chaotic state of the Buck converter to the fixed point and the various periodic orbits by only adjusting one external parameter, that is, the feedback coefficient. The results of this paper provide a new thought for controlling chaotic phenomena in the power electronic circuits.

This paper is concerned with the optimal Markov jump linear filter problem for Markov jumping parameter system with multi-channel observation delays and packet losses, where the jumping parameters are assumed to be known up to the present time. Firstly, by defining an equivalent delay-free observation sequence, the optimal jumping filter problem is transformed into the one of delay-free system which is just with jumping parameters and multi-channel multiplicative noises. Then an optimal Markov jump linear filter is presented based on the mean squared method, in which the filter gains are determined by solving a set of generalized coupled Riccati equations based on a set of coupled Lyapunov equations. And the Riccati equations developed in this paper are not only relevant to the transition probability of the jumping parameters but also concerned with the probability of the packet losses. This is the main difference btween the result proposed in this paper and our previous works.

In this paper, we consider a class of discrete-time multi-input inhomogeneous bilinear systems. Controllability and near-controllability of such systems have been well studied in our previous work, where controllable subspaces and nearly-controllable subspaces have also been derived. We continue the study on the controllable subspaces and nearly-controllable subspaces of the systems in this paper and make some remarks. New controllable subspaces and nearly-controllable subspaces are thus derived and analyzed.

With rapid development of highway transportation and increasing number of vehicles, the incidence rate of road traffic accidents also sets record. Among them, fatigue driving has become the main reason of traffic accidents. In this paper, we utilize OpenCV (Open Source Computer Vision) to implement face detection as well as eyes detection on Android platform and to compare the frequency of eyes closed. A fatigue detection system which is able to determine the examinee's fatigue state accurately was established. Firstly, this paper improve the traditional face detection algorithm (Adaboost algorithm) and increase the efficiency of face detection. Secondly, we use the template matching method to track the eye (iris) in real time, constantly updating the iris template to improve the accuracy of matching. Finally, the implementation of system is described in detail under the Android platform. Experiments show that the system can make full use of real time nature and robustness of AdaBoost face detection algorithm, to achieve high detection rate.

In this paper, the application of independent component analysis (ICA) to statistical process monitoring is studied. This paper mainly focuses on studying on the fault detection and isolation principle based on the data model of ICA. Contributions of this paper are: (1) for the purpose of fault detection, two monitoring statistics are designated by detailed analysis on the data model of ICA; (2) a semi-supervised Laplacian regularization (SLR) kernel density estimation approach is proposed to determine the normal operation region; (3) Under fault condition, a fault isolation method based on expert knowledge is introduced; (4) Hot Galvanizing Pickling Waste Liquor Treatment Process (HGPWLTP) is taken to evaluate the monitoring performance of the proposed approaches. Encouraging experimental results of the proposed process monitoring approaches are achieved.

In this paper, a robust position/force set-point control strategy based on a sliding-mode is designed for constrained underactuated manipulator. At first, the reduced underactuated position/force dynamic model is obtained by considering a coordination transformation. Then, a controller based on a sliding-mode is designed to drive the closed-loop system states to reach the adopted sliding model surface. Considering the parametric uncertainties of the manipulator, a robust control strategy is applied to bound compensation. Furthermore, we can obtain a global asymptotic stability result for the positions error in the Lyapunov sense and the error of constraint force is bounded with a controllable bound. Finally, the simulation results of a three-DOF manipulator system illustrate the validity and the feasibility of the proposed method in this paper.

In the use of statistical analysis methods to do fault diagnosis, influenced by the designed parameters, the information may be hidden. As a result, we could not extract the whole information from the data. In order to reduce the loss of information, it is necessary for us to use the deep structure to describe the structure of the data. In this paper, the method of deep learning is used to deal with the vibration signal of rolling bearing. At the same time, an unsupervised multiple layer stacked sparse auto encoder network is proposed and used to learn the complex structure of the vibration signal because it has a good learning ability. In this paper, the vibration signal processing is processed in two aspects: the time domain and the frequency domain. Deep learning algorithm can learn the hidden information from the data, then the classifier could get very good results. Experimental results show that the algorithm proposed in this paper performs well for the bearing vibration signal of the Case Western Reserve University. Especially in frequency domain, the results is almost perfect.

Combining with the characteristics of smart distribution grid, this paper expounded the architecture, control strategy and key technology of self-healing control in smart distribution grid and proposed the “three-layer with seven-unit” architecture and the option method of control strategy. Combining with the characteristics of smart distribution grid containing distributed power, this paper introduced the key technology of self-healing control, so as to make up the defects of current self-healing control that the smart grid features were not fully considered at present, which provided the theoretical basis for the construction of self-healing control.

The problem of observer design for a class of linear time-varying system is researched in this paper. An augmented linear time-varying system is proposed to extent the application of the previous results. Moreover, using the observable block companion form of the augmented linear time-varying system, this paper presents an algorithm of observer design. Compared with other observer design approaches, the prominent feature of the algorithm is that of simplifying the computational complexity. Finally, simulations are designed in linear time-varying system and results are presented to demonstrate validity of the proposed algorithm.

This paper studies the problem of determining sensor locations in a large sensor network using only relative distance (range) measurement. In this paper, we consider a synchronous communication protocol when 1) the communication delay between sensors is uncertain; 2) successive packet loss is finite. Based on the barycentric coordinate representation, our work generalizes the DILOC algorithm in the communication environment with time delays and packet loss. First, a synchronous communication protocol is developed, and the distributed localization algorithm is mathematically modeled as a linear difference equation with time-varying communication delays. Next, the time-varying-delay model is proved to be globally convergent to the true coordinates under consensus framework with quotient graphic theory. Finally, simulation examples are provided to demonstrate the effectiveness of the theoretical result.

Source seeking in emergency environment is a vital issue, which has attracted the attention of researchers recently. In this paper, we consider a robot traveling with a constant speed and guided by sensor measurements. This control law does not depend on any gradient dependent information. The robot is navigated using an angular velocity limited by a maximum and minimum constants. The robot location is calculated using the sensor neighborhood topology coordinates, which is an alternative to the physical coordinate system. In topology coordinate system, actual physical distances are hidden. Thus, the proposed control law does not incorporate any distance information. Moreover, the performance of the proposed method is evaluated using computer simulation and presented in this paper.

This paper investigates the list coloring problem, and presents a number of new results and algorithms. By using the matrix semi-tensor product, the list coloring problem is studied, and three necessary and sufficient conditions are proposed, based on which an algorithm is designed to find all the list coloring schemes for any simple graph. In addition, the effectiveness of the results/algorithms presented in this paper is shown by one illustrative example.

A class of linearly parameterized multiagent systems with unknown parameters and unknown control gain are investigated in this paper. An adaptive algorithm is implemented for such multiagent systems to obtain consensus. In most previous cases, consensus approaches usually induce uniformly ultimately bounded consensus error because of the uncertainties existing in the system dynamics. By contrast, this paper introduces a novel robust consensus algorithm which can ensure that the consensus error converges to zero asymptotically. More specifically, an adaptive controller will be designed for on-line unknown parameter identification, and a robust continuous term will be applied to ease the effects of the external disturbances. Besides, the control signal is insured continuously to neglect actuator bandwidth requirement and avoid the caused chattering phenomenon. Noting that, for multiagent systems studied in this paper, each agent only needs to exchange information with its neighbor agents, i.e., the proposed consensus algorithm is distributed. The theoretical verification of asymptotic consensus result is given through Lyapunov synthesis and several simulation tests on a linearly parameterized multiagent system are conducted to demonstrate the performance of the algorithm.

A stratospheric airship is difficult to achieve long duration flight at about 20km altitude at mid or low latitudes, where it endures pressure changes due to the heating and cooling of the day-night cycle. In order to overcome these difficulties, this paper proposed a novel stratospheric airship based on triple-gasbag principle, and investigated its static and dynamic behavior. Firstly, this paper briefly described the concept design of a triple-gasbag airship. Then, to investigate the dynamic flight performance of this airship, the paper built a vertical flight dynamics model coupled with distributed parameter thermal model and gas thermodynamic model. Moreover, the distributed thermal model was verified by comparing the simulation results with the flight experimental data. The comparison results show that the deviation between simulation and experimental data is less than 2.4%. Finally, to verify the effect of triple-gasbag principle on reducing differential pressure, this paper investigated the transient behavior of the thermodynamic and kinetic parameters of the triple-gasbag airship at floating conditions using the ammonia phase change combined with air exchange method. The simulation results indicate that if the volume of ammonia gasbag is 10% of the overall volume of the airship, the triple-gasbag method can reduce the differential pressure over 45%. In summary, the triple-gasbag principle is effective in regulating the pressure and altitude respectively, and can greatly reduce super pressure oscillation of the airship during diurnal cycle.

For distributed generation (DG) network, it is important to estimate the real-time states. The information-centric networking (ICN) is established to take charge of the communication of DG network. However, the assumption of ideal communication between sensors and the estimation center cannot be guaranteed due to the communication constraints of ICN with the increasing DG network. A conventional algorithm, which reduces the communication burden in ICN, is to drop the observation of each smart grid in a random way. However, the accuracy of this algorithm recession decays rapidly with the increasing drop rate. To guarantee an appropriate estimation accuracy when the drop rate increases, this paper introduces the event-trigger strategy into the estimation algorithm. An event-trigger extended Kalman filter (ET-EKF) is established in this paper to adapt the nonlinearity of DG system. ET-EKF reduces the communication burden and achieves an appropriate estimation accuracy at the same time. Finally, its feasibility and performance are demonstrated using the standard IEEE 39-bus system with phasor measurement units (PMUs).