Search

Being a widely used technology in papermaking industry, the fault diagnosis of the sequence batch reactor (SBR) wastewater treatment process has been a significant challenge owing to the batch characteristics. In order to decrease the complexity of monitoring the SBR process, the inherent multi–period characteristics has been considered in this study. The conventional multi– way principal component analysis (MPCA) method has been improved with sub–period division strategies (Sub–MPCA) to diagnose the faults in the SBR process. Aiming at identify the most applicative sub–period division strategy for the SBR process, four types of strategies (Scenarios 1–4) have been tested and compared. Beginning with the off–line modeling, the training set data was used to motivate the sub–MPCA models and acquire the control limits of T2 and SPE statistics. Subsequently, the fault alarm rates (FARs) of the developed models were estimated to verify the models. Finally, the fault diagnosis performances of the models were evaluated with the testing data set from an abnormal batch. After the examinations of the four sub–period division strategies in SBR process, the result revealed that a multi–sub–period algorithm based on the similarities of the loading matrices between the adjacent time slices (Scenario 3) demonstrated the best performance with fewer diagnostic error, which was identified the most accurate model for the fault diagnosis in the SBR wastewater treatment process.

In the process of urban sewage treating, reducing the energy consumption and improving the quality of the effluent are significantly meaningful. According to the activated sludge method, the key factors affecting the energy consumption and water quality of wastewater treatment are determined. In order to minimize the energy consumption of the activated sludge process and maximize the quality of the effluent, four different objective functions are modeled [i.e., the airflow rate, the carbonaceous biochemical oxygen demand (CBOD) of the effluent, the total phosphorus (TP) of the effluent, and the ammonia nitrogen of the effluent (NH4-N)]. These models are developed using a back propagation (BP) neural network based on industrial data, and dissolved oxygen (DO) is the controlled variable. A multi-objective model was evaluated by six evaluation indicators. Based on the analysis of the model and the mechanism of activated sludge process, the multi-objective particle swarm optimization(MOPSO) algorithm was used to optimize the energy consumption and water quality of the activated sludge process. The experimental results show that eventually reduce aeration energy consumption by 17%.

In this paper, we propose a robust fault detection scheme for second-order multi-agent systems with external disturbance on directed graph. As the open-loop system of the multi-agent system contains disturbance term, there doesn’t exist an observer whose error dynamics won’t be influenced by the disturbance term. Hence we design observer for the augmented closed-loop system of multi-agent system based on the idea of unknown input observer, which is formed by adding an active disturbance rejection term to the control law. Then utilize above observers to generate residuals corresponding to all the agents of the MAS, the generators are only sensitive to the agents which the fault doesn’t occur, which means that there exists one residual generator different from others. Through further residual dynamics analysis, the residual corresponding to the faulty agent will be zero after the fault occurs, the others will fluctuate. Then we prove the existence of above observers that we design. Finally, the efficiency of the proposed scheme is illustrated by some simulation examples. The rest of this paper is organized as follows. In section 2, some basic knowledge of graph theory and reduced-order UIO are introduced, then the problem formulation is given. Section 3 gives the main results of this paper.

The aero-engine faults diagnosis is essential to the safety of the long-endurance aircraft. The problem of fault diagnosis for aero-engines is essentially a sort of model classification problem. Due to the difficulty of the engine faults modeling, a data-driven approach is used in this paper, based on the Relevance Vector Machine for classification. However, the collection of the fault sample is so difficult that causes the imbalance learning problem. To solve this problem, a semi-supervised learning approach based on the Improved Wasserstein Generative Adversarial Networks and K-Means Cluster technique is proposed in this paper. The theoretical analysis and the experiment show that, compared with another sampling method synthetic minority oversampling technique (SMOTE), the proposed approach can better fit the fault sample distribution, generate much more appropriate new samples by learning from the small number of fault samples. It is more efficient to prevent over-fitting by training with the original samples that mixed with the Improved Wasserstein Generative Adversarial Networks generated samples.

With the development of Internet technology, IP video services have become more and more popular, and video traffic plays a more important role in Internet traffic. Future Internet technology needs to improve its ability to cope with the ever growing resource demands for video services, while being able to provide users with satisfied Quality-of-Experience (QoE). In the work of this paper, based on the distributed traffic allocation algorithm FDTAA which is proposed in the literature to maximize the utility function of the connectionless network, this paper extends FDTTA to fit into the asynchronous framework, and simultaneously considers the influence of communication delay on the asynchronous algorithm. In the synchronous algorithm, all regions need to wait for the slowest region due to that these regions are usually heterogeneous and require different amounts of calculation time, which results in the waste of computation resources. The results of the numerical experiment show that asynchronous FDTAA can improve this situation. In addition, the communication delay will deepen the severity of asynchronism of the algorithm. The greater the communication delay, the more severe asynchronism and the worse the convergence performance of the algorithm. In addition, The convergence performance of the algorithm can be improved by changing the step-size parameters of the algorithm, at the price of the increase of iterations and calculation time.

This paper considers the problem of robust static output feedback (SOF) $H_{\infty }$ control for uncertain switched Takagi-Sugeno (T-S) fuzzy systems with mode-dependent average dwell time (MDADT) switching. The T-S fuzzy model is used to represent every subsystem of the switched systems. More practical asynchronous SOF $H_{\infty }$ controllers are designed in this paper, which means that the switching of the SOF $H_{\infty }$ controllers to be designed has a lag to the switching of the system modes. By allowing the output matrices of the studied switched T-S fuzzy systems to be non-common and have uncertainties, new stabilization conditions with $H_{\infty }$ performance for the uncertain switched T-S fuzzy systems are obtained. Furthermore, all these conditions are expressed as a set of linear matrix inequalities (LMIs). The desired SOF $H_{\infty }$ controllers can be obtained by solving these LMIs.

In this paper, we study the performance of model prediction system(MPC) when data is missing and unpredictable disturbances occur. Compared the recent result in the literature, the proposed observer-based output feedback MPC is designed by the aid of developing a new switching idea. At each sampling instant, the considered optimization problem is transformed into an online minimization of the infinite time domain performance index function. More importantly, the developed switching idea is well carried out by exploiting the newly updated information and a better control performance can be achieved in this paper. Finally, the comparison of simulation graphics proves that the proposed method has better performance than before.

Two-stage inverters have been commonly applied in the distribution systems, where Total Harmonic Distortion (THD) is important for the judgement of the tacking effects. This paper focuses on the current tracking of the grid-tied two-stage inverter composed by the front buck-boost converter and the rear single-phase inverter. A cascaded Proportional-Integral (PI) Model Predictive Control (MPC) is applied for the intermediate voltage tracking of the front stage. Meanwhile, an adaptive MPC scheme is adopted for the output current tracking of the single-phase inverter stage by introducing an time-varying observer for the filter inductance and load resistance identification. By regulating the intermediate voltage, there will be a corresponding voltage which minimizes THD under the same other parameters. Hence, a reference voltage generator based on the circuit parameters, the input voltage, the reference current and grid voltage is employed for the voltage regulation to obtain lower THD. The simulation results show the effectiveness of the control scheme for the two-stage inverters applied in this paper.

The identification of AutoRegressive eXogenous (ARX) model by outliers is addressed in this paper. Shifted(non-centralized) asymmetric Laplace (SAL) distribution and expectation maximization (EM) algorithm are employed to estimate the unknown model parameters. Outliers are common in the signal acquisition process and have a serious impact on data-driven modeling method. In this paper, the probability method is used to solve the problem of outliers. When the noise parameter is regarded as a prior exponential distribution, the model output obeys the SAL distribution which is robust to outliers. The known statistical properties of SAL distribution are applied to calculate the M-step in the EM algorithm and get the iterative parametric formula. The accuracy of the proposed algorithm is verified by a numerical simulation example.

Knowledge of position and velocity is a practical requirement for all Earth-orbiting spacecraft. Some satellite systems have requirement of autonomous orbit determination onboard. This paper focuses on developing methods for estimating absolute positions of two satellites using relative range and bearings. These measurements obtained in spherical coordinate are converted into Cartesian coordinate with converted bias and covariance being calculated. With converted measurements, an unbiased CMKF algorithm is used to estimate the position and velocity of the satellites. Square-root filter is introduced to avoid the loss of positive definiteness of state covariance matrix. Simulation results show that the algorithm proposed in this paper converges faster and has higher estimation accuracy.

In this paper, we are dedicated to investigate the globally asymptotical stability almost surely (GAS a. s.) and almost surely exponential stability (ES a. s.) of dual switching linear continuous-time systems based on the probability analysis method and Multi-Lyapunov function. To extend the results already obtained and to investigate some new aspects of such systems, our main contributions in this paper are: i) Switching signal design, i.e., the smallest expected value is selected as the deterministic switching signal. ii) Sufficient conditions of the (GAS a. s.) and (ES a. s) for the dual switching linear continuous-time systems. Different from previous work, which is a special case of our study, the transition rate matrix for the random Markov process in our study is not fixed, but varies when a deterministic switching takes place.

Due to the flexibility of the steel track, self-excited vibration between the electromagnet and the track may occur when the EMS (electromagnetic suspension) urban maglev train is suspending above the track in a standstill. The self-excited vibration produces annoying high frequency noise, decreases the ride comfort of the maglev vehicle, and may cause fatigues to track structures. Therefore, in a commercial maglev line, the track induced high frequency self-excited vibration must be eliminated. In this paper, a new configuration of the gap sensors for the suspension system is proposed. In this configuration, two gap sensors are employed to measure the suspension gaps at different locations of the electromagnet, which is different from the conventional configuration of the urban maglev system in which only one gap sensor is used at the terminal of the electromagnet. Using two gap sensors, the deformations of the track can be measured at different locations, and by appropriate selection of the distance between these two sensors, the vibratory components of the track deformation included in the measured gap signals may be greatly cancelled by adding them together, leading to possible elimination of the self-excited vibration. The modeling of the multi-span maglev track and the principle of the proposed scheme are both investigated through theoretical analysis; a simulation is further undertaken to validate the effectiveness of this scheme. The configuration proposed in this paper is easy to realize in the urban maglev vehicle, and it is also benefit for decreasing the vibration of the suspension system in the presence of track irregularities when the train is traveling.

Mapping UAVs have played a more important role in regional mapping than ever before, and coverage path planning is an important part of mapping tasks. However, there are few studies on single-region path planning in complex polygonal region. In this paper, we proposed a method that searches the optimal yaw-angle and flight height of mapping UAVs based on particle swarm optimization algorithm. Compared with traditional algorithms such as cattle farming algorithm, there are less redundant coverage and turning, and the flight path planned by the algorithm in this paper is more reasonable and convenient for photographic processing.

Target detection is one of the core algorithms in robot applications, and the recognition speed has a significant impact on robot’s target capture. In this paper, the scene of long-distance and small targets is used as the test scene, and the purpose is to enhance the speed of detection without reducing the accuracy of detection. Consider that the mask branch and excessive full connection layer in the Mask Rcnn network will take up a lot of network detection time, and the feature map extracted by the convolutional neural network has a high dimension, which will occupy a large amount of computational memory. So, in this paper the Mask Rcnn network is improved: remove the mask branch; introduce Light-Head Rcnn into the Mask Rcnn network, increase R-CNN subnet and RoI warping; adjust the proportion of the Anchor in the RPN network. Finally, the improved model is applied to the tensorflow framework. These methods can save computer memory space and improve detection’s speed. In the end, the improved Mask Rcnn network has been verified in a service robot platform with Kinect II. The test results show that compared with Faster Rcnn, the improved Mask Rcnn can have a high accuracy of detection; Compared with the original Mask Rcnn, the improved Mask Rcnn network can greatly improve the speed of the algorithm while ensuring the detection accuracy. The detection time is reduced by more than 2 times, which helps to improve the efficiency of the service robot’s target capture task.

The order, algebra and topology of the Bourbaki group are the foundation of modern mathematics. On the basis of the proof assistant Coq, the formal system of these three major parental structures can be constructed completely. On the basis of the formalization of axiomatic set theory, this paper present the formal framework of general topology. This paper completes the formalization of some basic definitions of general topology, including base, subbase, topology, and compact space. In addition, the formal proof of the Alexander subbase theorem is proposed on the basis of Tukey’s lemma. Finally we complete the formal proof of Tychonoff’s theorem, which is famous in general topology. All formal processes have been verified in Coq, which demonstrates the reliability, readability and rigor.

As social network analysis becomes a popular topic, more and more people show interest in the network community evolution in the future. Currently existing algorithms usually use machine learning methods to predict the future event that a community will undergo. However, these algorithms can't achieve multi-step prediction. This paper proposes two novel algorithms named MC-based prediction and CC-based multi-step prediction. The first one is carried out based on Marcov Chain (MC). In addition to performing multi-step prediction, the MC algorithm is able to achieve higher single-step prediction accuracy than the algorithms using machine learning methods in some cases. The second algorithm proposed in this paper applies the thought of Classifier Chain (CC) to multi-step prediction. Experiments show that these two algorithms are valid and have good performance for multi-step prediction.

This paper mainly studies the path planning algorithm for multi-robot. It introduces a gain constraint and a random factor to improve the artificial potential field method, which respectively suppresses the path oscillation and avoid the local minimum. Based on the improved artificial potential field method, the B-spline curve optimization is used to smooth the planned path. The simulation results show that the algorithm improvement made in this paper is effective in solving the multi-robot path planning problem.

This paper implements the latest speech recognition technique and improves it so that it can be used on embedded mobile devices. In order to implement the speech recognition system, I build the two-layer Long Short-Term Memory networks (LSTMs) using the TensorFlow framework and train the model by using the Mel Frequency Cepstral Coefficients (MFCCs) vector extracted from the wave files. In order to speeds up the model, this paper innovatively applies the pruning method to the LSTMs by pruning the weights of each cell and regularizing the weight matrix of regression layer. After pruning, I retrain the model to preserve the accuracy. Finally, this new model discussed in this paper successfully improves the performance of the original model without reducing the precision of the model. It reduces the computational complexity and speeds up the speech recognition process.

This paper mainly solves an online path planning problem of messenger UAV in an air-ground collaborative system which is composed of a fixed-wing unmanned aerial vehicle(UAV) and multiple unmanned ground vehicles(UGVs). The UGVs play the role of mobile actuators, while the UAV acts as a messenger to transmit the information among the UGVs. The online path planning problem for the messenger UAV can be modeled as a Dynamic Dubins Traveling Salesman Problem (DDTSP). The goal of this problem is to find a shortest path which enables the UAV to access all UGVs periodically. This paper proposes an online path planning method based on heuristic rules. The method can be divided into two stages: firstly, the sequence for the UAV to access the UGVs are determined by the heuristic rules which consider the relative positions and movement characteristics of the UGVs and the UAV, and then the access positions of the UAV are determined. Computational results on instances with varying characteristics show that the proposed method can generate high-quality solutions in a very short time. The comparison experiments demonstrate that, compared with another advanced method, our method can achieve a better balance between the quality of the solution and time consumption.

This paper is concerned with design of a deviation correction control system for vertical drilling process. Notice that only inclination angle is considered by traditional deviation correction systems, which may cause that the drilling trajectory extends around the wellhead’s plumb line or gets away from the line. In this paper, focusing on the drilling trajectory itself, a deviation correction control scheme is provided to improve the vertical precision of the vertical drilling while maintaining small inclination angles. A hybrid modeling method for establishing the trajectory extension model and a model predictive control strategy based on the model are integrated in the scheme. It is expected that the phenomenon of spiral-like drilling trajectory can be attenuated by our control system over the traditional ones.

This paper briefly introduces the relative motion control problems of the wearable exoskeleton robot. By approximating the ideal human gait as a periodical signal, an iterative learning control algorithm with a D-type updating law is applied to the shank part of the exoskeleton robot to solve the tracking problem. As the non-linearity in the dynamic model of this exoskeleton robot is leading to a degradation in the efficiency of normal iterative learning methods, a feedback linearization method is introduced to improve the overall performance of the algorithm. Finally, the effects of measurement noises to the algorithm are discussed.

An adaptive sliding mode control (ASMC) method with moving sliding surface (MSS) is proposed for anti-swing and positioning control of two-dimensional underactuated overhead crane in this paper. The designed moving sliding surface can change its slope in real time according to displacement error and speed error, thereby tracking performance of the trolley is improved. At the same time, a dynamic exponential term is added into the moving sliding surface so as to drive initial state variables into the sliding surface quickly to ensure the global robustness of the crane system. Finally, the controller switching gain is adjusted by an adaptive control law, which can effectively reduce chattering phenomenon occurring in the phase of sliding mode control. The controller can realize rapid positioning of the trolley and anti-swing control of the load under the condition of rope length changing. The global stability analysis of the suggested controller is given in the paper. The final simulation results confirm the effectiveness of the proposed controller.

This paper studies the H ∞ non-fragile filter design problem of a class of switched systems with conic-type nonlinearities. By reconstructing the system, the filtering error dynamic system with unknown input is obtained. The aim of this paper is to design a suitable non-fragile filter such that the error dynamic system is finite-time bounded and satisfies the given H ∞ index. By using Lyapunov functional and linear matrix inequalities, a sufficient condition for the existence of the H ∞ non-fragile filter gain is given and proved. Finally, a numerical example is given to illustrate the effectiveness of the proposed approach.

This paper addresses the diagonal stability problem of a class of discrete-time nonlinear positive switched systems with delays. Different from previous methods, we first use homogeneous polynomial Lyapunov function method to present the diagonal stability condition of the corresponding linear system of the original nonlinear system, and then use the comparison method to establish its relationship with the nonlinear system, so as to obtain the final result. Finally, a numerical example is provided to illustrate the theoretical results obtained in this paper.

Neural networks (NNs) are always perturbed by random noise such as continuous noise and jump noise. This paper is concerned with the passivity problem of NNs perturbed by jump noise and continuous noise. Since Gaussian noise can depict continuous noise rather than jump noise, Poisson noise is utilized to give a description of jump noise. Then by employing the predictable projection operator and Lyapunov theory, a passivity criterion is presented by the linear matrix inequality (LMI) format. The main contribution of this paper is to extend the passivity research to NNs perturbed by continuous noise and jump noise. Two numerical examples are provided to illustrate the effectiveness of the proposed criterion.

This paper proposes an improved neural network structure based on LSTM for recognizing entities in sentences. The deep learning model of LSTM has a label offset problem when performing sequence labeling tasks, because it does not utilize the information of the model’s output layer. Therefore, this paper adds a layer of neurons on the output layer of the model to simulate the hidden state in the CRF to make full use of the label information contained in the output layer. On the other hand, the LSTM-based sequence labeling model cannot ensure that the recognized entities exist in the knowledge base. Besides, it has difficulty in word boundary recognition, so this paper introduces the entity dictionary in the model. Finally, part of speech is the core feature of the entity. The char vector in the LSTM-based sequence labeling model lacks part of speech information, so this paper modified the character vector incorporating the part of speech feature information. Experiments show that the improved LSTM-based model has achieved good results in entity mention recognition.

In this paper, a cooperative guidance law based on an undirected communication graph is presented, which can keep the target within a narrow field-of-view (FOV) all the flight time and achieve simultaneous attack against stationary targets without access to global information. To maintain the seeker’s lock-on condition, this paper considers the real look angle without neglecting the additional sideslip angle caused by wind turbulence. However, when the missile’s speed exceeds a certain value, the additional sideslip angle will be small enough to be neglected and this value is derived. Based on the derived value, the guidance law is adaptive to switch its structure between FOV constrained sliding mode control guidance law and proportional navigation guidance (PNG) law, which can achieve zero miss distance with terminal command converging to zero, theoretically. Numerical simulations are conducted and the results prove the proposed law’s validity.

In this article, linear quadratic tracking (LQT) problem for Ito stochastic system with time delay in control input is studied. The difference between LQT and linear quadratic regulation (LQR) is that the state and control input are desired to track a reference trajectory. It is noteworthy that the separation principle is not valid for Ito stochastic system. In this paper, we design an optimal controller based on the maximum principle, in which the optimal controller minimizes the linear quadratic performance index of state and control input tracking errors. Based on the coupled Riccati-like equation, the explicit expression of the optimal controller is given. The key technology of this paper is to solve forward and backward stochastic differential equations (FBSDEs). Finally, a numerical example is provided to validate the results.

For synchronous generator within multi-machine power system described by MIMO nonlinear Differential Algebraic Equations (DAE) system model, the robust control problem is considered in this paper. The controller design process can be divided into two steps: In the first step, nonlinear DAE system model is equivalently transformed by a coordinate transformation and a feedback control. In the second step, a robust controller is constructed through backstepping method based on the equivalent model, which makes the whole closed-loop system uniformly ultimately bounded (UUB). Finally, the synchronous generator excitation and valve control is studied, and the simulation results illustrate the utility of the method proposed in this paper.

This paper focuses on designing a distributed observer for multiple sensors to estimate the input and state of a linear-time-invariant (LTI) single-input-single-output (SISO) discrete-time system with completely unknown periodic input. A consensus based adaptive observer to estimate the unknown input. The observer assumes the form of an adaptive finite impulse response (FIR) filter, which requires the agent to get the model and output of the target. In this paper, an adaptive distributed estimation algorithm is designed for some agents that cannot get the target model and output to estimate the target state, output and unknown input simply by relying on the information of its neighbors. Experimental results are provided to further verify and testify the effectiveness of the algorithm.

With the increase in drilling depth, the encountered formations become more complicated, and industrial controller for weight on bit has limited precision due to the uncertain and varying formations. In this paper, we design a scheme of control system for weight on bit to achieve higher control precision in complex geological drilling process. A double-loop control structure is introduced, which involves with two system models: a multi-degree-of-freedom lumped parameter drill string model and a finite element bit-rock interaction model. The effect of varying formations on weight on bit is analyzed, and a control strategy for control system is presented in this paper.

Urban fog-haze level detection is the basis for controlling haze formation and controlling fog-haze. The multi-level regression and fuzzy C-means clustering theory were applied to study the haze-level detection algorithm. This paper puts forward the non-meteorological fixed effect and meteorological random effect theory of fog-haze formation, emphasize the effect of random effects on fixed effects and form fog-haze. The fog-haze and the non-meteorological effect function are introduced, and the fog-haze level detection model is established. At the same time, the detection algorithm of FCM fog-haze level is given as the large sample problem. The range of using the two algorithms is given. And the false positive rate test of the two algorithms achieves higher accuracy. The results of this paper have certain theoretical significance and application value to the urban fog-haze prevention and control.

In order to realize the non-cooperative target relative pose estimation of spacecraft under the ultra-close distance, this paper presents a pose estimation method based on spatial circle reconstruction and the edge eigenvectors of spacecraft ontology by using binocular vision. The method mainly consists of three parts, Firstly; the spatial circle image in binocular vision camera is positioned by feature circle extraction algorithm. Based on this, the circle normal vector on the spacecraft is obtained by using spatial circle reconstruction algorithm, and its reconstruction error is analyzed. Then the linear features on the space ontology are extracted by using the clustering linear iterative fitting algorithm, and the intersection and eigenvectors are solved. Finally, the pose parameters of spacecraft are estimated by using the two normal vectors using the intersection vectors algorithm. Experimental results show that the algorithm proposed in this paper can effectively estimate the pose parameters of spacecraft; its position measurement accuracy is up to 20mm when the measuring distance is 10m, and the attitude measurement accuracy can reach 2°, which satisfies the precision requirement of space task to estimate the pose.

In this paper we mainly studies the cooperative tracking issue of multi-UAV based on formation control. The theory and method of formation control are introduced in the paper to ensure the continuous and effective ground target tracking results. The tracking targets are regarded as the leader or the structural center of the formation. The multi-UAVs are considered as followers to form an unstable formation system. The formation stability can be maintained by constructing control law to ensure the continuous and effective tracking of intelligent targets by the multi-UAV. The formation control method based on the leader-follower system and the potential function are adopted to ensure the stability of formation consists of UAVs and the target, so that UAVs can keep tracking of the target. We prove the tracking strategy and the formation control results by MATLAB simulation.

It is well known there is no closed form for the roots of general polynomials with degree over four. Thus, in essence, all roots finding algorithms are approximate methods, which may also lead to incorrect results. However, Routh table method is strictly proved method for the root distribution of general polynomials and systems in the complex plane. On the other hand, when the special singular cases happen in the Routh table, the Routh table method may also meet difficulty. This paper presents a new Routh-type table method for all general polynomials and their systems. The proposed method has advantage when the first special case happens and/or both the first and second special cases happen in the Routh table. In these cases, the classical Routh table will give incorrect results and meet dilemma for the root distribution, while other existing methods in the literature had to use complex methods to solve it. The proposed method in this paper is not only easy, but also may reveal the characteristics of polynomial roots and coefficients well. Up to our best knowledge, this proposed new method is the easiest way and simplest method for root distribution analysis of general polynomials in the literature when the special cases happen.

In this paper we study delay robustness of PID controllers in stabilizing systems containing uncertain, variable delays. We consider second-order unstable systems and seek analytical characterization and exact computation of the PID delay margin, where by PID delay margin we mean the maximal range of delay values within which the system can be robustly stabilized by a PID controller. Our contribution is threefold. First, we show that the delay margin achieved by PID control coincides with that by PD controllers. Second, we show that other than helping stabilize the delay-free part of a plant, the proportional control contributes no action to increase the delay margin. Finally, we show that the PID delay margin can be computed efficiently by solving a unimodal problem, that is, a univariate optimization problem that admits a unique maximum and hence is a convex optimization problem in one variable. This unimodal problem is one of pseudo-concave optimization and hence can be solved using standard convex optimization or gradient-based methods. As such, from a computational perspective, the PID delay margin problem is completely resolved in this paper! The results not only insure that the PID delay margin problem be readily solvable, but also provide fundamental conceptual insights into the PID control of delay systems, and analytical justifications to long-held engineering intuitions and heuristics, thus lending useful guidelines in the tuning and analytical design of PID controllers.

Automation of goods transportation in warehousing system is essential to improve productivity and reduce costs for the modern logistic enterprises. Under this background, automatic warehousing is becoming a key to logistic systems due to its ability of scheduling and optimizing delivery tasks of robots and, consequently, increasing productivity and decreasing costs. This paper firstly describes briefly problems needed to be solved for working of multiple mobile robots in automatic warehousing systems. Then, we build a map model based on topological graph for the warehouse environment and choose a storage representation of adjacent matrix for it. This paper details the algorithm of task scheduling and path planning for multiple mobile robots in warehousing system. The presented algorithm applies the ${\mathrm {A}}^{*}$ shortest-path method to solve path planning problem of single robot and, further, introduces time window approach to solve traffic jams and collisions of multiple robots. Computer simulation tests were conducted to compare the efficiency of the ${\mathrm {A}}^{*}$ algorithm with other shortest-path methods of path planning for single robot. For logistics scenario of multiple robots, routing simulations were also carried out to verify performance of the purposed algorithm under different working conditions. Test results showed that the algorithm could generate collision-free and optimized paths for multiple mobile robots.

In order to deploy multi-robot into a 2-D desired formation during the seeking process, this paper proposes a hybrid seeking algorithm with the formation deployment. Firstly, this paper expands the 1-D deployment algorithm to 2-D, and then combines with the extreme seeking algorithm to obtain the hybrid seeking algorithm. With the algorithm, multi-robot can deploy into a desired 2-D formation during seeking process. Simulation results are provided to illustrate the effectiveness of the algorithm.

Drilling rigs are key devices in tunnel projects, which dig holes on a rock face by drilling arms. The quality and efficiency of tunnel constructions are directly determined by the positioning accuracy of the holes and time for traversing all the holes. This paper provides a complete solution for trajectory planning with accuracy and efficiency. First, this paper analyzes the structure of a drilling arm, establishes a forward kinematic model of the drilling arm by Denavit-Hartenberg convention. Next, to solve the inverse kinematics problem of drilling arm, a particle swarm optimization with adaptive inertia weight is proposed. Then, an optimal drilling sequence is found based on Hopfield network. Thereafter, a quintic polynomial method has been adopted to calculate the trajectory of the drilling arm. Finally, results of simulations show that the solution satisfies the accuracy of positioning and calculates the optimal trajectory.

The paper proposes a new encoding method inspired from the block chain technology, by which an operation sequence information and the corresponding machine information is integrated in an operation node and links all operation nodes to form an operation list with pointer technology of C++ program language. Accordingly, the paper designs an individual unit ID for every possible cases. Subsequently, it describes the process of searching the optimal solution of FJSP by implementing an improved UKPK-PSO algorithm, which hybridizes a varied PSO algorithm by getting the best information from K informants and the method of keeping the elitists for next generation in iterations. With the improved algorithm we can more efficiently traverse the solution space. After obtained some better solutions with the improved PSO, we continuously search the optimal solutions by tentatively changing the chosen machines of the dominant operations. Finally, it outputs the optimal solution through some limited iterations again. By analyzing the experiment results for the instances, the paper furtherly proves that the whole algorithm is effective for FJSP.

In order to realize obstacle intrusion detection on railway track, this paper proposes a detection scheme based on channel state information (CSI) that is exposed by commodity Wi-Fi AP. In this paper, the characteristics of multi-path propagation of wireless signals under obstacles are innovatively utilized. By collecting and processing CSI of wifi signals, the MUSIC algorithm is able to calculate the number of propagation paths of wifi signals with smoothed CSI, then it is easier to determine whether the wifi signal encounters obstacles in the process of propagation. In this way, obstacle intrusion detection of railway track can be realized. Simulation results verify the effectiveness and accuracy of this method. This is the first time to analyze CSI of wireless signals in railway operating environment and realize obstacle intrusion detection, which is also a research direction to be developed and has great research value.

The planar stable walking control of the biped robot with torso is studied in this paper. Firstly, improve the spring-mass model by adding a rigid torso to the hip joint and then get the spring-mass model with torso. Secondly, control the torque applied to the hip joint and regulate the stiffness of the springy leg to make the robot converge to the desired gait. Due to the nonlinear and strong coupling characteristics of the system model, use the feedback linearization to transform the nonlinear model into a linear model. Thirdly, the robot's torso angle and hip joint trajectory are controlled overall using PD control and P control. Finally, based on the theoretical analysis, the control method is simulated. The simulation results show that the feedback linearization and variable leg stiffness control methods used in this paper can achieve stable walking of the biped robot with torso, and have good anti-disturbance performance.

This paper investigates the necessary conditions on the number of unstable eigenvalues of the single-input positive systems for stabilizability. For a continuous-time linear positive system, it has been known that if there exists a stabilizing linear time-invariant controller, the number of nonnegative real eigenvalues cannot be greater than one. For the discrete-time systems with nonnegative state constraints, it is demonstrated in this paper that there can exist at most one unstable real eigenvalue as well if every state can be stabilized by a linear time-invariant controller. Regarding the unstable complex eigenvalues, a specific three-dimensional system is studied under the positivity constraint. In addition, the optimal control problem is discussed for the positive double-integrator systems, where only one state needs to be stabilized and the other one is considered in a cost functional.

This paper investigates the problem of switching neutral time-delay systems with actuator saturation. For switching systems, the switching law and input saturation influence the design of state feedback controller. A switching state feedback controller is designed to estimate the disturbance. The auxiliary time-delay feedback is introduced during this paper. By using Jensen integral inequalities, free-weighting matrix technique and augmented Lyapunov-Krasovskii functionals, some local stabilization conditions of the controller which are proposed in terms of linear matrix inequalities ensuring the asymptotically stable of the closed-loop system. The domain of attraction of which are estimated and optimized. Finally, a simulation example is presented to show the effectiveness of our proposed technique.

The precision motion stage exists the non-smooth nonlinear dead-zone varying with displacement and saturation simultaneously. In order to accurately describe such an electromechanical motion system, a novel Hammerstein-Wiener system (HWS) is firstly introduced to handle two non-smooth nonlinearities in this paper. Then, a three-step identification scheme based on two separable input signals is presented to estimate the parameters of HWS. Finally, the experimental results obtained by implementing the proposed identification scheme are also presented in this paper.

Virtual Traffic Simulation has always played an irreplaceable role in the study of urban transportation system, especially in intelligent vehicle infrastructure cooperative system (i-VICS), which is difficult to implement testing. A significant challenge is that the construction of large-scale simulation environment requires much time and the authenticity of the traffic simulation is doubted for various reasons. This paper presents a simulation scenario construction and fidelity evaluation method, which can build the traffic network efficiently and obtain the appraisal results easily. In this paper, the architecture and main function of traffic simulation system is analyzed firstly, which can achieve the extraction of road data and realize the construction of road network with 50 intersections covering about 18 square kilometers in 48.201 seconds. Besides the analysis on the function of system, this paper also evaluates the fidelity of simulation system. The evaluation system is mainly consisting of the construction of index system and determination of index weight. The evaluation result shows that the virtual models and traffic scenario of the simulation system has good fidelity. Thus, the traffic simulation system can meet the needs of the general simulation.

This paper mainly introduces a hybrid non-isolated DC-DC converter based on voltage lifting technology. This circuit is formed by introducing a new switching inductance structure on the basis of a unique voltage lifting structure. In this paper, the characteristics of this circuit in Continuous Inductor Current Mode (CCM) are studied, and the voltage and current are analyzed respectively. The voltage gain of this improved converter and other DC-DC converters in CCM mode is compared. Finally, this converter is simulated in Multisim to verify the conclusion of calculation.

Time-domain interpolation is often used to reduce harmonic analysis errors of voltage and current signals in power grids. In the time-domain interpolation process, a zero-crossing detection circuit is usually required to measure the fundamental frequency. However, zero-crossing detection usually increases hardware cost and measurement delay. In the zero-crossing detection, the existence of DC component will make the measurement result has great error, even lead to measurement failure. Therefore, this paper adjusts the Asynchronous Sampling data according to the sampling data and the fundamental frequency obtained by Taylor series expansion, and then adopts discrete Fourier transform(DFT) to carry out harmonic analysis. In addition, a detailed formula derivation process is given in this paper, and a simulation experiments is carried out. As a result, the method in this paper is verified experimentally.

International audience; In this paper, we present a new strategy to control the soft robot with elastic behavior, piloted by 4 actuators. The main contribution of this work is the use of neural network to get approximated model soft robots, based on which a robust controller is then proposed. In this paper, we proved that if the approximated model satisfies certain conditions, then the proposed robust controller can always drive any given point of interest of the robot to the desired position, without knowing the exact model. Finally, the proposed result is experimented and validated by a 3D printed silicone soft robot.

In this paper, we present a gain scheduled reference governor and its switching rules for the system with constraints on the state and/or input variables. The purpose of this paper is to shorten the settling time of the linear system output. For the linear system, we add two state feedback gains: K 1 and K 2 , which are computed off-line by solving linear matrix inequality. K 1 is designed to make the rise time of the linear system output faster (or shorter), while this usually leads to a big overshoot. K 2 is devised to make overshoot of the linear system output become smaller, while there exists the slower (or longer) rise time. In order to fulfill the constraints of the linear system involved $K_{i(i=1,2)}$, we take use of the control idea of the reference input governor method, which is based on the invariant set. The invariant sets of the linear system involved with the gains K i are obtained off-line. By choosing a right switch time, we transform the gains from K 1 to K 2 so that the settling time of linear system output will be shortened. Finally, we show some examples or application.