This study proposes an exosystem-observer approach to an output feedback problem with an output regulation constraint. The authors analyse the error covariance of a recently proposed exosystem observer and newly derive the degenerate Riccati equation and linear matrix inequalities for designing the optimal and suboptimal exosystem observer gains. Furthermore, they show that the output feedback controller based on the exosystem observer is equivalently implemented in a Youla-parameter-based structure, where the Youla parameter is realised as an observer for the exosystem state. By utilising the Youla-parameter-based structure, they provide output feedback controller design methods for achieving the optimal and suboptimal performances.

In this study, the finite-time cooperative tracking issue for non-linear multi-agent systems with stochastic disturbances is addressed. Different from the existing asymptotical stable control, the finite-time cooperative control can ensure the outputs of agents reach agree in finite time by raising a novel finite-time stability criterion. The radial basis function neural networks are employed to cope with the unknown non-linearities caused by the non-linear non-strict feedback form. Moreover, this study designs a new adaptive quantised control strategy to reduce communication burden. With such a control scheme, the tracking errors converge to a small area of the origin in finite time. Finally, some numerical simulation results are presented to testify the effectiveness of the approach proposed in this study.

To solve the problem that the frame of finite-time consensus in probability has not been proposed, this note discusses the finite-time consensus in probability for stochastic multi-agent systems. The main contributions exist in (i) a framework for constructing effectively distributed protocols is presented to improve the convergence rate of protocol, which can ensure the finite-time consensus in probability for stochastic multi-agent systems under undirected connected communication topology; (ii) the relationship is discussed between the convergence time and the influence factors including the protocol parameters, the algebraic connectivity, and the initial state. Finally, some simulations are performed to verify the effectiveness of the theoretical results.

The controllability and observability of networked systems are studied, where the network topology is directed and the nodes are time-invariant singular linear systems. Under the regularity assumption, a specific condition for the R-controllability of the network with single-input single-output node-systems is first established. Furthermore, necessary and sufficient conditions are derived for the R-controllability and C-controllability of the network with multi-input multi-output node systems. It is shown that the controllability of the overall system is an integrated result of the network topology, the subsystem dynamics, the external inputs, and the inner interactions. Additionally, corresponding observability criteria for such systems are also obtained, which indicate that the observability of the whole system depends only on the parameters of its subsystem. Finally, several examples are given to illustrate the proposed results.

This study proposes a distributed model predictive control (DMPC) scheme based on population games for a system formed by a set of sub-systems. In addition to considering independent operational constraints for each sub-system, the controller addresses a coupled constraint that involves the sum of all control inputs. This constraint models an upper bound on the total amount of energy supplied to the plant. The proposed approach does not need a centralised coordinator when having a coupled constraint involving all the decision variables. The proposed methodology, which takes advantage of evolutionary game theory concepts, provides an optimal solution for the described problem. Moreover, it is shown that the methodology has plug-and-play features, i.e. for each already designed local MPC controller nothing changes when more sub-systems are added/removed to/from the global constrained control problem. Furthermore, the stability analysis of the proposed DMPC scheme is presented.

This work presents a solution to the position-heading control problem of a small quadrotor. The main idea of this study is to exploit the advantages of the linear parameter varying (LPV) control approach together with a feedback linearisation approach for designing a cascade control structure without the need to perform simplifications with respect to the reference dynamic model used in the literature. The use of a LPV representation of the attitude subsystem allows to consider it as independent of the position one, allowing the use of a cascade control structure. For the inner attitude control loop, the proposed quasi-LPV–linear-quadratic regulator (LQR) controller and proportional–integral observer (PIO) are solved in an optimal manner by solving a set of linear matrix inequalities (LMIs). While for the outer position control loop, a feedback linearisation approach is developed before designing the corresponding controller and PIO also minimising the LQR problem by means of LMIs. Furthermore, another outer loop, interchangeable with the position one, such that controls the velocities in the body frame is proposed. This velocities controller is also solved in an optimal manner using a combination of feedback linearisation and LPV techniques. The proposed solutions are applied to an AscTec Hummingbird unmanned aerial vehicle in simulation.

In this study, the authors propose a stabilising data-based model predictive controller for systems subject to constraints in which the prediction model is inferred from experimental data of the plant using a machine learning technique. The inference method is a modification of the kinky inference tailored for model predictive control. In particular, the modified method has a lower computational effort and provides smoother predictions than the original method. The controller formulation considers soft constraints in the outputs, hard constraints in the inputs and guarantees closed-loop robust stability as well as performance by means of the use of different control and prediction horizons and a weighted terminal cost. Under the assumption that the model of the system is Hölder continuous, they prove that the closed-loop system is input-to-state stable with respect to the estimation errors. The results are demonstrated in a case study of a continuously stirred-tank reactor.

Prediction is an essential part of predictive control and is widely applied in control engineering. In a distributed control system configuration, there are signal transmissions between local subsystems. Communication delays impose a limit on the achievable prediction performance. Even though there is a plethora of literatures available for multiple step ahead prediction under the centralised framework, they are traditional techniques that cannot be applied to the distributed framework due to the interactions between the different subsystems and communication delays that have not been taken into account. In this study, a technique for determining the distributed multiple step ahead prediction is proposed which is optimal in sense of the lowest total prediction error variance. The proposed approach is useful for designing optimal controllers or assessing the performance of the implemented control loop while the controller structure has the distributed framework.

In this study, the authors address the problem of composite cooperative state estimation and system identification for linear multi-agent systems (MASs) under the leader–follower framework. This problem specifies an objective for each follower agent to estimate its local plant state and identify its plant dynamics simultaneously through interacting and communicating with its neighbours. A novel distributed adaptive estimation and identification protocol is proposed, which possesses five important properties to improve existing cooperative approaches: (i) it deals with general linear MASs with a completely unknown agent dynamics, allowing dynamics heterogeneity between the leader and the followers; (ii) it enables simultaneous exact state estimation and accurate system identification of MASs with guaranteed exponential convergence performance; (iii) it utilises only relative measurements from neighbouring agents, but without requiring any knowledge of absolute local plant states/outputs; (iv) it is fully distributed in the sense that its design and implementation do not involve any global information, including the overall network connectivity and the leader's dynamics; (v) exact system identification can be achieved under a cooperative persistent excitation (PE) condition, which significantly relaxes the classical PE condition. Simulation studies have been conducted to demonstrate the effectiveness of the proposed approach.

This study focuses on the state estimation problem for a linear system with unknown input and random false data injection attack. The unknown input is treated as a process with a non-informative prior. A residue-based detector is used to improve security of the linear system due to the randomness of the attack. Based on different detection information scenarios provided by the detector, a novel state estimator against the false data injection attack is proposed. Convergence and stability on the state estimation are investigated, and sufficient conditions are established to ensure boundedness of mean error covariance. Finally, the effectiveness of the proposed method is demonstrated by a numerical example.

Grasping and manipulation are common robotic tasks which require to properly control the interaction forces and position of multiple robots end-effectors. In this study, the authors focus on the problem of dexterous manipulation of rigid objects by means of a cooperative robotic system. Unlike several works, they consider the more complex case when the manipulated object is not mechanically attached to the robots' end-effectors. In addition, they address the control problem of designing a control law with continuous and bounded input torques for cooperative robots in constrained motion. To guarantee a stable grasp and a fine manipulation a centralised control algorithm based on the Orthogonalisation principle and a dynamic sliding mode control is proposed. The control algorithm does not require the knowledge of the dynamic model of the robot manipulators and manipulated object for implementation. Experimental results are presented to show the good performance of the proposed controller. In addition, the stability analysis of the closed-loop dynamics is developed.

In this paper, an adaptive control approach based on the multi-dimensional Taylor network (MTN) is proposed to control the non-linear uncertain time-varying systems with noise disturbances. MTNs are introduced to formulate adaptive filtering, non-linear identification and optimal control. First, an MTN filter is developed to eliminate the control interference and measurement noise, so that the model output without stochastic disturbance can be obtained. Then, an MTN identifier (MTNI) is so designed as to be capable of dynamic mapping and require fewer weights than traditional neural networks. On the basis of the above, the MTN controller (MTNC) is developed to realise the precise tracking control of the system. The non-linear uncertain time-varying system is identified by MTNI, which then provides sensitivity information of the plant to MTNC to make it adaptive. Furthermore, the skeletonisation algorithm is adopted to remove redundant inputs and redundant regression items from MTNI and MTNC for concise MTNs. Successful convergence and faster learning are guaranteed using the Lyapunov theorem, and the optimal learning rates are identified. Simulation results demonstrate that the proposed approach features its accurate identification, excellent tracking and better anti-interference capability for the adaptive real-time control of uncertain, stochastic and time-varying non-linear systems.

The stability problem of a three-phase voltage source PWM AC–DC converter is addressed through an improved adaptive backstepping method, which combines the following contents: input–output linearisation, adaptive backstepping method, error compensation and sliding mode control. The state equation of the three phase system is obtained by the power balance relationship between the input current and the output voltage, and then the coordinate conversion is carried out in order to design the controller. When the input–output linearisation method transfers the system model to a strict-feedback form, an improved adaptive backstepping control approach is given to stabilise the converter. Finally, the simulation results show that the improved method has three benefits: retain a faster stability of rectifier, reduce overshoot of DC voltage and bring about a real-time parameter estimation.

It is well known that if a network topology is a path or line and the states of vertices or nodes evolve according to the consensus policy, then the network is Laplacian controllable by an input connected to its terminal vertex. In this work, a path is regarded as the resulting graph after interconnecting a finite number of two-vertex antiregular graphs and then possibly connecting one more vertex. It is shown that the single-input Laplacian controllability of a path can be extended to the case of interconnecting a finite number of k-vertex antiregular graphs with or without one more vertex appended, for any positive integer k. The methods to interconnect these antiregular graphs and to select the vertex for connecting the single input that renders the network Laplacian controllable are presented as well.

This work studies the mixed-delay-dependent asymptotical stability for a class of time-delay neutral systems and three less conservative criteria are presented in terms of linear matrix inequalities. Firstly, by proposing an improved dynamic Lyapunov method, an augmented Lyapunov–Krasovskii functional (LKF) is constructed, which does not only benefit from the information on each time-delay but also heavily depends on the interconnection between time-delays. Then, two effective Wirtinger-based integral inequalities and an extended reciprocal convex technique are employed to give much tighter bound on the LKF's derivative. In all, these treatments above can result in much less conservatism. Finally, two numerical examples are presented to demonstrate the effectiveness and advantages of our proposed criteria.

In this study, a distributed adaptive control strategy for distance-based formation and flocking control of multi-agent system with parametric uncertainty is proposed. A 3-agent leader–first-follower system is considered first. The agents are supposed to have non-identical dynamics, whereas with similar structure. By introducing two new variables that involve distance error and velocity error and utilising them in the controller design, global convergence of all agents to the desired formation and to the velocity of the leader is established. Finally, the three-agent case is inductively extended to N-agent case. The stability and the effectiveness of the proposed control strategy are analysed in theory and demonstrated through simulation results, respectively.