Search

In this work two main robust control strategies, the sliding mode control (SMC) and nonlinear H 8 control, are integrated to function in a complementary manner for tracking control tasks. The SMC handles matched L 8 [0,8) type system uncertainties with known bounding functions. H 8 control deals with unmatched disturbances of L 2 [0,8) type where the upper-bound knowledge is not available. The new control method is designed for a class of nonlinear uncertain systems with two cascade subsystems. Nonlinear H 8 control is applied to the first subsystem in the presence of unmatched disturbances. Through solving a Hamilton-Jacoby inequality, the nonlinear H 8 control law for the first subsystem well defines a nonlinear switching surface. By virtue of nonlinear H 8 control, the resulting sliding manifold in the sliding phase possesses the desired L 2 gain property and to a certain extend the optimality. Associated with the new switching surface, the SMC is applied to the second subsystem to accomplish the tracking task, and ensure the L 2 gain robustness in the reaching phase. Two illustrative examples are given to show the effectiveness of the proposed robust control scheme. [ABSTRACT FROM AUTHOR]

In this paper, a modular control approach is applied to a permanent-magnet synchronous motor (PMSM) speed control. Based on the functioning of the individual module, the modular approach enables the powerfully intelligent and robust control modules to easily replace any existing module which does not perform well, meanwhile retaining other existing modules which are still effective. Property analysis is first conducted for the existing function modules in a conventional PMSM control system: proportional-integral (PI) speed control module, reference current-generating module, and P1 current control module. Next, it is shown that the conventional PMSM controller is not able to reject the torque pulsation which is the main hurdle when PMSM is used as a high-performance servo. By virtue of the internal model, to nullify the torque pulsation it is imperative to incorporate an internal model in the feed-through path. This is achieved by replacing the reference current-generating module with an iterative teaming control (ILC) module. The LLC module records the cyclic torque and reference current signals over one entire cycle, and then uses those signals to update the reference current for the next cycle. As a consequence, the torque pulsation can be reduced significantly. In order to estimate the torque ripples which may exceed certain bandwidth of a torque transducer, a novel torque estimation module using a gain-shaped sliding-mode observer is further developed to facilitate the implementation of torque learning control. The proposed control system is evaluated through real-time implementation and experimental results validate the effectiveness. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, the robust filtering problem is investigated for networked stochastic systems with norm bounded uncertainties and imperfect multiple transmitted measurements. The considered imperfect measurements contain randomly occurring sensor nonlinearities and packet dropouts, which are represented by multiple independent Markov chains with partially unknown transition probabilities. A one to one mapping is constructed to map the multiple independent Markov chains to an augmented one for facilitating the resultant system analysis. A sufficient condition is established to guarantee the exponential mean-square stability with fast decay rate and a certain performance level of the filtering error systems. Then, the parameters of the full-order filter are expressed in terms of linear matrix inequalities (LMIs). Finally, a numerical example is shown to demonstrate the effectiveness of the proposed method. [Copyright &y& Elsevier]

This article considers the output tracking control problem of nonidentical linear multiagent systems (MASs) using a model-free reinforcement learning (RL) algorithm, where partial followers have no prior knowledge of the leader's information. To lower the communication and computing burden among agents, an event-driven adaptive distributed observer is proposed to predict the leader's system matrix and state, which consists of the estimated value of relative states governed by an edge-based predictor. Meanwhile, the integral input-based triggering condition is exploited to decide whether to transmit its private control input to its neighbors. Then, an RL-based state feedback controller for each agent is developed to solve the output tracking control problem, which is further converted into the optimal control problem by introducing a discounted performance function. Inhomogeneous algebraic Riccati equations (AREs) are derived to obtain the optimal solution of AREs. An off-policy RL algorithm is used to learn the solution of inhomogeneous AREs online without requiring any knowledge of the system dynamics. Rigorous analysis shows that under the proposed event-driven adaptive observer mechanism and RL algorithm, all followers are able to synchronize the leader's output asymptotically. Finally, a numerical simulation is demonstrated to verify the proposed approach in theory.