This paper addresses a high-order sliding mode control strategy for output tracking of nonholonomic mobile robots. First, we introduce the dynamic model of robots, driving motors and nonslipping kinematics constraint conditions. Second, we decompose the system into linear and nonlinear components via diffeomorphism and nonlinear input transformation. Also we consider parameter variations of robots and deduce the uncertain model of robots. Third, we design a high order sliding mode controller for output tracking of known and uncertain systems, respectively. Finally, we perform numerical simulations, demonstrating that the proposed high-order sliding mode control not only reduces the chattering problem of sliding mode systems, but also has certain robustness properties with respect to uncertainties of robots.

An adaptive backstepping and high order sliding modes control algorithm is proposed for output tracking of mobile robots. The controller can greatly reduce the chattering due to conventional sliding modes technique. The proposed algorithm has certain robustness with respect to the external random disturbances and good adaptability with respect to the parametric uncertainty. The effectiveness of the proposed control strategy is demonstrated by simulations studies.

Software-Defined Radio (SDR) receiver has the ability of operating in a multi-mode environment and has wide applications. However, efficient recognition of the currently active modulation system in real-time is a major problem faced by many applications. In this paper, an efficient method for the recognition of modulation system in a SDR receiver is proposed. The method is a classical two-stage approach based on (i) decision feature extraction and (ii) modulation system classification. In the first stage, decision features are extracted by the use of digital quadrature polyphase filter. In the second stage, an efficient parallel decision algorithm is proposed to classify the active modulation type. This proposed algorithm is proof to be more efficient than the conventional type of decision-tree approach. The complete recognition system is implemented using MATLAB. Simulation result shows that the proposed method achieved good robustness even with the presence of band-limited Additive White Gaussian Noise (AWGN). The overall successful recognition rate of 98.5% can be achieved even at a low signal-to-noise ratio (SNR) of 8 dB.