Neural network control-based adaptive design for a class of DC motor systems with the full state constraints

An Adaptive Speed Control Approach for DC Shunt Motors

“…Moreover, a nonlinear dynamic system could be subjected to unknown endogenous and exogenous disturbances affecting the performance indicators specified for the system response. There are several controllers based on sliding modes [2], predictive control [4], conventional proportional-integral-derivative (PID) techniques [5][6][7][8], robust control [9], neural networks [1,10], and fuzzy logic [11]. Some relevant aspects of these contributions are described in the next paragraphs.…”

“…Some relevant aspects of these contributions are described in the next paragraphs. Nevertheless, most of them require full or partial information of the mathematical model and motor parameters, limiting their application because controllers are depending on the availability of these values [4,10,11]. Motors as electric actuators are employed in many practical engineering systems; therefore, obtaining this information is an additional task, and it is possible that the controller performance could be degraded.…”

“…Motors as electric actuators are employed in many practical engineering systems; therefore, obtaining this information is an additional task, and it is possible that the controller performance could be degraded. Moreover, some of them have a high computational cost which could restrict their operation for real time applications [10]. For instance, radial basis function neural networks are formed by three layers (input-hidden-output), and the number of neurons in each layer is also an important issue for correct performance.…”

“…Moreover, simulation results are exhibited for design performance evaluation. In particular, the analysis and design of DC motor controllers is emphasized for so-called direct current permanent magnet, or separate excitation, reaching simple linear models, and justifying its performance against certain operating conditions [4,8,10,11]. Its operation is guaranteed around some equilibrium point and is also highly dependent on the knowledge of the motor parameters.…”

“…The use of adaptive control algorithms offers an attractive alternative for speed tracking of DC motors [10,11,13,14]. A direct adaptive fuzzy logic controller is exposed in [11]; it is estimated from two levels-one uses a Mamdani fuzzy controller and the other is an inverse model based on a Takagi-Sugeno method.…”