Multi-Input Variable Structure Controllers

Sliding mode control of a DC/DC PWM converter with PFC implemented by neural networks

“…Traditionally, to implement these kind of control laws it has been necessary to use a digital microprocessor [14] to evaluate all the expressions introduced in the previous section, due to the complexity of these calculations. Moreover, it is also necessary several analog to digital converters to measure state variables in the power circuit.…”

“…where are nonlinear and complex functions of the inputs variables as it can be seen in the following equations: The parameters of the power circuit and the eigenvalues selected have been obtained as suggested in [14].…”

“…The shape of the input current should match the shape of a signal proportional to the input voltage as close as possible, depending on the power delivered by the converter (load variation); • the output voltage should be equal to the reference voltage A switching function has been designed for this application as follows: The required system behavior has been developed using eigenvalue assignment [12]. If the eigenvalues and the corresponding eigenvectors are selected as suggested in [14] the corresponding switching surface is defined by the following equations:…”

“…The variables used to control the converter have been the input current the capacitor voltage the inductance current and the output voltage The power converter can be described by the following equations [14], obtained from the circuit shown in Fig. 1:…”

“…Classical controllers, which use a PI approach, are very sensitive to variation of system parameters. Other techniques try to optimize the system response using neural networks [1]- [10], sliding modes [12]- [14], etc.…”