Control System Design

Control System Design Process

The first step in designing a controller is the development of a high-fidelity model of the plant to be controlled. Then a controller can be designed to control the plant within design specifications. To do this, VisSim supplies tools like VisSim/Analyze, a frequency domain analysis model and VisSim/OptimizePro, a global parameter optimizer.

Modeling the Plant

The VisSim mathematical engine provides fast and accurate solutions for linear, nonlinear, continuous time, discrete time, time varying, and hybrid systems. VisSim lets you easily build plant models by simply selecting and connecting function blocks. Models can be linear, nonlinear, continuous, discrete, hybrid, SISO or MIMO.

VisSim supports hierarchical design by letting you group blocks into compound block subsystems. You can also create custom blocks using C, Fortran or Pascal and add them to the VisSim block library. There is virtually no limit to model size or complexity.

Once assembled, a click of the GO button simulates the plant model. VisSim data import blocks allow easy comparison of model output to actual plant output. In VisSim, responses are immediately visualized with plots, stripcharts, meters and gauges. The plant model can be refined until the model accurately reflects the desired response. Plant models can also be derived through system identification methods which "reverse engineer" the plant model from measured plant data.

Frequency Domain Analysis

After the plant model is verified, the dynamics of a plant can be analyzed with VisSim's Analyze option. VisSim/Analyze approximates the dynamics of a nonlinear system by linearizing the system about the current operating point. Linearized systems can be represented in ABCD state-space or transfer function form. With VisSim/Analyze, you can easily access transfer function information, edit zeros and poles, and obtain Nyquist, Bode, and root locus plots.

E+Bike
Parameter optimization of PID gains using the VisSim global optimizer

Controller Design

The next step is to design a controller for the plant model by interactively editing compensator zeroes and poles, and observing their combined behavior in Bode and root locus plots. Once the desired responses are obtained, the pole placement controller block is simply inserted into the VisSim diagram.

The preliminary pole placement controller is then connected to the plant model creating a feedforward or a feedback control loop. A simulation is run and the results of the simulation can be viewed in the form of plots. The stability of the closed-loop system can then be determined using Nyquist plots. In addition to pole placement controllers, VisSim provides pre-configured PI, PD and PID controllers that can be easily customized or optimized for specific user requirements.

Simulation

Simulations can be set up to run in interactive, batch, or single-step modes. VisSim's highly interactive interface makes it easy to perform "what if" simulations. For example, you can dynamically change parameter values like controller gains and VisSim will immediately display the corresponding changes in system behavior.

Both linear and nonlinear systems can be simulated with VisSim. Nine different integration algorithm options are available that offer a trade-off between speed and accuracy.

VisSim/OptimizePro can determine optimal values for design variables subject to user-defined constraints starting from initial user guess values. For example, VisSim/OptimizePRO can automatically calculate optimal PID controller gains that give minimal time to setpoint, plus minimal overshoot. It allows user-specified cost functions that can consider controller behavior, such as steady-state error, overshoot and rise and settling times. VisSim/OptimizePRO can even work with a physical system when coupled with VisSim's hardware analog I/O interface (VisSim/Real-TimePRO).

Download

VisSim Trial