sim
Simulate a jMPC MPC Controller in a given jSIM Simulation Environment.
Syntax
jSIMobj = sim(jMPCobj,jSIMobj)
jSIMobj = sim(jMPCobj,jSIMobj,mode)
Description
jSIMobj = sim(jMPCobj,jSIMobj,mode) applies the MPC Controller in jMPCobj to the Simulation Environment in jSIMobj, using the given mode. There are currently four modes available for simulation:
'MEX' | Default High speed C implementation of the jMPC engine and QP solver. Useful for multiple MPC simulation studies. |
'Matlab' | Simulation mode uses an m-file implementation of the jMPC engine. Best for debugging and algorithm development. |
'Simulink' | Uses the Discrete MPC C S Function within Simulink as the jMPC engine. |
'MPCToolbox' | Converts the jMPC object to a MATLAB MPC Toolbox object and simulates it using the MPC Toolbox®. Useful for comparison studies of jMPC results against a commercial product. |
From all simulation modes the returned jSIM object will be tagged as a results object, and will contain the simulation results.
Functional Overview
The function performs 6 steps at every sample of the simulation:
| Step | Operation | Description |
| 1 | State Estimator Update | The current model states are updated with the state estimation gain, Kest. This infers a current step rather than a predictor step state estimator. |
| 2 | Solve MPC Control Law | As described in mpcsolve, the MPC Control law is computed here and u solved for. |
| 3 | Saturate Inputs | The calculated rate of change of input, and absolute input are saturated to the limits of the constraints (if they are not already within these limits). |
| 4 | Simulate Model | The model is simulated over one time step. |
| 5 | Apply Input to Plant | The input is applied to the plant for one time step. |
| 6 | Save Sample Results & Repeat | The results of the above are saved and the simulation repeats until specified. |