Define MPC controller state


xmpc = mpcstate(MPCobj)
xmpc = mpcstate(MPCobj,xp,xd,xn,u,p)
xmpc = mpcstate


xmpc = mpcstate(MPCobj) creates a controller state object compatible with the controller object, MPCobj, in which all fields are set to their default values that are associated with the controller’s nominal operating point.

xmpc = mpcstate(MPCobj,xp,xd,xn,u,p) sets the state fields of the controller state object to specified values. The controller may be an implicit or explicit controller object. Use this controller state object to initialize an MPC controller at a specific state other than the default state.

xmpc = mpcstate returns an mpcstate object in which all fields are empty.

mpcstate objects are updated by mpcmove through the internal state observer based on the extended prediction model. The overall state is updated from the measured output ym(k) by a linear state observer (see State Observer).

Input Arguments


MPC controller, specified as either a traditional MPC controller (mpc) or explicit MPC controller (generateExplicitMPC).


Plant model state estimates, specified as a vector with Nxp elements, where Nxp is the number of states in the plant model.


Disturbance model state estimates, specified as a vector with Nxd elements, where Nxd is the total number of states in the input and output disturbance models. The disturbance model states are ordered such that input disturbance model states are followed by output disturbance model state estimates.


Measurement noise model state estimates, specified as a vector with Nxn elements, where Nxn is the number of states in the measurement noise model.


Values of the manipulated variables during the previous control interval, specified as a vector with Nu elements, where Nu is the number of manipulated variables.


Covariance matrix for the state estimates, specified as an N-by-N matrix, where N is the sum of Nxp, Nxd and Nxn).

Output Arguments


MPC state object, containing the following properties.




Vector of state estimates for the controller’s plant model. Values are in engineering units and are absolute, i.e., they include state offsets.

If the controller’s plant model includes delays, the Plant field of the MPC state object includes states that model the delays. Therefore length(Plant) > order of undelayed controller plant model.

Default: controller’s Model.Nominal.X property.


Vector of unmeasured disturbance model state estimates. This comprises the states of the input disturbance model followed by the states of the output disturbances model.

Disturbance models may be created by default. Use the getindistand getoutdistcommands to view the two disturbance model structures.

Default: zero, or empty if there are no disturbance model states.


Vector of output measurement noise model state estimates.

Default: zero, or empty if there are no noise model states.


Vector of manipulated variables used in the previous control interval, u(k–1). Values are absolute, i.e., they include manipulated variable offsets.

Default: nominal values of the manipulated variables.


n-by-n symmetrical covariance matrix for the controller state estimates, where n is the dimension of the extended controller state, i.e., the sum of the number states contained in the Plant, Disturbance, and Noise fields.

Default: If the controller is employing default state estimation the default is the steady-state covariance computed according to the assumptions in Controller State Estimation. See also the description of the P matrix in the Control System Toolbox kalmd command. If the controller is employing custom state estimation, this field is empty (not used).


collapse all

Create a Model Predictive Controller for a single-input-single-output (SISO) plant. For this example, the plant includes an input delay of 0.4 time units, and the control interval to 0.2 time units.

H = tf(1,[10 1],'InputDelay',0.4);
MPCobj = mpc(H,0.2);
-->The "PredictionHorizon" property of "mpc" object is empty. Trying PredictionHorizon = 10.
-->The "ControlHorizon" property of the "mpc" object is empty. Assuming 2.
-->The "Weights.ManipulatedVariables" property of "mpc" object is empty. Assuming default 0.00000.
-->The "Weights.ManipulatedVariablesRate" property of "mpc" object is empty. Assuming default 0.10000.
-->The "Weights.OutputVariables" property of "mpc" object is empty. Assuming default 1.00000.

Create the corresponding controller state object in which all states are at their default values.

xMPC = mpcstate(MPCobj)
-->Converting the "Model.Plant" property of "mpc" object to state-space.
-->Converting model to discrete time.
-->Converting delays to states.
-->Assuming output disturbance added to measured output channel #1 is integrated white noise.
-->The "Model.Noise" property of the "mpc" object is empty. Assuming white noise on each measured output channel.
MPCSTATE object with fields
          Plant: [0 0 0]
    Disturbance: 0
          Noise: [1x0 double]
       LastMove: 0
     Covariance: [4x4 double]

The plant model, H, is a first-order, continuous-time transfer function. The Plant property of the mpcstate object contains two additional states to model the two intervals of delay. Also, by default the controller contains a first-order output disturbance model (an integrator) and an empty measured output noise model.

View the default covariance matrix.

ans = 4×4

    0.0624    0.0000    0.0000   -0.0224
    0.0000    1.0000   -0.0000   -0.0000
    0.0000   -0.0000    1.0000    0.0000
   -0.0224   -0.0000    0.0000    0.2301

Introduced before R2006a