Save and Restore Operating Points for Stateflow Charts

An operating point is a snapshot of a Simulink^® model during simulation. If your model contains a Stateflow^® chart, the operating point includes information about:

• Active states

• Chart output data

• Chart, state, and function local data

• Persistent variables in MATLAB^® functions and truth tables

You can save the state of a model at a certain time step as an operating point and then use that operating point as the initial state for a simulation. For example, suppose that your simulation starts with an initialization phase. If you save an operating point after the model completes the initialization phase, you can use the operating point as the initial state for your model and get the results of a continuous simulation without starting from time t = 0. For more information, see Use Operating Points to Specify Initial State of Simulation.

Save Operating Points

To save the final operating point of a Simulink model, enable the configuration parameters Final states (Simulink) and Save final operating point (Simulink), optionally specify a variable for the operating point for the model, and run the simulation. Then, access the operating point information for the Stateflow chart by calling the get (Simulink) function and specifying the operating point for the model and the block path to the chart.

For example, suppose you want to save an operating point for the model sf_car, which contains the chart shift_logic.

Open the model.

mdl = "sf_car";
openExample(mdl);

Set the model to save the final operating point. Then, run the simulation.

set_param(mdl,"SaveFinalState","on","SaveOperatingPoint","on");
sim(mdl);

If you do not specify a name for the operating point, the default name is xFinal. To access the operating point information for this chart, enter:

op = get(xFinal,"sf_car/shift_logic")

op = 

Block: "shift_logic"    (handle)    (active)
  Path:     sf_car/shift_logic

  Contains:

    + gear_state                     "State (AND)"                        (active)       
    + selection_state                "State (AND)"                        (active)       
      gear                           "State output data"                  gearType [1, 1]
      down_th                        "Local data"                         double [1, 1]  
      up_th                          "Local data"                         double [1, 1]  
      datasetWrittenToVector         "Data set written to vector"         logical 4x1

The operating point lists the states, boxes, functions, and data in the chart in hierarchical order. If name conflicts exist, the operating point appends underscores to the end of these names so that the objects have unique identifiers.

To access the operating point information for an object in the chart, use dot notation. For example, to access the operating point information for the top-level state gear_state, enter:

op.gear_state

ans = 

State: "gear_state"    (handle)    (active)
  Path:     sf_car/shift_logic/gear_state

  Contains:

    + first          "State (OR)"                 
    + fourth         "State (OR)"                 
    + second         "State (OR)"                 
    + third          "State (OR)"         (active)

To access the operating point information for the chart output gear, enter:

op.gear

ans = 

     Description: 'State output data'
        DataType: 'gearType'
            Size: '[1, 1]'
           Range: [1×1 struct]
    InitialValue: []
           Value: third
            Root: [1×1 Stateflow.op.BlockOperatingPoint]
          Parent: [1×1 Stateflow.op.BlockOperatingPoint]

For more information about this example, see Model a Car with Automatic Transmission.

Note

Stateless flow charts have an empty operating point because they do not contain states or persistent data.

Copy Operating Points

To create a copy of the operating point for a Stateflow chart, call the clone function. For example, suppose that you try to copy the operating point op by entering:

op1 = op;
op2 = clone(op);

In this case, op1 refers to the same operating point as op, so modifying op1 also modifies op. In contrast, op2 refers to a different operating point that you can modify without modifying op.

Note

The clone function copies the operating point information for the entire Stateflow chart. You cannot copy the operating point information for a state or data object.

Modify Operating Point Values

To test the response of your Stateflow chart to different settings, you can modify an operating point by changing the values of the local, output, or persistent data, the states that are currently active, and the previous state activity recorded by history junctions during a simulation.

Modify Data Values

To modify the values of local, output, or persistent data in an operating point, change the Value property of the data. For example, to change the down throttle threshold in the operating point op, enter:

op.down_th.Value = 25;

When you modify data values:

• You cannot change the data type or size.

• Refer to elements of a vector or matrix by using one-based indexing delimited by parentheses and commas, even if your chart uses C as the action language.

• New values for numeric data must be within the range that you specify in the Minimum and Maximum parameters. For more information, see Limit range.

• For enumerated data types, you can choose only enumerated values from the type definition.

• You cannot modify the values of persistent data in custom C code or external MATLAB code.

Modify Current State Activity

To change the states that are currently active in an operating point, call the setActive function using a leaf state as an argument. This function maintains state consistency by:

• Exiting and entering the appropriate states

• Resetting temporal counters for newly active states

• Updating values of active state data

• Enabling or disabling function-call subsystems and Simulink functions that bind to states

However, the chart does not perform exit actions for the previously active states or entry actions for the newly active state. Additionally, the state does not reinitialize any state-parented local data. If you want these actions to occur, you must execute them separately. For example, if your state actions assign values to data, you must assign the values explicitly as described in Modify Data Values.

Modify Previous State Activity

To change the previously active state recorded by a history junction, call the setPrevActiveChild function using the state that contains the history junction as an argument. The state must not be active when you call this function.

Load Modified Operating Point Information

To load modified operating point information for a Stateflow chart into the operating point for a Simulink model, call the set (Simulink) function. For example, if xFinal is the operating point for the model sf_car and op contains the modified operating point information for the chart shift_logic, you can save the modified operating point xModified by entering:

xModified = set(xFinal,"sf_car/shift_logic",op);

Note

The set function loads the operating point information for the entire Stateflow chart. You cannot load the operating point information for a state or data object.

Restore Operating Points

To use an operating point as the initial state for a simulation, set the configuration parameter Initial state (Simulink) and specify the variable name for the operating point for the model. When you simulate your model, the simulation starts at the time of the operating point. For more information, see Speed Up Simulation Workflows by Using Model Operating Points (Simulink).

Limitations on Operating Points

Continuous-Time Charts

Operating points for continuous-time charts are read-only. You can save an operating point for a continuous-time chart and use it as the initial state for a simulation. However, you cannot modify the state activity or any data values in the operating point. For more information on continuous-time charts, see Continuous-Time Modeling in Stateflow.

Charts That Use Edge Detection

Stateflow charts that use edge detection operators do not support operating points. If your model contains a chart that uses edge detection, attempting to save the final operating point results in a compile-time error. For more information on edge detection operators, see Detect Changes in Data and Expression Values.