System and method for simulating branching behavior

• US 8,805,664 B1
• Filed: 10/01/2010
• Issued: 08/12/2014
• Est. Priority Date: 08/11/2003
• Status: Active Grant

• Current Assignee: The MathWorks, Inc.
• Original Assignee: The MathWorks, Inc.
• Inventors: Han, Zhi, Yeddanapudi, Murali, Mosterman, Pieter J., Lin, Xiaocang, Sunkari, Rajesh Pavan
• Primary Examiner(s): Rivas, Omar Fernandez
• Assistant Examiner(s): MOLL, NITHYA JANAKIRAMAN
• Application Number:
  US12/896,407
• Time in Patent Office: 1,411 Days
• Field of Search: None
• US Class Current: 703/13
• CPC Class Codes: G06F 2111/10 Numerical modelling
• CPC Class Codes: G06F 2111/12 Symbolic schematics
• CPC Class Codes: G06F 30/20 Design optimisation, verification or simulation optimisation, verification or simulation of circuit designs G06F30/30

Abstract

In an embodiment, a method of establishing directed relationships between states in a simulation is disclosed. The directed relationships may allow the simulation to proceed from an initial state according to two or more divergent behaviors. The simulation may merge if two or more behaviors result in states that are equivalent. The method may further allow a state of the simulation which has not been stored to be interpolated from one or more states that have been stored. In one embodiment, a system may receive a request to revert to a previous state, and the system may identify a saved state that is closest to the requested state. The system may simulate from the identified state to arrive at the requested state. In one embodiment, the simulation may be a hybrid simulation which is advanced in both discrete and continuous increments.

Claims

1. A method comprising:

capturing, using a processor of a computing device, a first simulation state during a simulation of a model;

identifying that continued simulation of the model from the first simulation state may follow a first pattern of behavior and a second pattern of behavior;

simulating the model from the first simulation state according to the first pattern of behavior to arrive at a second simulation state;

simulating the model from the first simulation state according to the second pattern of behavior to arrive at a third simulation state;

establishing a first directed relationship between the first simulation state and the second simulation state, the first directed relationship representing;

a change of a state of the model from the first simulation state to the second simulation state, or

a change of the state of the model from the second simulation state to the first simulation state;

establishing a second directed relationship between the first simulation state and the third simulation state, the second directed relationship representing;

a change of the state of the model from the first simulation state to the third simulation state, or

a change of the state of the model from the third simulation state to the first simulation state; and

storing information allowing at least the first simulation state to be restored.

2. The method of claim 1, wherein the model is simulated in continuous time and the first directed relationship and the second directed relationship are temporal relationships.

3. The method of claim 1, wherein the identifying further comprises:

receiving an input specifying that the model is to be simulated in at least two different ways.

4. The method of claim 3, wherein the input specifies a simulation time from which the model is to be simulated in the at least two different ways.

5. The method of claim 3, wherein the input specifies one or more conditions which, if met, cause the model to be simulated in the at least two different ways.

6. The method of claim 1, wherein the simulating of the model from the first simulation state according to the first pattern of behavior and the simulating of the model from the first simulation state according to the second pattern of behavior are performed in parallel.

7. The method of claim 1, wherein the simulating of the model from the first simulation state according to the first pattern of behavior to arrive at the second simulation state of the model is performed by a first processor and the simulating of the model from the first state according to the second pattern of behavior to arrive at the third simulation state of the model is performed by a second processor.

8. The method of claim 1, wherein the simulating of the model according to the first pattern of behavior and the simulating of the model according to the second pattern of behavior result in two or more outcomes for the simulation, and the simulating further comprising:

displaying a graph of the two or more outcomes.

9. The method of claim 1, further comprising:

simulating the model according to the first pattern of behavior to arrive at a fourth simulation state, and simulating the model according to the second pattern of behavior to arrive at a fifth simulation state; and

identifying that the fourth simulation state and the fifth simulation state are equivalent.

10. The method of claim 9, wherein the fourth simulation state and the fifth simulation state are identified as being equivalent based on at least an outcome of performing a differencing between the fourth simulation state and the fifth simulation state.

11. The method of claim 9, wherein the fourth simulation state and the fifth simulation state are identified as being equivalent if any outcome of simulating the fourth simulation state is equivalent to an outcome of simulating the fifth simulation state.

12. The method of claim 9, wherein the fourth simulation state and the fifth simulation state each comprise a value of an attribute associated with the model, and wherein the fourth simulation state and the fifth simulation state are identified as equivalent if the value of the attribute in the fourth simulation state is equal to, or within a predetermined tolerance of, the value of the attribute in the fifth simulation state.

13. The method of claim 12, wherein the identifying that the fourth simulation state and the fifth simulation state are equivalent further comprises:

generating a checksum for the fourth simulation state;

generating a checksum for the fifth simulation state; and

identifying that the fourth simulation state and the fifth simulation state are equivalent if the checksum for the fourth simulation state is equal to the checksum of the fifth simulation state.

14. The method of claim 12, further comprising merging the fourth simulation state and the fifth simulation state when the fourth simulation state and the fifth simulation state are equivalent.

15. The method of claim 1, further comprising:

presenting a user with a choice of simulating from the second simulation state or simulating from the third simulation state;

receiving a selection of the second simulation state or the third simulation state; and

simulating the system from the selected state.

16. The method of claim 1, wherein simulating according to the first pattern of behavior further comprises:

reaching a predetermined acceptance point; and

terminating the simulation according to the second pattern of behavior before the simulation according to the second pattern of behavior is complete.

17. A method comprising:

capturing, using a processor of a computing device, a first simulation state during a simulation of a model;

capturing a second simulation state during the simulation;

establishing a directed relationship between the first simulation state and the second simulation state, the directed relationship describing;

a change of the first simulation state into the second simulation state, or a change of the second simulation state into the first simulation state;

storing at least a portion of the first simulation state, at least a portion of the second simulation state, and information based on the directed relationship; and

deriving a third simulation state of the model, the deriving comprising;

using the directed relationship to determine the third simulation state, wherein;

the third simulation state represents a state of the model that occurs between the first simulation state and the second simulation state.

18. The method of claim 17, wherein the model is simulated in continuous time and the directed relationship is a temporal relationship.

19. The method of claim 17, further comprising:

identifying a portion of the first simulation state that is redundant with a portion of the second simulation state, wherein the storing comprises storing the redundant portion only once.

20. The method of claim 17, wherein the storing at least a portion of the first simulation state is performed after a triggering condition is met.

21. The method of claim 17, wherein the simulation is a hybrid simulation which is advanced in both discrete and continuous increments.

22. The method of claim 21, wherein the first simulation state and the second simulation state are non-consecutive states.

23. The method of claim 22, further comprising:

storing a plurality of simulation states;

receiving a request to restore the model to a selected simulation state, the selected simulation state being different from the stored plurality of simulation states;

identifying at least one simulation state in the plurality of simulation states that approximates the selected simulation state; and

simulating from the identified simulation state to arrive at the selected simulation state.

24. The method of claim 23, wherein the plurality of simulation states are each associated with a simulation time and the selected simulation state is identified by the selected simulation state'"'"'s simulation time, and wherein identifying one of the plurality of simulation states that approximates the selected simulation state comprises:

identifying a simulation state among the plurality of simulation states whose simulation time is nearest to the simulation time of the selected state.

25. A non-transitory electronic device readable storage medium storing instructions, the instructions comprising:

instructions that, when executed by a processor, cause the processor to;

capture a first simulation state during a simulation of a model;

identify that continued simulation of the model from the first simulation state may follow a first pattern of behavior and a second pattern of behavior;

simulate the model from the first simulation state according to the first pattern of behavior to arrive at a second simulation state of the model;

simulate the model from the first simulation state according to the second pattern of behavior to arrive at a third simulation state of the model;

establish a first directed relationship between the first simulation state and the second simulation state, the directed relationship representing a change of a state of the model from the first simulation state to the second simulation state or a change of the state of the model from the second simulation state to the first simulation state;

establish a second directed relationship between the first simulation state and the third simulation state, the directed relationship representing a change of the state of the model from the first simulation state to the third simulation state or a change of the state of the model from the third simulation state to the first simulation state; and

store information allowing at least the first simulation state to be restored.

26. A system comprising:

a storage for storing a model of a system; and

a processor for;

capturing a first simulation state during a simulation of the model;

identifying that continued simulation of the model from the first simulation state may follow a first pattern of behavior and a second pattern of behavior;

simulating the model from the first simulation state according to the first pattern of behavior to arrive at a second simulation state of the model;

simulating the model from the first simulation state according to the second pattern of behavior to arrive at a third simulation state of the model;

establishing a first directed relationship between the first simulation state and the second simulation state, the directed relationship representing a change of a state of the model from the first simulation state to the second simulation state or a change of the state of the model from the second simulation state to the first simulation state;

establishing a second directed relationship between the first simulation state of the model and the third simulation state of the model, the directed relationship representing a change of the state of the model from the first simulation state to the third simulation state or a change of the state of the model from the third simulation state to the first simulation state; and

storing information allowing at least the first simulation state to be restored.