Continuous Time Mode-Driven Simulations in Graphical Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional state diagramming environments in technical computing environments are limited in modeling systems with arbitrary or ill-conditioned control flow and continuous time systems, failing to accurately simulate mode changes and transitions.
Innovation Solution
A modeling environment that supports continuous time state-based systems by using state diagram models to accurately model mode changes and continuous states, incorporating tools like Stateflow and Simulink to generate and execute state diagram models, and facilitating recognition and handling of state diagram mode transitions within continuous time simulation models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional state diagramming environments are used, then well-formed control flow systems can be modeled, but arbitrary or ill-conditioned control flow and continuous time systems cannot be supported
Solution Approach 1:
The patent implements dynamic sample time adjustment in the simulation engine, allowing the system to adapt between discrete and continuous time processing based on the current state. The sample time is dynamically modified based on the active state's configuration, enabling accurate simulation of continuous time systems while maintaining efficiency for discrete events.
Solution Approach 2:
The patent changes the time parameter from fixed discrete intervals to variable continuous intervals. Each state can be configured with specific sample time parameters, allowing the simulation to transition between discrete and continuous time modes. This parameter flexibility enables support for arbitrary control flow and continuous time systems simultaneously.
2Productivity
If discrete time simulation is used, then computational efficiency is maintained, but accuracy of continuous state simulation deteriorates
Solution Approach 1:
The simulation engine dynamically adjusts the sample time parameter based on which state is currently active. When a state requiring continuous simulation is active, the sample time is reduced to maintain accuracy. When discrete states are active, larger sample times are used to maintain efficiency. This dynamic adaptation resolves the contradiction between speed and accuracy.
Solution Approach 2:
Instead of using uniformly small time steps throughout the entire simulation (which would ensure accuracy but reduce efficiency), the patent applies fine-grained time steps only when and where continuous state accuracy is needed. This partial application of excessive precision maintains accuracy for continuous states while preserving overall simulation efficiency.
3Stability of the object's composition
If fixed sample times are used, then simulation stability is maintained, but ability to capture fast transient behavior is lost
Solution Approach 1:
The patent implements dynamic sample time adjustment where the simulation engine modifies the time step size based on the current state's requirements and detected transient behavior. When fast transients are detected or when entering states requiring high fidelity, the sample time is automatically reduced. This maintains simulation stability through adaptive control rather than fixed parameters.
4Measurement precision
If detailed state modeling is implemented, then accuracy of mode change representation improves, but system complexity increases
Solution Approach 1:
The patent segments the modeling functionality into distinct components: state definition, transition conditions, sample time configuration, and continuous state variables. Each aspect is handled separately through dedicated interfaces and data structures. This segmentation allows detailed modeling capability while maintaining system manageability through modular organization.
Solution Approach 2:
The patent introduces an intermediary simulation engine that handles the complexity of continuous time processing, sample time management, and mode transition detection. This intermediary layer shields the user from implementation complexity while providing accurate mode change representation. The engine mediates between the state model and the simulation execution, managing the detailed state modeling internally.
Data Source
AI summary
A system, computer-readable medium, and method includes providing a control flow graph including at least one condition. An update control flow graph is automatically generated based on the provided control flow graph, where execution of the update control flow graph includes an evaluation of the condition. The update control flow graph and a value corresponding to the evaluated condition is stored. An output control flow graph based on the provided control flow graph is automatically generated, where execution of the output control flow graph is based on the stored value corresponding to the evaluated condition. A mode transition control flow graph configured to identify mode transitions for the control flow graph at one or more time intervals is automatically generated and the update control flow graph is re-executed based on the generated mode transition control flow graph.