Synchronous AI Simulation for Control Software Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for testing and developing control software with AI components are complex, costly, and lack reproducibility due to the integration of AI components within simulation models, requiring extensive recompilation and halting tests for changes.
Innovation Solution
A method where the AI component is executed synchronously with a simulation model outside the simulation model, allowing independent modification and real-time capability, using a computing environment with a virtual machine and just-in-time compilation of byte code.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the AI component is statically integrated into the simulation model, then the simulation model can be used for hardware-independent software tests, but the entire simulation model must be translated anew into executable machine code when there is a change to the AI component
Solution Approach 1:
The system segments the AI component from the simulation model, allowing the AI component to be independently loaded and executed within the running simulation. This separation enables modification of the AI component without requiring recompilation of the entire simulation model, thus resolving the contradiction between test reproducibility and recompilation time.
Solution Approach 2:
The AI component is extracted from the static simulation model structure and implemented as a separate, dynamically loadable module. This extraction allows the AI component to be changed independently while the simulation model remains intact and executable, eliminating the need for full recompilation when AI components are updated.
2Reliability
If the AI component is statically integrated into the simulation model, then the simulation model can interact with the control software, but the AI component cannot be changed during the test process without stopping the test
Solution Approach 1:
The system implements dynamic loading of AI components during test execution. The AI component can be replaced or updated while the simulation is running, allowing flexibility and adaptability without stopping the test process, while maintaining test reproducibility through controlled loading mechanisms.
Solution Approach 2:
The AI component is designed to support hot-swapping and dynamic replacement during test execution. This allows the useful action of testing to continue uninterrupted while AI components are changed, resolving the contradiction between test continuity and component flexibility.
3Reliability
If hardware-in-the-loop tests are used, then the ECU can be tested with real electrical signals, but the development of an ECU must already be relatively far advanced
Solution Approach 1:
The system introduces a virtual AI component as an intermediary that can be loaded into the running simulation. This intermediary allows early-stage software testing with realistic AI behavior without requiring physical ECU hardware, thus improving development efficiency while maintaining test realism through virtual simulation.
Data Source
AI summary
A computer-implemented method for testing and/or developing control software, having a computing environment, and a simulation model calculated in the computing environment. The simulation model interacts with the control software by outputting input values to the control software and by receiving output values from the control software. The control software includes an AI component, which can be executed synchronously with the simulation model by the computing environment. In addition, the invention relates to a computing environment for data processing comprising a computer system for carrying out the above method, a computer program product, and a corresponding computer-readable data carrier.