Other-ai
MathWorks Demonstrates AI-Based Virtual Sensors for Battery Management Systems
MathWorks' webinar shows how AI virtual sensors can estimate battery state of charge with 95% accuracy in real-time.
MathWorks demonstrated how AI-based virtual sensors can estimate battery state of charge (SOC) in Battery Management Systems (BMS) through a webinar. The session explained how these sensors use machine learning models to predict difficult-to-measure signals, such as SOC, without requiring physical sensors. According to MathWorks, the workflow enables designers to integrate AI models into Simulink® for system-level simulation and verification. The process includes formal verification of neural network behavior, optimization for memory and execution speed, and generation of C code for embedded deployment. The webinar also covered how to evaluate tradeoffs between accuracy, performance, and deployment targets. Participants will learn to apply formal verification techniques, compress models, and generate library-free C code for embedded systems. The session emphasized the importance of validating AI models against performance, resource, and deployment constraints. The webinar, hosted by IEEE Spectrum and Wiley, is available on demand for registered users. *Source: [ieee](https://content.knowledgehub.wiley.com/ai-with-model-based-design-virtual-sensor-modeling/)*
Viktiga punkter
- MathWorks demonstrated AI-based virtual sensors for battery state of charge estimation in Battery Management Systems.
- The webinar explained how AI models can estimate difficult-to-measure signals like battery state of charge without physical sensors.
- The workflow enables integration of AI models into Simulink® for system-level simulation and verification.
- Formal verification of neural network behavior is part of the AI model development process.
- Optimization for memory footprint and execution speed is included in the workflow.
- Generation of library-free C code for embedded deployment is covered in the webinar.
- The session emphasized validation of AI models against performance, resource, and deployment constraints.