Hi Peter,
When generating C code for Simulink models, the principles for storing data in accumulators generally follow these guidelines:
- Fixed Size Inputs: Initially, you generate C code that can accept only inputs of a fixed preassigned size.
- Variable Size Inputs: You can also generate C code that can accept inputs of many different sizes.
- Array Size Assertions: Asserting the maximum size of arrays is a recommended practice for generating efficient C code. For instance, use assert(N<25); to limit the size of an array.
- Passing by Reference: In C, arrays are always passed by reference, which means the original data can be modified by the called function.
To handle the process of generating data into smaller segments in Simulink, you can utilize the "For Each Subsystem" block. This block allows you to partition your data and process each segment individually. Here’s a sample approach to achieve the same:
- Add a "For Each Subsystem" block to your model.
- Configure the "Partition dimension" and "Partition width" to segment the data.
- Inside the "For Each Subsystem", add and connect the processing blocks (e.g., Gain, Sum).
- Optionally, add "Data Store Memory" for data handling.
- Use "Data Store Read" and "Data Store Write" blocks to manage data within the For Each Subsystem.
You may refer to the following MathWorks documentation links to have a better understanding on generating data into smaller segments:
- For Each Subsystem: https://www.mathworks.com/help/releases/R2020b/simulink/slref/foreachsubsystem.html
- Data Store Memory: https://www.mathworks.com/help/releases/R2020b/simulink/slref/datastorememory.html
- Data Store Read: https://www.mathworks.com/help/releases/R2020b/simulink/slref/datastoreread.html
- Data Store Write: https://www.mathworks.com/help/releases/R2020b/simulink/slref/datastorewrite.html
