Execution rates
Overview of how execution rates are handled within Typhoon HIL Control Center software
Signal Processing components currently support only discrete time execution rates.
- Forward propagation: sample time is inherited from the components connected to the input terminals. Always performed first.
- Backward propagation: sample time is inherited from the components connected to the output terminals. Performed in cases when the execution rate could not be resolved using the forward propagation method.
Models can contain components with different execution rates. The fastest component execution rate has to be an integer multiple of the base simulation step (step of the FPGA solver). All other execution rates have to be integer multiples of the fastest execution rate.
The maximum number of execution rates is limited and depends on the HIL platform as defined in Table 1.
User CPU | Max execution rates |
---|---|
ARM Cortex A53 | 2 |
ARM Cortex A9 | 4 |
When the model is executed in real time it is required that all execution-rate-related computations are completed within the defined simulation step. In case that the model takes longer to compute, an overrun condition will occur resulting in inaccurate simulation results.
To detect such a situation, the workload on every execution rate is monitored by dedicated hardware logic. Time-slot-utilization information is visualized in the HIL SCADA application. Computing interval overrun conditions are explicitly signaled using the dedicated CIO flag. For more information, please refer to the HIL SCADA Status Bar documentation.