[fmusim] Feedthrough outputs are calculated one step too late #480
Description
According to the Feedthrough_in.csv file inputs shall be applied at t=1.0.
Therefore, using Co-Simulation mode, I expect the FMU outputs to change accordingly one step later.
However, the outputs only change two steps later:
| "time" | "Float32_continuous_output" | "Float32_discrete_output" | "Float64_continuous_output" | "Float64_discrete_output" | "Int8_output" | "UInt8_output" | "Int16_output" | "UInt16_output" | "Int32_output" | "UInt32_output" | "Int64_output" | "UInt64_output" | "Boolean_output" | "Enumeration_output" |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 0 | 0 | 0 | 0 | -128 | 0 | -32768 | 0 | -2147483648 | 0 | -9223372036854775808 | 0 | 0 | 1 |
| 1.004 | 0 | 0 | 0 | 0 | -128 | 0 | -32768 | 0 | -2147483648 | 0 | -9223372036854775808 | 0 | 0 | 1 |
| 1.008 | 0 | 0 | 0 | 0 | 127 | 255 | 32767 | 65535 | 2147483647 | 4294967295 | 9223372036854775807 | 18446744073709551615 | 0 | 1 |
I ran the following fmusim command:
fmusim.exe --interface-type cs --output-file Feedthrough.csv --input-file Feedthrough_in.csv --output-variable Float32_continuous_output --output-variable Float32_discrete_output --output-variable Float64_continuous_output --output-variable Float64_discrete_output --output-variable Int8_output --output-variable UInt8_output --output-variable Int16_output --output-variable UInt16_output --output-variable Int32_output --output-variable UInt32_output --output-variable Int64_output --output-variable UInt64_output --output-variable Boolean_output --output-variable Enumeration_output Feedthrough.fmu