3aaa38fd-5fad-455c-9e3b-92a3a2e1048e 63 themzlab 4 5 decimal

&[DATE] &[TIME] - &[FILENAME]

The loop speed and desired settling time determine how fast to run the filter KI_settling_time 7 s * : pid_loop_speed 20 Hz * : 21 sys

t0 1 time :

x floor x .5 - 0 round :

loops 200 :

KI_settling_time pid_loop_speed * 3 / floor 46

This should be declared using the 'floor' formula above but that caused a slowdown KI.FILTER_LENGTH KI_settling_time pid_loop_speed * 3 / floor eval :

command k 120 range_command k el 100 : for k 21 loops range_command k el 500 : for k 21 loops range_command k el_command k el k .2 * sin 10 * + eval : for_command 41 line :

<--- 2nd loop not really needed, can be merged in the 3rd

command command :

<--- evaluate procedure defined above

This loop shows the initialization of the filter
and then the operation of it.  I think it is the ROOT
cause for slow calculations filter_results SUM 1 el command 1 el KI.FILTER_LENGTH * : temp command 1 el KI.FILTER_LENGTH * : previous command 1 el : k 2 loops range SUM k el command k el temp + previous - eval : filter_out k el SUM k el KI.FILTER_LENGTH / eval : temp SUM k el eval : previous filter_out k el eval : 41 line for filter_out 1 el command 1 el : filter_out 61 line :

filter_results filter_results :

<--- evaluate procedure defined above

idx 1 command rows range :

idx command augment idx filter_results augment 2 1 sys

1 time t0 - 0.9 s *

k 1 loops range x_axis k el k pid_loop_speed 1 * s * / : for

x_axis length 200

Here is where the slow calculation starts filter_results length 200

I'd like to add this to the plot, except it becomes SLOW I_term x_axis filter_results augment :

Simulated_PID_result x_axis command augment :

This plot calculates QUICKLY

Simulated_PID_result

This is a STEP response, it calculates SLOWLY

Simulated_PID_result I_term 2 1 sys