neerajb wrote:
Shubham, the low refresh rate of physical actuators is obvious. I guess you didn't get my question right. The author wrote that an unstable configuration can't be flown by a human because of slow reaction time of 500 ms. Further he says that aircraft would go out of control in a quarter of a second i.e. 250ms unless a correction in control surfaces (actuators) is applied before that. So for Tejas a refresh rate of atleast 4 cycles per second or 4 Hz is required whereas the actuators are moved every 500ms or at a frequency of 2 Hz.
Cheers....
My explanation would be that(assuming what is said in the book is correct
) , We know that we have a constraint that we can't provide new input to the actuators before 500ms (according to the book), and also the sensor and computers are continuously running their computation and say one iteration of that computation takes 100ms.
So after end of every 100ms I have a new values of various errors and also new input values for the actuators. But I don't give this new input value to the actuators and keep on adding the calculated errors.
At the end of 5th iteration (ie 500ms) I have a value of error which is cumulative sum of the error derived from 5 previous iterations and based on this cumulative value, I find out the corresponding input value for the actuators.
The whole point being that with DFCC, we can hold back the input to the actuators till a certain time and calculate the input corresponding to the cumulative error.
In case of human, it would not work because he can't make sure of the "cumulative" part of error calculation, the moment you give a correcting input, the actual value of correcting input required has changed and this lag will make the flight uncontrollable.