amardeep mishra
FULL MEMBER
- Joined
- Mar 8, 2012
- Messages
- 1,323
- Reaction score
- 26
- Country
- Location
Hi @The DeterrentUm no thats not the case. Pakistan is advanced enough in the INS domain now to ensure strategic precision in a GPS-denied environment, and surgical precision with integrated GNSS corrections.
Not much is known about "indigenous" pakistani research in the field of INS(I'm not talking about the algorithms but the hardware). Can you shed some light?
Also as I guess @JamD pointed out, filtering techniques might play some but not major role in increasing the range. In my opinion though, range can be extended if either somehow drag is reduced, or if they somehow get a highly efficient turbofan engine or if they go for all composite missile. Drag reduction might seem easy but it requires passing your model through wind tunnels a lot and wind tunnel experiments are really pain in the **** sometimes(at least in india)--I am sure such experiments in pakistan wouldnt be any comfortable. Second option requires pakistan to actually get a turbofan from china(which in turn used ukrainian expertise to design their own).
Speaking of INS and filters,I have personally implemented -
(a) Complementary filer-1st order
(b) Complimentary filter-2nd order and
(c) Kalman Filter
in python for simple MPU6050 as a hobby stuff. Here is the a real time graph plotted in python.
Although this is not from MPU6050 but from ADXL335. ADXL335 is notoriously sensitive to even small vibrations as you can see above(kindly note this is the result without any filtering). Some sort of high pass filter needed in order to smoothen the ADXL335 signal.
Now the reason why we need any kind of filtering is quite evident from my personal graph above-- accelerometer is jerky(yet doesnt drift),whereas gyro(rate gyro to be precise) is stable(yet it drifts over time). So in order to get a stable estimate we implement a filter-- one that combines the features of both HPF and LPF and the easiest, one could implement in either C or python happens to be a complementary filter. It gives fairly reasonable values in most of the regimes. For more info you can follow me on github
Last edited: