Hello all,As I see, the one of most important thing is a problem to solve a wind direction.It uses for a right plane stabilisation.So, I've some questions about it1. What kind of algorithms do we have (DCM, kalman filter, ??? )2. Which advantages and disadvantages for all those algorithms? (For example adv: simple to solve; disadv: bad accuracy)3. Is it a good idea to try a make hybrid algorithm, that will be use for example average of kalman filter and DCM solution?Becouse as I see we dont have a absolutly right method to solve this problem.
You need to be a member of diydrones to add comments!
Are you asking about gust alleviation, or navigation compensation for (quasi) steady winds? In the latter case, The key variables are going to be your ground speed and track (available by GPS), and your airspeed and heading (Available by pitot and compass, if you have them). I should note that there are ways to find the wind vector with less than this, but you usually have to fly patterns to generate differences in the GPS data. The algorithms involved are really the same as the ones we're already using in state estimation, but differencing the GPS and other sensor data to back out an error term.
Dealing with gusts is a lot more about your sensor and actuator bandwidth, since that generally happens in the low-level stabilization code. Those are usually proportional-integral-derivative (PID) loops with a set point in terms of whatever sensor you're using.
Replies
Are you asking about gust alleviation, or navigation compensation for (quasi) steady winds? In the latter case, The key variables are going to be your ground speed and track (available by GPS), and your airspeed and heading (Available by pitot and compass, if you have them). I should note that there are ways to find the wind vector with less than this, but you usually have to fly patterns to generate differences in the GPS data. The algorithms involved are really the same as the ones we're already using in state estimation, but differencing the GPS and other sensor data to back out an error term.
Dealing with gusts is a lot more about your sensor and actuator bandwidth, since that generally happens in the low-level stabilization code. Those are usually proportional-integral-derivative (PID) loops with a set point in terms of whatever sensor you're using.