I'm not talking about the entire auto pilot system, but only the thrust. Because I think, when experiencing crosswind during landing, the speed of wind may fluctuate and on 3 digit approach altitudes and below, it is hard for pilots to balance the forces with counter-forces without missing the runway. Is auto thrust precise enough to balance those forces in such short distances and short amount of time, and keep the plane aligned with runway after pilot points the plane against the wind?
EDIT: I take your question as "when crabbing the aircraft for landing in strong crosswind, is the auto-thrust precise enough to maintain horizontal alignment with the extended runway centerline?"
No. Auto-thrust is not precise nor responsive enough for this capability. Furthermore, auto-thrust is not used in this manner. To track the extended runway centerline when the crosswind changes, the pilots would increase or decrease the crab angle using rudder inputs.
Let us for a moment consider using auto-thrust to track the centerline. As an example, imagine an aircraft landing on runway 36 (pointed North) with a strong crosswind blowing East to West. The plane's heading will be somewhere between 360 and 90 degrees, while its track is 360. Suddenly, the crosswind increases, and the plane experiences a deviation to its left.
The plane is drifting to its left because the velocity component in the East-West direction is nonzero. If this is countered by increasing engine power, the plane can, theoretically, increase its velocity component pointing towards the East and get back on track. The problem is, flying a plane is three-dimensional! If engine power is increased, airspeed will increase, lift will increase, and the plane will drift above the glide slope vertically.
The correct response is this case is to input right rudder. Staying on localizer and glide slope requires frequent, precise but small changes. The right tool for that is the control surfaces. When using the crab technique in a crosswind landing:
From your question, it appears that you may have some misconceptions.
Thrust is forward. It controls the airspeed, which is also related to the angle of attack. Crosswind is wind blowing either left-to-right or right-to-left. Therefore, thrust and crosswind are unrelated; auto-thrust will not compensate for crosswind because thrust can only make the plane go faster/slower or go up/down, but crosswind blows side to side.
If your question is whether the autopilot uses asymmetrical thrust to control the plane in these conditions, then the answer is no: thrust is not responsive enough to control a plane's attitude, asymmetrical thrust is only used in emergency situations when no other options are available.
My bet is you're asking about gusts and wind shear, which can be experienced as either a headwind, tailwind or crosswind, so it may affect an aircraft's airspeed. Wind shear is defined as a sudden change of wind speed and/or wind direction.
The calculated speed of a landing is called Vref. In a normal landing, the pilots usually input Vref+5 into the autopilot. In gusty conditions, the pilots may input Vref+10 into the autopilot. A less flap setting may also be selected, again for the purpose of flying the approach at a higher airspeed. A higher airspeed is desired because it increases the stall margin.
When using the autothrottle, position command speed to VREF + 5 knots. Sufficient wind and gust protection is available with the autothrottle connected because the autothrottle is designed to adjust thrust rapidly when the airspeed drops below command speed while reducing thrust slowly when the airspeed exceeds command speed. In turbulence, the result is that average thrust is higher than necessary to maintain command speed. This results in an average speed exceeding command speed.
If a manual landing is planned with the autothrottle connected in gusty or high wind conditions, consider positioning the command speed to VREF + 10 knots. This helps protect against a sudden loss of airspeed during the flare.
quote from Boeing 777 Flight Crew Training Manual
Auto-thrust itself is precise enough to handle gusty winds to an extent, but it has its limitations. IIRC using auto-thrust is recommended in these scenarios, as it reduces the pilots' workload in a challenging situation. Plane manufacturers and airline companies often have limitations regarding the maximum gust / maximum crosswind. Attempting a landing outside these limitations may exceed the auto-thrust's capability to respond to rapid airspeed changes; it may also exceed the landing gear's strength or the wing's loading and therefore is dangerous.
From your question, it seems that you have the following situation:
An aircraft is having a constant angle w.r.t. the runway to compensate for a crosswind. When this crosswind increases, the aircraft will as a result drift off the centerline of the runway. To compensate this, you imagine the aircraft speeding up to 'catch' the centerline.
Note that I'm not talking about 'forces' - it is helpful to think in velocities only when thinking about crosswinds, and assume the aircraft responds instantly so it only has a forward velocity relative to the wind.
Now, while this is in theory a way to deal with crosswinds, it's far from ideal. Much better is to simply change your crabbing angle, so that the sideways velocity component (w.r.t. the runway centerline) matches the crosswind. This is much quicker, although it does mean banking is required. You will see that, when the changes in crosswind become too sudden on short final, a go around will be necessary to avoid excessive banking near the ground. Airspeed is kept constant with auto-thrust, not 'crabbing speed'.
From a automation perspective, your suggested way of dealing with crosswinds is entirely possible - an aircraft is capable to detect its lateral and angular deviation from the centerline, and could calculate thrust accordingly. However, physically, a jet engine needs time to spool up to produce the thrust required, and spool down before less thrust can be produced (there is a lot of inertia, and the combustion chamber does not have very relaxed margins as to suddenly injecting extra or removing all fuel).
Furthermore, the airspeed will fluctuate, requiring different angle of attack continuously, and possibly excess velocity on touchdown. Note that when the initial crabbing angle is very small (say, a 1kt crosswind), you would require ludicrous amounts of speed to account for a small increase (the extra speed required varies inversely proportional with the crabbing angle). Conversely, when the initial angle is quite large, you would need to stop in mid-air when the crosswind dies down. All in all, a simple bank to change the crabbing angle is a much more effective way of dealing with a crosswind.
The question was not whether the autothrust/autothrottle DOES keep the aircraft aligned, but whether or not it CAN keep aircraft aligned. There is a very interesting, but not well known story that took place after United Airlines Flight 232 crashed in Sioux City in 1989.
After having lost all hydraulic power on a DC-10 aircraft due to a failed disc on the #2 engine which severed all the hydraulic lines, Captain Al Haynes, First Officer William R. Records, Second Officer Dudley J. Dvorak, and some random check airman who happened to be on board to offer his assistance, Dennis E. Fitch, managed to cohesively work together to guide the aircraft to Sioux City. Although the plane crashed, they exclusively used the remaining engines (engines 1 and 3) to get the aircraft there. Although many lives were lost, many were also saved.
Following that, an engineer at NASA came up with the idea to implement a sort of reversionary mode of the autopilot/autothrottle systems that could safely perform an auto land with all hydraulic power lost. Although never implemented, it was successfully tested on an F-16 and then on a passenger version of an MD-11 and B747. The story, an excellent one to read, is chronicled here. Although it includes many technical details, even further technical information can be found here.
This is a video of the F-15 demonstration. There is actually a video of the MD-11 performing a landing, but I can't seem to find it.
So, to answer your question. Is the auto throttle capable of everything you asked? Absolutely! Is it installed/implemented on aircraft? No.
