The NASA Helios prototype solar powered UAV used very efficient wide chord propellers that were optimized for high altitude (upwards of 100k feet). It's possible something similar was utilized here.
Helios has 14 engines and not much drag, the Chinese balloon has 4 and a lot of drag. On the other hand, Helios is faster, which increases air resistance.
The wing span is 75m, twice the diameter of the balloon.
I see 48 separate solar panels.
Assuming those panels generate 350w to 600w each, that's roughly 17kw to 24kw of power generation. That's quite a bit of power.
Peak power? In winter, 2/3 of that power is spent charging the batteries during the day. So the average power draw would be 8kW at most; an 8kW engine has 10 horsepower. For comparison, the smallest engine in this B22 Bantam (Gross weight: 430 kg=948 lbs) has 48 kW.
Helios's engine power is 14×1.5kW; if China copied the design, 4×1.5kW seems sustainable by the solar power installed on that balloon.
Besides motors, what else would require that type of power?
if it's using air as ballast, the ballast pump would be a big low-pressure pump.
Passive ELINT receivers generally don't draw a lot of current (transmitters on the other hand...).
I'm not disagreeing with you necessarily, it's just another thing to think about.
How much power would be needed for heating?
As for bringing the solar cells into a better angle to the Sun... Wouldn't rotating the props eat up more power than the slight increase you'd get by adjusting the angle? Wouldn't they just add reserve generation capacity by adding more cells? The additional weight would be supported by a bigger balloon and the "free" lift.
Adjusting the angle horizontally would be mostly frictionless.
Single-axis trackers follow the position of the sun as it moves from east to west. These are usually used in utility-scale projects. A single-axis tracker can increase production between 25% to 35%. [...] Dual-axis trackers can increase energy production by about 40%.
I wonder if the platform is intended to be reusable. The typical "drop it on a parachute" method of returning the equipment back to ground seems a bit rough for a structure that big. If that balloon can land (on a calm day) by ballasting, the engines would certainly help it hit its designated landing spot.