Rebirth: Starting from Lighting Up the Tech Tree

Chapter 158 Technical Pre-research

In the second week after the space photovoltaic project was approved, the third floor of the research institute was brightly lit.

Qiu Pei has not left the lab before 2 a.m. for four consecutive days. There are three monitors on his workstation: the left one is running orbit simulation, the middle one is the parameter tuning interface for the beamforming algorithm, and the right one is a terminal window that monitors the output of Weijia's atmospheric correction model in real time.

The problem is the beamforming algorithm.

Traditional microwave transmission schemes use a single antenna array with a fixed beam direction, and the transmission efficiency fluctuates greatly due to atmospheric influence. The phased array scheme proposed by Yu Ying at the strategic meeting requires hundreds of antenna elements to work together, and the phase and amplitude of each element must be precisely controlled, resulting in an exponential increase in computational load.

Qiu Pei tried seven different algorithm architectures, but the first six either lacked accuracy or had too high computational latency.

The seventh approach is a hybrid architecture he designed himself, combining global optimization and local adaptation. It's theoretically feasible, but he needs a high-precision simulation environment to verify it.

At one o'clock in the morning, Yu Ying pushed open the laboratory door.

She was wearing a white sweatshirt, her hair tied in a ponytail, and holding two cups of coffee. This was the seventh time she had appeared in the lab late at night since joining the company.

"Drink." She placed a cup of coffee on Qiu Pei's desk.

Qiu Pei took the cup and rubbed his eyes. "Dean Yu, the seventh architecture has produced preliminary results."

Yu Ying pulled up a chair and sat down, staring at the data on the screen.

"How's the delay?"

"The single-shot beamforming time has been reduced from twelve milliseconds to three and a half milliseconds, but accuracy is still an issue. The pointing deviation under complex atmospheric conditions reached 0.3 degrees, far exceeding the design requirement of 0.05 degrees."

Yu Ying remained silent for a few seconds.

What are the sources of this bias?

"The main issue is that the input data for the atmospheric correction model isn't refined enough. Wei Jia's database covers 92% of the geographical area, but its temporal resolution is insufficient. Her data is updated hourly, while beamforming requires minute-level real-time corrections."

Yu Ying stood up and walked to the whiteboard. She drew a flowchart with a marker.

"The problem isn't with the algorithm itself, it's with the input. Qiu Pei, your hybrid architecture direction is correct, but it needs better atmospheric data to drive it. We'll use AI to compensate for the time resolution issue on Weijia's end."

She picked up her phone and sent a message to Shen Yiming.

Three minutes later, Shen Yiming replied. He was working on something in the AI ​​division: transferring the multi-sensor data fusion technology that 402 had accumulated in the drone field to atmospheric prediction. Drones need to process environmental data such as wind speed, temperature, and humidity in real time during flight. This algorithm, after two years of practical refinement, has been validated in 80,000 flights.

After reading the reply, Yu Ying handed the phone to Qiu Pei.

"Shen Yiming has an environmental perception fusion algorithm that can improve the temporal resolution of multi-source sensor data from the hourly level to the minute level. If we combine it with Weijia's atmospheric database, theoretically it could solve your input accuracy problem."

Qiu Pei stared at the screen for a long time, then slowly nodded.

"The handover will begin tomorrow."

The next morning, Wei Jia brought her three laptops to the AI ​​division. Shen Yiming assigned two engineers to assist her in encapsulating the core code of the environmental perception fusion algorithm into an interface. The integration process took a full day.

That evening, Qiu Pei fed the merged data into his hybrid architecture.

The result was unexpected by everyone.

The beam pointing deviation has been reduced from 0.3 degrees to 0.08 degrees. Although it has not yet reached the design requirement of 0.05 degrees, it is very close.

More importantly, transmission efficiency has been significantly improved. Simulation results under standard atmospheric conditions show that the microwave transmission efficiency reaches 79.3%.

We are only 0.7 percentage points away from the target of 80%.

After reviewing the data, Yu Ying neither praised it nor made any requests. She simply said, "Let's continue tomorrow."

Three more days passed.

Qiu Pei successfully tested the eighth version of the beamforming algorithm. This time, he completely reconstructed the global optimization module, deeply embedding Shen Yiming's environment fusion algorithm into the real-time beam pointing correction link.

At four o'clock in the morning, the simulation results came out.

Microwave transmission efficiency: 80.2 percent.

Beam pointing deviation: 0.04 degrees.

Qiu Pei stared at the numbers on the screen for a full ten seconds. Then he picked up his phone and sent a message to Yu Ying.

Yu Ying replied four minutes later: "Apply for a patent and prepare the technical report. Report to Zuo Cheng at nine o'clock tomorrow morning."

The next morning at nine o'clock, when Zuo Cheng entered the research institute's conference room, he saw that the whiteboard was covered with formulas and arrows. Yu Ying, Li Guodong, Qiu Pei, Wei Jia, and Shen Yiming were already waiting for him.

Qiu Pei finished explaining the three weeks' worth of work in ten minutes.

Zuo Cheng listened without speaking immediately. He walked to the whiteboard and glanced at the number "80.2 percent".

How big is the gap between this result and that of the United States?

Qiu Pei replied, "The US ground simulation experiment in 2019 achieved an efficiency of 68%. Our results are simulation verification; we haven't conducted any ground simulations yet. However, based on the theoretical model, it's at least 12 percentage points ahead of the current international level."

Zuo Cheng nodded.

"Fang Ze is already following up on the patent matter. Qiu Pei, prepare a complete technical report for the NDRC meeting next week."

Li Guodong spoke first. He was the quietest person throughout the entire process, but every time he spoke, it carried weight.

"An 80.2% simulation efficiency is good, but don't forget that this is only under ideal conditions. The real space environment is far more complex than simulations; extreme temperature fluctuations, cosmic ray interference, and micrometeorite collisions all affect actual performance. My suggestion is to initiate ground-based simulation experiments as soon as possible while advancing the technical solutions. Before actually launching into space, a complete microwave transmission link must be validated on the ground."

Zuo Cheng glanced at Li Guodong and then nodded.

"You'll lead the ground simulation experiment. Coordinate the venue and equipment with Han Lu. Approve an initial budget of 20 million; add more if needed."

Li Guodong grunted in response and said nothing more.

He paused for a moment, his gaze sweeping over everyone present.

"This is our first major battle in space photovoltaics. We must win it."

After the meeting, Zuo Cheng stayed behind alone. He stood in front of the whiteboard, looking at the densely packed formulas. He didn't need to derive or calculate these things. All he could do was ensure that these people had sufficient resources, sufficient time, and sufficient space.

He opened his phone and glanced at the system panel. The progress bar for the Space Solar Power branch of the tech tree lit up slightly, as if in response to something.

Points: 397. No change. But Zuo Cheng knew that some things were more important than points.