Genius of the Rules-Style System

Chapter 924: Chapter 472 It’s okay to not understand anything!_1

Liu Hemin steered the conversation towards the Intelligent Control System, and Zhao Yi began to inquire about it, aiming to understand the areas they were struggling with.

Quickly.

They proceeded to the office together.

Liu Hemin, Guo Jinhua, and Deng Minrong highlighted several algorithm packages that were difficult to comprehend and discussed some overall issues with the algorithms.

Zhao Yi nearly grasped the situation.

Their concerns were mainly related to the core of intelligent control, particularly in the part that involved adapting to the pilot’s intent. Adapting to the pilot’s intent required not only data from the control room sensors but also a reference to the current flight parameters and environment, and it necessitated the coordination of various component controls. This included not only the variable externals but also the control of core internal components such as the engine.

The reason this part was not easy to understand was largely due to the coordination control algorithms; they involved calculations related to intelligent control —calculations that went beyond adjusting based on parameters and included the intelligent monitoring aspect, which is using algorithms to make intelligence more accurate.

Intelligent monitoring is about assessing the accuracy of intent judgement, as success is not 100% guaranteed and must be evaluated through computation.

Two intelligent monitoring detection parameters were introduced, representing indicators of intelligent monitoring, each derived from different computational methods.

Before adjusting the variable parts according to the pilot’s intent, the Intelligent Control System calculates the intelligent monitoring detection parameters using all available parameters. It then compares the two parameters, and adjustments to the variable parts are made only if the probability of a correct judgement exceeds 98%.

If the comparison of the two parameters falls between 80% to 98%, the speed of adjustment to the variable parts is reduced, allowing the pilot some reaction time.

If the comparison result is less than 80%, no adjustments are made. Instead, a confirmation message is sent to the pilot, and the changes are executed only after the pilot’s confirmation.

This ensures the accuracy of the Intelligent Control System and reduces the potential for accidents due to discrepancies between the algorithms and external variables.

Zhao Yi explained this content for over three hours in a very detailed and meticulous manner. He connected the most critical dozen algorithm packages, enlightening everyone present.

"So that’s how it is!"

"I could never understand the calculation method of this detection parameter. Two different methods cover almost all possibilities you can think of, quite impressive!"

"Academician Zhao, you must have put a lot of research into fighter jets, haven’t you? Coming up with so much, you seem even more professional than a real pilot!"

"With intelligent detection, even if something unexpected occurs, the control system will automatically halt and change tactics, allowing the pilot to issue commands."

"It’s so user-friendly!"

Liu Hemin was particularly interested in the ’introduction of intelligent monitoring detection parameters’ and the ’calculation methods of these parameters’.

After Zhao Yi had finished his explanations, Liu Hemin asked, clearly not yet satisfied, whether such a parameter could be introduced into the evolving algorithm, like a child encountering an uncertain point of knowledge, being able to find the correct answers through alternative means and then recording it.

"That might be possible,"

Zhao Yi considered and replied, "But introducing this parameter into the evolutionary algorithm would make it complex, as it involves too much content. The parameters of this Intelligent Control System, after all, are finite."

Liu Hemin pondered and nodded in agreement.

Following the detailed explanation of the core algorithms of the Intelligent Control System, Zhao Yi’s task here concluded. Guo Jinhua and Deng Minrong, being the heads of the acceptance team, now fully understood, and they could delegate the verification process. Most of the remaining checks were just about debugging and conducting result tests.

This part of the work was simplified.

Consequently, Liu Hemin’s work here was also finished, as his team had only come to assist with the acceptance. If the acceptance team encountered no issues, they wouldn’t need to intervene directly in the work.

Liu Hemin, having some free time, stayed with Zhao Yi and discussed research topics.

Suddenly remembering what Liu Jiankun had said, that Zhao Yi seemed to have been engaged in mathematical research recently, Liu Hemin asked with interest, "Why have you taken up an interest in mathematics again?"

"I’ve just had some spare time lately, waiting for the acceptance process is taking too long," Zhao Yi said with some helplessness. What he looked forward to the most was for the acceptance to finish soon, not only to receive the substantial Study Coin rewards but also to start manufacturing prototypes of the WZ-A1 engine.

Design is just design, after all. Turning a design into a tangible thing is very exciting.

"What kind of research have you been doing?" Liu Hemin continued to inquire, adding, "Any results?"

"There are, I’ve finished it."

Zhao Yi first stated the outcome, then nonchalantly added, "The NP problem."

"The NP prob..."

Liu Hemin repeated before he could express his opinion on the topic raised by Zhao Yi, but he stopped abruptly and gaped at Zhao Yi, "The NP problem? You’ve completed it?" freeweɓnovēl.coɱ

"Yes?" Confusion appeared on Zhao Yi’s face as if he couldn’t comprehend why Liu Hemin was so shocked.

"..."

Liu Hemin continued gaping, at a loss for words. Eventually, he sighed deeply, didn’t even bid farewell, and walked off alone towards the cafeteria.