I started with a ship full of max-level veterans, and I'm just a complete noob.

Inside the No. 3 furnace room of the Institute of Physics, the quartz boat loaded with the verification silicon material has entered the high-temperature zone. After the furnace temperature exceeds 1,300 degrees Celsius, the response of the aged silicon carbide rod begins to deviate from the no-load calibration.

The ink line on the recorder first rises upward, then crosses the reference line and swings back downward, with the peaks and troughs gradually forming a pattern, and the temperature range expanding to ±0.5 degrees.

Da Liu was guarding the control cabinet. He checked the table and found that the feedforward had given the reference power according to the material position. The fixed PID was also continuously being corrected, but each action fell after the peak was formed.

Jiang Ming left the Russian parameter package in the locker. When he rushed back to the furnace room, he first checked the thermocouple output and propulsion position to confirm that the measurement link was normal. Then he put his hand on the disturbance loop switch.

"Connect the identification loop, set the disturbance level to the first level, and keep the parameter update upper limit at last night's calibration value."

When the switch is pressed, the cam contact delivers a small step voltage, and the pointers of the multiplier network swing simultaneously, with the gain estimate and inertia estimate moving to the new scale at different speeds.

The servo motor is first supplied with a power, and then withdraws after the response slope increases. The first ripple on the paper tape is actually wider than before it was connected.

Jiang Ming stared at the updated voltage parameters. The first round of disturbances was only enough to measure the local gain. The thermal inertia of the furnace body was still slowly forming in the integrating capacitor. The controller was still getting to know the furnace.

Lin Lanying led Huang Kun through the door and saw the ammeter swinging continuously, and the temperature curve on the recording paper had crossed the original ±0.5 degree boundary.

She strode towards the distribution cabinet, her gaze falling on the newly connected control box, her hand already reaching for the test circuit breaker.

"First, cut off the new loop; the high temperature of the material cannot withstand continuous trial and error."

Jiang Ming stepped aside to block the switch, handed her the first round of response records, and pointed to the still changing inertia estimate.

"We will leave two more disturbance cycles, and the amplitude will still be within the range of process error. We will observe the execution action after the second round begins."

Lin Lanying looked at the observation window of the quartz boat. The bright band on the surface of the silicon material was still intact. She then withdrew her hand from beside the circuit breaker and stood to the side of the recorder.

The second disturbance pulse enters the system, and the identification network compares the current temperature rise slope with the previous result. The furnace gain estimate shifts downward, and the equivalent hysteresis is extended compared to the no-load parameter.

The proportional branch output is reduced accordingly, the integral time is automatically lengthened, the servo motor that originally chased the deviation reduces its movement, and the control rod stops its force in advance before the overshoot and head rises.

As soon as the ink line on the paper tape left the reference position, the power correction was already pressed in the opposite direction, and the peak was cut back into the constant temperature zone before it could unfold.

Da Liu watched as the two parameter meters stopped near the new scale. The tuning process, which used to require Jiang Ming to repeatedly calculate and manually replace resistors, was now completed automatically by the capacitor, resistor, and multiplier.

When the third disturbance pulse arrives, the parameter update amplitude has already decreased, and the controller only makes corrections to the minor changes caused by the continued aging of the silicon carbide rod.

The ink line extends forward along the recording paper, and the previously clearly visible peaks gradually become lower. After the third cycle, only a thin line remains, running along the reference scale.

Fang Xudong cut off this section of paper tape, first checked the paper feed speed and temperature scale of the recorder, and then used vernier calipers to measure the upper and lower boundaries of the constant temperature zone.

The first reading was +0.09 degrees and -0.08 degrees. He moved to another location to measure, and the result still fell within +0.09 degrees.

Da Liu recalculated the thermocouple potential, actuator power, and table output to confirm that the recorder's range matched the calibration source and that the temperature wire was not caused by the pen tip getting stuck.

Jiang Ming requested that the temperature be kept constant, and the disturbance loop should operate according to the predetermined cycle. After each injection, the parameter voltage should only move by a very small increment, and the furnace temperature line should always remain within the same narrow band.

Through the combined effects of material heat load, heating rod attenuation, and end contact drift, the constant temperature range of 1420 degrees Celsius was ultimately maintained at ±0.09 degrees Celsius.

Huang Kun took the paper tape from Fang Xudong, walked under the light to check the recording time, and then checked the disturbance points and parameter update positions grid by grid along the ink line.

After reading it once, he turned the paper tape over to check the recorder's indentation, and then rechecked the conversion process from the beginning by comparing it with the original thermoelectric potential table.

Lin Lanying stood to the side, still holding the pre-firing and material-carrying process card in her hand. She had originally worried that active disturbance would damage the melting zone, but now the paper tape had given a direct result.

Huang Kun handed her the record sheet that had been reviewed twice, pressing his finger on the center of the 1,420-degree constant temperature range.

Have you ever seen a furnace in the country that remains this stable above 1400 degrees Celsius?

Lin Lanying took the paper tape, first looked at the disturbance injection location, then turned to the cam contact on the control cabinet that was still working slowly, and only spoke after a long time.

"The existing high-temperature furnaces at the Institute of Physics cannot reach this number even when unloaded, and the difference is even greater when carrying materials."

Huang Kun took off his glasses, wiped the lenses, put them back on, and walked to the newly added control box. He asked Fang Xudong to list the drawings of the disturbance generator, response extractor, and parameter update network.

"The materials and processes were left in the joint project, while the schematic diagram, experimental conditions, and safety limits for the variable gain adaptive network were compiled separately. I personally wrote a letter to the director of the Institute of Automation, Chinese Academy of Sciences."

The weight of those words made Liu put down the tools he had just picked up. The No. 3 furnace was originally designed to purify silicon, but now this control method has gone beyond the scope of materials equipment.

When Jiang Ming thought of the double-ring road on Wiener's blackboard, his first thought wasn't the ease of having his achievement recommended, but rather the need to reorganize the public sources and technical descriptions.

The continuous test of Furnace No. 3, the decay curve of silicon carbide rods, and the pure analog circuit can all form a complete chain of evidence, while the cybernetics ideas need to be grounded in verifiable engineering derivations.

"Tonight I will organize the publicly available principles and test records, retaining all the disturbance amplitudes, limiting ranges, and component tolerances, so that the Institute of Automation can reproduce them independently."

Huang Kun nodded and put away the paper tape. He then turned the process card in Lin Lanying's hand to the formal purification page. After the material limit test was passed, the control blockade of the silicon zone furnace had been opened.

At this moment, Fang Xudong rushed back to the furnace room from outside and handed a copy of the Russian parameter package to Jiang Ming. The cover was attached with a temporary access slip for Room 502.

"Ren Xinmin has arrived. He's waiting for you in the secure room. He had someone recalculate those two pages about the aerodynamic deflection coefficients."

Jiang Ming looked at the No. 3 furnace, which was still running smoothly, and handed the formal purification process card to Da Liu, asking him to maintain the current temperature and start the process according to the established schedule after the silicon material condition was confirmed.

Old Sun took over the mechanical end inspection, Fang Xudong left a record of the adaptive parameters, and the on-site task had been completely handed over. Jiang Ming then put the Russian dossier into the file bag.

When he left the Institute of Physics, it was already dark. In the car, he flipped to the section on aerodynamic data, where the six identical variables were still listed in the same column.

The problem with Furnace No. 3 allows for active perturbation to expose the system's true parameters, but the Soviet case files can only rely on cross-verification from different sources, and any judgment must withstand the scrutiny and recalculation at the demonstration meeting.

Under the dim light of the confidential room in the west of the city, Ren Xinmin sat at the end of the long table, his pencil pressed against the pneumatic deflection coefficient gauge.

When Jiang Ming pushed the door open, he turned the page halfway around, and a line of numbers on one side had been pressed into a clear indentation by the pen tip.

"If these coefficients are real data, the air in the wind tunnel would probably be more obedient than a ruler."