Chapter 326: Mathematical Theory of Error Checking

Style: Historical Author: braggartWords: 2934Update Time: 24/01/11 23:20:12
Qian Yuzhi's eyes began to be trance-like at first, and Li Janai also started to lose focus after persisting until this point. Only Feng Shan was still listening.

"Binary search searches for specific values ​​in an ordered list. It is essentially a divide-and-conquer strategy, which means dividing a large problem into several similar sub-problems, and then either solving it directly or continuing to divide it. Why does it require an ordered list? ? is to ensure that each operation can solve all sub-problems at the same time. For example, if the median value of the ascending list is less than the value being searched, I can ensure two conclusions at the same time. One, the value being searched is not in the first half of the ordered list. , two, the value being looked up is in the second half of the ordered list - then I just repeat the above operation in the second half of the ordered list."

"Our problem is similar. In terms of probability, first we can reasonably assume that there is and only 1 card is wrong. Then, we count half of all the cards that are known to contain wrong cards each time. If we count The results showed that the wrong card was not in this half, so it must be in the other half, and vice versa. So I narrowed down the 'suspect range' of half of the wrong cards. I repeated the half operation to narrow the suspect range, and when it was narrowed to a certain level, the problem It won’t be a problem anymore.”

"I told you before that the actual capabilities of the punched card computer we are making are not limited to what we see in front of us. My half-folding operation just now was very mechanical - I always cut out half, input it, and then check the result. Take the stack containing the wrong card and repeat."

"So if one day, we design a machine to replace the repetitive mechanical operation I just did, and combined with the tabulation machine, more things can be done. Many big problems will be broken down into small problems, and then the same operating process will be used solve."

"The idea of ​​decomposing a seemingly complex problem into smaller problems that are similar to the original problem and repeatedly using a similar series of mechanical operations to solve it so that a computer can complete it is called 'recursion'. This is what we You should think carefully about using computers, which is an essential idea. In particular, when thinking about this kind of problem, don't take into account the running speed of existing mechanical computers, which are not as fast as human power. The key is to think about it. Without intervention, what problems can a computer solve by just running according to the rules? That is, what kind of problems can be solved by a computer? We call them 'computable problems'. As for speed, that is not a problem - bread will have it. .”

Feng Nuo stopped and asked Feng Shan to chew on this passage carefully. For her, this kind of thinking mode is similar to mathematics, but it is quite different from the mathematics she studied before. Li Janai and Qian Yuzhi's mathematics are only at the level of four arithmetic operations. It is really difficult for them to understand. Because he didn't sleep well last night, he was already very confused at this time. These words were nothing more than a lullaby - Qian Yuzhi was still wondering what this had to do with bread.

"Okay, you two go to bed. Let me see what's going on with this card." Feng Nuo left Feng Shan aside who was still thinking blankly, and said to Qian Yuzhi and Li Janai, he pointed In the back room, "You can sleep on that bed." After saying that, he picked up the card with the wrong hole on the table.

Li Janai woke up from the bed and found Qian Yuzhi still sitting on the chair, sleeping soundly against the wall. She looked around and saw that Feng Shan was reading and calculating something at the workbench, but Elder Feng was not in the room and was probably having a meeting somewhere else.

She yawned and got out of bed, poked Qian Yuzhi, and said, "Go to bed." Unexpectedly, he just snorted and stopped moving. Li Janai dragged him to the bed - they said they were working in shifts, but in fact Qian Yuzhi stayed up late last night and was on duty much longer than she did.

Probably feeling the remaining body temperature on the bed, Qian Yuzhi turned over, as if he wanted to curl up into the depression where Li Janai had just slept, and continue to sleep soundly.

Li Janai walked to the workbench and poured a glass of water. At this time, she saw Feng Shan put down her book and rubbed her eyes, as if she was going to take a rest, so she started chatting with her. After a while, they agreed to go to a gathering of Li Janai and his classmates from the School of Arts and Sciences next Monday - it was said that several seniors would also attend.

In the next few days, the research team repeatedly improved and tested the duplicating machine. At Feng Nuo's request, a card reading mechanism was added to the row of the duplicating machine's punching mechanism. This change is not big, but it can make the duplicating machine more efficient. The machine also functions as a partial verifier. After the copy is completed, a small amount of wiring can be changed so that the machine function can automatically check whether the perforations of the two stacks of cards are consistent, and stop the machine and light up the alarm when inconsistent perforations are found.

No matter how fast the binary search is, it is still faster than having to go through the machine directly.

In addition, the improved version can also fix a master card in the card reader mechanism, punch and verify a stack of cards. In practical applications, this function can pre-punch the common hole positions of a batch of cards, reducing the workload of manual punching.

However, although the copying machine has been improved and the problem of punching wrong holes has been fixed, the phenomenon of missing holes still occasionally occurs. Finally, the team slowed down the card transmission and processing speeds and found that the problem disappeared.

"It seems that the problem is the reaction speed of the relay." Sun Li threw the pen in his hand on the table, "Slow down the speed and run it."

Everyone looked at Feng Nuo, and Feng Nuo could only nodded silently. But the elders have now become quite thick-skinned - everyone is about the same level, so no one should laugh at anyone. The embarrassment quickly recovered, and he proposed a plan to develop a decoder.

According to the plan, the tabulating machine should have integrated printing and summary punching functions. For the purpose of simplifying the complexity of a single machine, the tabulating machine does not include these parts at this stage. Therefore, Fonnuo planned to build the most basic digital decoder first. Since the printing of Latin letters was not considered, it could not even be called "decoding". It was just printing. Its mechanical structure was similar to that of a copying machine, which was nothing more than a hole inspection. Then the relay is activated to drive the arc-shaped font to print the numbers 0-9 on the top of the card. Now that the copying machine has been developed, the decoding machine is just a matter of course.

The elders of the machine shop felt that the problem was not a big one and promised to build a prototype and send it over as soon as possible.

The meeting ended here, and Feng Nuo returned to the office: He thought that he would have to prepare lessons again today - with the development of mechanical computers, he felt that he had to teach some mathematical issues related to software engineering in depth - first Today's programmers are almost all mathematicians.

He handled the daily official documents of the data center in the office, walked around the computer room as usual, and performed daily equipment maintenance. He browsed through the "to-be-repaired catalog" again and checked several "urgent" level projects with a red pencil. These could only be repaired after finishing class and sacrificing sleep time - ever since he got a mechanical computer. project, the original work has been reduced a lot. Now that Xu Laowu has done most of the paperwork for him, it would be unreasonable for him to continue to neglect his job.

After finally finishing the things at hand one by one, he got up and went to the workroom.

The "workroom" is now packed to the brim. Various "engineering prototypes" manufactured from the machinery factory are filled with this workshop. Some of them are not the products of "technical meetings" at all, but the products of the brains of some veterans with strong hands-on ability - to say it is completely This is not necessarily the case with imagination, because these things are roughly the different technical ideas in the development process of mechanical computers at that time.

Naturally, engineering prototypes are not very sophisticated. Because of the tight schedule, and mostly just to verify whether the design idea has engineering feasibility, there is no optimization in design and manufacturing. It adheres to the usual silly, black and rough appearance of the products of the machinery factory. Many equipment parts are also exposed, which saves manufacturing time and facilitates troubleshooting and debugging at any time.

The workshop is full of machines with iron mouths and teeth, and there are also many materials and parts scattered on the ground. Therefore, Feng Nuo has always taken care of his three helpers and must take precautions when entering the workshop. He sets an example himself, wearing a rattan safety helmet, coarse work clothes, and protective leather shoes.

Passing through the equipment area, at the other end of the workshop is the research area. A heavy and large "Holy Ship" 12-person conference table stands in the middle. The table is filled with drawings, documents and draft calculation papers, and seven or eight sheets are placed around it. Folding chairs. In the corner is a huge blackboard with formulas and numbers written in chalk. Against the wall is a row of open bookshelves, layer upon layer filled with various technical and reference materials - so much paperwork has been accumulated in just a few months. In accordance with the regulations of the Planning Institute, these materials were organized and bound by category by Feng Shan and Li Janai - they were all to be archived to provide reference for future technical workers.

But without a set of efficient retrieval equipment, these technical materials would probably just lie dormant in a corner of the big library and eventually be slowly forgotten. Thinking of this, Feng Nuo felt more and more that he had a heavy responsibility on his shoulders.

The ground in the research area was covered with scattered calculation drafts, making it look like the scene in the old movie when the gunners were defeated. The research team spent the whole day forgetting to eat and sleep, so naturally they didn't bother to clean up. After all, everyone had to wash their lunch boxes and take them away, otherwise they would be so sloppy that they would attract rats.

Feng Nuo sat down on the chair and took out a professional textbook on discrete mathematics - it was reprinted locally in Lingao and reviewed by the Office of Truth, so no confidentiality measures were required and it could be inserted directly on the bookshelf.

Just as I was about to open it, I suddenly saw that the wastebasket under the table was filled with scraps of paper - not ordinary scraps of paper, but torn and shredded scraps of paper. This was not the calculation paper they commonly used, but letter paper. It seems to be densely covered with writing.

-------------------------------------------------- -----------------------

Next update: Volume 7, Guangdong and Guangxi Strategy Chapter 53