With the accumulation of bits and pieces of knowledge, we should start to build our own knowledge system. Systematic learning strategies can help us network fragmented knowledge, which is almost a crushing existence compared to traditional learning.
When I was studying in high school, my mathematics teacher was a high-achieving student of Beijing Normal University and an excellent school-level teacher. The systematic learning strategy was taught by the teacher, which helped me a lot in my subsequent study work. Any knowledge learning process tends to show discrete characteristics at the beginning, such as learning trigonometric functions during this period, and learning analytic geometry after a period of time. When it is hard to learn analytic geometry, trigonometric functions have been almost forgotten, leading to many people’s learning careers It is to oscillate repeatedly between continuous learning and forgetting.
How to overcome this problem? The systematic learning strategy requires continuous summary and reflection when learning, and the newly learned knowledge needs to be related to the original knowledge as much as possible, or in other words, the original knowledge system should be used to collide and dissolve the new knowledge. For example, there are many complex formula inferences in trigonometric functions, many of which can be verified or deepened by analytic functions. More importantly, this process will help you establish a connection between abstract numbers and intuitive shapes. The most classic example of this in my memory is to prove that a cubic equation in one unknown must have a real number solution. Everyone knows that a quadratic equation in one variable may have one solution, two solutions, or no solution at all (within the real number domain), but why is it transformed into a cubic equation in one variable? Because the corresponding curve of a cubic equation in analytic geometry always tends to positive infinity at one end and negative infinity at the other end, at this time, there must be a real number solution that is not very intuitive!
I majored in computer science, but I was surprised that a friend invited me to consult on electricity. I personally guess that it may be because he learned that I had made a power system product alone without the assistance of power professionals! In fact, behind this is a rather bumpy way to learn about the power system, which may also make everyone understand the sadness of not being supported by the system.
Some friends may be curious, if you are a computer learner and write software honestly, why do you learn about power systems? But the key point is that you will be curious, and I will be curious too! At the beginning of my work, I did research and development of microcomputer protection products. The following picture is the knowledge structure of microcomputer protection products in my eyes:
I participated in the research and development of microcomputer protection devices for power systems in the early days. At that time, the whole company was divided into three teams: software, hardware and protection. At the beginning, I was assigned to the software group and followed the leader to write the embedded software specification program, but it didn’t take long for me to become interested in the underlying data model. After reading the source code, I began to have a sense of the lower-level driver. Interested.
I hit a small threshold in the driver section. At that time, most of the product drivers were already formed. Although I read the code carefully, there was still a hazy feeling. A chance coincidence, I need to use a very new chip for product preview (In 2005, Motorola recommended their latest V2 core chip in China), and I need to complete the driver program as soon as possible. My English is very bad, and I am really dumbfounded in the face of thousands of pages of English materials. But this chip is too new, and there is not much valuable information on the Internet, only the official document Huashan is left.
I remember that it was a difficult time. At the beginning, I had to look up many words in a short paragraph, but for some reason, I was able to read ten lines at a glance. I felt that I was not reading English, but looking up information. Three months later, I was able to drive the chip, and the early hazy feelings disappeared naturally. Moreover, debugging the chip driver has brought me an extra bonus, and I am no longer afraid to read English materials (only in the IT field, such as bus in my eyes, the first meaning of the word is bus, not bus), I often wander github, stack overflow Other websites like to pay attention to the development trend of IT frontier.
At this point, the software project team has nothing I am curious about. On the contrary, because I have strong PC software programming ability, I can even start to combine multiple fields to do some innovative work, such as scripts, GUI modules, etc. Because of the reason of debugging the driver, I began to slowly become interested in the hardware part. In the past, when debugging the driver, I only read the data of the digital chip. Now I read the data of other chips on the printed circuit board, and then I can start to understand the complete hardware schematic diagram.
Reading the hardware schematic diagram requires the use of protel software, and because of the use of protel software, I took advantage of the trend to draw on pcb board for a while. The PCB board drawn by a colleague of mine has many grid filling areas. If you modify a line at random, you need to refresh it again, which is very annoying. I wrote a set of macro scripts to help him complete the automatic refresh. I remember a sense of accomplishment.
The above-mentioned learning experience basically didn't encounter too many twists and turns, and even felt that I didn't take much effort. The journey was smooth and smooth, but the bitter journey began immediately.
◇◇◇
A good colleague of mine, the university is studying electric power system, which belongs to the protection group. Once he was so excited to show me what he simulated with matlab. He told me that he had overcome a very critical algorithm. At that time, I could only share his emotions. As for what he did, well, I couldn't understand a word. Not long after, the colleague’s patent was published and the company received a large financial reward.
Envy and jealousy, I feel as if there are not a few lines of matlab program, it is really worthy of words, and then look at the large and large sections of code I wrote, and my psychology is instantly imbalanced. Moreover, if I can understand the basic protection principles, I can fully understand the protection products. From then on, I started a journey of studying the power system.
In my eyes, there are three main parts of power knowledge in microcomputer protection: microcomputer protection principle, sampling algorithm and matlab simulation. My colleague’s invention patent is a new sampling algorithm. I found a book written by academician Yang Qixun of our school on my colleague’s desk to try.
This is a book that explains the basic algorithms of microcomputer protection. The first chapter is very easy, but the second chapter immediately gave me a head, full of mathematical formulas, I don’t know what they are. I don’t understand, there must be some basic knowledge I don’t have. I analyzed these formulas and found that most of them have integral and imaginary symbols, as shown in the figure below.
I remember that I learned calculus and complex functions when I was in college. Although I gave them back to the teacher, I naively thought that as long as I went back and reviewed them quickly, I could understand the meaning of the above formulas. Driven by the desire to make money, I opened up the dusty textbooks on college calculus and complex functions.
I don’t know if I have gone through many disasters, I finally finished the review of calculus and complex functions, and even recorded a thick study note. However, I still don’t understand the above formula. What’s more depressing is that I found that not all The integrals based on complex numbers are the integrals of the complex plane. The complex variable function has been reviewed for nothing, and my assumption is of course bankrupt. What’s more distressing is that recalling functions of complex variables makes me very interested in the nature of complex numbers. This nihilistic thing that does not exist in the real world has important applications in many practical problems. I have read many books. Later, I became more and more confused, adding a period of trouble.
After one path failed, I started to try another path to study the principle of microcomputer protection. Continuing from buying a book, I began to read the most classic book in this field: "Principles of Power System Relay Protection".
It’s a pity that I still don’t understand this book. There are many strange symbols and many nouns that I only know when they are split into words. I don’t know what the words are. There is no way. I searched through Baidu for the concepts one by one, and I found that there are many types of equipment used in the power system, which is not fun at all. There are still many formulas that I don’t understand. I remember that some of them seemed to be taught in the circuit class in college, but the college textbooks have been sold as waste paper by me (I only kept the math books for the college textbooks), so I had to buy another one . I found the most popular college circuit textbooks at present, as follows:
"The crowd looked for him a thousand Baidu, suddenly looking back, but that person is there, the lights are dim." Do you have the feeling that the stuck problem is refreshed after being broken by a trick? ? This brand new circuit book explains the analysis process of the first-order and second-order circuits in the time domain and the frequency domain in turn. After reading these contents and truly understanding the concept of the frequency domain, the complex formulas of the original learning sampling algorithm are instantly melted. Up.
Since then, my learning of power system knowledge seems to have entered the fast lane, and I have made power system transformer protection and low-cycle load shedding automatic equipment based on the content of the textbook. The following figure shows the sampling waveform matlab when I did a sudden change in frequency. Simulation diagram:
The direct consequence of the overwhelming self-confidence is that it is easy to be slapped in the face. When I started to call myself a "cottage version" of the electric power expert, several on-site accident analysis experiences severely slapped me in the face. When analyzing a power system accident, it is not only necessary to analyze the characteristics of various electrical quantities at the time of the failure, but also to be familiar with the mathematical models of various electrical equipment, and even to understand the difference in characteristics of various faults or unsteady operation. Without in-depth power system stability and transient analysis capabilities, my so-called fragmentary knowledge about power systems are basically castles in the sky.
◇◇◇
Two similar learning journeys, one was smooth and the other was quite sad. What is the reason behind this? I certainly can't answer with superficial reasons such as difficulty, easy and boring. In fact, many algorithms in the IT field are quite complex and boring, and complex mathematical processes are often used.
After careful consideration, I think the key reason is familiarity. In the beginning of the software and hardware learning process, although I was also learning new knowledge, it was always related to my original knowledge system. For example, many excellent algorithms use various B-trees. Although the specific details are forgotten, because I have studied data structure in university, knowing the general working principle and the problem to be solved, so it is easy to overcome. Correspondingly, when studying the power system, it is a headache when I see the analysis of various power models. Although I know a lot of superficial concepts, I always feel dimly lacking foundations. Perhaps these foundations are electromagnetics and power system analysis. Wait for the power system basic course.
This is the difference between systematic and unsystematic learning. One side can continue to accumulate based on compound interest, while the other side is like a monkey picking a stick while picking it up and throwing it away. The main courses studied at the university level, such as data structure, computer composition and computer architecture, operating system, network, compilation principle and other courses, we will feel boring and not practical. It is better to learn python or java to write a small program. Value, but in fact, it is these boring courses that can become the support point of the system.
If you are a college student and are lucky enough to read this article, I strongly recommend that you take the basic courses of the university seriously. Python is easy to learn after work, but the data structure is not so easy to add. Moreover, in our work practice, we found that although the trainees of the training institutions started to get started quickly, they were far less powerful than the graduates of the regular schools. As a result, everyone did not dare to recruit trainees from the training institutions.
From then on, I started to stop losses decisively, and no longer spend too much energy on learning the power system, because the pay is not proportional to the return. Of course, any experience is valuable. For me, its greatest value is to help me build a complete product view, and it is also easy to communicate with power experts.