How to write academic papers well

I feel that the quality of manuscripts is not as good as before in recent years. When the manuscript was reviewed a few years ago, the pass rate of the first trial was over 90%; In the past year, only about 20%, about 60% had various problems that needed to be corrected and retried, and about 20% could not be hired. I often think, is it because I read too many manuscripts, my eyes are a little awkward and I don't like anything; Not if you think about it. It's just that the quality of the manuscript has dropped too obviously recently. Once, the editorial department of a periodical held a forum on how to improve the level of the periodical. It is recognized that the premise of improving the level is the high quality of the manuscript. Everyone has the same feeling about the decline in the quality of manuscripts in recent years. As for the reason, different people have different opinions. Some comrades think that it is the rejuvenation of the author team; Some comrades believe that in some evaluation systems for individuals or units, the number of papers is set as an index, which leads to the simple pursuit of quantity and shoddy construction. Admittedly, these can be the reasons, but some of them are independent of our own will. We should see that this is an inevitable trend and a good thing for young writers, especially young writers are in a period of vigorous creation, which is an important guarantee for a good article. In my opinion, the decline of thesis quality is serious on the surface, but the main reason is that the basic training of thesis writing is not enough. This situation can be changed through hard work. So I summed up the problems encountered in the review, hoping to help some authors write. First, the content of scientific papers was learned in middle school. The main genres of the article are narrative and argumentative. Narrative is about the background, process and influence of an event, and some author's feelings can also be added. Argumentative writing is about the demonstration of some problems and the verification of proving the correctness of arguments, that is, providing arguments, reasoning and finally drawing conclusions. Therefore, these two schools are very different. There is also a genre close to argumentative writing, but it is not in the strict sense, and only expresses its own views or impressions on certain issues or phenomena. Although there is an opinion, it does not deliberately prove the correctness of the opinion. This kind of article belongs to the essay in prose, and this article belongs to this kind of article. Scientific papers should be argumentative, at least with opinions. Generally speaking, the main problems to be solved in scientific papers are not "what" but "how to do it" and "why". It's better to have "why do it" in the article "How to do it". Some authors can easily write argumentative essays into narratives, especially concluding articles after doing research on a certain project, and only say how to do it, and rarely say why. The reason may be that this writing is very smooth, because the work is made by him (them), and the process is very clear, so you can make it clear without blowing off dust. Such an article is not deep enough. In recent years, software has been concerned by everyone, but most articles about software belong to this situation. In addition, there are many clues in the software, which should be made clear, regardless of thickness, covering everything, and the length is not short, which is a bit confusing. A scientific paper is neither a work summary nor a manual. The correctness of the content of the paper is of course very important. However, the errors in different fields are very different, so it is difficult to generalize. However, there are five points to note: (1) must be innovative, at least innovative. Whether there is innovation is the most important starting point for many publications to consider employment, especially authoritative publications like China Science. Without innovation, it is impossible to hire them. It can be said that there are two kinds of innovation: original innovation and integrated innovation. In engineering papers, there are few original innovations, and most of them are to propose some new methods and algorithms, or to analyze a problem in a way that others have never used before, which belongs to integrated innovation. Although this is also desirable, the paper must eloquently explain the results obtained by adopting the new method. There is a paper about the road extraction method of color image based on hermit spline. There is nothing wrong with the method itself, but there are two conclusions that are problematic. Firstly, hermit spline is used to fit the extracted discontinuous road signs to get continuous road signs. In fact, ordinary cubic spline or polynomial fitting can also solve this problem. Secondly, hermit spline is more suitable to express curved road markings on images, but there is no data to explain why other fitting curves are not suitable. In this way, although others have not used hermit spline in road extraction, the novelty of this paper is gone. The writing of the paper must highlight the key points. There is an article about dexterous hands of robots. The project itself is well done, but this article covers the structure of dexterous hand, finger drive and grasping control, and all parts are not thorough and have no depth. It would be much better if this article could focus on the driving of dexterous hands with limited size. There is also an article about tele-technology, which first talks about the structure of micro-camera, and then extracts the basic concepts of fuzzy control from some books. There is no organic connection between them, so this article has no content at all. If we focus on the problem solved by miniaturization of visual system, I am afraid we can write something. (3) The content of the paper should be true and correct. This is very important. Not cheating is a good scientific morality. If people see the fake, this article will definitely not be used. There is an article about control algorithm, which is a simulation algorithm. The joint angle function used in the simulation is q=0. 1sin(3πt), and the period is obviously (2/3)s, but the period of the simulation curve is 2.8s, and the maximum angular velocity is much smaller. This result at least makes people suspect that the author has not really made a simulation. (4) About comprehensive articles. The content of the summary article is mainly about what research predecessors have done on a particular topic, what problems have been solved, and what problems need further study. The most important thing is to point out the direction of further research on this subject. In this sense, writing a comprehensive article is actually very difficult and requires a lot of information. Moreover, it is necessary to analyze information, discard the rough and retain the essence, discard the false and retain the true, and build a strategic position. Never write a summary after reading a few published papers. 5] About the formula in this paper. Formulas are generally indispensable in scientific papers. It is important to deduce the formula correctly, but it is not necessary to write down the step-by-step derivation process clearly. Although some articles have written formulas, such formulas are useless if we do not pay attention to the interpretation of symbols used in the formulas and the applicable conditions of the formulas. There is an article about a parallel mechanism with three artificial muscles as actuators. The author tries to establish its mathematical model. The relationship between input air pressure and muscle length of artificial muscle is written in front, and the force balance equation is listed. Then it is said that a nonlinear state equation is obtained by combining the former with the latter, and it is regarded as a mathematical model. There are many coefficients in the equation, which are obviously related to the mechanism and muscle parameters. The author just didn't write the relationship between coefficients and parameters, which makes people wonder whether this model is derived. Even so, such a model is just a general formula and useless. Second, the title of the paper The title of the paper has the function of making the finishing point. The title should be very relevant to the content of the article. This is often overlooked. Some titles are too big; Some of them are too limited. There is an article entitled "Research on the Kinematics of Humanoid Arm of Service Robot", which is the author's kinematic analysis of a seven-degree-of-freedom arm when developing a service robot. "Humanoid" is not important, what is important is that the author puts forward the inverse kinematics solution method of this redundant degree-of-freedom arm, which can not only be used for service robots. If the topic is changed to "a solution to the inverse kinematics of a humanoid arm with seven degrees of freedom", it is of both academic significance and the author's research project. There is also a paper entitled "Research on Joint Lubrication Technology of Lunar Landing Robot", which is about the preparation and characteristics of a solid lubrication film. The topic of this paper is too big. Moreover, if the robot landing on the moon is explicitly mentioned in the title, experiments simulating the ultra-low vacuum, ultra-low temperature and strong dust environment of the moon are needed, and the conditions are not yet available. If the title is changed to "Study on Preparation Method and Characteristics of Molybdenum Sulfide-based Solid Lubricating Film", the scope will be narrowed and the experiment that cannot be carried out at present will be avoided. There are also some problems with the title of the paper itself. For example, there is a doctoral thesis entitled "Dynamic Walking Control of Humanoid Robot". Is walking static? Of course not. Even standing still is dynamic. Isn't such a title a joke? Third, the abstract is a summary of the article. It should be concise and to the point, and only need to explain the purpose of writing the paper, the methods used and the results obtained. In a poorly written abstract, there are often some useless words. For example, "with the development of robot technology, the application field is broader, and a certain problem has become a hot topic of research", and so on are useless. Four. The purpose of introduction is to put forward the problems to be solved in this paper. The introduction should come straight to the point and be concise. Some people write articles about robots. At first, a robot slave in a play written by a Czech writer was named Roberta. The first industrial robot was made in the United States in 1950, which went around too many circles and was a bit "one word must be called Greece". It is necessary for many papers to briefly describe the previous work in this field in the introduction. In particular, those articles that propose improvements to the previous methods are more necessary. It should be noted that the summary of predecessors' work should not be taken out of context, deliberately distorting the meaning of others and highlighting the advantages of their own methods, which is even more undesirable. In a paper, the summary of predecessors' work should be put in the introduction as far as possible. In the text, if it is not necessary, there should be no more such paragraphs. The citation of the literature should be correct. Your article quotes some literature, and others' articles may also quote your article. If you don't pay attention to the correctness when quoting, it may be passed down. An article about teleoperation quoted an article in the American journal Nature, in which it was mentioned that teleoperation was carried out in two places 7000km apart, and it only took 150μs to feed back information from the operation end to the execution end. This is completely impossible. Even if the radio wave travels in a straight line, it will take at least 46.67 ms, and later it was found that the author made an error in quoting. If the article is published, with the authority of Nature magazine and the influence of the author of this article, this erroneous data will definitely be cited again. Some papers have also started to quote literature and news on the Internet. Online documents are more reliable; For various reasons, the reliability of news is not high. We developed a remote-controlled target vehicle for a test base with armored vehicles, which was reported in Weapons Knowledge. Because some contents are inconvenient to make public, some technical treatment has been done to the article. The news spread to the Internet and became "China successfully developed a remote-controlled armored vehicle" and "ant" became "elephant". So if you want to quote online news, you must verify the news through other channels. Many papers also explain the structure of the article in the introduction. Although there are not many words, it is not necessary. For the paper, it is appropriate to explain the structure of the whole paper in the introduction because of its large space. Articles published in journals are not necessary. V. Experimental Verification The purpose of the experiment in the paper is to verify the correctness, feasibility and effectiveness of the theory or method proposed in the paper. At one stage, I don't agree to call it simulation experiment, but with the progress of simulation technology, it can at least become a means of verification. The correctness of the theory is not always confirmed by experiments. Those new theorems that have been proved by accepted theorems do not need to be verified. The verification of the feasibility of the method is relatively simple, and the experiment only needs to show that the method used solves the problem. The verification of methods (especially algorithms) is not well done in many papers. The so-called effective, should be to solve the problem faster or simpler than other methods, or low computational complexity, or higher computational speed, or occupy less memory. To illustrate the effect, first, there must be contrast, not "Pharaoh selling melons"; The second is to have corresponding data. In this sense, the experiments in the paper are often specially designed to illustrate the problem. The design of the experiment is very important. To explain the function of a certain factor, we must try to isolate it. I have reviewed four papers about building a robot simulation team in succession. These four papers are well written except for narrative content. The content of the article involves personal skills, decision-making mechanism and overall coordination, so I won't go into details. Moreover, the team participated in the simulation group match of the Robot Soccer World Cup twice, and both achieved good results as the runner-up. In other words, the technical measures taken in the team building era are still effective. However, these four articles all use the scores of competitions to illustrate the effectiveness of technical measures, which is inappropriate. Because the result of the football match is only a ranking, it can only explain the relative strength of the participating teams. If the opponent's strength is too low, even if he wins the championship, it can't effectively explain that the measures he has taken are correct. Moreover, as mentioned above, whether a team can win is related to many factors, such as personal skills, decision-making ability, overall coordination and so on. It is often unclear which factors play a role in achieving better results. In two recent articles, the author used the same competition results to illustrate the role of different technical measures, which is obviously unconvincing. If the same opponent is beaten twice with or without the measures in the text, the result of the game can better illustrate the role of this measure. For various reasons, many papers can't prove the correctness and effectiveness of this method with strict theory, and they can't do experiments for the time being, so they use simulation methods to illustrate it. At this time, it should be noted that although only a few simulation examples can be given in the article, as many examples as possible should be given in the simulation, because the simulation results of one or two examples are likely to be overturned by another example. There is a paper trying to find the shortest traversal path between several points with known mutual distance. The length of the paper is very long, and the method used has circled many times. The method is correct, but it has not been proved. Finally, an example is used for simulation. I wrote a simpler method when reviewing the manuscript, which is consistent with the result obtained by the thesis method. In this way, although the method proposed in this paper is good, it is meaningless at all. Sixth, there are not many conclusions, but there are not many wonderful conclusions. Due to the convenience of "copying" and "pasting" provided by word and other word processing software, some words in the text, introduction and abstract of the paper are copied into the conclusion, and we know what the conclusion says before we see it, so the conclusion is boring. However, occasionally, problems that are not involved in the article will pop up in the conclusion. Rarely, the conclusions of some papers overturn some or all of the discussions in the article. Seven, when writing with a pen, I often heard people say that "words are people's faces", which means that a good hand will add luster to you and make you look comfortable. Now the words on the paper are printed by the printer, so whether the article is fluent or not is very prominent and becomes a "face". As the saying goes, "a writing is like a person". If there are too many questions in an article, the author will not give a good impression. In the manuscript, the outstanding problems in terms of words and punctuation are: (1) Every word, every sentence, wherever you think and say, your tone and meaning are incoherent. (2) the meaning is repeated, the words are wordy, and pronouns are not used well. ⑶ Fictional words such as "Er", "Li", "Ran" and "Qi" are used awkwardly. (4) Improper use of technical terms or coinage terms is the last word of this paper. If there are national standards for terms in a certain field, although such standards are generally recommended standards, the terms recognized by the standards should be used first in order to be consistent with other people's languages. Common names should not be used in papers, even if such nouns have been used by more people. Terms have connotations, and each term has a strict definition when formulating terminology standards. If you have to create a new term in your paper, you must make it clear and have a strict definition. I questioned "trajectory tracking control" in a paper. On the surface, "tracking control" still makes sense. Come to think of it, there are only two words that can be associated with "control". One is the object, such as temperature control, pressure control, position control and force control. The other is methods, such as pid control, adaptive control and fuzzy control. What is "trajectory tracking control"? Trajectory tracking is neither a control object nor a control method. The tracking of the actual trajectory to the expected trajectory is the effect of controlling the trajectory. Therefore, "trajectory tracking" and "trajectory control" are available terms, while "trajectory tracking control" is untenable. Also, don't use commercial and Hong Kong-Taiwan-oriented terms such as "computer" and "CD-ROM" in academic articles. 5] Indiscriminate use of punctuation. The most common mistakes are periods, or long sentences, or improper sentence breaks. The most difficult mistakes are question marks and exclamation marks. In recent years, there is a very fashionable and overused word "based on". Sometimes when you open a magazine, there are two or three articles with the word "based on" in the title. The English for "based on x" is "x-based" or "Based on X". It should be said that the word "based on" is translated well. "Rule-based system" sounds better than the earlier translations of "Rule-based system" and "Rule-based system". The problem is to use "based on" in a necessary and appropriate way. You don't have to use it, so you don't have to use it to pursue the taste of wrinkles. Moreover, since it is "based on X", then X should be a real thing that can be used as a basis. An article uses the phrase "based on task level ……", and this "task level" does not really exist. Another inappropriate word is "intelligence", and some things that have no intelligence at all have also been put on this hat. In fact, it is not difficult to solve the problem of writing. As long as the author reads the article once or twice, he can find most of the text problems. However, if the author is not standardized and rigid in oral expression, it will not achieve much. Eight, the special problems of English manuscripts English is not our mother tongue, writing papers in English will of course have some problems. Most people have no ability to think in English. In this case, it is better to write a Chinese manuscript first and then translate it into English, so as to at least avoid the problem of semantic incoherence that is easy to occur when writing an English manuscript directly. The most common problem in using words in English manuscripts is: (1) hard translation in Chinese, resulting in so-called "Chinglish". Although "Pidgin" such as "study hard and make progress every day" is rarely seen, hard translation is still common. One paper translated "vehicle-mounted" into "tank-mounted". In fact, the word "vehicle" means vehicle-borne. (2) Improper use of prepositions, more use of "of" and "to", less use of other prepositions. (3) Pronouns "this" and "that" are used much, while "it" is used little, and the latter is only used much in scientific articles. (4) The sentence pattern is monotonous, and I like (or have to use "to be" to make sentences. 5. Don't pay attention to the part of speech of verbs. Some verbs can be both transitive verbs and intransitive verbs, so we should give priority to using transitive verbs to make sentences instead of using the passive voice of transitive verbs. Articles "a" and "The" are used improperly, especially the definite article "the" is easily forgotten. (7) Pay no attention to the singular and plural nouns and the person collocation of subject and predicate. ⑻: The words used in the paper should be more formal, and words with multiple meanings should be used as little as possible. For example, "get" means "get" in spoken language, but "objective" is best used in papers. The differences between Chinese and Western cultures often make English manuscripts have "China characteristics". The author of a manuscript is very modest. At the end of the article, he analyzed the shortcomings of the proposed method, and said that these shortcomings will be gradually overcome in future research. Foreigners don't say that. They always look forward. Even if they see the shortcomings, they will say that this method will have a broader application prospect with the deepening of research. In the introduction of some articles, the author insists on adding a "doctoral tutor" after "professor", so foreigners can't tell what a professor who is not a doctoral tutor looks like. Nine. There is no doubt that the first author of the paper should be the author. This is both respect for his labor and responsibility for the article. Many articles are written by graduate students, followed by the name of the tutor, which is beyond reproach. But from some articles, it is obvious that the tutor has not read it before submitting it. Even some articles have been published, and the tutor doesn't know yet. This situation is not good. Even if the tutor participated in the comment before writing, he didn't read it after writing, which means that the tutor didn't take responsibility; If students contribute in the name of the tutor without the tutor's knowledge, it is respect for the tutor on the good side, and it is suspected of "raising the flag as a tiger skin" on the bad side. Recently, there are more and more signatures on papers, and even a short article has five or six names. This situation is more common in the summary article of a project. Admittedly, it is difficult to distinguish the ideas of project participants in the research process, but the paper is not a work summary, and writing a paper is unlikely to concentrate many people's ideas. As for the practice of writing more gifts from others or the names of leaders who have never participated in the work at all, it is even more undesirable. X. How to face the opinions of peer review Generally speaking, the delivered manuscript must go through at least one technical review, and the English manuscript also has a text review. This review is usually conducted by experts in the same field hired by the editorial department of a periodical or magazine. The responsibility of the editorial department is to unify the paper format, review the text and handle peer review opinions. The responsibility of the reviewer is to review the innovation and correctness of the paper, and the review opinions should generally include suggestions for revision to improve the quality of the manuscript. The author should first attach great importance to the review opinions put forward by the reviewers and consider why he put forward these suggestions. After all, the critic's point of view comes from a bystander. As the saying goes, "bystanders see clearly", there is always some truth in his views. Some authors think that it is wrong for reviewers not to understand their own articles (I don't rule out this possibility) and not to seriously consider his opinions. The reviewer is the first reader of the paper. If he doesn't understand, the author should also consider what is wrong with his article so that others don't understand. Otherwise, how can he face more readers after publication? Of course, we should also analyze peer review opinions. Although the reviewers hired by the editorial department are all experts in the same field, it is not surprising that the reviewers may not be familiar with a specific topic studied by the author of the article and put forward some opinions that are not very pertinent. Therefore, you don't have to follow the reviewer's advice completely. After reviewing the manuscript, if you want to revise it, you must be realistic and can't cope with the reviewers. I commented on a paper about the drag reduction effect of bristle surface for a journal. At the first trial, I raised some questions and asked for a retrial after revision. The author is modest, admits that all the questions are reasonable, and makes some modifications. However, the most critical question I raised is about the experiment. The author is a little perfunctory and shows the experimental results in another curve form different from the first draft, which is obviously not available from the previous experimental results. I have a bigger question, and I suggest retrial after revision. In the third draft, the author changed the way, with more loopholes. Later, this manuscript never appeared again. Many editorial departments adopt a double-blind system for peer review, that is, the reviewer does not know who the author of the paper is; The author doesn't know who the reviewer is. No matter what the starting point of this system is, I think it limits the communication between authors and reviewers to words, and some editorial departments even relay some opinions of reviewers to the authors. This kind of communication is often insufficient, which is likely to become an obstacle to improving the quality of manuscripts. Since we are studying, we should not have scruples. The editorial departments of some magazines don't seem to review manuscripts. As soon as you publish the manuscript, write that you want to publish it in a certain issue and pay a page fee. This is an irresponsible editorial department and should be kept away from it. Speaking of the title, the title of this article is too big, but this article is just an essay. Using such a title is to pursue an eye-catching effect, which does not mean that you can write a good paper by paying attention to what is mentioned in this article. To make an inaccurate analogy, a paper is like a tree, and its content is its trunk and branches. Title, introduction, experiment, text, etc. What is described in this article may be regarded as some leaves, and this tree is rooted in the fertile soil of real talent and practical learning. If you want to write a good paper, study hard and increase your knowledge is the key. This paper was watered with painstaking efforts.