MathorCup, commonly known as the Mother Cup, is the competition with the largest number of participants other than the Mathor National Competition, and our Mother Cup also started today as scheduled. The difficulty of this year's Mom Cup, at least in my opinion, should be the most difficult game so far in 2023. The setting of the question, the selection of the background and other aspects all reveal the thoughts of I want to kill you. Difficulty is constant. Difficulty is difficult for everyone, and everyone can treat it with a normal mind. Below I will bring you a brief analysis of each competition question, so that you can choose the topic conveniently, and predict the difficulties you will face in advance.
Optimization problem ABC, difficulty B>A>C. Other type D, data processing + correlation analysis + comprehensive evaluation + prediction D questions involve many types, but the questions are simple and the model selection is not difficult, which is perfect for teams who are not familiar with optimization models. Comprehensive difficulty, personally think B>A>C>D. The number of people who choose the topic should be similar to this, and the statistics of the number of people who choose the topic are expected to be announced tomorrow.
The amount of questions is a bit large except for question D, but each question is not difficult. The amount of ABC questions is very small and can be overcome slowly. Those who intend to choose D questions need to start working as soon as possible. The amount of D questions is really not small. I will also first update the information of question D to facilitate modeling.
Question D is a series of question settings based on QAR data. The biggest advantage of question D is that it does not require optimization knowledge. The Ma Cup itself is a competition around optimization problems. Therefore, for teams who are not familiar with optimization models or have difficulty programming in the direction of optimization, question D is undoubtedly the best. s Choice. Question D, in my opinion, can be regarded as a data analysis, that is, it can be understood as a quantitative analysis, which is similar to the term of language modeling.
Question D Aviation Safety Risk Analysis and Flight Technology Assessment Questions
Question 1 : Some QAR data has errors, and it is necessary to preprocess the data to remove the false and preserve the true, so as to reduce the impact of the wrong data on the research and analysis. Please invite your team to conduct reliability research on the data quality in Annex 1, extract some key data items related to flight safety, and analyze their importance.
Question 1. The standard data preprocessing topic requires us to remove the false and preserve the true, so as to reduce the impact of wrong data on research and analysis. Therefore, we mainly focus on data preprocessing for this problem. Data preprocessing is nothing more than two aspects, outliers and missing values. For missing values, we can choose to directly eliminate this sample, or we can choose to supplement it by means of interpolation. This processing method varies from person to person. For outliers, we usually choose the first step to determine the outliers to determine what kind of value is considered an outlier. Usually, even if the 3σ principle and box diagram are used for judgment, gray systems can also be used for judgment. Outliers, are all doable.
Question 2 : During the whole flight process from take-off to landing, the aircraft ensures flight safety through a series of flight controls, these controls mainly include roll control, pitch control, etc. At present, domestic airlines monitor flight maneuvers through overruns. Although this monitoring method can quickly identify aircraft state deviations, it can only tell safety managers what happened, and cannot immediately determine the cause of such deviations. Therefore, the cause of this deviation can be analyzed through the process change of the joystick. According to Annex 1, please provide a reasonable quantitative description of the flight control. The figure below shows the change curve of the stick position during the 3 landings. The red curve describes the process of a heavy landing (landing G value exceeds a given limit value). The heavy landing is mainly due to an improper release of the stick at low altitude by the flight crew As a result, there is an obvious sag in the 5 seconds before touchdown in the red curve, which is a release of the stick that needs to be described quantitatively.
The second problem is that we need to analyze the cause of this deviation through the process change of the joystick. Question two currently seems to me similar to a correlation analysis. The number of selected variables corresponds to different methods. Here, you can choose according to the variable data and relationship you choose according to the table below.
Question 3: The reasons for different overruns are different, sometimes specific airports are prone to specific overruns, sometimes specific weather is prone to specific overruns, and sometimes specific pilots are prone to specific overruns. Please study the data in Appendix 2, analyze the different situations of overruns, and study the basic characteristics of different overruns, such as analyzing which routes the aircraft is on or which airports are prone to overruns, etc.
Question 3. The analysis we conduct is basically quantitative analysis. Therefore, for the data given in Appendix 2, the first thing we need to do is to convert the table into data to facilitate the establishment of our model and analysis. According to our converted data, we can analyze and perform language modeling, which is not difficult
Question 4 : The research on aircraft operation data is generally divided into two categories, one is the operational performance of pilots obtained through the Line Operations Safety Audit (LOSA), and the other is based on the suggestions of relevant scholars, based on flight parameters Conduct flight technology assessments. According to Annex 3, please establish a mathematical model to discuss a flight technique evaluation method based on flight parameters, and analyze the pilot's flying technique. The "different qualifications" in the data table represent the different skill levels of the pilot.
Question 4. We need to establish an evaluation model based on the data to analyze the pilot's flight technology evaluation model. Here, everyone can understand it as an evaluation model. According to the data set in Appendix 3, there are many, many indicators. Here, it is recommended that you use the principal component analysis method to analyze the evaluation of many indicators. Of course, this is just a suggestion, and you can choose the method that suits you, or the method you want to use. Entropy weight method, rank sum ratio, etc.
Question 5 : With the advancement of technology, it is possible to install a real-time transmission QAR data recording system on civil aviation aircraft in the future. This "real-time flight data" technology can transmit flight data to the ground for analysis in near real-time system, greatly improving risk identification capabilities and prevention levels. Assuming that the flight data can be transmitted in real time between land and air, if you are the safety manager of the airline, please establish a real-time automatic early warning mechanism for the airline to prevent possible safety accidents. Combined with the data in Appendix 1, give the simulation results.
Question 5. The standard early warning model requires us to build a prediction model, and it is enough to give an early warning when the model situation is reached. It's not difficult. For the selection of prediction models, you can choose according to your team's situation. Some long-term predictions such as machine learning LSTM are recommended here, and gray predictions can also be used. They are all possible, depending on the strength of the modeler and the actual situation.
