With the continuous development of science and technology, unmanned ship has gradually become a hot spot in the field of marine research. In practical applications, unmanned ships need to have long-range navigation and high efficiency, so that they can accurately perform tasks in the marine environment. As one of the key technologies of unmanned ship, path planning is of great significance to improve the endurance of unmanned ship. In order to meet the requirements, this paper proposes a path planning method for long distance unmanned ships based on reinforcement learning angle precedence ant colony improvement algorithm. Firstly, canny operator is used to automatically extract navigation environment information, and then MAKLINK graph theory is applied for environment modelling. Finally, the basic ant colony algorithm is improved and applied to the path planning of unmanned ship to generate an optimal path. The experimental results show that, compared with the traditional ant colony algorithm, the path planning method based on the improved ant colony algorithm can achieve a voyage duration of nearly 7 km for unmanned ships under the same sailing environment, which has certain practicability and popularization value.

In order to solve the problem of improper thrust distribution of each thruster of underwater vehicle, the PSO optimization algorithm is used to solve the problem of thrust distribution. According to the spatial layout of the thruster, the algorithm model of the underwater vehicle propulsion system is established. The thrust input is carried out under the broken line search trajectory, and the simulation verifies the thrust allocation results of the PSO algorithm and the traditional pseudo-inverse method. The simulation results show that compared with the traditional algorithm. First of all, the PSO algorithm can set the physical threshold for each thruster to prevent the thruster from having too much thrust. Secondly, it can ensure that the thruster can turn with a reasonable torque to prevent the robot from drifting due to the large thrust gap. This paper provides a theoretical reference for thrust distribution of underwater salvage robot, and has practical engineering significance.

This paper is proposing a novel machine scheduling model for the unrelated parallel machine scheduling problem without setup times to minimize the total completion time, also known as “makespan”. This problem is a NP-complete problem, and to date, most approaches for real-life situations are based on the operator’s experience or simple heuristics. The new model based on the Memetic Algorithm, which was proposed by P. Moscato in 1989, is a hybrid algorithm that includes genetic algorithm and local search optimization. The new model is tested on randomly generated datasets, and is compared to optimal solution, and four scheduling models; three rule-based heuristic algorithms, and a genetic algorithm based scheduling model from literature; the test results show that the new model performed better than scheduling models from literature.

The increasing use of drones in terrorist attacks highlights the need for effective strategies to prevent and respond to drone terrorism. This study uses machine learning approach to identify factors that predict the success of drone terrorism and suggests policy alternatives for preventing such acts. Drone terrorism is becoming increasingly accessible due to advancements in information and communication technology, and events such as North Korea’s drone infiltration and the Russia-Ukraine war demonstrate the potential threat of drone attacks on Important National Facilities, including nuclear power plants. Using the Global Terrorism Database (GTD), this study analyzed drone terrorism incidents that occurred worldwide from 2016 to 2020. The study employed the Random Forest algorithm, which can incorporate multiple factors and their interactions, making it particularly suitable for social science research. The study provides new insights by deriving predictors that were previously overlooked in empirical analyses of drone terrorism. The findings of this study can aid in the establishment of anti-terrorism policies aimed at addressing the growing threat of drone terrorism. This can include the organization and expansion of the crisis management governance terrorism response council, the creation of a working manual through the partial revision of laws concerning drone terrorism response, and the implementation of anti-drone equipment and systems. Ultimately, the insights gained from this study can provide development of effective strategies aimed at preventing and responding to drone attacks. The study highlights the importance of proactive measures to mitigate the risks posed by drone technology in the context of terrorism.

Establishing a ship's passage plan is an essential step before it starts to sail. The research related to the automatic generation of ship passage plans is attracting attention because of the development of maritime autonomous surface ships. In coastal water navigation, the land, islands, and navigation rules need to be considered. From the path planning algorithm's perspective, a ship's passage planning is a global path-planning problem. Because conventional global path-planning methods such as Dijkstra and A* are time-consuming owing to the processes such as environmental modeling, it is difficult to modify a ship's passage plan during a voyage. Therefore, the D* algorithm was used to address these problems. The starting point was near Busan New Port, and the destination was Ulsan Port. The navigable area was designated based on a combination of the ship trajectory data and grid in the target area. The initial path plan generated using the D* algorithm was analyzed with 33 waypoints and a total distance of 113.946 km. The final path plan was simplified using the Douglas–Peucker algorithm. It was analyzed with a total distance of 110.156 km and 10 waypoints. This is approximately 3.05% less than the total distance of the initial passage plan of the ship. This study demonstrated the feasibility of automatically generating a path plan in coastal navigation for maritime autonomous surface ships using the D* algorithm. Using the shortest distance–based path planning algorithm, the ship's fuel consumption and sailing time can be minimized.

A sample size calculation algorithm was developed in a prototype version to select inspection samples in domestic bulk handling facilities. This algorithm determines sample sizes of three verification methods satisfying target detection probability for defected items corresponding to one significant quantity (8 kg of plutonium, 75 kg of uranium 235). In addition, instead of using the approximation equation-based algorithm presented in IAEA report, the sample size calculation algorithm based on hypergeometric density function capable of calculating an accurate non-detection probability is adopted. The algorithm based the exact equation evaluates non-detection probability more accurately than the existing algorithm based on the approximation equation, but there is a disadvantage that computation time is considerably longer than the existing algorithm due to the large amount of computational process. It is required to determine sample size within a few hours using laptop-level performance because sample size is generally calculated with an inspector’s portable laptop during inspection activity. Therefore, it is necessary to improve the calculation speed of the algorithm based on the exact equation. In this study, algorithm optimization was conducted to improve computation time. In order to determine optimal sample size, the initial sample size is calculated first, and the next step is to perform an iterative process by changing the sample size to find optimal result. Most of the computation time occurs in sample size optimization process performing iterative computation. First, a non-detection probability calculation algorithm according to the sample sizes of three verification methods was improved in the iterative calculation process for optimizing sample size. A computation time for each step within the algorithm was reviewed in detail, and improvement approaches were derived and applied to some areas that have major effects. In addition, the number of iterative process to find the optimal sample size was greatly reduced by applying the algorithm based on the bisection method. This method finds optimal value using a large interval at the beginning step and reduces the interval size whenever the number of repetitions increases, so the number of iterative process is less than the existing algorithm using unit interval size. Finally, the sample sizes were calculated for 219 example cases presented by the IAEA report to compare computation time. The existing algorithm took about 15 hours, but the improved algorithm took only about 41 minutes using high performance workstation (about 22 times faster). It also took 87 minutes for calculating the cases using a regular laptop. The improved algorithm through this study is expected to be able to apply the sample size determination process, which was performed based on the approximate equation due to the complexity and speed issues of the past calculation process, based on the accurate equation.

Path planning is necessary for mobile robots to perform precise and rapid tasks. A collision avoidance function must be included so that the robot can move safely during work, and it must be able to create an optimal path to reduce work execution time and save energy. In this paper, we propose a smart route generation algorithm that searches for global route with an algorithm that can speed up route search and integrates the TEB algorithm that can search for regional optimum routes in real time according to the situation. The performance of the proposed algorithm was verified through actual driving experiments of mobile robots.

Recently, research on prediction algorithms using deep learning has been actively conducted. In addition, algorithmic trading (auto-trading) based on predictive power of artificial intelligence is also becoming one of the main investment methods in stock trading field, building its own history. Since the possibility of human error is blocked at source and traded mechanically according to the conditions, it is likely to be more profitable than humans in the long run. In particular, for the virtual currency market at least for now, unlike stocks, it is not possible to evaluate the intrinsic value of each cryptocurrencies. So it is far effective to approach them with technical analysis and cryptocurrency market might be the field that the performance of algorithmic trading can be maximized. Currently, the most commonly used artificial intelligence method for financial time series data analysis and forecasting is Long short-term memory(LSTM). However, even t4he LSTM also has deficiencies which constrain its widespread use. Therefore, many improvements are needed in the design of forecasting and investment algorithms in order to increase its utilization in actual investment situations. Meanwhile, Prophet, an artificial intelligence algorithm developed by Facebook (META) in 2017, is used to predict stock and cryptocurrency prices with high prediction accuracy. In particular, it is evaluated that Prophet predicts the price of virtual currencies better than that of stocks. In this study, we aim to show Prophet's virtual currency price prediction accuracy is higher than existing deep learning-based time series prediction method. In addition, we execute mock investment with Prophet predicted value. Evaluating the final value at the end of the investment, most of tested coins exceeded the initial investment recording a positive profit. In future research, we continue to test other coins to determine whether there is a significant difference in the predictive power by coin and therefore can establish investment strategies.

Recently, many studies are being conducted to extract emotion from text and verify its information power in the field of finance, along with the recent development of big data analysis technology. A number of prior studies use pre-defined sentiment dictionaries or machine learning methods to extract sentiment from the financial documents. However, both methods have the disadvantage of being labor-intensive and subjective because it requires a manual sentiment learning process. In this study, we developed a financial sentiment dictionary that automatically extracts sentiment from the body text of analyst reports by using modified Bayes rule and verified the performance of the model through a binary classification model which predicts actual stock price movements. As a result of the prediction, it was found that the proposed financial dictionary from this research has about 4% better predictive power for actual stock price movements than the representative Loughran and McDonald’s (2011) financial dictionary. The sentiment extraction method proposed in this study enables efficient and objective judgment because it automatically learns the sentiment of words using both the change in target price and the cumulative abnormal returns. In addition, the dictionary can be easily updated by re-calculating conditional probabilities. The results of this study are expected to be readily expandable and applicable not only to analyst reports, but also to financial field texts such as performance reports, IR reports, press articles, and social media.

Research and interest in sustainable printing are increasing in the packaging printing industry. Currently, predicting the amount of ink required for each work is based on the experience and intuition of field workers. Suppose the amount of ink produced is more than necessary. In this case, the rest of the ink cannot be reused and is discarded, adversely affecting the company's productivity and environment. Nowadays, machine learning models can be used to figure out this problem. This study compares the ink usage prediction machine learning models. A simple linear regression model, Multiple Regression Analysis, cannot reflect the nonlinear relationship between the variables required for packaging printing, so there is a limit to accurately predicting the amount of ink needed. This study has established various prediction models which are based on CART (Classification and Regression Tree), such as Decision Tree, Random Forest, Gradient Boosting Machine, and XGBoost. The accuracy of the models is determined by the K-fold cross-validation. Error metrics such as root mean squared error, mean absolute error, and R-squared are employed to evaluate estimation models' correctness. Among these models, XGBoost model has the highest prediction accuracy and can reduce 2134 (g) of wasted ink for each work. Thus, this study motivates machine learning's potential to help advance productivity and protect the environment.

An Ant Colony Optimization Algorithm(ACO) is one of the frequently used algorithms to solve the Traveling Salesman Problem(TSP). Since the ACO searches for the optimal value by updating the pheromone, it is difficult to consider the distance between the nodes and other variables other than the amount of the pheromone. In this study, fuzzy logic is added to ACO, which can help in making decision with multiple variables. The improved algorithm improves computation complexity and increases computation time when other variables besides distance and pheromone are added. Therefore, using the algorithm improved by the fuzzy logic, it is possible to solve TSP with many variables accurately and quickly. Existing ACO have been applied only to pheromone as a criterion for decision making, and other variables are excluded. However, when applying the fuzzy logic, it is possible to apply the algorithm to various situations because it is easy to judge which way is safe and fast by not only searching for the road but also adding other variables such as accident risk and road congestion. Adding a variable to an existing algorithm, it takes a long time to calculate each corresponding variable. However, when the improved algorithm is used, the result of calculating the fuzzy logic reduces the computation time to obtain the optimum value.

In this paper, we proposed and tested an indoor obstacle recognition and avoidance algorithm using vision and ultrasonic sensors for effective operation of drone with low-power. In this paper, the indoor flight of a drone is mainly composed of two algorithms. First, for the indoor flight of the drone, the vanishing point and the center point of the image were extracted through Hough transform of the input image of the vision sensor. The drone moves along the extracted vanishing point. Second, we set an area of interest so that the drone can avoid obstacles. The area of interest is a space where the drone can fly after recognizing an obstacle at a distance from the ultrasonic sensor. When an obstacle is recognized in the drone's area of ​​interest, the drone performs an obstacle avoidance action. To verify the algorithm proposed in this paper, a simple obstacle was installed in an indoor environment and the drone was flown. From the experimental results, the proposed algorithm confirmed the indoor flight and obstacle avoidance behavior of the drone according to the vanishing point.

In this paper, we design a basic algorithm enabling recognition of surrounding environment and collision avoidance among elemental technologies for autonomous driving, also applies sensor theoretical data and actual road performance to robo-racing system based on experimental data obtained through driving tests to enable sophisticated collision avoidance. For this study, a commercial autonomous driving patform(ERP-42), LiDAR and GPS sensors were used to implement efficient comunication systems and autonomous driving algorithms between each module.

The management of algal bloom is essential for the proper management of water supply systems and to maintain the safety of drinking water. Chlorophyll-a(Chl-a) is a commonly used indicator to represent the algal concentration. In recent years, advanced machine learning models have been increasingly used to predict Chl-a in freshwater systems. Machine learning models show good performance in various fields, while the process of model development requires considerable labor and time by experts. Automated machine learning(auto ML) is an emerging field of machine learning study. Auto ML is used to develop machine learning models while minimizing the time and labor required in the model development process. This study developed an auto ML to predict Chl-a using auto sklearn, one of most widely used open source auto ML algorithms. The model performance was compared with other two popular ensemble machine learning models, random forest(RF) and XGBoost(XGB). The model performance was evaluated using three indices, root mean squared error, root mean squared error-observation standard deviation ratio(RSR) and Nash-Sutcliffe coefficient of efficiency. The RSR of auto ML, RF, and XGB were 0.659, 0.684 and 0.638, respectively. The results shows that auto ML outperforms RF, and XGB shows better prediction performance than auto ML, while the differences between model performances were not significant. Shapley value analysis, an explainable machine learning algorithm, was used to provide quantitative interpretation about the model prediction of auto ML developed in this study. The results of this study present the possible applicability of auto ML for the prediction of water quality.

본 논문에서는 인력에 의한 외관 조사의 단점을 해결하고 터널 안전 점검의 자동화를 위하여 터널 스캐닝 영상을 통 한 영상접합 자동화 알고리즘을 제시한다. 터널 스캐닝 영상을 통한 안전 점검은 기존 인력에 의한 외관 조사에 비해 조사 기 간과 인력을 크게 줄일 수 있으며 조사자의 안전사고와 교통체증에 따른 사회적 비용을 절감할 수 있다는 장점이 있다. 터널 스캐닝 영상 기반 안전 점검을 위해서는 터널 스캐닝 영상의 접합을 통하여 평면 전개 이미지 자동화 생성이 핵심이다. 터널 스캐닝 영상 기반 평면 전개 이미지 생성의 자동화를 위하여 특징점 추출 및 특징점 매칭을 통한 다중촬영 이미지 간 접합 과 정이 주요한 요소이다. 본 연구에서는 터널 평면 전개 이미지 자동화 생성의 주요 요소인 이미지 접합의 성능을 높이고 기존 접합 기술에서 발생하는 오류를 해결하기 위하여 특징점 매칭 선분의 물리적인 특성을 고려하여 매칭 정확도를 높인 기술을 제 안하였다. 터널 이미지 중 약80∼90%를 이루는 타일부와 콘크리트부를 대상으로 기존기술의 특징점 매칭 결과와 제안 기술의 특징점 매칭 결과를 비교분석 하였으며 제안 기술을 통해 매칭 성능이 향상된 것을 확인하였다.

In this study, the load fluctuation of the main engine is considered to be a disturbance for the jacket coolant temperature control system of the low-speed two-stroke main diesel engine on the ships. A nonlinear PID temperature control system with satisfactory disturbance rejection performance was designed by rapidly transmitting the load change value to the controller for following the reference set value. The feed-forwarded load fluctuation is considered the set points of the dual loop control system to be changed. Real-coded genetic algorithms were used as an optimization tool to tune the gains for the nonlinear PID controller. ITAE was used as an evaluation function for optimization. For the evaluation function, the engine jacket coolant outlet temperature was considered. As a result of simulating the proposed cascade nonlinear PID control system, it was confirmed that the disturbance caused by the load fluctuation was eliminated with satisfactory performance and that the changed set value was followed.

Recently, most software development uses object-oriented method. The core of object-oriented method is class, and encaptualition and inheritance structure based on class improves development efficiency. However, if the method is encapsulated, it can be developed more effectively. In this paper, we examine the differences and effects from existing methods when encapsulating a method by applying the Template Method Pattern.

한국 천일염 생산 지역의 인구는 빠르게 고령화되고 있어 생산 노동자가 줄고 있는 추세이다. 소금 포집 작업은 천일염 생산 과정에서 가장 많은 노동력을 필요로 한다. 기존의 포집 장치는 사람의 작동 및 운전이 필요하여 상당한 노동력이 필요해서, 천일염 무인 포집장치를 개발하여 생산 노동자의 노동력을 감소시키고자 한다. 천일염 포집장치는 색상 검출을 통해 소금의 포집 상황과 염전에서의 위치를 파악하도록 설계되었기 때문에, 포집장치의 색상 검출 성능이 중요한 요소이다. 그래서 색상 검출 성능 향상을 위해 이미지 처리 를 이용한 알고리즘을 연구하였다. 알고리즘은 입력 이미지를 크기 재조정, 회전 및 투시 변환을 이용하여 around-view 이미지를 생성하고, RoI를 설정하여 해당 영역만 HSV 색상 모델로 변환하고 논리곱 연산을 통해 색상 영역을 검출한다. 검출 된 색상영역은 형태학적 연산을 이용하여 검출 영역을 확장하고 노이즈를 제거하여 컨투어와 이미지 모멘트를 이용하여 검출영역의 면적을 계산하고 설정된 면적과 비 교하여 염판에서 포집장치의 위치 경우를 결정한다. 성능 평가는 알고리즘을 적용한 최종 검출 색상의 계산 면적과 알고리즘의 각 단계 의 검출 색상의 면적을 비교하여 평가하였다. 평가 결과 소금을 검출하는 흰색의 경우 최소 25%에서 최대 99% 이상, 빨간색의 경우 최소 44%에서 최대 68%, 파란색과 녹색은 평균적으로 각각 7%와 15% 검출면적 증가가 있어 색상 검출 성능이 향상되었음을 확인할 수 있었으 며, 이를 무인 천일염 포집장치의 무인작업 수행을 위한 위치 확인에 적용 가능할 것으로 사료된다.

For a plastic diffusion lens to uniformly diffuse light, it is important to minimize deformation that may occur during injection molding and to minimize deformation. It is essential to control the injection molding condition precisely. In addition, as the number of meshes increases, there is a limitation in that the time required for analysis increases. Therefore, We applied machine learning algorithms for faster and more precise control of molding conditions. This study attempts to predict the deformation of a plastic diffusion lens using the Decision Tree regression algorithm. As the variables of injection molding, melt temperature, packing pressure, packing time, and ram speed were set as variables, and the dependent variable was set as the deformation value. A total of 256 injection molding analyses were conducted. We evaluated the prediction model's performance after learning the Decision Tree regression model based on the result data of 256 injection molding analyses. In addition, We confirmed the prediction model's reliability by comparing the injection molding analysis results.

본 연구는 데이터 과학의 과정에 따른 파이썬 기반의 외부고리 은하 영상 분석 알고리즘 개발을 목적으로 한다. 잠재적 사용자는 학생과 교사를 포함한 시민 과학자로 정하였다. 은하의 실제 데이터를 이용한 분류 연구는 IRAF 라는 전문 소프트웨어가 이용되고 있어 일반인이 접근하기에 한계가 있다. 이에 IRAF를 사용한 선행 연구의 결과와 비교 검 증이 가능한 외부고리 은하를 분석 대상 천체로 정하여, 영상 분석 알고리즘을 개발하고 그 결과를 검증하였다. 검증 결과 총 69개의 외부고리 은하 중 50개(72.5%)가 IRAF 결과와 높은 일치를 보였다. 남은 19개(27.5%)는 시선 방향에 겹친 밝은 별의 존재 혹은 은하 내부의 약한 밝기로 인해 IRAF 결과와 다른 낮은 일치를 보였다. 보완 과정을 거친 최종 결과물은 공유 및 교육 자료의 활용도를 높이기 위해 전체 사용된 데이터와 알고리즘, 파이썬 코드 파일 및 사용 설명서를 GitHub에 탑재하였다.