• 전기학회논문지
• /
• 제66권5호
• /
• pp.833-842
• /
• 2017

In this study, the design methodology of radial basis function neural networks based on incremental K-means clustering is introduced for learning and processing the big data. If there is a lot of dataset to be trained, general clustering may not learn dataset due to the lack of memory capacity. However, the on-line processing of big data could be effectively realized through the parameters operation of recursive least square estimation as well as the sequential operation of incremental clustering algorithm. Radial basis function neural networks consist of condition part, conclusion part and aggregation part. In the condition part, incremental K-means clustering algorithms is used tweights of the conclusion part are given as linear function and parameters are calculated using recursive least squareo get the center points of data and find the fitness using gaussian function as the activation function. Connection s estimation. In the aggregation part, a final output is obtained by center of gravity method. Using machine learning data, performance index are shown and compared with other models. Also, the performance of the incremental K-means clustering based-RBFNNs is carried out by using PSO. This study demonstrates that the proposed model shows the superiority of algorithmic design from the viewpoint of on-line processing for big data.

• 산업경영시스템학회지
• /
• 제39권2호
• /
• pp.88-102
• /
• 2016

This paper considers the allocation and engagement scheduling problem of interceptor missiles, and the problem was formulated by using MIP (mixed integer programming) in the previous research. The objective of the model is the maximization of total intercept altitude instead of the more conventional objective such as the minimization of surviving target value. The concept of the time window was used to model the engagement situation and a continuous time is assumed for flying times of the both missiles. The MIP formulation of the problem is very complex due to the complexity of the real problem itself. Hence, the finding of an efficient optimal solution procedure seems to be difficult. In this paper, an efficient genetic algorithm is developed by improving a general genetic algorithm. The improvement is achieved by carefully analyzing the structure of the formulation. Specifically, the new algorithm includes an enhanced repair process and a crossover operation which utilizes the idea of the PSO (particle swarm optimization). Then, the algorithm is throughly tested on 50 randomly generated engagement scenarios, and its performance is compared with that of a commercial package and a more general genetic algorithm, respectively. The results indicate that the new algorithm consistently performs better than a general genetic algorithm. Also, the new algorithm generates much better results than those by the commercial package on several test cases when the execution time of the commercial package is limited to 8,000 seconds, which is about two hours and 13 minutes. Moreover, it obtains a solution within 0.13~33.34 seconds depending on the size of scenarios.

• 전기학회논문지
• /
• 제64권7호
• /
• pp.1114-1124
• /
• 2015

In this paper, precipitation / non-precipitation pattern classification of meteorological radar data is conducted by using neuro-fuzzy algorithm. Structure expression of meteorological radar data information is analyzed in order to effectively classify precipitation and non-precipitation. Also diverse input variables for designing pattern classifier could be considered by exploiting the quantitative as well as qualitative characteristic of meteorological radar data information and then each characteristic of input variables is analyzed. Preferred pattern classifier can be designed by essential input variables that give a decisive effect on output performance as well as model architecture. As the proposed model architecture, neuro-fuzzy algorithm is designed by using FCM-based radial basis function neural network(RBFNN). Two parts of classifiers such as instance classifier part and echo classifier part are designed and carried out serially in the entire system architecture. In the instance classifier part, the pattern classifier identifies between precipitation and non-precipitation data. In the echo classifier part, because precipitation data information identified by the instance classifier could partially involve non-precipitation data information, echo classifier is considered to classify between them. The performance of the proposed classifier is evaluated and analyzed when compared with existing QC method.

• 한국군사과학기술학회지
• /
• 제23권4호
• /
• pp.407-416
• /
• 2020

Recent complex battle field demands Network Centric Warfare(NCW) ability to control various parts into a cohesive unit. In path planning filed, the NCW ability increases complexity of path planning algorithm, and it has to consider a communication coverage map as well as traditional parameters such as minimum radar exposure and survivability. In this paper, pros and cons of various propagation models are summarized, and we suggest a coverage map generation method using a Longley-Rice propagation model. Previous coverage map based on line of sight has significant discontinuities that limits selection of path planning algorithms such as Dijkstra and fast marching only. If there is method to remove discontinuities in the coverage map, optimization based path planning algorithms such as trajectory optimization and Particle Swarm Optimization(PSO) can also be used. In this paper, the Longley-Rice propagation model is used to calculate continuous RF strengths, and convert the strength data using smoothed leaky BER for the coverage map. In addition, we also suggest other types of rough coverage map generation using a lookup table method with simple inputs such as terrain type and antenna heights only. The implemented communication coverage map can be used various path planning algorithms, especially in the optimization based algorithms.

• 상하수도학회지
• /
• 제37권6호
• /
• pp.375-382
• /
• 2023

The binary oxide adsorbent using Fe and Mn (Fe-Mn) has been prepared by precipitation method to enhance the removal of phosphate. Different amounts of chitosan, a natural organic polymer, were used during preparation of Fe-Mn as a stabilizer to protect an aggregation of Fe-Mn particles. The optimal amount of chitosan has been determined considering the separation of the Fe-Mn particles by gravity from solution and highest removal efficiency of phosphate (Fe-Mn10). The application of Fe-Mn10 increased removal efficiency at least 15% compared to bare Fe-Mn. According to the Langmuir isotherm model, the maximum uptake (q_m) and affinity coefficient (b) were calculated to be 184 and 240 mg/g, and 4.28 and 7.30 L/mg for Fe-Mn and Fe-Mn10, respectively, indicating 30% and 70% increase. The effect of pH showed that the removal efficiency of phosphate was decrease with increase of pH regardless of type of adsorbent. The enhanced removal efficiency for Fe-Mn10 was maintained in entire range of pH. In the kinetics, both adsorbents obtained 70% removal efficiency within 5 min and 90% removal efficiency was achieved at 1 h. Pseudo second order (PSO) kinetic model showed higher correlation of determination (R²), suggesting chemisorption was the primary phosphate adsorption for both Fe-Mn and Fe-Mn10.

• 한국지능시스템학회논문지
• /
• 제22권6호
• /
• pp.722-728
• /
• 2012

도심지역의 하수관거 시스템은 우수 수용능력 및 하수 월류 발생 등의 시스템의 한계점을 가지고 있어, 강우시 우수 유출수로 인한 침수저감과 더불어 도시비점오염원의 저감에 모두 대응할 수 있는 저류시설의 도입이 주목받고 시작하였다. 최근 환경부에서는 방재적 우수관리와 더불어 합류식 하수관거 월류수, 분류식 우수관거 유출수 처리를 포함하는 다기능 저류시설을 "하수저류시설"이라 통칭하고, 이의 도입을 적극 추진하고 있는 실정이다. 반면 대규모 단일 저류시설 설치의 경우에는 공간 확보의 문제가 발생할 수 있으며, 이에 대안으로는 중 소규모의 분산형 저류시설 설치 및 운영을 들 수 있다. 본 연구에서는 분산형 저류시설-하수관망 네트워크 시스템의 최적 운용을 위한 모델 예측 제어기법을 제안한다. 이를 위해 첫째로 네트워크 시스템의 각 구성 요소의 수리모델을 제시함으로써 보다 정밀한 하수관망 네트워크의 거동을 모사하고자 한다. 둘째로 제안된 모델을 기반으로 현재의 강우 유입량을 고려하여 각 저류조의 수위, 하수관로의 유입/유출량을 예측하여, 입자군집 최적화 알고리즘을 이용한 모델 예측 제어기법을 바탕으로 주어진 제약조건을 만족하며 상황을 바탕으로 제안된 제어기법의 사용여부에 따른 효과를 비교 분석하고, 이의 타당성을 검증하고자 한다.

• Journal of Electrical Engineering and Technology
• /
• 제12권6호
• /
• pp.2388-2398
• /
• 2017

In this study, we propose a robust face recognition system to pose variations based on automatic pose estimation. Radial basis function neural network is applied as one of the functional components of the overall face recognition system. The proposed system consists of preprocessing and recognition modules to provide a solution to pose variation and high-dimensional pattern recognition problems. In the preprocessing part, principal component analysis (PCA) and 2-dimensional 2-directional PCA ($(2D)^2$ PCA) are applied. These functional modules are useful in reducing dimensionality of the feature space. The proposed RBFNNs architecture consists of three functional modules such as condition, conclusion and inference phase realized in terms of fuzzy "if-then" rules. In the condition phase of fuzzy rules, the input space is partitioned with the use of fuzzy clustering realized by the Fuzzy C-Means (FCM) algorithm. In conclusion phase of rules, the connections (weights) are realized through four types of polynomials such as constant, linear, quadratic and modified quadratic. The coefficients of the RBFNNs model are obtained by fuzzy inference method constituting the inference phase of fuzzy rules. The essential design parameters (such as the number of nodes, and fuzzification coefficient) of the networks are optimized with the aid of Particle Swarm Optimization (PSO). Experimental results completed on standard face database -Honda/UCSD, Cambridge Head pose, and IC&CI databases demonstrate the effectiveness and efficiency of face recognition system compared with other studies.

• 전기학회논문지
• /
• 제64권6호
• /
• pp.922-934
• /
• 2015

In this paper, The classification between precipitation echo(PRE) and non-precipitation echo(N-PRE) (including ground echo and clear echo) is carried out from weather radar data using neuro-fuzzy algorithm. In order to classify between PRE and N-PRE, Input variables are built up through characteristic analysis of radar data. First, the event classifier as the first classification step is designed to classify precipitation event and non-precipitation event using input variables of RBFNNs such as DZ, DZ of Frequency(DZ_FR), SDZ, SDZ of Frequency(SDZ_FR), VGZ, VGZ of Frequency(VGZ_FR). After the event classification, in the precipitation event including non-precipitation echo, the non-precipitation echo is completely removed by the echo classifier of the second classifier step that is built as Type-2 FCM based RBFNNs. Also, parameters of classification system are acquired for effective performance using PSO(Particle Swarm Optimization). The performance results of the proposed echo classifier are compared with CZ. In the sequel, the proposed model architectures which use event classifier as well as the echo classifier of Interval Type-2 FCM based RBFNN show the superiority of output performance when compared with the conventional echo classifier based on RBFNN.

• 한국항공우주학회지
• /
• 제43권4호
• /
• pp.326-333
• /
• 2015

최적화 프로그램과 연동시키기 위한 터보팬 엔진 성능해석 프로그램을 개발하고, 최적화 기법인 Particle Swarm Optimization을 이용하여 전투기 엔진의 주요 설계변수인 바이패스비, 팬 압축비, 고압압축기 압축비 및 버너출구온도에 대한 성능 최적화를 수행하였다. 최적화 목표는 순추력과 비연료소모율을 다목표 함수로 설정하였으며, 두 개의 목표에 대해 가중치를 주어 각 가중치별 최적 설계점을 도출하였다. 기본 모델은 F-18 전투기와 T-50 고등훈련기에 쓰이고 있는 F404 터보팬 엔진을 선정하여 분석을 수행하였다. 본 연구 결과로 네 개의 변수에 대한 최적 조건을 도출하고, 다양한 설계조건에 대한 최적 설계점 추이를 분석하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권10호
• /
• pp.3834-3857
• /
• 2021

The design of cluster-based routing protocols is necessary for Wireless Sensor Networks (WSN). But, due to the lack of features, the traditional methods face issues, especially on unbalanced energy consumption of routing protocol. This work focuses on enhancing the security and energy efficiency of the system by proposing Energy Efficient Based Secure Routing Protocol (EESRP) which integrates trust management, optimization algorithm and key management. Initially, the locations of the deployed nodes are calculated along with their trust values. Here, packet transfer is maintained securely by compiling a Digital Signature Algorithm (DSA) and Elliptic Curve Cryptography (ECC) approach. Finally, trust, key, location and energy parameters are incorporated in Particle Swarm Optimization (PSO) and meta-heuristic based Harmony Search (HS) method to find the secure shortest path. Our results show that the energy consumption of the proposed approach is 1.06mJ during the transmission mode, and 8.69 mJ during the receive mode which is lower than the existing approaches. The average throughput and the average PDR for the attacks are also high with 72 and 62.5 respectively. The significance of the research is its ability to improve the performance metrics of existing work by combining the advantages of different approaches. After simulating the model, the results have been validated with conventional methods with respect to the number of live nodes, energy efficiency, network lifetime, packet loss rate, scalability, and energy consumption of routing protocol.