• Architectural research
• /
• v.24 no.1
• /
• pp.1-11
• /
• 2022

As the occurrence of a crime is dependent on different factors, their correlations are beyond the ordinary cognitive range. Owing to this limitation, systems face difficulty in correlating various factors, thereby requiring the assistance of artificial intelligence (AI) to overcome such limitations. Therefore, AI has become indispensable for crime prediction. Crimes can cause severe and irrevocable damage to a society. Recently, big data has been introduced for developing highly accurate models for crime prediction. Prediction of night crimes should be given significant consideration, because crimes primarily occur during nights, when the spatiotemporal characteristics become vulnerable to crimes. Many environmental factors that influence crime rate are applied for crime prediction, and their influence on crime rate may differ based on temporal characteristics and the nature of crime. This study aims to identify the environmental factors that influence sex and theft crimes occurring at night and proposes an artificial neural network (ANN) model to predict sex and theft crimes at night in random areas. The crime data of A district in Seoul for 12 years (2004-2015) was used, and environmental factors that influence sex and theft crimes were derived through multiple regression analysis. Two types of crime prediction models were developed: Type A using all environmental factors as input data; Type B with only the significant factors (obtained from regression analysis) as input data. The Type B model exhibited a greater accuracy than Type A, by 3.26 and 9.47 % higher for theft and sex crimes, respectively.

• Smart Structures and Systems
• /
• v.30 no.4
• /
• pp.397-410
• /
• 2022

It is essential to monitor the working conditions of bolt-connected joints, which are widely used in various kinds of steel structures. The looseness of bolts may directly affect the stability and safety of the entire structure. In this study, a guided wave-based method for bolt looseness detection and localization is presented for a joint structure with multiple bolts. SH waves generated and received by a small number (two pairs) of magnetostrictive transducers were used. The bolt looseness index was proposed based on the changes in the reconstructed responses excited by the time reversal signals of the measured unit impulse responses. The damage locations and local damage severities were estimated using the damage indices from several wave propagation paths. The back propagation neural network (BPNN) technique was employed to identify the local damages. Numerical and experimental studies were conducted on a lap joint with eight bolts. The results show that the total damage severity can be successfully detected under the effect of external force and measurement noise. The local damage severity can be estimated reasonably for the experimental data using the BPNN constructed by the training patterns generated from the finite element simulations.

• Journal of the Society of Naval Architects of Korea
• /
• v.60 no.2
• /
• pp.110-119
• /
• 2023

The prediction of ship resistance performance is typically obtained by Computational Fluid Dynamics (CFD) simulations or model tests in towing tank. However, these methods are both costly and time-consuming, so hull-form designers use statistical methods for a quick feed-back during the early design stage. It is well known that results from statistical methods are often less accurate compared to those from CFD simulations or model tests. To overcome this problem, this study suggests a new approach using a Convolution Neural Network (CNN) with voxelized hull-form data. By converting the original Computer Aided Design (CAD) data into three dimensional voxels, the CNN is able to abstract the hull-form data, focusing only on important features. For the verification, suggested method in this study was compared to a parametric method that uses hull parameters such as length overall and block coefficient as inputs. The results showed that the use of voxelized data significantly improves resistance performance prediction accuracy, compared to the parametric approach.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.9 no.2
• /
• pp.83-89
• /
• 2005

In this paper a new algorithm of learning and evolving artificial neural networks using gene expression programming (GEP) is presented. Compared with other traditional algorithms, this new algorithm has more advantages in self-learning and self-organizing, and can find optimal solutions of artificial neural networks more efficiently and elegantly. Simulation experiments show that the algorithm of evolving weights or thresholds can easily find the perfect architecture of artificial neural networks, and obviously improves previous traditional evolving methods of artificial neural networks because the GEP algorithm imitates the evolution of the natural neural system of biology according to genotype schemes of biology to crossover and mutate the genes or chromosomes to generate the next generation, and the optimal architecture of artificial neural networks with evolved weights or thresholds is finally achieved.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• v.2
• /
• pp.101-106
• /
• 2006

The geoidal undulations are needed for determining the orthometric heights from the Global Positioning System GPS-derived ellipsoidal heights. There are several methods for geoidal undulation determination. The paper presents a method employing a simple architecture Two Layer Multiquadric-Biharmonic Artificial Neural Network (TLMB-ANN) to approximate an area of 4200 square kilometres quasigeoid surface with GPS-levelling data. Hardy’s Multiquadric-Biharmonic functions is used as the hidden layer neurons’ activation function and Levenberg-Marquardt algorithm is used to train the artificial neural network. In numerical examples five surfaces were compared: the gravimetric geometry hybrid quasigeoid, Support Vector Machine (SVM) model, Hybrid Fuzzy Neural Network (HFNN) model, Traditional Three Layer Artificial Neural Network (ANN) with tanh activation function and TLMB-ANN surface approximation. The effectiveness of TLMB-ANN surface approximation depends on the number of control points. If the number of well-distributed control points is sufficiently large, the results are similar with those obtained by gravity and geometry hybrid method. Importantly, TLMB-ANN surface approximation model possesses good extrapolation performance with high precision.

• Journal of IKEEE
• /
• v.27 no.4
• /
• pp.430-438
• /
• 2023

The Modular Multilevel Converter (MMC) has emerged as a key component in HVDC systems due to its ability to efficiently transmit large amounts of power over long distances. In such systems, accurate estimation of the MMC capacitor voltage is of utmost importance for ensuring optimal system performance, stability, and reliability. Traditional methods for voltage estimation may face limitations in accuracy and robustness, prompting the need for innovative approaches. In this paper, we propose a novel distributed neural network observer specifically designed for MMC capacitor voltage estimation. Our observer harnesses the power of a multi-layer neural network architecture, which enables the observer to learn and adapt to the complex dynamics of the MMC system. By utilizing a distributed approach, we deploy multiple observers, each with its own set of neural network layers, to collectively estimate the capacitor voltage. This distributed configuration enhances the accuracy and robustness of the voltage estimation process. A crucial aspect of our observer's performance lies in the meticulous initialization of random weights within the neural network. This initialization process ensures that the observer starts with a solid foundation for efficient learning and accurate voltage estimation. The observer iteratively updates its weights based on the observed voltage and current values, continuously improving its estimation accuracy over time. The validity of proposed algorithm is verified by the result of estimated voltage at each observer in capacitor of MMC.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.5
• /
• pp.1-9
• /
• 1996

In this paper, a parameter identification method for a discrete-time linear system using multi-layer neural network is proposed. The parameters are identified with the combination of weights and the output of neuraons of a neural network, which can be used for a linear and a nonlinear controller. An adaptive output tracking architecture is designed for the linear controller. And, the nonlinear controller. A sliding mode control law is applied to the stabilizing the nonlinear controller such that output errors can be reduced. The effectiveness of the proposed control scheme is illustrated through simulations.

• Proceedings of the KIEE Conference
• /
• 1999.07b
• /
• pp.764-767
• /
• 1999

In this paper, to bring under robust and accurate control of auto-equipment systems which disturbance, parameter alteration of system, uncertainty and so forth exist, neural network controller called dynamic neural processor(DNP) is designed. Also, the architecture and learning algorithm of the proposed dynamic neural network, the DNP, are described and computer simulations are provided to demonstrate the effectiveness of the proposed learning method using the DNP.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 1996.10a
• /
• pp.211-216
• /
• 1996

This paper propose a fuzzy regression method using fuzzy neural networks when a membership value is attached to each input-output pair. First, an architecture of fuzzy nerual networks with fuzzy weights and fuzzy biases is shown. Next a cost function is defined using the fuzzy output from the fuzzy neural network and the corresponding target output with a membership value.A learning algorithm is derived from the cost function. The derived learning algorithm trains the fuzzy neural network so that the level set of the fuzzy output includes the target output. Last, the proposed method is applied to the quality evaluation problem of injection molding.

• Proceedings of the KIEE Conference
• /
• 1998.11b
• /
• pp.533-536
• /
• 1998

The intent of this paper is to describe a neural network structure called dynamic neural processor(DNP), and examine how it can be used in developing a learning scheme for computing robot inverse kinematic transformations. The architecture and learning algorithm of the proposed dynamic neural network structure, the DNP, are described. Computer simulations are provided to demonstrate the effectiveness of the proposed learning using the DNP.