• Journal of the Korean Society for Precision Engineering
• /
• v.14 no.6
• /
• pp.121-127
• /
• 1997

Multilayer feedforward networks may be applied to identify the deterministic relationship between input and output data. When the results from the network require a high level of assurance, consideration of the stochastic relationship between the input and output data may be very important. Variance is one of the effective parameters to deal with the stochastic relationship. This paper presents a new algroithm for a multilayer feedforward network to learn the variance of dispersed data without preliminary calculation of variance. In this paper, the network with this learning algorithm is named as a variance learning neural network(VALEAN). Computer simulation examples are utilized for the demonstration and the evaluation of VALEAN.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.166-168
• /
• 2015

Line of Balance (LOB) method is suitable to schedule construction projects composed of repetitive activities. Since existing LOB based repetitive project scheduling methods are deterministic, they do not lend themselves to handle uncertainties involved in repetitive construction process. Indeed, existing LOB scheduling dose not handle variability of project performance indicators. In order to bridge the gap between reality and estimation, this study provides a stochastic LOB based scheduling method that allows schedulers for effectively dealing with the uncertainties of a construction project performance. The proposed method retrieves an appropriate probability distribution function (PDF) concerning project completion times, and determines favorable start times of activities. A case study is demonstrated to verify and validate the capability of the proposed method in a repetitive construction project planning.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.2
• /
• pp.133-139
• /
• 2011

Estimation of biomass carbon stock is an important research for estimation of public benefit of forest. Previous studies about biomass carbon stock estimation have limitations, which come from the used deterministic models. The most serious problem of deterministic models is that deterministic models do not provide any explanation about the relevant effects of errors. In this study, the effects of location errors were analyzed in order to estimation of biomass carbon stock of Danyang area using Monte Carlo simulation method. More specifically, the k-Nearest Neighbor(kNN) algorithm was used for basic estimation. In this procedure, random and systematic errors were added on the location of Sample plot, and effects on estimation error were analyzed by checking the changes of RMSE. As a result of random error simulation, mean RMSE of estimation was increased from 24.8 tonC/ha to 26 tonC/ha when 0.5~1 pixel location errors were added. However, mean RMSE was converged after the location errors were added 0.8 pixel, because of characteristic of study site. In case of the systematic error simulation, any significant trends of RMSE were not detected in the test data.

• The Journal of the Korea Contents Association
• /
• v.16 no.8
• /
• pp.671-680
• /
• 2016

This research suggested a new real-time traffic signal control algorithm using fusion data of the travel time and the occupancy rate. This research applied the travel time data of traffic information system to traffic signal operation, and developed the signal control process using the degree of saturation that was estimated from the travel time data. This algorithm estimates a queue length from the travel time based on a deterministic delay model, and includes the process to change from the queue length to the degree of saturation. In addition, this model can calculate the traffic signal timings using fusion data of the travel time and the occupancy rate based on the saturation degree. The micro simulation analysis was conducted for effectiveness evaluation. We checked that the average delay decreased by up to 27 percent. In addition, we checked that this signal control algorithm could respond to a traffic condition of oversaturation and detector breakdown effectively and usefully. This research has important contribution to apply the traffic information system to traffic signal operation sectors.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.10
• /
• pp.786-790
• /
• 2009

In this paper, we present a new heuristic search algorithm, HSCP, that can solve conditional/contingent planning problems with nondeterministic actions as well as partial observations. The algorithm repeats its AND-OR search trials until a complete solution graph can be found. However, unlike existing heuristic AND-OR search algorithms such as$AO^*$ and $LAO^*$, the AND-OR search trial conducted by HSCP concentrates on only a single candidate of solution subgraphs to expand it into a complete solution graph. Moreover, unlike real-time dynamic programming algorithms such as RTDP and LRTDP, the AND-OR search trial of HSCP finds a solution immediately when it possible without delaying it until the estimated value of every state converges. Therefore, the HSCP search algorithm has the advantage that it can find a sub-optimal conditional plan very efficiently.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.6
• /
• pp.1433-1442
• /
• 1997

Algorithm design is needed to optimized bandwidth which satisfy quality of service(QoS) requirements of vary traffic classes for Multimedia service in ATM networks. The diverse flow characteristics e.g., burstiness, bit rate and burst length, have to guarantee the different quality of service(QoS) requirements in Asynchronous Transfer Mode(ATM). The QoS parameter may be measured in terms of cell loss probability and maximum cell dely. In this paper, we consider the ATM networks which the virtual path(vip) concept is implemented by applying the Markov Modulated Deterministic Process method. We develop an efficient algorithm to computer the minimum capacity required to satisfy all the QoS requirements when multiple classes of on-off source are multiplexed on single VP. Using above the result, we propose a simple algorithm to determine the VP combination to achieve the near optimum of total capacity required for satisfying the individual QoS requirements. Numerical results are also presented to demonstrate the performance of the algorithm, when compared to the optimal total capacity required.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.3
• /
• pp.11-25
• /
• 2012

We present the method to infer Context-Free Grammars by applying genetic algorithm to the Binary Third-order Recurrent Neural Networks(BTRNN). BTRNN is a multiple-layered architecture of recurrent neural networks, each of which is corresponding to an input symbol, and is combined with external stack. All parameters of BTRNN are represented as binary numbers and each state transition is performed with any stack operation simultaneously. We apply Genetic Algorithm to BTRNN chromosomes and obtain the optimal BTRNN inferring context-free grammar of positive and negative input patterns. This proposed method infers BTRNN, which includes the number of its states equal to or less than those of existing methods of Discrete Recurrent Neural Networks, with less examples and less learning trials. Also BTRNN is superior to the recent method of chromosomes representing grammars at recognition time complexity because of performing deterministic state transitions and stack operations at parsing process. If the number of non-terminals is p, the number of terminals q, the length of an input string k, and the max number of BTRNN states m, the parallel processing time is O(k) and the sequential processing time is O(km).

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.40 no.4
• /
• pp.211-220
• /
• 2017

The application of the theoretical model to real assembly lines has been one of the biggest challenges for researchers and industrial engineers. There should be some realistic approach to achieve the conflicting objectives on real systems. Therefore, in this paper, a model is developed to synchronize a real system (A discrete event simulation model) with a theoretical model (An optimization model). This synchronization will enable the realistic optimization of systems. A job assignment model of the assembly line is formulated for the evaluation of proposed realistic optimization to achieve multiple conflicting objectives. The objectives, fluctuation in cycle time, throughput, labor cost, energy cost, teamwork and deviation in the skill level of operators have been modeled mathematically. To solve the formulated mathematical model, a multi-objective simulation integrated hybrid genetic algorithm (MO-SHGA) is proposed. In MO-SHGA each individual in each population acts as an input scenario of simulation. Also, it is very difficult to assign weights to the objective function in the traditional multi-objective GA because of pareto fronts. Therefore, we have proposed a probabilistic based linearization and multi-objective to single objective conversion method at population evolution phase. The performance of MO-SHGA is evaluated with the standard multi-objective genetic algorithm (MO-GA) with both deterministic and stochastic data settings. A case study of the goalkeeping gloves assembly line is also presented as a numerical example which is solved using MO-SHGA and MO-GA. The proposed research is useful for the development of synchronized human based assembly lines for real time monitoring, optimization, and control.

• Journal of Korean Society of Transportation
• /
• v.28 no.5
• /
• pp.107-116
• /
• 2010

The vehicle travel time between locations in a downtown is greatly influenced by both complex road conditions and traffic situation that changes real time according to various external variables. The customer's demands also stochastically change by time period. Most vehicle routing problems suggest a vehicle route considering travel distance, average vehicle speed, and deterministic demand; however, they do not consider the dynamic external environment, including items such as traffic conditions and stochastic demand. A realistic vehicle routing problem which considers traffic (smooth, delaying, and stagnating) and stochastic demands is suggested in this study. A mathematical programming model and hybrid genetic algorithm are suggested to minimize the total travel time. By comparing the results considering traffic and stochastic demands, the suggested algorithm gives a better solution than existing algorithms.

• Structural Engineering and Mechanics
• /
• v.86 no.5
• /
• pp.647-661
• /
• 2023

It is recognized that the installation of energy dissipation devices, such as the tuned mass damper (TMD), decreases the dynamic response of structures, however, the best parameters of each device persist hard to determine. Unlike many works that perform only a deterministic optimization, this work proposes a complete methodology to minimize the dynamic response of footbridges by optimizing the parameters of multiple tuned mass dampers (MTMD) taking into account uncertainties present in the parameters of the structure and also of the human excitation. For application purposes, a steel footbridge, based on a real structure, is studied. Three different scenarios for the MTMD are simulated. The proposed robust optimization problem is solved via the Circle-Inspired Optimization Algorithm (CIOA), a novel and efficient metaheuristic algorithm recently developed by the authors. The objective function is to minimize the mean maximum vertical displacement of the footbridge, whereas the design variables are the stiffness and damping constants of the MTMD. The results showed the excellent capacity of the proposed methodology, reducing the mean maximum vertical displacement by more than 36% and in a computational time about 9% less than using a classical genetic algorithm. The results obtained by the proposed methodology are also compared with results obtained through traditional TMD design methods, showing again the best performance of the proposed optimization method. Finally, an analysis of the maximum vertical acceleration showed a reduction of more than 91% for the three scenarios, leading the footbridge to acceleration values below the recommended comfort limits. Hence, the proposed methodology could be employed to optimize MTMD, improving the design of footbridges.