• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1305-1308
• /
• 2003

A new technique is presented to construct a predictive model of plasma etch process. This was accomplished by combining a backpropagation neural network (BPNN) and a genetic algorithm (GA). The predictive model constructed in this way is referred to as a GA-BPNN. The GA played a role of controlling training factors simultaneously. The training factors to be optimized are the hidden neuron, training tolerance, initial weight magnitude, and two gradients of bipolar sigmoid and linear functions. Each etch response was optimized separately. The proposed scheme was evaluated with a set of experimental plasma etch data. The etch process was characterized by a $2^3$ full factorial experiment. The etch responses modeled are aluminum (A1) etch rate, silica profile angle, A1 selectivity, and dc bias. Additional test data were prepared to evaluate model appropriateness. The GA-BPNN was compared to a conventional BPNN. Compared to the BPNN, the GA-BPNN demonstrated an improvement of more than 20% for all etch responses. The improvement was significant in the case of A1 etch rate.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.599-602
• /
• 2003

A new technique is presented to construct predictive models of plasma etch processes. This was accomplished by combining a backpropagation neural network (BPNN) and a random generator (RC). The RG played a critical role to control neuron gradients in the hidden layer, The predictive model constructed in this way is referred to as a randomized BPNN (RG-BPNN). The proposed scheme was evaluated with a set of experimental plasma etch process data. The etch process was characterized by a 2$^3$ full factorial experiment. The etch responses modeled are 4, including aluminum (Al) etch rate, profile angle, Al selectivity, and do bias. Additional test data were prepared to evaluate model appropriateness. The performance of RC-BPNN was evaluated as a function of the number of hidden neurons and the range of gradient. for given range and hidden neurons, 100 sets of random neuron gradients were generated and among them one best set was selected for evaluation. Compared to the conventional BPNN, the proposed RC-BPNN demonstrated about 50％ improvements in all comparisons. This illustrates that the RG-BPNN of multi-valued gradients is an effective way to considerably improve the predictive ability of current BPNN of single-valued gradient.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.52 no.5
• /
• pp.290-294
• /
• 2003

Using a backpropagation neural network (BPNN), a high power semiconductor device was empirically modeled. The device modeled is a n-LDMOSFET and its electrical characteristics were measured with a HP4156A and a Tektronix curve tracer 370A. The drain-source current $(I_{DS})$ was measured over the drain-source voltage $(V_{DS})$ ranging between 1 V to 200 V at each gate-source voltage $(V_{GS}).$ For each $V_{GS},$ the BPNN was trained with 100 training data, and the trained model was tested with another 100 test data not pertaining to the training data. The prediction accuracy of each $V_{GS}$ model was optimized as a function of training factors, including training tolerance, number of hidden neurons, initial weight distribution, and two gradients of activation functions. Predictions from optimized models were highly consistent with actual measurements.

• Structural Engineering and Mechanics
• /
• v.89 no.3
• /
• pp.309-321
• /
• 2024

In this paper, the effect of semi-rigid connections on the stability bearing capacity of cross-bracings in steel tubular transmission towers is investigated. Herein, a prediction method based on the hybrid model which is a combination of particle swarm optimization (PSO) and backpropagation neural network (BPNN) is proposed to accurately predict the stability bearing capacity of cross-bracings with semi-rigid connections and to efficiently conduct its probabilistic assessment. Firstly, the establishment of the finite element (FE) model of cross-bracings with semi-rigid connections is developed on the basis of the development of the mechanical model. Then, a dataset of 7425 samples generated by the FE model is used to train and test the PSO-BPNN model, and the accuracy of the proposed method is evaluated. Finally, the probabilistic assessment for the stability bearing capacity of cross-bracings with semi-rigid connections is conducted based on the proposed method and the Monte Carlo simulation, in which the geometric and material properties including the outer diameter and thickness of cross-sections and the yield strength of steel are considered as random variables. The results indicate that the proposed method based on the PSO-BPNN model has high accuracy in predicting the stability bearing capacity of cross-bracings with semi-rigid connections. Meanwhile, the semi-rigid connections could enhance the stability bearing capacity of cross-bracings and the reliability of cross-bracings would significantly increase after considering semi-rigid connections.

• Journal of Institute of Control, Robotics and Systems
• /
• v.13 no.4
• /
• pp.315-319
• /
• 2007

A technique to model plasma processes was presented. This was accomplished by combining the backpropagation neural network (BPNN) and genetic algorithm (GA). Particularly, the GA was used to optimize five training factor effects by balancing the training and test errors. The technique was evaluated with the plasma etch data, characterized by a face-centered Box Wilson experiment. The etch outputs modeled include Al etch rate, AI selectivity, DC bias, and silica profile angle. Scanning electron microscope was used to quantify the etch outputs. For comparison, the etch outputs were modeled in a conventional fashion. GABPNN models demonstrated a considerable improvement of more than 25% for all etch outputs only but he DC bias. About 40% improvements were even achieved for the profile angle and AI etch rate. The improvements demonstrate that the presented technique is effective to improving BPNN prediction performance.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.46.2-46
• /
• 2002

Plasma processing plays a crucial role in fabricating integrated circuits (ICs). Manufacturing ICs in a cost effective way, it is increasingly demanded a computer model that predicts plasma properties to unknown process inputs. Physical models are limited in the prediction accuracy since they are subject to many assumptions. Expensive computation time is another hindrance that prevents their widespread used in manufacturing site. To circumvent these difficulties inherent in physical models, neural networks have been used to learn nonlinear plasma data [1]. Among many types of networks, a backpropagation neural network (BPNN) is the most widely used architecture. Many training variables are...

• Journal of The Korean Society of Agricultural Engineers
• /
• v.60 no.6
• /
• pp.43-54
• /
• 2018

The accurate estimation of reference crop evapotranspiration ($ET_o$) is essential in irrigation water management to assess the time-dependent status of crop water use and irrigation scheduling. The importance of $ET_o$ has resulted in many direct and indirect methods to approximate its value and include pan evaporation, meteorological-based estimations, lysimetry, soil moisture depletion, and soil water balance equations. Artificial neural networks (ANNs) have been intensively implemented for process-based hydrologic modeling due to their superior performance using nonlinear modeling, pattern recognition, and classification. This study adapted two well-known ANN algorithms, Backpropagation neural network (BPNN) and Generalized regression neural network (GRNN), to evaluate their capability to accurately predict $ET_o$ using daily meteorological data. All data were obtained from two automated weather stations (Chupungryeong and Jangsu) located in the Yeongdong-gun (2002-2017) and Jangsu-gun (1988-2017), respectively. Daily $ET_o$ was calculated using the Penman-Monteith equation as the benchmark method. These calculated values of $ET_o$ and corresponding meteorological data were separated into training, validation and test datasets. The performance of each ANN algorithm was evaluated against $ET_o$ calculated from the benchmark method and multiple linear regression (MLR) model. The overall results showed that the BPNN algorithm performed best followed by the MLR and GRNN in a statistical sense and this could contribute to provide valuable information to farmers, water managers and policy makers for effective agricultural water governance.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.6
• /
• pp.111-121
• /
• 2019

Evapotranspiration (ET) of vegetation is one of the major components of the hydrologic cycle, and its accurate estimation is important for hydrologic water balance, irrigation management, crop yield simulation, and water resources planning and management. For agricultural crops, ET is often calculated in terms of a short or tall crop reference, such as well-watered, clipped grass (reference crop evapotranspiration, $ET_o$). The Penman-Monteith equation recommended by FAO (FAO 56-PM) has been accepted by researchers and practitioners, as the sole $ET_o$ method. However, its accuracy is contingent on high quality measurements of four meteorological variables, and its use has been limited by incomplete and/or inaccurate input data. Therefore, this study evaluated the applicability of Backpropagation Neural Network (BPNN) model for estimating $ET_o$ from less meteorological data than required by the FAO 56-PM. A total of six meteorological inputs, minimum temperature, average temperature, maximum temperature, relative humidity, wind speed and solar radiation, were divided into a series of input groups (a combination of one, two, three, four, five and six variables) and each combination of different meteorological dataset was evaluated for its level of accuracy in estimating $ET_o$. The overall findings of this study indicated that $ET_o$ could be reasonably estimated using less than all six meteorological data using BPNN. In addition, it was shown that the proper choice of neural network architecture could not only minimize the computational error, but also maximize the relationship between dependent and independent variables. The findings of this study would be of use in instances where data availability and/or accuracy are limited.