• Proceedings of the KIEE Conference
• /
• 1998.07b
• /
• pp.578-580
• /
• 1998

The paper presents an approach to the structure identification based on the CART (Classification And Regression Tree) algorithm and to the parameter identification by hybrid learning method in neuro-fuzzy system. By using the CART algorithm, the proposed method can roughly estimate the numbers of membership function and fuzzy rule using the centers of decision regions. Then the parameter identification is carried out by the hybrid learning scheme using BP (Back-propagation) and RLSE (Recursive Least Square Estimation) from the numerical data. Finally, we will show it's usefulness for fuzzy modeling to truck backer upper control.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.25 no.4
• /
• pp.124-134
• /
• 2011

A new dead time compensation scheme that can exactly estimate the dead time and inverter nonlinearity under parameter variations is proposed for a PWM inverter-fed PMSM drive. The proposed scheme uses the fact that the sixth harmonic component in total disturbance estimated under the presence of various uncertainties is mainly caused by the dead time and inverter nonlinearity. The total disturbance due to the parameter variations as well as the dead time and inverter nonlinearity is estimated by the adaptive scheme. The sixth harmonic component is extracted from this total disturbance through harmonic analysis. The obtained sixth harmonic is processed by the PI controller to estimate the disturbance caused by the dead time and inverter nonlinearity in the stationary reference frame. The effectiveness of the proposed scheme is verified. Without requiring an additional hardware, the proposed scheme can effectively compensate the dead time and inverter nonlinearity even under the parameter variations.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.12
• /
• pp.76-84
• /
• 1999

This paper presents an integrated method for aircraft position estimation using sequential aerial images. The proposed integrated system for position estimation is composed of two parts: relative position estimation and absolute position estimation. Relative position estimation recursively computes the current position of an aircraft by accumulating relative displacement estimates extracted from two successive aerial images. Simple accumulation of parameter values decreases reliability of the extracted parameter estimates as an aircraft goes on navigating, resulting in large position error. Therefore absolute position estimation is required to compensate for the position error generated in relative position estimation. Absolute position estimation algorithms by image matching or digital elevation model (DEM) matching are presented. In image matching, a robust oriented Hausdorff measure (ROHM) is employed whereas in DEM matching an algorithm using multiple image pairs is used. Computer simulation with four real aerial image sequences shows the effectiveness of the proposed integrated position estimation algorithm.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.26 no.4
• /
• pp.241-248
• /
• 2021

A battery state-of-health (SOH) estimation algorithm using a machine learning-based linear regression method is proposed for estimating battery aging. The proposed algorithm analyzes the change trend of the open-circuit voltage (OCV) curve, which is a parameter related to SOH. At this time, a section with high linearity of the SOH and OCV curves is selected and used for SOH estimation. The SOH of the aged battery is estimated according to the selected interval using a machine learning-based linear regression method. The performance of the proposed battery SOH estimation algorithm is verified through experiments and simulations using battery packs for electric vehicles.

• Journal of Biomedical Engineering Research
• /
• v.22 no.3
• /
• pp.293-301
• /
• 2001

A patient-specific musculoskeletal model, whose parameters can be identified noninvasively, was developed for the automatic generation of patient-specific stimulation pattern in FES. The musculotendon system was modeled as a torque-generator and all the passive systems of the musculotendon working at the same joint were included in the skeletal model. Through this, it became possible that the whole model to be identified by using the experimental joint torque or the joint angle trajectories. The model parameters were grouped as recruitment of muscle fibers, passive skeletal system, static and dynamic musculotendon systems, which were identified later in sequence. The parameters in each group were successfully estimated and the maximum normalized RMS errors in all the estimation process was 8%. The model predictions with estimated parameter values were in a good agreement with the experimental results for the sinusoidal, triangular and sawlike stimulation, where the normalized RMS error was less than 17%, Above results show that the suggested musculoskeletal model and its parameter estimation method is reliable.

• Journal of IKEEE
• /
• v.22 no.3
• /
• pp.558-564
• /
• 2018

In distributed estimation where each node can collect only partial information on the parameter of interest without communication between nodes and quantize it before transmission to a fusion node which conducts estimation of the parameter, we consider a novel quantization technique employed at local nodes. It should be noted that the performance can be greatly improved if each node can transmit its measurement to one designated node (namely, head node) which can quantize its estimate using the total rate available in the system. For this case, the best strategy at the head node would be simply to partition the parameter space using the generalized Lloyd algorithm, producing the global codewords, one of which is closest to the estimate is transmitted to a fusion node. In this paper, we propose an iterative design algorithm that seeks to efficiently assign the codewords into each of quantization partitions at nodes so as to achieve the performance close to that of the system with the head node. We show through extensive experiments that the proposed algorithm offers a performance improvement in rate-distortion perspective as compared with previous novel techniques.

• Proceedings of the KIPE Conference
• /
• 1996.06a
• /
• pp.17-20
• /
• 1996

An adaptive current control scheme of a permanent magnet (PM) synchronous motor with the simultaneous estimation of the magnitude of the flux linkage and stator resistance is proposed. The adaptive parameter estimation is achieved by a model reference adaptive system (MRAS) technique. The adaptive laws are derived by the Popov's hyperstability theory and the positivity concept. The predictive control scheme is employed for the current controller with the estimated parameters. The robustness of the proposed current control scheme is compared with the conventional one through the computer simulations.

• Proceedings of the KOSOMBE Conference
• /
• v.1995 no.11
• /
• pp.193-196
• /
• 1995

An algorithm using topological mapping has been developed for a real-time detection of the QRS complexes of ECG signals. As a measurement of QRS complex energy, we used topological mapping from one dimensional sampled ECG signals to two dimensional vectors. These vectors are reconstructed with the sampled ECG signals and the delayed ones. In this method, the detection rates of CRS complex vary with the parameters such as R-R interval average and peak detection threshold coefficient. We use mean, median, and iterative method to determint R-R interval average and peak estimation. We experiment on various value of search back coefficient and peak detection threshold coefficient to find optimal rule.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.526-526
• /
• 2000

The present has a double objective. The first one is to compare and estimate sixteen continuous-time methods through the identificatiun of a system consisted with an RLC electrical circuit. These sixteen methods are classified into three groups that are the linear filters, the modulating functions and the integral methods. The second objective is to estimate directly the physical parameters of the RLC circuit, without resorting to a discrete-time model. The system is consisted of a coil with inductance L and resistance H, and of a capacitor with capacitance C. Having written the physical equations which describe the behavior of the system, the transfer function in where the initial conditions appear is given. These initial conditions should be taken into account during the parameter estimation phase, because they are inevitable within the framework of real signals. A physical interpretation of the identified models is tempted by the direct estimation of the physical parameters L and C. In conclusion, a classification of the studied methods is proposed.

• Proceedings of the KIEE Conference
• /
• 2005.04a
• /
• pp.207-209
• /
• 2005

Recently, the use of IPMSM is coming to be active, in many industrial applications. In sensorless drive of IPMSM, it is important to know the exact information of the initial rotor position, because the wrong estimation of the initial rotor position brings about the decrease of the starting torque, or a temporary reverse revolution, In addition, PMSM is necessary to use the accurate information of the inductance for the precise torque control owing to the reluctance torque. In this paper presents initial rotor position estimation method and, measure method of the each-axis inductance. And to minimize the speed estimations error, the estimated speeds are compensated by using an instantaneous reactive power in synchronously rotating reference frame.