http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Koichi Fujiwara,Manabu Kano,Shinji Hasebe 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
Although linear regression is a simple and useful method to build process models, they do not always function well in practice due to not only changes in process characteristics but differences of specifities between the equipments when multiple equipments are operated in parallel. To cope with them, the correlation between variables should beconsidered. In the present work, a new pattern recognition method, referred to as Nearest Correlation(NC) method that can select samples whose correlations are similar to the query point without supervised signal is proposed. The proposed procedures are as follows:1) Subtract the query point from all the other samples.2) Calculate the correlation coefficient between all pairs of arbitrary two subtracted samples, and the pairs whos ecorrelation coefficients are close to-1 are selected.4) Derive the subspace containing the query point from the selected samples.4) The Qstatistics between all samples and the derived subspace are calculated, and the samples whose Qstatistic is small are selected as the similar samples to the query point. In addition, a new soft-sensor design method integrating the NC method and Just-In-Time(JIT) modeling is proposed. This method is referred to as Correlation-based JIT(C-JIT) modeling, and it cope with the changes inprocess characteristics and the differences of specifities between the equipments. The usefulness of the proposed NC method and C-JIT modeling are demonstrated through case studies of CSTR process.
State and Parameter Estimation for Tubular Microreactors Using Particle Filter
Jun-ichi Kano,Osamu Tonomura,Manabu Kano,Shinji Hasebe 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
Micro Chemical Processes (MCPs) are constructed of micrometer scale channels. The characteristics of MCPs are rapid mixing due to short diffusion distances and accurate temperature control due to large surface-volume ratio. Therefore, MCPs make it possible to realize the production of specialty chemicals, which cannot be handled in conventional processes. To realize stable long-term operation of MCPs, it is necessary to detect the catalyst deterioration and the blockage which are the critical problems in the operation of MCPs. For example, the catalyst deterioration and the blockage are detected through the concentration meters and the flow meters installed in microchannels respectively. However, the installation of such sensors sometimes disturbs the flow. In addition, the existing miniaturized sensors are too expensive. Therefore, it is necessary to develop a monitoring system of MCPs using the state and parameter estimation. In this work, a monitoring system based on physical models and wall temperature measurements for Tubular Microreactor (TMR) is developed. It is described that Particle Filter (PF) can detect the catalyst deterioration of TMR more rapidly and accurately than Extended Kalman Filter (EKF).
Manabu Kano,Sanghong Kim,Ryota Okajima,Shinji Hasebe 제어로봇시스템학회 2012 제어로봇시스템학회 국제학술대회 논문집 Vol.2012 No.10
Virtual sensing technology is crucial to realize high product quality and productivity in any industry, but model maintenance is the most critical issue because the estimation accuracy deteriorates due to changes in processes characteristics and operating conditions. In order to realize maintenance-free high-performance virtual sensing, locally weighted partial least squares (LW-PLS) was proposed and has been successfully applied to various industrial processes. In this article, the algorithm of LW-PLS is explained focusing on the effect of similarity measures on the estimation performance. In addition, several industrial applications of LW-PLS are presented.