Online Performance Evaluation Research Articles

On-line evaluation of systolic performance by densitometry in digital left ventriculography.

The angiocardiographic evaluation of left ventricular end-diastolic (LVEDV) and end-systolic (LVESV) volumes and ejection fraction (EF) is routinely performed by the area-length method (ALM) but may lead to erroneous results. Digital imaging in real time allows densitometric procedures of determining left ventricular (LV) performance to be applied alternatively. In this study, we present densitometric algorithms for the analysis of LVEDV, LVESV, and EF from digital image data, establish accuracy and reproducibility, and determine value and limitations in comparison with ALM in single-plane 30 degrees right anterior oblique (RAO) projection. A linear relationship between iodine depth and measured densities is mainly burdened with scatter radiation and beam hardening which reduce primary radiation and suppress iodine depth. However, facilities such as deconvolution and correction algorithms are capable of reducing these sources of error. In the present study, computer-analyzed contrast images of iodine-filled wedges and spheres showed a near-linear relationship between iodine depth between 50-100 mg/cm2 and measured densities. Contrast images of heart casts and LV angio-grams of 54 patients were obtained with a digital image acquisition and processing system, and evaluated by two independent observers. The phantom study resulted in significantly (p < or = 0.01) better densitometric standard errors of estimate for volumes [3.3 ml densitometry (DENS) vs. 8.9 ml (ALM)] and simulated EF [4.3% (DENS) vs. 7.8% (ALM)] than ALM. The standard error of estimate for the comparison between both methods was 8.4 ml for volumes and 7.5% for EF. Densitometric volumes tended to underestimate volumes calculated by ALM. The angiographic study of patients demonstrated significant correlations between both methods (LVEDV r = 0.78, LVESV r = 0.83, total volumes: r = 0.89; EF r = 0.88). The standard errors of estimate can be ascribed to systematic, method-related errors of both DENS and ALM (LVEDV +/- 28.9 ml, LVESV +/- 23.4 ml, total volumes (EDV and ESV) +/- 27.1 ml; EF +/- 8.1%). The intra- and interobserver variability, respectively, exhibited significantly smaller (p < or = 0.01 and p < or = 0.05, respectively) standard errors of estimate for densitometric EF [4.6% (DENS) vs. 8.5% (ALM) and 7.1% (DENS) vs. 10.3% (ALM), respectively]. Inclined but not significant differences were found for LVEDV and LVESV. In conclusion, the data presented indicate that the calculation of LV volumes and EF in digital left ventriculography may be performed accurately by densitometric calculation in single-plane 30 degrees RAO projection. Minor underestimations in densitometric volume determination may be anticipated in the evaluation of LV geometry.

On-line performance evaluation of multiloop digital control systems

A real-time controller, developed to implement flutter suppression and rolling maneuver load alleviation control laws digitally, was tested on an aeroelastic wind-tunnel model. A controller-per formance-evaluation (CPE) methodology to evaluate various multi-input/multi-output digital control systems on line using sampled digital data was also developed and tested. Modern signal processing methods were used to generate appropriate transfer matrices of the control system from the sampled time-domain data of the whole system (controller and plant). Matrix procedures were used to calculate singular values of return-difference matrices at the plant input and output points to evaluate the performance of the control system. The CPE procedures confirmed satisfactory performance of stabilizing control laws and effectively identified potentially destabilizing ones.

Bilinear state-disturbance composite observer and its application

ABSTRACT Like linear systems, the state and disturbance observers for bilinear systems must be considered in order to realize the state feedback and/or disturbance feedforward control in the case where some of the states and/or disturbances are unmeasurable. This paper first addresses the design issue of minimal order state observers for bilinear systems; then discusses a new method of designing minimal order state-disturbance composite observers. Finally, the application of this state-disturbance composite observer to the headbox control system in the papermaking process is presented as an example. The simulation results and on-line performance evaluation for this composite observer are discussed.

A unified approach for on-line performance evaluation of CCNs

The primary trade-off in a large computer-communication network (CCN) is the balance between computation processing capability and communication capability. Essentially, a CCN model should utilize operational optimization (rather than mathematical optimization) allowing the user (mostly a non-programme network expert) to call upon his experience and insight in dealing with real-life networks to decide upon proper direction for modifications. We propose, in this paper, a unified approach which will hopefully allow an on-line user to make real-time modifications to CCNs to obtain operationally optimal system trade-offs in such areas as network overall reliability, communication response, computation response, and optimum channel capacities etc.

Calculational Model Based on Influence Function Method for Power Distribution and Control Rod Worth in Fast Reactors

A model for calculating the power distribution and the control rod worth in fast reactors has been developed. This model is based on the influence function method.The characteristics of the model are as follows:1. Influence functions for any changes in the control rod insertion ratio are expressed by using an influence function for an appropriate control rod insertion in order to reduce the computer memory size required for the method.2. A control rod worth is calculated on the basis of a one-group approximation in which cross sections are generated by bilinear (flux-adjoint) weighting, not the usual flux weighting, in order to reduce the collapse error.3. An effective neutron multiplication factor is calculated by adjoint weighting in order to reduce the effect of the error in the one-group flux distribution.The results obtained in numerical examinations of a prototype fast reactor indicate that this method is suitable for on-line core performance evaluation because of a short computing time and a small memory size.

On-Line Performance Evaluation of a TMP Refining Process - A Feasibility Study

The feasibility of an on-line performance evaluation of a disc refining process is presented. The spectral moments of the refiner torque signal dynamics in combination with a recently developed Dynamic Data System (DDS) modeling technique is being proposed for the analysis and evaluation of the TMP refining operation. The zeroth and second order spectral moments of the torque signal were found to be closely related to the process performance. Lower values of the zeroth moment (variance) and higher values of the second moment are generally associated with a good refining operation.The application of these two spectral moments is extended by formulating an “On Line Quality Index” (ONQI), which represents a suitable combination of the above moments together with other statistical quantities and the average torque itself. In a parallel development, a quality control chart is also presented which could be useful as a surveillance tool during the actual refining process.

Optimal adaptive estimation: Structure and parameter adaption

Optimal structure and parameter adaptive estimators have been obtained for continuous as well as discrete data Gaussian process models with linear dynamics. Specifically, the essentially nonlinear adaptive estimators are shown to be decomposable (partition theorem) into two parts, a linear nonadaptive part consisting of a bank of Kalman-Bucy filters and a nonlinear part that incorporates the adaptive nature of the estimator. The conditional-error-covariance matrix of the estimator is also obtained in a form suitable for on-line performance evaluation. The adaptive estimators are applied to the problem of state estimation with non-Gaussian initial state, to estimation under measurement uncertainty (joint detection-estimation) as well as to system identification. Examples are given of the application of the adaptive estimators to structure and parameter adaptation indicating their applicability to engineering problems.

Probability of Error Bounds

Abstract : Simplified upper and lower bounds to the probability of error for general M-ary hypotheses pattern recognition are obtained. The bounds, applicable to general non-gaussian densities and especially mixture densities encountered in adaptive pattern recognition, are simple to calculate and hence valuable for on-line performance evaluation of pattern recognition system. Computer evaluation of the bounds, established their tight nature and computational simplicity. Based on the bounds, feature extraction criteria are derived for supervised as well as parametric adaptive pattern recognition. (Author)