Initial data truncation for multivariate output of discrete-event simulations using the Kalman filter

Data truncation is a commonly accepted method of dealing with initialization bias in discrete-event simulation. An algorithm for determining the appropriate initial-data truncation point for multivariate output is proposed. The technique entails averaging across independent replications and estimating a steady-state output model in a state-space framework. A Bayesian technique called Multiple Model Adaptive Estimation (MMAE) is applied to compute a time varying estimate of the output's steady-state mean vector. This MMAE implementation features the use, in paralle, of a bank of Kalman filters. Each filter is constructed under a different assumption concerning the output's steady-state mean vector. One of the filters assumes that the steady-state mean vector is accurately reflected by an estimate, called the assumed steady-state mean vector, taken from the last half of the simulation data. As the filters process the output through the effective transient, this particular filter becomes more likely (in a Bayesian sense) to be the best filter to represent the data and the MMAE mean estimator is influenced increasingly towards the assumed steady-state mean vector. The estimated truncation point is selected when a norm of the MMAE mean vector estimate is within a small tolerance of the assumed steady-state mean vector. A Monte Carlo analysis using data from simulations of open and closed queueing models is used to evaluate the technique. The evaluation criteria include the ability to construct accurate and reliable confidence regions for the mean response vector based on the truncated sequences.     

Modelling of the processes of ultrafine SiO2 production under thermal plasma conditions

The production of ultrafine silica particles is examined by modelling the processes in a flow plasma reactor. The model is based on the authors' experimental data on silica sand processing by use of thermal arc plasma. The free-molecular coagulation is assumed to be the dominant process for particle growth. This is carried out at fast cooling of the vapour, during its mixing with oxygen. The particle size distribution functions are calculated, and the influence of the chemical monomer generation and the mixing on their behavior is investigated. Comparison of the calculated and the experimental mean particle size is made.List of symbols C _nl collision frequency function, cm³/s - d _m mean median diameter, nm - F _LN(d) integral form of the particle size distribution function,% - g(t) mixing function - k Boltzmann's constant - k ₁,k,k rate coefficients, cm³/s - k ₂,k ₃ rate coefficiens, cm⁶/s - m mass of the monomer, g - M number of groups - T absolute temperature, K - T ₀ temperature of the hot vapour, K - T _E temperature of SiO₂ condensation, K - t time, s - t _m mixing time, s - t _c colling time, s - z ₀ monomer concentration, cm^–3 - z _n ,z _l concentrations of particles in groups, cm^–3 - z _n total number concentration of particles (n=0, 1, M), cm^–3 - z _n /(z _n )(log d _n+1 –logd _n) differential form of particle size distribution function - w _i reaction rates (i=1, 2, 3), cm^–3/s - , cooling rates, K/s - coefficient of proportionality - _SiO conversion rate of SiO,% - coefficient determined from the cooling rate, s^–1 - mass density of the particles, g/cm³ - standard deviation The authors gratefully acknowledge Prof. L. S. Polak from the Institute of Petrol Chemical Synthesis, Russian Academy of Science, for helpful discussions.