No abstract

A simulated verification and validation of the neural-network rate-function (NNRF) approach to modeling the nonlinear dynamic systems is provided. The NNRF modeling scheme utilizes some a priori process knowledge and experimental data to develop a dynamic neural-network model. Based on the obtained neural-network model, an optimal temperature trajectory was computed via the two-step method to drive a batch free-radical polymerization reaction to a prescribed molecular weight distribution (MWD). Evaluation of the quality of the end product suggests that the proposed NNRF modeling approach can be applied in dynamic modeling of a complex and nonlinear reaction system.

We propose an integrated method for optimization and control of semibatch reactors. Based on the desired control objective, dynamic programming is applied to obtain optimal operating trajectories. Yield optimization is assured for a real plant by tracking model-dependent optimal trajectories according to the proposed modified globally linearizing control (MGLC) structure. The behavior of the proposed MGLC structure is predictable and reliable, with tuning parameters based on the proposed tuning method if the manipulated variables are not constrained.

Highly dispersed silver nanoparticles could be used as catalysts, as staining pigments for glasses and ceramics, as antimicrobial materials, in surface-enhanced Raman spectroscopy, as transparent conductive coatings, in electronics, etc. Consequently, the versatility of such particles provides strong incentives to the development of a systematic way to produce their dispersions. In this work, the optimization of the synthesis of nanosized silver particles by chemical reduction using formaldehyde in aqueous solution was studied. Effects of the possible processing variables such as the reaction temperature T, the mole ratio of [formaldehyde]/[AgNO3], [NaOH]/[AgNO3], the weight ratio of PVP/AgNO3, and the molecular weight (MW) of protective agent PVP (Polyvinyl−pyrrolidone) were considered. The data-driven model on the basis of the 44 designed experimental runs provided us the optimal conditions for closely achieving the product with the specified mean particle size and conversion of silver nitrate.

No abstract

A new oxygenate additive for diesels (bio or petroleum) was manufactured using glycerol, dimethyl sulfate (DMS), and sodium hydroxide pellets as raw materials. By feeding the dimethyl sulfate into the batch reactor containing the sodium glycerate, a semibatch mode operation enhanced the effective methylation of glycerol. A conventional stirred tank reactor that can produce large quantities of oxygenate additives under a normal atmospheric pressure operation became the main feature of the methylation process. With a 3:2 molar ratio of DMS to glycerol, a 3:1 molar ratio of sodium hydroxide to glycerol, a 0.43:1 molar ratio of water to sodium hydroxide, and a temperature of 343 K at the reaction time of 24 h with the feeding time of DMS under 12 h, the conversion of glycerol (93.5%) and a combined yield of GDMEs and GTME of 71.2% were achieved for a once-through operation. A product mixture of GDME (20 wt %) and GTME (80 wt %) served as a new oxygenate additive for (bio or petroleum) diesels.

A parameter estimation procedure was developed to fit a kinetic model with the experimental data obtained from methyl methacrylate (MMA) solution polymerization in a 10 L batch reactor. For molecular weight control, we applied the modified two-step method in calculating the optimal temperature trajectory on the basis of the identified kinetic model. Application of a conventional proportional-integral (PI) controller to track the temperature trajectory proved to be acceptable in a conventional batch reactor with both heating and cooling systems. Experimental results revealed that the identified model is accurate; therefore, one can rely on the calculated optimal temperature trajectories to ensure that the polymer product quality conforms to the specification.

The application of the uniform design (UD) method to nonlinear multivariate calibration by an artificial neural network (ANN) can be used to build a model for an unknown process efficiently because it allows many levels for each factor. If the cost of each experiment is high, low partitioned levels are usually proposed first to carry out the experiments. However, if a reliable ANN model cannot be obtained from the designed experiments, the sequential pseudo-uniform design (SPUD) method developed here can be employed to locate additional experiments in the experimental region. An information free energy index is used to validate the identified ANN model. Once the identified model is verified as reliable, the optimal operating conditions can be determined to guide the process to the desired objective. The simulation results demonstrate that the product and process development based on the proposed method require only a reasonable number of experiments.