The Optimization of Technological Condition of Glutamate Fermentation by Pattern Recognition Method
谷氨酸发酵工艺参数的模式识别优化设计

The technological condition of the fermention process of glutamate (such as pH value, temperature, ventilation rate et al. ) were optimized by computerized pattern recognition method. The optimization region, which is visual, were found based on the mapping from multi-dimensional pattern space into plane, then it is transformed by reciprocal direction into the original data space using Monte Carlo simulation, so the deric-tion of optimization and best combination of all factors can be determined. A new mathematical model is being proposed based on experimental evidence taken from actual production. Using the optimization method the conversion yield of carbonhydrate into glutamate increases 2.90% , the production capacity increases 1. 45%, and the glutamate output increases 2. 65%. The method has great potentialities for the decrement of the expense of raw materials and that of production cost. ,