谷氨酸发酵工艺参数的模式识别优化设计

应用模式识别调优法，对谷氨酸发酵的各项可控工艺参数(如pH值，温度，风量等)进行优化设计。以多维空间图象处理技术揭示模式空间的可视优化区，结合Monte Carlo模拟法从低维映照平面逆照到高维原始空间，设计了若干组优化工艺条件。经生产试验后补充新的生产记录，再行设计新的数学模型。对比试验表明：按模式识别优化设计的工艺条件操作，糖酸转化率提高2．90％，罐产量提高1 45％，产酸率提高2.65％。优化设汁方案已投入实际生产应用，该方法对降低原料单耗、产品成本具有很大潜力。

The Optimization of Technological Condition of Glutamate Fermentation by Pattern Recognition Method

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. ,