Evolutionary operation process

On-line investigation methods for statistical analysis are used when the performances of a continuous process carried out in a pilot unit or in an apparatus, have to be improved. The Evolutionary Operation Process (EVOP) method [5.7, 5.27, 5.28, 5.31] is the most famous method for on-line process analysis. The name of this method comes from its analogy with biological evolution. This analogy is based on the observation of the natural selection process in which a small variation in independent life factors is responsible for genetic mutations and thus for the evolution of species. 

Until about the second World War chemical processes were developed in an evolutionary way by building plants of increasing size and capacity. The capacity of the next plant in the series was determined by a scale-up factor that depended mainly upon experience gained from scale-ups of similar plants. Due to a lack of predictive models for chemical processes and operations, processes had to be scaled up in many small steps. This procedure was very expensive and the results unreliable. Therefore, large safety margins were incorporated in scale-up procedures, which often resulted in a significant unintended overcapacity of the designed plant. 

Carpenter, B.H., Sweeney, H.C. Process Improvement with Simplex Self Directing Evolutionary Operation, Chemical Engineering, July 5, 1965, P- D7. 

In any genetic algorithm application, the physical problem must be translated into a form suitable for manipulation by the evolutionary operators. Choice of coding is an important part of this process. 

Hahn, G.J. (1976b), Process Improvement Using Evolutionary Operation, CHEMTECH, 6, 204-206. 

Box G, Draper NR. Evolutionary Operation A Statistical Method for Process... 

A handbook of practical statistics would not be complete without at least a mention of evolutionary operation (EVOP). Although strictly speaking evolutionary operation is a method for operating plant processes, and perhaps less applicable to chemical research, a research worker may never know when he will be called upon to help the production department out of a tight spot. 

Evolutionary operation is a system for obtaining information on a going process, which will permit stepwise improvement of the process without risking a serious deviation from standard operating conditions. The method consists in observing a property or properties of the material such as quality, yield, etc, as a function of two or more operating variables. 

Box, G.EP., and Draper, N.R. (1969). Evolutionary Operation. A Statistical Method for Process Improvement. Wiley, New York. 

Besides analyzing and correlating data by statistical means, the chemical engineer also uses statistics in the development of quality control to establish acceptable limits of process variables and in the design of laboratory, pilot plant, and process plant (evolutionary operation) experiments. In the latter application, statistical strategy in the design of experiments enables the engineer to set experimental variables at levels that will yield maximum information with a minimum amount of data. 

The technique of evolutionary operation (EVOP) involves systematic small changes in process variables during the operation of the process. The results of... 

Evolutionary operation is a technique for fine-tuning a process by systematically varying process parameters. Here is how you can obtain valuable information on how to improve a product, while making it salable. 

Evolutionary operation (EVOP) is a simple technique that allows operators of industrial processes to continually improve efficiency during routine production. The basic philosophy behind EVOP is that it is inefficient to run an industrial process in such a way that only a product is produced, and that a process should be operated so as to produce not only a product but also information on how to improve the product [1 ... 

Validation of clinical trial formulations Validation of in vitro tests on solid dosage forms Experimental strategy for validation of industrial processes Optimization of industrial processes by evolutionary operation (EVOP)... 

D. Optimization of Industrial Processes by Evolutionary Operation (EVOP)... 

Evolutionary operation (EVOP) was proposed by Box and Draper to answer this problem (22). Any number of variables may be treated, but in general it is limited to the two or three critical factors known already, from the pilot scale process study, to have an influence on the properties or yield of the product. By very slightly altering the values of these variables in a systematic manner - they should remain within the limits already defined as acceptable - the dependence of the product on the operating conditions can be assessed, and in some cases the process can be improved. Two level factorial designs, or the simplex (1), are most commonly used. 

G. E. P. Box and N. R. Draper, Evolutionary Operation a Statistical Method for Process Improvement, John Wiley Sons, N. Y., 1969. 

The truth is that considerations like this are rarely made in practice. In most cases, the industrial plant operates indefinitely under the same conditions established at its start-up, and process efficiency tends to decrease with time. This way of operating, which is without doubt the usual, might be called static operation, to contrast it with what is called evolutionary operation (Box, 1957), which we address in this section. 

BOX, G. E. P. and DRAPER, N. R. Evolutionary operation a statistical method for process improvement. New York, Wiley, 1969. 

However, all of these devices require that the compounder has a reasonable understanding of the relationship between the power consumed by the mixer and a desirable process control strategy. This can be a reasonable requirement if the compounder has to consider only a limited number of formulations, or if runs are significantly long for each of the mixes. The compounder may use the technique known as evolutionary operation, EVOP , to determine a control strategy, and an appropriate instrument to effect that control action. 

SPI can be a valuable response to study in a statistically designed study of the effect of varying manufacturing process parameters, such as in evolutionary operation. 

In some cases, an actual process variable (such as yield) can be the objective function, and no process model is required. Instead, the process variables are varied systematically to find the best value of the objective function from the specific data set, sometimes involving design of experiments as discussed by Myers and Montgomery (2002). In this way, improvements in the objective function can be obtained gradually. Usually, only a few variables can be optimized in this way, and it is limited to batch operations. Methods used in industrial batch process applications include EVOP (evolutionary operation) and response surface analysis (Edwards and Jutan, 1997 Box and Draper, 1998 Myers and Montgomery, 2002). 

The evolutionary process of a genetic algorithm is accomplished by genetic operators which translate the evolutionary concepts of selection, recombination or crossover, and mutation into data processing to solve an optimization problem dynamically. Possible solutions to the problem are coded as so-called artificial chromosomes, which are changed and adapted throughout the optimization process until an optimrun solution is obtained. 

---