Below follows a couple of example scenarios with a description of how db Collection can help solving the problem. These are only two examples of the vast array of applications were

db Collection can be used.

 

 

Industrial application - 1st example

 

db Collection contains intuitive and powerful tools that can be used for a vast array of applications. One example is a typical control problem in a paper mill where a quality parameter fluctuates outside the accepted tolerances. There is a large number of parameters that might affect the quality and it is not clear which are the most important. The goal is to identify the most important parameters and propose a way to improve their feedback control loops in order to reduce the quality variations. The db Collection workflow is described below:

 

• Identify all parameters that might affect the process.

 

• Collect measurement data from these parameters during normal operating conditions.

 

• Use the data to create a numerical model of the process in db Model Builder.

 

• Calculate correlations between different signals in the model with db Analyzer. In this way the most important parameters can be identified. Unimportant parameters can be removed in order to simplify the model.

 

• Changes in the process can now be tested in the model with db Simulator instead of in the real System.

 

• When a promising solution has been found it can be tested on the real process.

 

The result of the above procedure is a substantial cost reduction. Production loss can be minimized since most changes can be tested on the numerical model instead of on the real process. The numerical approach is also much less time consuming so more tests can be performed which improves the final quality of the solution.

 

 

 

Industrial application - 2nd example

 

A product in a processing plant has to be produced with a specified quality. The goal is to minimize the consumption of certain chemicals but still satisfy the quality requirements. The

db Collection workflow is described below:

 

• Identify the input and output signals of the system.

 

• Collect data under normal operating Conditions.

 

• Build a model in db Model Builder.

 

• Determine which inputs can be controlled.

 

• Define operating limits on output signals and controllable input signals.

 

• Run db Optimizer.

 

The result from db Optimizer is a set of optimized signals which suggest how the process can be improved. If necessary, the above procedure can be iterated but the total number of real tests can be sharply reduced compared with a fully manual approach. This will shorten lead times, increase profits, and reduce the environmental impact.

 

A graph showing the original quality vs. optimized quality.