Prediction of cut size for pneumatic ­classification based on a back propagation (BP) neural network

A new method is provided for predicting the cut size of a turbo air classifier using a BP neural network and a three layer BP neural network prediction model for cut size based on the training samples has been established. The MATLAB neural network toolbox can be used conveniently to calculate the prediction model. Verification results from the testing samples indicate that the predicted cut size using this BP neural network model is closer to the experimental cut size than the calculated cut size using a theoretical formula from other findings.

Cut size is one of the most important classification performance indices for powder classification equipment. To a certain extent it reflects the fineness of the product. For the same structure the cut size is determined mainly by such operating parameters as air inlet velocity, feed rate, and rotary speed of the rotor cage. A three layer (BP) neural network prediction model for calculating the cut size was introduced with the aim of predicting the cut size for pneumatic classification accurately using a turbo air classifier as an example. The calcium carbonate classification experiments were...

Related articles:


Universal classifier head for robust and efficient classification of soft to medium-hard materials

Hosokawa Alpine will present a flexible solution for the ultrafine classification of different soft to medium-hard materials, namely the TTD classifier head. Fineness values to x97 < 3 ?m at high...

Issue 2017-5 RETSCH GMBH

New XL cutting mill accepts large sample pieces and generates high throughput

Retsch have added a new model to their cutting mill family: the SM 400 XL is suitable for primary size reduction of very large sample pieces measuring up to 17 x 22 cm, but can also achieve the...

Issue 2018-6 WO 2018/061239 A1

Prediction method for cement quality or manufacturing conditions

(22) 03.02.2017 (43) 05.04.2018 (57) Provided is a method by which cement quality can be predicted in a short time and with a high degree of accuracy. In a prediction method for cement quality or...


Mill and classifier operation: The Pfeiffer Seminar 2010

Following the very successful realization of our last process-technological training courses in the past, customers are welcome to participate in our next three days intensive training held at...

Issue 2011-12

Correct analysis by representative sample division

A faultless and comparable analysis is closely linked to accurate sample handling. Only a sample representative of the initial material can provide meaningful analysis results. This can be achieved by...