Learning The Neural Network - Siemens SINUMERIK 840D sl Function Manual

Figure 5-23
5.6.3.3

Learning the neural network

Learning process sequence
A certain type of response is impressed upon the neural network during the learning phase.
The relation between the input and output signals is learnt.
The learning process is controlled entirely by NC programs and is divided into the following
areas:
1. Preset the QEC system variables for the learning process
2. Activate QEC system variables (by starting the NC program)
3. Parameterize the learning cycle
4. Start the learning cycle
The result is the learnt compensation characteristic that is stored in the non-volatile user
memory.
The results achieved must be checked using the circularity test (Section "Circularity test").
Extended Functions
Function Manual, 09/2009, 6FC5397-1BP20-0BA0
Effect of fine quantization with "Detailed learning" = active
5.6 Friction compensation (quadrant error compensation)
K3: Compensation
429