i want to control the speed of IM using ANFIS. i gathered the data for training by simulating the block using the conventional PI controller with out load and i trained these data then i develop the neuro-fuzzy controller.it is expected to track the same as without load sine am using artificial intelligence controller but when adding a load torque the speed is not tracking the reference speed unlike without load. can any one help me to fix this problem?
thank you
Hi If your speed is not folowed according to the desired one, means your load torque is higher than the motor torque. You can also check about the Kp you designed using neurofuzzy inference editor. You should increase Kp when you have load torque. You can also design the controller with the load load. Regards, Rabia
1) one solution is u can use adaptive pid controller.
2) train the neural network(u r using supervised learning so give the sufficient data and train the network) with and without load torque.
3) Increase the gains of the controller or tune PI gains until it reaches desired performances.
4) load torque plays major role in the speed and torque so consider the load torque and collect the data and train the network.
- Badges
- Science topic
- Similar topics
- Statistics
- Survey