7
MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus Andrew Seltzman Dec 14, 2010

MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

  • Upload
    boris

  • View
    32

  • Download
    0

Embed Size (px)

DESCRIPTION

MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus. Andrew Seltzman. Dec 14, 2010. Background and Project Description :. - PowerPoint PPT Presentation

Citation preview

Page 1: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

Andrew Seltzman

Dec 14, 2010

Page 2: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

Background and Project Description :

MST is a plasma physics experiment currently running in the physics department. MST operates in a pulsed mode, taking plasma shots that last approximately 40-60ms. Before the shot, gas input at 15 different points in time during the shot and plasma current is set. Density values drift from one shot to the next, requiring operator input to adjust the gas input profile.

Currently a human operator is required to manually set gas input valves to stabilize density.

Density fluctuates around desired level due to human error

No automatic control system exists due to the complexity and non-linearity of the system in question

A MLP will be designed to emulate and eventually replace the skilled human operator.

Page 3: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

ANN Control System Setup:

• Control system given system parameters and requested output• Error computed from actual output• BP learning

– Eventual convergence ?

MST

(system to control)

Density

Current

Feature space compression

MLP

Gas input

Human input

Requested density

Error

Page 4: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

MLP Design:

• Simplest MLP required to adequately classify output data– 40 element feature space

• <= 40 elements in first hidden layer• <= 2 hidden layers

– 20 element output layer• Pre-processing of raw date to compress feature space prior to MLP input• BP training algorithm

– Initially train MLP from human operator responses– Eventual self learning when integrated into control system

• Calculation of error (requested density – actual density)• BP learning to allow real time adaptability to varying system

conditions• Optimization of initial training parameters and network design in progress

– Number of layers, Number of elements– Training parameters: learning rate, momentum, etc…

Page 5: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

Feature Space Data:

0 2 4 6 8 10 12 14 16 18 20-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

1.2co2

0 2 4 6 8 10 12 14 16 18 200

1

2

3

4

5

6

7

8

9

10gas

0 2 4 6 8 10 12 14 16 18 20-0.5

0

0.5

1

1.5

2

2.5

3

3.5ip

0 1000 2000 3000 4000 5000 6000 7000 8000-50

0

50

100

150

200

250

300

350ip

0 2 4 6 8 10 12 14 16 18 200

1

2

3

4

5

6

7

8

9

10gas

0 1000 2000 3000 4000 5000 6000 7000 8000-0.5

0

0.5

1

1.5

2

2.5co2

Initial Feature Space

Compressed Feature Space

•High speed digitizer captures experiment data

•30000 data points for Ip and Ne

•8000 actual relevant data points

•21 data points for Gas input

•ANN complexity with ~60000 input neurons would be wasteful of computing resourced

•Real time operation required (~1 update every 2 min)

•Feature space compressed by averaging points in a given data set

•20 data points for Ip and Ne

•2 data points for Gas input

•Relevant data extracted without loss accuracy

Page 6: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

Expected Results / Initial Testing:

• MLP given requested plasma density and plasma current

• System successfully generates gas waveform

• Gas waveform very similar to training set• Not integrated with actual system

– Gas output appears similar to human operator

• Accurate model?• Further testing / training required with

multiple data sets

0 5 10 15 200

0.5

1gas

0 5 10 15 200

100

200

300testip

0 5 10 15 200

0.5

1testco2

Page 7: MLP Based Feedback System for Gas Valve Control in a Madison Symmetric Torus

Discussion / Problems / Future Modifications:

• Control method– Output types:

• Output gas values based on requested density– Larger training data set

» Data available from every shot– Known density result

• Output change in gas valued to error in density– Smaller training data set

» Data only when operator changes gas settings– Requested density value not known

• Over specification of fitted data– Training sometimes results of MLP copying human gas output regardless of

density