- By
**deion** - Follow User

- 570 Views
- Uploaded on

Download Presentation
## PowerPoint Slideshow about 'MATLAB Feedback' - deion

**An Image/Link below is provided (as is) to download presentation**

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript

### DEMO

- New Block Diagram
- Demo
- New Requirements
- Tradeoffs and other Issues
- Questions / suggestions

D. Fairley & J. Wu

D. Fairley & J. Wu

- Additional Feedbacks
- Bunch Charge
- Longitudinal will expand to include BC2 Energy and BC2 Bunch Length
- No DL2 energy for now, see SLC BPMs below
- L2 and L3 Launch
- Creating new feedback loops
- Who is the target user? User needs to be familiar with EPICS databases
- I suggest a team; physicist+controls sw
- Reading SLC BPMs
- Longer response time
- Throwaway code
- We have 3 EPICS BPMs for longitudinal; allows 5x5 fully EPICS
- Controlling SLC Magnets
- There was discussion about leaving the correctors under SLC control, what are the tradeoffs there and is this a real option?

D. Fairley & J. Wu

- Automatically switch configurations
- MATLAB script must read these conditions from PVs and make decision on each iteration of the algorithm, since it cannot be ‘interrupted’; a cost in processing time
- not ‘standard’ across feedbacks. May not be an issue now, but specialization means maintenance difficulties, and implementation difficulties looking ahead to fast-feedback
- Runtime adjustment of gains, weighting factors, mask matrix
- MATLAB script must read and apply these parameters from PVs at each iteration of the algorithm, whether they’ve changed or not. Also not ’standard’, see above.
- Cascading Feedbacks
- Longitudinal already satisfies this
- Is cascading required for transverse?

D. Fairley & J. Wu

- Introduce a Mask Matrix MM
- The user can design a specific feedback, may be needed in commissioning stage
- The Mask Matrix can be multiplied on the Gain Matrix MG
- Minimum modification to existing longitudinal algorithm
- Performs unnecessary calculation to get the correction on the actuator
- Yet, the correction is not applied (ie. Becomes 0 change)
- The Mask Matrix can be “multiplied” on getting the measurement, so to stop the measurement and to stop unnecessary calculation
- Yet, the framework stays the same

D. Fairley & J. Wu

LCLS – Accelerator and Bunch Compressor Schematic

D. Fairley & J. Wu

Cascade Loops: Injector, L2, and L3 Launch

- Identify transport matrices from loop i (=1 … n) to loop j (=i+1 … n)
- Based on Model? How accurate is the model, if the n—loop and n+1—loop are separated too far physically?
- We can measure the transport matrices
- Injector launch is commissioned and worked well
- For each stage n, similar implementation as that for injector launch
- However, the measurement should subtract what comes from the upstream loop disturbance
- If n is too large, this becomes cumbersome
- Minimum modification, and the framework stays the same

D. Fairley & J. Wu

Download Presentation

Connecting to Server..