Size: 92
Comment:
|
Size: 5172
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 3: | Line 3: |
=== Kick-off meeting === | == Project definition == === Deliverables of the work === A document to answer the following questions 1. Does the expected actuator and sensor noise level satisfy the KAGRA requirement ? 2. Can we lock the interferometer with the modified actuators ? 1. Is lock acquisition possible ? (ignore noise) 2. Can we keep the lock with weak damping ? i.e., not injecting excess control noise from strong damping 3. Can we damp all the resonant modes of the CRY payload within reasonable time (like 1 minutes) ? === Methods === * Create a combined model of SUMCON and Simulink for the Type-A suspension. * The model includes a LTI version of suspension mechanical model and feedback loops around it. * Copy the current control topology into the baseline model * Try to reproduce the current behavior of the suspensions (validation) * Update the baseline model to the modified CRY payload * Optimize the control topology and filters to find a configuration which allows us to control the suspension without saturating the actuators and not injecting excess noise to DARM === Assignment === * Create a baseline model: Shoda, Ushiba, Miyo * Optimization of the control topology and filters: Ushiba, Yamada, Nakano, Shoda, Miyo, Terrence * Give advice from the viewpoint of interferometer: Nakano, Michimura === Deadlines === * End-of-June: Create a base model * End-of-July: Make a recommendation for the CRY payload modification ---- == Kick-off meeting == 2020/5/27 10:30 - 12:00 Zoom: KAGRA [[https://zoom.us/j/6676627462|Zoom2]] === Minutes === * The main work of this project is to create a simulation model of the suspension including the control loops, then optimize the control topology. * We also identified additional tasks related to this project ==== Simulation of the Type-A suspension to find optimum control strategy ==== * We will create a model (SUMCON + Simulink) to simulate the current Type-A suspension, called the baseline model. * Using the baseline model, we will try to reproduce the current suspension behavior: * Close the local damping loops with the current control topology and filteres * Inject known noises, such as sensor noise, actuator noise, seismic noise, green laser noise. * Check the actuator loads (saturated or not) and find out what is limiting the current performance. * After validating the baseline model, update it to the modified actuators and sensors * Optimize the control topology and filters to find a configuration which allows us to control the suspension without saturating the actuators and not injecting excess noise to DARM ==== Additional tasks ==== ===== Largest magnet possible ? ===== * It is relatively easy to modify the coil drivers, but not easy to change magnets in the vacuum. * Therefore, it may be better to install magnets as large as possible into the suspensions * We need to identify what is the limit, and is there any drawback from this. * Will start email discussion on this. ===== Actuator efficiency of the currently installed actuators ===== * Measure the in-situ actuation efficiency and compare it with the theoretical predictions or ex-situ measurements using scales * We can probably take the measurements by CAL of the actuation efficiency, which is in the unit of m/V, then convert it into N/A using the suspension mechanical model. ===== Seismic noise coupling of the PF actuator ===== * Coils for the new PF actuators will be mounted on the ground. * We need to estimate the seismic motion coupling of this actuator. * Measurement of the coupling using a scale will be necessary * Probably an action item for CRY ---- === Meeting Materials === ==== Current status and plan of CRY payload by Ushiba ==== [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=11669|Slides]] === Goal of the meeting === Decide the followings === Deliverables of the work === A document to answer the following questions 1. Does the expected actuator noise level satisfy the KAGRA requirement ? 2. Can we lock the interferometer with the modified actuators ? 2. Is lock acquisition possible ? (ignore noise) 2. Can we keep the lock with weak damping ? i.e., not injecting excess control noise from strong damping 3. Can we damp all the resonant modes of the CRY payload within reasonable time (like 1 minutes) ? === Methods === * Create a SUMCON model of modified CRY payload * Implement local damping loops on simulation * Check the damping efficiency of resonant modes * Implement a hierarchical control scheme on simulation * Check the controllability of the suspension * How can you evaluate the feasibility of lock acquisition ? === Assignment === * Create a SUMCON model: Shoda, Ushiba, Miyo * Damping and hierarchical control: Ushiba, Yamada, Nakano, Shoda, Miyo, Terrence * Set requirements from IFO: Nakano, Michimura === Deadline === * Mid-June: intermediate report on the progress at a meeting * End-of-June: final report (too aggressive ?) |
CRY payload actuator upgrade in 2020
Contents
Project definition
Deliverables of the work
A document to answer the following questions
- Does the expected actuator and sensor noise level satisfy the KAGRA requirement ?
- Can we lock the interferometer with the modified actuators ?
- Is lock acquisition possible ? (ignore noise)
- Can we keep the lock with weak damping ? i.e., not injecting excess control noise from strong damping
- Can we damp all the resonant modes of the CRY payload within reasonable time (like 1 minutes) ?
Methods
- Create a combined model of SUMCON and Simulink for the Type-A suspension.
- The model includes a LTI version of suspension mechanical model and feedback loops around it.
- Copy the current control topology into the baseline model
- Try to reproduce the current behavior of the suspensions (validation)
- Update the baseline model to the modified CRY payload
- Optimize the control topology and filters to find a configuration which allows us to control the suspension without saturating the actuators and not injecting excess noise to DARM
Assignment
- Create a baseline model: Shoda, Ushiba, Miyo
- Optimization of the control topology and filters: Ushiba, Yamada, Nakano, Shoda, Miyo, Terrence
- Give advice from the viewpoint of interferometer: Nakano, Michimura
Deadlines
- End-of-June: Create a base model
- End-of-July: Make a recommendation for the CRY payload modification
Kick-off meeting
2020/5/27 10:30 - 12:00 Zoom: KAGRA Zoom2
Minutes
- The main work of this project is to create a simulation model of the suspension including the control loops, then optimize the control topology.
- We also identified additional tasks related to this project
Simulation of the Type-A suspension to find optimum control strategy
- We will create a model (SUMCON + Simulink) to simulate the current Type-A suspension, called the baseline model.
- Using the baseline model, we will try to reproduce the current suspension behavior:
- Close the local damping loops with the current control topology and filteres
- Inject known noises, such as sensor noise, actuator noise, seismic noise, green laser noise.
- Check the actuator loads (saturated or not) and find out what is limiting the current performance.
- After validating the baseline model, update it to the modified actuators and sensors
- Optimize the control topology and filters to find a configuration which allows us to control the suspension without saturating the actuators and not injecting excess noise to DARM
Additional tasks
Largest magnet possible ?
- It is relatively easy to modify the coil drivers, but not easy to change magnets in the vacuum.
- Therefore, it may be better to install magnets as large as possible into the suspensions
- We need to identify what is the limit, and is there any drawback from this.
- Will start email discussion on this.
Actuator efficiency of the currently installed actuators
- Measure the in-situ actuation efficiency and compare it with the theoretical predictions or ex-situ measurements using scales
- We can probably take the measurements by CAL of the actuation efficiency, which is in the unit of m/V, then convert it into N/A using the suspension mechanical model.
Seismic noise coupling of the PF actuator
- Coils for the new PF actuators will be mounted on the ground.
- We need to estimate the seismic motion coupling of this actuator.
- Measurement of the coupling using a scale will be necessary
- Probably an action item for CRY
Meeting Materials
Current status and plan of CRY payload by Ushiba
Goal of the meeting
Decide the followings
Deliverables of the work
A document to answer the following questions
- Does the expected actuator noise level satisfy the KAGRA requirement ?
- Can we lock the interferometer with the modified actuators ?
- Is lock acquisition possible ? (ignore noise)
- Can we keep the lock with weak damping ? i.e., not injecting excess control noise from strong damping
- Can we damp all the resonant modes of the CRY payload within reasonable time (like 1 minutes) ?
Methods
- Create a SUMCON model of modified CRY payload
- Implement local damping loops on simulation
- Check the damping efficiency of resonant modes
- Implement a hierarchical control scheme on simulation
- Check the controllability of the suspension
- How can you evaluate the feasibility of lock acquisition ?
Assignment
- Create a SUMCON model: Shoda, Ushiba, Miyo
- Damping and hierarchical control: Ushiba, Yamada, Nakano, Shoda, Miyo, Terrence
- Set requirements from IFO: Nakano, Michimura
Deadline
- Mid-June: intermediate report on the progress at a meeting
- End-of-June: final report (too aggressive ?)