Control Technologies at ESOC: Current Projects and Future Perspectives - PowerPoint PPT Presentation

Loading...

PPT – Control Technologies at ESOC: Current Projects and Future Perspectives PowerPoint presentation | free to download - id: 11366a-ODliN



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Control Technologies at ESOC: Current Projects and Future Perspectives

Description:

Goal: take a picture of the region X of the Earth in a certain window time [t1, t2] ... control processes in terms of cost efficiency and augmented functionalities: ... – PowerPoint PPT presentation

Number of Views:103
Avg rating:3.0/5.0
Slides: 27
Provided by: ad18
Learn more at: http://www.informatik.uni-ulm.de
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Control Technologies at ESOC: Current Projects and Future Perspectives


1
Control Technologies at ESOC Current Projects
and Future Perspectives
  • European Space Operations Centre of ESA
    Darmstadt, Germany

Contact Point Alessandro.Donati_at_esa.int (TOS-OSC
Control Technologies Unit _at_ ESA/ESOC)
PLANET Technologietag 16.6.2003, Ulm
2
Content
  • Definition of Terms
  • Effectively Introducing Innovation
  • The Motivation
  • The Environment
  • The Methodology
  • TOS-OSC Modus Operandi
  • Problem Cases Recent Developments
  • ENVISAT Gyro Monitoring Tool
  • Optimal INTEGRAL Reaction Wheels Bias Manoeuvre
  • XMM/INTEGRAL Radiation Monitoring Operational
    Adjustment
  • PROBA Autonomous (Re-)Scheduler
  • Prototype Scheduler for Ground Station Operations
  • Expected Benefits and Lessons Learnt
  • Medium Long Term Vision

3
Definitions of Terms
  • Mission Control Processes include
  • Planning Scheduling,
  • Monitoring, Diagnostic Control
  • Resource Management Off-line Analysis
  • Simulation Training
  • Mission Control RD Process is
  • the process of efficiently and effectively
    introducing innovation in specific Mission
    Control Processes where this is justified and
    needed.
  • Mission Control Teams include
  • Flight Control Team (spacecraft)
  • Ground Control Team (ground segment)
  • Flight Dynamics Team

4
Effectively Introducing Innovation Motivation
  • Enhancement of the overall Mission Control system
    performance
  • For meeting increasing demands (new functions)
    from ongoing and future missions
  • For reducing cost and/or risk
  • Contribution to the modernisation of ESOCs
    mission control approach
  • Provide European and National Mission Control
    Centres Innovative and Validated Technical
    Solutions to Specific Problems

5
Effectively Introducing Innovation Automation of
Control Processes
  • Mission Control Processes encompass
  • Humans
  • Machines (Hardware Software)
  • Procedures
  • Innovation in Mission Control Processes can
    support automation of routine activities

Humans
Machines
Procedures
6
Effectively Introducing Innovation the
Environment
  • Mission Control Teams members are the users and
    final beneficiaries of the improvements provided
    by the Mission Control RD Process
  • Mission Control Teams are very sensitive to
    potential risks introduced by system changes,
    for obvious reasons related to the criticality of
    spacecraft operations
  • Mission Control Teams members devote marginal
    manpower resources to support innovation due to
    current limitation of available manpower
  • Availability of financial resources is very
    limited

7
Effectively Introducing Innovation the
Methodology
Approach derived from the Dynamic System
Development Method() and adapted to our
environment
  • Iterative incremental prototyping
  • Competitive selection of User-defined cases
  • User is part of the development team
  • Frequent time-fixed deliveries with features
    implemented according to priorities negotiated
    for each time-box()
  • Iterative risk assessment
  • Maximum re-use of available resources (open
    source S/W, infrastructure)
  • Scalable solutions

() see next chart
() http//www.dsdm.org
8
Effectively Introducing Innovation the
Methodology
DSDM Time Boxing
Development slots fixed in allocated time and
resources variable in implemented
functionalities Time box content priorities is
negotiated at each iteration.
  • Must haves are essential, the minimum usable
    subset, without them the objectives are missed
  • Should haves are really needed, but when you miss
    them you can define a workaround
  • Could haves still included, but you can easily do
    without them
  • Won't haves Not enough value to include in this
    increment of development (these are usually
    greater than 40)

() Time boxing is a technique based on the fact
that over 40 of custom built software is NEVER
used. This together with the fact that most
projects traditionally deliver too late to
address the needs they were designed to address.
9
Effectively Introducing Innovation TOS-OSC modus
operandi
Future MissionsStudy Teams
Project Case
RD Spin-in Universities Industry
Technology
Flight/Ground Control Teams Project Teams
Conferences Seminars
Prototype Implementation
In-houseLectures,Training
Operational Validation
ProofedSolution
  • Library of reusable solutions, algorithms and
    techniques

Infrastructure
10
Case 1 ENVISAT Gyro Performance Monitoring Tool
  • PROBLEM enhance the capability to monitor
    ENVISAT gyros behaviour, detect the anomaly
    signatures at an earlier stage, when no standard
    alarming (OOL or FDIR) is yet triggered
    automate the reporting process.
  • EXPECTED BENEFITS automate the gyro monitoring
    tasks, early detection of potential degradation
    patterns, smoothing the possible gyro replacing
    process w/o affecting spacecraft payload
    productivity and reducing operators stress level
    vs. sudden unexpected degradations
  • IMPLEMENTED SOLUTION operational prototype
    making use of past ERS-1/2 operational
    experience, coded with fuzzy logic diagnostic
    inference engine off-line history database for
    all ENVISAT gyros data
  • STATUS currently under extended operational
    validation in March 03 a correct and punctual
    detection of a slight noise increase affecting
    gyroscope 1 measurements validated the
    capability of the tool.

11
Case 1 ENVISAT Gyro Performance Monitoring Tool
12
Case 2 Optimal XMM/INTEGRAL Reaction Wheels
Bias Manoeuvre
  • PROBLEM enhance the current optimisation
    approach for identifying the initial reaction
    wheels speed at the beginning of each orbit
    (perigee), able to support all scheduled
    observations, at minimum resource usage.
  • EXPECTED BENEFITS save onboard fuel, extend
    mission lifetime increase mission return
  • IMPLEMENTED SOLUTION operational prototype with
    data import export for FD formats, for XMM
    INTEGRAL missions. Optimisation algorithms using
    either classic genetic algorithm or
    multi-objective genetic algorithm
  • STATUS initial operational validation completed
    with equivalent fuel saving of around 35, using
    multi-objective GA (XMM and INTEGRAL cases)
    Flight Dynamics plan to use the tool in middle
    2003 for INTEGRAL.

13
Case 2 Optimal XMM/INTEGRAL Reaction Wheels
Bias Manoeuvre
14
Case 3 Radiation Monitoring Operational
Adjustments
  • PROBLEM enhance the capability to monitor
    INTEGRAL/XMM radiation environment and support
    operational decision making process for payload
    sensors reconfigurations.
  • EXPECTED BENEFITS reduce uncertainty gap of
    instrument operability in heavy radiation
    conditions forecast short term radiation level
    evolution (e.g. Van Allen crossing, solar flares
    waves) enhance spacecraft safety and
    productivity levels
  • reduced instrument exposure risk, increased
    observation return
  • IMPLEMENTED SOLUTION initial prototype making
    use of on-board radiation history and real-time
    data, complemented by external measurements
    (NOAA) forecast engine based on dynamic
    numerical modelling techniques and artificial
    neural network
  • STATUS version for XMM delivered and under
    acceptance phase on-going fine tuning of the
    implemented algorithm human-machine interface.

15
Case 4 PROBA Autonomous (Re-) Scheduler
  • PROBLEM Allow the on-board data handling
    subsystem to constantly monitor the successful
    execution of scheduled tasks and in case of
    resources unavailability or new activity requests
    autonomously reschedule the tasks on queue,
    respecting the stated constraints.
  • EXPECTED BENEFITS introduce a higher level of
    on-board autonomy, increase the spacecraft
    productivity
  • IMPLEMENTED SOLUTION ground based dynamic
    scheduler prototype with conflict detection,
    multi criteria decision making capability and
    dynamic context-sensitive ranking (conflict
    resolution)
  • potential for upgrading to an on-board software
    implementation and to validate it within the
    extended Proba operational lifetime.
  • STATUS final prototype delivered to Redu.
    Operational validation campaign due to start.

16
Case 4 PROBA Autonomous (Re-) Scheduler
  • Implementation of an onboard smart scheduler
    which can
  • Allocate activities to satisfy a set of given
    goals.
  • Identify and solve conflicts between resources
    and housekeeping or requested activities with no
    person in the loop.
  • Reschedule activities, whenever necessary, by
    working in almost real-time.
  • Requisites the final allocation must always
  • Be consistent with current temporal and resource
    constraints.
  • Converge in a finite time.

17
Case 4 PROBA Autonomous (Re-) Scheduler
  • Users Goal users can specify high level goals
    and the system should be able to achieve them
    considering all the resource constraints.
  • Example
  • Goal take a picture of the region X of the Earth
    in a certain window time t1, t2.
  • Output
  • Reschedule the pending activities so the new
    request can be scheduled.
  • If the memory is full plan first a downlink to
    dump it.
  • Perform attitude manoeuvre, lower vibration in
    the spacecraft, warm up the payload, make sure
    the required energy is available, etc

18
Case 4 PROBA Autonomous (Re-) Scheduler
  • Architecture
  • The knowledge base is pre-processed off-line.
  • New goals (user-defined or self-defined) are
    inserted in the current scheduling definition.
  • A first allocation of the activities is done
    using constructive methods (CSP). Usually the new
    scenario is over constrained, so a cost function
    is used to guide the search.
  • Then, the conflicts are solved using
    Multi-Attribute Decision Making (MADM) and
    Approximate Reasoning (fuzzy logic).
  • After that, the state transitions are checked to
    obtain a consistent scheduling scenario.
  • The new scheduling scenario is executed.

19
Case 4 PROBA Autonomous (Re-) Scheduler
  • Current Status
  • Final prototype delivered to Redu. Operational
    validation campaign due to start.
  • Potential Future Work
  • Design and implementation of an onboard
    autonomous scheduler

20
Case 5 Prototype Scheduler for Ground
Station Operations
  • The Kiruna Ground Station tracks several
    spacecrafts using two antennas and ground station
    connection equipments.
  • Every mission team issues a request to the Kiruna
    ground station to book a certain number of
    temporal slots.
  • The mission team is aware of the next passes.
  • Every mission team is totally unaware of the
    requests issued by the other mission teams and
    issues its request as if it was the only user of
    the ground station services.
  • Users Goal check that all requests are
    satisfied and produce a schedule completely free
    of resource conflicts. If this is not possible,
    it allocates first the activities with higher
    priority.

21
Case 5 Prototype Scheduler for Ground
Station Operations
  • The resource conflict detection module checks the
    existence of conflicts in the proposed schedule.
  • The detected conflicts are solved using two
    different approaches
  • Back-Tracking Approach
  • Allocate activities until there is a conflict. In
    this case it goes one step back and tries with a
    different activity. If none of the remaining
    activities produces a feasible schedule it goes
    another step back, and so on.
  • Slow algorithm tries every possibility with the
    use of heuristics to improve the performance.

22
Case 5 Prototype Scheduler for Ground
Station Operations
  • Genetic Algorithms Approach
  • Artificial intelligence technique based on
    natural evolution.
  • Codification (potential solution) ordered list
    of the activities to perform in the next slots
    (schedule).
  • Optimization minimize the sum of priorities of
    the activities not allocated.
  • Mutation swaps activities or groups of
    activities.
  • Fast algorithm in finding the best solution,
    however it is not guaranteed that every
    possibility is tried.

23
Case 5 Prototype Scheduler for Ground
Station Operations
  • Current Status
  • Initial prototype finalized and tested with
    operational data.
  • Potential Future Work
  • Design and implementation of a conflict detection
    and resolution module for an integrated ground
    station planning and scheduling tool

24
Expected Benefits Lessons Learnt
  • Artificial Intelligent techniques CAN provide
    benefits in improving Mission Control Processes
    in
  • efficiency
  • capabilities
  • A major area is decision making process in
    existence of unsharp input parameters or activity
    conflicts
  • User-driven fast iterative prototyping
    operational prototype final delivery are
    instrumental to bridge the gap between Academic
    world and Operational world
  • Facilitate focusing on the highest priority
    functions
  • Enable operational use of risk mitigation
  • Rationalise use of limited resources
  • Availability of historical data is often a
    pre-requisite

25
Medium Long-term Vision
  • Positive experience and encouraging results
    generate expectation of increase of Project Cases
    in number and complexity
  • In 5-year time we expect to have a consolidated
    class of solved problem cases to become an
    infrastructure asset ready for re-use
  • Expected increase of level of automation and
    performance of ground systems, at acceptable risk
  • Expected integration of currently split
    functional systems
  • Migration of proven and validated intelligent
    solution from ground to space augmented on-board
    autonomy capability

26
Conclusions
  • The European Space Operations Centre of ESA is
    pursuing continuous improvement of its mission
    control processes in terms of cost efficiency and
    augmented functionalities
  • Artificial intelligence and advanced control
    technologies play a significant role in specific
    problem cases
  • Positive measurable results provide comfort in
    further exploitation of artificial intelligence
    to serve mission control processes

Thank you for your interest !
Feedback alessandro.donati_at_esa.int
About PowerShow.com