A Telecommand Verification of KOMPSAT2 using Rulebased and Casebased Reasoning

1
A Telecommand Verification of KOMPSAT-2 using
Rule-based and Case-based Reasoning

• Hee- Sook Mo, Won-Chan Jung, and Jae-Hoon Kim
• presenter wcjung_at_etri.re.kr 82-42-860-5367
• Electronics and Telecommunications Research
  Institute
• The SpaceOps 2002 Conference, Oct. 9 - 12, 2002

2
Contents

• Introduction
• Knowledge-Based System
• KOMPSAT-2 Telecommand Processing
• Design of K2 Automatic TC Verification
• Prototyping Implementation and Results
• Conclusion

3
Introduction (1)

• Satellite Control System
• Mission Analysis and Planning
• Control and Monitoring the Station Equipment
• Monitoring the Satellite State-of-health using
  Telemetry Data
• Preparation and Transmission of Telecommands for
  Mission Operation
• Archiving and Trend Analysis of State of health
• Satellite Simulation

4
Introduction (2)

• Satellite Control
• Repeated operations on every pass
• Emergency control if necessary
• Performed by transmitting telecommands with
  operator support
• Restrictions
• Many telemetry values changes on real-time rate
• Restricted monitoring screen
• Hard to monitor and control at the same time

5
Introduction (3)

• Solution
• Automation to support the operators operation

6
Operations to be Automated (1)

• Off-line operations to be automated
• Preparation of mission planning and scheduling
• Data backup
• File handling
• Start/stop control scheduling of ground station

7
Operations to be Automated (2)

• Real-time operations to be automated
• Monitoring of the status-of-health
• Emergency notification to operator
• Automatic transmission of the telecommand on
  reserved time
• Verification of the telecommand execution

8
Knowledge-Based System (1)

• Rule-Based Reasoning RBR
• Store the knowledge of experts into the rule base
• Rule-based inference for solving the problem
• Advantage
• Easy to express declaration knowledge module
• Disadvantages
• Hard to acquire the rule that explains the real
  domain correctly without exception
• Time consuming for the experts to convert the
  knowledge into rule
• Bottle-neck for acquisition of knowledge

9
Knowledge-Based System (2)

• Case-Based Reasoning CBR
• Store the cases that are used for solving the
  problem in the past into case base
• Use the case to solve the new problem
• Disadvantages
• Should maintain the index for cases and Searching
  time cost is high
• Need to have many cases in the case base to be
  effective

10
Knowledge-Based System (2)

• Hybrid System
• Combined the rule-based and case-based reasoning
• Apply rule-base first, and if the answer is not
  satisfactory, then apply case-base
• Can compensate the disadvantages of the methods

11
KOMPSAT-2 (KOrea Multi-Purpose SATellite-2)

• Second Satellite of KOMPSAT Series
• Earth Observation Satellite
• Cartography of Korean Peninsula
• Ocean Color Monitoring
• Ionospheric Information Collection
• KOMPSAT-1 Launched in 1999
• KOMPSAT-2 Will Be Launched in 2004
• Development of KOMPSAT-2 Using KOMPSAT-1 Heritage

12
KOMPSAT-2 Telecommand Processing

• TC Processing
• TC Generation Using Command Plan
• Transmission after packetized
• Verification after transmission
• Verified with changed telemetry

13
KOMPSAT-2 Telecommand Verification

• Real-Time Command
• Transmission check using command counter
• Function check using related telemetry value
• Special Command
• Check the related command counter
• Non real-time Command
• On-board CIB buffer dump and check with CRC
• Check related command counter without CRC

14
Design of K2 Automatic TC Verification

• Adopts Knowledge-Based System
• Rule-Based
• If-then-else Telemetry verifier
• Real-Time and Special Command
• Case-Based
• an Exception handling of rule-based reasoning
• Usage of the knowledge gained from KOMPSAT-I
  operation
• Non Real-Time Command, Fault Diagnosis
• Data mining by Reasoning
• If no knowledge is compatible with rules then
  case based reasoning

15
Class View for TC Verification
16
Interaction Diagram for TC Verification
17
Prototyping on TC Verification

• Prototype Environment
• Adopts CLIPs to express knowledge-based system
• CLIPs is used as an embedded application of SOS
• XML is used to manage the database
• OOP language C
• Object Communication over the Orbix 2000

18
RBR Prototyping

• To Check the Rule-Based Reasoning
• TC Verification Pattern
• Telemetry Items 200/second
• Rules in Knowledge-Base 50
• Syntax of the rules
• (defrule tc-cond-check
• (send-tc TC1)
• (post-TM1-value 4)
• (post-TM3-value 10)
• (assert(tmcheck-true)))

Patterns
19
RBR Prototyping Implementation

• Test
• The various number of Facts and Rules CLIPs
• Performance
• Pattern Matching(condition) Fact Assert
• Rule Activation Run the Rule

20
RBR Prototyping Results

• Result
• All of the telecommand verification with CLIPs
  finished within the 0.2 second.
• Telemetry frame rate is 1 second, so time for
  reasoning is acceptable to verify for KOMPSAT-2.

21
CBR Prototyping

• To Check the Case-Based Reasoning
• Start with the data in table form
• Generate Rules from the Cases

• Rules
• TC 1 ?TM-A, TM-C
• TC 2 ?TM-A, TM-B
• TC 3 ?TM-B, TM-C

Cases
?
22
Conclusion (1)

• Described the possibility of adopting TC
  verification using knowledge based system.
• The time for rule-based reasoning compared to
  total TC processing is acceptable.
• Case-based reasoning produced the rules that can
  be used in CLIPs.
• To improve the performance and flexibility,CLIPs
  will be embedded in a system as an object.

23
Conclusion (2)

• Future Work
• Development of knowledge-based system for
  KOMPSAT-2
• Researched on customized pattern matching
  strategy for KOMPSAT-2
• Decision strategy between cases and searching
  rule
• Cases learning, and so on.