first principles planning in bdi systems
play

First Principles Planning in BDI Systems Sebastian Sardina - PowerPoint PPT Presentation

Feb 06, 2023 •471 likes •1.5k views

First Principles Planning in BDI Systems Sebastian Sardina Department of Computer Science and Information Technology RMIT University, Melbourne, Australia Joint work with Lavindra de Silva and Lin Padgham 1 / 25 Motivation The B elief, D

  1. A Mars Rover BDI Agent ExploreSoilLocation (? dst ) Event-goal P 0 − → BDI plan Navigate (? src , ? dst ) PerformSoilExperiment (? dst ) Action P 1 − → Move (? src , ? dst ) Calibrate P 2 − → ObtainSoilResults (? dst ) TransmitSoilResults (? dst ) P 3 − → PickSoilSample (? dst ) AnalyseSoilSample (? dst ) DropSoilSample P 5 − → P 4 SendResults (? dst ) EstablishConn BreakConn − → GetMoistureContent (? dst ) GetSoilParticleSize (? dst ) 14 / 25

  2. A Mars Rover BDI Agent ExploreSoilLocation (? dst ) Event-goal P 0 − → BDI plan Navigate (? src , ? dst ) PerformSoilExperiment (? dst ) Action P 1 − → Move (? src , ? dst ) Calibrate P 2 − → ObtainSoilResults (? dst ) TransmitSoilResults (? dst ) P 3 − → PickSoilSample (? dst ) AnalyseSoilSample (? dst ) DropSoilSample P 5 − → P 4 SendResults (? dst ) EstablishConn BreakConn − → GetMoistureContent (? dst ) GetSoilParticleSize (? dst ) 14 / 25

  3. A Mars Rover BDI Agent ExploreSoilLocation (? dst ) Event-goal P 0 − → BDI plan Navigate (? src , ? dst ) PerformSoilExperiment (? dst ) Action P 1 − → Move (? src , ? dst ) Calibrate P 2 − → ObtainSoilResults (? dst ) TransmitSoilResults (? dst ) P 3 − → PickSoilSample (? dst ) AnalyseSoilSample (? dst ) DropSoilSample P 5 − → P 4 SendResults (? dst ) EstablishConn BreakConn − → GetMoistureContent (? dst ) GetSoilParticleSize (? dst ) 14 / 25

  4. A Mars Rover BDI Agent ExploreSoilLocation (? dst ) Event-goal P 0 − → BDI plan Navigate (? src , ? dst ) PerformSoilExperiment (? dst ) Action P 1 − → Move (? src , ? dst ) Calibrate P 2 − → ObtainSoilResults (? dst ) TransmitSoilResults (? dst ) P 3 − → PickSoilSample (? dst ) AnalyseSoilSample (? dst ) DropSoilSample P 5 − → P 4 SendResults (? dst ) EstablishConn BreakConn − → GetMoistureContent (? dst ) GetSoilParticleSize (? dst ) 14 / 25

  5. A Mars Rover BDI Agent ExploreSoilLocation (? dst ) Event-goal P 0 − → BDI plan Navigate (? src , ? dst ) PerformSoilExperiment (? dst ) Action P 1 − → Move (? src , ? dst ) Calibrate P 2 − → ObtainSoilResults (? dst ) TransmitSoilResults (? dst ) P 3 − → PickSoilSample (? dst ) AnalyseSoilSample (? dst ) DropSoilSample P 5 − → P 4 SendResults (? dst ) EstablishConn BreakConn − → GetMoistureContent (? dst ) GetSoilParticleSize (? dst ) 14 / 25

  6. Hybrid-Solutions in the Mars Rover Goal state = locations rock 2 and rock 3 have been explored. 1. Navigate(rock1, rock2) 2. PerformSoilExperiment(rock2) 3. Navigate(rock2, rock3) 4. PerformSoilExperiment(rock3) Hybrid-solution h 15 / 25

  7. Hybrid-Solutions in the Mars Rover Goal state = locations rock 2 and rock 3 have been explored. 1. Navigate(rock1, rock2) (A) Calibrate (B) Move(rock1, rock2) 2. PerformSoilExperiment(rock2) (A) ObtainSoilResults(rock2) 1. Navigate(rock1, rock2) (i) PickSoilSample(rock2) 2. PerformSoilExperiment(rock2) (ii) AnalyseSoilSample(rock2) 3. Navigate(rock2, rock3) (a) GetMoistureContent(rock2) 4. PerformSoilExperiment(rock3) (b) GetSoilParticleSize(rock2) (iii) DropSoilSample (B) TransmitSoilResults(rock2) (i) EstablishConn (ii) SendResults(rock2) (iii) BreakConn 3. Navigate(rock2, rock3) (A) Calibrate (B) Move(rock2, rock3) 4. PerformSoilExperiment(rock3) (A) ObtainSoilResults(rock3) (i) PickSoilSample(rock3) (ii) AnalyseSoilSample(rock3) (a) GetMoistureContent(rock3) (b) GetSoilParticleSize(rock3) (iii) DropSoilSample (B) TransmitSoilResults(rock3) (i) EstablishConn (ii) SendResults(rock3) (iii) BreakConn Hybrid-solution h Execution of h 15 / 25

  8. Hybrid-Solutions in the Mars Rover Goal state = locations rock 2 and rock 3 have been explored. 1. Navigate(rock1, rock2) (A) Calibrate (B) Move(rock1, rock2) 2. PerformSoilExperiment(rock2) (A) ObtainSoilResults(rock2) 1. Navigate(rock1, rock2) (i) PickSoilSample(rock2) 2. PerformSoilExperiment(rock2) (ii) AnalyseSoilSample(rock2) 3. Navigate(rock2, rock3) (a) GetMoistureContent(rock2) 4. PerformSoilExperiment(rock3) (b) GetSoilParticleSize(rock2) (iii) DropSoilSample (B) TransmitSoilResults(rock2) (i) EstablishConn (ii) SendResults(rock2) (iii) BreakConn ⇐ redundant 3. Navigate(rock2, rock3) (A) Calibrate (B) Move(rock2, rock3) 4. PerformSoilExperiment(rock3) (A) ObtainSoilResults(rock3) (i) PickSoilSample(rock3) (ii) AnalyseSoilSample(rock3) (a) GetMoistureContent(rock3) takes too long (b) GetSoilParticleSize(rock3) (iii) DropSoilSample (B) TransmitSoilResults(rock3) (i) EstablishConn ⇐ redundant (ii) SendResults(rock3) (iii) BreakConn Hybrid-solution h Execution of h 15 / 25

  9. Hybrid-Solutions in the Mars Rover Goal state = locations rock 2 and rock 3 have been explored. 1. Navigate(rock1, rock2) (A) Calibrate (B) Move(rock1, rock2) 2. PerformSoilExperiment(rock2) 1. Navigate(rock1, rock2) (A) ObtainSoilResults(rock2) 2. ObtainSoilResults(rock2) 1. Navigate(rock1, rock2) (i) PickSoilSample(rock2) 3. EstablishConn 2. PerformSoilExperiment(rock2) (ii) AnalyseSoilSample(rock2) 4. SendResults(rock2) 3. Navigate(rock2, rock3) (a) GetMoistureContent(rock2) 5. Navigate(rock2, rock3) 4. PerformSoilExperiment(rock3) (b) GetSoilParticleSize(rock2) 6. ObtainSoilResults(rock3) (iii) DropSoilSample 7. SendResults(rock3) (B) TransmitSoilResults(rock2) 8. BreakConn (i) EstablishConn ⇑ Non-redundant + maximally- (ii) SendResults(rock2) abstract (iii) BreakConn ⇐ redundant 3. Navigate(rock2, rock3) (A) Calibrate (B) Move(rock2, rock3) 4. PerformSoilExperiment(rock3) (A) ObtainSoilResults(rock3) (i) PickSoilSample(rock3) (ii) AnalyseSoilSample(rock3) (a) GetMoistureContent(rock3) takes too long (b) GetSoilParticleSize(rock3) (iii) DropSoilSample (B) TransmitSoilResults(rock3) (i) EstablishConn ⇐ redundant (ii) SendResults(rock3) (iii) BreakConn Hybrid-solution h ′ Hybrid-solution h Execution of h 15 / 25

  10. Creating Abstract Operators We translate event-goals into abstract planning operators by using an adaptation and extension of the “summary algorithms” of [Clement&Durfee,07]. 16 / 25

  11. Creating Abstract Operators We translate event-goals into abstract planning operators by using an adaptation and extension of the “summary algorithms” of [Clement&Durfee,07]. • The name of the event-goal becomes the name of the operator. • The disjunction of the context conditions of the event-goal’s associated plans are taken as the precondition of the operator. • The postcondition of an operator is computed based on the structure of the corresponding event-goal’s hierarchy, combined with knowledge about the effects of primitive actions. • In particular, we use the definite effects of event-goals as the postconditions of operators. 16 / 25

  12. Computing Preconditions and Postconditions e 3 Event-goal OR BDI plan φ 1 P 1 φ 2 P 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  13. Computing Preconditions and Postconditions e 3 φ 1 ∨ φ 2 Event-goal OR BDI plan φ 1 P 1 φ 2 P 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  14. Computing Preconditions and Postconditions e 3 φ 1 ∨ φ 2 Event-goal OR BDI plan φ 1 P 1 φ 2 P 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  15. Computing Preconditions and Postconditions e 3 φ 1 ∨ φ 2 Event-goal OR BDI plan P 1 ¬ r φ 1 φ 2 P 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  16. Computing Preconditions and Postconditions e 3 φ 1 ∨ φ 2 Event-goal OR BDI plan P 1 ¬ r φ 1 φ 2 P 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  17. Computing Preconditions and Postconditions e 3 φ 1 ∨ φ 2 Event-goal OR BDI plan P 1 ¬ r P 2 q ∧ ¬ r φ 1 φ 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  18. Computing Preconditions and Postconditions e 3 φ 1 ∨ φ 2 Event-goal OR BDI plan P 1 ¬ r P 2 q ∧ ¬ r φ 1 φ 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  19. Computing Preconditions and Postconditions e 3 ¬ r φ 1 ∨ φ 2 Event-goal OR BDI plan P 1 ¬ r P 2 q ∧ ¬ r φ 1 φ 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r 17 / 25

  20. Computing Preconditions and Postconditions e 3 ¬ r φ 1 ∨ φ 2 Event-goal OR BDI plan P 1 ¬ r P 2 q ∧ ¬ r φ 1 φ 2 → Action a 1 ¬ r a 2 q ∧ r a 1 ¬ r Abstract operators computed are used for first principles planning. Hybrid-plans obtained need to be validated because we do not take into account possible effects (e.g., q ) of event-goals. 17 / 25

  21. Basic Notions for Hybrid-Solutions Definition (Perfect Justification [Fink & Yang,‘92]) A primitive solution σ for some planning problem is a perfect justification if there does not exist a proper subsequence σ ′ of σ such that σ ′ is a primitive solution for the problem. 18 / 25

  22. Basic Notions for Hybrid-Solutions Definition (Perfect Justification [Fink & Yang,‘92]) A primitive solution σ for some planning problem is a perfect justification if there does not exist a proper subsequence σ ′ of σ such that σ ′ is a primitive solution for the problem. Definition (Non-redundant Hybrid-Solutions) A hybrid-solution is a non-redundant hybrid-solution if it can produce at least one perfect justification via decomposition. 18 / 25

  23. Basic Notions for Hybrid-Solutions Definition (Perfect Justification [Fink & Yang,‘92]) A primitive solution σ for some planning problem is a perfect justification if there does not exist a proper subsequence σ ′ of σ such that σ ′ is a primitive solution for the problem. Definition (Non-redundant Hybrid-Solutions) A hybrid-solution is a non-redundant hybrid-solution if it can produce at least one perfect justification via decomposition. Definition (Maximally-Abstract) A hybrid-plan h is maximally-abstract if there is no other hybrid-plan h ′ such that one can reach h from h ′ via decomposition, and h ′ produces any of the solutions of interest that are produced by h . 18 / 25

  24. Basic Notions for Hybrid-Solutions Definition (Perfect Justification [Fink & Yang,‘92]) A primitive solution σ for some planning problem is a perfect justification if there does not exist a proper subsequence σ ′ of σ such that σ ′ is a primitive solution for the problem. Definition (Non-redundant Hybrid-Solutions) A hybrid-solution is a non-redundant hybrid-solution if it can produce at least one perfect justification via decomposition. Definition (Maximally-Abstract) A hybrid-plan h is maximally-abstract if there is no other hybrid-plan h ′ such that one can reach h from h ′ via decomposition, and h ′ produces any of the solutions of interest that are produced by h . Ideal hybrid-solution = non-redundant + maximally-abstract 18 / 25

  25. Results Theorem 1 Whenever a hybrid planning problem can be solved, there is at least one ideal hybrid-solution. • Unfortunately, it is not clear how ideal hybrid-solutions can be computed in polynomial time. • Instead, we focus on improving a given hybrid-solution. • To this end, we extract a non-redundant and most abstract hybrid-solution from the decomposition tree. • An improvement of a hybrid-solution is called a preferred specialization. 19 / 25

  26. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  27. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  28. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  29. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  30. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  31. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  32. Example Consider initial state p and goal state t . e 0 e 5 Event-goal P 0 P 6 BDI plan − → − → e 1 e 2 s a 5 r s a 6 t Action P 1 P 2 − → p a 1 q ∧ u e 3 e 4 OR P 3 P 4 P 5 − → q a 3 r p a 2 q q a 3 r r a 4 s Hybrid-solution Non-redundant Maximally-absract Ideal √ √ √ e 2 · a 6 √ e 0 · e 5 × × √ e 1 · e 2 · a 6 × × e 0 · a 5 · a 6 × × × 20 / 25

  33. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  34. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  35. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  36. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  37. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  38. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  39. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 4 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  40. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 4 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  41. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 4 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  42. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 5 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  43. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 5 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  44. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 5 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  45. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 6 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  46. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 6 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  47. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 7 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  48. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 7 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  49. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 7 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  50. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  51. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 2 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  52. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 2 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  53. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 2 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  54. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 3 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  55. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 3 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  56. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  57. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 1 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  58. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . u= e 1 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  59. Preferred Specializations Find-Preferred-Specialization( h , H , T ) e 1 h= 1 Input : Hybrid-solution h ; hybrid problem H ; and ∈ π decomposition tree T of h . 2 Output : A preferred specialization of h w.r.t. H and T . 3 σ ⇐ leaf-level primitive solution in T . e 1 4 σ ′ ⇐ Get-Perfect-Justification ( σ, H ). 5 π ⇐ { a i | a i ∈ σ ′ } e 2 e 3 6 for ℓ ⇐ 1 to height ( T ) do e 4 e 5 e 6 e 7 • for each node u at level ℓ in tree T do if children ( u , T ) ⊆ π , then π ⇐ ( π \ children ( u , T )) ∪ { u } a 1 a 2 a 3 a 4 a 5 a 6 a 7 a 8 7 Return π 21 / 25

  60. Results Theorem 2 Given a perfect justification, the algorithm computes a preferred specialization in polynomial time on the size of the given decomposition tree. 22 / 25

  61. Results Theorem 2 Given a perfect justification, the algorithm computes a preferred specialization in polynomial time on the size of the given decomposition tree. Theorem 3 Whenever a hybrid-solution exists for some hybrid planning problem, there is at least one preferred specialisation for some decomposition tree of the hybrid-solution. 22 / 25

  62. Results Theorem 2 Given a perfect justification, the algorithm computes a preferred specialization in polynomial time on the size of the given decomposition tree. Theorem 3 Whenever a hybrid-solution exists for some hybrid planning problem, there is at least one preferred specialisation for some decomposition tree of the hybrid-solution. Theorem 4 Whenever an ideal hybrid-solution occurs in a decomposition tree of a given hybrid-solution h , and the ideal hybrid-solution yields a perfect justification in the decomposition tree, then the ideal hybrid-solution is guaranteed to be a preferred specialisation of h . 22 / 25

  63. Conclusion • We provided algorithms for converting BDI event-goals into (abstract) planning operators. • We characterized different notions of “goodness” for hybrid-plans, such as non-redundancy and maximal-abstraction. • We proved certain properties of “good” hybrid-plans, in particular, that ideal hybrid-plans are also preferred specializations in certain situations. • We provided algorithms for obtaining preferred specializations. 23 / 25

Recommend

PLANNING AND DESIGNING OPERATING SYSTEMS A 86012 Management and Principles of 1 Accoun:ng 8-2

PLANNING AND DESIGNING OPERATING SYSTEMS A 86012 Management and Principles of 1 Accoun:ng 8-2

PLANNING AND DESIGNING OPERATING SYSTEMS A 86012 Management and Principles of 1 Accoun:ng 8-2 Planning and Designing Operations Systems Opera:ons planning involves making the following decisions What will we produce? Who are our customers?

503 views • 13 slides
A Formal Approach to Embedding First-Principles Planning in BDI Agent Systems Mengwei Xu, Kim

A Formal Approach to Embedding First-Principles Planning in BDI Agent Systems Mengwei Xu, Kim

A Formal Approach to Embedding First-Principles Planning in BDI Agent Systems Mengwei Xu, Kim Bauters, Kevin McAreavey , Weiru Liu Agent Frameworks Programming-Based Model-Based Belief-Desire-Intention (BDI) Hierarchical Task Networks (HTN)

568 views • 20 slides
State Revenue Systems and State Revenue Systems and Principles of Tax Policy Principles of Tax

State Revenue Systems and State Revenue Systems and Principles of Tax Policy Principles of Tax

State Revenue Systems and State Revenue Systems and Principles of Tax Policy Principles of Tax Policy Blue Ribbon Tax Structure Commission Vermont General Assembly Bert Waisanen National Conference of State Legislatures September 1, 2009

1.2k views • 72 slides
SYSTEMS PLANNING UNIT CRESSON SLOTTEN, P .E., MANAGER CSLOTTEN@A2GOV.ORG SYSTEMS PLANNING UNIT

SYSTEMS PLANNING UNIT CRESSON SLOTTEN, P .E., MANAGER CSLOTTEN@A2GOV.ORG SYSTEMS PLANNING UNIT

SYSTEMS PLANNING UNIT CRESSON SLOTTEN, P .E., MANAGER CSLOTTEN@A2GOV.ORG SYSTEMS PLANNING UNIT Contact: Systems Planning City Hall - 4 th and 5 th Floor (734)794-6430 x 43701 a2gov.org/departments/systems-planning/ Long-range &

204 views • 17 slides
Classical Planning Systems ICS 271 Fall 2014 Outline: Planning Planning environments

Classical Planning Systems ICS 271 Fall 2014 Outline: Planning Planning environments

Set 9: Planning Classical Planning Systems ICS 271 Fall 2014 Outline: Planning Planning environments Classical Planning: Situation calculus PDDL: Planning domain definition language STRIPS Planning Planning graphs

677 views • 54 slides
Classical Planning Systems Chapter 10 R&N ICS 271 Fall 2016 Outline: Planning Planning

Classical Planning Systems Chapter 10 R&N ICS 271 Fall 2016 Outline: Planning Planning

Set 9: Planning Classical Planning Systems Chapter 10 R&N ICS 271 Fall 2016 Outline: Planning Planning environments Classical Planning: Situation calculus PDDL: Planning domain definition language STRIPS Planning SAT

927 views • 60 slides
Design Principles for Secure Systems Systems Driving Ideas for Security Principles Saltzer

Design Principles for Secure Systems Systems Driving Ideas for Security Principles Saltzer

1 Design Principles for Secure Systems Systems Driving Ideas for Security Principles Saltzer and Schroeder (1975) defined 8 principles that are based on the ideas of simplicity and restriction are based on the ideas of simplicity and

468 views • 17 slides
Set 9: Planning Classical Planning Systems Chapter 10 R&N ICS 271 Fall 2018 Outline:

Set 9: Planning Classical Planning Systems Chapter 10 R&N ICS 271 Fall 2018 Outline:

Set 9: Planning Classical Planning Systems Chapter 10 R&N ICS 271 Fall 2018 Outline: Planning Planning environments Classical Planning: Situation calculus PDDL: Planning domain definition language STRIPS Planning SAT

860 views • 59 slides
Chapter 5: I/O Systems Input/Output Principles of I/O hardware Principles of I/O software

Chapter 5: I/O Systems Input/Output Principles of I/O hardware Principles of I/O software

Chapter 5: I/O Systems Input/Output Principles of I/O hardware Principles of I/O software I/O software layers Disks Clocks Character-oriented terminals Graphical user interfaces Network terminals Power management

1.28k views • 63 slides
Facility Planning Committee Guiding Principles, Budgeting 101 & Learning January 31, 2017

Facility Planning Committee Guiding Principles, Budgeting 101 & Learning January 31, 2017

North Wasco Co. SD Facility Planning Committee Guiding Principles, Budgeting 101 & Learning January 31, 2017 tonights objectives 1. Review the Schedule 2. Group Activity Crafting the Guiding Principles 3. Condition assessment

914 views • 41 slides
Quantum Systems (Lecture 3: The principles of quantum computation) Lu s Soares Barbosa

Quantum Systems (Lecture 3: The principles of quantum computation) Lu s Soares Barbosa

Quantum Systems (Lecture 3: The principles of quantum computation) Lu s Soares Barbosa Universidade do Minho The principles State space Evolution Composition Measurement The principles If quantum computation explores the laws of

811 views • 45 slides
CIS 4930/6930: Principles of Cyber-Physical Systems Chapter 4: Hybrid Systems Hao Zheng

CIS 4930/6930: Principles of Cyber-Physical Systems Chapter 4: Hybrid Systems Hao Zheng

CIS 4930/6930: Principles of Cyber-Physical Systems Chapter 4: Hybrid Systems Hao Zheng Department of Computer Science and Engineering University of South Florida H. Zheng (CSE USF) CIS 4930/6930: Principles of CPS 1 / 50 Hybrid Systems

794 views • 50 slides
CIS 4930/6930: Principles of Cyber-Physical Systems Chapter 4: Hybrid Systems Hao Zheng

CIS 4930/6930: Principles of Cyber-Physical Systems Chapter 4: Hybrid Systems Hao Zheng

CIS 4930/6930: Principles of Cyber-Physical Systems Chapter 4: Hybrid Systems Hao Zheng Department of Computer Science and Engineering University of South Florida H. Zheng (CSE USF) CIS 4930/6930: Principles of CPS 1 / 19 Hybrid Automata

602 views • 22 slides
Design Principles for Security-conscious Systems 1 Overview Design principles from Saltzer

Design Principles for Security-conscious Systems 1 Overview Design principles from Saltzer

Design Principles for Security-conscious Systems 1 Overview Design principles from Saltzer & Schroeders 1975 paper A few case studies What did Saltzer-Schroeder overlook? (Personal opinions) 2 Saltzer and Schroeders

780 views • 45 slides
The Role of POLST in Advance Care Planning End-of-Life Principles End-Of-Life Care Is About:

The Role of POLST in Advance Care Planning End-of-Life Principles End-Of-Life Care Is About:

The Role of POLST in Advance Care Planning End-of-Life Principles End-Of-Life Care Is About: Compassion at the bedside Providing comfort Honoring patients preferences Advance Care Planning "Advance Care Planning Is Not An

662 views • 28 slides
THE PLANNING SYSTEMS OF THE UK DON GOBBETT - WHUK OUTLINE OF TALK The Planning System

THE PLANNING SYSTEMS OF THE UK DON GOBBETT - WHUK OUTLINE OF TALK The Planning System

THE PLANNING SYSTEMS OF THE UK DON GOBBETT - WHUK OUTLINE OF TALK The Planning System Who makes the decisions? The underlying philosophy THE PLANNING SYSTEM PLANNING DEVELOPMENT + POLICY MANAGEMENT YOU NEED TO ENGAGE WITH BOTH

184 views • 15 slides
Presentation of Draft Principles & Strategies The Planning Partnership November 25, 2015

Presentation of Draft Principles & Strategies The Planning Partnership November 25, 2015

Brock District Plan Presentation of Draft Principles & Strategies The Planning Partnership November 25, 2015 What is a District Plan ? A proactive planning tool and development strategy for a defjned area Intended to be a Regional

732 views • 47 slides
1 s Markov Decision Processes Graphical View of MDP a s, a s,a,s s S t S t+1 S t+2 $R

1 s Markov Decision Processes Graphical View of MDP a s, a s,a,s s S t S t+1 S t+2 $R

Logistics 1 HW 1 Monte Monte- -Carlo Planning: Carlo Planning: Basic Principles and Recent Progress Basic Principles and Recent Progress Dan Weld UW CSE 573 October 2012 Most slides by Alan Fern Consistency &

518 views • 13 slides
Overview of Complex Systems Orientation Principles of Complex Systems Course Information

Overview of Complex Systems Orientation Principles of Complex Systems Course Information

Overview Overview of Complex Systems Orientation Principles of Complex Systems Course Information CSYS/MATH 300, Fall, 2011 Major Complexity Centers Resources Projects Topics Fundamentals Prof. Peter Dodds Complexity Emergence

1.12k views • 70 slides
Principles Principles Principles Principles of a well of a well of a well of a well- - -

Principles Principles Principles Principles of a well of a well of a well of a well- - -

High quality independent research into competition policy and regulation Principles Principles Principles Principles of a well of a well of a well of a well- - - -functioning functioning functioning functioning retail energy retail

495 views • 7 slides
Planning as Satisfiability Enrico Giunchiglia Laboratory of Systems and Technologies for

Planning as Satisfiability Enrico Giunchiglia Laboratory of Systems and Technologies for

Planning Classical Planning Satisfiability of Boolean formulas (SAT) and Planning QBFs satisfiability and Planning Planning as Satisfiability Enrico Giunchiglia Laboratory of Systems and Technologies for Automated Reasoning (STAR-Lab) DIBRIS

1.5k views • 137 slides
Advanced Systems Security: Principles Trent Jaeger Systems and Internet Infrastructure Security

Advanced Systems Security: Principles Trent Jaeger Systems and Internet Infrastructure Security

Systems and Internet Infrastructure Security Network and Security Research Center Department of Computer Science and Engineering Pennsylvania State University, University Park PA Advanced Systems Security: Principles Trent Jaeger Systems and

589 views • 23 slides
Advanced Systems Security: Principles Trent Jaeger Systems and Internet Infrastructure Security

Advanced Systems Security: Principles Trent Jaeger Systems and Internet Infrastructure Security

Systems and Internet Infrastructure Security Network and Security Research Center Department of Computer Science and Engineering Pennsylvania State University, University Park PA Advanced Systems Security: Principles Trent Jaeger Systems and

420 views • 24 slides
Opera&ng Systems: Principles and Prac&ce Tom Anderson

Opera&ng Systems: Principles and Prac&ce Tom Anderson

Opera&ng Systems: Principles and Prac&ce Tom Anderson How This Course Fits in the UW CSE Curriculum CSE 333: Systems Programming Project

486 views • 21 slides

More recommend