cobre-sddp 0.1.2

Stochastic Dual Dynamic Programming (SDDP) for hydrothermal dispatch and energy planning
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94

//! # cobre-sddp
//!
//! Stochastic Dual Dynamic Programming for the [Cobre](https://github.com/cobre-rs/cobre) power systems ecosystem.
//!
//! This crate implements the SDDP algorithm (Pereira & Pinto, 1991) for
//! long-term hydrothermal dispatch and energy planning:
//!
//! - **Forward pass**: scenario-based simulation with policy evaluation.
//! - **Backward pass**: Benders cut generation for the cost-to-go function.
//! - **Cut management**: single-cut and multi-cut strategies, cut selection,
//!   and dominance pruning.
//! - **Risk measures**: expected value, `CVaR`, and convex combinations.
//! - **Convergence**: statistical stopping criteria and bound gap monitoring.
//! - **Parallelism**: designed for hybrid MPI + thread-level parallelism
//!   via [ferrompi](https://github.com/cobre-rs/ferrompi).
//!
//! Built on `cobre-core` for system data, `cobre-stochastic` for inflow
//! modeling, and `cobre-solver` for LP subproblems.
//!
//! ## Status
//!
//! This crate is in early development. The API **will** change.
//!
//! See the [repository](https://github.com/cobre-rs/cobre) for the full roadmap.

// Relax strict production lints for test builds (normal in test contexts).
#![cfg_attr(
    test,
    allow(
        clippy::unwrap_used,
        clippy::expect_used,
        clippy::panic,
        clippy::float_cmp,
    )
)]

pub mod backward;
pub mod config;
pub mod context;
pub mod convergence;
pub mod cut;
pub mod cut_selection;
pub mod cut_sync;
pub mod error;
pub mod estimation;
pub mod forward;
pub mod horizon_mode;
pub mod indexer;
pub mod inflow_method;
pub mod lower_bound;
pub mod lp_builder;
pub(crate) mod noise;
pub mod risk_measure;
pub mod simulation;
pub mod state_exchange;
pub mod stopping_rule;
pub mod training;
pub mod training_output;
pub mod trajectory;
pub mod workspace;

pub use backward::{BackwardPassSpec, BackwardResult, run_backward_pass};
pub use config::TrainingConfig;
pub use context::{StageContext, TrainingContext};
pub use convergence::ConvergenceMonitor;
pub use cut::{CutPool, FutureCostFunction};
pub use cut_selection::{
    CutMetadata, CutSelectionStrategy, DeactivationSet, parse_cut_selection_config,
};
pub use cut_sync::CutSyncBuffers;
pub use error::SddpError;
pub use estimation::EstimationError;
pub use forward::{ForwardResult, SyncResult, run_forward_pass, sync_forward};
pub use horizon_mode::HorizonMode;
pub use indexer::StageIndexer;
pub use inflow_method::InflowNonNegativityMethod;
pub use lower_bound::{LbEvalSpec, evaluate_lower_bound};
pub use lp_builder::{PatchBuffer, StageTemplates, ar_dynamics_row_offset, build_stage_templates};
pub use risk_measure::{BackwardOutcome, RiskMeasure};
pub use simulation::{
    CategoryCostStats, EntityCounts, ScenarioCategoryCosts, SimulationBusResult, SimulationConfig,
    SimulationContractResult, SimulationCostResult, SimulationError, SimulationExchangeResult,
    SimulationGenericViolationResult, SimulationHydroResult, SimulationInflowLagResult,
    SimulationNonControllableResult, SimulationOutputSpec, SimulationPumpingResult,
    SimulationScenarioResult, SimulationStageResult, SimulationSummary, SimulationThermalResult,
    StageSummaryStats, accumulate_category_costs, aggregate_simulation, assign_scenarios,
    extract_stage_result, simulate,
};
pub use state_exchange::ExchangeBuffers;
pub use stopping_rule::{MonitorState, StoppingMode, StoppingRule, StoppingRuleSet};
pub use training::{TrainingResult, train};
pub use training_output::build_training_output;
pub use trajectory::TrajectoryRecord;
pub use workspace::{BasisStore, BasisStoreSliceMut, SolverWorkspace, WorkspacePool};