Package neqsim.process.ml.surrogate

package neqsim.process.ml.surrogate

Surrogate models and physics constraint validation for AI/ML integration.

This package provides infrastructure for hybrid physics-ML systems:

• Surrogate Registry: Cache and manage trained ML models
• Physics Validation: Verify AI actions against physical constraints
• Fallback Support: Automatic fallback to rigorous physics

Design Principles:

1. Physics First: ML augments, never replaces, thermodynamic rigor
2. Safety by Design: Constraints enforced before action execution
3. Explainability: All decisions traceable to physical constraints

Usage Pattern:

// Register surrogate
SurrogateModelRegistry.getInstance().register("flash-model", myNeuralNet);

// Use with physics fallback
double[] result = registry.predictWithFallback("flash-model", input, physicsModel::calculate);

// Validate AI actions
PhysicsConstraintValidator validator = new PhysicsConstraintValidator(process);
ValidationResult check = validator.validate(proposedAction);
if (!check.isValid()) {
  System.out.println("Rejected: " + check.getRejectionReason());
}

Version:

1.0

Author:

ESOL

See Also:

• SurrogateModelRegistry
• PhysicsConstraintValidator

Related Packages

Package	Description
neqsim.process.ml	Machine Learning and AI integration for NeqSim.
neqsim.process.ml.controllers	Simple control algorithms for testing RL environments from Java.
neqsim.process.ml.examples	Example RL environments demonstrating NeqSim ML integration.
neqsim.process.ml.multiagent	Multi-agent reinforcement learning infrastructure for coordinated process control.

Class	Description
PhysicsConstraintValidator	Validates AI-proposed actions against physics constraints.
PhysicsConstraintValidator.Constraint	Base interface for constraints.
PhysicsConstraintValidator.ConstraintCheckResult	Internal result of checking a single constraint.
PhysicsConstraintValidator.ConstraintViolation	Details of a constraint violation.
PhysicsConstraintValidator.EquipmentConstraint	Constraint on specific equipment.
PhysicsConstraintValidator.PhysicalBoundConstraint	Constraint on physical bounds (e.g., positive temperature).
PhysicsConstraintValidator.ValidationResult	Result of a validation check.
SurrogateModelRegistry	Registry for managing trained surrogate (machine learning) models.
SurrogateModelRegistry.SurrogateMetadata	Metadata for a surrogate model.
SurrogateModelRegistry.SurrogateModel	Interface for surrogate model implementations.
SurrogateModelRegistry.SurrogateModelEntry	Internal entry combining model and metadata.