Package neqsim.process.ml.multiagent

package neqsim.process.ml.multiagent

Multi-agent reinforcement learning infrastructure for coordinated process control.

This package provides infrastructure for multi-agent systems where multiple RL agents coordinate to control interconnected process equipment:

• Agent - Individual agent interface
• MultiAgentEnvironment - Coordinated environment
• ProcessAgent - Base process agent implementation
• SeparatorAgent - Separator control agent

Use Cases:

• Train separation with downstream compression (level agent + anti-surge agent)
• Heat integration networks (multiple heat exchanger agents)
• Gas processing plants (multiple unit operations)

Version:

1.0

Author:

ESOL

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

Class	Description
Agent	Interface for an individual agent in a multi-agent system.
CompressorAgent	RL agent for compressor control with anti-surge protection.
MultiAgentEnvironment	Multi-agent environment for coordinated process control.
MultiAgentEnvironment.CoordinationMode	Coordination mode for multi-agent systems.
MultiAgentEnvironment.MultiAgentStepResult	Step result for multi-agent environment.
ProcessAgent	Base class for process equipment control agents.
SeparatorAgent	RL agent for separator level and pressure control.