sproclib

SPROCLIB - Standard Process Control Library

A comprehensive Python library for chemical process control, providing essential classes and functions for PID control, process modeling, simulation, optimization, and advanced control techniques.

SPROCLIB provides a semantic API for chemical plant design that uses familiar patterns from machine learning frameworks like TensorFlow and Keras.

Installation

pip install sproclib

Features

• Semantic Plant Design: Intuitive API similar to ML frameworks for building complex chemical processes
• Process Units: CSTR, pumps, heat exchangers, distillation columns, reactors, and tanks
• Economic Optimization: Built-in optimization algorithms for cost minimization and profit maximization
• PID Controllers: Classical and advanced PID control implementations with auto-tuning
• Analysis Tools: Transfer functions, simulation, and system identification
• Advanced Control: Model predictive control, state-space controllers, and IMC
• Transport Models: Continuous and batch transport for liquids and solids

Quick Start

Create and optimize a chemical plant in just a few lines:

from sproclib.unit.plant import ChemicalPlant
from sproclib.unit.pump import CentrifugalPump
from sproclib.unit.reactor import CSTR

# Define plant
plant = ChemicalPlant(name="Process Plant")

# Add units
plant.add(CentrifugalPump(H0=50.0, eta=0.75), name="feed_pump")
plant.add(CSTR(V=150.0, k0=7.2e10), name="reactor")

# Connect units
plant.connect("feed_pump", "reactor", "feed_stream")

# Configure optimization
plant.compile(
   optimizer="economic",
   loss="total_cost",
   metrics=["profit", "conversion"]
)

# Optimize operations
plant.optimize(target_production=1000.0)

Advanced Example

# Traditional PID control example
import sproclib as spc

# Create a PID controller
controller = spc.PIDController(kp=1.0, ki=0.1, kd=0.05)

# Create a tank model
tank = spc.Tank(volume=100, area=10)

# Simulate step response
response = spc.step_response(tank, time_span=100)

Requirements

• Python 3.8+
• NumPy >= 1.20.0
• SciPy >= 1.7.0
• Matplotlib >= 3.3.0

License

MIT License

Author

Thorsten Gressling [e-mail]