optivgi 0.2.0

An EV smart charging management application designed to optimize electric vehicle charging based on power or pricing constraints.

Opti-VGI: EV Smart Charging Scheduler Application

Opti-VGI is an EV smart charging management application designed to optimize electric vehicle charging based on power or pricing constraints. It provides a flexible framework whose scheduling logic can integrate with Charge Station Management Systems (CSMS). The specific integration method, potentially supporting protocols like OCPP 1.6 or OCPP 2.0.1 / ISO 15118, depends on the implementation of its communication layer.

Documentation: https://argonne-vci.github.io/Opti-VGI

---

Key Features

• Constraint-Based Optimization: Schedules EV charging considering factors like overall site power limits or dynamic pricing signals.
• Modular Architecture: Clearly separates communication logic (Translation layer) from optimization strategies (Algorithm layer).
• Pluggable Algorithms: Supports different optimization approaches. Includes implementations like:
  • PulpNumericalAlgorithm: Uses linear programming via the PuLP library.
  • GoAlgorithm: A custom heuristic-based algorithm.
• CSMS Integration Interface: The Translation abstract class defines the necessary methods to fetch data (EVs, constraints) and send charging commands, allowing integration with various external systems. The implementation of this layer determines how Opti-VGI communicates (e.g., via API calls, database interaction, or specific protocols like OCPP).
• Handles Active & Future EVs: Considers both currently connected EVs and planned future reservations in its scheduling.
• Asynchronous Operation: Designed to run scheduling logic periodically or in response to events using background worker threads.

---

Architecture Overview

Opti-VGI follows a modular design:

1. Translation Layer (optivgi.translation): An abstract interface (Translation) for communication with external systems (e.g., a CSMS). Implementations handle fetching EV data, power constraints, and sending charging profiles. This is where specific communication protocols (like OCPP methods via a library, or custom API calls) would be implemented.
2. SCM Runner (optivgi.scm_runner): Orchestrates the scheduling cycle, using a Translation instance to get inputs, an Algorithm instance to calculate schedules, and the Translation instance again to dispatch results.
3. SCM Algorithm (optivgi.scm.algorithm): An abstract interface (Algorithm) for charging strategies. Concrete implementations contain the core optimization logic (e.g., GoAlgorithm, PulpNumericalAlgorithm).
4. EV Data Structure (optivgi.scm.ev): A standard dataclass (EV) representing vehicles and their charging parameters.
5. Worker Threads (optivgi.threads): Helpers to run the SCM logic periodically or on events without blocking the main application.

Sequence Diagram

The diagram shows the flow of data and commands between the EVSEs, Opti-VGI, and the CSMS, highlighting the roles of the Translation and Algorithm layers in this interaction. It illustrates how Opti-VGI fetches data from the CSMS, processes it using the selected algorithm, and sends back charging commands.

---

Installation

You can install the latest stable release from PyPI:

pip install optivgi

For development, clone the repository and install with development dependencies:

git clone https://github.com/argonne-vci/Opti-VGI.git
cd Opti-VGI
pip install -e ".[dev]"

---

Usage & Examples

Opti-VGI provides the core scheduling framework. To use it, you need to:

1. Implement the Translation interface: Create a concrete class that inherits from optivgi.translation.Translation and implements the get_evs, get_peak_power_demand, and send_power_to_evs methods to communicate with your specific CSMS or data source.
2. Choose an Algorithm: Select one of the provided algorithms (e.g., GoAlgorithm, PulpNumericalAlgorithm) or implement your own inheriting from optivgi.scm.algorithm.Algorithm.
3. Run the scm_worker: Use the optivgi.threads.scm_worker function in a separate thread, providing your Translation implementation and chosen Algorithm class. Trigger the worker using an event queue.

Examples

• HTTP API: Integration via a simple HTTP API and WebSocket notifications.
• CitrineOS: Full OCPP 1.6J integration with CitrineOS CSMS, including a Docker Compose stack with charger simulator, event-driven scheduling via RabbitMQ, and an automated curtailment demo.

See the Examples Documentation for more details.

---

Contributing

Contributions are welcome! Please feel free to open an issue on the GitHub Issues page to report bugs, suggest features, or discuss potential improvements.

---

License

This project is licensed under the Apache License, Version 2.0. See the LICENSE file for details.

---

Links

• Documentation: https://argonne-vci.github.io/Opti-VGI
• PyPI Package: https://pypi.org/project/optivgi/
• GitHub Repository: https://github.com/argonne-vci/Opti-VGI
• Issue Tracker: https://github.com/argonne-vci/Opti-VGI/issues