UW Solar Power Monitor

Energy Meter Data Collector

Overview

The Energy Meter Data Collector (EMDC) is a daemon that connects to an energy meter over a network using the Modbus TCP protocol. It was designed to periodically query a meter for a set of metrics and record them to a database.

The daemon exposes three HTTP endpoints:

Endpoint	Description
GET /ping	Pings the server to check if it is operating.
GET /metric	Retrieves the most recent value of a metric.
POST /collect	Begins a data collection cycle

Currently an external process (e.g. a cron job) is required to trigger a data collection cycle by calling "POST /collect." In the future this functionality may be integrated directly into the daemon.

Meter Support

The Eaton Power Xpert Meter 4000 and Nexus 1272 meters are presently supported. Support for additional meters can be added by creating spreadsheets mapping the meter's Modbus registers to metric names. The spreadsheet should be placed in the src/collector/maps directory. An example from the Nexus 1272 meters spreadsheet is shown below:

Metric Name	Metric Description	Metric Address	Metric Size	Scaling Factor	Data Type
Voltage_AN	1s Phase A-N Voltage	179	2	1.52588E-05	INT32
Current_N	1s Neutral Current (Secondary)	193	2	1.52588E-05	INT32

In the first row, the Voltage_AN metric is mapped to Modbus register 179 on the Nexus 1272 meter. The metric is two registers in length and should be interpreted as a 32-bit integer. The value must be multiplied by a scaling factor to be converted to its true value.

Database Utilities

The following scripts and utilities help manage the EMDC database:

1. The gendata directory contains a tool that can be used to populate the database with sample data.
2. The migration directory contains scripts that can be used to make incremental changes to the database schema.
3. The sql directory contains sample SQL scripts that were used to create the original database schema.

Installation

Virtual Environment

EMDC depends on several third-party libraries. These libraries should be installed in a virtual environment.

$ git clone https://github.com/kjiwa/lotus-leaf.git
$ cd lotus-leaf
$ python3 -m venv env
$ source env/bin/activate
(env) $ pip install -r requirements.txt

Database Setup

EDMC defaults to using an in-memory SQLite database. This is so that an instance of the daemon can be run easily out of the box. Developers may also find using SQLite convenient since it does not require a separate installation.

Database schema changes are managed by Alembic.

(env) $ cd src/migration
(env) migration$ python migrate.py --db_host=sqlite.db

The following command can be used to prepare a MySQL database:

(env) migration$ python migrate.py --db_type=mysql+mysqlconnector --db_host=localhost

Deployment

Development

A command line client is made available for developers. The client uses the web server built into the Bottle framework. This server is not considered production ready and using this client should be avoided in production.

The following parameters are needed to run EMDC:

• panel_host: The meter host address.
• panel_topic_prefix: The meter topic prefix (e.g. UW/Mercer/nexus_meter).
• panel_metrics_workbook: The workbook containing solar panel metrics data (e.g. collector/maps/nexus-meter.xlsx).

(env) $ cd src
(env) src$ PYTHONPATH=. python collector/main.py \
    --panel_host=10.154.120.13 \
    --panel_topic_prefix=UW/Mercer/nexus_meter \
    --panel_metrics_workbook=collector/maps/nexus-metrics.xlsx

Production

It is recommended that a WSGI application server such as uWSGI be used to deploy EMDC in production. The file wsgi_main.py serves as an entrypoint for production deployments. Configuration of EMDC is done through environment variables.