GridBuddy — Solar Microgrid Monitoring & Management

GridBuddy is a solar microgrid monitoring and management system designed to retrofit existing solar trailers and help operators (often non-technical volunteers) understand system status at a glance and extend runtime by intelligently shedding non-essential loads.

Developed as part of the Georgia Tech undergraduate capstone project in Mechanical Engineering. Spring 2025

---

What’s in this repo

This repository contains multiple PlatformIO firmware modules that together form the GridBuddy prototype:

• monitoring_module/ — DC-side sensing + on-device display
• priority_switch_module/ — smart-plug / outlet control
• remote_module/ — an optional hand-held monitor that sync the system information

Each folder is its own PlatformIO project (see each module’s platformio.ini).

---

System Overview (Technical)

GridBuddy is split into two major subsystems:

1. Monitoring Module (DC side)
   Measures PV/battery electrical quantities using non-invasive Hall-effect current sensing and a voltage divider, digitized with a high-resolution external ADC and displayed locally on an ESP32-based UI.

2. Smart Plug Module (AC side)
   Controls multiple AC outlets through relays (with manual override + automatic mode) to shed non-essential loads when battery conditions require it.

The modules communicate via ESP32 wireless telemetry/control.

---

Monitoring Module (DC-side sensing)

Hardware

• Microcontroller: ESP32
• External ADC: ADS1115 (16-bit) over I²C for improved resolution vs. typical on-board ADCs
• Voltage measurement: resistive divider scales high DC bus voltage into the ADC input range
• Current measurement: non-invasive Hall-effect current sensor(s) (PV input; optional battery charge/discharge channels)
• Power estimation: computed from measured V and I
• Retrofit-friendly wiring: inline install with MC4 connectors

Monitoring wiring / signal flow (block diagram)

Monitoring physical prototype (packaging / interfaces)

Sensor calibration (used in firmware)

These plots show the calibration relationships used to convert raw analog/ADC readings into physical units:

Hall-effect current sensor calibration: analog reading (mV) → current (A) using a fitted linear model.

Voltage divider calibration: analog reading → bus voltage (V) using a fitted linear model.

---

Smart Plug Module (AC-side load shedding)

Hardware + behavior

• Microcontroller: ESP32
• Power chain: 110 VAC input → AC/DC supply → regulation for logic/relays
• Outputs: multiple relay-controlled outlets (prototype: 4)
• Control modes:
  • Manual override via local buttons (force on/off)
  • Automatic mode based on a finite-state load-shedding strategy
• Stability: uses hysteresis to reduce outlet “chatter”

Smart plug wiring / signal flow (block diagram)

Smart plug physical prototype (assembly)

Load-shedding state machine

How the state machine works (high level):

• The controller classifies the system using:
  • Time-of-day (day / transitional / night)
  • Battery state-of-charge (sufficient → at risk)
  • Rate-of-charge (charging vs. discharging)
• These inputs map to a discrete system state (NORMAL, ALERT 1, ALERT 2, CRITICAL).
• An operator override can force “always on” or “always off.”
• In automatic mode, the module either continues Normal Operation or issues a Request to Shed Load, which translates to switching one or more outlets off based on priority.

---