OPTHOME

Hourly energy consumption with load needed from the grid
Convergence of the algorithm towards the best cost

Inspiration

With buildings consuming 40% of global energy and the transition to electric vehicles looming, there's an urgent need for smarter home energy management. The vision was to create an autonomous system that could optimize both cost and grid stability while seamlessly integrating renewable energy and storage systems.

What it does

OPTHOME is a Home Energy Management System (HEMS) that:

Automatically schedules appliance usage to minimize energy costs
Intelligently manages energy flow between solar panels, batteries, and the grid
Optimizes when to store/use/sell energy based on electricity pricing
Reduces peak demand on the power grid while maintaining user comfort
Handles both shiftable appliances (like washing machines) and non-shiftable ones (like refrigerators)

How we built it

Developed a binary particle swarm optimization (BPSO) algorithm to handle the complex scheduling problem
Created a mathematical model accounting for all energy flows between components
Implemented both single-objective (cost) and multi-objective (cost + peak reduction) optimization
Built in real-world constraints like battery efficiency and appliance operating times
Used MATLAB for simulation and optimization

Challenges we ran into

Balancing multiple competing objectives (cost vs peak reduction)
Handling the complexity of binary variables for appliance scheduling
Managing physical constraints while keeping computation time reasonable
Implementing particle corrections instead of penalty functions
Dealing with unclear aspects of previous research implementations

Accomplishments that we're proud of

Reduced peak-to-average ratio by 15% compared to existing systems
Successfully integrated renewable energy and storage optimization
Developed a more efficient binary optimization approach
Created a solution that performs better with less solar capacity

What we learned

The power of meta-heuristic algorithms for complex real-world problems
Trade-offs between solution quality and computation speed
Importance of proper constraint handling in optimization
Benefits of binary vs continuous variable approaches
Critical role of energy storage in future smart grids

Next steps

Incorporating next day estimations for energy prices from public resources
Developing a user interface and smartphone app to allow the user to monitor, add new appliances and estimate future costs