Solar Farm Placement and How AI Can Help

The efficiency of solar panels rely on many different factors, which can be hard to take into account when designing solar farms. With AI, this can be made easier.

The sun is one of the most accessible resources available. Across the world, 430 quintillion Joules of energy hits the Earth every hour. For comparison, all humans combined use that amount of energy yearly. That is such an impressive number, and as a society, we are underutilizing its amazing capabilities.

For obvious reasons, we cannot put solar farms everywhere and anywhere. We cannot put one in the middle of LA, for example, because while it is sunny there, there is not enough land available. We also cannot put one in the middle of Alaska, even though there is plenty of space due to the lack of light during parts of the year.

Less obvious factors also contribute to how much power is generated, however. High amounts of wind can cool down panels and help them to absorb more energy. On the other hand, water droplets formed by high humidity can reflect light off of panels, decreasing the amount of electricity produced.

Solar farm developers need to take into account all of this information to choose the best location for their panels. With the factors mentioned above and more, however, it can be hard to get a rough estimate of how much electricity each solar cell will produce. Enter AI, and this problem can become more manageable.

How AI Can Help

One of the biggest features of AI is its ability to recognize patterns and use them to make predictions. Data is inputted, analyzed, and used as a source to inference off of when exposed to new information. By doing this with data on climate conditions in a particular area, developers can get a rough estimate of how much energy will be produced from a solar cell.

My Software

I used an idea developed by Neural Designer and data from Kaggle to replicate the AI model mentioned above. It takes in the following parameters:

• Distance to solar noon, in radians
• Temperature, in Celsius
• Wind direction, in degrees
• Wind speed, in meters per second
• Sky cover, on a five-step scale from 0 to 4
• Humidity, in percentage
• Average wind speed, in meters per second
• Average pressure, in mercury inches

And ultimately gives an estimate of how many Joules of electricity each solar cell will produce within a three-hour time frame. For developers this is huge — even if it’s just an estimate, this technology can not only make their solar panels more efficient, but it can also save a significant amount of time.

The above picture shows the code I wrote to create this model. I worked in Google Collab and used TensorFlow, a popular software for AI, to create the body for my model. The accuracy rate is only forty-six percent for new data, but that is relatively high for an estimate.

While building this project, one of the major challenges I faced was figuring out the combination of dense amounts and activation shapes that led to the highest accuracy rate. This was my first time creating an AI model, so I spent a lot of time learning how to write the needed code and make it as accurate as possible.

If you are interested in checking out the code itself, here is the link :)

Potential Problems

While something like this can make a big impact, it can also lead to some issues if relied on too heavily. If data is collected on a day with unusual weather conditions for the area, the estimate may not reflect the normal amount of energy that will be collected. This could lead to financial loss and less overall electricity generated, which is negative for both the developer and the electricity consumers.

The Future

There is so much room for AI in the clean energy space, and solar power is just the start of its potential. In the future, I want to look into the impact it could have on other forms of clean energy in addition to solar.

Imagine a world where clean energy is the globe’s main source of electricity. The risks of fossil fuels, severe global warming, and environmental disasters would be minimized, allowing us to shift our focus toward other important issues. While we still have a long way to go, technology is helping society inch closer and closer to that goal.