☀️ New AI Perfects Heliostat Aiming to Boost Solar Tower Power
In a breakthrough that could reshape solar energy, researchers at Spain's Plataforma Solar de Almería (PSA) have developed an AI-powered system that significantly improves how solar mirrors (heliostats) aim at tower receivers.
The AI uses reinforcement learning to adjust each mirror’s angle in real time, based on sunlight, cloud cover, temperature, and time of day — all without human intervention.
🧠 Why This Is a Big Deal
Traditional heliostat setups rely on fixed aiming points. PSA’s new AI system:
-
Learns from experience using a method called Soft Actor-Critic (SAC)
-
Adjusts mirror positions dynamically for maximum energy efficiency
-
Reduces mechanical wear and tear by minimizing unnecessary movements
🧪 Testing & Results
Researchers ran simulations using 114 heliostats at PSA. In three rounds of experiments, the AI strategy:
-
Increased power absorption during midday
-
Extended operation hours in the morning and evening
-
Improved energy output by up to 8.7% annually
-
Optimized aiming to reduce stress on solar receivers
They even taught the AI to balance cost vs. benefit, penalizing excessive movement to keep maintenance costs down.
“We don’t want mirrors to move constantly for tiny gains,” explained co-author Jose Carballo. “The AI learns to get the most energy with the least movement.”
🛠 What’s Next?
Although the AI system is still in the simulation phase, it is now being tested in a real experimental solar tower. The researchers aim to make it adaptable for different tower designs worldwide.
“Every plant is unique,” said Javier Bonilla. “So the AI needs to learn and optimize for each one.”
🌍 Why It Matters
This is part of the broader shift to smart, AI-driven solar infrastructure — making clean energy more reliable, efficient, and affordable. As global industries seek fossil-free heating and fuel options, smart solar plants may be key to delivering scalable, carbon-free power.
Comments
Post a Comment