Core Insights - A research team from the University of Michigan has developed a new algorithm based on physics that allows micro nuclear reactors to autonomously adjust their output power on demand [1] - The research indicates that several startups and traditional companies in the U.S. are actively promoting the short-term deployment of micro nuclear reactors, making this new development economically viable [1] Group 1: Micro Nuclear Reactor Advantages - Micro nuclear reactors can achieve a thermal output of up to 20 megawatts and are advantageous due to their transportability and versatility, suitable for remote communities, disaster areas, military bases, and cargo ships [1] - When connected to the grid, these reactors can provide stable zero-carbon energy [1] Group 2: Challenges and Solutions - A key challenge for micro nuclear reactors is the need for "load-following" capability, which allows them to dynamically adjust output power based on actual demand, unlike large nuclear power plants that rely on manual adjustments [1] - The research focuses on high-temperature gas-cooled reactors, which can be flexibly scaled, and a simplified model was constructed to retain core parameters such as power density and coolant inlet temperature [1] Group 3: Control Algorithm Development - The team employed Model Predictive Control (MPC) to predict and optimize system behavior within set constraints and time frames [2] - A specially designed MPC controller adjusts power by regulating the rotation angle of the reactor's central control drum, with inward rotation decreasing power and outward rotation increasing it [2] - Testing showed that the control algorithm maintained an error of less than 0.234% in tasks requiring a 20% power increase or decrease per minute, demonstrating high accuracy without the use of artificial intelligence [2]