A Neural Network-Based Model Predictive Approach for Controlled Electrochemical Green Ammonia Production

Green ammonia is a promising alternative to conventional ammonia production methods and has garnered increasing attention in recent years due to its potential to facilitate the transition toward sustainable energy systems. Moreover, green ammonia holds significant potential as a clean energy carrier for various applications, including transportation, energy storage, and fertilizer production. In large-scale commercial green ammonia production systems, an efficient and robust control strategy is required. Therefore, this research presents a novel Neural Network-based Model Predictive Control (NNMPC) approach for the regulation of the electrochemical ammonia synthesizer (EAS) output. The NNMPC utilizes the neural network model of the EAS to generate a control signal that makes the EAS ammonia output production can meet the ammonia demand. Unlike a traditional MPC, which uses a mathematical plant model for predictive optimization, an NN model demonstrates superior accuracy in encapsulating plant dynamics. As the volume of the plant input and output data increases, the precision of the NN model is further enhanced. Also, the NN model effectively addresses mathematical intricacies in MPC models, especially as plant complexities increase. Conversely, NN models’ intrinsic characteristics enable streamlined modeling of the complex EAS plant dynamics. The simulation results validate the efficacy of the NNMPC in controlling EAS ammonia production and meeting the demand, thereby highlighting its effectiveness in practical applications.