An online parameter identification and real-time optimization platform for thermal systems and its application

Real-time performance optimization of thermal systems is crucial for energy conservation but also challenging because of the high system complexity and time sensitivity. Herein, an online parameter identification and real-time optimization platform for thermal systems is developed, and a gas-steam combined cycle cogeneration system is used to demonstrate its capability. The platform integrates data collection, parameter identification, system simulation, and real-time optimization modules. The data preprocessing and parameter identification modules collects real-time operation parameters for optimization. The system simulation module applies the heat current method to model the cogeneration system precisely, and a high-efficiency simulation procedure is proposed using the hierarchical and categorized (H&C) algorithm. The system optimization module introduces the artificial neural network technology to ensure the response time of real-time optimization. Meanwhile, the H&C algorithm and genetic algorithm are combined to update the database to improve the optimization performance gradually. The platform is first validated on three typical conditions. It is further deployed at a power plant, where a field test is conducted for the practical verification. Field test results show that the standard coal consumption of the cogeneration system could be reduced by 0.415 g/kWh by using the platform, which proves its practicability.