Application of Silicon Carbide Photodiode Flame Temperature Sensors in an Active Combustion Pattern Factor Control System

Impact Technologies, in cooperation with General Electric’s Global Research Center, propose an intelligent pattern factor control concept that will utilize an innovative Silicon Carbide photodiode-based UV Flame Temperature Sensor (FTS) in an adaptive/learning controller that can compensate for degradations/errors in the combustion system. The core developments will be based on the integration of the first-time use of the FTS in an engine application, coupled with a learning neural network control algorithm that can accommodate for errors in the combustion monitoring process. This innovative engine control concept consists of the following key elements: 1) Learning Neural Network using Model Inversion Compensation control system - takes into account errors in the combustion balancing and actuation model; 2) Signal health estimation – if the sensors are unreliable or not giving consistent information as a group, then the system takes this into account; and 3) Nozzle health estimation and accommodation – if the combustion hardware is unhealthy (e.g. nozzle coking), the adaptive controller will alter system behavior accordingly. The Phase I concept will be demonstrated within a dSPACE hardware-in-the-loop simulation that can simulate real-time behavior of relevant combustor pattern faults.