Gross Weight and Center-of-Gravity Estimation System for the V-22

An evaluation of a combination Artificial Neural Network and Kalman filter system to estimate gross weight and center of gravity for the V-22 is presented. A sampling of V-22 flight test data is used to develop the estimation models, and typical event-driven recorded flight data is used to test the performance of the estimation method. Estimation results of airplane mode gross weight using the combined methods indicate an improvement over using a neural network method alone. The estimated gross weight is able to follow the recorded data without being subject to large gaps or spikes in the event-driven recorded data, as would be the case using a neural network method alone with this type of data. Results of airplane mode CG, helicopter mode GW, and helicopter mode CG estimation are also presented, but using a neural network only at this time. These estimation results match the recorded data well and will provide a good starting estimation for use in a Kalman filter in future efforts.