Study and design of an autotuning system for a helicopter flight simulator

Flight simulators for pilot training are constantly evolving, and their role in the training process is becoming increasingly important. The synthetic environment is used during the initial phases of training, in familiarisation with the aircraft, and then maintains its importance in the subsequent phases, culminating in training for mission-critical and safety-critical procedures. The simulation devices, although having different levels of complexity, realism and fidelity compared to the real aircraft, are linked by a common set of standards and requirements, coming from the designated National and Supranational Authorities, which must be met to allow the certification of simulators. This suitability is also demonstrated through objective tests that link the behaviour of the simulated aircraft within the synthetic environment with the real aircraft performances, by comparing flight manoeuvres accomplished in the simulator with the same manoeuvres recorded in flight. This tests, named QTGs (Qualification Test Guide), are executed by a software model and normally take a lot of time due to the difficulty in tuning the gain of the control system embedded in the software. Hence, the main goal of this project is to develop an automatic control system, to be used during the validation tests of a helicopter flight simulator, which is able to faithfully replicate the behaviour of the pilot and which is able to auto-tune, through an automatic process, in order to reduce the time required for the definition and execution of the tests, lowering the man-time currently required by the process and consequently reducing costs. This goal was achieved starting from a state-of-the-art analysis to understand the today environment in auto-tuning control system and to find the more suitable solution for this project. Then, the next step consisted in studying the helicopter flight model, in particular the MathPilot section that is the core part analysed during this work, to figure out the context in which the model would be implemented. Consequently, the modelling section began using a dedicated calculation environment. Hence, it followed a part of analysis in which the correct performances of the model were verified. Finally, the project ended with conclusions and some considerations on the whole project. This work was developed in a model-based environment that runs together with the helicopter simulation model in order to perfectly integrate itself in its loop.