Extended Behaviour Prediction Framework in Complex System Development

The research presented in this paper presents a framework with an algorithm intended for the development of complex technical systems during their operation in uncertain situations. It is based on the prediction of deviations in the behaviour of complex engineering systems with uncertain operating parameters relative to the behaviour of the same system with the predicted operating parameters. In this case, the unexpected behaviour of the system in a changing working environment is modelled with: the architecture model of a complex technical system and the behaviour model of the same technical system. The model of complex system architecture is based on a matrix representation of the system components using a component Design Structure matrix (DSM) component. A mathematical model with distributed parameters and a model predictive control (MPC) method is used to describe the behavioural model of a complex system. Observed system stability is also verified using the direct Lyapunov method. Bilateral mapping of the obtained data between these two models allows describing and modelling the system behaviour in uncertain situations. The recording of the behaviour of the created complex system is performed using the rules of fuzzy logic. For this purpose, the Adaptive-Network-based Fuzzy Inference System (ANFIS) is used. Verification of the research results was carried out on a real example of a complex technical system - an air handling unit.