Fractional order modelling and robust multi-model intelligent controllers’ synthesis for ACP1000 nuclear power plant
Abstract
In this research work, a third generation Pressurized Water Reactor (PWR) type Nuclear Power Plant (NPP) of 1100 MWe rating called Advanced Chinese PWR (ACP1000) NPP is adopted. The research and development work is mainly encompassed of addition of third steam generator loop, hybrid programming platforms, fractional order modular modelling and intelligent controllers. The ACP1000 NPP is comprised of primary loop, secondary loop and balance of plant (BOP). The ACP1000 nonlinear dynamics is precisely modelled with greater efficiency using fractional order differential equations. The entire plant is modelled into eighteen modules in a fractional order framework in Visual Basic (VB) environment. The eighteen modules are governed by nine controllers. Controllers are configured in a fashion of an adaptive neuro-fuzzy inference system (ANFIS). An innovative multi input single output (MISO) ANFIS virtual instrument (VI) is designed for training and configuring controllers in LabVIEW. The developed MISO ANFIS VI in combination is implemented for multi-input multi output (MIMO) ANFIS controllers in parallel computing framework. The variable transfer between fractional order multi- modular framework in VB (FO- MMF-VB) and robust intelligent ANFIS controllers in LabVIEW (RI-ANFIS-LV) is established in client server configuration and system exec VI for calling external code in VB. This hybrid programming framework of Visual Basic and LabVIEW is actually a development of LabVIEW Visual Basic Integration (LabVb) Toolkit used to evaluate the closed loop transient performance of ACP1000 nuclear power plant. Various parameters are simulated for dynamic transient simulations under load manoeuvring from 100% power to 50% power. The results are analysed and validated against benchmark design data and Preliminary Safety Analysis Report (PSAR). All the results are within the protection systems design limits under normal operating conditions.