Research On Modeling And Advanced Control For Low Temperature Waste Heat Power Generation Processes

Organic Rankine cycle (ORC) is an effective technology for recovering low temperature waste heat, which has significance for reducing fossil energy consumption, improving utilization efficiency of energy, reducing industrial operation cost and pollution. High performance and reliable controllers are essential to enhance the competitiveness of ORC based waste heat power generation technology. This dissertation focuses on the study of modeling and advanced control of ORC based waste heat energy conversion systems (WHECSs). The main work of this dissertation are summarized as follows:1. Moving-boundary method is used to build the simplified models of evaporator and condenser based on the equations of mass and energy conservation and reasonable simplifying assumptions. The steady state models of scroll expander and pump are obtained based on the mechanism analysis of them. Then the input-output relationships betweem different components are analyzed. Combining all the components together, a control-oriented dynamic model of waste heat energy conversion system is established. The model with low order can represent a wide range of operating conditions. So it is suitable for control system design and synthesis.2. The ORC based WHECS operating in following the waste heat mode presents highly nonlinear behaviors over a wide range of operating conditions. The linear parameter varying (LPV) gain scheduling control method is applied to the ORC system. The nonlinear dynamics of WHECS are formulated by a convex polytopic LPV model using Jacobian linearization method by adopting mass flow rate and temperature of waste heat as scheduling variables. Then the vertex controllers are designed using linear matrix inequalities (LMI) method, and the gain scheduling LPV controller is obtained by synthezing each vertex contoller. Simulation results demonstrate the proposed strategy outperforms convetional PI control schemes.3. Most industrial processes are multivariable systems with strong couplings. Also, it is difficult to obtain accrurate mathmatical models of pocesses due to uncertaities caused by variations of ambient and operating conditions. A multiloop robust control strategy is proposed based on Hx control and a partial least squares (PLS) model, especially for ill-condition plants and non-square systems. Taking advantage of automatic pairing and orthogonality of variables in PLS latent space, the multi-input-multi-output (MIMO) system can be decomposed into several independent single-input-single-output (SISO) subsystems. The effectiveness of the proposed approach are illustrated by the simulation results of a distillation column and a mixing tank process.4. The developed multiloop robust control strategy is applied to the ORC based waste heat energy conversion system operating in following the electric load mode. The dynamic filters are used to build the dynamic PLS model of WHECS. Taking into account model mismatch and disturbances from the heat source, multiloop robust controllers are designed using mixed sensitivity approach in each PLS subspace independently. The simulation results demonstrate the effectiveness and superiority of the proposed control method.