EMPC Systems: Computational Efficiency and Real-Time Implementation
In this chapter, three economic model predictive control (EMPC) schemes are presented that broadly address the issues of computational efficiency and real-time implementation. In the first section, a composite control structure featuring EMPC is presented
- PDF / 1,603,787 Bytes
- 57 Pages / 439.37 x 666.142 pts Page_size
- 43 Downloads / 165 Views
EMPC Systems: Computational Efficiency and Real-Time Implementation
7.1 Introduction While the two-layer EMPC structures of Chap. 6 were shown to successfully reduce the on-line computation time relative to that required for a centralized, one-layer EMPC scheme, EMPC optimization problems typically found within the context of chemical processes are nonlinear and non-convex because a nonlinear dynamic model is embedded in the optimization problem. Although many advances have been made in solving such problems and modern computers may efficiently perform complex calculations, it is possible that computation delay will occur that may approach or exceed the sampling time. If the computational delay is significant relative to the sampling period, closed-loop performance degradation and/or closed-loop instability may occur. In this chapter, three EMPC design methodologies are presented to further address computational efficiency. In the first section, a composite control structure featuring EMPC is designed for systems with explicit two-time-scale dynamic behavior. Owing to the fact that the class of dynamic models describing such systems are typically stiff, a sufficiently small time step is required for forward numerical integration with explicit methods, which subsequently affects the computation time required to solve the EMPC problem. On the other hand, the composite control structure allows for larger time steps because it avoids the use of the stiff dynamic model embedded in the MPC problems of the composite control structure. In the second section, distributed EMPC (DEMPC), which computes the control actions by solving a series of distributed EMPC problems, is considered. Specifically, an application study whereby several DEMPC schemes are applied to a benchmark chemical process example is presented to evaluate the ability of the resulting DEMPC schemes to reduce the computation time relative to a centralized EMPC system. The closed-loop performance under DEMPC is compared with that achieved under a centralized EMPC approach. In the third section, a real-time implementation strategy for Lyapunov-based EMPC (LEMPC) is presented which addresses potentially unknown and time-varying
© Springer International Publishing Switzerland 2017 M. Ellis et al., Economic Model Predictive Control, Advances in Industrial Control, DOI 10.1007/978-3-319-41108-8_7
233
234
7 EMPC Systems: Computational Efficiency and Real-Time Implementation
computational time for control action calculation. Closed-loop stability under the real-time LEMPC strategy is rigorously analyzed.
7.2 Economic Model Predictive Control of Nonlinear Singularly Perturbed Systems The development of optimal process control, automation, and management methodologies while addressing time-scale multiplicity due to the strong coupling of slow and fast phenomena occurring at different time-scales is an important issue in the context of chemical process control. For multiple-time-scale systems, closed-loop stability as well as controller design are usually addres
Data Loading...
