Explicit Nonlinear Model Predictive Control: Theory and by Alexandra Grancharova

By Alexandra Grancharova

Nonlinear version Predictive keep an eye on (NMPC) has turn into the permitted technique to resolve complicated keep an eye on difficulties concerning approach industries. the most motivation in the back of explicit NMPC is that an explicit state suggestions legislations avoids the necessity for executing a numerical optimization set of rules in genuine time. the advantages of an explicit solution, as well as the effective online computations, contain additionally verifiability of the implementation and the prospect to layout embedded keep watch over platforms with low software program and complexity.

This ebook considers the multi-parametric Nonlinear Programming (mp-NLP) ways to explicit approximate NMPC of restricted nonlinear systems, built by way of the authors, in addition to their functions to varied NMPC challenge formulations and several other case reviews. the next kinds of nonlinear systems are thought of, leading to varied NMPC challenge formulations:

Ø Nonlinear systems defined via first-principles versions and nonlinear systems defined through black-box models;

- Nonlinear systems with non-stop keep an eye on inputs and nonlinear systems with quantized keep an eye on inputs;

- Nonlinear systems with no uncertainty and nonlinear systems with uncertainties (polyhedral description of uncertainty and stochastic description of uncertainty);

- Nonlinear systems, along with interconnected nonlinear sub-systems.

The proposed mp-NLP ways are illustrated with functions to a number of case reviews, that are taken from different parts reminiscent of automobile mechatronics, compressor keep watch over, combustion plant regulate, reactor regulate, pH preserving process regulate, cart and spring process keep watch over, and diving computers.

Show description

Read or Download Explicit Nonlinear Model Predictive Control: Theory and Applications PDF

Similar information theory books

Communication Researchers and Policy-making: An MIT Press Sourcebook (MIT Press Sourcebooks)

Because the international info infrastructure evolves, the sector of conversation has the chance to resume itself whereas addressing the pressing coverage want for brand new methods of considering and new facts to consider. communique Researchers and Policy-making examines diversified relationships among the verbal exchange study and coverage groups over greater than a century and the problems that come up out of these interactions.

Additional resources for Explicit Nonlinear Model Predictive Control: Theory and Applications

Example text

E. 7. 54) with H 0, X convex. Then the set of feasible parameters X f ⊆ X is convex, the optimizer z∗ (x) : X f → Rs is continuous and piecewise linear and the value function V ∗ (x) : X f → R is continuous, convex and piecewise quadratic. 3. Exact mp-QP. Step 1. Let the current region be the whole polyhedron X ⊆ Rn . Step 2. 58). Step 3. For x = x0 , compute the corresponding optimal solution (z0 , λ0 ) by solving a QP. ˜ Step 4. Determine the set of active constraints when z = z0 , x = x0 , and build G, ˜ ˜ W , S.

As discussed in [72, 73, 67, 69], the method in [12, 13] is effective to handle these special cases. 2. Example. 73) The partitioning starts with finding the region where no constraints are active. As the mp-QP is created from a feasible MPC problem, the empty active set will be op˜ W˜ and S˜ are empty matrices, timal in some full-dimensional region (A0 = 0/ and G, ∗ z (x) = 0). This critical region is then described by 0 ≤ W + Sx which contains 8 inequalities. Two of these inequalities are redundant with degree 0 (#2 and #4), the remaining 6 hyperplanes are facet inequalities of the polyhedron (see Fig.

54) is solved with x = x0 in order to obtain the corresponding optimal solution z0 . 54). Let G, S and W denote the rows of G, S and W corresponding to the active constraints. 5. Let H 0. Consider a combination of active constraints G, ˜ and assume that the rows of G are linearly independent. Let CR0 be the set of all vectors x for which such a combination is active at the optimum (CR0 is referred to as critical region). Then, the optimal z and the associated vector of Lagrange multipliers λ are uniquely defined linear functions of x over CR0 .

Download PDF sample

Rated 4.73 of 5 – based on 39 votes

Published by admin