Title:
METHOD AND APPARATUS FOR MODEL PREDICTIVE CONTROL
Document Type and Number:
WIPO Patent Application WO/2017/024583
Kind Code:
A1
Abstract:
A method and apparatus for a model predictive control. The method comprises: determining a parameter variable θ(k) at a moment k (S110); by means of a single-layer recursion neural network algorithm, determining a control increment Δu(k) which satisfies a performance index of a predictive model according to a control variable u(k-1) at a moment k-1, and a state variable x(k) and the parameter variable θ(k) at the moment k (S120); determining a sum of the control variable u(k-1) and the control increment Δu(k) of the predictive model to be a control variable u(k) of the predictive model at the moment k (S130); and performing model predictive control according to the control variable u(k) of the predictive model (S140). A control variable u(k) satisfying a performance index of a predictive model is determined by means of a single-layer recursion neural network algorithm according to a state variable x(k) and a parameter variable θ(k) at a current moment, i.e. a moment k, and a control variable u(k-1) at a previous moment, and model predictive control is performed according to the control variable u(k), which can improve the calculation efficiency, and may guarantee the reliability of the performance index.
Inventors:
YAN ZHENG (CN)
BI SHUZHAN (CN)
BI SHUZHAN (CN)
Application Number:
PCT/CN2015/086848
Publication Date:
February 16, 2017
Filing Date:
August 13, 2015
Export Citation:
Assignee:
HUAWEI TECH CO LTD (CN)
International Classes:
G05B13/04
Foreign References:
| CN103345161A | 2013-10-09 | |||
| CN101923083A | 2010-12-22 | |||
| CN104156628A | 2014-11-19 | |||
| CN104268650A | 2015-01-07 | |||
| CN103998999A | 2014-08-20 | |||
| US20090299933A1 | 2009-12-03 |
Attorney, Agent or Firm:
LONGSUN LEAD IP LTD. (CN)
Download PDF: