Method for predicting and controlling steam temperatures of boilers on basis of state observation model

Abstract

The invention relates to a method for predicting and controlling steam temperatures of boilers on the basis of a state observation model. The method includes 1, replacing heat exchangers by multiple sections of concentration parameter links by a principle of natural segmentation or virtual segmentation, and using outlet parameters of media of various sections as representative parameters of the concentration links to build a mathematical model of enthalpy and temperature channels of the heat exchangers; 2, creating a state observer on the basis of output feedback and estimating outlet temperatures of the various sections; 3, connecting an MPC (multivariate model prediction controller) and a PID (proportion, integration and differentiation) regulator in a cascade manner to form a serial control circuit, acquiring step response coefficients according to the model of the enthalpy and temperature channels of the heat exchangers and acquiring set values of lead temperatures; 4, regulating the steam temperatures by the inner-loop PID regulator according to difference among the lead temperatures and output of the MPC and acquiring control signals of attemperation water valves. The method has the advantages that model building and verification are carried out, so that field test requirements can be reduced; states of inertial and lagged controlled objects are reconstructed via the state observer, so that characteristics of the objects can be modified; prediction and control are implemented, so that targets can be accurately and quickly controlled.

Description

technical field [0001] The invention relates to a predictive control method based on a state observation model, specifically a method for effectively segmenting large inertia and large lag control objects, using state observation methods to construct a mathematical model, and performing model predictive control. Boiler steam temperature control. Background technique [0002] Due to the use of the mathematical model of the control object, the model predictive control enables the controller to directly process some typical characteristics of the control object in the dynamic process, and predict the role of disturbance factors in a period of prediction time domain, and make correct decisions under constraints. control effect. Both theoretical derivation and numerical simulation have proved that model predictive control is effective, and more and more appear in practical attempts. [0003] However, if model predictive control is applied alone in field practice, the effect is ...

AI Technical Summary

Problems solved by technology

[0005] In order to make full use of the good characteristics of predictive control, it is necessary to be able to obtain the mathematical model more conveniently and accurately, and to effectively overcome the problem of long stable periods caused by large inertia and large lag objects.

Images