Dynamic soft measuring and form establishing method base pulse response formwork and parameter optumization

A technology of impulse response and modeling method, which is applied in the direction of measuring devices, biological neural network models, instruments, etc., and can solve the problems that soft measurement methods cannot be used and are difficult to promote and use.

CN100507460CInactive Publication Date: 2009-07-01TSINGHUA UNIV
4 Cites 2 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Publication Date
2009-07-01
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention belongs to the technical field of production process soft measurement. It is characterized in that the convolution model is established based on the impulse response principle, and the impulse response parameters are used as the main identification object. A set of impulse response templates and template parameter constraint ranges are generated on the basis of the production process mechanism analysis or actual experience, and the impulse response is identified using the optimization method. Responding to the template parameters, the optimization solution of the parameters of the dynamic soft sensor model is completed. The invention has the advantages of: 1) adapting to the requirements of soft sensor modeling and application under the dynamically changing production process. 2) It is suitable for dynamic soft sensor modeling when the difficult-to-measure dominant variable is often manually sampled and tested. For the easy-to-measure auxiliary variable, only the corresponding sampling data during the sampling of the difficult-to-measure dominant variable and the previous transition process are needed to solve the problem. The existing dynamic soft sensor method based on the experimental model represented by the neural network model can only be used for the problem of continuous equal-period sampling. 3) The impulse response parameter identification is limited to the specified template class and template parameter constraints that can be determined based on mechanism analysis and experience, greatly reducing the learning parameters to save learning time.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to a dynamic soft sensing method based on an impulse response template and an optimization method in the production process. This method is suitable for general soft sensor modeling of production process. Background technique

[0002] With the development of production technology and the increasingly complex production process, in order to ensure the safety of production equipment, ensure product quality, and promote the wide application of direct closed-loop control of product quality, quality constraint and safety constraint control, product quality indicators are currently not visible. The important process variables of the production device put forward the requirement of real-time measurement or estimation of some important difficult-to-measure process variables. Due to technical or economical reasons, it is difficult to measure through online process instruments. Therefore, soft measurement and its application technology are...

Examples

Embodiment Construction

[0103] The present invention will be further described below in conjunction with a simulation example.

[0104] In order to be close to the actual production process, 2-input-single-output is selected to generate a process system including two subsystems. The subsystems are second-order inertia plus lagging links. The model parameters of each subsystem are:

[0105] Ω 1 = { K p 1 , T 1 1 , T 2 1 , τ 1 1 } = { 120,15,12,2 } , Ω 2 = { K p 2 ...