CN103878186A - Method for determining hot rolled strip steel laminar cooling temperature - Google Patents

Abstract

The invention discloses a method for determining the laminar flow cooling temperature of hot-rolled strip steel, which belongs to the field of iron and steel metallurgy. The invention includes: collecting a plurality of data in the laminar cooling process to establish a database, and calculating the curling temperature in the laminar cooling process by using a laminar cooling temperature mixed model. The invention can not only predict the temperature value that cannot be collected in practice, but also combine it with actual production data to accurately calculate the curling temperature of the steel plate. At the same time, the parameters can be continuously corrected through self-learning according to the aging of the production line and the renewal of equipment, so as to reduce the error of laminar cooling temperature calculation to a very low level. The invention can make the production data have high precision and consistency with the production line, and has a good guiding effect on the laminar flow cooling production process.

Description

A Method for Determining Laminar Cooling Temperature of Hot-rolled Steel Strip

technical field

The invention belongs to the field of iron and steel metallurgy, and more specifically relates to a method for determining the laminar flow cooling temperature of hot-rolled strip steel.

Background technique

The iron and steel industry is an important pillar industry supporting the development of the national economy, and the development level of the modern iron and steel industry is an important manifestation of a country's technological progress and comprehensive national strength. Laminar flow cooling of hot-rolled strip steel is a process of cooling high-temperature steel plates by water cooling. During this process, changes in steel plate crystals will occur. These changes can directly affect the performance of steel plates. Therefore, the precise control of the temperature of steel plate laminar cooling process It plays an important role in the production of high-quality steel plates. Its performance depends not only on the hot rolling process, but also on the controlled cooling technology after rolling. Whether the hot coil temperature can be controlled within the required range mainly depends on the control of the cooling system of the hot strip after the finishing mill.

The laminar flow cooling process of hot-rolled strip steel is completed by a complex nonlinear, multi-input-output, and variable-coupling system, which involves strip temperature, strip running speed, strip thickness, water spray flow, and strip internal organization. Structure and phase transition and many other issues. How to control such a complex production process and accurately produce strip steel products with different specifications and properties according to requirements is a long-term research problem in the field of iron and steel production. Among them, establishing a perfect and accurate temperature model is the key to achieve good control.

There are two main methods for calculating the laminar cooling temperature. One is based on the mechanism model and supplemented by the identification model. It uses the mechanism formula in the laminar cooling process of the steel plate to establish, and then uses the identified time-varying parameters for calculation. ; The second is based on the experimental identification model, supplemented by the mechanism model. The actual measurement value and the calculation result of the mechanism model are used as the input and output parameters of the experimental identification model to predict the coiling temperature, and the realization is based on a large amount of historical data and current field production. The coiling temperature prediction of .

However, both the mechanism model and the identification model have their own advantages and disadvantages. The mechanism model is closer to the theoretical value of the industrial process, and more physical quantities that cannot be measured in practice can be calculated, and the mechanism analysis can be used to understand the entire system more comprehensively; For physical and chemical changes, there must be actual phenomena that are unknown at this stage, so there are loopholes in the mechanism model built in this way and it is difficult to make up for it. The identification model is driven by data and only establishes the relationship between input and output. The obtained data is closer to the actual situation and more time-sensitive; however, the identification model cannot detect and calculate data that is not detected in the production process.

Contents of the invention

Aiming at the deficiencies of the prior art, the present invention provides a method for determining the laminar flow cooling temperature of hot-rolled strip steel, the purpose of which is to accurately calculate the coiling temperature of the steel plate, and at the same time, it can be passed according to the aging of the production line and the renewal of equipment. Self-learning and continuous correction parameters have a good guiding effect on the production process of laminar cooling.

The invention provides a method for determining the laminar cooling temperature of hot-rolled strip steel, comprising:

Step 1 collect multiple data during the laminar cooling process to establish a database, wherein the multiple data include the trace element content of the steel plate, the initial temperature T ₀ of the steel plate, the thickness of the steel plate, the width of the steel plate, the set coiling temperature T′ _{c of} the steel plate, and the water flow density q _w ;

Step 2 calculates the crimp temperature of the laminar cooling process using a laminar cooling temperature mixing model, wherein the laminar cooling temperature mixing model includes:

Mechanism model, calculate the heat value of air-cooled heat exchange and heat value of heat radiation heat dissipation according to the input amount of steel plate, and calculate the total heat loss in the laminar cooling process in combination with the heat value of water-cooled heat exchange, so as to calculate the laminar flow The coiling temperature in the cooling process, wherein the input amount of the steel plate includes the trace element content of the steel plate, the initial temperature T ₀ of the steel plate, the thickness of the steel plate, the width of the steel plate and the set coiling temperature T′ _c of the steel plate, The calorific value of the water-cooled heat transfer is calculated according to the water-cooled heat transfer coefficient α _w ;

The TS fuzzy model uses fuzzy rules to calculate the corresponding water-cooling heat transfer coefficient α _w according to the input membership function u _i , consequent parameter Θ and input matrix, and calculates the calculated water-cooling heat transfer coefficient α _w is input into the mechanism model to calculate the calorific value of the water-cooled heat transfer, wherein the input matrix is x(k)=[q _w T ₀ T′ _c ], the water flow density q _w , the Said steel plate initial temperature T ₀ and said steel plate set crimping temperature T ' _c composition; And

The self-learning model selects the latest N data from the database for C clustering processing each time it runs, where 10000≤N≤40000, and calculates the membership function u _i of the latest N data and the subsequent parameter Θ, and the calculated membership function u _i and the subsequent parameter Θ are input to the TS fuzzy model

Since the time-varying parameter that has the greatest impact on temperature in the laminar cooling process is the water-cooling heat transfer coefficient, a TS fuzzy model with the water-cooling heat transfer coefficient as the output is established. Using the algorithm based on C clustering to identify the former and subsequent parameters of the TS fuzzy model; using the actual production parameters in the laminar cooling process to identify the TS fuzzy model, where the actual production parameters include water flow density q _w , steel plate initial temperature T _0. Set the coiling temperature T′ _c for the steel plate. When the TS fuzzy model is running, use the fuzzy rules to calculate the corresponding water-cooling heat transfer coefficient α _w according to the input quantity input into the TS fuzzy model.

Further, the step 2 specifically includes the following sub-steps:

(2.1) The self-learning model selects the latest N data from the database for C clustering processing at each run, wherein, 10000≤N≤40000, calculates the number of the latest N data The membership function u _i and the consequent parameter Θ, and the calculated membership function u _i and the consequent parameter Θ are input to the TS fuzzy model;

(2.2) The TS fuzzy model calculates the corresponding water-cooling heat transfer coefficient α _w according to the membership function u _i , the consequent parameter Θ and the input matrix, and the calculated The water cooling heat transfer coefficient α _w is input to the mechanism model;

(2.3) The mechanism model calculates the calorific value of the water-cooled heat transfer according to the water-cooled heat transfer coefficient α _w ;

(2.4) The mechanism model calculates the heat value of the air-cooled heat exchange and the heat value of the heat radiation heat dissipation according to the input amount of the steel plate, and calculates the laminar cooling by combining the heat value of the water-cooled heat exchange said total heat loss during the process;

(2.5) The mechanism model calculates the crimp temperature of the laminar cooling process according to the total heat loss.

Furthermore, the step of calculating the water-cooling heat transfer coefficient α _w by the TS fuzzy model includes the following sub-steps:

(2.2.1) Build i TS fuzzy rules according to the calculated membership function u _i , where the i-th TS fuzzy rule is expressed as:

$\begin{array}{c}{\mathrm{<span\; class="notranslate">\; R</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; :</span>\mathrm{<span\; class="notranslate">\; If</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; is</span>}{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; and</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; is</span>}{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; and</span>}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>\mathrm{<span\; class="notranslate">\; and</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; is</span>}{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}\\ \mathrm{<span\; class="notranslate">\; then</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>\end{array}\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1,2</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>\mathrm{<span\; class="notranslate">\; c</span>}$

为代表各模糊子空间的具有线性隶属度函数u_i的模糊集，用来进行所述第i条模糊规则的模糊推理；

为所述第i条模糊规则的后件参数；yⁱ(k+1)为所述TS模糊模型的输出；Wherein, c is the total number of fuzzy rules; n is the number of input variables of the TS fuzzy model; x ₁ (k), x ₂ (k),..., x _n (k) is the input data at the kth moment; x (k)=[x ₁ (k), x ₂ (k), ..., x _n (k)] is the input matrix of the TS fuzzy model;

is a fuzzy set with a linear membership function u _i representing each fuzzy subspace, and is used for fuzzy inference of the i-th fuzzy rule;

is the consequent parameter of the i fuzzy rule; y ⁱ (k+1) is the output of the TS fuzzy model;

(2.2.2) Define β _i as the fitness of the i-th fuzzy rule, then:

${\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; c</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1,2</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; c</span>}$

Then the calculation formula of the (k+1)th output y(k+1) of the TS fuzzy model is:

$\begin{array}{c}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; c</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; c</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>\\ <span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; c</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>{<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}\end{array}$

Define the consequent parameter Θ(k) and the antecedent parameter Φ(k) as:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; \&Theta;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; \&theta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&theta;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&theta;</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 20</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; cn</span>}}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>\\ \mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>\end{array}\right.$

Among them, r=c·(n+1), you can get:

y(k+1)=Φ(k) ^T Θ(k);

(2.2.3) Define the output y(k+1)=a _w (k), then calculate the water-cooled heat transfer corresponding to the input matrix x(k) through the formula in step (2.2.2) Coefficient α _w (k).

The heat dissipation methods in the laminar cooling process mainly include: thermal radiation heat dissipation and convective heat transfer. The calculation method of each heat dissipation method is as follows:

(1) Thermal radiation

The heat radiation energy per unit area and unit time of high-temperature hot-rolled parts follows the Stefen-Boltzman law, and the heat radiation heat of the steel plate per unit time is described as follows:

${\mathrm{<span\; class="notranslate">\; wxya</span>}}_{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; [</span>{<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 273</span>}}{\mathrm{<span\; class="notranslate">\; 100</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 4</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 273</span>}}{\mathrm{<span\; class="notranslate">\; 100</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 4</span>}}<span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; d\&tau;</span>}$

Among them, _Ar is the surface area of the steel plate, the unit is m ² ; dQ _R is the heat radiation heat of the steel plate per unit time, the unit is J/s; Take 0.8 when there are many scales, and take 0.55~0.65 for smooth surface; σ is the thermal radiation coefficient, σ=5.67W/(m ² ·K ⁴ ); T _s is the surface temperature of the strip steel, in ℃; T _a is the ambient temperature , the unit is °C.

(2) Convective heat transfer

Convective heat transfer refers to the heat transfer phenomenon between the fluid and the solid surface when the fluid (including gas and liquid) flows through the solid. In the laminar cooling process, the water-cooled heat dissipation and air convection heat dissipation of the steel plate are both forms of convective heat transfer, both of which can be described by Newton's cooling formula and determined by different heat transfer coefficients. The heat change of convective heat transfer per unit time includes two aspects: water-cooled heat transfer and air-cooled heat transfer. Among them, the heat value of air-cooled heat transfer per unit time is:

dQ _ca ＝-F·α _a ·(T _s -T _w )·dτ

The heat value of the water-cooled heat exchange per unit time is:

dQ _cw ＝-F·α _w ·(T _s -T _w )·dτ

Among them, F is the contact surface area of steel plate and cooling water, the unit is ^m2 ; T _s is the strip temperature, the unit is ℃; T _w is the cooling water temperature, the unit is ℃; dτ is the contact cooling time, the unit is h; _a is the air-cooling heat transfer coefficient, which takes a value of 20; α _w is the water-cooling heat transfer coefficient, which is calculated by the TS fuzzy model.

The formula for the mechanistic model to calculate the crimp temperature in the laminar cooling process is:

T _c =-ΔQ/cm-T ₀

Among them, T _c represents the crimping temperature of the steel plate; ΔQ represents the total heat loss during the laminar cooling process, which is obtained from the heat value of thermal radiation heat dissipation, the heat value of air-cooled heat transfer and the heat value of water-cooled heat transfer; c represents the ratio of steel plate Heat; m represents the quality of the steel plate; T ₀ represents the initial temperature of the steel plate.

The beneficial effect of the present invention is reflected in: it can be closer to the theoretical value of the industrial process, can calculate more physical quantities that cannot be measured in practice, use mechanism analysis to understand the whole system more comprehensively, and can make the output data and actual production have the same Consistency, and the components that occur in practice but are unknown in theory are included in the water-cooling heat transfer coefficient, so that the water-cooling heat transfer coefficient has the function of making up for theoretical errors. The invention can also continuously track the production line through data update and parameter identification, so that the output data is always consistent with the production line.

In a word, the present invention can not only obtain the whole process change of the dynamic temperature of the laminar flow cooling steel plate, but also make the production data have high precision and consistency with the production line, which has good guiding value for the production process of laminar flow cooling.

Description of drawings

Fig. 1 is the schematic diagram of laminar flow cooling temperature mixing model of the present invention;

Fig. 2 is a flow chart of the method for determining the laminar cooling temperature of the hot-rolled strip in the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not constitute a conflict with each other.

Fig. 1 shows the schematic diagram of the laminar flow cooling temperature mixing model of the present invention, specifically comprises three parts:

(1) Mechanism model

Conservation of mass and energy in the laminar cooling process of hot-rolled strip steel is used to build, and the empirical function relationship is also used to represent the time-varying parameters in the laminar cooling process. The energy conservation mainly includes: thermal radiation heat dissipation, air convection heat dissipation And water cooling heat exchange and other parts. Among them, the water-cooling heat transfer coefficient α _w required to calculate the calorific value of water-cooling heat transfer is provided by TS fuzzy model calculation.

(2) TS fuzzy model

Since the time-varying parameter that has the greatest impact on temperature in the laminar cooling process is the water-cooling heat transfer coefficient, a TS fuzzy model with the water-cooling heat transfer coefficient as the output is established. Use the algorithm based on C clustering to identify the former and subsequent parameters of the TS fuzzy model; use the actual production parameters in the laminar cooling process, that is, the water flow density q _w , the initial temperature T ₀ of the steel plate and the set coiling temperature T′ _c of the steel plate , to identify to build the TS fuzzy model. When the TS fuzzy model is running, use the fuzzy rules to calculate the corresponding water-cooling heat transfer coefficient α _w according to the input quantity input into the TS fuzzy model, and input the calculation results into the mechanism model to perform the heat value of the water-cooling heat transfer calculation. Among them, the key parameters for calculating the water-cooling heat transfer coefficient α _w , that is, the membership function u _i and the consequent parameter Θ are provided by the self-learning model.

(3) Self-learning model:

The self-learning model includes a database and a parameter identifier. The database is used to store the data of the laminar flow cooling production process, namely the trace element content of the steel plate, the initial temperature T ₀ of the steel plate, the thickness of the steel plate, the width of the steel plate, the set coiling temperature T′ _c of the steel plate and the water flow density q _w , and continuously collect new data Instead of the oldest data, the latest 20,000 data are kept in the database. The parameter identifier performs clustering analysis on the data in the database through the C clustering algorithm, calculates the membership function u _i of the data and the subsequent parameter Θ, and inputs the results of each calculation into the TS fuzzy model .

After each production process, the self-learning model imports the newly collected production data into the laminar flow cooling database, and uses the latest data to calculate the front and rear parameters of the water-cooled heat transfer coefficient TS model, so that the TS fuzzy model is consistent with the actual production line consistency.

Fig. 2 shows the flow chart of the method for determining the laminar cooling temperature of hot-rolled strip in the present invention.

Collect multiple data in the laminar cooling process, namely the trace element content of the steel plate, the initial temperature T ₀ of the steel plate, the thickness of the steel plate, the width of the steel plate, the set coiling temperature T′ _c of the steel plate, and the water flow density q _w , and establish a database; The learning model identifies the latest 20,000 data in the database, and obtains two important parameters required for calculation of the TS fuzzy model, namely, the membership function u _i and the consequent parameter Θ, and inputs these two important parameters into the TS In the fuzzy model; the TS fuzzy model calculates the water cooling heat transfer coefficient α _w under this condition according to the input information of the steel plate, namely the water flow density q _w , the initial temperature T ₀ of the steel plate and the set coiling temperature T′ _c of the steel plate, and converts the water cooling The thermal coefficient a _w is input into the mechanism model; the mechanism model calculates the calorific value of the water-cooling heat transfer according to the water-cooling heat transfer coefficient α _w ; the mechanism model also calculates the heat value of the water-cooling heat transfer according to the input amount of the steel plate, that is, the trace element content of the steel plate, the initial temperature of the steel plate, the thickness of the steel plate, and the width of the steel plate Set the curling temperature with the steel plate, and calculate the heat value of air-cooled heat transfer and heat value of heat radiation heat dissipation; the mechanism model combines the calculated heat value of air-cooled heat transfer, heat value of heat radiation heat transfer and heat value of water-cooled heat transfer to get The total heat lost during laminar cooling is used to calculate the crimp temperature for laminar cooling.

The hot-rolled strip steel laminar cooling temperature hybrid model provided by the present invention combines the advantages of the identification model and the mechanism model to propose a hybrid model combining the mechanism model and the identification model, which can calculate the temperature that cannot be achieved in the laminar cooling production process. The measured physical quantities can also make the model output data include unknown physical and chemical processes at this stage. Moreover, this model has self-learning ability, and can adjust the model parameters according to the changes of the production line, so that the calculation output of the model has a good consistency with the output of the production line.

It is easy for those skilled in the art to understand that the above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention, All should be included within the protection scope of the present invention.

Claims (5)

1. a method for definite TEMPERATURE FOR HOT STRIP LAMINAR chilling temperature, comprising:

Step 1 gathers the multiple data building databases in laminar cooling process, and wherein, described multiple data comprise steel plate micronutrient levels, steel plate initial temperature T ₀, steel plate thickness, steel plate width, steel plate set curling temperature T ' _cwith jet density q _w;

Step 2 utilizes laminar flow chilling temperature mixed model to calculate the curling temperature of described laminar cooling process, and wherein, described laminar flow chilling temperature mixed model comprises:

Mechanism model, calculate the calorie value of air cooling heat exchange and the calorie value of heat-radiation heat-dissipating according to steel plate input quantity, and the total amount of heat calculating in laminar cooling process in conjunction with the calorie value of water-cooled heat exchange is scattered and disappeared, thereby calculate the curling temperature of described laminar cooling process, wherein, described steel plate input quantity comprises described steel plate micronutrient levels, described steel plate initial temperature T ₀, described steel plate thickness, described steel plate width and described steel plate set curling temperature T ' _c, the calorie value of described water-cooled heat exchange is according to water-cooled coefficient of heat transfer α _wcalculate;

TS fuzzy model, utilizes fuzzy rule, according to the membership function u of input _i, consequent parameter Θ and input matrix calculate corresponding described water-cooled coefficient of heat transfer α _w, and by the described water-cooled coefficient of heat transfer α calculating _wbe input to the calculating of carrying out the calorie value of described water-cooled heat exchange in described mechanism model, wherein said input matrix is x (k)=[q _wt ₀t ' _c], by described jet density q _w, described steel plate initial temperature T ₀with described steel plate set curling temperature T ' _ccomposition; And

Self learning model selects a up-to-date N data to carry out C clustering processing from described database when each run, wherein, 10000≤N≤40000, calculate the described membership function u of a described up-to-date N data _iand described consequent parameter Θ, and by the described membership function u calculating _iand described consequent parameter Θ inputs to described TS fuzzy model.

2. the method for claim 1, is characterized in that, described step 2 specifically comprises following sub-step:

(2.1) from described database, selected a described up-to-date N data to carry out C clustering processing by described self learning model when each run, wherein, 10000≤N≤40000, calculate the described membership function u of a described up-to-date N data _iand described consequent parameter Θ, and by the described membership function u calculating _iand described consequent parameter Θ inputs to described TS fuzzy model;

(2.2) described TS fuzzy model is according to described membership function u _i, described consequent parameter Θ and described input matrix, calculate corresponding described water-cooled coefficient of heat transfer α _w, and by the described water-cooled coefficient of heat transfer α calculating _wbe input to described mechanism model;

(2.3) described mechanism model is according to described water-cooled coefficient of heat transfer α _wcalculate the calorie value of described water-cooled heat exchange;

(2.4) described mechanism model calculates the calorie value of described air cooling heat exchange and the calorie value of described heat-radiation heat-dissipating according to described steel plate input quantity, and the described total amount of heat calculating in described laminar cooling process in conjunction with the calorie value of described water-cooled heat exchange is scattered and disappeared;

(2.5) described mechanism model scatters and disappears and calculates the described curling temperature of described laminar cooling process according to described total amount of heat.

3. method as claimed in claim 1 or 2, is characterized in that, described TS fuzzy model calculates described water-cooled coefficient of heat transfer α _wstep comprise following sub-step:

(2.2.1) according to the described membership function u calculating _iset up i bar TS fuzzy rule, wherein, i article of TS fuzzy rule is expressed as:

$\begin{array}{c}{R}^i:\mathrm{If}{x}_1\left(k\right)\mathrm{is}{A}_1^i\mathrm{and}{x}_2\left(k\right)\mathrm{is}{A}_2^i\mathrm{and}...\mathrm{and}{x}_n\left(k\right)\mathrm{is}{A}_n^i\\ \mathrm{Then}{y}^i(k+1)=p_0^i+p_1^i{x}_1\left(k\right)+...+p_n^i{x}_n\left(k\right);\end{array}i=\mathrm{1,2},...c$

Wherein, c is fuzzy rule sum; N is the input variable number of described TS fuzzy model; x ₁(k), x ₂(k) ..., x _n(k) be the input data in k moment; X (k)=[x ₁(k), x ₂(k) ..., x _n(k)] be the described input matrix of described TS fuzzy model;

for representing the linear membership function u of having of each fuzzy subspace _ifuzzy set, be used for carrying out the fuzzy reasoning of described i article of fuzzy rule;

for the consequent parameter of described i article of fuzzy rule; y ⁱ(k+1) be the output of described TS fuzzy model;

(2.2.2) definition β _ifor the fitness of described i article of fuzzy rule, have:

${\mathrm{\&beta;}}_i=\underset{j=1}{\overset{c}{\mathrm{\&Sigma;}}}\left(\frac{u_i}{u_j}\right),i=\mathrm{1,2},...,c$

Described TS fuzzy model in the computing formula of (k+1) inferior output y (k+1) is:

$\begin{array}{c}y(k+1)=\underset{i=1}{\overset{c}{\mathrm{\&Sigma;}}}{\mathrm{\&beta;}}_i\&CenterDot;y^i(k+1)=\underset{i=1}{\overset{c}{\mathrm{\&Sigma;}}}{\mathrm{\&beta;}}_i\&CenterDot;(p_0^i+p_1^i{x}_1\left(k\right)+...+p_n^i{x}_n\left(k\right))\\ =\underset{i=1}{\overset{c}{\mathrm{\&Sigma;}}}(p_0^i+p_1^i+...+p_n^i)({\mathrm{\&beta;}}_i+{\mathrm{\&beta;}}_i{x}_1\left(k\right)+...+{\mathrm{\&beta;}}_i{x}_n\left(k\right){)}^T\end{array}$

Definition consequent parameter Θ (k) and former piece parameter Φ (k) are:

$\left\{\begin{array}{c}\mathrm{\&Theta;}\left(k\right)={[{\mathrm{\&theta;}}_1,{\mathrm{\&theta;}}_2,...,{\mathrm{\&theta;}}_r]}^T={[p_{10},p_{20},...,p_{c0},p_{11},p_{12},...,p_{c1},...,p_{\mathrm{cn}}]}^T;\\ \mathrm{\&Phi;}\left(k\right)={[{\mathrm{\&beta;}}_1,...,{\mathrm{\&beta;}}_c,{\mathrm{\&beta;}}_1{x}_1\left(k\right),...,{\mathrm{\&beta;}}_c{x}_1\left(k\right),...,{\mathrm{\&beta;}}_1{x}_n\left(k\right),...,{\mathrm{\&beta;}}_c{x}_n\left(k\right)]}^T;\end{array}\right.$

Wherein, r=c (n+1), can obtain:

y(k+1)＝Φ(k) ^T·Θ(k)；

(2.2.3) define described output y (k+1)=a _w(k), calculate by the formula in step (2.2.2) the described water-cooled coefficient of heat transfer α that described input matrix x (k) is corresponding _w(k).

4. method as claimed in claim 1 or 2, is characterized in that, the formula that described mechanism model calculates the calorie value of described heat-radiation heat-dissipating is:

${\mathrm{dQ}}_R=A_r\&CenterDot;\mathrm{\&epsiv;}\&CenterDot;\mathrm{\&sigma;}[{\left(\frac{T_s+273}{100}\right)}^4-{\left(\frac{T_a+273}{100}\right)}^4]\mathrm{d\&tau;}$

Wherein, dQ _rthe thermal-radiating heat of steel plate in the representation unit time, unit is J/s; A _rfor surface of steel plate area, unit is m ²; ε is the blackness with steel, and its value is 0～1, when belt steel surface oxide skin is more, gets 0.8, and surface smoothing gets 0.55～0.65; σ is heat emissivity coefficient, σ=5.67W/ (m ²k ⁴); T _sfor belt steel surface temperature, unit is DEG C; T _afor environment temperature, unit is DEG C;

The formula of the calorie value of the air cooling heat radiation in the described mechanism model unit of account time is:

dQ _ca＝-F·α _a·(T _s-T _w)·dτ

The formula of the calorie value of the water-cooled heat exchange in the described mechanism model unit of account time is:

dQ _cw＝-F·α _w·(T _s-T _w)·dτ

Wherein, F is steel plate and cooling water contact surface area, and unit is m ²; T _sfor belt steel surface temperature, unit is DEG C; T _wfor cooling water temperature, unit is DEG C; D τ is the cooling time of contact, and unit is h; α _afor the air cooling coefficient of heat transfer, value is 20; α _wfor the described water-cooled coefficient of heat transfer, calculate by described TS fuzzy model;

The formula that described mechanism model calculates the described curling temperature of described laminar cooling process is:

T _c＝-ΔQ/cm-T ₀

Wherein, T _crepresent the curling temperature of steel plate; Δ Q represents that the total amount of heat in laminar cooling process is lost; C represents the specific heat of steel plate; M represents the quality of steel plate; T ₀represent steel plate initial temperature.

5. the method for claim 1, it is characterized in that, former piece parameter and the consequent parameter of TS fuzzy model described in the algorithm identification of utilization based on C cluster, utilize the actual production parameter in described laminar cooling process to carry out identification to set up described TS fuzzy model, wherein said actual production parameter comprises described jet density q _w, described steel plate initial temperature T ₀with described steel plate set curling temperature T '.

Images