CN112989687B - Method for forecasting transverse crack defect of continuous casting billet angle based on metallurgical principle - Google Patents
Method for forecasting transverse crack defect of continuous casting billet angle based on metallurgical principle Download PDFInfo
- Publication number
- CN112989687B CN112989687B CN201911307077.2A CN201911307077A CN112989687B CN 112989687 B CN112989687 B CN 112989687B CN 201911307077 A CN201911307077 A CN 201911307077A CN 112989687 B CN112989687 B CN 112989687B
- Authority
- CN
- China
- Prior art keywords
- refers
- continuous casting
- model
- casting billet
- crystallizer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000009749 continuous casting Methods 0.000 title claims abstract description 57
- 230000007547 defect Effects 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000003062 neural network model Methods 0.000 claims abstract description 27
- 238000004519 manufacturing process Methods 0.000 claims abstract description 14
- 229910000831 Steel Inorganic materials 0.000 claims description 25
- 239000010959 steel Substances 0.000 claims description 25
- 238000001816 cooling Methods 0.000 claims description 21
- 238000012937 correction Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 17
- 230000008859 change Effects 0.000 claims description 16
- 230000004907 flux Effects 0.000 claims description 14
- 238000012549 training Methods 0.000 claims description 11
- 238000005266 casting Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 9
- 239000007921 spray Substances 0.000 claims description 9
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 9
- 230000035945 sensitivity Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000004069 differentiation Effects 0.000 claims description 6
- 239000002893 slag Substances 0.000 claims description 6
- 238000005336 cracking Methods 0.000 claims description 5
- 230000001133 acceleration Effects 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 230000005484 gravity Effects 0.000 claims description 3
- 210000002364 input neuron Anatomy 0.000 claims description 3
- 239000007788 liquid Substances 0.000 claims description 3
- 210000004205 output neuron Anatomy 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims 2
- 238000013528 artificial neural network Methods 0.000 claims 1
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 6
- 229910052742 iron Inorganic materials 0.000 description 3
- 229910000742 Microalloyed steel Inorganic materials 0.000 description 2
- 229910052758 niobium Inorganic materials 0.000 description 2
- 239000010955 niobium Substances 0.000 description 2
- GUCVJGMIXFAOAE-UHFFFAOYSA-N niobium atom Chemical compound [Nb] GUCVJGMIXFAOAE-UHFFFAOYSA-N 0.000 description 2
- 229910001339 C alloy Inorganic materials 0.000 description 1
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- CFJRGWXELQQLSA-UHFFFAOYSA-N azanylidyneniobium Chemical compound [Nb]#N CFJRGWXELQQLSA-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000002436 steel type Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Continuous Casting (AREA)
Abstract
The invention belongs to the technical field of metallurgical continuous casting, and particularly relates to a method for forecasting transverse crack defects of a continuous casting billet angle based on a metallurgical principle. Aiming at the defect of on-line prediction of continuous casting billet angle transverse crack defects in the prior art, the method for predicting the continuous casting billet angle transverse crack defects based on the metallurgical principle is provided, and comprises the following steps: firstly, establishing a principle model for key factors of the continuous casting billet which generate the angle transverse crack defect through analyzing the metallurgical principle of the continuous casting billet; taking the principle model as a node of an input layer, and establishing a BP neural network model; learning the established BP neural network model by using the existing continuous casting billet metallurgical sample data until a verified BP neural network model is obtained; and (3) taking the actual production continuous casting billet sample parameters into the verified BP neural network model to obtain a result of predicting the continuous casting billet angle transverse crack defect, and providing reference for production and manufacture.
Description
Technical Field
The invention belongs to the technical field of metallurgical continuous casting, and particularly relates to a method for forecasting transverse crack defects of a continuous casting billet angle based on a metallurgical principle.
Background
With the rapid development of iron and steel enterprises, the hot charging and hot feeding and direct rolling technology of continuous casting blanks enables continuous casting steel to become an important research field of the iron and steel enterprises, and the hot charging and hot feeding of continuous casting can greatly reduce the production cost and capital equipment investment of the iron and steel enterprises, and can remarkably improve the product quality and the enterprise competitiveness. However, in the continuous casting process, as the brittleness of a casting blank is improved due to precipitation of fine carbon and niobium nitride in the steel, the high-temperature mechanical property of the niobium-containing steel is greatly changed, the third brittleness area of the steel is widened and deepened, the crack sensitivity of the steel is enhanced, and the occurrence rate of transverse crack of the continuous casting slab is far higher than that of other steel types.
Chinese patent document CN108145113a discloses a two-cold width cutting control method for reducing the transverse crack of the angle of the continuous casting slab of the niobium-containing microalloyed steel; chinese patent document CN103846401B discloses a two-cooling process for improving the surface quality of extra thick slabs. The above patent proposes a solution to the transverse crack of the casting blank corner from the standpoint of the two-stage cooling process. Chinese patent document CN104259412a discloses a mold flux for continuous casting of low-carbon alloy steel in extra-thick slabs and a preparation method thereof, and describes a method for solving the corner transverse crack from the aspect of mold flux. Chinese patent CN 107287487a discloses a method for avoiding the occurrence of billet corner cracking by adjusting the composition of the J55 grade microalloyed steel.
In summary, in the prior art, an analysis method for solving the transverse crack of the continuous casting billet is attempted based on the principle of metallurgical foundation, but in the continuous casting billet production process, online prediction of the transverse crack defect of the continuous casting billet is required, and the method is hoped to play a role in guiding production.
Disclosure of Invention
The technical problem to be solved by the invention is that the online prediction of the continuous casting billet angle transverse crack defect by the prior art method is insufficient, and the production process is difficult to guide.
The technical scheme provided by the invention for solving the technical problems is that a method for forecasting the transverse crack defect of the continuous casting billet angle based on the metallurgical principle is provided, and comprises the following steps:
step 1: establishing a principle model for key factors of the continuous casting billet which generate the angle transverse crack defect through analyzing the metallurgical principle of the continuous casting billet;
step 2: taking the principle model as a node of an input layer, and establishing a BP neural network model;
step 3: performing a learning process on the BP neural network model established in the step 2 by using the existing continuous casting billet metallurgical sample data until a verified BP neural network model is obtained;
step 4: and (3) carrying the sample parameters of the practically produced continuous casting billet into the BP neural network model verified in the step (3), and obtaining the result of predicting the transverse crack defect of the continuous casting billet through a judging process.
Further, the principle model in the step 1 comprises a crack sensitivity index model, a molten steel superheat degree and blank drawing speed change model, a crystallizer vibration model, a crystallizer cooling uniformity model, a mold flux control model and a secondary cooling control model.
Crack sensitivity index model: x's' 1 ＝b′ 1 +b 1 x 1 +b 2 x 2 +...+b n x n Wherein x' 1 Refers to the probability of cracking of a continuous casting billet, x 1 ,x 2 ...x n Refers to the probability of crack influence of various elements on the continuous casting billet, b' 1 ,b 1 ,b 2 ...b n Refers to the impact weights of the various elements.
Molten steel superheat degree and blank drawing speed change model: x's' 2 ＝c′ 1 +c 1 N+c 2 a, wherein x' 2 Refers to the index of variation of the superheat degree of molten steel and the withdrawal speed, N refers to the offline data of the superheat degree of molten steel, and a refers to the variation of the withdrawal speedDeltav is the speed change amount, deltat is the time required for the speed change, and the change of the drawing speed has an upper limit value A, c' 1 ,c 1 ,c 2 Refers to influencing the weights.
Mold vibration model: x's' 3 ＝d′ 1 +d 1 t f +d 2 A f +d 3 f, wherein x' 3 Refers to the vibration index of the crystallizer, t f Refers to the negative slip time of the vibration of the crystallizer, A f The vibration deflection amount is f, the internal friction force f=2μ rhoghb of the crystallizer, μ is the transverse friction coefficient (μ=0.4-0.6) of the shell, b is the width of the casting blank, ρ is the density of molten steel, g is the gravity acceleration, h is the depth from the liquid level of the crystallizer, and d '' 1 ,d 1 ,d 2 ,d 3 Refers to influencing the weights.
Crystallizer cooling uniformity model x' 4 ＝e′ 1 +e 1 ×{A 1 ,A 2 }+e 2 ×Max{B 1 ,B 2 ,B 11 ,B 22 }+e 3 ×Max{C 1 ,C 2 -x 'where' 4 Refers to the cooling uniformity index of a crystallizer, A 1 ,A 2 Respectively refers to the left and right taper of the crystallizer, B 1 ,B 2 Respectively refers to the heat flow density of the wide surface of the crystallizer, B 11 ,B 22 Respectively refers to the heat flow density, C, of the narrow surface of the crystallizer 1 Refers to the heat flow density inside and outside the wide surface of the crystallizer, C 2 Refers to the heat flow density, e 'inside and outside the narrow surface of the crystallizer' 1 ,e 1 ,e 2 ,e 3 Refers to influencing the weights.
Mold flux control model: x's' 5 ＝f′ 1 +f 1 N x +f 2 R+f 3 f, wherein x' 5 Refers to the suitability index of the covering slag, N x Refers to the consumption of the protecting slag, R refers to the alkalinity of the protecting slag, f refers to the viscosity of the protecting slag, f' 1 ,f 1 ,f 2 ,f 3 Refers to influencing the weights.
And (3) a secondary cooling control model: x's' 6 ＝g′ 1 +g 1 W+g 2 Q+g 3 S, where x' 6 Refers to the influence index of the secondary cooling zone, W refers to the water spray intensity, Q refers to the water spray quantity, S refers to the water spray area, g' 1 ,g 1 ,g 2 ,g 3 Refers to influencing the weights.
Further, the BP neural network model in the step 2 comprises three layers of structures of an input layer, an hidden layer and an output layer, wherein the input neurons of the input layer represent a principle model, the input layer and the hidden layer are associated through the connection weights of the two middle layers, the hidden layer and the output layer are associated through the connection weights of the two middle layers, and 1 output neuron of the output layer represents a result of continuous casting billet angle transverse crack defect prediction.
Further, the learning process in step 3 includes the steps of:
1) Network initialization: (1) initializing the number of network input nodes, the number of hidden layer nodes and the number of output nodes; (2) initializing weights and thresholds (weights ω) ij 、ω j And threshold a j B, respectively assigning random numbers in a section (-1, 1); (3) selecting K samples as training samples, each node of each sampleInput value x i (k) I= (1, 2 …, n), k= (1, 2 …, K); the expected output value is y (k), and the actual output value is O (k); (4) setting a local error function for a single sample (5) Setting the global error function of all samples to +.> Initializing a calculation precision value epsilon and a maximum iteration number T, and initializing the current iteration number t=1; the initial learning rate η.
2) At the t-th iteration, each training sample is circularly read, and the input value of the kth training sample is x i (k)，i＝(1，2…，n)，k＝(1，2…，K)。
3) The actual output value O (k) of each output node of the obtained sample is calculated layer by layer, and the calculation process is as follows: calculating the output value H of each node of the hidden layer j ：
Calculating an actual output value O (k) of the output node:
4) Calculating an error function E (k) for a single sample
5) Correcting network weight and threshold
The correction of the weight in the BP algorithm is in direct proportion to the partial differentiation of the error E (k) to the weight, and the correction of the threshold is in direct proportion to the partial differentiation of the error E (k) to the threshold; thereby calculating the weight correction amount and the threshold correction amount.
ω j Is a correction amount of (a):
ω j corrected values: omega j ＝ω j +Δω j ＝ω j +ηH j (y(k)-O(k))
Correction amount of b:
b corrected value: b=b+Δb=b+η (y (k) -O (k))
If the following steps are made: δ=y (k) -O (k)
Then:
ω j ＝ω j +ηH j δ
b＝b+ηδ
in the above formula: j=1, 2, …, L;
δ j ＝H j (1-H j )δω j
ω ij ＝ω ij +ηδ j x i (k)
a j ＝a j +ηδ j
in the above formula: i=1, 2, …, n; j=1, 2, …, L;
6) If K is less than K, let k=k+1, repeat steps 3, 4, 5; otherwise, continuing.
7) After all samples have been trained, a global error E is calculated
Stopping the iteration when E < epsilon or t=t; otherwise, let k=1, t=t+1, repeat steps 2), 3), 4), 5), 6), 7).
Further, the determining process in step 4 uses the weight ω obtained by the learning process in step 3) ij 、ω j And threshold a i B, predicting a sample of an actual production continuous casting billet, which specifically comprises the following steps:
1) The input parameter expression of the sample of the actual production continuous casting billet is: x's' i (i＝1，2，…，n)。
2) Calculating the output value of each node of the hidden layer of the BP neural network model:
3) Calculating an expected output value O of a node of the BP neural network model output layer:
4) When O is less than 0.5, judging that the defect does not exist; when O >0.5, it is judged that it has a defect.
The method has the beneficial effects that the method for forecasting the transverse crack defect of the continuous casting billet angle based on the metallurgical principle is provided, the reasons for the defect generation are essentially forecasted based on the metallurgical principle by combining a statistical method and a neural network model, and the key factors or key control links for the defect generation are positioned so as to provide references for production and manufacture or feedforward control.
Drawings
Fig. 1 is a schematic diagram of a three-layer BP network structure.
Reference sign, X is input variable, n is input layer node number, namely casting blank quality defect main influence parameter number, Y is output variable, the output node number is only 1, namely whether casting blank quality defect exists, H is hidden layer output value, the hidden layer node number is L, H j Output value of j-th output node representing hidden layer, j= (1, 2, …, L), ω ij Is the connection weight between the i node of the input layer and the j node of the hidden layer, i= (1, 2)，…，n)，j＝(1，2，…，L)、ω j As the connection weight between the j-th node of the hidden layer and the node of the output layer, j= (1, 2, …, L), a j J= (1, 2, …, L) and b are thresholds of nodes of the output layer.
Detailed Description
The specific embodiment of the invention is described in detail with reference to the accompanying drawings, and the method for forecasting the transverse crack defect of the continuous casting billet angle based on the metallurgical principle comprises the following steps:
step 1: by analyzing the metallurgical principle of the continuous casting billet, a principle model is established for key factors of the angle transverse crack defect of the continuous casting billet, wherein the principle model comprises a crack sensitivity index model, a molten steel superheat degree and billet drawing speed change model, a crystallizer vibration model, a crystallizer cooling uniformity model, a casting powder control model and a secondary cooling control model.
Crack sensitivity index model: x's' 1 ＝b′ 1 +b 1 x 1 +b 2 x 2 +…+b n x n Wherein x' 1 Refers to the probability of cracking of a continuous casting billet, x 1 ，x 2 …x n Refers to the probability of crack influence of various elements on the continuous casting billet, b' 1 ，b 1 ，b 2 …b n Refers to the impact weights of the various elements.
Molten steel superheat degree and blank drawing speed change model: x's' 2 ＝c′ 1 +c 1 N+c 2 a, wherein x' 2 Refers to the index of variation of the superheat degree of molten steel and the withdrawal speed, N refers to the offline data of the superheat degree of molten steel, and a refers to the variation of the withdrawal speedDeltav is the speed change amount, deltat is the time required for the speed change, and the change of the drawing speed has an upper limit value A, c' 1 ，c 1 ，c 2 Refers to influencing the weights.
Mold vibration model: x's' 3 ＝d′ 1 +d 1 t f +d 2 A f +d 3 f, wherein x' 3 Refers to the vibration index of the crystallizer, t f Refers to the negative slip time of the vibration of the crystallizer, A f The vibration deflection amount is f, the internal friction force f=2μ rhoghb of the crystallizer, μ is the transverse friction coefficient (μ=0.4-0.6) of the shell, b is the width of the casting blank, ρ is the density of molten steel, g is the gravity acceleration, h is the depth from the liquid level of the crystallizer, and d '' 1 ，d 1 ，d 2 ，d 3 Refers to influencing the weights.
Crystallizer cooling uniformity model x' 4 ＝e′ 1 +e 1 ×{A 1 ，A 2 }+e 2 ×Max{B 1 ，B 2 ，B 11 ，B 22 }+e 3 ×Max{C 1 ，C 2 X, where x 4 Refers to the cooling uniformity index of a crystallizer, A 1 ，A 2 Respectively refers to the left and right taper of the crystallizer, B 1 ，B 2 Respectively refers to the heat flow density of the wide surface of the crystallizer, B 11 ，B 22 Respectively refers to the heat flow density, C, of the narrow surface of the crystallizer 1 Refers to the heat flow density inside and outside the wide surface of the crystallizer, C 2 Refers to the heat flow density, e 'inside and outside the narrow surface of the crystallizer' 1 ，e 1 ，e 2 ，e 3 Refers to influencing the weights.
Mold flux control model: x's' 5 ＝f 1 ′+f 1 N x +f 2 R+f 3 f, wherein x' 5 Refers to the suitability index of the covering slag, N x Refers to the consumption of the mold flux, R refers to the alkalinity of the mold flux, f refers to the viscosity of the mold flux, f 1 ′，f 1 ，f 2 ，f 3 Refers to influencing the weights.
And (3) a secondary cooling control model: x's' 6 ＝g′ 1 +g 1 W+g 2 Q+g 3 S, where x' 6 Refers to the influence index of the secondary cooling zone, W refers to the water spray intensity, Q refers to the water spray quantity, S refers to the water spray area, g' 1 ，g 1 ，g 2 ，g 3 Refers to influencing the weights.
Step 2: using the principle model as the node of the input layer, building three-layer BP neural network model as shown in figure 1X is an input variable, n is the number of nodes of an input layer, namely the number of main influencing parameters of casting blank quality defects, Y is an output variable, the number of the output nodes is only 1, namely whether casting blank quality defects exist or not, H is an output value of an hidden layer, and the number of the nodes of the hidden layer is L, H j Output value of j-th output node representing hidden layer, j= (1, 2, …, L), ω ij Is the connection weight between the i node of the input layer and the j node of the hidden layer, i= (1, 2, …, n), j= (1, 2, …, L), ω j As the connection weight between the j-th node of the hidden layer and the node of the output layer, j= (1, 2, …, L), a j J= (1, 2, …, L) and b are thresholds of nodes of the output layer. The BP neural network model comprises a multi-input layer, a single hidden layer and a single output layer, wherein the input neurons of the input layer represent a principle model, the input layer is associated with the hidden layer through the connection weight of the middle layer of the input layer and the hidden layer, the hidden layer is associated with the output layer through the connection weight of the middle layer of the hidden layer and the output layer, and 1 output neuron of the output layer represents the result of continuous casting billet angle transverse crack defect prediction.
Step 3: and (3) carrying out a learning process on the BP neural network model established in the step (2) by using the existing continuous casting billet metallurgical sample data until a verified BP neural network model is obtained, wherein the learning process comprises the following steps:
1) Network initialization: (1) initializing the number of network input nodes, the number of hidden layer nodes and the number of output nodes; (2) initializing weights and thresholds (weights ω) ij 、ω j And threshold a j B, respectively assigning random numbers in a section (-1, 1); (3) k samples are selected as training samples, and the input value of each node of each sample is x i (k) I= (1, 2 …, n), k= (1, 2 …, K); the expected output value is y (k), and the actual output value is O (k); (4) setting a local error function for a single sample (5) Setting upThe global error function for all samples is +.> Initializing a calculation precision value epsilon and a maximum iteration number T, and initializing the current iteration number t=1; the initial learning rate η.
2) At the t-th iteration, each training sample is circularly read, and the input value of the kth training sample is x i (k)，i＝(1，2…，n)，k＝(1，2…，K)。
3) The actual output value O (k) of each output node of the obtained sample is calculated layer by layer, and the calculation process is as follows: calculating the output value H of each node of the hidden layer j ：
Calculating an actual output value O (k) of the output node:
4) Calculating an error function E (k) for a single sample
5) Correcting network weight and threshold
The correction of the weight in the BP algorithm is in direct proportion to the partial differentiation of the error E (k) to the weight, and the correction of the threshold is in direct proportion to the partial differentiation of the error E (k) to the threshold; thereby calculating the weight correction amount and the threshold correction amount.
ω j Is a correction amount of (a):
ω j corrected values: omega j ＝ω j +Δω j ＝ω j +ηH j (y(k)-O(k))
Correction amount of b:
b corrected value: b=b+Δb=b+η (y (k) -O (k))
If the following steps are made: δ=y (k) -O (k)
Then:
ω j ＝ω j +ηH j δ
b＝b+ηδ
in the above formula: j=1, 2, …, L;
δ j ＝H j (1-H j )δω j
ω ij ＝ω ij +ηδ j x i (k)
a j ＝a j +ηδ j
in the above formula: i=1, 2, …, n; j=1, 2, …, L;
6) If K is less than K, let k=k+1, repeat steps 3, 4, 5; otherwise, continuing.
7) After all samples have been trained, a global error E is calculated
Stopping the iteration when E < epsilon or t=t; otherwise, let k=1, t=t+1, repeat steps 2), 3), 4), 5), 6), 7).
Step 4: and (3) bringing sample parameters of the practically produced continuous casting billet into the BP neural network model verified in the step (3), and obtaining a result of predicting the transverse crack defect of the continuous casting billet through a judging process, wherein the judging process comprises the following steps:
1) Sample input parameters of actual production continuous casting billets: x's' i (i＝1，2，…，n)。
2) Calculating the output value of each node of the hidden layer of the BP neural network model:
3) Calculating an expected output value O of a node of the BP neural network model output layer:
4) When O is less than 0.5, judging that the defect does not exist; when O >0.5, it is judged that it has a defect.
The present invention is not limited to the specific technical solutions described in the above embodiments, and other embodiments may be provided in addition to the above embodiments. Any modifications, equivalent substitutions, improvements, etc. made by those skilled in the art, which are within the spirit and principles of the present invention, are intended to be included within the scope of the present invention.
Claims (4)
1. The method for forecasting the transverse crack defect of the continuous casting billet angle based on the metallurgical principle is characterized by comprising the following steps:
step 1: establishing a key factor principle model for key factors of the angle transverse crack defect of the continuous casting billet, wherein the key factors comprise crack sensitivity index, molten steel superheat degree, billet drawing speed change, crystallizer vibration, crystallizer cooling uniformity, mold flux control and secondary cooling control;
step 2: establishing a BP neural network model by taking the principle model as a node of an input layer;
step 3: performing a learning process on the BP neural network model established in the step 2 by using the existing continuous casting billet metallurgical sample data until a verified BP neural network model is obtained;
step 4: carrying the parameters of the sample of the practically produced continuous casting billet into the verification BP neural network model in the step 3, and judging to obtain the result of predicting the transverse crack defect of the continuous casting billet; the determination is as follows:
when the expected output value O of the node of the verification BP neural network model output layer is less than 0.5, judging that the node has no defect; when O >0.5, judging that the defect exists; the key factor principle model comprises a crack sensitivity index model, a molten steel superheat degree and blank drawing speed change model, a crystallizer vibration model, a crystallizer cooling uniformity model, a mold flux control model and a secondary cooling control model;
the crack sensitivity index model is expressed by the following formula: x's' 1 ＝b′ 1 +b 1 x 1 +b 2 x 2 +...+b n x n Wherein x' 1 Refers to the probability of cracking of a continuous casting billet, x 1 ,x 2 ...x n Refers to the probability of crack influence of various elements on the continuous casting billet, b' 1 ,b 1 ,b 2 ...b n The influence weights of various elements are referred;
the molten steel superheat degree and blank drawing speed change model is expressed by the following formula: x's' 2 ＝c′ 1 +c 1 N+c 2 a, wherein x' 2 Refers to the index of variation of the superheat degree of molten steel and the withdrawal speed, N refers to the offline data of the superheat degree of molten steel, and a refers to the variation of the withdrawal speedDeltav is the speed change amount, deltat is the time required for the speed change, and the change of the drawing speed has an upper limit value A, c' 1 ,c 1 ,c 2 Refers to influencing weights;
the mold vibration model is expressed by the following formula: x's' 3 ＝d′ 1 +d 1 t f +d 2 A f +d 3 f, wherein x' 3 Refers to the vibration index of the crystallizer, t f Refers to the negative slip time of the vibration of the crystallizer, A f The vibration deflection amount is f, the internal friction force f=2μ rhoghb of the crystallizer, μ is the transverse friction coefficient (μ=0.4-0.6) of the shell, b is the width of the casting blank, ρ is the density of molten steel, g is the gravity acceleration, h is the depth from the liquid level of the crystallizer, and d '' 1 ,d 1 ,d 2 ,d 3 Refers to influencing weights;
the mold cooling uniformity model is expressed by the following formula:
x' 4 ＝e′ 1 +e 1 ×{A 1 ,A 2 }+e 2 ×Max{B 1 ,B 2 ,B 11 ,B 22 }+e 3 ×Max{C 1 ,C 2 -x 'where' 4 Refers to the cooling uniformity index of a crystallizer, A 1 ,A 2 Respectively refers to the left and right taper of the crystallizer, B 1 ,B 2 Respectively refers to the heat flow density of the wide surface of the crystallizer, B 11 ,B 22 Respectively refers to the heat flow density, C, of the narrow surface of the crystallizer 1 Refers to the heat flow density inside and outside the wide surface of the crystallizer, C 2 Refers to the heat flow density, e 'inside and outside the narrow surface of the crystallizer' 1 ,e 1 ,e 2 ,e 3 Refers to influencing weights;
the mold flux control model is expressed by the following formula: x's' 5 ＝f′ 1 +f 1 N x +f 2 R+f 3 f, wherein x' 5 Refers to the suitability index of the covering slag, N x Refers to the consumption of the mold flux, R refers to the alkalinity of the mold flux, f refers to the viscosity of the mold flux, f 1 ',f 1 ,f 2 ,f 3 Refers to influencing weights;
the two-cold control model is expressed by the following formula: x's' 6 ＝g′ 1 +g 1 W+g 2 Q+g 3 S, where x' 6 Refers to the influence index of the secondary cooling zone, W refers to the water spray intensity, Q refers to the water spray quantity, S refers to the water spray area, g' 1 ,g 1 ,g 2 ,g 3 Refers to influencing the weights.
2. The method for forecasting the continuous casting billet angle transverse crack defect based on the metallurgical principle according to claim 1, wherein the BP neural network model comprises a multi-input layer, a single hidden layer and a single output layer three-layer structure, input neurons of the input layer represent the principle model, the input layer and the hidden layer are associated through connection weights of intermediate layers of the input layer and the hidden layer, the hidden layer and the output layer are associated through connection weights of intermediate layers of the hidden layer and the output layer, and 1 output neuron of the output layer represents a continuous casting billet angle transverse crack defect forecasting result.
3. A method for predicting a billet corner crack defect based on metallurgical principles according to claim 1, wherein the learning process comprises the steps of:
1) Network initialization: (1) initializing the number of input nodes, the number of hidden layer nodes and the number of output nodes of the BP neural network; (2) initializing weights and thresholds (weights ω) ij 、ω j And threshold a j B, respectively assigning random numbers in a section (-1, 1); (3) k samples are selected as training samples, and the input value of each node of each sample is x i (k) I= (1, 2 …, n), k= (1, 2 …, K); the expected output value is y (k), and the actual output value is O (k); (4) setting a local error function for a single sample(5) Setting the global error function of all samples asInitializing a calculation precision value epsilon and a maximum iteration number T, and initializing the current iteration number t=1; an initial learning rate η;
2) At the t-th iteration, each training sample is circularly read, and the input value of the kth training sample is x i (k)，i＝(1,2…,n),k＝(1,2…,K)；
3) And calculating layer by layer, and finally calculating to obtain an actual output value O (k) of each output node of the training sample, wherein the calculation process is as follows: calculating the output value H of each node of the hidden layer j ：
Calculating an actual output value O (k) of the output node:
4) Calculating an error function E (k) for a single training sample
5) Correcting network weight and threshold
The correction of the weight in the BP algorithm is in direct proportion to the partial differentiation of the error E (k) to the weight, and the correction of the threshold is in direct proportion to the partial differentiation of the error E (k) to the threshold; thereby calculating a weight correction amount and a threshold correction amount:
ω j is a correction amount of (a):
ω j corrected values: omega j ＝ω j +Δω j ＝ω j +ηH j (y(k)-O(k))
Correction amount of b:
b corrected value: b=b+Δb=b+η (y (k) -O (k))
Let delta=y (k) -O (k)
Then:
ω j ＝ω j +ηH j δ
b＝b+ηδ
in the above formula: j=1, 2, …, L;
δ j ＝H j (1-H j )δω j
ω ij ＝ω ij +ηδ j x i (k)
a j ＝a j +ηδ j
in the above formula: i=1, 2, …, n; j=1, 2, …, L;
6) If K is less than K, let k=k+1, repeat steps 3, 4, 5; otherwise, continuing;
7) After all samples have been trained, a global error E is calculated
When E < epsilon or t=t, stopping the iteration; otherwise, let k=1, t=t+1, repeat steps 2), 3), 4), 5), 6), 7).
4. The method for predicting billet corner cracking defect based on metallurgical principle according to claim 1, wherein the determining step uses the weight ω obtained by the learning step in step 3 ij 、v j And threshold a j B, predicting a sample of an actual production continuous casting billet, which specifically comprises the following steps:
1) Sample input parameters of actual production continuous casting billets: x is x i ′(i＝1,2,…,n)；
2) Calculating the output value of each node of the hidden layer of the BP neural network model:
3) Calculating an expected output value O of a node of the BP neural network model output layer:
4) When O <0.5, judging that the defect does not exist; when O >0.5, it is judged that it has a defect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911307077.2A CN112989687B (en) | 2019-12-18 | 2019-12-18 | Method for forecasting transverse crack defect of continuous casting billet angle based on metallurgical principle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911307077.2A CN112989687B (en) | 2019-12-18 | 2019-12-18 | Method for forecasting transverse crack defect of continuous casting billet angle based on metallurgical principle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112989687A CN112989687A (en) | 2021-06-18 |
| CN112989687B true CN112989687B (en) | 2023-11-07 |
Family
ID=76343773
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911307077.2A Active CN112989687B (en) | 2019-12-18 | 2019-12-18 | Method for forecasting transverse crack defect of continuous casting billet angle based on metallurgical principle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112989687B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114603090B (en) * | 2022-03-11 | 2023-06-16 | 北京海卓博尔科技有限公司 | Crystallizer vibration driving device, control method and control system |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299032A (en) * | 2008-06-06 | 2008-11-05 | 重庆交通大学 | Structural crack bionic monitoring system and monitoring method thereof |
| CN101995436A (en) * | 2010-10-26 | 2011-03-30 | 江苏大学 | Neural network based drawing part crack monitoring method |
| CN102319883A (en) * | 2011-10-09 | 2012-01-18 | 北京首钢自动化信息技术有限公司 | Method for controlling on-line prediction of continuous casting blank quality |
| CN102653835A (en) * | 2012-05-09 | 2012-09-05 | 东北大学 | Method for reducing transverse cracks at corner of continuous casting billet of boron-steel-containing wide-thick plate |
| CN102937784A (en) * | 2012-10-30 | 2013-02-20 | 中冶南方工程技术有限公司 | Artificial neural network based method for controlling online prediction of casting billet quality |
| CN103920859A (en) * | 2013-01-14 | 2014-07-16 | 中冶南方工程技术有限公司 | Continuous casting sheet billet internal crack online prediction method |
| CN104889358A (en) * | 2014-03-05 | 2015-09-09 | 鞍钢股份有限公司 | Method for controlling surface crack of continuous cast slab |
| CN105328155A (en) * | 2015-10-08 | 2016-02-17 | 东北电力大学 | Steel leakage visualized characteristic forecasting method based on improved neural network |
| CN105911095A (en) * | 2016-05-04 | 2016-08-31 | 东北电力大学 | Visual recognition method of continuous casting billet surface longitudinal cracks |
| CN106077555A (en) * | 2016-08-12 | 2016-11-09 | 湖南千盟物联信息技术有限公司 | A kind of continuous casting coordinating and optimizing control method |
| CN107609647A (en) * | 2017-10-16 | 2018-01-19 | 安徽工业大学 | One kind is based on BP neural network roll alloy mechanical property Forecasting Methodology |
| CN110202106A (en) * | 2019-06-04 | 2019-09-06 | 甘肃酒钢集团宏兴钢铁股份有限公司 | The method for controlling CSP thin sheet continuous casting machine production medium carbon alloy steel slab surface cracks |
-
2019
- 2019-12-18 CN CN201911307077.2A patent/CN112989687B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299032A (en) * | 2008-06-06 | 2008-11-05 | 重庆交通大学 | Structural crack bionic monitoring system and monitoring method thereof |
| CN101995436A (en) * | 2010-10-26 | 2011-03-30 | 江苏大学 | Neural network based drawing part crack monitoring method |
| CN102319883A (en) * | 2011-10-09 | 2012-01-18 | 北京首钢自动化信息技术有限公司 | Method for controlling on-line prediction of continuous casting blank quality |
| CN102653835A (en) * | 2012-05-09 | 2012-09-05 | 东北大学 | Method for reducing transverse cracks at corner of continuous casting billet of boron-steel-containing wide-thick plate |
| CN102937784A (en) * | 2012-10-30 | 2013-02-20 | 中冶南方工程技术有限公司 | Artificial neural network based method for controlling online prediction of casting billet quality |
| CN103920859A (en) * | 2013-01-14 | 2014-07-16 | 中冶南方工程技术有限公司 | Continuous casting sheet billet internal crack online prediction method |
| CN104889358A (en) * | 2014-03-05 | 2015-09-09 | 鞍钢股份有限公司 | Method for controlling surface crack of continuous cast slab |
| CN105328155A (en) * | 2015-10-08 | 2016-02-17 | 东北电力大学 | Steel leakage visualized characteristic forecasting method based on improved neural network |
| CN105911095A (en) * | 2016-05-04 | 2016-08-31 | 东北电力大学 | Visual recognition method of continuous casting billet surface longitudinal cracks |
| CN106077555A (en) * | 2016-08-12 | 2016-11-09 | 湖南千盟物联信息技术有限公司 | A kind of continuous casting coordinating and optimizing control method |
| CN107609647A (en) * | 2017-10-16 | 2018-01-19 | 安徽工业大学 | One kind is based on BP neural network roll alloy mechanical property Forecasting Methodology |
| CN110202106A (en) * | 2019-06-04 | 2019-09-06 | 甘肃酒钢集团宏兴钢铁股份有限公司 | The method for controlling CSP thin sheet continuous casting machine production medium carbon alloy steel slab surface cracks |
Non-Patent Citations (2)
| Title |
|---|
| 基于神经网络的连铸板坯质量在线诊断系统;郭贤利等;《冶金自动化》;第37卷(第3期);16-22 * |
| 江中块等.梅钢连铸机板坯角横裂缺陷解决实践.《 2014年全国炼钢—连铸生产技术会论文集》.2014,548-552. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112989687A (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108637020B (en) | Self-adaptive variation PSO-BP neural network strip steel convexity prediction method | |
| US20090138223A1 (en) | On-Line Quality Prediction System for Stainless Steel Slab and the Predicting Method Using It | |
| EP3912740B1 (en) | Production specification determination method, production method, and production specification determination apparatus for metal material | |
| CN105911095A (en) | Visual recognition method of continuous casting billet surface longitudinal cracks | |
| Zheng et al. | Optimization of high-pressure die-casting process parameters using artificial neural network | |
| CN112989687B (en) | Method for forecasting transverse crack defect of continuous casting billet angle based on metallurgical principle | |
| CN102937784A (en) | Artificial neural network based method for controlling online prediction of casting billet quality | |
| Cuartas et al. | Machine learning algorithms for the prediction of non-metallic inclusions in steel wires for tire reinforcement | |
| CN113343576B (en) | Prediction method of calcium yield in calcium treatment process based on deep neural network | |
| CN109396375A (en) | A kind of crystallizer bleedout prediction electric thermo method based on feature vector and hierarchical clustering | |
| CN105363794B (en) | A kind of finish rolling energy-saving control method based on mechanical properties forecast and required power model | |
| EP3916651A1 (en) | Design assistance method for metal material, prediction model generation method, metal material manufacturing method, and design assistance device | |
| Jiang et al. | BP neural network of continuous casting technological parameters and secondary dendrite arm spacing of spring steel | |
| CN101306438A (en) | Accelerated stream cooling device cooling technique integration method | |
| CN112036081A (en) | Method for determining addition amount of silicon-manganese alloy in converter tapping based on yield prediction | |
| Gong et al. | Optimization of steel casting feeding system based on BP neural network and genetic algorithm | |
| CN107716560B (en) | A kind of Hot Strip Rolling load distribution method | |
| CN116513273A (en) | Train operation scheduling optimization method based on deep reinforcement learning | |
| Lipiński et al. | The effect of the impurities spaces on the quality of structural steel working at variable loads | |
| CN113849020A (en) | Billet heating curve design method and device based on artificial intelligence algorithm | |
| CN115608793A (en) | Precision rolling temperature regulation and control method for mechano-fusion data | |
| CN112926622B (en) | Crystallizer breakout prediction method for generating countermeasure network based on feature vector and SWGAN-GP | |
| CN113063916A (en) | Hot-dip galvanized strip steel coating aluminum content prediction method based on PSO-SVR model | |
| CN115401178B (en) | Reduction process determination method for improving internal quality of gear steel | |
| CN111861041B (en) | Method for predicting dynamic recrystallization type rheological stress of Nb microalloyed steel |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |