CN113515042B - Multi-mode dry quenching Jiao Shaosun rate real-time computing system and computing method - Google Patents
Multi-mode dry quenching Jiao Shaosun rate real-time computing system and computing method Download PDFInfo
- Publication number
- CN113515042B CN113515042B CN202110681546.8A CN202110681546A CN113515042B CN 113515042 B CN113515042 B CN 113515042B CN 202110681546 A CN202110681546 A CN 202110681546A CN 113515042 B CN113515042 B CN 113515042B
- Authority
- CN
- China
- Prior art keywords
- dry quenching
- calculation
- coke
- layer
- output
- 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
- 238000010791 quenching Methods 0.000 title claims abstract description 127
- 230000000171 quenching effect Effects 0.000 title claims abstract description 127
- 238000004364 calculation method Methods 0.000 title claims abstract description 107
- 238000000034 method Methods 0.000 claims abstract description 103
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims abstract description 47
- 229910052799 carbon Inorganic materials 0.000 claims abstract description 40
- 238000004519 manufacturing process Methods 0.000 claims abstract description 30
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 18
- 239000000571 coke Substances 0.000 claims description 101
- 210000002569 neuron Anatomy 0.000 claims description 61
- 230000008569 process Effects 0.000 claims description 44
- 238000012549 training Methods 0.000 claims description 37
- 230000006870 function Effects 0.000 claims description 24
- 238000013528 artificial neural network Methods 0.000 claims description 23
- 238000012937 correction Methods 0.000 claims description 16
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 15
- 238000002485 combustion reaction Methods 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 230000009897 systematic effect Effects 0.000 claims description 8
- 239000000463 material Substances 0.000 claims description 7
- 239000013598 vector Substances 0.000 claims description 6
- 238000007599 discharging Methods 0.000 claims description 5
- 238000001704 evaporation Methods 0.000 claims description 3
- 230000008020 evaporation Effects 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 210000005036 nerve Anatomy 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000000946 synaptic effect Effects 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000012886 linear function Methods 0.000 claims description 2
- 230000008859 change Effects 0.000 abstract description 4
- 238000009434 installation Methods 0.000 abstract description 3
- 238000012423 maintenance Methods 0.000 abstract description 3
- 238000012986 modification Methods 0.000 abstract 1
- 230000004048 modification Effects 0.000 abstract 1
- 101100190621 Arabidopsis thaliana PLC4 gene Proteins 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 239000000446 fuel Substances 0.000 description 3
- 230000001603 reducing effect Effects 0.000 description 3
- 239000007787 solid Substances 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 238000004939 coking Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000035484 reaction time Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 239000006104 solid solution Substances 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- C—CHEMISTRY; METALLURGY
- C10—PETROLEUM, GAS OR COKE INDUSTRIES; TECHNICAL GASES CONTAINING CARBON MONOXIDE; FUELS; LUBRICANTS; PEAT
- C10B—DESTRUCTIVE DISTILLATION OF CARBONACEOUS MATERIALS FOR PRODUCTION OF GAS, COKE, TAR, OR SIMILAR MATERIALS
- C10B39/00—Cooling or quenching coke
- C10B39/02—Dry cooling outside the oven
Abstract
The invention provides a multi-mode dry quenching Jiao Shaosun rate real-time computing system and a computing method, wherein the computing system comprises a physical layer and a software layer, and the two parts are connected into a dry quenching automatic control system through an industrial Ethernet to obtain related production data in a read-only mode, so that the original dry quenching automatic control system is not rewritten; the calculation method comprises three calculation modes of artificial intelligence, a carbon method and a thermal method, and the server can calculate the three modes simultaneously, wherein the calculation results of the three modes are mutually verified; the PLC can only perform burning loss rate calculation in the carbon method mode as a hard backup of the system, and the burning loss rate calculation in the carbon method mode performed in the PLC is used as a backup when a large system breaks down and the system cannot perform artificial intelligence and thermal method calculation. The method calculates the rate of the dry quenching Jiao Shaosun in real time, does not change the original software and hardware system of the dry quenching, does not need to stop the furnace for modification, and has short installation time and convenient maintenance and upgrading.
Description
Technical Field
The invention relates to the technical field of metallurgical dry quenching Jiao Shaosun rate calculation, in particular to a multi-mode dry quenching Jiao Shaosun rate real-time calculation system and a calculation method.
Background
Coke is an important raw fuel for industrial production, especially as one of the main fuels for long-process ironmaking, and the yield and consumption of coke in China are in the first place in the world for a long time. The dry quenching is an important energy-saving and emission-reducing technology in Jiao Lugong sequence, and in recent years, under the advocacy of energy conservation and environmental protection, the dry quenching process has been unprecedented, and the dry quenching process has been used as the standard configuration of a newly-built coke oven in China.
In the process of the dry quenching process, red coke is filled into a dry quenching furnace, low-temperature circulating gas is conveyed into the dry quenching furnace by a circulating fan, the low-temperature circulating gas absorbs red Jiao Xianre in the dry quenching furnace to cool red Jiao Jiangwen, high-temperature circulating gas after absorbing sensible heat of the red coke is discharged from a flue of the dry quenching furnace to a boiler of the dry quenching process for heat exchange, steam generated by the boiler can be provided for a power plant, a steel plant or other users for utilization, thereby recycling the heat of the red coke, the cooled low-temperature circulating gas is conveyed into the dry quenching furnace by the circulating fan again, and the cooled coke is discharged from the dry quenching furnace to be raw fuel. In the production operation process of the dry quenching device, the circulating gas can carry out solid solution reaction with red hot coke, a part of coke is gasified into the circulating gas, and solid coke is lost. In order to ensure the safety of the system, a certain amount of air is usually introduced into the dry quenching, combustible materials in the circulating gas are burnt out to avoid explosion hazard, and part of solid coke powder is burnt out in the process, so that solid coke is lost. In the dry quenching process, the coke burning rate is directly related to the yield of coke, and the lower the coke burning rate is, the more coke is discharged from the dry quenching furnace, so the coke burning rate is an important index for guiding the operation and production of the dry quenching equipment.
The existing coke burning loss rate calculation method is generally as follows:
1) Mass conservation calculation method. The mass conservation is the simplest statistical calculation method, the carbon loss rate is roughly calculated through the record of the coke loading amount, the coke discharging amount and the coke powder amount in the shift operation, meanwhile, the mutual verification is carried out on the carbon loss rate calculated by the carbon balance, the change in the mass of the coke is analyzed, and the calculation mode is that the coke burning loss rate= (coke loading amount-coke discharging amount-coke powder amount)/the coke loading amount is multiplied by 100 percent.
2) A calculation method for coal consumption per ton of coke. When the coke oven adopts wet quenching, the burning loss rate is zero, and when the coke oven adopts dry quenching, the coke yield is reduced and the corresponding ton coke consumption is increased due to the burning loss, and the burning loss rate is calculated by analysis: coke burn rate= (dry quenching ton coke consumption-wet quenching ton coke consumption)/dry quenching ton coke consumption x 100%.
3) An intake air amount calculation method. Assuming that the dry quenching circulation system is sealed, the normally produced and introduced air reacts with carbon elements in the furnace to be in a balanced state, and the burning loss rate of the dry quenching can be calculated according to the introduced air quantity. The calculation mode is as follows: coke burn rate = mass of carbon element reacted per hour +.hourly coke charge.
4) Ash content calculation of coke breeze. Ash produced after coke is burnt in the furnace enters the coke powder, so that the ash content of the coke powder is greatly increased, and the burning loss rate can be calculated through the ash content of the coke powder. The ash balance calculation method comprises the following steps: coke burn-out rate=coke powder content× (coke powder ash content-coke ash content)/(coke content×coke ash content) ×100%.
When domestic coking enterprises adopt the methods to calculate the dry quenching Jiao Shaosun rate, real-time online monitoring is not realized, and operators cannot timely and effectively command production to reduce the burning loss of coke according to results, so that the method has little regulation and control guiding effect on the running production operation of the dry quenching equipment, and the burning loss rate of domestic dry quenching cannot be effectively reduced.
Disclosure of Invention
In order to solve the technical problems of the background technology, the invention provides a multi-mode dry quenching Jiao Shaosun rate real-time computing system and a computing method, which are used for computing the dry quenching Jiao Shaosun rate in real time, and meanwhile, the original software and hardware system of the dry quenching is not changed, the dry quenching does not need to be stopped for furnace transformation, the installation time of the system is short, and the maintenance and the upgrading are convenient.
In order to achieve the above purpose, the invention is realized by adopting the following technical scheme:
a multi-mode dry quenching Jiao Shaosun rate real-time computing system comprises two parts, namely hardware and software, wherein the two parts are divided into a physical layer and a software layer from a real object, and the two parts are connected into a dry quenching automatic control system through an industrial Ethernet to obtain related production data in a read-only mode without any rewriting of an original dry quenching automatic control system;
the physical layer comprises a PLC, a server, a router and a client computer;
the software layer comprises a PLC program, a data acquisition program, a data calculation program and a data management program;
the method comprises the steps that a data acquisition program, a data calculation program and a data management program are installed in a server, the PLC program is installed in the PLC, when a multi-mode dry quenching Jiao Shaosun rate real-time calculation system containing artificial intelligence operates, the multi-mode dry quenching Jiao Shaosun rate real-time calculation system needs to be connected to an Ethernet controlled in dry quenching and communicated with the PLC controlled in dry quenching, after production parameters are acquired from the Ethernet controlled in dry quenching, the data are respectively transmitted to the server and the PLC through a router, the server calculates the dry quenching Jiao Shaosun rate by using the data calculation program, the PLC calculates the burning loss rate by using the PLC program through a carbon method, and calculated results are transmitted to a client terminal through an industrial Ethernet through the router to be displayed or output.
The calculation method of the multi-mode dry quenching Jiao Shaosun rate real-time calculation system comprises three calculation modes of artificial intelligence, a carbon method and a thermal method, and the dry quenching Jiao Shaosun rate is calculated in real time; the server can simultaneously calculate three modes, and the calculation results of the three modes are checked; the PLC can only perform burning loss rate calculation in the carbon method mode as a hard backup of the system, and the burning loss rate calculation in the carbon method mode performed in the PLC is used as a backup when a large system breaks down and the system cannot perform artificial intelligence and thermal method calculation.
Further, the calculation method of the artificial intelligence calculation mode adopts a dry quenching Jiao Shaosun rate real-time calculation method based on a BP neural network;
the method comprises the following steps:
step one, determining parameters of an input layer and an output layer
The parameters input by the input layer are basic operation data of the dry quenching system, and the parameters comprise: coke temperature, dry quenching furnace material level, air guide valve opening, prestoring section negative pressure, circulating fan rotating speed, circulating air CO content and circulating air CO 2 Content, circulating air H 2 Content, boiler inlet air temperature, boiler water feed, dry quenching grate Jiao Wendu;
the output parameters of the output layer are only one, namely the coke burning loss rate;
step two, constructing a neural network
The method comprises the steps of constructing a dry quenching furnace coke real-time calculation neural network system according to the foregoing, wherein the input layer has 11 parameters, including: coke temperature, dry quenching furnace material level, air guide valve opening, prestoring section negative pressure, circulating fan rotating speed, circulating air CO content and circulating air CO 2 Content, circulating air H 2 Content, boiler inlet air temperature, boiler water feed, dry quenching grate Jiao Wendu; corresponding to 11 neurons, the middle layer has n neurons, and the output layer has one parameter: coke burn rate, corresponding to 1 neuron;
step three, determining training samples
Selecting one month as a training sample data acquisition period, calibrating corresponding meters before sample data acquisition, and then carrying out data acquisition on 11 input layer parameters and 1 output layer parameter;
| numbering device | (Code) | Content | Data source | |
| Input layer parameters | ||||
| 1 | a 1 | Coke oven number | Production record | |
| 2 | a 2 | Dry quenching furnace charge level | Central control room PLC record | |
| 3 | a 3 | Opening degree of air-guide valve | Central control room PLC record | |
| 4 | a 4 | Prestoring section negative pressure | Central control room PLC record | |
| 5 | a 5 | Rotation speed of circulating fan | Central control room PLC record | |
| 6 | a 6 | CO content of circulating air | Central control room PLC record | |
| 7 | a 7 | Circulating wind CO 2 Content of | Central control room PLC record | |
| 8 | a 8 | Circulating wind H 2 Content of | Central control room PLC record | |
| 9 | a 9 | Boiler inlet air temperature | Central control room PLC record | |
| 10 | a 10 | Boiler water supply | Central control room PLC record | |
| 11 | a 11 | Dry quenching grate Jiao Wendu | Central control room PLC record | |
| Output layer parameters | ||||
| 1 | y | Burn out rate of coke | Track scale and belt scale production record |
Input layer samples are represented by vectors A, A k ＝(a 1 ，a 2 ，……a 11 ) The method comprises the steps of carrying out a first treatment on the surface of the The output layer samples are represented by vectors Y, Y k = (y); where k=1, 2, … … 30, is the pair of samples takenNumber of pieces;
and fourthly, learning and training the BP neural network.
In the fourth step, the learning training of the BP neural network comprises the following processes:
1) A mode forward propagation process; the process is that the input values of the input and the output of the known mode, namely the known network, are provided for the neurons of the input layer of the network, and the network carries out propagation calculation to the output layer according to the mathematical model of the nerve; the 11 neurons of the input layer are directly output to the neurons of the middle layer without calculation, the neurons of the input layer are represented by i, i=1, 2, and the neurons of the … … middle layer are represented by j, and the input value isj=1, 2, … … 11, where w ij Is the synaptic connection coefficient or weighting value of neurons i through j; t (T) j Is the threshold for neuron j; the output transfer function of the intermediate layer neuron is one of a linear function, a ramp function, a threshold function, a unipolar S function, a bipolar S function, a tanH function, a ReLU function, or a Swish function, such that the output of the intermediate layer neuron is->
The output of the middle layer neuron is used as the input signal of the output layer neuron, the output conversion function of the output layer neuron adopts a unipolar S function, so that the input signal of the output layer neuron is sigma=v j o j -gamma, output signals of the output layer neuronsWherein v j Is the connection weight from the middle layer j neuron to the output layer neuron, and gamma is the threshold value of the output layer neuron;
the one-pass forward propagation process is completed for one input mode;
2) An error back propagation process; the back propagation of errors is the calculation of the network forward direction of the resulting output value, i.e. the error between the network response value and the desired outputMultiplying a correction factor, and then propagating according to a reverse network to obtain correction errors of each neuron; the correction error of the output layer is delta k ＝(Y k -Φ k )f′(σ k )＝(Y k -Φ k )Φ k (1-Φ k ) Correction error of each neuron in the middle layer isk＝1，2，……30；
3) Training; the training process is a process of repeatedly learning by a network, and continuously adjusting the connection weight and the threshold value according to the correction error obtained in the previous process; repeating 1) and 2) continuously, wherein the output value of the network gradually approaches the expected output value, namely a training sample; during training, the input connection right of the output layer becomesThe threshold value input by the output layer becomes gamma=gamma-beta delta k The input connection right of the middle layer becomes +.>The input threshold of the middle layer becomes +.>Alpha and beta are learning coefficients, the values of which are larger than 0 and smaller than 1, and the value range of the learning coefficients is between 0.25 and 0.75 under the general condition; the learning coefficient can be obtained with constant or can be modified continuously in the whole training process;
4) A convergence process of the network; for the learning training process of 30 pairs of modes, when all the mode pairs are converged, the network calculates convergence; the sum of squares of the errors is chosen as the systematic error, which is expressed asWhen the systematic error E is less than 0.001, the network is considered to converge, and training is ended, otherwise training is continued until the systematic error is less than 0.001.
Further, the calculation method of the thermal calculation mode comprises the following steps:
the dry quenching boiler generates steam by absorbing heat of combustion of red Jiao Xianre and coke through circulating gas, and in normal production, the loading amount of red coke is fixed, the coke discharging temperature is fixed, the red Jiao Xianre is fixed, and the evaporation amount of the boiler and the burning loss rate form a linear relation; the method for calculating the burning loss rate by using the steam production rate comprises the following steps:
wherein G is m To discharge 1t of coke burned in the cold coke; and:
wherein phi is Actual practice is that of Is the actual steam yield, phi No burning loss Is the steam yield without coke burning loss, q Steam generation Is the vapor enthalpy value, q Water supply Is the enthalpy value of water supply, q Combustion process Is the combustion heat value of coke, q Diffusing Is to release the heat of the gas.
Further, the calculation method of the carbon method calculation mode comprises the following steps:
assuming that the dry quenching circulation system is sealed, normally producing the reaction of the introduced air and carbon element in the furnace, and in a balanced state, calculating the burning loss rate of the dry quenching according to the introduced air quantity;
the calculation mode is as follows: coke burn rate = mass of carbon element reacted per hour +.o. coke charge per hour;
here, the mass of carbon element reacted per hour=the mass of air introduced×the carbon element content of the recycle gas;
coke charge per hour = coke discharge amount of dry quenching x correction factor;
thus, the real-time calculation of the burning loss rate can be performed by utilizing the existing production parameters of the dry quenching.
Compared with the prior art, the invention has the beneficial effects that:
1) The system is used for calculating the dry quenching Jiao Shaosun rate in real time, does not change the original software and hardware system of the dry quenching, does not need to stop the furnace for transformation, and has short installation time and convenient maintenance and upgrading;
2) The method comprises three calculation modes of artificial intelligence, a carbon method and a thermal method, and calculates the rate of the dry quenching Jiao Shaosun in real time; the server can simultaneously perform three modes of calculation, the calculation results of the three modes are mutually verified, and the calculation results are more accurate; the method has the characteristics of instantaneity, multi-mode complementarity, high accuracy, high reliability and the like, and can provide effective data basis for reducing the rate of dry quenching Jiao Shaosun; the PLC is used as a hard backup of the system, the burning loss rate calculation of the carbon method mode can only be carried out, the burning loss rate calculation of the carbon method mode carried out in the PLC is used as a backup when a large system breaks down and the system cannot carry out artificial intelligence and thermal method calculation, and the real-time calculation is ensured;
3) The artificial intelligence calculation mode adopts a calculation scheme based on BP neural network for calculating the coke burning loss rate of the dry quenching system which belongs to a nonlinear fuzzy system, and has the advantages of rapid calculation, accurate result, less influence factors and strong pertinence. The characteristics of the neural network determine: the accuracy of the calculation result of the data of which dry quenching furnace is used for training is high, so that each set of dry quenching system has a unique neural network system corresponding to the data, and the method is suitable for the actual state of the dry quenching system.
Drawings
FIG. 1 is a schematic diagram of a multi-mode dry quenching Jiao Shaosun rate real-time computing system architecture of the present invention;
FIG. 2 is a schematic diagram of a multi-mode dry quenching Jiao Shaosun rate real-time computing system data flow in accordance with the present invention;
fig. 3 is a structural diagram of the BP neural network of the present invention.
In the figure, (1) physical layer, (2) server, (3) router, (4) PLC, (5) client terminal, (6) software layer, (7) data acquisition program, (8) data calculation program, (9) data management program, (10) PLC program, (11) equipment layer, (12) data layer, (13) management layer, and (14) display layer
Detailed Description
The following detailed description of the embodiments of the invention is provided with reference to the accompanying drawings.
A multi-mode dry quenching Jiao Shaosun rate real-time computing system comprises two parts, namely hardware and software, which are connected into a dry quenching automation control system through an industrial Ethernet to obtain relevant production data in a read-only mode, and the original dry quenching automatic control system is not rewritten.
As shown in fig. 1, a multi-mode dry quenching Jiao Shaosun rate real-time computing system includes a physical layer 1, a software layer 6. The physical layer 1 comprises a server 2, a router 3, a PLC4, a client terminal 5 and other devices, wherein the server 2, the router 3 and the PLC4 are arranged in the same cabinet, and the client terminal 5 is arranged at a designated position and is connected with a system. The software layer 6 comprises a data acquisition program 7, a data calculation program 8, a data management program 9 and a PLC program 10. The data acquisition program 7, the data calculation program 8, and the data management program 9 are installed in the server 2, and the PLC program 10 is installed in the PLC4. When the multi-mode dry quenching Jiao Shaosun rate real-time computing system containing artificial intelligence operates, the system needs to be connected to an Ethernet of the dry quenching central control, is communicated with a PLC of the dry quenching central control, and transmits data to a server 2 and the PLC4 through a router 3 after acquiring production parameters. The server 2 uses the data calculation program 8 to calculate the dry quenching Jiao Shaosun rate simultaneously in three modes, the PLC4 uses the PLC program 10 to calculate the burning loss rate by adopting a carbon method, and the calculated result is transmitted to the client terminal 5 for display or output through the router 3 by the industrial Ethernet.
As shown in fig. 2, the multi-mode dry quenching Jiao Shaosun rate real-time computing system, the data flow architecture comprises a device layer 11, a data layer 12, a management layer 13 and a display layer 14. The equipment layer 11 comprises a server 2, a router 3, a PLC4, a client terminal 5 and other equipment, the data layer 12 comprises a data acquisition program 7 and a data management program 9, and after being transmitted in through the router 3 of the equipment layer 11, the dry quenching production data firstly enters the data layer 12 for storage and conversion. The control layer 13 includes a data calculation program 8, the converted production data transferred from the data layer 12 is calculated in real time by using the data calculation program 8 in three modes of artificial intelligence, carbon method and thermal method, and the calculation result is stored in a data management program 9 of the data layer 12, and the data management program 9 classifies, stores and indexes the production data and the calculation result and outputs the data through the data layer 12. The display layer 14 includes the client terminal 5, and can display, inquire, and output the production data and calculation results of the data management program 9.
The calculation method of the multi-mode dry quenching Jiao Shaosun rate real-time calculation system comprises three calculation modes of artificial intelligence, a carbon method and a thermal method, and the dry quenching Jiao Shaosun rate is calculated in real time; the server can simultaneously calculate three modes, and the calculation results of the three modes are checked; the PLC can only perform burning loss rate calculation in the carbon method mode as a hard backup of the system, and the burning loss rate calculation in the carbon method mode performed in the PLC is used as a backup when a large system breaks down and the system cannot perform artificial intelligence and thermal method calculation.
The calculation method of the artificial intelligence calculation mode adopts a dry quenching Jiao Shaosun rate real-time calculation method based on a BP neural network; the neural network is utilized to calculate the dry quenching Jiao Shaosun rate in real time, so that the method has the advantages of rapid calculation, accurate result, interference resistance, strong pertinence and the like, and can play a great role in optimizing the dry quenching operation and predicting the burning loss rate reducing effect.
The method specifically comprises the following steps:
step one, determining parameters of an input layer and an output layer
The parameters input by the input layer are basic operation data of the dry quenching system, and the parameters comprise: coke temperature, dry quenching furnace material level, air guide valve opening, prestoring section negative pressure, circulating fan rotating speed, circulating air CO content and circulating air CO 2 Content, circulating air H 2 Content, boiler inlet air temperature, boiler water feed, dry quenching grate Jiao Wendu;
the output parameters of the output layer are only one, namely the coke burning loss rate;
step two, constructing a neural network
The method comprises the steps of constructing a dry quenching furnace coke real-time calculation neural network system according to the foregoing, wherein the input layer has 11 parameters, including: coke temperature, dry quenching furnace material level, air guide valve opening, prestoring section negative pressure, circulating fan rotating speed, circulating air CO content and circulating air CO 2 Content, circulating air H 2 Content, boiler inlet air temperature, boiler water feed, dry quenching grate Jiao Wendu; corresponding to 11 neurons, there are 11 neurons in the middle layer, and one parameter of the output layer: coke burn rate, corresponding to 1 neuron; the topology is as in fig. 3.
Step three, determining training samples
Selecting one month as a training sample data acquisition period, calibrating corresponding meters before sample data acquisition, and then carrying out data acquisition on 11 input layer parameters and 1 output layer parameter;
| numbering device | (Code) | Content | Data source | |
| Input layer parameters | ||||
| 1 | a 1 | Coke oven number | Production record | |
| 2 | a 2 | Dry quenching furnace charge level | Central control room PLC record | |
| 3 | a 3 | Opening degree of air-guide valve | Central control room PLC record | |
| 4 | a 4 | Prestoring section negative pressure | Central control room PLC record | |
| 5 | a 5 | Rotation speed of circulating fan | Central control room PLC record | |
| 6 | a 6 | CO content of circulating air | Central control room PLC record | |
| 7 | a 7 | Circulating wind CO 2 Content of | Central control room PLC record | |
| 8 | a 8 | Circulating wind H 2 Content of | Central control room PLC record | |
| 9 | a 9 | Boiler inlet air temperature | Central control room PLC record | |
| 10 | a 10 | Boiler water supply | Central control room PLC record | |
| 11 | a 11 | Dry quenching grate Jiao Wendu | Central control room PLC record | |
| Output layer parameters | ||||
| 1 | y | Burn out rate of coke | Track scale and belt scale production record |
Input layer samples are represented by vectors A, A k ＝(a 1 ，a 2 ，......a 11 ) The method comprises the steps of carrying out a first treatment on the surface of the The output layer samples are represented by vectors Y, Y k = (y); where k=1, 2, &..30, the number of pairs of samples collected;
and fourthly, learning and training the BP neural network.
In the fourth step, the learning training of the BP neural network comprises the following processes:
1) A mode forward propagation process; the process is that the input values of the input and the output of the known mode, namely the known network, are provided for the neurons of the input layer of the network, and the network carries out propagation calculation to the output layer according to the mathematical model of the nerve; 11 neurons of the input layer directly output to neurons of the middle layer without calculation, and the input value of j neurons of the middle layer isj=1, 2, &..11, wherein w is ij Is the synaptic connection coefficient or weighting value of neurons i through j; t (T) j Is the threshold for neuron j; the output transfer function of the intermediate layer neuron is a unipolar S function, so that the output of the intermediate layer neuron is +.>
The output of the middle layer neuron is used as the input signal of the output layer neuron, the output conversion function of the output layer neuron adopts a unipolar S function, so that the input signal of the output layer neuron is sigma=v j o j -gamma, output signals of the output layer neuronsWherein v j Is the connection weight from the middle layer j neuron to the output layer neuron, and gamma is the threshold value of the output layer neuron;
the one-pass forward propagation process is completed for one input mode;
2) An error back propagation process; the back propagation process of the error is to multiply the output value obtained by forward calculation of the network, namely the error between the network response value and the expected output value, and then propagate according to the back network to obtain the correction error of each neuron; the correction error of the output layer is delta k ＝(Y k -Φ k )f′(σ k )＝(Y k -Φ k )Φ k (1-Φ k ) Correction error of each neuron in the middle layer isk＝1，2，……30；
3) Training; the training process is a process of repeatedly learning by a network, and continuously adjusting the connection weight and the threshold value according to the correction error obtained in the previous process; repeating 1) and 2) continuously, wherein the output value of the network gradually approaches the expected output value, namely a training sample; during training, the input connection right of the output layer becomesThe threshold value input by the output layer becomes gamma=gamma-beta delta k The input connection right of the middle layer becomes +.>The input threshold of the middle layer becomes +.>Alpha and beta are learning coefficients, the values of which are larger than 0 and smaller than 1, and the value range of the learning coefficients is between 0.25 and 0.75 under the general condition; the learning coefficient can be obtained with constant or can be modified continuously in the whole training process;
4) A convergence process of the network; for the learning training process of 30 pairs of modes, when all the mode pairs are converged, the network calculates convergence; the sum of squares of the errors is chosen as the systematic error, which is expressed asWhen the systematic error E is less than 0.001, the network is considered to converge, and training is ended, otherwise training is continued until the systematic error is less than 0.001.
The method collects enough accurate original data, constructs a proper neural network computing system, and trains the neural network computing system by using the original data, so that the computing error of the neural network computing system on the burning loss rate of the coke in the coke dry quenching furnace is less than or equal to 1%.
2. The calculation method of the thermal method calculation mode comprises the following steps:
the dry quenching boiler generates steam by absorbing heat of combustion of red Jiao Xianre and coke through circulating gas, and in normal production, the loading amount of red coke is fixed, the coke discharging temperature is fixed, the red Jiao Xianre is fixed, and the evaporation amount of the boiler and the burning loss rate form a linear relation; the method for calculating the burning loss rate by using the steam production rate comprises the following steps:
wherein G is m To discharge 1t of coke burned in the cold coke; and:
wherein phi is Actual practice is that of Is the actual steam yield, phi No burning loss Is the steam yield without coke burning loss, q Steam generation Is the vapor enthalpy value, q Water supply Is the enthalpy value of water supply, q Combustion process Is the combustion heat value of coke, q Diffusing Is to release the heat of the gas.
The calculation method not only depends on the measurement accuracy of the steam yield and the coke loading amount, but also has longer reaction time of a steam system, and the change of the calculation result along with time is not as fast as that of the carbon balance calculation mode, but can help to analyze the direction of reducing the burning loss rate from the view of the heat loss of the system.
3. The calculation method of the carbon method calculation mode comprises the following steps:
assuming that the dry quenching circulation system is sealed, normally producing the reaction of the introduced air and carbon element in the furnace, and in a balanced state, calculating the burning loss rate of the dry quenching according to the introduced air quantity;
the calculation mode is as follows: coke burn rate = mass of carbon element reacted per hour +.o. coke charge per hour;
here, the mass of carbon element reacted per hour=the mass of air introduced×the carbon element content of the recycle gas;
coke charge per hour = coke discharge amount of dry quenching x correction factor;
thus, the real-time calculation of the burning loss rate can be performed by utilizing the existing production parameters of the dry quenching.
This calculation method relies heavily on the accuracy of the measurement of gas parameters and coke charge, but reacts rapidly to systematic changes.
The above examples are implemented on the premise of the technical scheme of the present invention, and detailed implementation manners and specific operation processes are given, but the protection scope of the present invention is not limited to the above examples. The methods used in the above examples are conventional methods unless otherwise specified.
Claims (1)
1. A calculation method of a multi-mode dry quenching Jiao Shaosun rate real-time calculation system comprises two layers of a physical layer and a software layer, wherein the two layers are connected into a dry quenching automatic control system through an industrial Ethernet to obtain related production data in a read-only mode, and no rewrite is made to an original dry quenching automatic control system;
the physical layer comprises a PLC, a server, a router and a client computer;
the software layer comprises a PLC program, a data acquisition program, a data calculation program and a data management program;
the method comprises the steps that a data acquisition program, a data calculation program and a data management program are installed in a server, the PLC program is installed in the PLC, when a multi-mode dry quenching Jiao Shaosun rate real-time calculation system containing artificial intelligence operates, the multi-mode dry quenching Jiao Shaosun rate real-time calculation system needs to be connected to an Ethernet controlled in dry quenching and communicated with the PLC controlled in dry quenching, after production parameters are acquired from the Ethernet controlled in dry quenching, the data are respectively transmitted to the server and the PLC through a router, the server calculates the dry quenching Jiao Shaosun rate by using the data calculation program, the PLC calculates the burning loss rate by using the PLC program through a carbon method, and calculated results are transmitted to a client through an industrial Ethernet through the router to be displayed or output;
the method is characterized in that the calculation method of the calculation system comprises three calculation modes of artificial intelligence, a carbon method and a thermal method, and the dry quenching Jiao Shaosun rate is calculated in real time; the server calculates three modes simultaneously, and the calculation results of the three modes are checked; the PLC is used as a hard backup of the system, the burning loss rate calculation of the carbon method mode can only be carried out, and the burning loss rate calculation of the carbon method mode carried out in the PLC is used as a backup when a large system breaks down and the system cannot carry out artificial intelligence and thermal method calculation;
the calculation method of the artificial intelligence calculation mode adopts a dry quenching Jiao Shaosun rate real-time calculation method based on a BP neural network;
the method comprises the following steps:
step one, determining parameters of an input layer and an output layer
The parameters input by the input layer are basic operation data of the dry quenching system, and the parameters comprise: coke temperature, dry quenching furnace material level, air guide valve opening, prestoring section negative pressure, circulating fan rotating speed, circulating air CO content and circulating air CO 2 Content, circulating air H 2 Content, boiler inlet air temperature, boiler water feed, dry quenching grate Jiao Wendu;
the output parameters of the output layer are only one, namely the coke burning loss rate;
step two, constructing a neural network
Constructing a dry quenching furnace coke real-time calculation neural network system, wherein the input layer has 11 parameters, including: coke temperature, dry quenching furnace material level, air guide valve opening, prestoring section negative pressure, circulating fan rotating speed, circulating air CO content and circulating air CO 2 Content, circulating air H 2 Content, boiler inlet air temperature, boiler water feed, dry quenching grate Jiao Wendu; corresponding to 11 neurons, the middle layer has n neurons, and the output layer has one parameter: coke burn rate, corresponding to 1 neuron;
step three, determining training samples
Selecting one month as a training sample data acquisition period, calibrating corresponding meters before sample data acquisition, and then carrying out data acquisition on 11 input layer parameters and 1 output layer parameter;
Input layer samples are represented by vectors A, A k ＝(a 1 ，a 2 ，......a 11 ) The method comprises the steps of carrying out a first treatment on the surface of the The output layer samples are represented by vectors Y, Y k = (y); where k=1, 2, &..30, the number of pairs of samples collected;
step four, learning and training of the BP neural network;
in the fourth step, the learning training of the BP neural network comprises the following processes:
1) A mode forward propagation process; the process is that the input values of the input and the output of the known mode, namely the known network, are provided for the neurons of the input layer of the network, and the network carries out propagation calculation to the output layer according to the mathematical model of the nerve; the 11 neurons of the input layer are directly output to the neurons of the middle layer without calculation, the neurons of the input layer are represented by i, i=1, 2,..Wherein w is ij Is the synaptic connection coefficient or weighting value of neurons i through j; t (T) j Is the threshold for neuron j; the output transfer function of the intermediate layer neuron is one of a linear function, a slope function, a threshold function, a unipolar S function, a bipolar S function, a tanH function, a ReLU function or a Swish function, and the output of the intermediate layer neuron is->
The output of the middle layer neuron is used as the input signal of the output layer neuron, the output conversion function of the output layer neuron adopts a unipolar S function, so that the input signal of the output layer neuron is sigma=v j o j -gamma, output signals of the output layer neuronsWherein v j Is the connection weight from the middle layer j neuron to the output layer neuron, and gamma is the threshold value of the output layer neuron;
the one-pass forward propagation process is completed for one input mode;
2) An error back propagation process; the back propagation process of the error is to multiply the output value obtained by forward calculation of the network, namely the error between the network response value and the expected output value, and then propagate according to the back network to obtain the correction error of each neuron; the correction error of the output layer is delta k ＝(Y k -Φ k )f′(σ k )＝(Y k -Φ k )Φ k (1-Φ k ) Correction error of each neuron in the middle layer is
3) Training; the training process is a process of repeatedly learning by a network, and continuously adjusting the connection weight and the threshold value according to the correction error obtained in the previous process; repeating 1) and 2) continuously, and gradually approaching the output value of the networkThe desired output value, i.e., training samples; during training, the input connection right of the output layer becomesThe threshold value input by the output layer becomes gamma=gamma-beta delta k The input connection right of the middle layer becomes +.>The input threshold of the middle layer becomes +.>Alpha and beta are learning coefficients, the values of which are larger than 0 and smaller than 1, and the value range of the learning coefficients is between 0.25 and 0.75 under the general condition; the learning coefficient is obtained with constant or is continuously modified in the whole training process;
4) A convergence process of the network; for the learning training process of 30 pairs of modes, when all the mode pairs are converged, the network calculates convergence; the sum of squares of the errors is chosen as the systematic error, which is expressed asWhen the system error E is smaller than the set value, the network is considered to be converged, the training is finished, otherwise, the training is continued until the system error is smaller than the set value;
the calculation method of the thermal method calculation mode comprises the following steps:
the dry quenching boiler generates steam by absorbing heat of combustion of red Jiao Xianre and coke through circulating gas, and in normal production, the loading amount of red coke is fixed, the coke discharging temperature is fixed, the red Jiao Xianre is fixed, and the evaporation amount of the boiler and the burning loss rate form a linear relation; the method for calculating the burning loss rate by using the steam production rate comprises the following steps:
wherein G is m To discharge 1t of coke burned in the cold coke; and:
wherein phi is Actual practice is that of Is the actual steam yield, phi No burning loss Is the steam yield without coke burning loss, q Steam generation Is the vapor enthalpy value, q Water supply Is the enthalpy value of water supply, q Combustion process Is the combustion heat value of coke, q Diffusing Is to release the heat of the gas;
the calculation method of the carbon method calculation mode comprises the following steps:
assuming that the dry quenching circulation system is sealed, normally producing the reaction of the introduced air and carbon elements in the furnace, and in a balanced state, calculating the burning rate of the dry quenching according to the introduced air;
the calculation mode is as follows: coke burn rate = mass of carbon element reacted per hour +.o. coke charge per hour;
here, the mass of carbon element reacted per hour=the mass of air introduced×the carbon element content of the recycle gas;
coke charge per hour = coke discharge amount of dry quenching x correction factor;
and (5) calculating the burning loss rate in real time by utilizing the existing production parameters of the dry quenching.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110681546.8A CN113515042B (en) | 2021-06-18 | 2021-06-18 | Multi-mode dry quenching Jiao Shaosun rate real-time computing system and computing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110681546.8A CN113515042B (en) | 2021-06-18 | 2021-06-18 | Multi-mode dry quenching Jiao Shaosun rate real-time computing system and computing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113515042A CN113515042A (en) | 2021-10-19 |
| CN113515042B true CN113515042B (en) | 2024-03-08 |
Family
ID=78065713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110681546.8A Active CN113515042B (en) | 2021-06-18 | 2021-06-18 | Multi-mode dry quenching Jiao Shaosun rate real-time computing system and computing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113515042B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103163041A (en) * | 2011-12-16 | 2013-06-19 | 景德镇开门子陶瓷化工集团有限公司 | Method and system for accurately obtaining dry quenching coke burn-out rate |
| RU2013111881A (en) * | 2013-03-15 | 2014-09-27 | Открытое акционерное общество "ЕВРАЗ Нижнетагильский металлургический комбинат" (ОАО "ЕВРАЗ НТМК") | METHOD OF DRY EXTINGUISHING OF COX |
| CN110003923A (en) * | 2019-05-16 | 2019-07-12 | 华泰永创(北京)科技股份有限公司 | It is a kind of for measuring the device and measurement method of coke burning in coke dry quenching furnace |
| CN112410045A (en) * | 2020-11-06 | 2021-02-26 | 铜陵泰富特种材料有限公司 | Method and device for recovering waste heat of coke powder generated by primary dust removal of dry quenching |
| CN113515043A (en) * | 2021-06-18 | 2021-10-19 | 上海源矩技术有限公司 | Method for calculating dry quenching loss rate in real time based on BP neural network |
-
2021
- 2021-06-18 CN CN202110681546.8A patent/CN113515042B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103163041A (en) * | 2011-12-16 | 2013-06-19 | 景德镇开门子陶瓷化工集团有限公司 | Method and system for accurately obtaining dry quenching coke burn-out rate |
| RU2013111881A (en) * | 2013-03-15 | 2014-09-27 | Открытое акционерное общество "ЕВРАЗ Нижнетагильский металлургический комбинат" (ОАО "ЕВРАЗ НТМК") | METHOD OF DRY EXTINGUISHING OF COX |
| CN110003923A (en) * | 2019-05-16 | 2019-07-12 | 华泰永创(北京)科技股份有限公司 | It is a kind of for measuring the device and measurement method of coke burning in coke dry quenching furnace |
| CN112410045A (en) * | 2020-11-06 | 2021-02-26 | 铜陵泰富特种材料有限公司 | Method and device for recovering waste heat of coke powder generated by primary dust removal of dry quenching |
| CN113515043A (en) * | 2021-06-18 | 2021-10-19 | 上海源矩技术有限公司 | Method for calculating dry quenching loss rate in real time based on BP neural network |
Non-Patent Citations (2)
| Title |
|---|
| Development of Thermal Schemes for Dry Coke Quenching Using Carbonates in Order to Save Energy;I.S. Gordeeva 等;2020 International Multi-Conference on Industrial Engineering and Modern Technologies;全文 * |
| 焦炉烟气余热回收烟风系统分析;徐正 等;冶金能源;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113515042A (en) | 2021-10-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yang et al. | Prediction of the NOx emissions from thermal power plant using long-short term memory neural network | |
| CN107016455B (en) | Prediction system and method for oxygen content of flue gas at hearth outlet of circulating fluidized bed household garbage incineration boiler | |
| CN103728879B (en) | A kind of flue gas in power station boiler flexible measurement method | |
| CN111829003B (en) | Power plant combustion control system and control method | |
| CN106224939B (en) | Circulating fluid bed domestic garbage burning boiler bed temperature Forecasting Methodology and system | |
| CN102778538A (en) | Soft measuring method based on improved SVM (Support Vector Machine) for measuring boiler unburned carbon content in fly ash | |
| CN111754093A (en) | Fly ash carbon content prediction method based on coal quality prediction and PSO-SVM | |
| CN110207094B (en) | IQGA-SVR boiler heating surface contamination characteristic identification method based on principal component analysis | |
| CN111881405B (en) | Real-time calculation method for content of combustible in fly ash of coal-fired boiler | |
| CN112016754A (en) | Power station boiler exhaust gas temperature advanced prediction system and method based on neural network | |
| CN113515043B (en) | BP neural network-based real-time calculation method for dry quenching Jiao Shaosun rate | |
| CN108021027B (en) | Output power prediction system and method for supercritical circulating fluidized bed unit | |
| CN113515042B (en) | Multi-mode dry quenching Jiao Shaosun rate real-time computing system and computing method | |
| CN109140474A (en) | A kind of coal-fired power station boiler manually intelligent sootblowing | |
| CN105808945A (en) | Mixed intelligent boiler combustion efficiency optimization method | |
| CN111931436A (en) | Burner nozzle air quantity prediction method based on numerical simulation and neural network | |
| CN110684547A (en) | Optimized control method for biomass pyrolysis carbonization kiln | |
| CN115656461A (en) | Coal electric unit real-time carbon emission monitoring method based on coal quality soft measurement | |
| CN101929684A (en) | Coal consumption calculating method of composite firing low heat value gas unit | |
| CN114675607A (en) | Fired coal heat value prediction and correction control method combining flame temperature | |
| Tian | Predictive control of coke oven flue temperature based on orthogonal neural network | |
| CN112036091A (en) | Method for predicting air quantity of nozzle of four-corner tangential burner by using regression algorithm | |
| Shang et al. | Bagging-ELM Model for Heating Furnace Thermal Efficiency Prediction | |
| Zheng et al. | Prediction of NOx concentration from coal combustion using LS-SVR | |
| CN117649102B (en) | Optimal scheduling method of multi-energy flow system in steel industry based on maximum entropy reinforcement learning |
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 |