CN114265885A - Automatic control method and system for moisture of sintering return powder - Google Patents
Automatic control method and system for moisture of sintering return powder Download PDFInfo
- Publication number
- CN114265885A CN114265885A CN202111620431.4A CN202111620431A CN114265885A CN 114265885 A CN114265885 A CN 114265885A CN 202111620431 A CN202111620431 A CN 202111620431A CN 114265885 A CN114265885 A CN 114265885A
- Authority
- CN
- China
- Prior art keywords
- fuzzy
- moisture content
- water
- moisture
- deviation
- 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.)
- Pending
Links
Images
Classifications
-
- 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
- Manufacture And Refinement Of Metals (AREA)
Abstract
The invention provides a fuzzy feedforward water content automatic control method and system based on prediction model assistance, wherein the method comprises the following steps: acquiring production data of sintering return powder and establishing a production database; taking the feeding frequency and the feeding port temperature in a production database as input variables, taking the clear water flow as an output variable, and establishing a fuzzy control word table between the input quantity and the output quantity; carrying out fuzzy control word table query according to the currently acquired feeding frequency and feeding port temperature data to obtain the current fuzzy water addition amount; constructing an optimal historical data set according to the acquired production data, establishing a water content prediction model based on a time series neural network by using the optimal historical data set, and inputting the currently acquired data into the prediction model to output a water content prediction value; dividing the deviation category of the predicted value of the moisture content and a preset moisture content value according to a box dividing method; and correcting the fuzzy water adding amount at this time on line according to the deviation type to obtain the target water adding amount at this time.
Description
Technical Field
The invention belongs to the technical field of automatic monitoring and intelligent control of moisture of ore sintering materials, and particularly provides a method and a system for automatically controlling moisture of sintering return powder.
Background
The returned powder cooling process of the lead-zinc ore sintering process in a smelting plant is a subsequent process of the sintering process, namely, sintering ingredients which do not meet the requirements of the sintering process are recycled. The process of recycling the returned powder materials needs to convey the returned powder materials to the powder returning bin through the processes of buffering in the buffering bin, crushing in the crusher, water adding and cooling in the cooling cylinder, belt transmission and the like so as to carry out re-batching. The invention is just applied to the process that the returned powder is cooled to be conveyed by a belt after being added with water in the cooling cylinder. Water is added at the feed end of the cooling cylinder, and the materials are stirred by the cooling cylinder in a rotating way so as to achieve the purpose of cooling. At present, smelting plants adopt two control modes of manual water adding, wherein the water adding mode mainly comprises adding bottom mud (mineral processing waste liquid) and secondarily adding clear water. The technical operation key points are that the cooling cylinder aims to cool and mix materials from the powder return bin, the quantity of fed materials is positively correlated with the frequency of the vibrating feeder, different frequencies approximately correspond to different opening degrees of the bottom sediment valve, field workers adjust the opening degrees of the bottom sediment valve according to the existing operation experience, then the moisture content of the materials on the conveying belt is visually observed, if the moisture content is higher, the adding amount of the bottom sediment is reduced, and if the moisture content is lower, the adding amount of the bottom sediment is increased. However, the bottom mud is unstable in property, and a bottom mud pipe is easy to block, and clear water is required to be added for assistance or directly adopted to replace the bottom mud. In addition, in order to avoid raising dust as far as possible to cause environmental pollution and threat the personal health of production workers, a temporary emergency water adding spraying device is arranged on the conveying belt, once the workers detect that the moisture content is too low, the dust will start to raise, the standby water device is immediately opened, and water is directly sprayed onto the conveying belt to cover the raising dust. By analyzing the field process and the current operation situation, the current water adding mode mainly has the following difficulties or defects:
1) the size of the initial incoming material is unstable, the powder return bin is closed, and the mechanism of the initial incoming material is difficult to master, so that the blanking amount cannot be accurately calculated even if the frequency of the vibrating feeder is fixed, and the blanking amount cannot be accurately controlled.
2) Because the cooling cylinder needs to cool the incoming material, the temperature fluctuation of the incoming material under different working conditions is large, no specific rule exists, the blanking speed is high, the cooling water adding amount is not only influenced by the incoming material amount, but also influenced by the incoming material temperature, and the influence factor is not considered in the current production mode.
3) The nature of bed mud is unstable, because the bed mud is the waste liquid that mineral processing formed, is the mixture of different mineral residues, and the proportion can't be confirmed moreover, can also cause the jam of bed mud flowmeter and valve sometimes, can't realize accurate control to it.
4) Workers visually observe the moisture content of the materials of the conveying belt according to experience to determine the amount of bottom mud and the addition amount and the adjustment amount of clean water, and manual adjustment by experience has the characteristics of slow reaction, long reaction period, poor adjustment effect and the like and is high in labor intensity.
5) The returned powder material needs to go through about 3 minutes from the inlet to the outlet of the cooling cylinder, and the effect can be seen only after bottom mud or clear water is added, which is a nonlinear, large-lag and slow time-varying process, and the difficulty is great when the moisture value of the material at the outlet of the cooling cylinder is controlled to be stable within a required range.
6) The manual valve lags behind the mode of controlling the water adding amount. When workers perceive that the moisture content of the conveyor belt materials does not meet requirements, the operators need to manually adjust the valves from the duty room to the field, the operation is stopped after the moisture content of the conveyor belt materials meets the requirements, and the backward water adding mode increases the workload of the operators.
7) Because the cooling cylinder is similar to an externally closed black box, the friction force is large in the material transmission process, and a detection device is difficult to install in the cooling cylinder, so that the physical and chemical change mechanisms in the cooling cylinder are difficult to master, and the operation parameters such as material inlet and outlet temperature, material real-time moisture value and the like can be detected only outside the cooling cylinder. The detection has the problems of untimely and inaccurate detection.
In summary, the water feeding system of the cooling cylinder is influenced by many factors, and it is difficult to achieve the ideal target of stable moisture content and accurate control of the conveyor belt material.
Therefore, a novel method and a novel system for controlling the moisture content of the sintering return powder are urgently needed in the industry.
Disclosure of Invention
The invention provides a method and a system for automatically controlling moisture of sintered powder return, which solve the problems of unstable moisture and inaccurate control of manual water addition, and realize the target control, automatic control and stable control of the material cooling and mixing process in a cooling cylinder, thereby achieving the purpose of keeping the moisture content of the material on a conveyor belt in a stable range after mixing, avoiding the phenomena of conveyor belt slippage caused by overhigh moisture content of the material and dust emission caused by overlow moisture content of the material, improving the sintering efficiency and reducing the labor intensity of workers.
An automatic moisture control method for sintered powder return comprises the following steps:
acquiring production data of the sintering return powder, and establishing a production database; the production database comprises the feeding frequency of returned powder, the temperature of a feeding hole, the flow of bottom mud, the flow of clear water, the flow of standby water, the temperature of a discharging hole and the real-time moisture content of materials on a conveyor belt;
adopting a multi-factor control method to take the feeding frequency and the feeding port temperature which have the greatest influence on the moisture content of the material in the production database as input variables, taking the clear water flow as an output variable, and establishing a fuzzy control word table between the input quantity and the output quantity;
carrying out fuzzy control word table query according to the currently acquired feeding frequency and feeding port temperature data, and outputting the current fuzzy water addition amount;
constructing an optimal historical data set according to the acquired production data of the sintering return powder, and establishing a water content prediction model by adopting the optimal historical data set; inputting the currently collected data into a prediction model, and outputting a moisture content prediction value;
dividing the deviation category of the predicted water content value and a preset water content value according to a box dividing method;
and correcting the current fuzzy water adding amount output by the fuzzy control word table on line according to the deviation type to obtain the current target water adding amount.
Further, a multi-factor control method is adopted to take the feeding frequency and the feeding port temperature which have the greatest influence on the moisture content of the material in the production database as input variables, take the flow of clean water as an output variable, and establish a fuzzy control word table between the input quantity and the output quantity, wherein the method specifically comprises the following steps:
assuming that the sampling obtained feeding frequency is f and the temperature of the feeding hole is t, calculating the clear water output u:
wherein the content of the first and second substances,
and "·" are fuzzy matrix operators, R denotes fuzzy relations, u1Control outputs obtained for the first fuzzy relation, NBuAs a fuzzy set NBf，NMfOr fuzzy set NBt，NMtAnd meanwhile, in a fuzzy mapping set when the temperature is established, if the membership function values of the feeding frequency f and the feeding port temperature t correspond to the quantized grades and take 1, the rest are all taken as 0, and the formula (2) is simplified as follows:
in the formula (I), the compound is shown in the specification,
are fuzzy sets NB respectivelyf，NMfThe degree of membership of the ith element (i.e., the measured feed frequency is the ith grade;
are fuzzy sets NB respectivelyt，NMtThe membership degree of the jth element, namely the measured temperature is the jth grade; similarly, the control quantity u is obtained2,u3,……unThen the control quantity is represented as a fuzzy set u as:
u＝u1+u2+u3+……+un (4)
and converting the controlled variable from the fuzzy variable to the accurate variable by adopting a maximum membership method to obtain a fuzzy control word table.
Further, constructing an optimal historical data set according to the acquired production data of the sintering return powder, and establishing a water content prediction model by adopting the optimal historical data set; inputting the currently collected data into a prediction model, and outputting a moisture content prediction value, specifically:
the preferred historical data set comprises: feeding frequency, bottom mud flow, clear water flow, feeding port temperature and discharging port material moisture content;
the water content prediction model is a water content prediction model based on a time series neural network; the currently acquired data comprise bottom mud flow, clear water flow, feeding frequency and feeding port temperature which are used as input variables of the prediction model, and the moisture content of the material at the discharging port is used as an output variable of the prediction model.
Further, the construction of the water content prediction model based on the time series neural network specifically comprises the following steps: and the preferred historical data set is processed according to the following steps of 6: 2: and 2, dividing the ratio into a training set, a verification set and a test set, fully utilizing the training set data of the optimized historical data set through proper off-line training, and training by adopting a levenberg marquardt algorithm to obtain the model.
Further, dividing the deviation category of the predicted moisture content value and a preset moisture content value according to a box dividing method, specifically:
mapping a historical data set M of deviation between a predicted value of the moisture content of a discharge port and a preset moisture content value in the material mixing process to a moisture content deviation classification set N in a mode of function mapping by a box separation method to obtain a deviation classification function y (f) (x) of the moisture content corresponding to the deviation x of the moisture content of the material mixing, wherein x belongs to M;
because the measurement range of the moisture meter is between 0% and 10%, if the difference between the predicted value of the moisture content and the set value is less than-0.8%, -0.8% -0.2% -1%, and greater than 1%, the deviation categories of the moisture content are 1, 2, 3, 4, and 5, respectively.
Further, the current fuzzy water adding amount output by the fuzzy control word table is corrected on line according to the deviation type to obtain the current target water adding amount, and the method comprises the following steps:
if the deviation type y is 1, the target water adding amount P is a x (1+ 15%);
if the deviation type y is 2, the target water adding amount P is a x (1+ 7%);
if the deviation type y is 3, the target water adding amount P of the time is a;
if the deviation type y is 4, the target water adding amount P is a x (1-7%);
if the deviation type y is 5, the target water adding amount P is a x (1-15%);
wherein, P is the target water adding amount of this time, and a is the fuzzy water adding amount of this time output by the fuzzy control word table.
The invention also provides an automatic moisture control system of the sintering returned powder, which comprises a processor and a memory connected with the processor, wherein the memory stores an automatic moisture control program of the sintering returned powder, and the automatic moisture control program of the sintering returned powder is executed by the processor to realize the steps of the automatic moisture control method of the sintering returned powder.
Further, the automatic moisture control system for the sintering return powder further comprises: the device comprises a cooling cylinder, a clear water storage device, a bottom sediment storage device, a vibrating feeder, a material conveyor belt and an infrared thermometer for measuring the temperature of a feeding port and the temperature of a discharging port of the cooling cylinder;
the clear water storage device, the bottom sediment storage device and the vibrating feeder are connected with the feeding end of the cooling cylinder;
the material conveying belt is connected with the discharge end of the cooling cylinder;
a clean water electric valve and a clean water flowmeter are sequentially arranged on a pipeline between the clean water storage device and the cooling cylinder;
a manual valve and a bottom mud flowmeter are sequentially arranged on a pipeline between the bottom mud storage device and the cooling cylinder;
and a moisture meter is also arranged on the material conveying belt.
Further, still be equipped with the standby water device on the material conveyer belt, install standby water sprinkler additional on the standby water device for avoid the appearance of raise dust phenomenon when moisture content is on the low side.
The invention has the following beneficial effects:
1. according to the automatic control method for the moisture of the sintering returned powder, provided by the invention, the production database is established according to the past operation experience, the problem that the control system has no initial data is solved, and the automatic control method has reference significance for domestic water adding systems of the same type of sintering process. In addition, the invention builds a water content prediction model by utilizing the neural network, and then carries out online correction on the fuzzy output quantity according to the predicted deviation between the water content and the set value, thereby limiting the control output in a certain range in advance, greatly improving the control accuracy and precision and greatly improving the control delay phenomenon caused by large time lag.
2. The invention provides an automatic moisture control method for sintered powder, which adopts a fuzzy feedforward-closed loop control principle and constructs and trains a moisture content prediction model based on a time series neural network according to a historical data set. When the fuzzy controller is designed, the feeding temperature of the cooling cylinder, the vibration frequency of the conveyor and the flow of clean water are used as input quantities, the system output is used for determining the feeding temperature of the cooling cylinder, the vibration frequency of the conveyor, the universe of flow of clean water and a word set of fuzzy language variables and membership functions, a multi-factor control method is adopted, two variables are used as input, and the error change rate of a single variable are not used as input.
3. The automatic moisture control system for the sintering return powder, provided by the invention, adopts a mode of outputting and correcting the output value of the fuzzy controller by the prediction model, greatly improves the problems of large lag, nonlinearity, slow time variation and strong coupling which cannot be independently solved by single fuzzy control, obtains the deviation from the set moisture value after introducing the moisture predicted value, can predict the water adding amount of this time in advance, adopts a method of classifying moisture deviation, enables the system to correct the water adding amount more accurately on the premise of not frequently acting, limits the effect of the prediction model by a mode of leading fuzzy control, and avoids the instability of the system caused by sudden change of the predicted value due to sudden change of industrial environment. Meanwhile, the problems that the moisture content of the conveyor belt materials is too low and dust is raised, the temperature of the materials at the discharge port is too high, the equipment is damaged, the moisture content of the conveyor belt materials is too high and the materials slip and the like caused by inaccurate judgment, untimely adjustment and unstable moisture due to manual experience operation in the prior art are solved, the labor intensity of workers is greatly reduced, the labor efficiency is improved, the operation environment of operators is improved, and the influence of the dust on the surrounding environment is reduced.
In addition to the objects, features and advantages described above, other objects, features and advantages of the present invention are also provided. The present invention will be described in further detail below with reference to the drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
FIG. 1 is a flow chart of a method for automatically controlling the moisture content of a sintered powder lot according to the present invention;
FIG. 2 is a schematic illustration of the corrected fuzzy output of the prediction model of the present invention;
FIG. 3 is a training model response of a time series neural network-based moisture content prediction model employed in the present invention;
FIG. 4 is a regression curve of a training model of a water content prediction model based on a time series neural network employed in the present invention;
FIG. 5 is an additional test response of a model of a time series neural network based moisture content prediction model employed in the present invention;
FIG. 6 is a diagram of a fuzzy feedforward water content automatic control system based on prediction model assistance.
Detailed Description
Embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways, which are defined and covered by the claims.
The sintering powder returning and water adding model comprises the following concrete steps: after the sintering and roasting process is finished, crushing the powder by a single shaft (first crushing) and a toothed roller (second crushing), screening, conveying the oversize material which is qualified sinter cake to a blast furnace for smelting, and preparing the undersize material into qualified return powder after passing through a corrugated roller crusher (third crushing) and a smooth roller crusher (fourth crushing). In the automatic control system for adding water into the sintering return powder, the return powder materials sent by the conveyor are cooled and mixed, and the return powder materials comprise lead ore, zinc ore, mixed ore, limestone, dust and ash collection, blue powder, zinc hypoxide, electric dust and the like. An NIR1 moisture meter is adopted to detect the moisture content of the mixed materials at the outlet of the cooling cylinder on line in real time, and a moisture content relation model is as follows:
in the above formula, m (t) is the moisture content of the material at the outlet of the cooling cylinder at time t, s (t) is the moisture content of the bottom mud, w (t) is the total mass of the bottom mud, d (t) is the moisture content of the sintered return powder raw material, f (t) is the mass of the sintered return powder raw material, and u (t) is the water addition amount.
Therefore, the moisture content of the material at the discharge port is influenced by various factors.
Based on the above, as shown in fig. 1 and fig. 2, the present invention provides a method for automatically controlling moisture of sintered powder return, comprising the steps of:
And 102, adopting a multi-factor control method to take the feeding frequency and the feeding port temperature which have the greatest influence on the moisture content of the material in the production database as input variables, taking the flow of clean water as an output variable, and establishing a fuzzy control word table between the input quantity and the output quantity.
103, carrying out fuzzy control word table query according to the currently acquired feeding frequency and feeding port temperature data, and outputting the current fuzzy water adding amount.
104, constructing an optimal historical data set according to the acquired production data of the sintering return powder, and establishing a water content prediction model by adopting the optimal historical data set; and inputting the currently acquired data into a prediction model, and outputting a predicted value of the moisture content. Currently collected data include feed inlet temperature, feed frequency, bed mud flow and clean water flow.
And 105, dividing the deviation category of the predicted water content value and a preset water content value according to a box dividing method.
And 106, correcting the fuzzy water adding quantity output by the fuzzy control word table at this time on line according to the deviation type to obtain the target water adding quantity at this time.
In step 101, the production database is specifically: the actual moisture measured by the material outlet moisture meter is compared with a set value, and the operating parameters with the deviation within five percent are stored to obtain an empirical operation database which can keep the moisture of the material at the discharge port stable and accurate under various conditions. The method comprises the steps of feeding port temperature, feeding frequency, bottom mud flow, clear water flow (water adding amount), adding amount of standby water, indication of a moisture meter and a current moisture set value. The feeding frequency is determined by the vibration frequency of the vibrating feeder. According to the past operation experience, a production database is established, the problem that the control system has no initial data is solved, and the method has reference significance for domestic sintering process water adding systems of the same type.
In step 102, unlike previous studies, in this study, the error and the error change rate between the actual output value and the set value are not directly used as fuzzy variables, but a multi-factor control method is used to take the feeding frequency f and the temperature t which have the greatest influence on the moisture content of the material in the production process as input variables, the clear water flow u as output variables, a triangular membership function is selected, and a domain quantization formula is adopted:
wherein m and n respectively represent two end points of the discretization discourse domain, and a and b respectively represent the minimum value and the maximum value of the actual physical quantity. According to the variation range (30 HZ-42 HZ) of the feeding frequency, thirteen grades of discourse domains are obtained after quantification: [ -6, -5, -4, -3, -2, -1,0,+1,+2,+3,+4,+5,+6]. The fuzzy set is { NB, NM, NS, O, PS, PM, PB }, and the fuzzy assignment of frequency change is obtained as shown in Table 1:
TABLE 1 fuzzy value-giving table for feeding frequency f
According to the variation range of the feeding temperature (60 ℃ -200 ℃), fourteen grades of domains are obtained after quantification: [ -6, -5, -4, -3, -2, -1,0-,0+,+1,+2,+3,+4,+5,+6]. The fuzzy set is { NB, NM, NS, NO, PO, PS, PM, PB }, and the fuzzy assignment of the feed temperature change is obtained as shown in Table 2:
TABLE 2 fuzzy assignment table for feeding temperature t
According to the variation range (0-9 m) of the output quantity of the clean water3H), obtaining fifteen levels of discourse domain after quantization: [ -7, -6, -5, -4, -3, -2, -1,0,+1,+2,+3,+4,+5,+6,+7]. The fuzzy set is { NB, NM, NS, O, PS, PM, PB } to obtain fuzzy assignments of changes in the clear water flow as shown in Table 3:
TABLE 3 clear water flow u fuzzy value-assigning table
The fuzzy rule is described in the form of "IF a THEN B" according to expert experience (to name only four examples):
(1)if t＝NB or NM and f＝NB or NM then u＝NB
(2)if t＝NB or NM and f＝NS or O then u＝NB
(3)if t＝NB or NM and f＝PS then u＝NM
(4)if t＝NB or NM and f＝PB or PM then u＝O
assuming that the sampling obtained feeding frequency is f and the temperature of the feeding hole is t, calculating the clear water output u:
wherein the content of the first and second substances,
and "·" are fuzzy matrix operators, R denotes fuzzy relations, u1Control outputs obtained for the first fuzzy relation, NBuAs a fuzzy set NBf，NMfOr fuzzy set NBt，NMtSimultaneous presentation when in effectAnd (3) shooting a fuzzy set, if the membership function values of the feeding frequency f and the feeding port temperature t correspond to the quantized grades and take 1, and the rest are all taken as 0, simplifying the formula (2) as follows:
in the formula (I), the compound is shown in the specification,
are fuzzy sets NB respectivelyf，NMfThe degree of membership of the ith element (i.e. the measured feed frequency is the ith order),
are fuzzy sets NB respectivelyt，NMtThe degree of membership of the jth element (i.e., the measured temperature is in the jth order), and the control quantity u is obtained by the same method2,u3,……unThen the control quantity is represented as a fuzzy set u as:
u＝u1+u2+u3+……+un (4)
and converting the controlled quantity from the fuzzy quantity to an accurate quantity by adopting a maximum membership method, namely selecting the element with the maximum membership in the fuzzy subset as the controlled quantity, and taking the average value of the controlled quantity if the elements with the maximum membership are several. From this, a fuzzy control word table is obtained as shown in table 4:
table 4 output control word table
In step 103, through the control word table generated in the previous step, when we perform real-time control again, we can directly query the control strategy that should be adopted in the current state according to the currently collected feeding frequency and feeding temperature data to obtain the corresponding control word u, and then multiply by the scaling factor k to obtain the current fuzzy water addition amount a.
In step 104, a moisture content prediction model is first constructed, specifically:
acquiring historical data of a lead-zinc ore sintering return powder water adding and mixing process in an industrial environment, selecting parameter information such as conveyor vibration frequency, bottom mud flow, clear water flow, cooling cylinder feed inlet temperature and discharge port material moisture content after data preprocessing to construct a preferred historical data set, and selecting 40000 groups of conveyor vibration frequency, bottom mud flow, clear water flow, cooling cylinder feed inlet temperature and discharge port material moisture content data in the historical data set according to 6: 2: and 2, dividing the data into a training set, a verification set and a test set in proportion, and performing model training by fully utilizing the data of the preferred historical data set through proper off-line training and adopting a levenberg marquardt algorithm to obtain a basic time series neural network model. Specifically, for the watering control process, the input to the offline time-series neural network is represented by the sediment flow weight vector x1(n), feed frequency weight vector x2(n), feed inlet temperature weight vector x3(n), the weight of the flow of the added clear water a (n) and the moisture content at the discharge outlet y (n) and their delayed composition can be expressed as:
Input(n)＝[x(n-dx),...x(n),a(n-dx),...a(n),y(n-dx),...y(n-dx)] (5)
the output (n) of the time series neural network model is defined as:
the Levenberg Marquardt (LM) algorithm combines the advantages of the Newton method and the Gauss-Newton method, and can realize effective supervised learning in the training process of the neural network model. The weight updating formula is as follows:
Wh+1＝Wh-Jhrh(Hn+μI)-1 (7)
in the formula: w is a weight matrix; h is the number of iterations; j. the design is a squarehIs a Jacobian matrix; r ishIs a residual error vector; h is a Hessian matrix; mu.sIs a penalty factor; and I is an identity matrix.
slmIs the iteration step size, when mu is more than 0, the covariance matrix
Are always positive numbers, which ensures slmCan represent the descending direction of f (x):
the definition of the falling direction of the function results from the expansion of the Taylor function if slm< 0, this means that the function f (x) decreases for each small step:
f(x)≈f(x0)+f(x0)(x-x0)＝f(x0)slm (10)
in order to obtain a time series neural network model, the flow rate of bottom mud, the flow rate of clear water, the feeding frequency and the temperature of a feeding hole are used as input variables, the moisture content of a discharging hole is used as an output variable, a single hidden layer is arranged, the number of neurons in the hidden layer is 10, the delay is 2 periods, and an activation function is a sigmoid function
According to the following steps of 6: 2: 2, dividing the data into a training set, a verification set and a test set, selecting data on part of the training set to obtain time series neural network output response, wherein the test result is shown in figure 3, and the graph clearly shows that the model output has good following performance with a target value, and most of moisture errors are far lower than +/-0.2.
The scatter plot of FIG. 4 shows the correlation of the time series neural network model output values with the target values in the training set, validation set, test set, and overall data. In fig. 4, we can respectively see that the regression rate R of the training set is 0.9972, the regression rate R of the verification set is 0.99883, the regression rate R of the test set is 0.9951, and the overall regression rate R is 0.99713, so we consider that the model training works well and has fully learned all the information between the data.
In order to check whether the training of the model achieves the ideal effect again, 2000 groups of data are added to carry out additional tests on the trained model, the response curve is shown in fig. 5, and it can be seen that the performance of the model output value following the target curve is good, and the error is limited within the range of +/-0.1.
In step 105, mapping a historical data set M of deviation between a predicted value of the moisture content of the discharge port and a set value in the material mixing process to a moisture content deviation classification set N in a mode of function mapping according to manual experience by a box separation method, and obtaining a deviation classification function y (f (x) of the moisture content corresponding to the deviation x (x belongs to M) of the moisture content of the material mixing. Because the measurement range of the moisture meter is between 0% and 10%, if the difference between the predicted value of the moisture content and the set value is less than-0.8%, -0.8% -0.2% -, 1%, and greater than 1%, the deviation categories of the moisture content are 1, 2, 3, 4, and 5, respectively, and the classification results are shown in table 5:
TABLE 5 moisture content deviation classification set
In step 106, the current fuzzy water adding amount output by the fuzzy control word table is corrected on line according to the deviation classification table in the previous step, and the current target water adding amount is obtained. Collecting data of raw materials in the process of mixing and adding water into lead-zinc ore sintered return powder next time, inputting the vibration frequency of a current conveyor, the flow rate of bottom mud, the flow rate of clear water and the temperature of a feed inlet of a cooling cylinder into a trained moisture content prediction model to obtain a predicted value of the moisture content of a material at a discharge outlet after water is added, further obtaining deviation classification of the predicted value of the moisture content and a preset moisture content value, and then adjusting the current fuzzy water adding amount a according to the deviation classification and manual operation experience:
if the deviation type y is 1, the target water adding amount P is a x (1+ 15%);
if the deviation type y is 2, the target water adding amount P is a x (1+ 7%);
if the deviation type y is 3, the target water adding amount P of the time is a;
if the deviation type y is 4, the target water adding amount P is a x (1-7%);
if the deviation type y is 5, the target water adding amount P is a x (1-15%);
wherein, P is the target water adding amount of this time, and a is the fuzzy water adding amount of this time output by the fuzzy control word table.
The invention also provides an automatic moisture control system of the sintering returned powder, which comprises a processor and a memory connected with the processor, wherein the memory stores an automatic moisture control program of the sintering returned powder, and the automatic moisture control program of the sintering returned powder is executed by the processor to realize the steps of the automatic moisture control method of the sintering returned powder.
As shown in fig. 6, the automatic moisture control system for the sintering return powder further includes: the device comprises a cooling cylinder, a clear water storage device, a bottom sediment storage device, a vibrating feeder, a material conveyor belt and an infrared thermometer for measuring the temperature of a feeding port and the temperature of a discharging port of the cooling cylinder;
the clear water storage device, the bottom sediment storage device and the vibrating feeder are connected with the feeding end of the cooling cylinder;
the material conveying belt is connected with the discharge end of the cooling cylinder;
a clean water electric valve and a clean water flowmeter are sequentially arranged on a pipeline between the clean water storage device and the cooling cylinder;
a manual valve and a bottom mud flowmeter are sequentially arranged on a pipeline between the bottom mud storage device and the cooling cylinder;
and a moisture meter is also arranged on the material conveying belt.
Still be equipped with the standby water device on the material conveyer belt, install standby water sprinkler additional on the standby water device for the raise dust phenomenon's when avoiding moisture content on the low side appearance.
In view of the characteristics that the bottom mud is unstable in property and sometimes blocks the valve, the control system of the invention keeps the original manual adjustment mode of the bottom mud, the clear water electric adjustment valve is replaced, the bottom mud flow meter and the clear water flow meter are installed to measure the real-time bottom mud flow and the clear water flow, and the part with insufficient bottom mud is supplemented by the clear water.
The control system of the invention adopts a more advanced NIR moisture meter which can directly display the moisture content in a numerical form, thereby facilitating the determination of the moisture range meeting the requirements of workers. The design is equipped with water and adds the function, mainly be the safe operation problem of system of considering, the water of preparing is to discharge gate material moisture content on the low side or cross low and design, moisture on the low side not only can influence the efficiency of next process sintering pelletization, it is also possible to take place the conveyer belt raise dust, cause great pollution to the environment, and if in case the cold section of thick bamboo appears not adding water or the water addition is too little because of various reasons before leading to discharge gate material moisture to be low excessively, can direct burn the conveyer belt even, and bring very unfavorable interference for follow-up technology, it is unexpected, the loss that causes to production will be immeasurable.
Install the service water sprinkler additional on the material conveyer belt, the appearance of raise dust phenomenon when avoiding moisture content to hang down. Meanwhile, a signal cable is laid for data transmission, real-time monitoring water data are transmitted into the PLC system and the sintering plant server, real-time data parameters are provided for the control system, and meanwhile production operators can know the real-time changing water data. In addition, a control program is compiled according to the water control model, a man-machine interaction interface, namely an automatic water adding control picture, is installed on a sintering duty room industrial control computer, and the upper control interface adopts WINCC configuration software which is friendly in interface and powerful in function. The returned powder cooling and water adding control cabinet is designed and manufactured, and an SMARTLINE touch screen is configured and installed on a control cabinet panel, so that the running state of a cooling cylinder is displayed in real time, the returned powder cooling and water adding amount is controlled in real time, and the informatization and intellectualization of the production process are improved.
The working mode of the control system of the invention is as follows: an expected moisture content value of a conveyor belt material is set on a touch screen or WINCC operation interface, the blanking amount is roughly calculated according to the collected frequency of a vibrating feeder, according to the blanking amount and the temperature of a feed inlet, an operator adjusts the flow rate of bottom mud to a proper position by adjusting a manual valve of the bottom mud (the specific flow rate of the bottom mud is related to the property of the bottom mud on site, not only a certain usage amount of the bottom mud is ensured, but also a flowmeter and a pipeline are blocked due to overhigh concentration of the bottom mud, generally, about fifty percent of the total water addition amount is properly adjusted, the rest deficiency is supplemented by clear water, namely the actual control output amount), then the collected flow rate of the bottom mud, the feeding temperature, the frequency of the vibrating feeder and the clear water flow are subjected to feedforward compensation by utilizing a fuzzy control principle, the current fuzzy water addition amount is calculated by combining with the experience of an expert, and a water content prediction model is established by utilizing a neural network, and then the fuzzy output quantity is corrected on line according to the predicted deviation of the water content and a set value, so that the target water adding quantity of the time is obtained. The fuzzy controller transmits a control signal of the corrected target water adding amount to the fresh water electric valve, and the fresh water electric valve immediately acts after receiving the adjusting signal to complete the first step of water regulation; and quantifying the deviation to the addition amount of clear water by applying a PID control theory according to the deviation of the real-time measurement result of the conveyor belt material moisture meter and the set moisture content set value, and finishing the second step of moisture regulation until the moisture content of the conveyor belt material is basically realized. For avoiding the unexpected condition to take place, because of the moisture content change that the force of inequality leads to, install the water installation of reserve on the conveyer belt and be used for emergent, when detecting conveyer belt material moisture content on the low side, discharge gate temperature on the high side, the hierarchical action of water electric valve of reserve receives immediately according to fuzzy control rationale after the adjustment signal to avoid under the extreme environment because of the too low raise dust that leads to of conveyer belt material moisture content and the equipment damage that the material temperature leads to of material height. So far, the staged function of the whole lead-zinc ore sintering return powder water content automatic control system is completely realized.
The system adopts a mode of outputting and correcting the output value of the fuzzy controller by the prediction model, greatly improves the problems of large lag, nonlinearity, slow time variation and strong coupling which cannot be independently solved by single fuzzy control, obtains the deviation from the set moisture value after introducing the moisture predicted value, can predict the water adding amount at this time in advance, adopts a method of classifying moisture deviation, ensures that the system can more accurately correct the water adding amount on the premise of not frequently acting, limits the action of the prediction model by a mode of leading fuzzy control, and avoids the instability of the system caused by sudden change of the predicted value due to sudden change of industrial environment. Meanwhile, the problems that the moisture content of the conveyor belt materials is too low and dust is raised, the temperature of the materials at the discharge port is too high, the equipment is damaged, the moisture content of the conveyor belt materials is too high and the materials slip and the like caused by inaccurate judgment, untimely adjustment and unstable moisture due to manual experience operation are solved, the labor intensity of workers is greatly reduced, the labor efficiency is improved, the operation environment of operating personnel is improved, the influence of the dust raised on the surrounding environment is reduced, the environmental management level of a factory is improved, the requirements of relevant laws and regulations of environmental protection in China are met, the efficiency and the quality of sintering production are improved, the labor cost is reduced, and the coordination and unification of ecological benefits and economic benefits are realized.
The system can realize automatic and dynamic control of the water adding amount of the sintering return powder under different working conditions, so that the moisture of the return powder meets the requirements of sintering production, and displays the data of real-time moisture, water adding amount and the like of materials in the production process, the operation mode and the like. The system is suitable for controlling the water adding amount in the raw material mixing process in various sintering production processes, can be designed into an independent control system or realized by using a PLC (programmable logic controller) control system based on automatic production process control, is less restricted by field conditions, has low cost, solves the problem of environmental pollution caused by the production process, improves the production efficiency of a factory, reduces the labor intensity of workers, and realizes the coordination and unification of ecological benefits and economic benefits.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (9)
1. An automatic moisture control method for sintered powder return is characterized by comprising the following steps:
acquiring production data of the sintering return powder, and establishing a production database; the production database comprises the feeding frequency of returned powder, the temperature of a feeding hole, the flow of bottom mud, the flow of clear water, the flow of standby water, the temperature of a discharging hole and the real-time moisture content of materials on a conveyor belt;
adopting a multi-factor control method to take the feeding frequency and the feeding port temperature which have the greatest influence on the moisture content of the material in the production database as input variables, taking the clear water flow as an output variable, and establishing a fuzzy control word table between the input quantity and the output quantity;
carrying out fuzzy control word table query according to the currently acquired feeding frequency and feeding port temperature data, and outputting the current fuzzy water addition amount;
constructing an optimal historical data set according to the acquired production data of the sintering return powder, and establishing a water content prediction model by adopting the optimal historical data set; inputting the currently collected data into a prediction model, and outputting a moisture content prediction value;
dividing the deviation category of the predicted water content value and a preset water content value according to a box dividing method;
and correcting the current fuzzy water adding amount output by the fuzzy control word table on line according to the deviation type to obtain the current target water adding amount.
2. The method for automatically controlling the moisture of the sintered powder lot according to claim 1, wherein a multi-factor control method is adopted to take the feeding frequency and the feeding port temperature which have the greatest influence on the moisture content of the material in the production database as input variables, take the clear water flow as an output variable, and establish a fuzzy control word table between the input quantity and the output quantity, specifically:
assuming that the sampling obtained feeding frequency is f and the temperature of the feeding hole is t, calculating the clear water output u:
wherein the content of the first and second substances,
and "·" are fuzzy matrix operators, R denotes fuzzy relations, u1Control outputs obtained for the first fuzzy relation, NBuAs a fuzzy set NBf，NMfOr fuzzy set NBt，NMtAnd meanwhile, in a fuzzy mapping set when the temperature is established, if the membership function values of the feeding frequency f and the feeding port temperature t correspond to the quantized grades and take 1, the rest are all taken as 0, and the formula (2) is simplified as follows:
in the formula (I), the compound is shown in the specification,
are fuzzy sets NB respectivelyf，NMfThe membership of the ith element (i.e., the measured feed frequency is the ith grade);
are fuzzy sets NB respectivelyt，NMtThe membership degree of the jth element, namely the measured temperature is the jth grade; similarly, the control quantity u is obtained2,u3,……unThen the control quantity is represented as a fuzzy set u as:
u＝u1+u2+u3+……+un (4)
and converting the controlled variable from the fuzzy variable to the accurate variable by adopting a maximum membership method to obtain a fuzzy control word table.
3. The method for automatically controlling the moisture of the returned sintering powder according to claim 1, wherein a preferred historical data set is constructed according to the acquired production data of the returned sintering powder, and a moisture content prediction model is established by using the preferred historical data set; inputting the currently collected data into a prediction model, and outputting a moisture content prediction value, specifically:
the preferred historical data set comprises: feeding frequency, bottom mud flow, clear water flow, feeding port temperature and discharging port material moisture content;
the water content prediction model is a water content prediction model based on a time series neural network; the currently acquired data comprise bottom mud flow, clear water flow, feeding frequency and feeding port temperature which are used as input variables of the prediction model, and the moisture content of the material at the discharging port is used as an output variable of the prediction model.
4. The method for automatically controlling the moisture content of the sintered powder lot according to claim 3, wherein the construction of the time-series neural network-based moisture content prediction model specifically comprises: and the preferred historical data set is processed according to the following steps of 6: 2: and 2, dividing the model into a training set, a verification set and a test set in proportion, and training by using a levenberg marquardt algorithm by using training set data of an optimal historical data set through off-line training to obtain the model.
5. The method for automatically controlling the moisture content of the sintered powder lot according to claim 1, wherein the classification of the deviation between the predicted moisture content value and the preset moisture content value is divided according to a box separation method, and specifically comprises:
mapping a historical data set M of deviation between a predicted value of the moisture content of the material at a discharge port and a preset moisture content value in the material mixing process to a moisture content deviation classification set N in a mode of function mapping to obtain a deviation classification function y (f (x)) of the moisture content corresponding to the deviation x of the moisture content of the material to be mixed, wherein x belongs to M;
because the measurement range of the moisture meter is between 0% and 10%, if the difference between the predicted value of the moisture content and the set value is less than-0.8%, -0.8% -0.2% -1%, and greater than 1%, the deviation categories of the moisture content are 1, 2, 3, 4, and 5, respectively.
6. The method for automatically controlling the moisture of the sintering return powder according to claim 5, wherein the step of correcting the current fuzzy water addition amount output by the fuzzy control word table on line according to the deviation category to obtain the current target water addition amount comprises the following steps:
if the deviation type y is 1, the target water adding amount P is a x (1+ 15%);
if the deviation type y is 2, the target water adding amount P is a x (1+ 7%);
if the deviation type y is 3, the target water adding amount P of the time is a;
if the deviation type y is 4, the target water adding amount P is a x (1-7%);
if the deviation type y is 5, the target water adding amount P is a x (1-15%);
wherein, P is the target water adding amount of this time, and a is the fuzzy water adding amount of this time output by the fuzzy control word table.
7. An automatic moisture control system for returned sintering powder, comprising a processor and a memory connected to the processor, wherein the memory stores an automatic moisture control program for returned sintering powder, and the automatic moisture control program for returned sintering powder realizes the steps of the automatic moisture control method for returned sintering powder according to any one of claims 1 to 6 when executed by the processor.
8. The system of claim 7, further comprising: the device comprises a cooling cylinder, a clear water storage device, a bottom sediment storage device, a vibrating feeder, a material conveyor belt and an infrared thermometer for measuring the temperature of a feeding port and the temperature of a discharging port of the cooling cylinder;
the clear water storage device, the bottom sediment storage device and the vibrating feeder are connected with the feeding end of the cooling cylinder;
the material conveying belt is connected with the discharge end of the cooling cylinder;
a clean water electric valve and a clean water flowmeter are sequentially arranged on a pipeline between the clean water storage device and the cooling cylinder;
a manual valve and a bottom mud flowmeter are sequentially arranged on a pipeline between the bottom mud storage device and the cooling cylinder;
and a moisture meter is also arranged on the material conveying belt.
9. The system for automatically controlling the moisture of the sintering returned powder according to claim 7, wherein a standby water device is further arranged on the material conveying belt, and a standby water spraying device is additionally arranged on the standby water device and used for avoiding the occurrence of a dust emission phenomenon when the moisture content is relatively low.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111620431.4A CN114265885A (en) | 2021-12-28 | 2021-12-28 | Automatic control method and system for moisture of sintering return powder |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111620431.4A CN114265885A (en) | 2021-12-28 | 2021-12-28 | Automatic control method and system for moisture of sintering return powder |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114265885A true CN114265885A (en) | 2022-04-01 |
Family
ID=80831186
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111620431.4A Pending CN114265885A (en) | 2021-12-28 | 2021-12-28 | Automatic control method and system for moisture of sintering return powder |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114265885A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115090200A (en) * | 2022-05-27 | 2022-09-23 | 福建龙氟化工有限公司 | Automatic batching system for preparing electronic grade hydrofluoric acid and batching method thereof |
| CN115490100A (en) * | 2022-09-30 | 2022-12-20 | 深圳市中金岭南有色金属股份有限公司凡口铅锌矿 | Control method, device and system of lifting system |
| CN117469970A (en) * | 2023-12-25 | 2024-01-30 | 江苏美特林科特殊合金股份有限公司 | Accurate feeding system for smelting electron beam metal niobium |
-
2021
- 2021-12-28 CN CN202111620431.4A patent/CN114265885A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115090200A (en) * | 2022-05-27 | 2022-09-23 | 福建龙氟化工有限公司 | Automatic batching system for preparing electronic grade hydrofluoric acid and batching method thereof |
| CN115490100A (en) * | 2022-09-30 | 2022-12-20 | 深圳市中金岭南有色金属股份有限公司凡口铅锌矿 | Control method, device and system of lifting system |
| CN117469970A (en) * | 2023-12-25 | 2024-01-30 | 江苏美特林科特殊合金股份有限公司 | Accurate feeding system for smelting electron beam metal niobium |
| CN117469970B (en) * | 2023-12-25 | 2024-03-12 | 江苏美特林科特殊合金股份有限公司 | Accurate feeding system for smelting electron beam metal niobium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114265885A (en) | Automatic control method and system for moisture of sintering return powder | |
| Chai et al. | Optimal operational control for complex industrial processes | |
| Chai et al. | Hybrid intelligent control for optimal operation of shaft furnace roasting process | |
| US7139619B2 (en) | Kiln free lime control | |
| US20050154477A1 (en) | Kiln control and upset recovery using a model predictive control in series with forward chaining | |
| Zhou et al. | Intelligence-based supervisory control for optimal operation of a DCS-controlled grinding system | |
| US20040210325A1 (en) | Kiln thermal and combustion control | |
| CN106644013B (en) | Automatic weighing and batching system and control method | |
| CN108393146B (en) | Self-adaptive optimal decoupling control method for coal pulverizing system of steel ball coal mill | |
| CN101441443B (en) | Data processing method and device of sintering control parameter | |
| Korobiichuk et al. | Modern development technologies and investigation of food production technological complex automated systems | |
| CN103173584A (en) | Blast furnace burden-distribution control system with self-learning control function | |
| CN112317110A (en) | Ore grinding granularity prediction system and method based on deep learning | |
| Jiang et al. | Real-time moisture control in sintering process using offline–online NARX neural networks | |
| Chen et al. | Expert system based adaptive dynamic matrix control for ball mill grinding circuit | |
| CN115318431A (en) | Control method and device for heavy-medium shallow slot system and processor | |
| Chen et al. | A chance-constrained programming approach for a zinc hydrometallurgy blending problem under uncertainty | |
| CN102156405B (en) | Sintered ore chemical component prediction and intelligent control system under small sample poor information | |
| Chai | Optimal operational control for complex industrial processes | |
| Giridhar et al. | Intelligent process management for continuous operations in pharmaceutical manufacturing | |
| CN112070408A (en) | Agglomerate composition forecasting model based on big data and deep learning | |
| Bascur | Process control and operational intelligence | |
| KR101858860B1 (en) | Apparatus for controlling heat of blast furnace | |
| CN115329659A (en) | Method and system for real-time early warning and intelligent control of dioxin emission in waste incinerator | |
| CN109142167B (en) | Method for establishing online robust regular soft measurement model of ore grinding granularity |
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 |