Method for realizing closed-loop automatic spraying control by high-temperature cinder humidifying process

Abstract

The invention discloses a method for achieving closed-loop automatic spraying control in a high-temperature cinder humidifying process. The method comprises the following steps that S1, torque feedback of a humidifying roller frequency converter is detected in real time; s2, calculating a total basic value of real-time cinder in the roller according to torque feedback; s3, calculating a slag total amount correction value in real time according to the rotating speed of the roller and the real-time spray water flow data, and calculating a true value according to the basic value and the correction value; s4, according to the calculated real-time cinder amount, the basic spraying water flow needed by spraying is calculated; s5, calculating a spray water flow correction value according to dynamic data such as the temperature of the rear section of the roller and the real-time cinder amount change trend, and giving a water spray set value according to the basic value and the correction value; and S6, according to the water spraying set value, coordinated distribution and automatic control of a plurality of spraying water loops are achieved, and closed-loop control of automatic spraying is completed. The method has the beneficial effects that manual observation feedback is replaced by indirect parameter measurement such as roller electrode moment feedback, so that automatic control becomes possible.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sulfuric acid production emissions, in particular to a high-temperature cinder drum cooling and spraying device generated after reaction of a sulfuric acid production device, and particularly relates to a method for realizing closed-loop automatic spraying control of a high-temperature cinder humidification process. BACKGROUND

[0002] The process of sulfuric acid production is as follows: after being proportioned, the ore is uniformly thrown into the roasting furnace by the ore throwing machine, and boiling is formed in the roasting furnace to cause combustion reaction. After the reaction, the high-temperature flue gas is sent to the purification section after being cooled and collected by the waste heat boiler, the cyclone dust collector and the electric dust collector, and is used for producing sulfuric acid; and the high-temperature cinder generated after the reaction is discharged into the cinder cooler, and is sent to the cinder intermediate warehouse through the inclined scraper machine after being preliminarily cooled and cooled by the cinder cooler, and is mixed with the dust collected from the waste heat boiler and the cyclone dust collector in the cinder intermediate warehouse, and is further humidified and cooled in the humidifier.

[0003] The device involved in the present application is a drum humidifier. The drum of the humidifier is driven to rotate continuously by the motor, so that the cinder gradually moves backward in the drum. At the same time, the tail section of the drum sprays water on the cinder to cool it. The cinder that reaches the standard temperature and humidity falls from the tail of the drum by its own weight to the belt conveyor, and finally the cooled cinder powder is sent to the cinder storage by the belt conveyor. The approximate process is as follows Figure 1 .

[0004] The process requires that the outlet cinder temperature be not too high to prevent scalding the conveying belt, the moisture content of the cinder be not too high to prevent affecting the subsequent pellet process, and the humidification drum discharge port not emit dust. Before the completion of the present application, since the moisture content of the discharged cinder and the dust emission on the belt cannot be detected in real time, the spraying water control work of this part is completed manually by the operator. The operator adjusts the manual valve on the spraying pipeline in real time through visual observation beside the device to complete the final control. Since the site conditions are poor with high temperature and dust, which is very unfriendly to the operator, and the work intensity is very high, it is hoped that the automatic control of the device can be realized.

[0005] The Chinese utility model patent CN220026486U discloses a dust discharge port humidification drum device, which comprises a dust discharge port and a humidification assembly. The humidification assembly comprises a humidification channel, a water spraying unit, a plurality of drum units, two fixed inclined plates and a plurality of protrusions. The plurality of protrusions are fixedly connected with the corresponding fixed inclined plates, but the automatic control of the spraying pipeline is not involved. SUMMARY

[0006] The technical problems solved by the present application include at least one of the following: the high-temperature slag in the roller cannot be measured, so the water spraying control cannot be simply proportional to the amount of incoming materials; the temperature of the high-temperature slag cannot be calculated, because the slag comes from different sources such as a slag cooler and a cyclone drum, and the temperature of the incoming slag is uncertain due to real-time changes in the operating state, resulting in different water spraying requirements; the final control requirement is the humidity of the discharged slag and the dust emission on the belt, which needs to be observed by vision and cannot be fed back by online instruments. Therefore, a method for realizing closed-loop automatic spraying control of high-temperature slag humidification process is provided.

[0007] The technical solution of the present application is: a method for realizing closed-loop automatic spraying control of high-temperature slag humidification process, comprising the following steps: S1: real-time detection of torque feedback of the humidification roller frequency converter; S2: calculation of the real-time total amount of slag in the roller based on the torque feedback; S3: real-time calculation of the total amount of slag based on the roller speed and real-time spraying water flow data, and calculation of the real value based on the base value and the correction value; S4: calculation of the required base spraying water flow based on the real-time amount of slag; S5: calculation of the spraying water flow correction value based on the temperature of the rear section of the roller, the real-time slag amount change trend and other dynamic data, and calculation of the spraying water set value based on the base value and the correction value; S6: realization of coordinated distribution and automatic control of a plurality of spraying water circuits based on the spraying water set value, and completion of the closed-loop control of automatic spraying.

[0008] In the above-mentioned solution, S2 includes: 1) using the torque feedback of the roller motor frequency converter obtained by indirect measurement as the real-time slag amount calculation reference; 2) analyzing the slag storage data in the roller based on different roller speeds and lengths; 3) calculating the change amount data of the slag in the roller based on the differential of the storage data; and 4) using a soft measurement baseline model method based on time sequence characteristics to complete modeling and real-time calculation of the slag amount value.

[0009] In the above-mentioned solution, S5 includes: 1) using the real-time spraying amount calculated based on the real-time slag amount as the calculation basis; 2) judging the heating condition of the slag in the roller based on the temperature of the rear section of the roller, and correcting the spraying water amount in real time based on the heating condition; and 3) calculating the final appropriate spraying water set amount by changing the spraying water flow in advance based on the real-time slag amount change trend and the travel of the slag in the roller.

[0010] In the above-mentioned solution, 3) of step S2 refers to approximating the differential of the torque by the torque deviation of the fixed time slice t step The differential of the torque is approximated by the torque deviation of the fixed time slice t , where F d : the differential of the torque, , and then a certain number of F dThe mean filter is used to eliminate data disturbance and obtain stable torque change value F f The data for further modeling is as follows: The F calculated above is used to obtain approximate real-time slag amount through further modeling. f

[0011] The modeling model obtained in 4) of step S2 in the above scheme is as follows: Wherein, Z(s) is an instantaneous slag amount transfer function, K is a model static gain coefficient, T is an inertia time constant, and t is a pure lag time constant.

[0012] The correction value calculation formula of the total amount of burned slag in step S3 in the above scheme is as follows: Wherein, ZR(s) is the corrected instantaneous slag amount, W(s) is the water remaining in the burned slag after spraying cooling, W R : real-time spraying water amount, T R : real-time roller rear section temperature, and K is an empirical coefficient.

[0013] The final spraying water amount setting value WS(s) calculation formula of step S5 in the above scheme is as follows: Wherein, K is an empirical coefficient, and C is an empirical constant.

[0014] The beneficial effects of the present application are as follows: 1) indirect parameter measurement such as roller electrode torque feedback is used to replace artificial observation feedback, so that automatic control becomes possible; 2) automatic closed-loop control of the spraying system is realized, and the labor intensity of the on-site operator is reduced; and 3) through automatic spraying control implementation of multiple systems, the number of on-site operators is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 FIG. 1 is a schematic diagram of a roller spraying humidification system; Figure 2 FIG. 2 is a control flowchart of the present application; Figure 3 Torque signal-based inventory and change amount analysis; Figure 4 FIG. 3 is a soft measurement baseline model based on time sequence characteristics; Figure 5 FIG. 4 is a whole design scheme of a slag amount soft instrument system; Figure 6 FIG. 5 is an overall scheme of on-site system control; Figure 7 FIG. 6 is a control scheme in a limit situation; Figure 8 FIG. 7 is a torque feedback-actual spraying history curve diagram of an application example. DETAILED DESCRIPTION​​

[0016] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments that can be obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0017] As shown in the drawings, Figure 2 The present application comprises the following steps: S1: detecting the torque feedback of the humidifying roller frequency converter in real time; S2: calculating the total amount of real-time slag in the roller based on the torque feedback; S3: calculating the total amount of slag in real time based on the roller speed and real-time spray water flow and other data, and calculating the real value based on the basic value and the correction value; S4: calculating the basic spray water flow required for spraying based on the calculated real-time slag amount; S5: calculating the spray water flow correction value based on the temperature of the rear section of the roller, the real-time slag amount change trend and other dynamic data, and giving the spray setting value based on the basic value and the correction value; and S6: realizing the coordinated distribution and automatic control of a plurality of spray water circuits based on the spray setting value, and completing the closed-loop control of automatic spraying.

[0018] Specifically, S1 comprises: The original humidifying roller motor does not have a frequency converter, and has been running at a public frequency for a long time. In order to realize real-time feedback of the stress of the roller, a vector frequency converter is configured for the motor, and the real-time torque feedback sampling of the frequency converter is used to indirectly calculate the real-time slag amount by using a soft instrument.

[0019] Figure 1 After the roller electrode in the lower left corner of the spray system schematic diagram is configured with a vector frequency converter, its speed can be adjusted by the PLC and touch screen system arranged on site, and the torque and speed feedback of the frequency converter can be sampled by the PLC, so that the real-time torque feedback value of the spray system roller motor is obtained as the basis for further analysis.

[0020] S2 comprises: obtaining the total amount of slag in the humidifying roller and the real-time incoming amount by comprehensive calculation and analysis based on the data sampling of S1, and realizing subsequent control based on the same. For the humidification of the roller, we need to know the total amount of the roller at present, and also need to know the real-time incoming amount at present, or the real-time incoming amount is more important than the total amount of the storage.

[0021] As shown in the drawings, Figure 3 Assuming that the curve in the figure is the curve of the frequency converter torque feedback (ordinate) changing with time (abscissa) measured by us. Then the torque (solid line) more reflects the amount (current load), and the dashed line (derivative) more reflects the change amount (change trend) of the real-time incoming amount.

[0022] Here the differential of the torque is approximated by the torque deviation before and after a fixed time slice (t step ): , F d : the approximation of the torque differential, Then the F d is averaged to filter out the disturbance and get a more stable torque change value F f as the basis data for further modeling: Using the F f calculated above, we further try to get an approximate real-time slag amount by modeling. We use a "soft measurement baseline model based on time series features" to design the model. Figure 4 The soft measurement baseline model based on time series features mainly uses RNN (Recurrent Neural Network) series, which has strong time series feature extraction capability. As shown in the figure, the multi-layer RNN takes the output of the previous layer as the input of the next layer, and finally completes the establishment of the soft instrument model from data sampling.

[0023] The final on-site model is a typical first-order pure lag model. The instantaneous slag amount Z(s) transfer function calculated by the model is as follows, and its specific parameters are obtained by the above modeling method: K: model static gain coefficient, T: inertia time constant, t: pure lag time constant.

[0024] The obtained model is brought into the soft instrument system final scheme as shown in Figure 5 , and the original torque data obtained from the PLC is used as the original data of the model input. The established soft instrument model is used as the calculation algorithm, and the real-time estimated slag amount can be obtained. The calculated slag amount is used as the first-level soft instrument feedback to provide reference for the operator, and also as the input for the next step control.

[0025] At the same time, according to the belt scale installed on site, the soft instrument system retains the feedback mode of correcting the model, which can be corrected online through manual observation on site to ensure accuracy.

[0026] S3 includes: the real-time slag amount calculated in S2 is mainly calculated through the torque feedback of the roller, so there will be a certain deviation during operation, which needs to be further calibrated through a certain empirical formula.

[0027] For example, when the roller speed changes, it will cause the residence time of the slag in the roller to change, resulting in different real-time slag amounts corresponding to the same torque feedback. This part of the correction data is reflected in the pure lag data in the model in S2.

[0028] At the same time, since the torque feedback is the feedback caused by the gravity of the material in all the drums, it includes not only the cinder but also the water sprayed into the drum, so the weight of the water needs to be removed from the soft instrument when the water content of the cinder is different. This part of the calculation is mainly completed through the drum temperature and the physical instantaneous change amount and other parameters through an empirical formula. The correction here is based on experience. The higher the temperature of the rear section of the drum, the more the evaporation amount, and the higher the water amount to be deducted. Conversely, the lower the temperature of the rear section of the drum, the smaller the evaporation amount, and the lower the water amount to be deducted. The empirical formula for water correction is as follows: , , ZR(s): corrected instantaneous slag amount, W(s): water content remaining in the cinder after spraying cooling, W R : real-time spraying water amount, T R : real-time temperature of the rear section of the drum, K: empirical coefficient. The calculation process of S2 and S3 is embodied in the left upper side "feed amount estimation" module in the Figure 6 overall control scheme, Figure 6 which is the overall control scheme block diagram finally realized in the industrial field. Figure 6 The GPC in the above formula refers to the generalized predictive control, which is an advanced control algorithm. By establishing a prediction model, the future output behavior of the system is estimated, and then a series of control signals are calculated through rolling optimization, so that the system output can smoothly track the expected value. At the same time, the model error and external disturbance are compensated through feedback correction.

[0029] S4 includes: through the corrected real-time cinder amount in S3, the normal running data of the system can be represented. Through the above series of work, the comprehensive calculation of the online instrument replaces the current manual visual observation feedback. According to different process levels, the total spraying water amount required for spraying in the drum is calculated with different preset basic water distribution amounts. The spraying water set value of this step only needs to be calculated according to the empirical value through a first-order equation. The final water spraying amount set value WS(s) calculation formula is as follows: , K: empirical coefficient, C: empirical constant. The calculation process of S4 is embodied in the middle "water flow rate setting calculation" module in Figure 6 .

[0030] S5 includes: the spraying water amount calculated through S4 also needs to be further corrected. This part of the correction data is input into the "water flow rate setting calculation" module in Figure 6 for superposition and correction, as described below: "bias coefficient": through the naked eye observation of the cinder water content by the field operator, the spraying water bias amount is manually set to realize manual intervention in emergency situations; "Temperature coefficient": through the high and low of the rear section temperature of the roller, the experience coefficient is converted, the higher the temperature, the larger the water spraying ratio, the lower the temperature, the smaller the water spraying ratio, and finally multiplied by the basic water spraying amount calculated by S4; "Travel coefficient": multiplied by the proportion of the roller motor speed, in order to adapt to the influence of different motor speeds on the travel speed of the slag in the roller.

[0031] In addition, when the current section of the sulfuric acid system is unstable, sometimes it will cause the abnormal working condition of very large instantaneous slag amount, at this time the conventional control system will have the risk of insufficient water control, therefore when detecting this condition, abnormal processing is carried out, all the water supply of the rear section is opened to prevent the belt from running ash. The processing logic is as follows Figure 7 .

[0032] S6 includes: with the water supply amount setting of S5, only the automatic water supply regulating valve and the water flow feedback instrument of the final several fields are needed to complete the automatic closed loop. However, the field conditions are relatively poor, the water supply cleanliness is not enough, which affects the control effect due to the poor linearity of the valve.

[0033] Therefore, as Figure 6 The two branches on the right side of the "water flow setting calculation" module, namely GPC-up water valve and GPC-down water valve, and the upper section spraying belong to the water flow regulation module, a water flow regulation scheme of multiple spraying is designed, two water paths are used to control a target flow, and the problem of poor single-path water flow control effect is solved.

[0034] The final goal is to achieve: Through indirect measurement and the use of soft instrument calculation, the problem of only visual feedback in the previous field is solved, and the basis for automatic spraying control is provided.

[0035] Based on the soft instrument, combined with the experience formula, the closed loop control method of automatic spraying is realized, and long-term stable operation is carried out in the field.

[0036] Reduce the labor intensity of workers in the field, reduce the allocation of workers, and reduce the process phenomena such as ash and water running caused by operation errors.

[0037] As Figure 8 shown, it is a torque feedback-actual water spraying history curve diagram of the application example. Among them, the roller torque is used as the input of the algorithm, and the actual water spraying is the final output.

Claims

1. A method for realizing closed-loop automatic spray control of a high-temperature slag moistening process, characterized in that: The method comprises the following steps: S1: Real-time detection of the torque feedback of the humidifying roller frequency converter; S2: Calculating the total amount of real-time cinder in the roller based on the torque feedback; S3: Real-time calculation of the total amount of cinder correction value based on the roller speed and real-time spray water flow data, and calculating the real value based on the base value and the correction value; S4: Calculating the basic spray water flow required for spraying based on the calculated real-time cinder amount; S5: Calculating the spray water flow correction value based on the temperature of the rear section of the roller, the real-time cinder amount change trend and other dynamic data, and giving the spray setting value based on the base value and the correction value; S6: Realizing the coordinated distribution and automatic control of several spray water circuits based on the spray setting value, and completing the closed-loop control of automatic spraying.

2. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 1, characterized in that: The S2 comprises: 1) using the torque feedback of the roller motor frequency converter obtained by indirect measurement as the real-time cinder amount calculation benchmark; 2) analyzing the cinder stock data in the roller based on different roller speeds and lengths; 3) calculating the cinder change amount data in the roller based on the differential of the stock data; 4) using the soft measurement baseline model method based on time sequence characteristics to complete modeling and real-time calculation of the cinder amount value.

3. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 1, characterized in that: The S5 comprises: 1) using the real-time spraying amount calculated based on the real-time cinder amount as the calculation basis; 2) judging the cinder heating condition in the roller based on the temperature of the rear section of the roller, and correcting the spray water amount in real time based on the heating condition; 3) changing the spray water flow in advance based on the real-time cinder amount change trend and the travel of the cinder in the roller, and calculating the final appropriate spray water setting amount.

4. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 2, characterized in that: The step S2 3) refers to calculating the differential of the torque by fixing the time slice t step The differential of the torque is approximately calculated by the torque deviation before and after the time slice t: Wherein F d : torque differential approximation value, Then, the mean filtering is performed by a certain number of F d , and the relatively stable torque change value F f is obtained by eliminating data disturbance. The F f calculated above is used as the basis data for further modeling to obtain the approximate real-time slag amount.

5. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 2, characterized in that: The soft measurement baseline model based on time sequence characteristics in the step S2 4) refers to a recurrent neural network.

6. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 2, characterized in that: The modeling model obtained in the step S2 4) is where Z(s) is the instantaneous slag amount transfer function, K: model static gain coefficient, T: inertial time constant, t: pure lag time constant.

7. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 6, characterized in that: The slag total amount correction value of the step S3 is calculated by the following formula: , , wherein ZR(s): corrected instantaneous slag amount, W(s): water content in the slag after the spray cooling, W R : real-time spray water amount, T R : real-time rear drum temperature, K: empirical coefficient.

8. The method for realizing closed-loop automatic spraying control of high-temperature slag moistening process according to claim 7, characterized in that: The final water injection amount setting value WS(s) of the step S5 is calculated by the following formula: where K: an empirical coefficient, and C: an empirical constant.

Images