Intelligent control system, method and equipment for magnesium oxide rotary kiln

Abstract

The invention relates to the technical field of intelligent control of magnesium oxide rotary kilns, in particular to an intelligent control system, method and equipment of a magnesium oxide rotary kiln. The intelligent control method comprises the steps that the yield index Q1 and the energy consumption index K1 of the kiln condition are obtained through calculation according to historical data of controlled variables of the rotary kiln; real-time data of controlled variables of the rotary kiln are collected, and a yield correction coefficient Q2 and an energy consumption correction coefficient K2 of the kiln condition are obtained through fuzzy algorithm processing. And the Q1 is corrected according to the Q2 to obtain a real-time yield index Q3. And the K1 is corrected according to the K2 to obtain a real-time energy consumption index K3. And real-time data of the operation variables of the rotary kiln are collected, and the optimal combination of the operation variables is obtained through calculation by combining Q3 and K3. And the optimal combination of the operation variables is used as the input of an intelligent control system of magnesium oxide to realize intelligent control of the rotary kiln. The whole operation does not depend on manipulation and experience of operators, errors caused by manual operation are avoided, and energy consumption can be reduced while product quality can be guaranteed through real-time adjustment.

Description

Technical Field

[0001] The invention relates to the technical field of intelligent control of a magnesium oxide rotary kiln, and in particular to an intelligent control system, method and equipment of a magnesium oxide rotary kiln. Background Art

[0002] Magnesium oxide is an important inorganic material, which is widely used in refractory materials, building materials, chemical industry, medicine, agriculture and other fields. Among them, high temperature calcination in a rotary kiln is a very critical process in the production process of magnesium oxide, and this process is also a process with high energy consumption in the production process of magnesium oxide. As an important industrial equipment in this process, the magnesium oxide rotary kiln plays a vital role in the production process of magnesium oxide. However, under the current background of intelligent industrialization, there are still some challenges and problems in the control process of magnesium oxide rotary kiln. At present, most of the rotary kilns at home and abroad are in a manual control state, and some of them still use the most primitive manual fire-watching operation method, that is, by manually observing the "fire circle" in the kiln, judging the thermal state in the kiln, so as to adjust the feed fuel. Moreover, there are automatic control schemes for rotary kilns in the prior art, which all use multiple independent control loops to control multiple process parameters in the production process of the rotary kiln separately, and the control between each parameter is independent of each other, resulting in the existing rotary kiln control system being unable to achieve coordinated control of multiple parameters in the entire rotary kiln calcination process through global control, resulting in low production efficiency of the rotary kiln. Moreover, complex manual control operations are still required during the calcination process of the rotary kiln. In addition, due to the limitations of their own experience, reaction speed and operating accuracy, it is difficult for operators to timely and effectively overcome the impact of fuel system fluctuations and external interference on the production system, and different operators have different operating habits and methods. This results in the inability to use the multiple control variables of the rotary kiln to achieve the regulation of multiple process parameters in the production process of the rotary kiln even through the existing control scheme, which will affect the production efficiency and product quality of the rotary kiln. Summary of the invention

[0003] In order to solve the technical problem that the rotary kiln control system in the prior art controls multiple process parameters in the rotary kiln production process separately through multiple independent control loops, and the control of each parameter is independent of each other, resulting in the inability of the existing rotary kiln control system to achieve coordinated control of multiple parameters in the entire rotary kiln calcining process through global control, resulting in low production efficiency of the rotary kiln, the present invention provides a magnesium oxide rotary kiln control system, method and equipment.

[0004] The present invention is implemented by the following technical solution: a magnesium oxide rotary kiln control method, which comprises the following steps:

[0005] S1: Obtain the historical data of the controlled variables of the rotary kiln, and obtain the kiln output index Q1 and energy consumption index K1 through the model predictive control algorithm.

[0006] S2: Collect the real-time data of the controlled variables of the rotary kiln, and obtain the output correction coefficient Q2 and energy consumption correction coefficient K2 of the kiln condition through fuzzy algorithm processing based on the real-time data of the controlled variables.

[0007] S3: According to the output correction coefficient Q2, the output index Q1 of the kiln condition is corrected to obtain the real-time output index Q3 of the kiln condition; according to the energy consumption correction coefficient K2, the energy consumption index K1 of the kiln condition is corrected to obtain the real-time energy consumption index K3 of the kiln condition.

[0008] S4: Collecting real-time data of rotary kiln operating variables, and obtaining the optimal combination of operating variables according to the real-time data of rotary kiln operating variables, the real-time output index Q3 of kiln conditions, and the real-time energy consumption index K3 of kiln conditions through a finite time domain rolling optimization strategy.

[0009] The finite horizon rolling optimization strategy includes the following steps:

[0010] (1) An optimization model for optimizing the operating variables in the magnesium oxide rotary kiln is established based on the ARX prediction model; (2) Based on the input and output of the magnesium oxide rotary kiln at the current moment and the past moment, the predicted output value of the operating variables of the magnesium oxide rotary kiln is calculated through the optimization model, and then the output index and energy consumption index are used as constraints to obtain the objective function containing the future predicted input increment; (3) The predicted value of the operating variable is obtained by using the least squares method, and the predicted value is input into the optimization model to obtain the actual output value of the operating variable of the magnesium oxide rotary kiln; (4) The output predicted value of the operating variable at the next moment is calculated by the optimization model, and the solution is repeated in a rolling cycle to obtain the optimal combination of the operating variables.

[0011] S5: The optimal combination of operating variables is obtained as the input of the intelligent control system of magnesium oxide to realize intelligent control of the rotary kiln.

[0012] As a further improvement of the present invention, the expression of the ARX prediction model is:

[0013]

[0014] Where μ(t) is the input at time t; y(t) is the output at time t, t = 1, 2, 3…, N, N is the number of samples; α s and β i are all estimated parameters of the ARX prediction model; s is the number of output quantities, s = 1, 2, ..., e; i is the number of input quantities, i = 1, 2, ..., d.

[0015] As a further improvement of the present invention, the energy consumption index K1 is obtained by constructing a BP neural network prediction model with self-optimizing parameters. The BP neural network prediction model constructs an objective function with the minimum energy consumption index as the optimization target, and obtains the optimized value of the operating variable through a genetic algorithm, and then reduces its weight according to timeliness to obtain the energy consumption index K1.

[0016] As a further improvement of the present invention, the yield index Q1 is obtained by constructing a BP neural network prediction model with self-optimizing parameters. The BP neural network prediction model constructs an objective function with the maximum yield index as the optimization target, and obtains the optimized value of the operating variable through a genetic algorithm, and then increases its weight according to timeliness to obtain the yield index Q1.

[0017] As a further improvement of the present invention, the fuzzy algorithm includes the following processing steps:

[0018] (11) Determine the basic domain of the controlled variable, and quantify the basic domain of the controlled variable to obtain the fuzzy domain of the controlled variable.

[0019] (12) The fuzzy domain of the controlled variables is fuzzified by the maximum membership method to obtain the fuzzy set of the controlled variables.

[0020] (13) Establish fuzzy rules and perform fuzzy reasoning. Then use the centroid method to defuzzify the fuzzy set of controlled variables to obtain the precise set of controlled variables, and then normalize the precise set of controlled variables.

[0021] (14) The precise set of normalized controlled variables is used to obtain the output correction coefficient Q2 through the gradient descent method. The precise set of normalized controlled variables is used to obtain the energy consumption correction coefficient K2 through the LSTM algorithm.

[0022] As a further improvement of the present invention, the operating variables include the frequency of the kiln tail fan, the rotating speed of the rotary kiln, the feed rate, the gas input rate and the coal powder input rate.

[0023] As a further improvement of the present invention, the controlled variables include kiln tail NO x The concentration of CO at the kiln tail, the concentration of O2 at the kiln tail, the temperature at the kiln head, the temperature at the kiln tail, the negative pressure at the kiln head and the negative pressure at the kiln tail.

[0024] The present invention also includes an intelligent control system for a magnesium oxide rotary kiln, which is operated by the intelligent control method of the magnesium oxide rotary kiln described above. The intelligent control system includes: a data acquisition module and a control module. The input end of the data acquisition module is connected to the rotary kiln, and the output end is connected to the control module; the data acquisition module is used to obtain the process parameter data of the rotary kiln, and transmit the obtained data to the control module. The control module includes a feedback regulation module and a real-time optimization module. The real-time optimization module optimizes and calculates the data input by the data acquisition module through the least squares method and transmits the calculation result to the feedback regulation module. The feedback regulation module uses the calculation result of the real-time optimization module as the control target to realize intelligent control of the rotary kiln.

[0025] As a further improvement of the present invention, the process parameters include the frequency of the kiln tail fan, the speed of the rotary kiln, the feed rate, the gas input rate, the coal powder input rate, the kiln tail NO x The concentration of CO at the kiln tail, the concentration of O2 at the kiln tail, the temperature at the kiln head, the temperature at the kiln tail, the negative pressure at the kiln head and the negative pressure at the kiln tail.

[0026] The present invention also includes an intelligent control device for a magnesium oxide rotary kiln, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, an intelligent control system for a magnesium oxide rotary kiln as described above is created, thereby realizing intelligent regulation of the rotary kiln.

[0027] The technical solution provided by the present invention has the following beneficial effects:

[0028] (1) The intelligent control method provided by the present invention can collect the process parameters of the magnesium oxide rotary kiln in real time through the intelligent control system, automatically optimize the process parameters through the analysis of the intelligent control system, and adjust the process parameters of the magnesium oxide rotary kiln to match the kiln conditions through feedback adjustment. The entire operation does not rely on the operator's skills and experience, which not only avoids the errors caused by human operation, but also can ensure product quality while reducing energy consumption through real-time adjustment.

[0029] (2) The intelligent control system provided by the present invention has a faster adjustment speed, higher precision and no hysteresis compared with manual adjustment, thereby avoiding the problem of poor product quality caused by large fluctuations in the kiln conditions of the magnesium oxide rotary kiln due to manual operation, and improving the quality of products produced by the magnesium oxide rotary kiln. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 The present invention provides a flow chart of the steps of an intelligent control method for a magnesium oxide rotary kiln.

[0031] Figure 2This is a structural framework diagram of an intelligent control system for a magnesium oxide rotary kiln provided by the present invention.

[0032] Figure 3 This is a working framework diagram of an intelligent control system for a magnesium oxide rotary kiln provided by the present invention. DETAILED DESCRIPTION

[0033] The present invention is further described below in conjunction with specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0034] In the description of the present invention, it should be noted that for directional words, such as the terms "center", "lateral", "longitudinal", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicating directions and positional relationships are based on the directions or positional relationships shown in the drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and cannot be understood as limiting the specific protection scope of the present invention. The terms "first", "second", etc. in the specification and claims of the present invention are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. The terms "including" and "having" in the specification and claims of the present invention and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0035] Magnesium oxide is an important inorganic material, which is widely used in refractory materials, building materials, chemicals, medicine, agriculture and other fields. Among them, high-temperature calcination in a rotary kiln is a very critical process in the production process of magnesium oxide, and this process is also a process with high energy consumption in the production process of magnesium oxide. The existing magnesium oxide rotary kiln adjusts the parameters such as the feed amount, the rotation speed of the rotary kiln, and the frequency of the kiln tail fan by manual experience during calcination, which leads to fluctuations in the process in the magnesium oxide rotary kiln, and the fluctuation of the process will cause the quality of the produced magnesium oxide to fluctuate or increase the energy consumption. Therefore, a device that can intelligently adjust the calcination process of the rotary kiln is needed to improve the quality and efficiency of the produced magnesium oxide and reduce the energy consumption in the production process.

[0036] Therefore, the solution of the present application provides an intelligent control system and method for a magnesium oxide rotary kiln. The intelligent control system integrates existing sensors, data acquisition and model analysis technologies to realize automatic monitoring and optimized intelligent control of the magnesium oxide production process. The intelligent control system can accurately adjust key parameters such as temperature and negative pressure in the rotary kiln through a specific control method to ensure the production quality and efficiency of magnesium oxide.

[0037] The following is an introduction to the intelligent control method of the magnesium oxide rotary kiln and the intelligent control system of the magnesium oxide rotary kiln.

[0038] For the intelligent control method of magnesium oxide rotary kiln, please refer to Figure 1 , which comprises the following steps:

[0039] S1: Obtain the historical data of the controlled variables of the rotary kiln, and obtain the kiln output index Q1 and energy consumption index K1 through the model predictive control algorithm.

[0040] Among them, the yield index Q1 is obtained by constructing a BP neural network prediction model with self-optimizing parameters. The BP neural network prediction model constructs the objective function with maximum yield as the optimization goal, and obtains the optimized values ​​of the operating variables through genetic algorithm. The optimized values ​​of the operating variables are then increased according to timeliness to obtain the yield index Q1.

[0041] The energy consumption index K1 is obtained by constructing a BP neural network prediction model with self-optimizing parameters. The BP neural network prediction model constructs the objective function with minimum energy consumption as the optimization goal, and obtains the optimized value of the operating variable through the genetic algorithm. The optimized value of the operating variable is then reduced according to timeliness to obtain the energy consumption index K1.

[0042] Among them, the controlled variables in this embodiment include kiln tail NO x The operating variables include the concentration of CO at the kiln tail, the concentration of O2 at the kiln tail, the temperature at the kiln head, the temperature at the kiln tail, the negative pressure at the kiln head and the negative pressure at the kiln tail. The operating variables include the frequency of the kiln tail fan, the speed of the rotary kiln, the feed rate, the gas input rate and the coal powder input rate.

[0043] Since the output of magnesium oxide rotary kiln is affected by multiple process parameters, and multiple process parameters also affect each other and have a nonlinear relationship with the output, it is difficult to describe it using a simple model. In this application, BP neural network is used to model energy consumption, and then the initial weights and thresholds of BP neural network are optimized by genetic algorithm to obtain an improved BP neural network.

[0044] The database of the DCS system of the existing magnesium oxide rotary kiln can obtain the historical data of magnesium oxide production and key parameters related to production. These data are normalized separately, namely:

[0045]

[0046] Among them, y in formula (1) n is the normalized data value, x n is the original data value, x min is the minimum value of the data sequence, x max is the maximum value of the data series.

[0047] The initial value and threshold of the BP neural network are optimized by genetic algorithm, and the quality and parameters affecting the quality are used to establish the production model through the improved BP neural network. The specific process includes:

[0048] (1) Use the historical data of the operating variables as input to train the BP neural network and determine the BP neural network topology.

[0049] (2) Determine the individual length of the genetic algorithm based on the determined BP neural network topology and initialize the population.

[0050] (3) Determine the fitness function. First, obtain the initial weights and thresholds of the BP neural network through individual n. After training, obtain the predicted value output by the BP neural network. The sum of the difference between the predicted value and the actual value is multiplied by the coefficient K to obtain the fitness value A, that is:

[0051]

[0052] Wherein, m in formula (2) is the number of output nodes of BP neural network, o n is the predicted output of the nth node, y n is the expected output of the nth node of the BP neural network.

[0053] (4) Select the roulette wheel method as the selection strategy of the genetic algorithm, and the selection probability W of individual n n for:

[0054]

[0055] f n =a / F n (4)

[0056] In formula (3) and formula (4), F n is the fitness value of individual n, f n To find the inverse of the fitness value, N is the number of individuals in the population and a is the coefficient.

[0057] (5) Using the real number crossover method to perform the crossover operation, the kth chromosome β k and the cth chromosome β cThe crossover operation at position j is as follows:

[0058] β kj =β kn (1-b)+β cn b (5),

[0059] βcj=βcj(1-b)+βkjb (6),

[0060] In equations (5) and (6), b is a random number in the range [0, 1].

[0061] (6) Select the jth gene β of the nth individual nj , and perform corresponding mutation operations on it. The specific mutation operation methods are:

[0062]

[0063] f(g)=r2(1-g / G max ) (8),

[0064] In formula (7) and formula (8), β max and β min Gene β nj The upper and lower bounds of the , r is a random number in [0,1]. , r2 is a random number, g is the current iteration number, G max is the maximum number of evolutions.

[0065] (7) The optimal individual obtained through steps (1) to (6) is used as the initial weight and threshold assignment of the improved BP neural network. After the improved BP neural network is obtained, the yield model is established through the improved BP neural network.

[0066] It is understandable that the energy consumption model can also be established through the improved BP neural network in this embodiment.

[0067] The following describes the process of calculating the energy consumption index by constructing an energy consumption model through an improved BP neural network and a genetic algorithm. It includes: (11) Encoding the input of the improved BP neural network (i.e., the operating variable) through a real number encoding method to obtain a real number code, and use it as an individual of the population. (12) Since the genetic algorithm does not require a specific expression, the predicted value of the improved BP neural network can be directly used as the fitness value. (13) Select the roulette reading method as the selection operation, and use real number crossover for crossover operation, and then perform the corresponding mutation operation. (14) Finally, the optimized value of the operating variable can be found through the above steps. (15) Finally, the weight of the operating variable is reduced according to the timeliness to obtain the energy consumption index K1.

[0068] It can be understood that the yield index in this embodiment can also be constructed through the above-mentioned improved BP neural network with the maximum yield index as the optimization target to construct the objective function, and the optimized value of the operating variable is obtained by solving it through the genetic algorithm, and then the weight is increased according to the timeliness to obtain the yield index Q1.

[0069] S2: Collect the real-time data of the controlled variables of the rotary kiln, and obtain the output correction coefficient Q2 and energy consumption correction coefficient K2 of the kiln condition through fuzzy algorithm processing based on the real-time data of the controlled variables.

[0070] Among them, the production correction coefficient Q2 can be obtained through the following steps: (1.1) Determine the basic domain of the controlled variable, and quantize the basic domain of the controlled variable to obtain the fuzzy domain of the controlled variable. (1.2) Fuzzify the fuzzy domain of the controlled variable by the maximum membership method to obtain the fuzzy set of the controlled variable. (1.3) Establish fuzzy rules and perform fuzzy reasoning, and then use the centroid method to defuzzify the fuzzy set of the controlled variables to obtain the precise set of the controlled variables, and normalize the precise set of the controlled variables. (1.4) Use the gradient descent method to obtain the production correction coefficient Q2 from the precise set of the normalized controlled variables.

[0071] Among them, the energy consumption correction coefficient K2 can be obtained through the following steps: (2.1) Determine the basic domain of the controlled variable, and quantize the basic domain of the controlled variable to obtain the fuzzy domain of the controlled variable. (2.2) Fuzzify the fuzzy domain of the controlled variable by the maximum membership method to obtain the fuzzy set of the controlled variable. (2.3) Establish fuzzy rules and perform fuzzy reasoning, then use the centroid method to defuzzify the fuzzy set of the controlled variable to obtain the precise set of the controlled variables, and normalize the precise set of the controlled variables. (2.4) After the normalized precise set of the controlled variables is normalized, the energy consumption correction coefficient K2 is obtained by the LSTM algorithm.

[0072] In this step, a fuzzy model is selected, which is a rule-based modeling method with the ability to process uncertain information and is easy to model. It is suitable for industrial field objects with complex mechanisms, difficult mathematical models to obtain, dynamic characteristics that are difficult to grasp, or very significant changes.

[0073] S3: Correct the kiln condition output index Q1 according to the output correction coefficient Q2 to obtain the real-time kiln condition output index Q3. Correct the kiln condition energy consumption index K1 according to the energy consumption correction coefficient K2 to obtain the real-time kiln condition energy consumption index K3.

[0074] Specifically, the real-time production index Q3 of the kiln condition is obtained by multiplying the production index Q1 of the kiln condition by the production correction coefficient Q2; the real-time energy consumption correction coefficient K3 of the kiln condition is obtained by dividing the energy consumption index K1 of the kiln condition by the energy consumption correction coefficient K2.

[0075] S4: Collecting real-time data of rotary kiln operating variables, and obtaining the optimal combination of operating variables according to the real-time data of rotary kiln operating variables, the real-time output index Q3 of kiln conditions, and the real-time energy consumption index K3 of kiln conditions through a finite time domain rolling optimization strategy.

[0076] In this embodiment, the limited time domain rolling optimization strategy includes the following steps:

[0077] (41) Based on the ARX prediction model, an optimization model for optimizing the operating variables in the magnesium oxide rotary kiln is established; (42) Based on the input and output of the magnesium oxide rotary kiln at the current moment and the past moment, the predicted output value of the operating variables of the magnesium oxide rotary kiln is calculated through the optimization model, and then the output index and energy consumption index are used as constraints to obtain the objective function containing the future predicted input increment; (43) The predicted value of the operating variable is obtained by using the least squares method, and the predicted value is input into the optimization model to obtain the actual output value of the operating variable of the magnesium oxide rotary kiln; (44) After calculating the output predicted value of the operating variable at the next moment through the optimization model, the solution is repeated in a rolling loop to obtain the optimal combination of the operating variables.

[0078] Among them, the ARX forecasting model is a dynamic model, which is a time series analysis method with the advantages of simple calculation and small amount of calculation. It can be described by the following linear difference equation:

[0079]

[0080] In formula (9), μ(t) is the input at time t; y(t) is the output at time t, t = 1, 2, 3…, N, N is the number of samples. s and β i are the estimated parameters of the ARX prediction model. s is the number of outputs, s = 1, 2, ..., e. β i Estimated parameters of the ARX model, i is the number of input quantities; i = 1, 2, …, d.

[0081] The form of the incremental relationship can be described as:

[0082]

[0083] In formula (10), Δy(t) is the output increment, Δy(t)=y(t)-y(t-1). Δμ(t) is the input increment, Δμ(t)=μ(t)-μ(t-1).

[0084] In order to make the expression of formula (10) more concise, we define:

[0085] A(q)Δy(t)=B(q)Δμ (t) (11),

[0086] A(q)=1-α1q -1 -...+α e q -e (12)

[0087] B(q)=β1q -1 +...+β d q -d (13)

[0088] Wherein, formula (11) is the ARX model describing the incremental relationship between the input and output of the system, AR refers to the autoregressive part A(q)Δy(t), X refers to the external input B(q)Δμ(t), and q is the total output.

[0089] When the system is stimulated by a pseudo-random binary signal, the output of the system is measured. After obtaining the system input and output signal increment sequence {Δμ(t)}, {Δy(t)}, t = 1, 2, 3..., N, the least squares method can be used to estimate the parameters of the ARX model. Therefore, equation (10) can be rewritten as:

[0090]

[0091] In formula (14),

[0092] [Δμ(t-1),...,Δμ(td),Δy(t-1),...,Δy(te)],

[0093] θ=[β1 ,... ,βd ,α 1 ,...,α e]T,

[0094] Using the least squares method for identification, we can get:

[0095]

[0096] Among them, θ is the parameter vector. It is a matrix. is the least squares estimate of the parameter vector θ.

[0097] S5: The optimal combination of operating variables is obtained as the input of the intelligent control system of magnesium oxide to realize intelligent control of the rotary kiln.

[0098] The solution of this application uses the above-mentioned intelligent control method, which establishes an intelligent control system that simulates the operator's thinking, and then determines the target value of key production parameters in the intelligent control system through the product quality situation in the production process. It can effectively and intelligently control the magnesium oxide rotary kiln. Through the system to analyze and feedback the kiln conditions of the rotary kiln, it can automatically adjust the production parameters, reduce energy consumption, reduce human errors, and improve the consistency and stability of the product.

[0099] Example 2

[0100] The present application also provides an intelligent control system based on the intelligent control method of the magnesium oxide rotary kiln in Example 1, please refer to Figure 2 and Figure 3 , which includes a rotary kiln, a data acquisition module and a control module. The rotary kiln is used to calcine magnesium oxide, and the data acquisition module is used to collect data in the rotary kiln and transmit the collected data to the control module. In this embodiment, the data acquisition module can collect the following data: the feed amount of the rotary kiln, the frequency of the fan at the kiln tail, the input amount of the gas valve, the input amount of coal powder, the NO at the kiln tail x The concentration of CO at the kiln tail, the concentration of O2 at the kiln tail, the temperature at the kiln head, the temperature at the kiln tail, the flame temperature, the kiln current, the negative pressure at the kiln head, the negative pressure at the kiln tail, etc. Among them, the temperature at the kiln head and the temperature at the kiln tail can be collected by the temperature sensor. The negative pressure at the kiln head and the negative pressure at the kiln tail can be collected by the pressure sensor.

[0101] The control module includes a feedback adjustment module and a real-time optimization module. The real-time optimization module can optimize the data input by the data acquisition module through the least squares method, and transmit the calculation results to the feedback adjustment module. The feedback adjustment module uses the calculation results of the real-time optimization module as the adjustment target to realize intelligent adjustment of the rotary kiln. Through this adjustment, the intelligent control system can adjust the rotary kiln in real time, and at the same time, it will optimize the parameters that need to be adjusted according to the specific conditions in the kiln, so as to ensure the stable operation of the rotary kiln, and at the same time, it can reduce energy consumption while improving the consistency and stability of the product.

[0102] It can be understood that the feedback regulation module can be an APC advanced control module, and the real-time optimization module can be an RTO optimization decision module.

[0103] In addition, when communication interruption, data anomaly or other faults occur in the intelligent control system of this embodiment, it will exit automatic control and prompt the operator with relevant fault information through real-time voice alarm so that the operator can intervene and adjust the operation to ensure the safe operation of the intelligent control system and the rotary kiln.

[0104] It is understandable that the installation and modulation of the intelligent control system in this embodiment are completed in the same server computer. And the intelligent control system is independent of the original DCS system of the rotary kiln and will not affect the original DCS system. It is connected to the rotary kiln by means of an external plug-in, which not only ensures the data security of the DCS system, but also realizes the intelligent control of rotary kilns with different production processes.

[0105] The operation interface of the intelligent control system can be organically combined with the operation interface of the existing DCS, making the control operation of the rotary kiln more convenient. Each controller module can be put into use and cut out separately on the operation interface.

[0106] The software structure of the intelligent control system is global and can meet the requirements of metal data docking of different DCS systems in the factory. It can meet the requirements of two-way data communication between different DCS systems (AB, ABB, Siemens, Yokogawa, Schneider, Hollysys, SUPCON, etc.) and the intelligent control system.

[0107] In this intelligent control system, multiple controllers are also constructed based on the above control method. Figure 3 , which mainly includes kiln condition controller, kiln negative pressure controller, output optimization controller, kiln speed optimization controller and kiln head combustion controller.

[0108] For the kiln condition controller, it collects NO at the kiln tail. x , flame temperature, kiln tail temperature, kiln tail O2 content and concentration, kiln current and other parameters, and then use the designed intelligent control system to judge the condition of the kiln. It is understandable that the kiln condition controller can output the fuzzy concept that originally characterizes the quality of the kiln condition as a digital quantity through the intelligent control system. For example, 0 can be used to represent the standard kiln condition, which is the lowest energy consumption point under qualified quality. Above 0 points means the kiln condition is good, which means that production can be increased or fuel supply can be slightly reduced. Below 0 points means the kiln condition is poor, which means that fuel needs to be increased or production needs to be reduced. Therefore, this embodiment can estimate and judge the real-time situation of the kiln condition through the kiln condition controller, and then adjust the kiln condition in real time by adjusting the output or the kiln head combustion controller, so as to achieve stable operation in the kiln condition.

[0109] As for the kiln negative pressure controller, since the speed of the kiln tail fan will affect the flame retardant air content in the rotary kiln (mainly affecting the O2 content in the kiln), the speed of the kiln tail fan is often set to a larger value during manual operation, which will not only lead to a large power consumption of the kiln tail fan, but also cause more heat loss in the rotary kiln. Therefore, in this application, a kiln negative pressure controller is designed by setting, which can be controlled by the intelligent control system of this embodiment. The intelligent control system can optimize the data in the rotary kiln and then input the optimized adjustment parameters to the kiln negative pressure controller. The kiln negative pressure controller can feedback and adjust the speed of the kiln tail fan by receiving the data transmitted by the intelligent control system to ensure that the kiln head is in a slightly negative pressure, the O2 content at the kiln tail is in a reasonable range of 3%-5%, and the combustion-supporting air in the rotary kiln is in the optimal amount. In this way, the problem of excessive heat loss in the rotary kiln caused by excessive combustion-supporting air can be avoided while ensuring that the fuel is fully burned, thereby improving the utilization rate of the fuel and reducing the energy consumption of the rotary kiln.

[0110] It can be understood that, according to the current status of the existing rotary kiln, the kiln negative pressure controller in this real-time can be designed as a controller that can control the negative pressure at the kiln head and the negative pressure at the kiln tail. For example, in the normal production process, we can give priority to the negative pressure at the kiln tail as the reference target of the kiln negative pressure controller. When the negative pressure at the kiln tail fails, the intelligent control system can automatically switch the kiln negative pressure controller to a state where the negative pressure at the kiln head is controlled. And when the negative pressure at the kiln tail returns to normal, the intelligent control system can switch the kiln negative pressure controller to a state where the negative pressure at the kiln tail is controlled. This adjustment can avoid the situation where the kiln negative pressure controller is unable to regulate the combustion-supporting air in the rotary kiln when the negative pressure at the kiln tail fails.

[0111] For the output optimization controller, its design idea can refer to the following description: When the intelligent control system adjusts the kiln head coal feeding amount and still cannot meet the requirements of the rotary kiln condition, the operator in the prior art will appropriately increase or decrease the feeding amount. When the rotary kiln is operating normally, if the feeding amount is suddenly increased or decreased, it will cause a large fluctuation in the entire rotary kiln system, so that the intelligent control system needs a long time of feedback adjustment to adapt to this fluctuation, which will lead to the poor quality of magnesium oxide produced in the rotary kiln. Therefore, we can set a production optimization controller in the intelligent control system, which can analyze and optimize the kiln condition of the rotary kiln, the content and concentration of O2 at the kiln tail, the kiln head temperature and the kiln tail temperature through the intelligent control system, and obtain an optimized feeding amount. Then the production optimization controller is used to slowly adjust the feeding amount, so that the rotary kiln can smoothly transition from one thermal balance to another, and the intelligent control system can be stably operated while improving product quality.

[0112] As for the kiln speed optimization controller, in this embodiment, the intelligent control system can be used to calculate the feed speed ratio between the feed amount and the rotation speed, and at the same time monitor the temperature, pressure and other data in the rotary kiln in real time, and then the control system processes and analyzes the above data and outputs an optimized kiln speed data. The kiln speed optimization controller is then used to adjust the speed of the rotary kiln so that it reaches the optimized kiln speed data. The above settings can automatically adjust the speed of the rotary kiln and accurately control the speed of the rotary kiln. This adjustment process is carried out in full consideration of the kiln conditions, the feed speed ratio and other parameters, making the entire adjustment more intelligent. In addition, the intelligent control system of this embodiment can also dynamically adjust the speed of the rotary kiln according to different production needs to achieve the purpose of optimizing the production process and reducing energy consumption.

[0113] As for the kiln head combustion controller, in the prior art, the operator generally adjusts the gas consumption to keep the kiln tail temperature and the kiln head temperature within the set range. And when the adjustment of the gas consumption cannot meet the set requirements for the kiln tail temperature and the kiln head temperature, the operator will increase the coal powder to make the kiln head temperature and the kiln tail temperature meet the system setting requirements. However, the above-mentioned adjustment method has a certain hysteresis, and this manual adjustment method has a large workload. The adjustment process is also prone to cause large fluctuations in the kiln tail temperature, thereby affecting the quality of the product in the rotary kiln. This embodiment designs a kiln head combustion controller to control the gas valve opening and the coal powder scale opening respectively to adjust the kiln head temperature and the kiln tail temperature. Its adjustment can refer to the following operations: the data acquisition module obtains the data of kiln head temperature, kiln tail temperature, feed amount, and kiln tail O2 content and concentration, and analyzes and processes these data to obtain a set of optimized gas valve opening angle and coal powder input amount data. The intelligent control system can control the gas valve opening and coal powder opening respectively through the kiln head combustion controller to achieve the data optimized by the intelligent control system. This adjustment method can not only control the input amount of gas and coal powder, but also ensure that the kiln head temperature and kiln tail temperature are always within the set range of the intelligent control system, so that the rotary kiln is always in a good combustion condition, thereby reducing fuel loss and improving product quality.

[0114] By introducing the intelligent control system of this embodiment into the magnesium oxide rotary kiln, the magnesium oxide rotary kiln production line can move smoothly, reduce the fluctuation of process parameters and product quality during the production process, and reduce the volatility of important process parameters in the magnesium oxide rotary kiln by about 30%.

[0115] By introducing the intelligent control system of this embodiment into the magnesium oxide rotary kiln, standardized operation of the magnesium oxide rotary kiln can be established, the workload of operating staff can be reduced, and the rotary kiln can operate smoothly for a long time, thereby reducing energy consumption while ensuring product quality.

[0116] Example 3

[0117] An intelligent control device for a magnesium oxide rotary kiln includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, an intelligent control system for a magnesium oxide rotary kiln as in Example 2 is created, thereby realizing intelligent regulation of the magnesium oxide rotary kiln.

[0118] The intelligent control system of Example 2 is running in the intelligent control device in this embodiment. The intelligent control device in this embodiment can be an embedded device integrated into the back end of the imaging device. It can also be an independent device independent of the imaging device. Specifically, the intelligent control device in this embodiment can be an intelligent terminal, a tablet computer, a laptop computer, a desktop computer, a rack server, a blade server, a tower server or a cabinet server (including an independent server, or a server cluster composed of multiple servers) that can execute programs.

[0119] The intelligent control device pointed out in this embodiment at least includes but is not limited to: a memory and a processor that can be connected to each other through a system bus. Among them, the memory (i.e., a readable storage medium) includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory can be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the memory can also be an external storage device of a computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. Of course, the memory can also include both the internal storage unit of the computer device and its external storage device. In this embodiment, the memory is generally used to store an operating system and various application software installed on the computer device. In addition, the memory can also be used to temporarily store various types of data that have been output or are to be output.

[0120] The processor may be a central processing unit (CPU), a graphics processing unit (GPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor is generally used to control the overall operation of a computer device. In this embodiment, the processor is used to run program codes stored in a memory or process data.

[0121] The above describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention to be protected. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. An intelligent control method for a magnesium oxide rotary kiln, characterized in that: It includes the following steps: S1: Obtain the historical data of the controlled variables of the rotary kiln, and obtain the kiln output index Q1 and energy consumption index K1 through the model predictive control algorithm; S2: Collect the real-time data of the controlled variables of the rotary kiln, and obtain the output correction coefficient Q2 and energy consumption correction coefficient K2 of the kiln condition through fuzzy algorithm processing; S3: Correct the kiln condition output index Q1 according to the output correction coefficient Q2 to obtain the real-time kiln condition output index Q3; Correct the kiln condition energy consumption index K1 according to the energy consumption correction coefficient K2 to obtain the real-time kiln condition energy consumption index K3; S4: collecting real-time data of rotary kiln operating variables, and obtaining the optimal combination of operating variables through a finite time domain rolling optimization strategy according to the real-time data of rotary kiln operating variables, the real-time output index Q3 of kiln conditions, and the real-time energy consumption index K3 of kiln conditions; The finite horizon rolling optimization strategy includes the following steps: (1) An optimization model for optimizing the operating variables in the magnesium oxide rotary kiln is established based on the ARX prediction model; (2) Based on the input and output of the magnesium oxide rotary kiln at the current moment and the past moment, the predicted output value of the operating variables of the magnesium oxide rotary kiln is calculated through the optimization model, and then the output index and energy consumption index are used as constraints to obtain the objective function containing the future predicted input increment; (3) The predicted value of the operating variable is obtained by using the least squares method, and the predicted value is input into the optimization model to obtain the actual output value of the operating variable of the magnesium oxide rotary kiln; (4) Calculating the output prediction value of the operating variables at the next moment through the optimization model, and solving it in a rolling cycle in sequence to obtain the optimal combination of the operating variables; S5: The optimal combination of operating variables is used as the input of the intelligent control system of magnesium oxide to realize intelligent control of the rotary kiln.

2. The intelligent control method of magnesium oxide rotary kiln according to claim 1, characterized in that: The expression of the ARX prediction model is: Where μ(t) is the input at time t; y(t) is the output at time t, t = 1, 2, 3…, N, N is the number of samples; α s and β i are all estimated parameters of the ARX prediction model; s is the number of output quantities, s = 1, 2, ..., e; i is the number of input quantities, i = 1, 2, ..., d.

3. The intelligent control method of magnesium oxide rotary kiln according to claim 1, characterized in that: The energy consumption index K1 is obtained by constructing a BP neural network prediction model with self-optimization parameters. The BP neural network prediction model constructs an objective function with the minimum energy consumption index as the optimization target, and obtains the optimal value of the operating variable by solving the genetic algorithm, and then reduces its weight according to timeliness to obtain the energy consumption index K1; And / or, the yield index Q1 is obtained by constructing a BP neural network prediction model with self-optimizing parameters. The BP neural network prediction model constructs an objective function with the maximum yield index as the optimization target, and obtains the optimized value of the operating variable through a genetic algorithm, and then increases its weight according to timeliness to obtain the yield index Q1.

4. The intelligent control method of magnesium oxide rotary kiln according to claim 1, characterized in that: The fuzzy algorithm includes the following processing steps: (11) Determine the basic domain of the controlled variable, and obtain the fuzzy domain of the controlled variable after quantifying the basic domain of the controlled variable; (12) Fuzzifying the fuzzy domain of the controlled variables by the maximum membership method to obtain the fuzzy set of the controlled variables; (13) Establish fuzzy rules and perform fuzzy reasoning, then use the centroid method to defuzzify the fuzzy set of controlled variables to obtain the precise set of controlled variables, and then normalize the precise set of controlled variables; (14) The precise set of normalized controlled variables is used to obtain the production correction coefficient Q2 through the gradient descent method; the precise set of normalized controlled variables is used to obtain the energy consumption correction coefficient K2 through the LSTM algorithm.

5. The intelligent control method of magnesium oxide rotary kiln according to claim 1, characterized in that: The real-time production index Q3 of the kiln condition is obtained by multiplying the production index Q1 of the kiln condition by the production correction coefficient Q2; the real-time energy consumption correction coefficient K3 of the kiln condition is obtained by dividing the energy consumption index K1 of the kiln condition by the energy consumption correction coefficient K2.

6. The intelligent control method of magnesium oxide rotary kiln according to claim 1, characterized in that: The operating variables include the frequency of the kiln tail fan, the rotating speed of the rotary kiln, the feed rate, the gas input rate and the coal powder input rate.

7. The intelligent control method of magnesium oxide rotary kiln according to claim 1, characterized in that: The controlled variables include kiln tail NO x The concentration of CO at the kiln tail, the concentration of O2 at the kiln tail, the temperature at the kiln head, the temperature at the kiln tail, the negative pressure at the kiln head and the negative pressure at the kiln tail.

8. An intelligent control system for a magnesium oxide rotary kiln, characterized in that: It is operated by the intelligent control method of the magnesium oxide rotary kiln as claimed in any one of claims 1 to 7; the intelligent control system comprises: a data acquisition module and a control module; The input end of the data acquisition module is connected to the rotary kiln, and the output end is connected to the control module; the data acquisition module is used to obtain the process parameter data of the rotary kiln and transmit the obtained data to the control module; The control module includes a feedback regulation module and a real-time optimization module. The real-time optimization module optimizes and calculates the data input by the data acquisition module through the least squares method and transmits the calculation result to the feedback regulation module. The feedback regulation module uses the calculation result of the real-time optimization module as the control target to realize intelligent control of the rotary kiln.

9. The intelligent control system of the magnesium oxide rotary kiln according to claim 8, characterized in that: The process parameters include the frequency of the kiln tail fan, the speed of the rotary kiln, the feed rate, the gas input, the coal powder input, the kiln tail NO x The concentration of CO at the kiln tail, the concentration of O2 at the kiln tail, the temperature at the kiln head, the temperature at the kiln tail, the negative pressure at the kiln head and the negative pressure at the kiln tail.

10. An intelligent control device for a magnesium oxide rotary kiln, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the intelligent control system for the magnesium oxide rotary kiln as claimed in claim 8 or 9 is created, thereby realizing intelligent regulation of the magnesium oxide rotary kiln.

Images