Intelligent smelting furnace temperature control system based on neural network

Abstract

The invention discloses a smelting furnace temperature intelligent control system and method based on a neural network, and the system comprises a parameter controller and an intelligent processor, the parameter controller is used for obtaining a current furnace temperature parameter and a current combustion control parameter from a smelting furnace, and generating a target furnace temperature parameter; the intelligent processor is used for receiving the current furnace temperature parameter, the current combustion control parameter and the target furnace temperature parameter sent by the parameter controller, generating a combustion adjustment parameter according to a neural network model, and sending the combustion adjustment parameter to the intelligent processor; and the combustion adjusting parameters are sent to the parameter controller. The furnace temperature which can be reached in the future time period can be predicted according to the current furnace temperature parameter and the combustion control parameter, so that the combustion adjustment parameter is optimized, the furnace temperature can be efficiently controlled in an energy-saving manner, and the oxidation burning loss of molten metal is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of adaptive control systems for industrial equipment, and particularly relates to a system and method for intelligently controlling the furnace temperature of a smelting furnace by using a neural network algorithm. The present invention is particularly applicable to regenerative aluminum smelting furnaces fueled by gas or oil. Background Art

[0002] For an increasing number of complex control objects, on the one hand, the required control performance is no longer simply limited to one or two indicators; on the other hand, various existing optimization methods are based on the premise that the optimization problem has an accurate mathematical model. However, it is difficult or impossible to obtain an accurate mathematical model for many actual engineering problems. This has restricted the practical application of existing optimization methods.

[0003] With the development of intelligent technologies such as fuzzy theory and neural networks, as well as computer technology, intelligent adaptive optimization methods have received attention and development. The research on artificial neural networks originated from the work of McCulloch and Pitts in 1943. In terms of optimization, in 1982, Hopfield first introduced the Lyapunov energy function to judge the stability of the network and proposed the Hopfield single-layer discrete model; Hopfield and Tank further developed the Hopfield single-layer continuous model. In 1986, Hopfield and Tank directly corresponded the electronic circuit with the Hopfield model to achieve hardware simulation; Kennedy and Chua proposed an analog circuit model based on nonlinear circuit theory and studied the stability of the electronic circuit using the Lyapunov function of the system differential equation. These works have strongly promoted the research on neural network optimization methods.

[0004] According to neural network theory, the minimum point of the neural network energy function corresponds to the stable equilibrium point of the system. Thus, the solution of the minimum point of the energy function is transformed into the solution of the stable equilibrium point of the system. As time evolves, the motion trajectory of the network always moves in the direction of decreasing energy function in space and finally reaches the equilibrium point of the system - that is, the minimum point of the energy function. Therefore, if the stable attractor of the neural network dynamic system is considered as the minimum point of an appropriate energy function (or augmented energy function), the optimization calculation starts from an initial point and reaches a certain minimum point along the system flow. If the concept of global optimization is applied to the control system, the objective function of the control system will ultimately reach the desired minimum point. This is the basic principle of neural optimization calculation.

[0005] The algorithm based on neural network can be used to solve the problems of hysteresis and poor accuracy existing in the control systems of actual industrial equipment. For example, Figure 1Displays the parameter setting interface of a monitoring terminal of a control system for a traditional aluminum melting furnace. This interface is used for users to monitor the status of the melting furnace and set the control parameters of the melting furnace. The melting furnace is a gas regenerative melting furnace, which has three burners (also called main guns), namely Burner No. 1, Burner No. 2, and Burner No. 3. Each burner sprays air and fuel into the furnace chamber during combustion and is used to discharge waste gas when not burning. It can be seen from this interface that the current control system only controls the combustion of the burners based on the judgment of whether the current value reaches a predetermined value, and each burner can only be set to two working modes, namely "high-fire setting" and "low-fire setting", and the settings of the sizes of the combustion air fan valve, fuel valve, and exhaust air fan valve are fixed, and the setting of the burner fuel change time is also fixed.

[0006] Therefore, the prior art cannot fully rely on the direct measurement of sensors and fixed control modes for the temperature control of the furnace chamber and hearth of the melting furnace. The control system lacks predictability for the furnace temperature and cannot perform more precise temperature control. On the one hand, this will lead to an increase in the oxidation and burning loss of the metal, and on the other hand, it will increase the energy consumption and cannot achieve the purpose of energy conservation and emission reduction. Summary of the Invention

[0007] (1) Technical problems to be solved The technical problem to be solved by the present invention is that the existing control system of the melting furnace cannot predict the influence of combustion parameters on the furnace temperature, so it cannot accurately control the furnace temperature, resulting in an increase in the oxidation and burning loss of the molten metal and an increase in energy consumption.

[0008] (2) Technical solutions To solve the above technical problems, on the one hand, the present invention proposes an intelligent control system for the furnace temperature of a melting furnace based on a neural network. The melting furnace is heated by fuel combustion and includes a furnace chamber and burners. The intelligent control system includes: a parameter controller for obtaining the current furnace temperature parameters and current combustion control parameters from the melting furnace, generating target furnace temperature parameters, and sending the current furnace temperature parameters, current combustion control parameters, and target furnace temperature parameters to an intelligent processor, and controlling the combustion parameters of the currently burning burners according to the combustion adjustment parameters obtained from the intelligent processor; an intelligent processor for receiving the current furnace temperature parameters, current combustion control parameters, and target furnace temperature parameters sent by the parameter controller, generating combustion adjustment parameters according to the neural network model, and sending the combustion adjustment parameters to the parameter controller.

[0009] According to a preferred embodiment of the present invention, the neural network model is trained on a training dataset, which is composed of multiple pieces of data obtained during the historical smelting production of the smelting furnace. Each piece of data therein includes input parameters and output parameters; the input parameters include the pre-regulation furnace temperature parameter, the pre-regulation combustion control parameter, and the post-regulation furnace temperature parameter, and the output parameter includes the combustion adjustment parameter; the trained neural network model inputs the obtained current furnace temperature parameter, current combustion control parameter, and target furnace temperature parameter as the pre-regulation furnace temperature parameter, pre-regulation combustion control parameter, and post-regulation furnace temperature parameter respectively, so as to output the combustion adjustment parameter.

[0010] According to a preferred embodiment of the present invention, the current furnace temperature parameter includes the current furnace chamber temperature and the current regenerator bed temperature; the current combustion control parameter includes the fuel flow rate, the combustion-supporting gas flow rate, and the exhaust gas flow rate; the combustion adjustment parameter includes the change amount of the fuel flow rate, the change amount of the combustion-supporting gas flow rate, and the change amount of the exhaust gas flow rate.

[0011] According to a preferred embodiment of the present invention, there are at least two burners, and the combustion adjustment parameter further includes whether to switch the combustion state of the burners.

[0012] According to a preferred embodiment of the present invention, the current combustion control parameter further includes the current burner angle, and the combustion adjustment parameter further includes the adjustment amount of the burner angle.

[0013] According to a preferred embodiment of the present invention, the smelting furnace further includes a fuel pipe, a combustion-supporting air pipe, and an exhaust pipe corresponding to the burners, as well as a fuel valve for controlling the fuel flow rate of the fuel pipe, a combustion-supporting air blower for providing combustion-supporting air, and an exhaust fan for exhausting gas; the parameter controller controls the burners to perform combustion switching according to the combustion adjustment parameter obtained from the intelligent controller, calculates the opening degree of the fuel valve corresponding to the burning burner, the rotation speeds of the combustion-supporting air blower and the exhaust fan according to the combustion adjustment parameter, and sends control signals to the fuel valve, the combustion-supporting air blower, and the exhaust fan corresponding to the burning burner.

[0014] On the other hand, the present invention proposes an intelligent control method for the furnace temperature of a smelting furnace based on a neural network. The smelting furnace uses fuel combustion for heating and includes a furnace chamber and burners. The intelligent control method includes: obtaining the current furnace temperature parameter and the current combustion control parameter, and generating a target furnace temperature parameter; generating a combustion adjustment parameter according to the current furnace temperature parameter, the current combustion control parameter, and the target furnace temperature parameter, and based on the neural network model; controlling the combustion parameters of the currently burning burners through the combustion adjustment parameter.

[0015] According to a preferred embodiment of the present invention, the neural network model is trained on a training data set, which is composed of multiple pieces of data obtained during the historical smelting production of the smelting furnace. Each piece of data includes input parameters and output parameters. The input parameters include the pre-adjustment furnace temperature parameter, the pre-adjustment combustion control parameter, and the post-adjustment furnace temperature parameter, and the output parameter includes the combustion adjustment parameter. The trained neural network model inputs the obtained current furnace temperature parameter, current combustion control parameter, and target furnace temperature parameter as the pre-adjustment furnace temperature parameter, pre-adjustment combustion control parameter, and post-adjustment furnace temperature parameter respectively, so as to output the combustion adjustment parameter.

[0016] According to a preferred embodiment of the present invention, the current furnace temperature parameter includes the current furnace chamber temperature and the current regenerator bed temperature; the current combustion control parameter includes the fuel flow rate, the combustion-supporting gas flow rate, and the exhaust gas flow rate; the combustion adjustment parameter includes the change amount of the fuel flow rate, the change amount of the combustion-supporting gas flow rate, and the change amount of the exhaust gas flow rate.

[0017] According to a preferred embodiment of the present invention, there are at least two burners, and the combustion adjustment parameter further includes whether to switch the combustion state of the burners.

[0018] According to a preferred embodiment of the present invention, the current combustion control parameter further includes the current burner angle, and the combustion adjustment parameter further includes the adjustment amount of the burner angle.

[0019] (III) Beneficial effects The present invention can predict the furnace temperature that can be reached in the future time period according to the current furnace temperature parameter and combustion control parameter, so as to optimize the combustion adjustment parameter. Therefore, it can efficiently and energy-savingly control the furnace temperature and reduce the oxidation and burning loss of molten metal. Brief description of the drawings

[0020] Figure 1 is the parameter setting interface of a monitoring terminal of the control system of a traditional aluminum smelting furnace, and this interface is used to monitor the state of the smelting furnace and set the control parameters of the smelting furnace.

[0021] Figure 2 is the schematic structural diagram of the intelligent control system for the furnace temperature of the smelting furnace based on neural network of the present invention.

[0022] Figure 3 is the schematic module structure diagram of the parameter controller of the intelligent control system for the furnace temperature of the smelting furnace based on neural network of the present invention.

[0023] Figure 4 is the schematic module structure diagram of the intelligent processor of the intelligent control system for the furnace temperature of the smelting furnace based on neural network of the present invention.

[0024] Figure 5It is a schematic diagram of an embodiment of the display interface of the monitoring terminal of the intelligent control system for the furnace temperature of the smelting furnace based on neural network according to the present invention.

[0025] Figure 6 It is a schematic diagram of the neural network of the intelligent control system for the furnace temperature of the smelting furnace based on neural network according to the present invention.

[0026] Figure 7 It is a flow chart of the intelligent control method for the furnace temperature of the smelting furnace based on neural network according to the present invention. Detailed implementation manners

[0027] To solve the above technical problems, the present invention proposes an intelligent control system for the furnace temperature of the smelting furnace based on neural network. The smelting furnace of the present invention generally refers to a flame furnace, that is, the smelting furnace uses fuel combustion for heating. The present invention is not limited to the type of fuel used, and fuel oil, gas, coal gas, etc. are all acceptable. In addition, although the present invention is not limited to the type of smelting furnace, the present invention is more preferably a regenerative smelting furnace with better energy saving.

[0028] Although the present invention is described with an aluminum smelting furnace as a specific embodiment, the present invention can also be used for smelting furnaces of other non-ferrous metals or other types of smelting furnaces with similar combustion control.

[0029] The present invention is not limited to the type of smelting furnace. Therefore, the structure of the applicable smelting furnace can be different, but it should at least include a hearth for containing molten metal and serving as a combustion chamber, and at least two regenerative beds and burners corresponding to the regenerative beds. The burners of the smelting furnace usually appear in pairs, but may also be other numbers. The present invention is not limited to the number of burners, but excludes the case of only one burner.

[0030] During the smelting process, the present invention is described by taking the combustion of one of the burners as an example. However, the present invention can actually be extended to the case of multiple burners burning. As long as there are other burners for exhaust without combustion, the present invention can generally be applied in principle, although the present invention does not tend to this.

[0031] Figure 2 It is a schematic diagram of the architecture of the intelligent control system for the furnace temperature of the smelting furnace based on neural network according to the present invention. The following parameters Figure 2 are described as follows.

[0032] The smelting furnace of the present invention includes a fuel pipe, a combustion air pipe and an exhaust pipe corresponding to each burner, as well as a fuel valve for controlling the fuel pipe flow rate, a combustion air blower for providing combustion air and an exhaust blower for exhausting gas. The combustion air blower for providing combustion air and the exhaust blower for exhausting gas can be shared by multiple burners, or independent combustion air blowers and exhaust blowers can be configured for different burners. The present invention is not limited to the number and connection mode of the combustion air blower and the exhaust blower.

[0033] As a specific implementation, the smelting furnace further includes a furnace temperature sensor and a regenerator temperature sensor. The furnace temperature sensor is used to detect the temperature of the furnace gas, and the regenerator temperature sensor is used to detect the temperature of the regenerator. In addition, the smelting furnace of the present invention also has an ignition mechanism and a corresponding ignition control system, as well as safety alarm sensor devices, etc., but they are not closely related to the innovative points of the present invention, so no detailed description will be given here.

[0034] Different from the prior art, the intelligent control system of the present invention includes a parameter controller and an intelligent controller. The parameter controller is similar to the PLC controller in the traditional control system, and the intelligent processor is a key component newly added in the present invention, which works in cooperation with the parameter controller to achieve the intelligent control of the present invention.

[0035] The parameter controller of the present invention is used to obtain the current furnace temperature parameter and the current combustion control parameter from the smelting furnace. At the same time, it is also used to generate a target furnace temperature parameter. Different from the prior art, after obtaining these parameters, the parameter controller does not process them, but sends the current furnace temperature parameter, the current combustion control parameter, and the target furnace temperature parameter to the intelligent processor for processing.

[0036] The current furnace temperature parameter can be directly generated by temperature sensors such as the furnace temperature sensor and the regenerator temperature sensor and sent to the parameter controller. The target furnace temperature parameter can be a predetermined default value or can be determined by user input. When determined by user input, the parameter controller can have a user input module for receiving user input information.

[0037] The current combustion control parameter generally refers to the parameters related to the combustion state of the burners of the smelting furnace, such as the opening degree of the fuel valve controlling the fuel pipe flow rate, the rotational speed of the combustion air blower for providing combustion air, and the rotational speed of the exhaust blower for exhausting gas, etc. These parameters can be directly obtained from these devices or components.

[0038] The parameter controller is also used to control the combustion parameters of the burners of the current combustion according to the combustion adjustment parameters obtained from the intelligent processor. The combustion adjustment parameter is the parameter change mode and / or change amount sent by the intelligent processor regarding the control of the combustion state of the burners of the smelting furnace. After receiving the combustion adjustment parameter, the intelligent processor converts it to generate a control signal and sends it to the corresponding mechanisms of the smelting furnace, such as the fuel valve controlling the fuel pipe flow rate, the combustion air blower for providing combustion air, and the exhaust blower for exhausting gas, to adjust the opening degree of the fuel valve, the rotational speeds of the combustion air blower and the exhaust blower, etc.

[0039] Figure 3 It is a schematic diagram of the module architecture of the parameter controller of the intelligent control system for the furnace temperature of the smelting furnace based on neural network of the present invention. As Figure 3As shown, the parameter controller of the present invention may include a user input module, a parameter transceiver module, and a parameter forwarding module. However, those skilled in the art should understand that although Figure 3 not shown in the figure, the actual parameter controller may also include other modules in addition to the above modules. For example, it may also include an output module for inputting to the terminal monitoring display for the user to view. The parameter controller may also have a storage module for recording and storing historical parameter information for subsequent use.

[0040] The user input module is used to receive the input information of the user to obtain the target furnace temperature parameter, so that the parameter controller will subsequently send the target furnace temperature parameter to the intelligent processor. The target furnace temperature parameter may include the target furnace chamber temperature and the target regeneration bed temperature. The target user input module generally refers to the user input interface, such as the interfaces of the keyboard and mouse, which can be connected to an external keyboard and / or mouse for the user to input.

[0041] The parameter transceiver module is connected to each temperature sensor, valve, blower, etc. of the smelting furnace, and is used to obtain the current furnace temperature parameter and the current combustion control parameter of the smelting furnace. As described above, the current furnace temperature parameter includes the current furnace chamber temperature and the current regeneration bed temperature, and the current combustion control parameter includes the fuel flow rate, the combustion-supporting gas flow rate, the exhaust gas flow rate, etc.

[0042] The parameter forwarding module is connected to the parameter transceiver module and the intelligent processor, and is used to send the various parameters obtained by the parameter transceiver module and the target furnace temperature parameter obtained by the user input module to the intelligent processor, and at the same time send the combustion adjustment parameter from the intelligent processor to the parameter transceiver module.

[0043] In specific implementation, the above parameter transceiver module and parameter forwarding module may be implemented by the same functional module or separately.

[0044] The intelligent processor of the present invention is a key component for implementing the neural network algorithm control. Generally speaking, it is used to receive the current furnace temperature parameter, the current combustion control parameter, and the target furnace temperature parameter sent by the parameter controller, generate the combustion adjustment parameter according to the neural network model, and send the combustion adjustment parameter to the parameter controller.

[0045] Figure 4 is a schematic diagram of the module architecture of the intelligent processor of the intelligent control system for the furnace temperature of the smelting furnace based on the neural network of the present invention. As Figure 4 shown, it mainly includes an input-output module, a model calculation module, and a storage module.

[0046] Among them, the input / output module is mainly used for data interaction with the parameter controller, receiving the current furnace temperature parameter, the current combustion control parameter, and the target furnace temperature parameter from the parameter controller, and sending the combustion adjustment parameter to the parameter controller. The model calculation module is mainly used for calculating the combustion adjustment parameter. The storage module is used for storing the data used by the model calculation module during calculation. The combustion adjustment parameter includes, for example, the change in fuel flow rate, the change in combustion-supporting gas flow rate, the change in exhaust gas flow rate, and so on.

[0047] The model calculation module can establish a neural network model and make the trained neural network model perform calculations. The neural network model can be trained externally to the intelligent processor and then deployed in the intelligent processor. During the training process of the neural network model, it needs to use a training data set. As a preferred embodiment, the training data set is composed of multiple pieces of data obtained during the historical smelting production of the smelting furnace, and each piece of data therein includes an input parameter and an output parameter.

[0048] As an embodiment, during the training of the model, the input parameters include the furnace temperature parameter before adjustment, the combustion control parameter before adjustment, and the furnace temperature parameter after adjustment, and the output parameter includes the combustion adjustment parameter. The trained neural network model then inputs the obtained current furnace temperature parameter, current combustion control parameter, and target furnace temperature parameter as the furnace temperature parameter before adjustment, the combustion control parameter before adjustment, and the furnace temperature parameter after adjustment respectively, so as to output the combustion adjustment parameter.

[0049] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the following further describes the present invention in detail with reference to specific embodiments and the accompanying drawings.

[0050] Figure 5 It is a schematic diagram of an embodiment of the display interface of the monitoring terminal of the intelligent control system for the furnace temperature of the smelting furnace based on a neural network according to the present invention.

[0051] As Figure 5 shown, different from the parameter setting interface of the monitoring terminal of the prior art shown Figure 1 in the figure, there is no option for setting combustion parameters in this display interface. Instead, there are options for turning on (ON) and turning off (OFF) the intelligent control mode. When the on option is selected, the intelligent control of the furnace temperature of the smelting furnace based on the neural network of the present invention is adopted. Otherwise, the traditional control method is directly executed on the parameter controller.

[0052] Different from the traditional control system that displays the set values and current values of the fuel valve opening, the rotational speeds of the combustion-supporting fan and the exhaust fan, the monitoring terminal of the present invention also displays the real-time dynamic values of the fuel valve opening, the rotational speeds of the combustion-supporting fan and the exhaust fan. That is to say, the fuel valve opening, the rotational speeds of the combustion-supporting fan and the exhaust fan of the present invention change dynamically under intelligent control, rather than working only under two settings of high fire and low fire as in the traditional technology.

[0053] In addition, the dynamic change values of the furnace temperature and the temperatures of each regeneration bed are also displayed below the display interface of the monitoring terminal of the present invention to facilitate the user to monitor the temperature change in the furnace in real time. This embodiment shows the situation of a melting furnace with three burners.

[0054] From Figure 5 It can also be seen that in this embodiment, the melting furnace includes a fuel pipe, a combustion-supporting air pipe and an exhaust pipe corresponding to the burner, as well as a fuel valve for controlling the fuel pipe flow rate, a combustion-supporting fan for supplying combustion-supporting air and an exhaust fan for exhausting gas; thus, the parameter controller controls the burner to perform combustion switching according to the combustion adjustment parameters obtained from the intelligent controller, calculates the opening of the fuel valve corresponding to the burning burner, the rotational speeds of the combustion-supporting fan and the exhaust fan according to the combustion adjustment parameters, and sends control signals to the fuel valve, the combustion-supporting fan and the exhaust fan corresponding to the burning burner.

[0055] In this embodiment, the number of burners is three, namely the 1# burner, the 2# burner and the 3# burner, which respectively correspond to the # bed, the 2# bed and the 3# bed. Here, the bed refers to the regeneration bed. However, the present invention is not limited to the number of burners, and the burners can be two or more. For heating balance and energy conservation, when one of the burners is burning, the other burners do not burn but are used to discharge the gas in the furnace.

[0056] In this embodiment, the combustion adjustment parameters further include whether the combustion state of the burner is switched. Since the combustion of the burners is carried out alternately, this involves the switching of the currently burning burner. As Figure 1 shown, the traditional control system switches the combustion of the burners in a timed manner, and the burner switching time can be set on the operation interface. However, this timed switching may cause uneven temperatures of each regeneration bed. For this reason, the present invention preferably also takes whether the combustion state of the burner is switched as a combustion adjustment parameter. That is to say, the neural network of the present invention can automatically calculate whether the burner needs to be switched at the current time. If switching is required, which burner to switch to.

[0057] In this embodiment, the current combustion control parameters further include the current burner angle, and the combustion adjustment parameters further include the adjustment amount of the burner angle. The burner angle refers to the inclination angle of the burner. Many practices have proven that different burner angles will affect the combustion effect of the burner, thereby affecting the change of the furnace temperature. However, the relationship between the burner angle and the change of the furnace temperature is not very clear. Therefore, in the prior art, it is impossible to utilize the change of the burner angle. Since the present invention adopts a neural network algorithm, it can learn the internal relationship between the burner angle and the change of the furnace temperature through a large amount of data training, thereby providing feasibility for the adjustment of the burner angle.

[0058] Figure 6 It is a schematic diagram of the neural network of the intelligent control system for the furnace temperature of the melting furnace based on the neural network of the present invention.

[0059] As Figure 6 shown, the neural network model includes an input layer, a hidden layer, and an output layer. The input parameters of the input layer include the furnace temperature parameters before adjustment (current furnace temperature parameters), the combustion control parameters before adjustment (current combustion control parameters), and the furnace temperature parameters after adjustment (target furnace temperature parameters). The output layer includes combustion adjustment parameters.

[0060] When the control system runs, the calculation of the neural network model can be executed regularly, thereby continuously adjusting the combustion control parameters. For example, the parameter controller obtains each parameter from the melting furnace every 0.5 s, and the intelligent processor sends it to the parameter controller after calculation.

[0061] Figure 7 It shows a flowchart of the intelligent control method for the furnace temperature of the melting furnace based on the neural network of the present invention. Corresponding to the dynamic control system for the furnace temperature of the melting furnace of the present invention, the control method includes: S1. Obtain the current furnace temperature parameters and the current combustion control parameters, and generate target furnace temperature parameters.

[0062] The melting furnace uses fuel combustion for heating, and includes a containing furnace chamber, at least two regenerative beds, and burners corresponding to the regenerative beds. As an example, the current furnace temperature parameters include the current furnace chamber temperature and the current regenerative bed temperature, and the current combustion control parameters include the fuel flow rate, the combustion-supporting gas flow rate, and the exhaust gas flow rate. More preferably, the combustion adjustment parameters further include whether the combustion state of the burner is switched. In addition, the current combustion control parameters further include the current burner angle.

[0063] S2. Generate combustion adjustment parameters according to the current furnace temperature parameters, the current combustion control parameters, and the target furnace temperature parameters, and according to the neural network model.

[0064] As an example, the combustion adjustment parameters include the change in fuel flow rate, the change in combustion-supporting gas flow rate, and the change in exhaust gas flow rate. More preferably, the combustion adjustment parameters further include the adjustment amount of the burner angle.

[0065] The neural network model is trained on a training data set, which is composed of multiple pieces of data obtained during the historical smelting production of the smelting furnace. Each piece of data therein includes input parameters and output parameters; the input parameters include the furnace temperature parameters before adjustment, the combustion control parameters before adjustment, and the furnace temperature parameters after adjustment, and the output parameters include the combustion adjustment parameters; the trained neural network model inputs the obtained current furnace temperature parameters, current combustion control parameters, and target furnace temperature parameters as the furnace temperature parameters before adjustment, the combustion control parameters before adjustment, and the furnace temperature parameters after adjustment respectively, so as to output the combustion adjustment parameters.

[0066] S3. Control the combustion parameters of the burner for the current combustion through the combustion adjustment parameters.

[0067] The smelting furnace further includes a fuel pipe, a combustion-supporting air pipe, and an exhaust pipe corresponding to the burner, as well as a fuel valve for controlling the fuel flow rate of the fuel pipe, a combustion-supporting air blower for supplying combustion-supporting air, and an exhaust fan for exhausting gas. Control the burner to perform combustion switching according to the obtained combustion adjustment parameters, calculate the opening degree of the fuel valve corresponding to the burner for combustion, the rotation speeds of the combustion-supporting air blower and the exhaust fan according to the combustion adjustment parameters, and send control signals to the fuel valve, the combustion-supporting air blower, and the exhaust fan corresponding to the burner for combustion.

[0068] It can be seen from the description of the above embodiments that the present invention solves the problem that the existing smelting furnace cannot predict the influence of combustion parameters on the furnace temperature and cannot accurately control the furnace temperature, resulting in an increase in oxidation and burning loss of molten metal and an increase in energy consumption. The present invention can accurately control the combustion mode of the burner, thereby making the furnace temperature control more accurate and balanced, so as to achieve the effects of reducing burning loss, reducing energy consumption, and improving the quality of the alloy.

[0069] The above specific embodiments have further detailed the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent control system for the temperature of a smelting furnace based on a neural network. The smelting furnace uses fuel combustion for heating and includes a furnace chamber and a burner, characterized in that, The intelligent control system includes: A parameter controller, which is used to obtain the current furnace temperature parameter and the current combustion control parameter from the smelting furnace, generate a target furnace temperature parameter, and send the current furnace temperature parameter, the current combustion control parameter, and the target furnace temperature parameter to the intelligent processor, and control the combustion parameters of the burners of the current combustion according to the combustion adjustment parameter obtained from the intelligent processor; An intelligent processor, which is used to receive the current furnace temperature parameter, the current combustion control parameter, and the target furnace temperature parameter sent by the parameter controller, generate a combustion adjustment parameter according to the neural network model, and send the combustion adjustment parameter to the parameter controller.

2. The intelligent control system for the furnace temperature of a smelting furnace based on a neural network according to claim 1, wherein: The neural network model is obtained by training on a training data set, and the training data set is composed of multiple data obtained by the smelting furnace during historical smelting production, and each piece of data therein includes an input parameter and an output parameter; The input parameters include the furnace temperature parameter before adjustment, the combustion control parameter before adjustment, and the furnace temperature parameter after adjustment, and the output parameter includes the combustion adjustment parameter; The trained neural network model inputs the obtained current furnace temperature parameter, current combustion control parameter, and target furnace temperature parameter as the furnace temperature parameter before adjustment, the combustion control parameter before adjustment, and the furnace temperature parameter after adjustment respectively, so as to output the combustion adjustment parameter.

3. The intelligent control system for the furnace temperature of a smelting furnace based on a neural network according to claim 2, wherein: The current furnace temperature parameter includes the current furnace chamber temperature and the current regenerator bed temperature; The current combustion control parameter includes the fuel flow rate, the combustion-supporting gas flow rate, and the exhaust gas flow rate; The combustion adjustment parameter includes the change amount of the fuel flow rate, the change amount of the combustion-supporting gas flow rate, and the change amount of the exhaust gas flow rate.

4. The intelligent control system for the temperature of the smelting furnace based on a neural network according to claim 3, characterized in that: There are at least two burners, and the combustion adjustment parameter also includes whether to switch the combustion state of the burners.

5. The intelligent control system for the temperature of the smelting furnace based on a neural network according to claim 4, characterized in that: The current combustion control parameter also includes the current burner angle, and the combustion adjustment parameter also includes the adjustment amount of the burner angle.

6. The intelligent control system for the furnace temperature of a smelting furnace based on a neural network according to claim 5, wherein: The smelting furnace further includes a fuel pipe, a combustion-supporting air pipe, and an exhaust pipe corresponding to the burners, as well as a fuel valve for controlling the fuel flow rate of the fuel pipe, a combustion-supporting air blower for providing combustion-supporting air, and an exhaust blower for exhausting gas; The parameter controller controls the burners to perform combustion switching according to the combustion adjustment parameter obtained from the intelligent controller, calculates the opening degree of the fuel valve corresponding to the burners of the combustion, the rotation speeds of the combustion-supporting air blower and the exhaust blower according to the combustion adjustment parameter, and sends control signals to the fuel valve, the combustion-supporting air blower, and the exhaust blower corresponding to the burners of the combustion.

7. An intelligent control method for the temperature of a smelting furnace based on a neural network. The smelting furnace uses fuel combustion for heating and includes a furnace chamber and burners, and is characterized in that, The intelligent control method includes: Obtaining the current furnace temperature parameter and the current combustion control parameter, and generating a target furnace temperature parameter; Generating a combustion adjustment parameter according to the current furnace temperature parameter, the current combustion control parameter, the target furnace temperature parameter, and according to the neural network model; Controlling the combustion parameters of the burners of the current combustion through the combustion adjustment parameter.

8. The intelligent control method for the furnace temperature of a smelting furnace based on a neural network according to claim 7, wherein: The neural network model is trained on a training dataset, which consists of multiple pieces of data obtained during the historical smelting production of the smelting furnace. Each piece of data includes input parameters and output parameters; The input parameters include the pre-regulation furnace temperature parameter, the pre-regulation combustion control parameter, and the post-regulation furnace temperature parameter, and the output parameter includes the combustion regulation parameter; The trained neural network model inputs the obtained current furnace temperature parameter, current combustion control parameter, and target furnace temperature parameter as the pre-regulation furnace temperature parameter, pre-regulation combustion control parameter, and post-regulation furnace temperature parameter respectively to output the combustion regulation parameter.

9. The intelligent control method for the furnace temperature of a smelting furnace based on a neural network according to claim 8, wherein: The current furnace temperature parameter includes the current furnace chamber temperature and the current regenerator bed temperature; The current combustion control parameter includes the fuel flow rate, the combustion-supporting gas flow rate, and the exhaust gas flow rate; The combustion regulation parameter includes the change in the fuel flow rate, the change in the combustion-supporting gas flow rate, and the change in the exhaust gas flow rate.

10. The intelligent control method for the temperature of the smelting furnace based on neural network according to claim 9, characterized in that: There are at least two burners, and the combustion regulation parameter further includes whether to switch the combustion state of the burners.

Images