Prediction and regulation method and device for copper smelting, medium and equipment

Abstract

The invention discloses a prediction, regulation and control method and device for copper smelting, a medium and equipment, and the method comprises the steps: collecting material flow and energy flow data in a copper smelting process in real time, predicting future material flow and energy flow states through a prediction model, and inputting the prediction data into an intelligent agent; and the intelligent agent dynamically generates a control strategy based on the optimization target of the reward function, and smelting process parameters are accurately adjusted. The technology has the beneficial effects that the problem that control parameters depend on artificial experience in traditional smelting is solved, the phenomena of extensive and lagged manual adjustment are avoided especially when process fluctuation is coped with, finer, more efficient and adaptive process optimization is achieved, the smelting quality is improved, and energy consumption and pollution are reduced.

Description

Technical Field

[0001] The present invention relates to the field of smelting control technology, and in particular to a prediction and control method, device, medium and equipment for copper smelting. Background Art

[0002] The existing copper smelting production mainly relies on distributed control systems, and production control is achieved through manually set process parameters. However, the smelting process is instantaneous and continuous, and the upstream and downstream process links are mutually restricted. Manual adjustment of control parameters mainly depends on personal experience, and there are often problems of extensive and lagging adjustments, which will further lead to problems such as unstable production and large fluctuations in working conditions. Summary of the invention

[0003] Based on this, it is necessary to provide prediction and regulation methods, devices, media and equipment for copper smelting to solve the problem that manual adjustment of control parameters mainly relies on personal experience, and the adjustment is often extensive and lagging.

[0004] A method for predicting and controlling copper smelting, characterized in that the method comprises:

[0005] Obtain current material flow data and current energy flow data in the copper smelting process;

[0006] Encoding the current material flow data and the current energy flow data in time sequence and inputting them into a converged prediction model to predict future material flow data and future energy flow data;

[0007] The future material flow data and the future energy flow data are input into the intelligent agent, so that the intelligent agent determines a control strategy based on the goal of maximizing a preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

[0008] In one embodiment, after obtaining the current material flow data and current energy flow data during the current copper smelting, the method further includes:

[0009] Removing noise of a preset frequency band in the current material flow data and the current energy flow data by using a filter;

[0010] Detect and eliminate abnormal values ​​in the current material flow data and the current energy flow data by using a multiple standard deviation method;

[0011] Filling missing values ​​in the current material flow data and the current energy flow data by linear interpolation;

[0012] Smoothing the current material flow data and the current energy flow data by using an exponentially weighted moving average method;

[0013] The current material flow data and the current energy flow data are normalized.

[0014] In one embodiment, after obtaining the current material flow data and current energy flow data during the current copper smelting, the method further includes:

[0015] Constructing an original matrix including the current material flow data and the current energy flow data; wherein each element in the original matrix corresponds to one current material flow data or one current energy flow data;

[0016] The original matrix is ​​standardized and the covariance matrix of the original matrix after the standardization is calculated; wherein the elements in the covariance matrix are used to express the correlation relationship between the data;

[0017] Performing eigenvalue decomposition on the covariance matrix to obtain D eigenvalues; wherein D is the total number of data types of the current material flow data and the current energy flow data;

[0018] The current material flow data and the current energy flow data corresponding to the largest d eigenvalues ​​among the D eigenvalues ​​are retained; wherein the cumulative contribution of the d eigenvalues ​​is greater than a preset contribution threshold.

[0019] In one embodiment, the prediction model includes a plurality of loop structures, each loop structure including a forget gate, an input gate, an output gate, and a unit state;

[0020] The cell state connects the cell state of the previous moment and the cell state of the next moment, and is used to transfer long-term memory between moments;

[0021] The forget gate is used to control the amount of data retained from the unit state at the previous moment to the current moment, so as to screen the long-term memory;

[0022] The input gate is used to control the amount of data in the current input data that is saved to the current unit state; wherein the current input data is the code corresponding to the current material flow data or the current energy flow data;

[0023] The output gate is used to adjust the output data amount of the current unit state to transmit the short-term memory of the current moment; wherein the short-term memory indicates future material flow data or future energy flow data.

[0024] In one embodiment, the method further comprises:

[0025] Substituting the future material flow data into the material balance equation, and judging whether the material flow has reached a balanced state through the calculation result;

[0026] Substitute the future energy flow data into the energy balance equation, and determine whether the energy flow reaches a balanced state based on the calculation result.

[0027] In one embodiment, the material balance equation is expressed as:

[0028]

[0029] In the above formula, ∑ m represents the sum of the substance m; E c，e Indicates the influence coefficient or conversion efficiency of material components c and e; It represents the amount of variable component c of substance m under input state; represents the concentration of the constant component c of substance m under the input state; represents the input amount of substance m; represents the input mass of substance m; Represents the amount of variable component c of substance m in the output state; C m，c represents the concentration of variable component c of substance m; Represents the output mass of substance m.

[0030] In one embodiment, the energy flow balance calculation formula is:

[0031]

[0032] ∑ i represents the sum of reactants i; ΔH 298，Ai Represents reactant A i Enthalpy change at 298K; Represents reactant A i Specific heat capacity at constant pressure; r i Represents reactant A i The final temperature; Represents reactant A i Heating from 298K to r i The heat required during the temperature process; j represents the sum of product j; ΔH 298，Bj Represents product B j Enthalpy change at 298K; Represents product B j The constant pressure specific heat capacity; T represents the final temperature of the product; Represents product B j The heat required to heat from 298K to T; Q Loss Represents the heat loss in the system.

[0033] A prediction and control device for copper smelting, the prediction and control device for copper smelting comprising:

[0034] A data acquisition module is used to acquire current material flow data and current energy flow data in the copper smelting process;

[0035] A prediction module, used for encoding the current material flow data and the current energy flow data in time sequence and inputting the encoded data into a converged prediction model to predict future material flow data and future energy flow data;

[0036] The control module is used to input the future material flow data and the future energy flow data into the intelligent agent, so that the intelligent agent determines the control strategy based on the goal of maximizing a preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

[0037] A computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to execute the steps of the above-mentioned copper smelting prediction and control method.

[0038] A terminal device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above-mentioned copper smelting prediction and control method.

[0039] The present invention provides a prediction and control method, device, medium and equipment for copper smelting. By real-time collection of material flow and energy flow data in the copper smelting process, and using a prediction model to predict the future material flow and energy flow state, the predicted data is input into an intelligent agent, and the intelligent agent dynamically generates a control strategy based on the optimization target of the reward function to accurately adjust the smelting process parameters. The beneficial effect of this technology is that it solves the problem of control parameters relying on manual experience in traditional smelting, especially when dealing with process fluctuations, it avoids the phenomenon of extensive and lagging manual adjustment, and realizes more detailed, efficient and adaptive process optimization, which helps to improve smelting quality and reduce energy consumption and pollution. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] in:

[0042] Figure 1 A schematic diagram of a process for prediction and control of copper smelting;

[0043] Figure 2It is a schematic diagram of the structure of the prediction and control device for copper smelting;

[0044] Figure 3 This is a structural block diagram of the terminal equipment. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0047] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0048] The inputs in the copper smelting process can be mainly divided into copper concentrate, slag-making agent and oxygen-enriched air. Copper concentrate includes two categories: copper sulfide ore and copper oxide ore. The main types of copper sulfide ore are Cu2S (chalcocite), CuS (covellite), CuFeS2 (chalcopyrite) and Cu5FeS4 (bornite). The main types of copper oxide ore are Cu2O (copperite) and CuO (copper ore). The output materials involved in the copper smelting process can be mainly divided into copper slag, copper matte, smoke and flue gas.

[0049] The existing copper smelting production mainly relies on distributed control systems, and production control is achieved through manually set process parameters (including furnace pressure, fuel consumption, etc.). However, the smelting process is instantaneous and continuous, and the upstream and downstream process links are mutually restricted. Manual adjustment of control parameters mainly depends on personal experience, and there are often problems of extensive and lagging adjustments, which will further lead to problems such as unstable production and large fluctuations in working conditions.

[0050] like Figure 1 As shown, Figure 1 The figure is a flow chart of a method for predicting and controlling copper smelting in one embodiment. The steps provided by the method for predicting and controlling copper smelting in this embodiment include:

[0051] S101, obtaining current material flow data and current energy flow data in the copper smelting process.

[0052] Among them, material flow data refers to the quality, concentration or flow rate of various raw materials, products and intermediates in the copper smelting process. For example, the flow rate of copper ore, oxygen input, output slag, etc. Energy flow data refers to data related to energy input and output in the smelting process, such as temperature, furnace pressure, fuel consumption, etc. These data reflect the distribution and changes of heat energy, electrical energy, etc. in the smelting process.

[0053] Optionally, the current material flow data and the current energy flow data can be obtained in the following ways: the total amount of materials entering and exiting the oxygen top-blown smelting furnace is obtained through the batching and conveying system or the on-site material total amount measuring instrument. The composition and mass flow of the ore entering the submerged top-blown smelting furnace are obtained through the online composition detector and the electronic belt scale, and real-time monitoring is performed. The smelting temperature in the furnace is obtained and monitored through the temperature sensor arranged in the smelting furnace. The outlet temperature of copper slag and copper matte is obtained through the armored thermocouple. The oxygen concentration in the furnace is obtained and monitored through the oxygen content sensor installed in the flue of the smelting furnace. The inlet flow rate of oxygen-enriched air, the flue gas flow rate and the cooling water flow rate of the water-cooling jacket are obtained through the flowmeter installed at the conveying inlet. The temperature changes of slag and flue gas are obtained and monitored through the infrared thermometer. The flue gas composition is obtained through the flue gas analyzer. The temperature and humidity of the reference environment are obtained through the dry-bulb thermometer.

[0054] Optionally, after obtaining the current material flow data and current energy flow data during the current copper smelting in S101, the following steps may be performed:

[0055] A1. Remove the noise of the preset frequency band in the current material flow data and the current energy flow data through the filter.

[0056] Specifically, a low-pass filter is used to remove high-frequency noise, and its mathematical expression is as follows: y(t)=x(t)*h(t), where: y(t) is the filtered signal, x(t) is the original signal, that is, the various real-time collected data input, h(t) is the impulse response of the low-pass filter, and * represents the convolution operation.

[0057] A2. Detect and eliminate abnormal values ​​in the current material flow data and the current energy flow data through the multiple standard deviation method.

[0058] Specifically, the statistical 3-times standard deviation method is used to detect and eliminate outliers to ensure the authenticity and representativeness of the data. In the 3-times standard deviation method, for normally distributed data, data points that exceed the mean ±3 times the standard deviation can be considered outliers. The outlier is represented by X 异常 ={x||x-μ|>3σ}, where μ is the data mean and σ is the standard deviation.

[0059] A3. Fill in the missing values ​​in the current material flow data and the current energy flow data through linear interpolation.

[0060] Specifically, for the missing parts in the data, linear interpolation is used to fill them in to ensure the continuity of the data. Linear interpolation uses two adjacent known points to estimate the missing values. The formula is as follows: Where: (x1, y1) and (x2, y2) are known data points adjacent to the missing value.

[0061] A4. Smooth the current material flow data and current energy flow data by using the exponentially weighted moving average method.

[0062] Specifically, the exponentially weighted moving average (EWMA) method is used to give exponentially decaying weights to historical data, which has a higher and faster response speed to short-term changes, so as to smooth the time series data and ensure that the trend characteristics are prominent. The formula is as follows: S t =αy t +(1-α)S t-1 , where: α is the smoothing coefficient, ranging from 0 to 1.

[0063] A5. Normalize the current material flow data and the current energy flow data.

[0064] Specifically, the collected data is checked for consistency to ensure that all data formats are unified and the dimensions are consistent. Since the dimensions of different data sets are different and there are large differences between different values, the accuracy of the data analysis results will be affected by the values. In order to reduce its impact on the accuracy of the analysis results, it is necessary to perform min-max standardization, linearly change the production data collected from the copper smelter, and map the result values ​​after the linear change to the range of [0, 1]: Where: max represents the maximum value of the sample data; min represents the minimum value of the sample data. In addition, the data should also conform to the expected physical range and logical relationship. Physical constraint verification: the oxygen concentration should be between 0% and 100%. If the data exceeds this range, further inspection and processing are required. Logical relationship verification: the smelting temperature should be consistent with the thermodynamic characteristics of the ore component reaction. If abnormal fluctuations occur, it is necessary to re-collect data or adjust the sensor.

[0065] Optionally, after obtaining the current material flow data and current energy flow data during the current copper smelting in S101, the following steps may be performed:

[0066] B1. Construct an original matrix containing current material flow data and current energy flow data.

[0067] Each element in the original matrix corresponds to a current material flow data or a current energy flow data. Specifically, take the data matrix X∈R under normal working conditions m×n , where m is the number of sampling times and n is the number of features.

[0068] In the above formula, X1=[x1(1),x1(2),…,x1(m)] T .

[0069] B2. Standardize the original matrix and calculate the covariance matrix of the standardized original matrix.

[0070] The elements in the covariance matrix are used to express the correlation between data. Specifically, the larger the value of the element in the covariance matrix, the more necessary it is to perform principal component analysis on the data.

[0071]

[0072] In the above formula, r ij ——is feature x i With feature x j The correlation coefficient of .

[0073] B3. Perform eigenvalue decomposition on the covariance matrix to obtain D eigenvalues.

[0074] Where D is the total number of data types of the current material flow data and the current energy flow data. Specifically, to calculate the eigenvalue of R, it is necessary to solve the characteristic equation:

[0075] |λI-R|=0

[0076] In the above formula, λ is the eigenvalue and I is the unit matrix.

[0077] B4. Keep the current material flow data and current energy flow data corresponding to the largest d eigenvalues ​​among the D eigenvalues.

[0078] Among them, the cumulative contribution of d eigenvalues ​​is greater than a preset contribution threshold.

[0079] Specifically, the eigenvalues ​​are arranged in order of magnitude. The eigenvalue is the average variance of each principal component, and its value reflects the contribution of the principal component and the cumulative contribution rate. The contribution rate is: The cumulative contribution rate is: Establish the initial factor loading matrix and explain the principal components. The factor loading is the correlation coefficient R between the principal component Zi and the original feature Xi, which reveals the degree of correlation between the principal components. By performing principal component analysis on the control variables and obtaining the minimum number of principal components / features with a cumulative variance contribution rate of more than 75%, the dimensionality reduction of the relevant variable parameters can be achieved.

[0080] S102, encoding the current material flow data and the current energy flow data in time sequence and inputting them into a converged prediction model to predict future material flow data and future energy flow data.

[0081] Specifically, the material flow and energy flow data at different time points in the copper smelting process are recorded in sequence to form a time series (such as recording flow, temperature, composition and other data at intervals of one minute), which are then encoded and input into a converged prediction model to accurately predict future data, including future material flow data and future energy flow data.

[0082] Optionally, the prediction model includes multiple loop structures, each loop structure includes a forget gate, an input gate, an output gate, and a unit state; the unit state connects the unit state of the previous moment and the unit state of the next moment, and is used to transfer long-term memory between moments; the forget gate is used to control the amount of data retained from the unit state of the previous moment to the current moment, so as to screen the long-term memory; the input gate is used to control the amount of data saved in the current unit state in the current input data; wherein the current input data is the encoding corresponding to the current material flow data or the current energy flow data; the output gate is used to adjust the output data amount of the current unit state to transfer the short-term memory of the current moment; wherein the short-term memory indicates future material flow data or future energy flow data.

[0083] Since the temperature, oxygen concentration, ore ratio, etc. in the smelting process are all time series data, the application of this prediction model can effectively capture the long-term dependencies in the sequence data and be used to predict the dynamic parameter changes in the smelting process. The various structures are represented as follows:

[0084] Forget gate: f t =σ(W f ·[h t-1 , x t ]+b f )

[0085] Input gate: i t =σ(W i ·[h t-1 , x t ]+b i )

[0086] Candidate memory cells:

[0087] Output gate: i t =σ(W o ·[h t-1 , x t ]+b o )

[0088] Update memory cells and hidden states:

[0089] h t =o t *tanh(C t )

[0090] S103, inputting the future material flow data and the future energy flow data into the intelligent agent, so that the intelligent agent determines the control strategy based on the goal of maximizing the preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

[0091] Specifically, during the smelting process, the agent inputs future material flow and energy flow data to obtain the state that the system is about to enter. The input of future data helps the agent understand possible changes in process parameters in advance. These data include key smelting process conditions, such as temperature, oxygen concentration, ore ratio and other sensor monitoring data for a period of time in the future. By mastering this information, the agent can "foresee" possible conditions in the smelting process and take corresponding measures based on these predictions.

[0092] The reward function is an important indicator used to measure the smelting effect. It is the core of defining the operation effect of the intelligent agent. Here, the reward function is comprehensively designed based on the following key indicators: Smelting quality: such as whether the copper content of the copper matte meets the expected standard. Energy consumption: The lower the energy consumption per unit output, the higher the reward. Pollutant emissions: For example, the lower the content of harmful gases in the flue gas, the higher the reward. The goal of the intelligent agent is to maximize the value of the reward function, that is, to optimize the various indicators in the smelting process. Through the predicted input of the future state, the intelligent agent dynamically finds a set of optimal operation strategies in the complex control space according to the reward function.

[0093] Based on these future data, the agent will determine the control strategy according to the current state and reward function, including adjusting the oxygen flow rate, oxygen-enriched air input, ore ratio, etc. Furthermore, the agent adjusts the key process parameters in the smelting process based on the control strategy until the requirements of the control strategy are met.

[0094] In the above embodiment, by collecting the material flow and energy flow data in the copper smelting process in real time, and using the prediction model to predict the future material flow and energy flow state, the prediction data is input into the intelligent agent, and the intelligent agent dynamically generates a control strategy based on the optimization target of the reward function to accurately adjust the smelting process parameters. The beneficial effect of this technology is that it solves the problem of control parameters relying on manual experience in traditional smelting, especially when dealing with process fluctuations, it avoids the phenomenon of extensive and lagging manual adjustment, and realizes more detailed, efficient and adaptive process optimization, which helps to improve smelting quality and reduce energy consumption and pollution.

[0095] Furthermore, in order to more scientifically ensure that the entire system reaches a balanced state in terms of input and output, the above copper smelting prediction and control method can also perform the following steps:

[0096] C1. Substitute the future material flow data into the material balance equation and determine whether the material flow has reached a balanced state based on the calculation results.

[0097] Among them, the material balance equation is used to confirm whether the amount of input and output materials in the smelting process is balanced to ensure that there is no abnormal loss or accumulation of materials. For example, the input amount of ore, combustion aid, etc. entering the smelting furnace can be compared with the output amount of output materials (such as copper matte, copper slag, etc.). The material flow data is substituted into the calculation using the balance equation to determine whether the equilibrium state is reached. If the equilibrium state is not reached, the relevant process parameters (such as the amount of input raw materials, oxygen concentration, etc.) can be automatically adjusted to make the system gradually tend to equilibrium.

[0098] Optionally, the material balance equation is expressed as:

[0099]

[0100] In the above formula, ∑ m represents the sum of the substance m; E c，e Indicates the influence coefficient or conversion efficiency of material components c and e; It represents the amount of variable component c of substance m under input state; represents the concentration of the constant component c of substance m under the input state; represents the input amount of substance m; represents the input mass of substance m; Represents the amount of variable component c of substance m in the output state; C m，c represents the concentration of variable component c of substance m; Represents the output mass of substance m.

[0101] C2. Substitute the future energy flow data into the energy balance equation and determine whether the energy flow has reached a balanced state based on the calculation results.

[0102] Among them, the energy balance equation is used to confirm whether the total amount of energy input and output in the smelting process is balanced. It usually includes energy consumption such as heating, cooling and material reaction. For example, the system will calculate the balance between input energy (fuel combustion, oxygen-enriched air, etc.) and consumed energy to ensure effective management of temperature and energy consumption in the system. Similarly, the energy flow data is substituted into the calculation using the balance equation to determine whether a state of equilibrium has been reached. If a state of equilibrium has not been reached, the relevant process parameters (such as energy supply, etc.) can be automatically adjusted so that the system gradually tends to equilibrium.

[0103] Optionally, the energy flow balance calculation formula is:

[0104]

[0105] ∑ i represents the sum of reactants i; ΔH 298，Ai Represents reactant A i Enthalpy change at 298K; Represents reactant A i Specific heat capacity at constant pressure; r i Represents reactant A i The final temperature; Represents reactant A i Heating from 298K to r i The heat required during the temperature process; j represents the sum of product j; ΔH 298，Bj Represents product B j Enthalpy change at 298K; Represents product B j The constant pressure specific heat capacity; T represents the final temperature of the product; Represents product B j The heat required to heat from 298K to T; Q Loss Represents the heat loss in the system.

[0106] It can be understood that the copper smelting process model obtained based on the material and energy balance equations will serve as the basis for adjusting the parameters in the prediction model, and will also serve as a reference model for reliability analysis of the prediction model.

[0107] In one embodiment, Figure 2 As shown, a prediction and control device for copper smelting is proposed, which includes:

[0108] The data acquisition module 201 is used to acquire the current material flow data and the current energy flow data in the copper smelting process;

[0109] A prediction module 202, for encoding the current material flow data and the current energy flow data in time sequence and inputting the encoded data into a converged prediction model to predict future material flow data and future energy flow data;

[0110] The control module 203 is used to input the future material flow data and the future energy flow data into the intelligent agent, so that the intelligent agent determines the control strategy based on the goal of maximizing the preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

[0111] Figure 3 FIG. 2 shows an internal structure diagram of a terminal device in an embodiment. Figure 3 As shown, the terminal device includes a processor, a memory and a network interface connected via a system bus. Among them, the memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the terminal device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor can implement the prediction and control method of copper smelting. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor can implement the prediction and control method of copper smelting. Those skilled in the art can understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the terminal device to which the scheme of the present application is applied. The specific terminal device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0112] A computer-readable storage medium stores a computer program, which implements the following steps when executed by a processor: obtaining current material flow data and current energy flow data in a copper smelting process; encoding the current material flow data and the current energy flow data in chronological order and inputting them into a convergent prediction model to predict future material flow data and future energy flow data; inputting the future material flow data and the future energy flow data into an intelligent agent, so that the intelligent agent determines a control strategy based on the goal of maximizing a preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

[0113] A terminal device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: obtaining current material flow data and current energy flow data in a copper smelting process; encoding the current material flow data and the current energy flow data in chronological order and inputting them into a convergent prediction model to predict future material flow data and future energy flow data; inputting the future material flow data and the future energy flow data into an intelligent agent, so that the intelligent agent determines a control strategy based on the goal of maximizing a preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

[0114] It should be noted that the above-mentioned copper smelting prediction and control method, device, equipment and computer-readable storage medium belong to a general inventive concept, and the contents of the copper smelting prediction and control method, device, equipment and computer-readable storage medium embodiments are applicable to each other.

[0115] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing related hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0116] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for predicting and controlling copper smelting, characterized in that: The method comprises: Obtain current material flow data and current energy flow data in the copper smelting process; Encoding the current material flow data and the current energy flow data in time sequence and inputting them into a converged prediction model to predict future material flow data and future energy flow data; The future material flow data and the future energy flow data are input into the intelligent agent, so that the intelligent agent determines a control strategy based on the goal of maximizing a preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

2. The method according to claim 1, characterized in that After obtaining the current material flow data and current energy flow data during the current copper smelting, the method further includes: Removing noise of a preset frequency band in the current material flow data and the current energy flow data by using a filter; Detect and eliminate abnormal values ​​in the current material flow data and the current energy flow data by using a multiple standard deviation method; Filling missing values ​​in the current material flow data and the current energy flow data by linear interpolation; Smoothing the current material flow data and the current energy flow data by using an exponentially weighted moving average method; The current material flow data and the current energy flow data are normalized.

3. The method according to claim 1, characterized in that After obtaining the current material flow data and current energy flow data during the current copper smelting, the method further includes: Constructing an original matrix including the current material flow data and the current energy flow data; wherein each element in the original matrix corresponds to one current material flow data or one current energy flow data; The original matrix is ​​standardized and the covariance matrix of the standardized original matrix is ​​calculated; wherein the elements in the covariance matrix are used to express the correlation between the data; Performing eigenvalue decomposition on the covariance matrix to obtain D eigenvalues; wherein D is the total number of data types of the current material flow data and the current energy flow data; The current material flow data and the current energy flow data corresponding to the largest d eigenvalues ​​among the D eigenvalues ​​are retained; wherein the cumulative contribution of the d eigenvalues ​​is greater than a preset contribution threshold.

4. The method according to claim 1, characterized in that: The prediction model includes a plurality of loop structures, each loop structure including a forget gate, an input gate, an output gate, and a unit state; The cell state connects the cell state of the previous moment and the cell state of the next moment, and is used to transfer long-term memory between moments; The forget gate is used to control the amount of data retained from the unit state at the previous moment to the current moment, so as to screen the long-term memory; The input gate is used to control the amount of data in the current input data that is saved to the current unit state; wherein the current input data is the code corresponding to the current material flow data or the current energy flow data; The output gate is used to adjust the output data amount of the current unit state to transmit the short-term memory of the current moment; wherein the short-term memory indicates future material flow data or future energy flow data.

5. The method according to claim 1, characterized in that The method further comprises: Substituting the future material flow data into the material balance equation, and judging whether the material flow has reached a balanced state through the calculation result; Substitute the future energy flow data into the energy balance equation, and determine whether the energy flow reaches a balanced state based on the calculation result.

6. The method according to claim 5, characterized in that The material balance equation is expressed as: In the above formula, ∑ m represents the sum of the substance m; E c，e Indicates the influence coefficient or conversion efficiency of material components c and e; It represents the amount of variable component c of substance m under input state; represents the concentration of the constant component c of substance m under the input state; represents the input amount of substance m; represents the input mass of substance m; Represents the amount of variable component c of substance m in the output state; C m，c represents the concentration of variable component c of substance m; Represents the output mass of substance m.

7. The method according to claim 5, characterized in that The energy flow balance calculation formula is: ∑ i represents the sum of reactants i; ΔH 298，Ai Represents reactant A i Enthalpy change at 298K; Represents reactant A i Specific heat capacity at constant pressure; r i Represents reactant A i The final temperature; Represents reactant A i Heating from 298K to r i The heat required during the temperature process; j represents the sum of product j; ΔH 298，Bj Represents product B j Enthalpy change at 298K; Represents product B j The constant pressure specific heat capacity; T represents the final temperature of the product; Represents product B j The heat required to heat from 298K to T; Q Loss Represents the heat loss in the system.

8. A prediction and control device for copper smelting, characterized in that: The prediction and control device for copper smelting includes: A data acquisition module is used to acquire current material flow data and current energy flow data in the copper smelting process; A prediction module, used for encoding the current material flow data and the current energy flow data in time sequence and inputting the encoded data into a converged prediction model to predict future material flow data and future energy flow data; The control module is used to input the future material flow data and the future energy flow data into the intelligent agent, so that the intelligent agent determines the control strategy based on the goal of maximizing a preset reward function, and adjusts the process parameters of copper smelting according to the control strategy.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

10. A terminal device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 7.

Images