Automatic control system for producing lithium carbonate

Abstract

The invention belongs to the technical field of production automation, and discloses an automatic control system for producing lithium carbonate, which comprises a data acquisition module used for acquiring production parameter data, equipment parameter data and environment parameter data within t time; the data processing module is used for preprocessing the acquired production parameter data, equipment parameter data and environment parameter data to obtain a production feature data set, an equipment feature data set and an environment feature data set; performing feature extraction on the production feature data set, the equipment feature data set and the environment feature data set to obtain a comprehensive feature data set; the quality evaluation module is used for constructing a quality evaluation model according to the comprehensive feature data set, and obtaining technological parameter errors of lithium carbonate production based on the quality evaluation model; the labor intensity of operators is reduced, the risk of human errors is reduced, and the overall efficiency and accuracy of the production process are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of production automation, and more specifically, to an automatic control system for producing lithium carbonate. Background Art

[0002] Patent publication number CN108287535A discloses an automated control system for lithium carbonate production. The present invention provides an automated control system for lithium carbonate production, including a mineral analyzer, a central control server, a roasting controller, a roasting device; a particle size analyzer, a grinding controller, a grinding device; a first flow meter, a second flow meter, and an electromagnetic valve controller. The embodiment of the present invention can realize automatic setting of the operation and temperature of the roasting device, automatic setting of the grinding time of the grinding device, and automatic control of the addition of concentrated sulfuric acid during the process of preparing natural spodumene into β-spodumene concentrate, grinding, and sulfuric acid acidification. The operation is simple and error-free, meeting people's demand for automated lithium carbonate production.

[0003] The existing control system for producing lithium carbonate has the following main problems:

[0004] Failure to consider using the EMD method to gradually extract IMF components of different frequencies and time scales may lead to confusion between short-term fluctuations and long-term trends in production data; making it difficult for the control system to identify and distinguish dynamic changes at different time scales during lithium carbonate production, which may lead to the adoption of incorrect long-term strategies for short-term fluctuations, affecting precise control of production; Failure to use the Cauchy convergence criterion as the iteration stop condition may cause the decomposition process to be excessive, leading to the introduction of new errors and even signal distortion. Excessive iteration will cause the noise components in the signal to be over-separated, destroying the main features of the original signal, affecting the accuracy of the data and the authenticity of the signal; due to the inability to accurately identify key IMF components and influencing factors, the production control strategy may lack specificity, resulting in inaccurate adjustments in the production process, unnecessary adjustments or delayed responses. This imprecise control will lead to an unstable production process, frequent fluctuations or deviations, affecting the efficient operation of production, and may even lead to downtime or reduced product quality;

[0005] Failure to use Hilbert transform to analyze the instantaneous frequency and amplitude of the IMF component will result in missing out on the detailed time-frequency characteristics of the signal during the production process. Lack of time-frequency characteristic analysis may make it impossible to accurately grasp the frequency changes and instantaneous characteristics during the production process, thus affecting the in-depth understanding of the production process and the formulation of optimization strategies. Lack of restrictions on proportional gain may lead to an uneven error adjustment process. Too high a proportional gain may cause the system to overreact, oscillate or become unstable; too low a proportional gain may cause the system to respond slowly. Without a reasonable proportional gain restriction, the system may not be able to effectively shorten the process parameter adjustment time, affecting production accuracy and process stability.

[0006] In view of this, the present invention proposes an automatic control system for producing lithium carbonate to solve the above problems. Summary of the invention

[0007] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: an automatic control system for producing lithium carbonate, comprising:

[0008] Data collection module, used to collect production parameter data, equipment parameter data and environmental parameter data within time t;

[0009] A data processing module is used to pre-process the acquired production parameter data, equipment parameter data and environmental parameter data to obtain a production feature data set, an equipment feature data set and an environmental feature data set;

[0010] Extracting features from the production feature data set, the equipment feature data set, and the environment feature data set to obtain a comprehensive feature data set;

[0011] The quality assessment module is used to build a quality assessment model based on the comprehensive feature data set, and obtain the process parameter error of lithium carbonate production based on the quality assessment model; compare the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard; if the product quality meets the standard, the production terminal is automatically controlled to issue a qualified instruction to continue the production of lithium carbonate; if the product quality does not meet the standard, the production terminal is automatically controlled to issue a warning instruction and adjust the production parameters of lithium carbonate;

[0012] The strategy adjustment module is used to adjust the production parameters of lithium carbonate according to the process parameter errors of lithium carbonate production to reduce the process parameter errors; each module is connected by wired and / or wireless means.

[0013] Furthermore, the production parameter data include: temperature data, humidity data, pressure data, reactant concentration data, gas flow data, liquid flow data and PID controller parameter data in the hydrogenation reactor; the equipment parameter data include: current data, voltage data, vibration data and cumulative operating time; the environmental parameter data include: ambient temperature data, ambient humidity data and ambient air pressure data;

[0014] The production parameter data, equipment parameter data and environmental parameter data are acquired through installed sensors and control terminals.

[0015] Furthermore, the method of preprocessing the acquired production parameter data, equipment parameter data and environmental parameter data to obtain the production feature data set, the equipment feature data set and the environmental feature data set includes:

[0016] The empirical mode decomposition method EMD is used to preprocess the production parameter data, equipment parameter data and environmental parameter data to obtain the production feature data set, equipment feature data set and environmental feature data set; the specific steps are as follows:

[0017] S31, presetting an original data sequence x(t), wherein the original data sequence includes a production feature data set, an equipment feature data set, and an environment feature data set;

[0018] S32, identifying all local extreme points; interpolating the original data sequence x(t) to find all local maximum points and local minimum points;

[0019] S33, construct upper and lower envelopes; perform cubic spline interpolation on the local maximum points to construct the upper envelope e max (t); Perform cubic spline interpolation on the local minimum points to construct the lower envelope e min (t);

[0020] S34, calculating the average envelope m(t);

[0021] The mathematical formula is: Where t is the index of the data point;

[0022] S35, calculating the intrinsic mode function IMF; subtracting the average envelope from the original data sequence to obtain the intrinsic mode function h(t)=x(t)-m(t);

[0023] Determine whether h(t) satisfies the IMF condition, specifically: in the entire data sequence, the number of extreme points and zero crossing points of h(t) must be equal or differ by at most one; at any time, the average value of the upper envelope formed by the local maximum points and the lower envelope formed by the local minimum points must be zero;

[0024] If h(t) satisfies the above conditions, no iteration is required; if h(t) does not satisfy the above conditions, h(t) is used as a new data sequence and steps S32-S35 are repeated until the IMF conditions are met;

[0025] S36, extracting IMF components; when h(t) satisfies the IMF condition, taking h(t) as the first IMF component c1(t); subtracting c1(t) from the original data sequence to obtain the residual trend r1(t); repeating the iteration to extract the next IMF component until the residual trend r1(t) satisfies the stopping criterion and stops the iteration;

[0026] After the above steps, all IMF components and residual trends are extracted, integrated and reconstructed to obtain production feature data sets, equipment feature data sets and environmental feature data sets.

[0027] Further, the method for stopping iteration when the residual trend meets the stopping criterion includes:

[0028] The Cauchy convergence criterion is used as the stopping criterion, and the iteration is stopped when the residual trend satisfies the stopping criterion; the specific mathematical formula of the Cauchy convergence criterion is: Among them, h k-1 (t) is the IMF component generated by the k-1th iteration, h k (t) is the IMF component generated by the kth iteration; N is the total number of data points; t is the index of the data point; when SD k When it is less than the preset threshold SD′, the iteration process stops.

[0029] Furthermore, the method for determining the value of the preset threshold SD′ includes:

[0030] The threshold SD′ is determined by a threshold limiting formula; the threshold limiting formula is specifically: Among them, σ is the standard deviation of the data sequence; N is the total number of data points; η is the noise ratio in the data sequence; α1 is the adjustment coefficient of the standard deviation; α2 is the adjustment coefficient of the noise ratio.

[0031] Furthermore, the method of extracting features from the production feature data set, the equipment feature data set and the environment feature data set to obtain a comprehensive feature data set includes:

[0032] For each IMF component c i′ (t), whose energy is defined as: Where, is the value of the IMF component at time t;

[0033] Calculate the relative energy percentage of each IMF component using the formula: Where, is the relative energy percentage of the i′th IMF component; m′ is the total number of IMF components; is the sum of the energies of all IMF components; j is the index of traversing all IMF components;

[0034] Based on the calculated relative energy percentage, the IMF component with the largest proportion is selected as the feature; the instantaneous frequency and amplitude of the IMF component are calculated by Hilbert transform. For an IMF component c i′ (t), whose Hilbert transform is defined as: Among them, H[c i′ (t)] is the component c i′ Hilbert transform of (t); P is the principal value integral; is the kernel function of Hilbert transform; τ is the integral variable;

[0035] Use principal component analysis to reduce the dimension of the selected features to obtain a production feature data set, an equipment feature data set, and an environmental feature data set after dimension reduction;

[0036] The reduced-dimensional production feature data set, equipment feature data set and environmental feature data set are fused through a weighted model to form a comprehensive feature data set. The production feature data set is denoted as F1, the equipment feature data set is denoted as F2, and the environmental feature data set is denoted as F3.

[0037] The weighted model is: SQ=F1·ω1+F2·ω2+F3·ω3; wherein ω1 is the weight coefficient of the production feature data set; ω2 is the weight coefficient of the equipment feature data set; and ω3 is the weight coefficient of the environment feature data set.

[0038] Furthermore, the method for constructing the quality assessment model includes:

[0039] The data set is divided into training set, validation set and test set, and the model is trained and the model performance is evaluated; the sample set is a subset of the data set, and each sample set includes a historical comprehensive feature data set and the corresponding process parameter error of lithium carbonate production; a quality assessment model is constructed, and the model includes an input layer, a GRU layer, a fully connected layer and an output layer;

[0040] The input layer of the quality assessment model is used to input the historical comprehensive feature dataset. The number of neurons in the input layer should match the number of features in the historical comprehensive feature dataset. The GRU layer is used to process the historical comprehensive feature dataset, and the number of GRU layers and neurons are adjusted according to the complexity of the task. The fully connected layer is used to provide additional nonlinear transformation.

[0041] The output layer of the model is used to output the process parameter error of lithium carbonate production. The number of neurons in the output layer corresponds to the number of prediction targets. The prediction value is output through one neuron, and the identity function is used as the activation function. The transmission error prediction model is a quality assessment model.

[0042] Define the loss function of the model, and use the mean square error loss function to measure the difference between the predicted process parameter error of lithium carbonate production and the actual process parameter error of lithium carbonate production;

[0043] Use the training set to train the quality assessment model and update the model parameters through the back propagation algorithm to minimize the loss function; use the validation set to evaluate the performance of the quality assessment model by calculating the accuracy index;

[0044] The Adam optimization algorithm is used as the optimizer. The model is tuned according to the performance feedback of the validation set. The model parameters are adjusted until the performance is no longer significantly improved or the preset number of iterations is reached.

[0045] The test set is used to evaluate the performance of the model in the prediction task, and the trained transmission error prediction model is used to predict the current comprehensive feature data set to obtain the process parameter error of lithium carbonate production.

[0046] Furthermore, the method of comparing the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard includes:

[0047] If the error of the process parameters predicted for lithium carbonate production is less than the preset error threshold of the process parameters for lithium carbonate production, the product quality of lithium carbonate is determined to be up to standard;

[0048] If the error of the predicted process parameters for lithium carbonate production is greater than or equal to the preset process parameter error threshold for lithium carbonate production, it is determined that the product quality of the lithium carbonate does not meet the standards.

[0049] Furthermore, the method for adjusting the production parameters of lithium carbonate to reduce the process parameter errors according to the process parameter errors of lithium carbonate production includes:

[0050] The feedback control algorithm is used to adjust the production parameters of lithium carbonate, thereby reducing the process parameter error; the specific method steps are as follows:

[0051] S71, preset the target process parameter for lithium carbonate production as SP; use the sensor to measure the actual process parameter at time t' in real time as PV(t'); calculate the error between the target process parameter and the actual process parameter e(t') = SP-PV(t');

[0052] S72, using a feedback control algorithm to calculate the control output, and adjusting the temperature, pressure, liquid level and stirring speed in the reactor during the lithium carbonate production process based on the control output; the specific mathematical expression is:

[0053] Where u(t′) is the control output; K p is the proportional gain; K i is the integral gain; K d is the differential gain; is the integral term, which represents the integral of the error e(t′) from time 0 to the current time t′. The integral term is used to accumulate errors to reduce long-term errors; is the rate of change of the error at time t′, that is, the derivative of the error e(t′) with time; t′ is the time index;

[0054] S73. Adjust the production parameters of lithium carbonate according to the control output u(t′), limit the proportional gain of the production parameters, thereby affecting the process parameters and reducing the error between the target process parameters and the actual process parameters; apply the adjusted lithium carbonate production parameters to the system.

[0055] Furthermore, the method for limiting the proportional gain includes:

[0056] The proportional gain is limited by a proportional gain control model, which is: Among them, T de is the time required for control output; C is the type of control output; β is the adjustment coefficient.

[0057] The technical effects and advantages of an automatic control system for producing lithium carbonate of the present invention are as follows:

[0058] The present invention can decompose the original data into signal components of different time scales by gradually extracting IMF components; each IMF component represents information of a specific frequency or time scale; this feature extraction method helps to distinguish dynamic changes at different frequencies in the lithium carbonate production process, such as short-term fluctuations and long-term trends, thereby optimizing the production control strategy; by calculating the residual trend and IMF components, the noise components in the data can be effectively removed;

[0059] The Cauchy convergence criterion is used as the iteration stop condition to ensure that the decomposition process will not be over-performed, avoid introducing new errors, and retain the main characteristics of the signal. The threshold limiting formula uses parameters such as standard deviation and noise ratio to ensure the adaptability of noise processing, thereby enhancing the robustness of the system and the accuracy of the data; by calculating the relative energy of the IMF components and extracting the key components, the system can better identify which factors in the production process have the greatest impact on the overall system stability and performance, and regulate and optimize these key factors; this data-driven control method can further enhance the stability of the system and ensure a smooth and efficient production process;

[0060] Using Hilbert transform to analyze the instantaneous frequency and amplitude of IMF components can help discover the time-frequency characteristics of signals in the production process. In this way, we can gain an in-depth understanding of the dynamic changes in production parameters, which is very helpful for control system optimization. The temperature, pressure, liquid level and stirring speed are continuously adjusted through the feedback control algorithm, so that the actual process parameters gradually approach the target process parameters. The limitation of proportional gain can make the error adjustment process smoother and effectively shorten the time required for process parameter adjustment, thereby reducing the error between the target parameters and the actual parameters and improving production accuracy. The key parameters in the production process are adjusted in real time through the automatic control system, which reduces dependence on operators and improves the degree of automation of the production process. This not only reduces operating errors, but also optimizes the production process and improves overall production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a schematic diagram of the structure of an automatic control system for producing lithium carbonate according to the present invention;

[0062] Figure 2 The present invention is a schematic flow chart of an automatic control method for producing lithium carbonate. DETAILED DESCRIPTION

[0063] 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.

[0064] Embodiment 1

[0065] See also Figure 1 As shown, an automatic control system for producing lithium carbonate described in this embodiment includes:

[0066] Data collection module, used to collect production parameter data, equipment parameter data and environmental parameter data within time t;

[0067] A data processing module is used to pre-process the acquired production parameter data, equipment parameter data and environmental parameter data to obtain a production feature data set, an equipment feature data set and an environmental feature data set;

[0068] Extracting features from the production feature data set, the equipment feature data set, and the environment feature data set to obtain a comprehensive feature data set;

[0069] The quality assessment module is used to build a quality assessment model based on the comprehensive feature data set, and obtain the process parameter error of lithium carbonate production based on the quality assessment model; compare the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard; if the product quality meets the standard, the production terminal is automatically controlled to issue a qualified instruction to continue the production of lithium carbonate; if the product quality does not meet the standard, the production terminal is automatically controlled to issue a warning instruction and adjust the production parameters of lithium carbonate;

[0070] The strategy adjustment module is used to adjust the production parameters of lithium carbonate according to the process parameter errors of lithium carbonate production to reduce the process parameter errors; each module is connected by wired and / or wireless means.

[0071] Production parameter data include: temperature data, humidity data, pressure data, reactant concentration data, gas flow data, liquid flow data and PID controller parameter data in the hydrogenation reactor; equipment parameter data include: current data, voltage data, vibration data and cumulative operating time; environmental parameter data include: ambient temperature data, ambient humidity data and ambient air pressure data;

[0072] The production parameter data, equipment parameter data and environmental parameter data are acquired through installed sensors.

[0073] The method of preprocessing the acquired production parameter data, equipment parameter data and environmental parameter data to obtain a production feature data set, an equipment feature data set and an environmental feature data set includes:

[0074] The empirical mode decomposition method EMD is used to preprocess the production parameter data, equipment parameter data and environmental parameter data to obtain the production feature data set, equipment feature data set and environmental feature data set; the specific steps are as follows:

[0075] S31, presetting an original data sequence x(t), wherein the original data sequence includes a production feature data set, an equipment feature data set, and an environment feature data set;

[0076] S32, identifying all local extreme points; interpolating the original data sequence x(t) to find all local maximum points and local minimum points;

[0077] S33, construct upper and lower envelopes; perform cubic spline interpolation on the local maximum points to construct the upper envelope e max (t); Perform cubic spline interpolation on the local minimum points to construct the lower envelope e min (t);

[0078] S34, calculating the average envelope m(t);

[0079] The mathematical formula is: Where t is the index of the data point;

[0080] S35, calculating the intrinsic mode function IMF; subtracting the average envelope from the original data sequence to obtain the intrinsic mode function h(t)=x(t)-m(t);

[0081] Determine whether h(t) satisfies the IMF condition, specifically: in the entire data sequence, the number of extreme points and zero crossing points of h(t) must be equal or differ by at most one; at any time, the average value of the upper envelope formed by the local maximum points and the lower envelope formed by the local minimum points must be zero;

[0082] If h(t) satisfies the above conditions, no iteration is required; if h(t) does not satisfy the above conditions, h(t) is used as a new data sequence and steps S32-S35 are repeated until the IMF conditions are met;

[0083] S36, extracting IMF components; when h(t) satisfies the IMF condition, taking h(t) as the first IMF component c1(t); subtracting c1(t) from the original data sequence to obtain the residual trend r1(t); repeating the iteration to extract the next IMF component until the residual trend r1(t) satisfies the stopping criterion and stops the iteration;

[0084] After the above steps, all IMF components and residual trends are extracted, integrated and reconstructed to obtain production feature data sets, equipment feature data sets and environmental feature data sets.

[0085] Methods for stopping iterations when the residual trend meets the stopping criteria include:

[0086] The Cauchy convergence criterion is used as the stopping criterion, and the iteration is stopped when the residual trend satisfies the stopping criterion; the specific mathematical formula of the Cauchy convergence criterion is: Among them, h k-1 (t) is the IMF component generated by the k-1th iteration, h k (t) is the IMF component generated by the kth iteration; N is the total number of data points; t is the index of the data point; when SD kWhen it is less than the preset threshold SD′, the iteration process stops.

[0087] The method for determining the value of the preset threshold SD′ includes:

[0088] The threshold SD′ is determined by a threshold limiting formula; the threshold limiting formula is specifically: Among them, σ is the standard deviation of the data sequence; N is the total number of data points; η is the noise ratio in the data sequence; α1 is the adjustment coefficient of the standard deviation; α2 is the adjustment coefficient of the noise ratio.

[0089] For example, the standard deviation σ of the data sequence is 20, the total number of data points N is 16, the noise ratio η in the data sequence is 0.5, the adjustment coefficient α1 of the standard deviation is 0.4, and the adjustment coefficient α2 of the noise ratio is 0.6, then the threshold SD′ is 2.3.

[0090] The method of extracting features from the production feature data set, the equipment feature data set and the environment feature data set to obtain a comprehensive feature data set includes:

[0091] For each IMF component c i′ (t), whose energy is defined as: Where, is the value of the IMF component at time t;

[0092] Calculate the relative energy percentage of each IMF component using the formula: Where, is the relative energy percentage of the i′th IMF component; m′ is the total number of IMF components; is the sum of the energies of all IMF components; j is the index of traversing all IMF components;

[0093] Based on the calculated relative energy percentage, the IMF component with the largest proportion is selected as the feature; the instantaneous frequency and amplitude of the IMF component are calculated by Hilbert transform. For an IMF component c i′ (t), whose Hilbert transform is defined as: Among them, H[c i′ (t)] is the component c i′ Hilbert transform of (t); P is the principal value integral; is the kernel function of Hilbert transform; τ is the integral variable;

[0094] Use principal component analysis to reduce the dimension of the selected features to obtain a production feature data set, an equipment feature data set, and an environmental feature data set after dimension reduction;

[0095] The reduced-dimensional production feature data set, equipment feature data set and environmental feature data set are fused through a weighted model to form a comprehensive feature data set. The production feature data set is denoted as F1, the equipment feature data set is denoted as F2, and the environmental feature data set is denoted as F3.

[0096] The weighted model is: SQ=F1·ω1+F2·ω2+F3·ω3; wherein ω1 is the weight coefficient of the production feature data set; ω2 is the weight coefficient of the equipment feature data set; and ω3 is the weight coefficient of the environment feature data set.

[0097] The methods for building the quality assessment model include:

[0098] The data set is divided into training set, validation set and test set, and the model is trained and the model performance is evaluated; the sample set is a subset of the data set, and each sample set includes a historical comprehensive feature data set and the corresponding process parameter error of lithium carbonate production; a quality assessment model is constructed, and the model includes an input layer, a GRU layer, a fully connected layer and an output layer;

[0099] The input layer of the quality assessment model is used to input the historical comprehensive feature dataset. The number of neurons in the input layer should match the number of features in the historical comprehensive feature dataset. The GRU layer is used to process the historical comprehensive feature dataset, and the number of GRU layers and neurons are adjusted according to the complexity of the task. The fully connected layer is used to provide additional nonlinear transformation.

[0100] The output layer of the model is used to output the process parameter error of lithium carbonate production. The number of neurons in the output layer corresponds to the number of prediction targets. The prediction value is output through one neuron, and the identity function is used as the activation function. The transmission error prediction model is a quality assessment model.

[0101] Define the loss function of the model, and use the mean square error loss function to measure the difference between the predicted process parameter error of lithium carbonate production and the actual process parameter error of lithium carbonate production;

[0102] Use the training set to train the quality assessment model and update the model parameters through the back propagation algorithm to minimize the loss function; use the validation set to evaluate the performance of the quality assessment model by calculating the accuracy index;

[0103] The Adam optimization algorithm is used as the optimizer. The model is tuned according to the performance feedback of the validation set. The model parameters are adjusted until the performance is no longer significantly improved or the preset number of iterations is reached.

[0104] The test set is used to evaluate the performance of the model in the prediction task, and the trained transmission error prediction model is used to predict the current comprehensive feature data set to obtain the process parameter error of lithium carbonate production.

[0105] The method of comparing the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard includes:

[0106] If the error of the process parameters predicted for lithium carbonate production is less than the preset error threshold of the process parameters for lithium carbonate production, the product quality of lithium carbonate is determined to be up to standard;

[0107] If the error of the predicted process parameters for lithium carbonate production is greater than or equal to the preset process parameter error threshold for lithium carbonate production, it is determined that the product quality of the lithium carbonate does not meet the standards.

[0108] According to the process parameter error of lithium carbonate production, the method of adjusting the production parameters of lithium carbonate to reduce the process parameter error includes:

[0109] The feedback control algorithm is used to adjust the temperature, pressure, liquid level and stirring speed in the reactor during the lithium carbonate production process, thereby reducing the process parameter error; the specific method steps are as follows:

[0110] S71, preset the target process parameter for lithium carbonate production as SP; use the sensor to measure the actual process parameter at time t' in real time as PV(t'); calculate the error between the target process parameter and the actual process parameter e(t') = SP-PV(t');

[0111] S72, using a feedback control algorithm to calculate the control output, and adjusting the temperature, pressure, liquid level and stirring speed in the reactor during the lithium carbonate production process based on the control output; the specific mathematical expression is:

[0112] Where u(t′) is the control output; K p is the proportional gain; K i is the integral gain; K d is the differential gain; is the integral term, which represents the integral of the error e(t′) from time 0 to the current time t′. The integral term is used to accumulate errors to reduce long-term errors; is the rate of change of the error at time t′, that is, the derivative of the error e(t′) with time; t′ is the time index;

[0113] S73. Adjust the temperature, pressure, liquid level and stirring speed in the reactor according to the control output u(t′), limit the proportional gain, thereby affecting the process parameters and reducing the error between the target process parameters and the actual process parameters; apply the adjusted temperature, pressure, liquid level and stirring speed in the reactor to the system.

[0114] Methods for limiting the proportional gain include:

[0115] The proportional gain is limited by a proportional gain control model, which is: Among them, T de is the time required for control output; C is the type of control output; β is the adjustment coefficient to ensure K p Within a reasonable range.

[0116] For example, the time required to control the output is T de is 15 seconds, there are 4 types of control output C, the adjustment coefficient is 0.2, and the error change rate at time t' is 0.8, then the proportional gain K p About 0.42.

[0117] The preset process parameter error threshold for lithium carbonate production is to collect process parameter errors of different lithium carbonate productions through the automatic control production terminal, and take the average value of the process parameter errors of multiple lithium carbonate productions as the preset process parameter error threshold for lithium carbonate production;

[0118] In this embodiment, by gradually extracting IMF components, EMD can decompose the original data into signal components of different time scales; each IMF component represents information of a specific frequency or time scale; this feature extraction method helps to distinguish dynamic changes at different frequencies in the lithium carbonate production process, such as short-term fluctuations and long-term trends, thereby optimizing the production control strategy; by calculating the residual trend and IMF components, the noise components in the data can be effectively removed;

[0119] The Cauchy convergence criterion is used as the iteration stop condition to ensure that the decomposition process will not be over-performed, avoid introducing new errors, and retain the main characteristics of the signal. The threshold limiting formula uses parameters such as standard deviation and noise ratio to ensure the adaptability of noise processing, thereby enhancing the robustness of the system and the accuracy of the data; by calculating the relative energy of the IMF components and extracting the key components, the system can better identify which factors in the production process have the greatest impact on the overall system stability and performance, and regulate and optimize these key factors; this data-driven control method can further enhance the stability of the system and ensure a smooth and efficient production process;

[0120] Using Hilbert transform to analyze the instantaneous frequency and amplitude of IMF components can help discover the time-frequency characteristics of signals in the production process. In this way, we can gain an in-depth understanding of the dynamic changes in production parameters, which is very helpful for control system optimization. The temperature, pressure, liquid level and stirring speed are continuously adjusted through the feedback control algorithm, so that the actual process parameters gradually approach the target process parameters. The limitation of proportional gain can make the error adjustment process smoother and effectively shorten the time required for process parameter adjustment, thereby reducing the error between the target parameters and the actual parameters and improving production accuracy. The key parameters in the production process are adjusted in real time through the automatic control system, which reduces dependence on operators and improves the degree of automation of the production process. This not only reduces operating errors, but also optimizes the production process and improves overall production efficiency.

[0121] Embodiment 2

[0122] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, which provides an automatic control method for producing lithium carbonate, comprising:

[0123] S1, collect production parameter data, equipment parameter data and environmental parameter data within time t;

[0124] S2. Preprocess the acquired production parameter data, equipment parameter data, and environmental parameter data to obtain a production feature data set, an equipment feature data set, and an environmental feature data set;

[0125] Extracting features from the production feature data set, the equipment feature data set, and the environment feature data set to obtain a comprehensive feature data set;

[0126] S3. Build a quality assessment model based on the comprehensive feature data set, and obtain the process parameter error of lithium carbonate production based on the quality assessment model; compare the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard; if the product quality meets the standard, the production terminal is automatically controlled to issue a qualified instruction to continue the production of lithium carbonate; if the product quality does not meet the standard, the production terminal is automatically controlled to issue a warning instruction and adjust the production parameters of lithium carbonate;

[0127] S4. According to the process parameter error of lithium carbonate production, the production parameters of lithium carbonate are adjusted to reduce the process parameter error.

[0128] Since the electronic device introduced in this embodiment is an electronic device used for implementing an automatic control system for producing lithium carbonate in the embodiment of the present application, based on an automatic control system for producing lithium carbonate introduced in the embodiment of the present application, a person skilled in the art can understand the specific implementation of the electronic device of the present embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not described in detail here. As long as a person skilled in the art implements an electronic device used for an automatic control system for producing lithium carbonate in the embodiment of the present application, it belongs to the scope protected by the present application.

[0129] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0130] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. An automatic control system for producing lithium carbonate, characterized in that: include: Data collection module, used to collect production parameter data, equipment parameter data and environmental parameter data within time t; A data processing module is used to pre-process the acquired production parameter data, equipment parameter data and environmental parameter data to obtain a production feature data set, an equipment feature data set and an environmental feature data set; Extracting features from the production feature data set, the equipment feature data set, and the environment feature data set to obtain a comprehensive feature data set; The quality assessment module is used to build a quality assessment model based on the comprehensive feature data set, and obtain the process parameter error of lithium carbonate production based on the quality assessment model; compare the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard; if the product quality meets the standard, the production terminal is automatically controlled to issue a qualified instruction to continue the production of lithium carbonate; if the product quality does not meet the standard, the production terminal is automatically controlled to issue a warning instruction and adjust the production parameters of lithium carbonate; The strategy adjustment module is used to adjust the production parameters of lithium carbonate according to the process parameter errors of lithium carbonate production to reduce the process parameter errors; each module is connected by wired and / or wireless means.

2. An automatic control system for producing lithium carbonate according to claim 1, characterized in that, The production parameter data include: temperature data, humidity data, pressure data, reactant concentration data, gas flow data, liquid flow data and PID controller parameter data in the hydrogenation reactor; the equipment parameter data include: current data, voltage data, vibration data and cumulative running time; the environmental parameter data include: ambient temperature data, ambient humidity data and ambient air pressure data; The production parameter data, equipment parameter data and environmental parameter data are acquired through sensors and control terminals.

3. An automatic control system for producing lithium carbonate according to claim 2, characterized in that, The method of preprocessing the acquired production parameter data, equipment parameter data and environmental parameter data to obtain a production feature data set, an equipment feature data set and an environmental feature data set comprises: The empirical mode decomposition method EMD is used to preprocess the production parameter data, equipment parameter data and environmental parameter data to obtain the production feature data set, equipment feature data set and environmental feature data set; the specific steps are as follows: S31, presetting an original data sequence x(t), wherein the original data sequence includes a production feature data set, an equipment feature data set, and an environment feature data set; S32, identifying all local extreme points; interpolating the original data sequence x(t) to find all local maximum points and local minimum points; S33, construct upper and lower envelopes; perform cubic spline interpolation on the local maximum points to construct the upper envelope e max (t); Perform cubic spline interpolation on the local minimum points to construct the lower envelope e min (t); S34, calculating the average envelope m(t); The mathematical formula is: Where t is the index of the data point; S35, calculating the intrinsic mode function IMF; subtracting the average envelope from the original data sequence to obtain the intrinsic mode function h(t)=x(t)-m(t); Determine whether h(t) satisfies the IMF condition, specifically: in the entire data sequence, the number of extreme points and zero crossing points of h(t) must be equal or differ by at most one; at any time, the average value of the upper envelope formed by the local maximum points and the lower envelope formed by the local minimum points must be zero; If h(t) satisfies the above conditions, no iteration is required; if h(t) does not satisfy the above conditions, h(t) is used as a new data sequence and steps S32-S35 are repeated until the IMF conditions are met; S36, extracting IMF components; when h(t) satisfies the IMF condition, taking h(t) as the first IMF component c1(t); subtracting c1(t) from the original data sequence to obtain the residual trend r1(t); repeating the iteration to extract the next IMF component until the residual trend r1(t) satisfies the stopping criterion and stops the iteration; After the above steps, all IMF components and residual trends are extracted, integrated and reconstructed to obtain production feature data sets, equipment feature data sets and environmental feature data sets.

4. An automatic control system for producing lithium carbonate according to claim 3, characterized in that: The method for stopping iteration when the residual trend meets the stopping criterion includes: The Cauchy convergence criterion is used as the stopping criterion, and the iteration is stopped when the residual trend satisfies the stopping criterion; the specific mathematical formula of the Cauchy convergence criterion is: Among them, h k-1 (t) is the IMF component generated by the k-1th iteration, h k (t) is the IMF component generated by the kth iteration; N is the total number of data points; t is the index of the data point; when SD k When it is less than the preset threshold SD′, the iteration process stops.

5. An automatic control system for producing lithium carbonate according to claim 4, characterized in that: The method for stopping iteration when the residual trend meets the stopping criterion includes: The threshold SD′ is determined by a threshold limiting formula; the threshold limiting formula is specifically: Among them, σ is the standard deviation of the data sequence; N is the total number of data points; η is the noise ratio in the data sequence; α1 is the adjustment coefficient of the standard deviation; α2 is the adjustment coefficient of the noise ratio.

6. An automatic control system for producing lithium carbonate according to claim 5, characterized in that: The method of extracting features from the production feature data set, the equipment feature data set and the environment feature data set to obtain a comprehensive feature data set includes: For each IMF component c i′ (t), whose energy is defined as: Where, is the value of the IMF component at time t; Calculate the relative energy percentage of each IMF component using the formula: Where, is the relative energy percentage of the i′th IMF component; m′ is the total number of IMF components; is the sum of the energies of all IMF components; j is the index of traversing all IMF components; Based on the calculated relative energy percentage, the IMF component with the largest proportion is selected as the feature; the instantaneous frequency and amplitude of the IMF component are calculated by Hilbert transform. For an IMF component c i′ (t), whose Hilbert transform is defined as: Among them, H[c i′ (t)] is the component c i′ Hilbert transform of (t); P is the principal value integral; is the kernel function of Hilbert transform; τ is the integral variable; Use principal component analysis to reduce the dimension of the selected features to obtain a production feature data set, an equipment feature data set, and an environmental feature data set after dimension reduction; The reduced-dimensional production feature data set, equipment feature data set and environmental feature data set are fused through a weighted model to form a comprehensive feature data set. The production feature data set is denoted as F1, the equipment feature data set is denoted as F2, and the environmental feature data set is denoted as F3. The weighted model is: SQ=F1·ω1+F2·ω2+F3·ω3; wherein ω1 is the weight coefficient of the production feature data set; ω2 is the weight coefficient of the equipment feature data set; and ω3 is the weight coefficient of the environment feature data set.

7. An automatic control system for producing lithium carbonate according to claim 6, characterized in that: The method for constructing the quality assessment model includes: The data set is divided into training set, validation set and test set, and the model is trained and the model performance is evaluated; the sample set is a subset of the data set, and each sample set includes a historical comprehensive feature data set and the corresponding process parameter error of lithium carbonate production; a quality assessment model is constructed, and the model includes an input layer, a GRU layer, a fully connected layer and an output layer; The input layer of the quality assessment model is used to input the historical comprehensive feature dataset. The number of neurons in the input layer should match the number of features in the historical comprehensive feature dataset. The GRU layer is used to process the historical comprehensive feature dataset, and the number of GRU layers and neurons are adjusted according to the complexity of the task. The fully connected layer is used to provide additional nonlinear transformation. The output layer of the model is used to output the process parameter error of lithium carbonate production. The number of neurons in the output layer corresponds to the number of prediction targets. The prediction value is output through one neuron, and the identity function is used as the activation function. The transmission error prediction model is a quality assessment model. Define the loss function of the model, and use the mean square error loss function to measure the difference between the predicted process parameter error of lithium carbonate production and the actual process parameter error of lithium carbonate production; Use the training set to train the quality assessment model and update the model parameters through the back propagation algorithm to minimize the loss function; use the validation set to evaluate the performance of the quality assessment model by calculating the accuracy index; The Adam optimization algorithm is used as the optimizer. The model is tuned according to the performance feedback of the validation set. The model parameters are adjusted until the performance is no longer significantly improved or the preset number of iterations is reached. The test set is used to evaluate the performance of the model in the prediction task, and the trained transmission error prediction model is used to predict the current comprehensive feature data set to obtain the process parameter error of lithium carbonate production.

8. An automatic control system for producing lithium carbonate according to claim 7, characterized in that: The method of comparing the predicted process parameter error of lithium carbonate production with the preset process parameter error threshold of lithium carbonate production to determine whether the product quality of lithium carbonate meets the standard includes: If the error of the process parameters predicted for lithium carbonate production is less than the preset error threshold of the process parameters for lithium carbonate production, the product quality of lithium carbonate is determined to be up to standard; If the error of the predicted process parameters for lithium carbonate production is greater than or equal to the preset process parameter error threshold for lithium carbonate production, it is determined that the product quality of the lithium carbonate does not meet the standards.

9. An automatic control system for producing lithium carbonate according to claim 8, characterized in that: The method for adjusting the production parameters of lithium carbonate to reduce the process parameter errors according to the process parameter errors of lithium carbonate production includes: The feedback control algorithm is used to adjust the production parameters of lithium carbonate, thereby reducing the process parameter error; the specific method steps are as follows: S71, preset the target process parameter for lithium carbonate production as SP; use the sensor to measure the actual process parameter at time t' in real time as PV(t'); calculate the error between the target process parameter and the actual process parameter e(t') = SP-PV(t'); S72. Calculate the control output using a feedback control algorithm, and adjust the production parameters of lithium carbonate based on the control output; the specific mathematical expression is: Where u(t′) is the control output; K p is the proportional gain; Ki is the integral gain; K d is the differential gain; is the integral term; is the rate of change of the error at time t′; t′ is the time index; S73. Adjust the production parameters of lithium carbonate according to the control output u(t′), limit the proportional gain of the production parameters, thereby affecting the process parameters and reducing the error between the target process parameters and the actual process parameters; apply the adjusted lithium carbonate production parameters to the system.

10. An automatic control system for producing lithium carbonate according to claim 9, characterized in that: The method for limiting the proportional gain of the production parameter comprises: The proportional gain is limited by a proportional gain control model, which is: Among them, T de is the time required for control output; C is the type of control output; β is the adjustment coefficient.

Images