Coal ash melting point prediction method and system, electronic equipment and storage medium

Abstract

The invention discloses a coal ash melting point prediction method and system, electronic equipment and a storage medium, and belongs to the technical field of coal ash melting point prediction. The method comprises the following steps: acquiring historical data of coal ash components and coal ash melting point temperature, and preprocessing to obtain a training sample; constructing a coal ash melting point prediction initial model based on a BP neural network according to the training sample; optimizing an initial weight and a threshold value of the coal ash melting point prediction initial model by adopting an improved particle swarm algorithm to obtain a coal ash melting point prediction model; and acquiring real-time data of coal ash components, and inputting the real-time data into the coal ash melting point prediction model for prediction to obtain a predicted coal ash melting point. Neural network training is completed by combining the improved particle swarm optimization and the BP neural network, the model has the advantages of high prediction accuracy, high training precision, high practicability and the like, accurate prediction of the coal ash melting point is realized, and a new effective method is provided for prediction of the coal ash melting point.

Description

Technical Field

[0001] The present invention relates to the technical field of coal ash melting point prediction, and in particular to a coal ash melting point prediction method, system, electronic equipment and storage medium. Background Art

[0002] During the operation of power station boilers, slagging and coking will occur in the tail flue of the water-cooled wall. This is due to the low ash melting point of coal powder. Therefore, the influence of the ash melting point of the coal used should be considered in the design and operation of the boiler. The chemical composition of coal ash is the main factor affecting the ash melting point. The deformation temperature (DT), softening temperature (ST), and flow temperature (FT) are generally used to characterize the fusibility of coal. However, its internal change process and the mutual influence between the components have not been clearly explored. There is no universal and accurate calculation method for the calculation of coal ash melting point.

[0003] At present, the common experimental determination method is the national standard GB / T 219-2008 "Determination of coal ash fusibility", but the test process specified in this standard is complicated and requires special instruments and equipment, which most power plants and small and medium-sized enterprises do not have. Therefore, establishing a mathematical model that can be used to predict the ash melting point of coal ash can solve this problem well.

[0004] The empirical regression formula is a method for predicting the melting point of coal ash. Its theoretical basis is to perform linear regression analysis based on certain characteristics of coal ash components. This method is relatively simple. The specific formula is as follows:

[0005] (1) When SiO2>60%, the calculation formula of coal ash melting point temperature is:

[0006] ST=10.75*SiO2+13.03*Al2O3-5.28*Fe2O3-5.88*CaO-10.28*MgO+3.75*a+453

[0007] (2) When SiO2≤60% and Al2O3>30%, the calculation formula of coal ash melting point temperature is:

[0008] ST=69.94*SiO2+71.01*Al2O3+65.23*Fe2O3+12.16*CaO+68.31*MgO+67.19*a-5485.7

[0009] (3) When SiO2≤60%, Al2O3≤30%, and Fe2O3≤15%, the calculation formula of the ash melting point temperature is:

[0010] ST=92.55*SiO2+97.83*Al2O3+84.52*Fe2O3+83.67*CaO+81.04*MgO+91.92*a-7891

[0011] (4) When SiO2≤60%, Al2O3≤30%, and Fe2O3>15%, the calculation formula of coal ash melting point temperature is:

[0012] ST=1531-3.01*SiO2+5.08*Al2O3-8.02*Fe2O3-9.69*CaO-5.86*MgO-3.99*a

[0013] Where a = 100-(SiO2+Al2O3+Fe2O3+CaO+MgO)

[0014] The above linear regression method is relatively simple in predicting the ash melting point, but in practice, large deviations sometimes occur, so it is necessary to find a more accurate prediction method. Summary of the invention

[0015] The present invention solves the problem of accuracy of the prediction results of the existing prediction of coal ash melting point and provides a coal ash melting point prediction method, system, electronic device and storage medium. The prediction method predicts the coal ash melting point by constructing a neural network model based on an improved particle swarm algorithm.

[0016] In order to achieve the above object, the present invention provides the following technical solutions:

[0017] The present invention provides a method for predicting the ash melting point, comprising the following steps:

[0018] Obtain real-time data on coal ash composition, input it into a coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point;

[0019] The coal ash melting point prediction model is constructed by the following steps:

[0020] Obtain historical data on coal ash composition and coal ash melting point temperature, and perform preprocessing to obtain training samples;

[0021] According to the training samples, the initial model for coal ash melting point prediction based on BP neural network is constructed;

[0022] The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction, and the coal ash melting point prediction model is obtained.

[0023] The fly ash components include nine oxides, namely silicon dioxide, aluminum oxide, iron oxide, sulfur trioxide, calcium oxide, potassium oxide, sodium oxide, titanium dioxide and magnesium oxide.

[0024] The historical data of coal ash composition and coal ash melting point temperature are obtained and the training samples are pre-processed, specifically:

[0025] The historical data of coal ash composition is normalized, and the normalization formula y = (xx min ) / (x max -x min ) is standardized and mapped to [0, 1];

[0026] Among them, y is the normalized data, x is the sample data; min is the minimum value of the sample data, x max is the maximum value of the sample data.

[0027] The BP neural network is used to construct an initial model for predicting the ash melting point; specifically:

[0028] According to the training samples, the initial weights and thresholds of the BP neural network and the number of iterations are set to train the BP neural network. The training method of the BP neural network adopts the gradient training method.

[0029] The BP neural network is selected to construct the initial model for predicting the coal ash melting point. A three-layer network structure is adopted. According to the types of coal ash components in the training samples, the input layer is determined to be 9 nodes, corresponding to the 9 oxides of coal ash components; the output layer is 1 node, which is the coal ash melting point temperature; the transfer function selected for the hidden layer input and output is the sigmoid function, and the hidden layer has 15 nodes.

[0030] The selection of the number of nodes in the hidden layer includes:

[0031] According to the formula Preliminarily determine the number of nodes in the hidden layer;

[0032] Where: n is the number of nodes in the hidden layer, p is the number of nodes in the output layer, q is the number of nodes in the input layer, and m is a constant between 0 and 10;

[0033] Some training samples are selected for training, and the change of the mean square error of the BP neural network is observed by increasing or decreasing the number of nodes; when the number of nodes increases and the mean square error no longer decreases, the number of nodes at this time is determined to be the number of nodes in the hidden layer.

[0034] The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for predicting the ash melting point, and the ash melting point prediction model is obtained, which is specifically:

[0035] S1: Assign each particle a random initial position and speed. The initial position of the particle is a random number between [0, 1];

[0036] S2: The mean square error between the predicted result and the actual value is used as the fitness function to calculate the fitness value of each particle;

[0037] S3: For each particle, compare the fitness value of its current position with the fitness value corresponding to its historical best position. If the fitness value of the current position is higher, update the historical best position with the current position.

[0038] S4: for each particle, compare the fitness value of its current position with the fitness value corresponding to its global best position. If the fitness value of the current position is higher, update the global best position with the current position.

[0039] S5: Update the speed and position of each particle according to the following formula

[0040]

[0041] in: is the optimized speed of particle i; is the current speed of particle i; w is the inertia factor; c1 and c2 are learning factors; r1 and r2 are random numbers between (0,1); pbest id is the best historical position of particle i; gbest d is the global optimal position of the particle swarm; is the current position of particle i;

[0042] S6: If the end condition is not met, return to S2 to continue iteration; if the end condition is met, the iteration ends, and the position and speed of the optimal particle are assigned to the weight and threshold of the BP neural network to obtain the coal ash melting point prediction model.

[0043] The method also includes selecting some of the training samples as test samples to input into the coal ash melting point prediction model, and evaluating the prediction effect of the coal ash melting point prediction model, specifically:

[0044] The test samples were input into the coal ash melting point prediction model for prediction, and the predicted values ​​were compared with the true values. The mean square error (MSE) and the coefficient of determination (R 2 ) is evaluated, and the expression is as follows:

[0045]

[0046] Where n is the number of samples, y i is the true value of the i-th sample, y i ' is the predicted value of the i-th sample;

[0047]

[0048] Where n is the number of samples, y i is the true value of the i-th sample, yi ' is the predicted value of the i-th sample, is the average of all true values.

[0049] The present invention also provides a coal ash melting point prediction system, comprising the following modules:

[0050] A prediction module is used to obtain real-time data of coal ash composition, input the coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point;

[0051] A data processing module is used to obtain historical data of coal ash composition and coal ash melting point temperature, and perform pre-processing and post-training samples;

[0052] A model building module is used to build an initial model for coal ash melting point prediction based on BP neural network according to training samples;

[0053] The model optimization module is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction by using an improved particle swarm algorithm to obtain a coal ash melting point prediction model.

[0054] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned coal ash melting point prediction method when executing the computer program.

[0055] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned coal ash melting point prediction method are implemented.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The present invention proposes a coal ash melting point prediction method based on the combination of an improved particle swarm algorithm and a neural network model. The present invention not only improves the accuracy and precision of the prediction in terms of the prediction of the coal ash melting point, but also enhances the practicality and computational efficiency of the model. By combining the global search capability of the improved particle swarm algorithm and the local optimization capability of the neural network, the present invention can more accurately mine and utilize the complex relationship between the chemical composition and the melting point of the coal ash, thereby significantly improving the prediction accuracy of the coal ash melting point. By using the global optimal value provided by the improved particle swarm algorithm as the initial value and further optimizing it using the nonlinear mapping capability of the neural network, it is possible to improve the computational efficiency and shorten the prediction time while ensuring the prediction accuracy. The coal ash melting point prediction model established by the present invention is based on the chemical composition of the coal ash and has clear physical meaning and practical application value. The prediction results are highly consistent with the actual results, providing a new and more effective method for the prediction of the coal ash melting point, which has broad industrial application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the following briefly introduces the drawings required for use in the embodiment:

[0059] Figure 1 This is a flow chart of coal ash melting point prediction based on improved particle swarm algorithm neural network of the present invention;

[0060] Figure 2 This is a schematic diagram of the process flow of the improved particle swarm algorithm neural network model of the present invention;

[0061] Figure 3 A neural network topology diagram of the improved particle swarm algorithm neural network model used for coal ash melting point modeling in the present invention;

[0062] Figure 4 A comparison chart of the results of coal ash melting point prediction in the embodiment;

[0063] Figure 5 It is a structural schematic diagram of the coal ash melting point prediction system based on the improved particle swarm algorithm neural network of the present invention;

[0064] Figure 6 The figure is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be further described below in conjunction with the accompanying drawings:

[0066] The improved particle swarm algorithm and neural network model adopted by the present invention. The improved particle swarm algorithm is a global optimization problem algorithm that has been widely used in recent years. It requires fewer parameters in the algorithm, is simple to write programs, and the iterative operation process is convenient and easy to understand. It can be solved by iterative cycles only by updating the particle's own speed direction, which is relatively easy to implement. Neural networks are a good way to solve the "black box" problem in the modeling process. The neural network is trained by the steepest descent method, and is continuously trained by establishing a nonlinear mapping mathematical model between chemical composition and ash melting point, adjusting the network connection weight coefficient and threshold, reducing the network error value, and achieving the expected error.

[0067] like Figure 1 As shown, a method for predicting the ash melting point is provided. The method is based on a BP neural network with an improved particle swarm algorithm and specifically includes the following steps:

[0068] Obtain real-time data on coal ash composition, input it into a coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point;

[0069] The coal ash melting point prediction model is constructed by the following steps:

[0070] Obtain historical data on coal ash composition and coal ash melting point temperature, and perform preprocessing on training samples;

[0071] According to the training samples, the initial model for coal ash melting point prediction based on BP neural network is constructed;

[0072] The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction, and the coal ash melting point prediction model is obtained.

[0073] The present invention completes the neural network training by combining the improved particle swarm algorithm with the BP neural network. The model has the advantages of high prediction accuracy, high training precision, strong practicality, etc., realizes the accurate prediction of the coal ash melting point, and provides a new and effective method for the prediction of the coal ash melting point.

[0074] The present invention is described in detail below with reference to the accompanying drawings:

[0075] (1) The historical data of coal ash composition and coal ash melting point temperature are obtained and preprocessed to obtain training samples, specifically:

[0076] The historical data of coal ash composition is normalized, and the normalization formula y = (xx min ) / (x max -x min ) is normalized and mapped to [0, 1].

[0077] Among them, y is the normalized data, x is the sample data; min is the minimum value of the sample data, x max is the maximum value of the sample data.

[0078] The coal ash components in the sample data include 9 oxides, namely silicon dioxide (SiO2), aluminum oxide (Al2O3), iron oxide (Fe2O3), sulfur trioxide (SO3), calcium oxide (CaO), potassium oxide (K2O), sodium oxide (Na2O), titanium dioxide (TiO2) and magnesium oxide (MgO).

[0079] To ensure the comprehensiveness of the sample, the coal ash sources include single coal and mixed coal. At the same time, the selection of sample data volume should be moderate. Too little sample data may lead to a large deviation between the prediction results and the actual results, while too much sample data may lead to a decrease in the learning rate.

[0080] (2) The initial model for predicting the ash melting point based on the BP neural network is constructed according to the training samples, specifically:

[0081] The initial weights, thresholds and iteration times of the BP neural network are set, and the BP neural network is trained with training samples. The training method of the BP neural network adopts the gradient training method.

[0082] like Figure 3As shown in the figure, the initial model for predicting the coal ash melting point uses a BP neural network with a three-layer network structure. According to the types of coal ash components in the training samples, the input layer is determined to be 9 nodes, corresponding to the 9 oxides of the coal ash components; the output layer is 1 node, which is the coal ash melting point temperature.

[0083] Choosing a reasonable number of nodes in the hidden layer is crucial to the performance and generalization ability of the BP neural network. The transfer function selected for the hidden layer input and output is the sigmoid function, and the formula is as follows:

[0084]

[0085] The number of nodes in the hidden layer can be selected, including:

[0086] First, according to the formula Preliminary determination of the number of nodes in the hidden layer

[0087] Where n is the number of nodes in the hidden layer, p is the number of nodes in the output layer, q is the number of nodes in the input layer, and m is a constant between 0 and 10.

[0088] Then, a small number of dragon training samples are selected for training, and the change of MSE (mean square error) of the BP neural network is observed by increasing or decreasing the number of nodes. When the number of nodes increases and the MSE no longer decreases, the number of nodes at this time is determined to be the most appropriate number of nodes. The preferred hidden layer of the present invention is 15 nodes.

[0089] (3) Figure 2 As shown, the improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for predicting the ash melting point, and the ash melting point prediction model is obtained, which is specifically:

[0090] S1: Each particle represents a BP neural network, and the particle swarm is initialized (the particle swarm has n particles in total), and the particles are given random initial positions and speeds. The initial position represents the weight and bias of the neural network, and the initial position of the particle is a random number between [0, 1]; the speed represents the search intensity in a certain direction;

[0091] S2: Calculate the fitness value of each particle according to the fitness function; the fitness function can choose accuracy, recall rate, etc. as measurement indicators. Therefore, the mean square error between the predicted result and the actual value is selected as the fitness function;

[0092] S3: For each particle, compare the fitness value of its current position with the fitness value corresponding to its best historical position (pbest). If the fitness value of the current position is higher, update the best historical position with the current position.

[0093] S4: for each particle, compare the fitness value of its current position with the fitness value corresponding to its global best position (gbest). If the fitness value of the current position is higher, update the global best position with the current position.

[0094] S5: Update the speed and position of each particle according to the following formula

[0095]

[0096] in: is the optimized speed of particle i; is the current speed of particle i; w is the inertia factor; c1 and c2 are learning factors; r1 and r2 are random numbers between (0,1); pbest id is the best historical position of particle i; gbest d is the global optimal position of the particle swarm; is the current position of particle i;

[0097] S6: If the end condition is not met, return to S2 to continue iteration. If the end condition is met, the iteration ends. The global best position (gbest) is the global optimal solution. The position and speed of the optimal particle are assigned to the weights and thresholds of the BP neural network to obtain the coal ash melting point prediction model.

[0098] (4) Select some of the training samples as test samples and input them into the coal ash melting point prediction model to evaluate the prediction effect of the coal ash melting point prediction model. Specifically:

[0099] The training process of the sample is carried out according to the gradient descent method. After the set accuracy is met, the test sample is input into the coal ash melting point prediction model for prediction. The predicted value is compared with the true value, and the mean square error (MSE) and determination coefficient (R 2 ) is evaluated, and the expression is as follows:

[0100]

[0101] Where n is the number of samples, y i is the true value of the i-th sample, y i ' is the predicted value of the i-th sample.

[0102]

[0103] Where n is the number of samples, y i is the true value of the i-th sample, y i ' is the predicted value of the i-th sample, is the average of all true values.

[0104] Example:

[0105] 150 groups of sample data were selected, corresponding to 150 different types of coal, including single coal and mixed coal. The sample data were normalized and used as training samples, and 25 groups of sample data were taken as test samples.

[0106] The relevant parameters of the neural network are set, the neural network learning rate is 0.03, the number of iterations is 50,000, and the initial model for coal ash melting point prediction is constructed; the relevant parameters of the improved particle swarm algorithm are set, the number of particles is 100, the particle position is a random number between [0, 1], the particle speed is 1, and the maximum number of iterations is 500;

[0107] The test results of 150 sets of sample data are as follows Figure 4 As shown, the predicted value is compared with the true value to obtain the training sample error; then 25 groups of test samples are taken and brought into the coal ash melting point prediction model for testing to obtain the test sample error. The test results are shown in Table 1.

[0108] Table 1 Prediction results of coal ash melting point using three models

[0109]

[0110] From the prediction results in Table 1, it can be seen that the prediction error of the ash melting point predicted by the method provided by the present invention is ≤100°C, whether it is the error of the training sample or the error of the test sample, which accounts for the highest proportion and is higher than the other two models. This shows that the ash melting point prediction by the present invention is more accurate.

[0111] like Figure 5 As shown, the third object of the present invention is to provide a coal ash melting point prediction system, comprising the following modules:

[0112] A prediction module is used to obtain real-time data of coal ash composition, input the coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point;

[0113] A data processing module is used to obtain historical data of coal ash composition and coal ash melting point temperature, and perform pre-processing and post-training samples;

[0114] A model building module is used to build an initial model for coal ash melting point prediction based on BP neural network according to training samples;

[0115] The model optimization module is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction by using an improved particle swarm algorithm to obtain a coal ash melting point prediction model.

[0116] The coal ash melting point prediction system of the present invention also includes a model testing module, which is used to input training samples and test samples into the optimized coal ash melting point prediction initial model for training and testing, so as to obtain a coal ash melting point prediction model.

[0117] like Figure 6 As shown, the third object of the present invention is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the coal ash melting point prediction method when executing the computer program.

[0118] The coal ash melting point prediction method comprises the following steps:

[0119] Obtain real-time data on coal ash composition, input it into a coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point;

[0120] The coal ash melting point prediction model is constructed by the following steps:

[0121] Obtain historical data on coal ash composition and coal ash melting point temperature, and perform preprocessing on training samples;

[0122] According to the training samples, the initial model for coal ash melting point prediction based on BP neural network is constructed;

[0123] The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction, and the coal ash melting point prediction model is obtained.

[0124] A fourth object of the present invention is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the coal ash melting point prediction method are implemented.

[0125] The coal ash melting point prediction method comprises the following steps:

[0126] Obtain real-time data on coal ash composition, input it into a coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point;

[0127] The coal ash melting point prediction model is constructed by the following steps:

[0128] Obtain historical data on coal ash composition and coal ash melting point temperature, and perform preprocessing on training samples;

[0129] According to the training samples, the initial model for coal ash melting point prediction based on BP neural network is constructed;

[0130] The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction, and the coal ash melting point prediction model is obtained.

[0131] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0132] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0133] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0134] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for predicting coal ash melting point, characterized in that: The following steps are involved: Obtain real-time data on coal ash composition, input it into a coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point; The coal ash melting point prediction model is constructed by the following steps: Obtain historical data on coal ash composition and coal ash melting point temperature, and perform preprocessing to obtain training samples; According to the training samples, the initial model for coal ash melting point prediction based on BP neural network is constructed; The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction, and the coal ash melting point prediction model is obtained.

2. A method for predicting coal ash melting point according to claim 1, characterized in that: The fly ash components include nine oxides, namely silicon dioxide, aluminum oxide, iron oxide, sulfur trioxide, calcium oxide, potassium oxide, sodium oxide, titanium dioxide and magnesium oxide.

3. A method for predicting coal ash melting point according to claim 1, characterized in that: The historical data of coal ash composition and coal ash melting point temperature are obtained and the training samples are pre-processed, specifically: The historical data of coal ash composition is normalized, and the normalization formula y = (xx min ) / (x max -x min ) is standardized and mapped to [0, 1]; Among them, y is the normalized data, x is the sample data; min is the minimum value of the sample data, x max is the maximum value of the sample data.

4. A method for predicting coal ash melting point according to claim 2, characterized in that: The BP neural network is used to construct an initial model for predicting the ash melting point; specifically: According to the training samples, the initial weights and thresholds of the BP neural network and the number of iterations are set to train the BP neural network. The training method of the BP neural network adopts the gradient training method. The BP neural network is selected to construct the initial model for predicting the coal ash melting point. A three-layer network structure is adopted. According to the types of coal ash components in the training samples, the input layer is determined to be 9 nodes, corresponding to the 9 oxides of coal ash components; the output layer is 1 node, which is the coal ash melting point temperature; the transfer function selected for the hidden layer input and output is the sigmoid function, and the hidden layer has 15 nodes.

5. A method for predicting coal ash melting point according to claim 4, characterized in that: The selection of the number of nodes in the hidden layer includes: According to the formula Preliminarily determine the number of nodes in the hidden layer; Where: n is the number of nodes in the hidden layer, p is the number of nodes in the output layer, q is the number of nodes in the input layer, and m is a constant between 0 and 10; Some training samples are selected for training, and the change of the mean square error of the BP neural network is observed by increasing or decreasing the number of nodes; when the number of nodes increases and the mean square error no longer decreases, the number of nodes at this time is determined to be the number of nodes in the hidden layer.

6. A method for predicting coal ash melting point according to claim 1, characterized in that: The improved particle swarm algorithm is used to optimize the initial weights and thresholds of the initial model for predicting the ash melting point, and the ash melting point prediction model is obtained, which is specifically: S1: Assign each particle a random initial position and speed. The initial position of the particle is a random number between [0, 1]; S2: The mean square error between the predicted result and the actual value is used as the fitness function to calculate the fitness value of each particle; S3: For each particle, compare the fitness value of its current position with the fitness value corresponding to its historical best position. If the fitness value of the current position is higher, update the historical best position with the current position. S4: for each particle, compare the fitness value of its current position with the fitness value corresponding to its global best position. If the fitness value of the current position is higher, update the global best position with the current position. S5: Update the speed and position of each particle according to the following formula in: is the optimized speed of particle i; is the current speed of particle i; w is the inertia factor; c1 and c2 are learning factors; r1 and r2 are random numbers between (0,1); pbest id is the best historical position of particle i; gbest d is the global optimal position of the particle swarm; is the current position of particle i; S6: If the end condition is not met, return to S2 to continue iteration; if the end condition is met, the iteration ends, and the position and speed of the optimal particle are assigned to the weight and threshold of the BP neural network to obtain the coal ash melting point prediction model.

7. A method for predicting coal ash melting point according to claim 1, characterized in that: The method also includes selecting some of the training samples as test samples to input into the coal ash melting point prediction model, and evaluating the prediction effect of the coal ash melting point prediction model, specifically: The test samples were input into the coal ash melting point prediction model for prediction, and the predicted values ​​were compared with the true values. The mean square error (MSE) and the coefficient of determination (R 2 ) is evaluated, and the expression is as follows: Where n is the number of samples, y i is the true value of the i-th sample, y i ' is the predicted value of the i-th sample; Where n is the number of samples, y i is the true value of the i-th sample, y i ' is the predicted value of the i-th sample, is the average of all true values.

8. A coal ash melting point prediction system, characterized in that: Includes the following modules: A prediction module is used to obtain real-time data of coal ash composition, input the coal ash melting point prediction model for prediction, and obtain the predicted coal ash melting point; A data processing module is used to obtain historical data of coal ash composition and coal ash melting point temperature, and perform pre-processing and post-training samples; A model building module is used to build an initial model for coal ash melting point prediction based on BP neural network according to training samples; The model optimization module is used to optimize the initial weights and thresholds of the initial model for coal ash melting point prediction by using an improved particle swarm algorithm to obtain a coal ash melting point prediction model.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the coal ash melting point prediction method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the method for predicting the ash melting point according to any one of claims 1 to 7.

Images