MSWI process combustion and heat exchange stage-oriented multipoint serial temperature prediction method

Abstract

The invention discloses a multi-point serial temperature prediction method for combustion and heat exchange stages of an MSWI process. The sample value of a solid-phase burnout area is the sum of a data set of the area and the predicted temperature of a solid-phase combustion area; the sample value of the gas-phase combustion area is the sum of the data set of the gas-phase combustion area, the predicted temperature of the solid-phase burnout area and the predicted temperature of the solid-phase combustion area; the sample value of the high-temperature heat exchange area is the sum of the data set of the area and the predicted temperature of the gas-phase combustion area; the sample value of the low-temperature heat exchange area is the sum of the data set of the area and the predicted temperature of the high-temperature heat exchange area. According to the method, through respective data sets and predicted temperatures of the solid-phase combustion area, the solid-phase burnout area, the gas-phase combustion area, the high-temperature heat exchange area and the low-temperature heat exchange area, the predicted temperature of each area is accurately predicted, and the mutual influence degrees among the multiple areas are effectively combined; and the temperature prediction precision of each region is improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of multi-point serial temperature prediction, in particular to a multi-point serial temperature prediction method for combustion and heat exchange stages of MSWI processes. BACKGROUND

[0002] Whether the temperature control of different regions in the combustion and heat exchange stages is within the process setting range is directly related to the economic benefits of MSWI enterprises and whether the environmental protection indicators meet the standards. In the process of municipal solid waste (MSW) combustion, combustion instability is prone to occur, such as high combustion temperature and uneven turbulent distribution. Stable incineration conditions can make MSW completely burn, and under the condition of high temperature of 850 DEG C to 950 DEG C, toxic gases such as carbon dioxide can be effectively decomposed. The 'denitration reaction' in the high-temperature zone of the furnace can reduce the emission of nitrogen oxides in the tail gas. If the temperature in the combustion stage is less than 850 DEG C or locally low, carbon dioxide decomposition will not be complete, and there is a risk of carbon dioxide recombination. In the heat exchange stage, the power generation efficiency increases by 0.5% to 0.1% for every 10 DEG C increase in superheated steam. From the perspective of physics and generation mechanism, the temperature of the MSWI process is strongly related to other variables and the temperature of the previous region. The existing temperature prediction methods for industrial processes basically focus on a single temperature point in a single region, and do not consider the degree of mutual influence between multiple key temperature points in multiple regions. SUMMARY

[0003] Therefore, it is necessary to propose a multi-point serial temperature prediction method for combustion and heat exchange stages of MSWI processes in view of the above problems.

[0004] A multi-point serial temperature prediction method for combustion and heat exchange stages of MSWI processes, the MSWI process combustion includes five regions: a solid-phase combustion region, a solid-phase ignition region, a gas-phase combustion region, a high-temperature heat exchange region and a low-temperature heat exchange region, each region is provided with a plurality of measuring points, and the actual detection values of each measuring point at different times constitute a parameter data set, a plurality of parameter data sets constitute original data, and the method comprises the following steps:

[0005] Obtaining the parameter data set in the original data of each region, and determining the PCC value of each parameter data set of each region and the predicted temperature;

[0006] Taking an absolute value of the PCC value, and taking the absolute value as a correlation coefficient of the parameter data set in each region;

[0007] Determining a threshold range according to the correlation coefficient, and determining a data set from the parameter data set in the threshold range;

[0008] The data set is divided into multiple leaf nodes and non-leaf nodes, the non-leaf nodes include left and right child nodes, the input and output sample set of the leaf node is determined according to the non-leaf node, the input and output sample set is composed of a temperature vector of a predicted temperature of each region and a detection value sample of each region, and the detection value sample is composed of an actual detection value of each region;

[0009] A weight vector of the leaf node is obtained, and a predicted temperature of each region is determined in combination with a sample value; wherein the sample value of the solid-phase combustion region is a data set of the region; the sample value of the solid-phase combustion region is a sum of the data set of the region and the predicted temperature of the solid-phase combustion region; the sample value of the gas-phase combustion region is a sum of the data set of the region, the predicted temperature of the solid-phase combustion region and the predicted temperature of the solid-phase combustion region; the sample value of the high-temperature heat exchange region is a sum of the data set of the region and the predicted temperature of the gas-phase combustion region; and the sample value of the low-temperature heat exchange region is a sum of the data set of the region and the predicted temperature of the high-temperature heat exchange region.

[0010] In one embodiment, the parameter data set in the original data of each region is obtained, and the PCC value of each parameter data set and the predicted temperature of each region is determined, including:

[0011] Each actual detection value in each parameter data set is determined;

[0012] A detection value average of multiple actual detection values in the parameter data set is determined;

[0013] A temperature value of a to-be-tested point at different time of each region is determined;

[0014] A temperature average of multiple temperature values of the to-be-tested point in the original data of each region is determined;

[0015] The PCC value of each parameter data set and the predicted temperature of each region is determined according to the actual detection value, the detection value average, the temperature value and the temperature average.

[0016] In one embodiment, the threshold range is determined according to the correlation coefficient, and the data set is determined from the parameter data set in the threshold range, including:

[0017] A lower limit value of the threshold is determined according to the number of the parameter data set and the correlation coefficient; and a maximum of multiple correlation coefficients is taken as an upper limit value of the threshold;

[0018] In the threshold range, a floating value is taken as an optimization interval from the lower limit value until an optimal threshold value is obtained; and the correlation coefficient greater than the optimal threshold value is determined, and the parameter data set corresponding to the correlation coefficient is retained to constitute a data set.

[0019] In one embodiment, the PCC value is obtained by the following expression:

[0020] (1)

[0021] wherein, PCC is the PCC value; is the actual detection value; is the average value of the detection value; is the temperature value at different time; is the average value of the temperature; represents the number of actual detection values; represents the number of parameter data sets, and the sample is a detection data set composed of a plurality of actual detection values corresponding to each measuring point.

[0022] In one embodiment, the threshold range is determined by the following expression:

[0023] (2)

[0024] (3)

[0025] (4)

[0026] wherein, is the correlation coefficient of the parameter data set in the i-th region; PCC is the PCC value; is the lower limit value of the threshold value; is the upper limit value of the threshold value; P is the number of parameter data sets; is the optimal threshold value; is an intermediate variable.

[0027] In one embodiment,

[0028] The expressions of the left child node and the right child node are as follows:

[0029] (5)

[0030] wherein, is the left child node; is the right child node; is the indicator function;

[0031] The expression of the input-output sample set is as follows:

[0032] (6)

[0033] wherein, is the input-output sample set; The detection value sample; The temperature vector; The detection value sample The number of; The detection value sample The dimension of.

[0034] In one embodiment,

[0035] The weight vector is obtained by determining the weight vector through a regularized least square loss function:

[0036] (7)

[0037] Let , we have:

[0038] (8)

[0039] After moving the term, there is:

[0040] (9)

[0041] (10)

[0042] Wherein, is the weight vector; is the detection value sample; is the temperature vector; is the regularization coefficient; is a unit matrix of dimension.

[0043] In one embodiment, the predicted temperature of each region is realized by the following expression:

[0044] (11)

[0045] Wherein, is the predicted temperature of each region; is the weight vector; is the detection value sample.

[0046] In one embodiment, the expression of the sample value of the solid phase combustion region, the sample value of the solid phase combustion region, the sample value of the gas phase combustion region, the sample value of the high temperature heat exchange region and the sample value of the low temperature heat exchange region is as follows:

[0047] (12)

[0048] (13)

[0049] (14)

[0050] (15)

[0051] (16)

[0052] wherein, is a sample value of the solid phase combustion region; is a data set of the solid phase combustion region; is a predicted temperature of the solid phase combustion region; is a sample value of the solid phase combustion region; is a data set of the solid phase combustion region; is a predicted temperature of the solid phase combustion region; is a sample value of the gas phase combustion region; is a data set of the gas phase combustion region; is a predicted temperature of the gas phase combustion region; is a sample value of the high temperature heat exchange region; is a data set of the high temperature heat exchange region; is a predicted temperature of the high temperature heat exchange region; is a sample value of the low temperature heat exchange region; is a data set of the low temperature heat exchange region.

[0053] The present application realizes accurate prediction of the predicted temperature of each region by the data set and the predicted temperature of the solid phase combustion region, the solid phase combustion region, the gas phase combustion region, the high temperature heat exchange region and the low temperature heat exchange region, effectively combines the mutual influence degree between multiple regions, and improves the predicted temperature accuracy of each region. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0055] wherein,

[0056] Figure 1 is an application environment diagram of the multi-point serial temperature prediction method for the MSWI process combustion and heat exchange stage in an embodiment;

[0057] Figure 2 is a flowchart of the multi-point serial temperature prediction method for the MSWI process combustion and heat exchange stage in an embodiment;

[0058] Figure 3 Figure 3 is a comparison chart of experimental results in one embodiment;

[0059] Figure 4 Figure 4 is a structural block diagram of a computer device in one embodiment. DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the scope of protection of the present application.

[0061] Whether the temperature control of the multi-point series different zones in the combustion and heat exchange stages is within the process setting range is directly related to the economic benefits of the MSWI enterprise and whether the environmental protection indicators meet the standards. The combustion of municipal solid waste (MSW) is prone to cause unstable combustion, resulting in high combustion temperature, uneven turbulent distribution, and other problems. Stable incineration conditions can ensure complete combustion of MSW, and under high temperature conditions of 850-950℃, toxic gases such as carbon dioxide can be effectively decomposed. The "denitration reaction" in the high-temperature zone of the furnace can reduce the emission of nitrogen oxides in the tail gas. If the temperature in the combustion stage is less than 850℃ or locally low, it will lead to incomplete decomposition of carbon dioxide and the risk of carbon dioxide recombination. In the heat exchange stage, for every 10℃ increase in superheated steam, the power generation efficiency increases by 0.5%-0.1%. From the perspective of physics and generation mechanism, the temperature of the MSWI process is strongly related to other variables and the temperature of the previous zone. Existing temperature prediction methods for industrial processes basically focus on a single temperature point in a single zone, without considering the degree of influence between multiple key temperature points in multiple zones. To solve the above technical problems, the present application provides a multi-point series temperature prediction method for the combustion and heat exchange stages of the MSWI process.

[0062] Figure 1 Figure 4 is a structural block diagram of a computer device in one embodiment. Figure 1The multi-point serial temperature prediction method for the combustion and heat exchange stages of the MSWI process is applied to a multi-point serial temperature prediction system for the combustion and heat exchange stages of the MSWI process. The multi-point serial temperature prediction system for the combustion and heat exchange stages of the MSWI process includes a terminal 110 and a server 120. The terminal 110 and the server 120 are connected through a network, and the terminal 110 can be a desktop terminal or a mobile terminal, and the mobile terminal can be at least one of a mobile phone, a tablet computer, a notebook computer, etc. The server 120 can be implemented by an independent server or a server cluster composed of multiple servers. The terminal 110 is used to obtain a parameter data set in original data of each region and determine a PCC value of each parameter data set of each region and a predicted temperature, the server 120 is used to take an absolute value of the PCC value and use the absolute value as a correlation coefficient of the parameter data set in each region; determine a threshold range according to the correlation coefficient and determine a data set from the parameter data set in the threshold range; divide the data set into multiple leaf nodes and non-leaf nodes, the non-leaf nodes include left child nodes and right child nodes, determine an input-output sample set of the leaf nodes according to the non-leaf nodes, the input-output sample set is composed of a temperature vector of the predicted temperature of each region and a detection value sample of each region, and the detection value sample is composed of an actual detection value of each region; obtain a weight vector of the leaf nodes and determine a predicted temperature of each region in combination with a sample value; wherein the sample value of the solid-phase combustion region is a data set of the region; the sample value of the solid-phase combustion region is a sum of the data set of the region and the predicted temperature of the solid-phase combustion region; the sample value of the gas-phase combustion region is a sum of the data set of the region, the predicted temperature of the solid-phase combustion region and the predicted temperature of the solid-phase combustion region; the sample value of the high-temperature heat exchange region is a sum of the data set of the region and the predicted temperature of the gas-phase combustion region; and the sample value of the low-temperature heat exchange region is a sum of the data set of the region and the predicted temperature of the high-temperature heat exchange region.

[0063] As shown in Figure 2 In one embodiment, a multi-point serial temperature prediction method for the combustion and heat exchange stages of the MSWI process is provided. The method can be applied to a terminal or a server, and the embodiment is exemplified by application to a terminal. The multi-point serial temperature prediction method for the combustion and heat exchange stages of the MSWI process specifically includes the following steps:

[0064] In this application, the MSWI process combustion includes five regions: a solid-phase combustion region, a solid-phase combustion region, a gas-phase combustion region, a high-temperature heat exchange region and a low-temperature heat exchange region, each region is provided with multiple measuring points, and the actual detection values of each measuring point at different times constitute a parameter data set, and multiple parameter data sets constitute original data, and the method includes:

[0065] S10: Obtain a parameter data set in original data of each region, and determine a PCC value of each parameter data set of each region and a predicted temperature;

[0066] S20: Take an absolute value of the PCC value, and take the absolute value as a correlation coefficient of the parameter data set in each region;

[0067] S30: Determine a threshold range according to the correlation coefficient, and determine a data set from the parameter data set in the threshold range;

[0068] S40: Divide the data set into a plurality of leaf nodes and non-leaf nodes, the non-leaf nodes include left and right child nodes, determine an input-output sample set of the leaf nodes according to the non-leaf nodes, the input-output sample set is composed of a temperature vector of a predicted temperature of each region and a detection value sample of each region, the detection value sample is composed of an actual detection value of each region;

[0069] S50: Obtain a weight vector of a leaf node, and determine a predicted temperature of each region in combination with a sample value; wherein, the sample value of the solid phase combustion region is a data set of the region; the sample value of the solid phase combustion region is a sum of the data set of the region and a predicted temperature of the solid phase combustion region; the sample value of the gas phase combustion region is a sum of the data set of the region, the predicted temperature of the solid phase combustion region and a predicted temperature of the solid phase combustion region; the sample value of the high-temperature heat exchange region is a sum of the data set of the region and a predicted temperature of the gas phase combustion region; and the sample value of the low-temperature heat exchange region is a sum of the data set of the region and a predicted temperature of the high-temperature heat exchange region.

[0070] The present application realizes accurate prediction of the predicted temperature of each region by using the data set and the predicted temperature of the solid phase combustion region, the solid phase combustion region, the gas phase combustion region, the high-temperature heat exchange region and the low-temperature heat exchange region, effectively combines the mutual influence degree between multiple regions, and improves the prediction temperature accuracy of each region.

[0071] In one embodiment, the obtaining a parameter data set in original data of each region, and determining a PCC value of each parameter data set of each region and a predicted temperature in step S10 comprises:

[0072] S101: Determine each actual detection value in each parameter data set;

[0073] S102: Determine a detection value average of a plurality of actual detection values in the parameter data set;

[0074] S103: Determine a temperature value of a to-be-tested point at different time of each region;

[0075] S104: determining a temperature average value of the multiple temperature values of each region in the original data;

[0076] S105: determining a PCC value of each parameter data set and a predicted temperature of each region according to the actual detection value, the detection value average value, the temperature value and the temperature average value.

[0077] In one embodiment, for the step S30, the determining a threshold range according to the correlation coefficients, and determining data sets from the parameter data sets within the threshold range comprises:

[0078] S301: determining a lower limit value of the threshold according to the number of the parameter data sets and the correlation coefficients; taking the maximum of the multiple correlation coefficients as an upper limit value of the threshold;

[0079] S302: within the threshold range, starting from the lower limit value and taking a floating value as an optimization interval until an optimal threshold value is obtained; determining the correlation coefficients greater than the optimal threshold value, and retaining the parameter data sets corresponding to the correlation coefficients to constitute data sets.

[0080] In one embodiment, for the step S20, the PCC value is obtained by the following expression:

[0081] (1)

[0082] wherein, is the PCC value; is the actual detection value; is the detection value average value; is the temperature value at different time; is the temperature average value; represents the number of the actual detection values; represents the number of the parameter data sets, and the sample is a detection data set composed of multiple actual detection values corresponding to each measuring point.

[0083] In one embodiment, for the step S30, the threshold range is determined by the following expression:

[0084] (2)

[0085] (3)

[0086] (4)

[0087] wherein, is the correlation coefficient of the parameter data set in the i-th region; is the PCC value; is the lower limit value of the threshold. is the upper limit of the threshold; P is the number of parameter datasets; The optimal threshold; This serves as an intermediate variable.

[0088] In one embodiment, the expressions for the left and right child nodes in step S40 are as follows: Given the first... Data sets for each region , For dataset The first in A sample vector, For sample vectors The first in One sample, For the first A temperature value, The number of parameters in the dataset. Intermediate nodes are based on sample vectors. Dataset Divide into left child nodes and right child node ,Right now

[0089] (5)

[0090] in, The left child node; The right child node; For indicator functions;

[0091] (6)

[0092] In the formula: The first of the sample vectors There are eigenvalues, and they exist. ; The node identifier representing the first non-leaf node is determined by the mean square error (MSE) and a loop traversal process, as follows:

[0093] (7)

[0094] (8)

[0095] In the formula: For dataset Left child node of the split and right child node The MSE values ​​between; For dataset Coordinates (i.e., at the 1st) In the nth iteration, the th The first sample (one eigenvalue); This is the function for solving MSE; For indicator functions; The first of all sample vectors in the left child node A temperature value; The first of all sample vectors in the right child node Temperature value; The left child node's first Average temperature; The right child node's first The average temperature at each temperature point. argmin represents the parameter that minimizes the objective function. ) is the optimization variable. A determiner to classify each child node as either a left or right child node:

[0096] (9)

[0097] In the formula, The maximum number of traversals ( When the number of non-leaf node samples is greater than the minimum number of samples. Then, the traversal and segmentation process of equations (5) to (9) is repeated until the number of actual detection values ​​after the child node is segmented is less than the minimum number of samples. The child node is then identified as a leaf node and the loop stops.

[0098] Through the above process, it can be determined that A path from the root node to a leaf node. non-leaf nodes and Each path contains leaf nodes, and each path also contains non-leaf nodes with significant differences.

[0099] The expression for the input and output sample sets is as follows:

[0100] (10)

[0101] in, For input and output sample sets; For the sample of test values; It is a temperature vector; For the test value sample Quantity; For the test value sample The dimension of.

[0102] In one embodiment, the process of obtaining the weight vector in step S50 is as follows: the weight vector is determined by a regularized least squares loss function, and the weight vector is calculated using a linear regression method. The predicted temperature of the lower leaf node of the path is calculated as follows:

[0103] (11)

[0104] In the formula: For the first Predicted temperature of the lower leaf node of the path; For the first The first path The weight vectors of the leaf nodes in each region are calculated using a regularized least squares loss function to ensure the weight vectors are optimized. Non-divergent, its definition is as follows:

[0105] (12)

[0106] In the formula, For the first The regularization coefficients of each model.

[0107] (13)

[0108] make We can obtain:

[0109] (14)

[0110] After rearranging, the following exists:

[0111] (15)

[0112] (16)

[0113] in, This is the weight vector; For the sample of test values; It is a temperature vector; The regularization coefficient is used. for An identity matrix of dimension 1.

[0114] In one embodiment, the predicted temperature for each region in step S50 is achieved by the following expression:

[0115] (17)

[0116] in, Predicted temperature for each region; This is the weight vector; This is a sample of the detection values.

[0117] In one embodiment, the expressions for the sample value of the solid phase combustion zone, the sample value of the solid phase burnout zone, the sample value of the gas phase combustion zone, the sample value of the high temperature heat exchange zone, and the sample value of the low temperature heat exchange zone in step S50 are as follows:

[0118] (18)

[0119] (19)

[0120] (20)

[0121] (21)

[0122] (22)

[0123] The final expression of the predicted temperature of each zone is as follows:

[0124] (23)

[0125] wherein, is the sample value of the solid phase combustion zone; is the data set of the solid phase combustion zone; is the predicted temperature of the solid phase combustion zone; is the sample value of the solid phase burnout zone; is the data set of the solid phase burnout zone; is the predicted temperature of the solid phase burnout zone; is the sample value of the gas phase combustion zone; is the data set of the gas phase combustion zone; is the predicted temperature of the gas phase combustion zone; is the sample value of the high temperature heat exchange zone; is the data set of the high temperature heat exchange zone; is the predicted temperature of the high temperature heat exchange zone; is the sample value of the low temperature heat exchange zone; is the data set of the low temperature heat exchange zone.

[0126] To compare model performance, taking the average temperature of the grate in the combustion section as an example, this paper selects Classification and Regression Tree (CART), Random Forest (RF), and Back-Propagation Neural Network (BPNN) to conduct experiments on modeling the average temperature of the grate in the combustion section, and compares their performance with the LRDT algorithm proposed in this paper. The hyperparameters of the above algorithms are defined as follows: CART, minimum number of samples in leaf nodes is 20, number of decision trees is 50; RF, minimum number of samples in leaf nodes is 10, number of feature selections per tree is 6, number of trees is 100; BPNN, maximum number of convergence iterations is 1500, number of hidden layer neurons is 15, learning rate is 0.2, convergence error is 0.001, and tanh activation function is used; LRDT, minimum number of samples in leaf nodes is 19, number of feature selections is 4, and regularization coefficient is 0.0027.

[0127] The experimental results of the method proposed in this paper are compared with those of other methods, for example... Figure 3 As shown in the figure. Since the results of BPNN and RF are affected by random parameters, to ensure the fairness of the comparative experiment, each algorithm was run 30 times, and the optimal and mean values ​​were calculated for comparison.

[0128] Table 1 Comparison of experimental statistical results of average grate temperature in the combustion section.

[0129]

[0130] Depend on Figure 3 As shown in Table 1, (1) the LRDT algorithm of this application exhibits the best performance, with MSE=0.0503 and MAE=0.1066 on the test set. =0.9973, which is due to its piecewise fitting ability of the fusion decision tree and the continuous advantage of linear regression. The dotted line graph shows that its error bar range is extremely small, indicating that it can accurately capture the continuous temperature change pattern in the 220-235℃ range, with a prediction error of only about 0.1℃ and extremely strong stability; (2) BPNN algorithm test set MSE=0.3244, =0.9824, the light blue dotted line can track the overall trend, but the local fluctuations are obvious, reflecting that the neural network has a strong fitting ability through nonlinear mapping, but the difference between the best value and the mean is significant. The model is greatly affected by the random initial weights and the parameters need to be fine-tuned to approach the LRDT level; (3) CART algorithm test set MSE=0.4932, MAE=0.5203, the green dotted line shows a step-like change, because its single tree structure is sensitive to noise, resulting in overfitting (training set Best MSE=0.2359 to test set degenerates to 0.4932); (4) RF algorithm test set MSE=0.5283 is the worst, the yellow dotted line deviates significantly from the true value after 300 seconds, the variance of the validation set and the test set is less than CART, indicating that although the ensemble strategy improves stability, its essential "piecewise constant" prediction characteristics are contrary to the continuous temperature change pattern, resulting in the inability to eliminate the system bias (best MSE=0.4932). (5) Experiments show that LRDT is only 0.9791); A precision of 0.9967 and MAE of 0.1066 is the optimal solution.

[0131] In summary, the LRDT algorithm proposed in this application has good fitting effect and high modeling accuracy in multi-point temperature modeling applications, meeting the needs of temperature prediction in actual industrial processes. Experimental results prove the effectiveness of the proposed algorithm.

[0132] Figure 4 An internal structural diagram of a computer device in one embodiment is shown. This computer device can specifically be a terminal or a server. Figure 4 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program. When executed by the processor, this computer program enables the processor to implement a multi-point serial temperature prediction method for the combustion and heat transfer stages of the MSWI process. The internal memory may also store a computer program, which, when executed by the processor, enables the processor to implement the multi-point serial temperature prediction method for the combustion and heat transfer stages of the MSWI process. Those skilled in the art will understand that... Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0133] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiment methods can be included. Any reference to memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0134] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0135] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process. The MSWI process combustion includes five regions: a solid-phase combustion region, a solid-phase burnout region, a gas-phase combustion region, a high-temperature heat transfer region, and a low-temperature heat transfer region. Multiple measuring points are set in each region. The actual detection values ​​of each measuring point at different times constitute a parameter dataset, and multiple parameter datasets constitute the raw data. The method is characterized by... The method includes: Obtain the parameter dataset from the raw data for each region, and determine the PCC value for each parameter dataset and the predicted temperature for each region; The absolute value of the PCC value is taken and used as the correlation coefficient of the parameter dataset in each region; A threshold range is determined based on the correlation coefficient, and a dataset is determined from the parameter dataset within the threshold range; The dataset is divided into multiple leaf nodes and non-leaf nodes. The non-leaf nodes include left child nodes and right child nodes. The input and output sample sets of the leaf nodes are determined based on the non-leaf nodes. The input and output sample sets consist of the temperature vector of the predicted temperature of each region and the detection value samples of each region. The detection value samples consist of the actual detection values ​​of each region. Obtain the weight vector of the leaf node and combine it with the sample values ​​to determine the predicted temperature of each region; wherein, the sample value of the solid-phase combustion region is the dataset of that region; the sample value of the solid-phase burnout region is the sum of the dataset of that region and the predicted temperature of the solid-phase combustion region; the sample value of the gas-phase combustion region is the sum of the dataset of that region, the predicted temperature of the solid-phase burnout region, and the predicted temperature of the solid-phase combustion region; the sample value of the high-temperature heat exchange region is the sum of the dataset of that region and the predicted temperature of the gas-phase combustion region; the sample value of the low-temperature heat exchange region is the sum of the dataset of that region and the predicted temperature of the high-temperature heat exchange region.

2. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 1, characterized in that, The process of obtaining the parameter dataset from the raw data of each region and determining the PCC value of each parameter dataset and the predicted temperature for each region includes: Determine each actual detection value in each of the parameter datasets; Determine the average value of the multiple actual detection values ​​in the parameter dataset; Determine the temperature values ​​of the measurement points in each region at different times; Determine the average temperature of multiple temperature values ​​for the test points in each region from the raw data; The PCC value for each parameter dataset and the predicted temperature for each region is determined based on the actual detected values, the average detected values, the temperature value, and the average temperature value.

3. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 1, characterized in that, The step of determining a threshold range based on the correlation coefficient and determining a dataset from the parameter dataset within the threshold range includes: The lower limit of the threshold is determined based on the number of parameter datasets and the correlation coefficients; the largest of the multiple correlation coefficients is taken as the upper limit of the threshold. Within the threshold range, starting from the lower limit, the optimization interval is adjusted with floating values ​​until the optimal threshold is obtained; and the correlation coefficients greater than the optimal threshold are determined, and the parameter datasets corresponding to the correlation coefficients are retained to form a dataset.

4. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 2, characterized in that, The PCC value is obtained through the following expression: （1） in, PCC value; This is the actual measured value; This represents the average value of the detected values. These are the temperature values ​​at different times; This represents the average temperature. Indicates the number of actual detected values; This indicates the number of parameter datasets, where the sample is a detection dataset consisting of multiple actual detection values ​​corresponding to each measurement point.

5. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 3, characterized in that, The threshold range is determined by the following expression: （2） （3） （4） in, Let be the correlation coefficient of the parameter dataset in the i-th region; PCC value; This is the lower limit of the threshold value; is the upper limit of the threshold; P is the number of parameter datasets; The optimal threshold; This serves as an intermediate variable.

6. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 1, characterized in that, The expressions for the left and right child nodes are as follows: （5） in, The left child node; The right child node; For indicator functions; The expression for the input and output sample sets is as follows: （6） in, For input and output sample sets; For the sample of test values; It is a temperature vector; For the test value sample Quantity; For the test value sample The dimension of.

7. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 1, characterized in that, The weight vector is obtained as follows: the weight vector is determined using a regularized least squares loss function. （7） make We can obtain: （8） After rearranging, the following exists: （9） （10） in, This is the weight vector; For the sample of test values; It is a temperature vector; The regularization coefficient is used. for An identity matrix of dimension 1.

8. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 1, characterized in that, The predicted temperature for each region is achieved using the following expression: （11） in, Predicted temperature for each region; This is the weight vector; This is a sample of the detection values.

9. The multi-point serial temperature prediction method for the combustion and heat transfer stages of an MSWI process according to claim 1, characterized in that, The expressions for the sample values ​​of the solid-phase combustion region, the solid-phase burnout region, the gas-phase combustion region, the high-temperature heat exchange region, and the low-temperature heat exchange region are as follows: （12） （13） （14） （15） （16） in, These are sample values ​​from the solid-phase combustion region; This is a dataset of solid-phase combustion regions; The predicted temperature of the solid-phase combustion region; These are sample values ​​from the solid-phase combustion region; This is a dataset of the solid-phase combustion region; The predicted temperature of the solid-phase combustion region; These are sample values ​​for the gas-phase combustion region; This is a dataset for the gas-phase combustion region; This is the predicted temperature of the gas-phase combustion region; These are sample values ​​for the high-temperature heat exchange region; Data set for high-temperature heat exchange regions; The predicted temperature for the high-temperature heat exchange region; These are sample values ​​for the low-temperature heat exchange region; This is a dataset for the low-temperature heat exchange region.

Images