A multi-modal data-driven optimization control method and system for fly ash dioxin low-temperature pyrolysis

Abstract

The present application relates to a kind of multi-modal data-driven fly ash dioxin low-temperature pyrolysis optimization control method and system, including the neural network model of multi-modal data multi-flow multi-scale feature sharing, simulate fly ash dioxin low-temperature pyrolysis process;Step two, the neural network model of multi-flow multi-scale feature sharing is optimized by data-driven mode;Current atmosphere data, fly ash feed data, heat treatment process data are input to the neural network model of multi-flow multi-scale feature sharing, and the estimated value of dioxin low-temperature pyrolysis rate is obtained;On the basis of current dioxin low-temperature pyrolysis rate estimate, maintain current atmosphere, fly ash feed condition unchanged, by the adjustment control of heat treatment process, optimize control fly ash dioxin low-temperature pyrolysis;The present application is modeled by data modeling mode to fly ash dioxin pyrolysis process, and pyrolysis process is adjusted adaptively to improve the pyrolysis rate of dioxin.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optimization control of fly ash dioxin low-temperature pyrolysis, and particularly relates to a multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control method and system. BACKGROUND

[0002] With the increasing amount of waste incineration year by year, the dioxin in fly ash generated by incineration has become a major problem restricting the popularization and application of waste incineration technology. The low-temperature pyrolysis technology has gradually become one of the main technologies for reducing the content of fly ash dioxin, which can effectively promote pollution prevention and fly ash resource utilization. However, the fly ash low-temperature pyrolysis technology is a complex system of atmosphere, fly ash feeding and heat treatment process multi-element coupling, and its working conditions and pyrolysis efficiency are variable, which is difficult to accurately control. For the modeling analysis of the pyrolysis process, it is early relied on mechanism modeling and experimental analysis. On the one hand, this kind of technology is difficult to accurately simulate the complex low-temperature pyrolysis process of dioxin, and the simulation accuracy is low, on the other hand, it is difficult to be applied to actual working conditions. Although the pilot technology is used to regulate and control the pyrolysis equipment system, it is also difficult to solve this problem, and there is still a problem of low output of single equipment, which is difficult to realize large-scale industrialization.

[0003] The modeling of fly ash dioxin pyrolysis process is the key to improve the pyrolysis efficiency. In order to get rid of the difficulty of early mechanism modeling, the data-driven pyrolysis process modeling gradually attracts the attention of academia and industry. However, the existing data-driven dioxin pyrolysis control technology only considers single or part of the key factors affecting the pyrolysis efficiency, and cannot model the multi-element of atmosphere, fly ash feeding and heat treatment process, which has serious deficiencies in the generality and accuracy of the pyrolysis process modeling. SUMMARY

[0004] The purpose of the present application is to provide a multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control method and system.

[0005] To have a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not an overview, nor is it intended to identify key / important elements or delineate the scope of the embodiments. Its sole purpose is to present some concepts in a simple form as a preamble to the detailed description that follows.

[0006] According to a first aspect of the embodiments of the present application, a multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control method is provided, which comprises:

[0007] Step one, a multi-modal data multi-flow multi-scale feature sharing neural network model is constructed to simulate the fly ash dioxin low-temperature pyrolysis process;

[0008] Step two, optimize the multi-stream multi-scale feature sharing neural network model in a data-driven manner;

[0009] Step three, input the current atmosphere data, fly ash feeding data, and heat treatment process data into the multi-stream multi-scale feature sharing neural network model to obtain an estimated value of the dioxin low-temperature pyrolysis rate;

[0010] Step four, based on the current estimated value of the dioxin low-temperature pyrolysis rate, maintain the current atmosphere and fly ash feeding conditions unchanged, and improve the dioxin low-temperature pyrolysis rate by adjusting and controlling the heat treatment process to optimize the control of fly ash dioxin low-temperature pyrolysis, forming a fly ash dioxin low-temperature pyrolysis optimization system that matches the boiler body and the optimization control method.

[0011] In an optional embodiment, in the multi-modal data multi-stream multi-scale feature sharing neural network model constructed in step one, the input data is multi-modal data, and the output data is the dioxin pyrolysis rate y; the input multi-modal data includes atmosphere data, fly ash feeding data, and heat treatment process data; the atmosphere data represents two attributes of atmosphere type and atmosphere concentration; the fly ash feeding data represents two attributes of fly ash dioxin concentration and fly ash feeding amount; the heat treatment process data represents two attributes of pyrolysis temperature and heat treatment time; wherein,

[0012] The expression of the atmosphere data is as follows:

[0013]

[0014] In the formula, represents the atmosphere type, represents the atmosphere concentration;

[0015] The expression of the fly ash feeding data is as follows:

[0016]

[0017] In the formula, represents the fly ash dioxin concentration, represents the fly ash feeding amount;

[0018] The expression of the heat treatment process data is as follows:

[0019]

[0020] In the formula, represents the pyrolysis temperature, represents the heat treatment time.

[0021] In an optional embodiment, in step two, the multi-modal data multi-stream multi-scale feature sharing neural network model simulates the pyrolysis process in a data-driven manner, and the model is optimized using training samples, including:

[0022] The initial multi-modal data is input, and the initial multi-modal data includes multi-modal data including atmosphere data, fly ash feed data, and heat treatment process data;

[0023] A first convolutional layer and a pooling layer are provided to extract and calculate features of the multi-modal data;

[0024] A second convolutional layer and a pooling layer are provided to extract and calculate features of the processed multi-modal data;

[0025] A feature sharing layer is provided to couple the multi-modal data after the first convolutional layer and the second convolutional layer to obtain a first layer feature coupling result and a second layer feature coupling result, respectively;

[0026] A feature concatenation layer is provided to concatenate the first layer feature coupling result and the second layer feature coupling result to form a concatenation feature under two convolutional scales;

[0027] A channel attention layer is provided to weight and fuse the concatenation feature under the two convolutional scales, and then input the feature into the feature concatenation layer again, and combine the feature with the atmosphere data in the initial input multi-modal data to form a final feature for estimating the fly ash dioxin low-temperature pyrolysis rate;

[0028] A pooling layer and a softmax layer are provided to input the final feature into the pooling layer and the softmax layer to obtain an estimation result of the dioxin pyrolysis rate under the current condition.

[0029] In an optional embodiment, the coupled data of the feature sharing layer includes: atmosphere concentration , fly ash dioxin concentration , fly ash feed amount , pyrolysis temperature , heat treatment time ;

[0030] After the first layer convolutional layer feature, the first layer feature coupling is:

[0031]

[0032] wherein, is the first layer convolutional feature, is the first layer feature coupling weight, is the first layer feature coupling result;

[0033] After the second layer convolutional layer feature, the second layer feature coupling is:

[0034]

[0035] wherein, is the second layer convolutional feature, coupling weights for the second layer of features, coupling results for the second layer of features.

[0036] In an optional embodiment, the convolution kernel size of the first layer of convolution layers is 1x7, the convolution kernel size of the second layer of convolution layers is 1x5, and the size of the pooling layer is 3;

[0037] wherein the spliced feature expression under the convolution scale of the first layer of convolution layers is:

[0038]

[0039] the spliced feature expression under the convolution scale of the second layer of convolution layers is:

[0040]

[0041] In an optional embodiment, the spliced features under the two convolution scales are weighted and fused by using a channel attention layer, and then re-input into the feature splicing layer, and combined with the atmosphere data in the initial input multi-modal data, to form a final feature expression for estimating the low-temperature pyrolysis rate of fly ash dioxin:

[0042] wherein, represents the atmosphere type, , , , , are all weighting coefficients.

[0043] In an optional embodiment, in step four, based on the current dioxin low-temperature pyrolysis rate estimate, the current atmosphere is maintained, the fly ash feeding conditions are unchanged, and the dioxin low-temperature pyrolysis rate is estimated, the adjustment control of the thermal treatment process is established, and the optimization function of the maximum dioxin low-temperature pyrolysis rate y is established, and the fly ash dioxin low-temperature pyrolysis thermal treatment process under the condition of the maximum dioxin low-temperature pyrolysis rate is solved by using the gradient descent algorithm, wherein,

[0044]

[0045] wherein, is the optimal pyrolysis temperature, is the optimal heat treatment time, is the optimal solution of the maximum dioxin low-temperature pyrolysis rate in the fly ash dioxin low-temperature pyrolysis thermal treatment process;

[0046] The optimal pyrolysis temperature and heat treatment time are fed back to the PLC control system of the low-temperature pyrolysis boiler body, so as to realize the optimization control of the fly ash dioxin low-temperature pyrolysis.

[0047] According to a second aspect of the embodiment of the present application, a multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control system is provided, comprising:

[0048] A collection module is configured to collect current atmosphere data, fly ash feed data and heat treatment process data, and transmit multi-modal data to the processing module.

[0049] The processing module is configured to construct a multi-modal data multi-flow multi-scale feature sharing neural network model, and optimize the model by the multi-modal data transmitted by the collection module, so as to realize the optimization solution of estimating the fly ash dioxin low-temperature pyrolysis emission.

[0050] The technical solution provided by the embodiment of the present application can include the following beneficial effects:

[0051] Through the multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control method and system of the embodiment of the present application, the fly ash dioxin pyrolysis process can be modeled by data modeling, a solution considering multiple factors is proposed for the complex system of multi-factor coupling in the heat treatment process, and the problems of mechanism modeling and existing machine learning technology only considering single or a small number of working conditions are solved. In addition, the heat treatment process can be adjusted to maximize the pyrolysis rate objective function under the conditions of atmosphere, fly ash feed determination and known conditions, and the pyrolysis process can be adjusted adaptively under different atmosphere and fly ash feed conditions to improve the pyrolysis rate of dioxin. BRIEF DESCRIPTION OF DRAWINGS

[0052] The above and other objects, features and advantages of the present application will become more apparent from the following description of the embodiments of the present application taken with reference to the accompanying drawings, in which:

[0053] Figure 1 is a structural schematic diagram of a fly ash dioxin low-temperature pyrolysis boiler;

[0054] Figure 2 is a step diagram of the method of the present application;

[0055] Figure 3 is a flowchart of optimizing the multi-flow multi-scale feature sharing neural network model by data driven method; DETAILED DESCRIPTION

[0056] The following description and drawings are illustrative of the specific embodiments of the present document and are not intended to be limiting. Portions and features of some embodiments can be included in, or alternative to, portions and features of other embodiments. The range of variations of the embodiments of the present document as described can be encompassed by the entire scope of the claims. In this document, the terms "first," "second," etc. are used merely as label to distinguish one element from another, and are not intended to impose numerical requirements on their objects. First and second elements could be replaced by third and fourth elements, respectively, without changing the present document. Furthermore, the terms "comprise", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a structure, device or apparatus that comprises a list of elements is not necessarily limited to those elements, but can include other elements not expressly listed or inherent to such structure, device or apparatus. In this document, the terms "include", "including", or any other variation thereof, cover a non-exclusive inclusion, such that a structure, device or apparatus that includes a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such structure, device or apparatus. Embodiments of the present document are described with reference to the attached drawings, in which various embodiments of the present document are shown for purposes of illustration.

[0057] It should be understood that, although the steps of the flow diagrams are shown in a sequential order, these steps are not necessarily performed in the order shown. Unless explicitly stated, as can be apparent from the disclosure, the steps of the flow diagrams can be performed in other orders. Further, at least some of the steps can include multiple sub-steps or multiple stages, which can be performed at the same time or in a different order than shown.

[0058] The modules in the apparatus or system of the present application can be implemented in whole or in part by software, hardware or a combination thereof. The modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory in the computer device in software form, so as to be invoked by the processor to perform the operations corresponding to the modules.

[0059] The embodiments in the present application and the features in the embodiments can be combined with each other in the case of no conflict.

[0060] Figure 1A structure diagram of a fly ash dioxin low-temperature pyrolysis boiler is shown in the figure, which comprises a feeding cover, a sealing device, a cylinder part, a heating cover, an internal blade, a discharging cover, a 2nd support device, a bottom steel frame, a transmission device, and a 1st support device; it should be noted that the present application provides a multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control method and system; in the training process of the model, multi-modal operation data of the designed boiler body are collected in a targeted manner, the model is converged through a data driven manner, and the control model can adapt to the designed boiler body; Figure 1 That is, a boiler structure capable of applying the method to optimize the fly ash dioxin pyrolysis process.

[0061] According to a first aspect of the embodiments of the present application, a multi-modal data driven fly ash dioxin low-temperature pyrolysis optimization control method is provided, Figure 2 The step process of the method of the present application is shown in the figure, and the method comprises:

[0062] Step 1, model construction: a multi-modal data multi-flow multi-scale feature sharing neural network model is constructed to simulate the fly ash dioxin low-temperature pyrolysis process; the input data is multi-modal data, and the output data is the dioxin pyrolysis rate y; the input multi-modal data includes atmosphere data, fly ash feeding data, and heat treatment process data;

[0063] Step 2, model training: the multi-flow multi-scale feature sharing neural network model is optimized through a data driven manner; specifically, experiments are carried out on the designed pyrolysis boiler body, and the working condition data and the pyrolysis rate of the atmosphere, the fly ash feeding, and the heat treatment process of the dioxin pyrolysis process are collected as training samples; a training set with a sample size of more than 6000 is established for training;

[0064] Step 3, model response: the normalized current atmosphere data, fly ash feeding data, and heat treatment process data are input into the multi-flow multi-scale feature sharing neural network model to obtain the estimated value of the dioxin low-temperature pyrolysis rate;

[0065] Step 4, optimal feedback control: based on the current dioxin low-temperature pyrolysis rate estimate, the current atmosphere and fly ash feeding conditions are maintained unchanged, the heat treatment process is adjusted and controlled to improve the dioxin low-temperature pyrolysis rate, and the fly ash dioxin low-temperature pyrolysis is optimized and controlled to form a fly ash dioxin low-temperature pyrolysis optimization system matched with the boiler body and the optimization control method.

[0066] In an optional embodiment, in the multi-modal data multi-flow multi-scale feature sharing neural network model constructed in step 1, the atmosphere data represents two attributes of atmosphere type and atmosphere concentration; the fly ash feeding data represents two attributes of fly ash dioxin concentration and fly ash feeding amount; and the heat treatment process data represents two attributes of pyrolysis temperature and heat treatment time; wherein,

[0067] The expression of the atmosphere data is as follows:

[0068]

[0069] In the formula, represents the atmosphere type, represents the atmosphere concentration;

[0070] The expression of the fly ash feed data is as follows:

[0071]

[0072] In the formula, represents the fly ash dioxin concentration, represents the fly ash feed amount;

[0073] The expression of the heat treatment process data is as follows:

[0074]

[0075] In the formula, represents the pyrolysis temperature, represents the heat treatment time.

[0076] In an optional embodiment, in step two, a multi-modal data multi-flow multi-scale feature sharing neural network model simulates the pyrolysis process in a data-driven manner, the model realizes multi-modal data multi-flow multi-scale feature sharing fusion and recognition, the multi-flow multi-scale features are calculated by multi-level multi-scale convolution kernels, and the model is optimized by using training samples, and specifically includes:

[0077] Input initial multi-modal data, the initial multi-modal data including atmosphere data, fly ash feed data, and heat treatment process data;

[0078] A first convolutional layer and a pooling layer are provided to extract and calculate features of the multi-modal data;

[0079] A second convolutional layer and a pooling layer are provided to extract and calculate features of the processed multi-modal data;

[0080] A feature sharing layer is provided to couple the multi-modal data after the first convolutional layer and the second convolutional layer to obtain a first layer feature coupling result and a second layer feature coupling result, respectively;

[0081] A feature splicing layer is provided to splice the first layer feature coupling result and the second layer feature coupling result to form spliced features in two convolutional scales;

[0082] The channel attention layer is used to weight and fuse the spliced features at two convolution scales, and then re-input the fused features into the feature splicing layer, and combine the atmosphere data in the initial input multi-modal data to form the final features for estimating the fly ash dioxin low-temperature pyrolysis rate.

[0083] The pooling layer and the softmax layer are provided, and the final features are input into the pooling layer and the softmax layer to obtain the estimation result of the dioxin pyrolysis rate under the current condition.

[0084] In an optional embodiment, the coupled data of the feature sharing layer includes: atmosphere concentration , fly ash dioxin concentration , fly ash feed amount , pyrolysis temperature , heat treatment time ;

[0085] Wherein, the first layer feature coupling after the first layer convolution layer feature is:

[0086]

[0087] Wherein, is the first layer convolution feature, is the first layer feature coupling weight, is the first layer feature coupling result.

[0088] The second layer feature coupling after the second layer convolution layer feature is:

[0089]

[0090] Wherein, is the second layer convolution feature, is the second layer feature coupling weight, is the second layer feature coupling result.

[0091] In an optional embodiment, the convolution kernel size of the first layer convolution layer is 1x7, the convolution kernel size of the second layer convolution layer is 1x5, and the size of the pooling layer is 3.

[0092] Wherein, the spliced feature expression under the convolution scale of the first layer convolution layer is:

[0093]

[0094] The spliced feature expression under the convolution scale of the second layer convolution layer is:

[0095]

[0096] In an optional embodiment, the spliced features at two convolution scales are weighted and fused by using a channel attention layer, and then re-input into the feature splicing layer, and combined with the atmosphere data in the initial input multi-modal data, to form a final feature expression for estimating the low-temperature thermal decomposition rate of fly ash dioxin:

[0097] In the formula, indicates the atmosphere type, , , , , are all weighted coefficients.

[0098] By collecting the atmosphere of the dioxin thermal decomposition process, the fly ash feed, the working condition data of the heat treatment process and the thermal decomposition rate as training samples, a training set with a sample amount of more than 6000 is established, and the model parameters are solidified after a large amount of training.

[0099] In an optional embodiment, in step four, the low-temperature thermal decomposition rate of dioxin is estimated based on the current low-temperature thermal decomposition rate of dioxin, the current atmosphere is maintained, and the fly ash feed conditions are unchanged. By adjusting and controlling the heat treatment process, an optimization function of maximizing the low-temperature thermal decomposition rate of dioxin y is established, and the fly ash low-temperature thermal decomposition heat treatment process under the condition of the maximum low-temperature thermal decomposition rate of dioxin is solved by using the gradient descent algorithm, wherein,

[0100]

[0101] In the formula, is the optimal pyrolysis temperature, is the optimal heat treatment time, is the optimal solution of the maximum low-temperature thermal decomposition rate of dioxin in the fly ash low-temperature thermal decomposition heat treatment process;

[0102] The optimal pyrolysis temperature and heat treatment time are fed back to the PLC control system of the low-temperature thermal decomposition boiler body, so as to realize the optimization control of the fly ash low-temperature thermal decomposition.

[0103] On the basis of accurate estimation of the dioxin thermal decomposition rate, under the existing working condition, the control variables of the optimal heat treatment process are fed back by maximizing the dioxin thermal decomposition rate under the condition of accurately measuring the atmosphere and fly ash feed, so as to optimize the objective function of maximizing the thermal decomposition rate, adjust the heat treatment process, and adaptively adjust the thermal decomposition process under different atmospheres and fly ash feed conditions, so as to improve the thermal decomposition rate of dioxin.

[0104] In an embodiment of the present application, a multi-modal data driven fly ash dioxin low-temperature thermal decomposition optimization control system is provided, comprising:

[0105] An acquisition module, the acquisition module is used to collect current atmosphere data, fly ash feed data and thermal treatment process data, and transmit multimodal data to the processing module;

[0106] A processing module is used to construct a neural network model that shares multi-modal data, multi-stream, and multi-scale features, and to optimize the model using multi-module data transmitted by the acquisition module to achieve an optimized solution for estimating the low-temperature pyrolysis emissions of fly ash dioxins.

[0107] In this embodiment, a computer device is provided, which may be a server. The computer device includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store static information and dynamic information data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the steps of the above-mentioned method embodiment are implemented.

[0108] Those skilled in the art will appreciate that the features described in the various embodiments of the present invention may be combined and / or coupled in various ways, even if such combinations and / or couplings are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention may be combined and / or coupled in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or couplings fall within the scope of the present invention.

Claims

1. A multimodal data driven fly ash dioxin low temperature pyrolysis optimization control method, characterized in that: The method includes: Step 1: Build a neural network model that shares multi-modal data, multi-stream, and multi-scale features to simulate the low-temperature pyrolysis process of fly ash dioxins; Step 2: Optimize the multi-stream and multi-scale feature sharing neural network model through data-driven approach; Step 3: Input the current atmosphere data, fly ash feed data, and thermal treatment process data into the multi-stream and multi-scale feature sharing neural network model to obtain an estimated value of the dioxin low-temperature thermal decomposition rate; Step 4: Based on the current estimated dioxin low-temperature pyrolysis rate, while maintaining the current atmosphere and the fly ash feed conditions, the dioxin low-temperature pyrolysis rate is increased through regulation and control of the heat treatment process, and the fly ash dioxin low-temperature pyrolysis is optimized and controlled, thereby forming a fly ash dioxin low-temperature pyrolysis optimization system in which the boiler body and the optimization control method are mutually matched; In the multimodal data multi-stream multi-scale feature sharing neural network model constructed in step 1, the input data is multimodal data, and the output data is dioxin pyrolysis rate y; the input multimodal data includes atmosphere data, fly ash feed data, and heat treatment process data; the atmosphere data represents two attributes: atmosphere type and atmosphere concentration; the fly ash feed data represents two attributes: fly ash dioxin concentration and fly ash feed amount; the heat treatment process data represents two attributes: pyrolysis temperature and heat treatment time; wherein, The expression of atmosphere data is as follows: Where, Indicates the type of atmosphere. Indicates the atmospheric concentration; The expression for fly ash feed data is as follows: Where, Indicates the dioxin concentration in fly ash, Indicates the fly ash feed amount; The expression of heat treatment process data is as follows: Where, represents the pyrolysis temperature, Indicates heat treatment time; In step 2, a neural network model with multi-modal data, multi-stream, and multi-scale feature sharing simulates the pyrolysis process in a data-driven manner and optimizes the model using training samples, including: Inputting initial multimodal data, the initial multimodal data includes multimodal data including atmosphere data, fly ash feed data, and heat treatment process data; Provide the first convolution layer and pooling layer to extract and calculate features of multimodal data; Provide the second convolution layer and pooling layer to further extract and calculate features of the processed multimodal data; Providing a feature sharing layer to couple the multimodal data after the first convolution layer and the second convolution layer respectively to obtain the first layer feature coupling results and the second layer feature coupling results respectively; Provide a feature splicing layer to splice the first layer feature coupling results and the second layer feature coupling results to form splicing features at two convolution scales; Providing a channel attention layer to weightedly fuse the spliced ​​features at the two convolutional scales and then re-inputting them into the feature splicing layer. The features are then combined with the atmosphere data in the initial input multimodal data to form the final features for estimating the low-temperature pyrolysis rate of fly ash dioxins. Provide a pooling layer and a softmax layer, input the final features into the pooling layer and the softmax layer, and make an estimate of the dioxin thermal decomposition rate under the current conditions; In step 4, based on the current estimated value of the dioxin low-temperature pyrolysis rate, the current atmosphere is maintained, the fly ash feed conditions remain unchanged, and the dioxin low-temperature pyrolysis rate is estimated. By regulating and controlling the thermal treatment process, an optimization function for maximizing the dioxin low-temperature pyrolysis rate y is established. The fly ash dioxin low-temperature pyrolysis heat treatment process under the condition of maximizing the dioxin low-temperature pyrolysis rate is solved by a gradient descent algorithm, wherein: Where, The optimal pyrolysis temperature For the optimal heat treatment time, It is the optimal solution for maximizing the low-temperature pyrolysis rate of dioxins in the process of low-temperature pyrolysis of dioxins in fly ash; The optimal pyrolysis temperature and heat treatment time are fed back to the PLC control system of the low-temperature pyrolysis boiler body to achieve optimized control of low-temperature pyrolysis of fly ash dioxins.

2. The multimodal data driven fly ash dioxin low temperature pyrolysis optimization control method according to claim 1, characterized in that: The coupled data of the feature sharing layer include: atmosphere concentration , fly ash dioxin concentration , fly ash feed rate , pyrolysis temperature , heat treatment time ; Among them, the first layer feature coupling after the first layer convolutional layer feature is: in, is the first layer of convolutional features, is the first layer feature coupling weight, is the first layer feature coupling result; After the second convolutional layer features, the second layer feature coupling is: in, is the second layer convolution feature, is the second layer feature coupling weight, It is the second-layer feature coupling result.

3. The multimodal data driven fly ash dioxin low temperature pyrolysis optimization control method according to claim 2, characterized in that: The convolution kernel size of the first convolution layer is 1×7, the convolution kernel size of the second convolution layer is 1×5, and the size of the pooling layer is 3; Among them, the splicing feature expression at the convolution scale of the first convolutional layer is: The convolution feature expression at the convolution scale of the second convolutional layer is:

4. The multimodal data driven fly ash dioxin low temperature pyrolysis optimization control method according to claim 3, characterized in that: The channel attention layer is used to perform weighted fusion of the splicing features at the two convolutional scales and then re-input them into the feature splicing layer. The features are then combined with the atmosphere data in the initial multimodal data to form the final feature expression for estimating the low-temperature pyrolysis rate of fly ash dioxins: Where, Indicates the type of atmosphere. 、 、 、 、 are all weighting coefficients.

5. A multimodal data driven fly ash dioxin low temperature pyrolysis optimization control system, used to implement a multimodal data driven fly ash dioxin low temperature pyrolysis optimization control method according to any one of claims 1 to 4, characterized in that: include: An acquisition module, the acquisition module is used to collect current atmosphere data, fly ash feed data and thermal treatment process data, and transmit multimodal data to the processing module; A processing module is used to construct a neural network model that shares multi-modal data, multi-stream, and multi-scale features, and to optimize the model using multi-module data transmitted by the acquisition module to achieve an optimized solution for estimating the low-temperature pyrolysis emissions of fly ash dioxins.

Images