Urban solid waste incineration operation optimization method and device, storage medium and computer equipment

Abstract

The invention provides an urban solid waste incineration operation optimization method and device, a storage medium and computer equipment. The optimization method comprises the steps that an urban solid waste incineration real-time data stream is obtained; a neural network integration model is trained according to the urban solid waste incineration real-time data flow, and an urban solid waste incineration optimization model is obtained; inputting the urban solid waste incineration operation variable and the related process variable at the optimization moment into an urban solid waste incineration optimization model to obtain an incineration optimization index; according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning, optimizing the incineration optimization index to obtain an optimal incineration decision result corresponding to the inflection point; and urban solid waste is incinerated according to the incineration decision result. According to the method, the optimal incineration decision result compatible with the advantages of high combustion stability and low pollutant emission can be obtained.

Description

Technical Field

[0001] The present application relates to the technical field of urban solid waste incineration, and specifically to an urban solid waste incineration operation optimization method, optimization device, storage medium and computer equipment. Background Art

[0002] Municipal solid waste incineration (MSWI) directly affects a city's environment. However, due to the complex composition of MSWI and the calorific value waveform during incineration, it is difficult to ensure the stability of incineration, which may lead to high pollutant emissions and low combustion efficiency. Summary of the invention

[0003] The purpose of this application is to provide a method, device, storage medium and computer equipment for optimizing the operation of urban solid waste incineration, which can overcome the shortcomings and deficiencies in the prior art.

[0004] A first aspect of an embodiment of the present application provides a method for optimizing the operation of municipal solid waste incineration, comprising:

[0005] S1: Obtain real-time data stream of municipal solid waste incineration;

[0006] S2: training a neural network integration model according to the real-time data stream of solid waste incineration to obtain an optimization model for urban solid waste incineration;

[0007] S3: Inputting the municipal solid waste incineration operation variables and related process variables at the optimization time into the municipal solid waste incineration optimization model to obtain the incineration optimization index;

[0008] S4: optimizing the multiple incineration optimization indicators according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain an optimal incineration decision result corresponding to the inflection point;

[0009] S5: Incinerate the urban solid waste according to the incineration decision result.

[0010] Further, the municipal solid waste incineration data stream includes continuously increasing continuous data blocks; the base model of the neural network integration model is RBFNN;

[0011] The step S2: training a neural network integration model according to the real-time data stream of municipal solid waste incineration to obtain a municipal solid waste incineration optimization model comprises:

[0012] S21: Set the maximum number of neural network ensemble base models, the learning rate and number of learning times of the neural network base model;

[0013] S22: According to the first data block, a base model based on RBFNN is established to obtain an initial incineration optimization model;

[0014] S23: according to the t-th data block, establish the t-th base model based on RBFNN, and test the performance of the t-1 historical base model according to the t-th data block;

[0015] S24: if t is less than or equal to the maximum number of integrated basis models, select all existing historical basis models; otherwise, select a historical basis model with a smaller test root mean square error for the t-th data block; and update the selected historical basis model according to the t-th data block;

[0016] S25: Determine the integration weights of each historical base model and the latest base model according to the test accuracy of the t-th data block to develop an optimization model for urban solid waste incineration;

[0017] S26: Increase the value of t and jump to step S23.

[0018] Further, the data block includes first data labeled as nitrogen oxide emission concentration and second data labeled as combustion efficiency.

[0019] Further, S4: optimizing the incineration optimization index according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain an incineration decision result corresponding to the inflection point, comprising:

[0020] S41: Obtaining convergence profile CA and diversity profile DA from the incineration optimization population through a dual-profile multi-objective particle swarm optimization search method;

[0021] S42: Obtaining a source domain training data set and a target domain training data set according to the convergence archive CA and the diversity archive DA;

[0022] S42: The source domain training data set and the target domain training data set are merged to construct a classifier for transfer learning, a high-quality solution is determined by the classifier to initialize the population, a dual-file multi-objective particle swarm optimizer is used to search and obtain the final optimization solution, and the optimal incineration decision result corresponding to the inflection point is determined.

[0023] A second aspect of the embodiment of the present application provides a device for optimizing the incineration of urban solid waste, comprising:

[0024] The real-time data stream acquisition module is used to execute S1: obtain the real-time data stream of urban solid waste incineration;

[0025] A neural network integrated model training module is used to execute S2: training a neural network integrated model according to the real-time data stream of urban solid waste incineration to obtain an urban solid waste incineration optimization model;

[0026] The incineration decision result acquisition module is used to execute S3: inputting the municipal solid waste incineration operation variables and related process variables at the optimization time into the municipal solid waste incineration optimization model to obtain the incineration optimization index;

[0027] A dynamic multi-objective particle swarm optimization module based on transfer learning is used to execute S4: optimizing the incineration optimization index according to the dynamic multi-objective particle swarm optimization module based on transfer learning to obtain the best incineration decision result corresponding to the inflection point;

[0028] The incineration module is used to execute S5: incinerate the urban solid waste according to the incineration decision result.

[0029] Further, the municipal solid waste incineration data stream includes continuously increasing continuous data blocks; the base model of the neural network integration model is RBFNN;

[0030] S2 performed by the network model training module: training a neural network integration model according to the real-time data stream of municipal solid waste incineration to obtain a municipal solid waste incineration optimization model, including:

[0031] S21: Set the maximum number of neural network ensemble base models, the learning rate and number of learning times of the neural network base model;

[0032] S22: According to the first data block, a base model based on RBFNN is established to obtain an initial incineration optimization model;

[0033] S23: according to the t-th data block, establish the t-th base model based on RBFNN, and test the performance of the t-1 historical base model according to the t-th data block;

[0034] S24: if t is less than or equal to the maximum number of integrated basis models, select all existing historical basis models; otherwise, select a historical basis model with a smaller test root mean square error for the t-th data block; and update the selected historical basis model according to the t-th data block;

[0035] S25: Determine the integration weights of each historical base model and the latest base model according to the test accuracy of the t-th data block to develop an optimization model for urban solid waste incineration;

[0036] S26: Increase the value of t and jump to step S23.

[0037] Further, the data block includes first data labeled as nitrogen oxide emission concentration and second data labeled as combustion efficiency.

[0038] Furthermore, the step S4 performed by the dynamic multi-objective particle swarm optimization module based on transfer learning: optimizing the incineration optimization index according to the dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain the incineration decision result corresponding to the inflection point includes:

[0039] S41: Obtaining convergence profile CA and diversity profile DA from the incineration optimization population through a dual-profile multi-objective particle swarm optimization search method;

[0040] S42: Obtaining a source domain training data set and a target domain training data set according to the convergence archive CA and the diversity archive DA;

[0041] S42: The source domain training data set and the target domain training data set are merged to construct a classifier for transfer learning, a high-quality solution is determined by the classifier to initialize the population, a dual-file multi-objective particle swarm optimizer is used to search and obtain the final optimization solution, and the optimal incineration decision result corresponding to the inflection point is determined.

[0042] A third aspect of an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the urban solid waste incineration operation optimization method as described above are implemented.

[0043] A fourth aspect of an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor implements the steps of the urban solid waste incineration operation optimization method as described above when executing the computer program.

[0044] Compared with the prior art, the urban solid waste incineration operation optimization method disclosed in the present application can use data training to establish an urban solid waste incineration optimization model, and then combine it with a dynamic multi-objective particle swarm optimization algorithm based on transfer learning for optimization and decision-making, so as to obtain the optimal incineration decision result that has the advantages of high combustion stability and low pollutant emissions.

[0045] In order to provide a clearer understanding of the present application, the specific implementation of the present application will be described below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a flow chart of a method for optimizing the operation of urban solid waste incineration according to an embodiment of the present application.

[0047] Figure 2 This is a schematic diagram of the step flow of a software program for a method for optimizing the operation of urban solid waste incineration according to an embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to make the objectives, technical solutions and advantages of the present application more clear, the embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0049] It should be clear that the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the embodiments of the present application.

[0050] When the following description relates to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. In the description of the present application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances. The singular forms of "a", "said" and "the" used in the present application and the appended claims are also intended to include the majority form, unless the context clearly indicates other meanings. The words "if" / "if" used herein can be interpreted as "at the time of" or "when" or "in response to determination".

[0051] In addition, in the description of this application, unless otherwise specified, "plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0052] See also Figure 1 , which is a flow chart of a method for optimizing the operation of urban solid waste incineration according to an embodiment of the present application, the method comprising:

[0053] S1: Obtain real-time data stream of municipal solid waste incineration;

[0054] S2: training a neural network integration model according to the real-time data stream of municipal solid waste incineration to obtain an optimization model for municipal solid waste incineration;

[0055] S3: Inputting the municipal solid waste incineration operation variables and related process variables at the optimization time into the municipal solid waste incineration optimization model to obtain the incineration optimization index;

[0056] S4: optimizing the incineration optimization index according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain an optimal incineration decision result corresponding to the inflection point;

[0057] S5: Incinerate the urban solid waste according to the incineration decision result.

[0058] In a feasible embodiment, the municipal solid waste incineration data stream includes continuously increasing continuous data blocks; the base model of the neural network integration model is RBFNN. The data blocks include first data labeled as nitrogen oxide (NOx) emission concentration and second data labeled as combustion efficiency (CE). Specifically, each data block in the municipal solid waste incineration data stream can be represented as Wherein, the subscript t is the serial number of the data block corresponding to the time, and the superscript r is the label. r=1 and 2 can be used to represent the first data with the label of nitrogen oxide emission concentration and the second data with the label of combustion efficiency respectively. Each data block can be represented as Among them, the subscripts 1, 2, and NS represent the first data, the second data, and the NSth data in the data block, and NS represents the maximum data size of the data block, X represents the input data in the data block, and Y represents the output data in the data block.

[0059] The step S2: training a neural network integration model according to the real-time data stream of solid waste incineration to obtain an optimization model for urban solid waste incineration comprises:

[0060] S21: Set the maximum number of neural network ensemble base models, the learning rate and number of learning times of the neural network base model;

[0061] S22: According to the first data block, a base model based on RBFNN is established to obtain an initial incineration optimization model;

[0062] S23: according to the t-th data block, establish the t-th base model based on RBFNN, and test the performance of the t-1 historical base model according to the t-th data block;

[0063] S24: if t is less than or equal to the maximum number of integrated basis models, select all existing historical basis models; otherwise, select a historical basis model with a smaller test root mean square error for the t-th data block; and update the selected historical basis model according to the t-th data block;

[0064] S25: Determine the integration weights of each historical base model and the latest base model according to the test accuracy of the t-th data block to develop an optimization model for urban solid waste incineration;

[0065] S26: Increase the value of t and jump to step S23.

[0066] In other embodiments, other neural network models besides RBFNN may be used.

[0067] In a feasible embodiment, S4: the step of optimizing the incineration optimization index according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain an incineration decision result corresponding to the inflection point includes:

[0068] S41: Obtaining convergence profile CA and diversity profile DA from the incineration optimization population through a dual-profile multi-objective particle swarm optimization search method;

[0069] S42: Obtaining a source domain training data set and a target domain training data set according to the convergence archive CA and the diversity archive DA;

[0070] Among them, the source domain training dataset is S his =CA t-1 ∪DA t-1 , the solution of CA and DA, namely CA t-1 and DA t-1 .

[0071] Target domain training dataset S cur =POP t,t-1 ∪POP t rand , POP t,t-1 The final population of the t-1th optimization is the population obtained by iterating once in the tth optimization, which is used as the training data of the target domain. t rand The population randomly generated for the t-th optimization, where t in the t-th optimization and the t-th data block has the same meaning, is used to perform the t-th optimization task based on the obtained t-th data block.

[0072] S42: The source domain training data set and the target domain training data set are merged to construct a classifier for transfer learning, a high-quality solution is determined by the classifier to initialize the population, a dual-file multi-objective particle swarm optimizer is used to search and obtain the final optimization solution, and the optimal incineration decision result corresponding to the inflection point is determined.

[0073] Compared with the prior art, the urban solid waste incineration operation optimization method disclosed in the present application can use data training to establish an urban solid waste incineration optimization model, and then combine it with a dynamic multi-objective particle swarm optimization algorithm based on transfer learning for optimization and decision-making, so as to obtain the optimal incineration decision result that has the advantages of high combustion stability and low pollutant emissions.

[0074] It should be noted that the urban solid waste incineration operation optimization method disclosed in this application is also a dynamic multi-objective operation optimization method based on transfer learning (DSE-TrMOPSO). On the one hand, its modeling is based on data stream integrated learning, and uses incremental update and selective integration strategies to adapt to the dynamic operation characteristics under operating condition drift. On the one hand, operating condition drift adaptation is achieved by gradually establishing new base models and updating historical members. On the other hand, the selective integration of historical base models can realize the knowledge utilization of historical data.

[0075] On the other hand, the optimization method adopted is the dynamic multi-objective particle swarm optimization algorithm based on online transfer learning (CTrDMOPSO), which is a dynamic MOPSO algorithm based on transfer learning. Considering the optimization characteristics of the MSWI process, the transfer learning technique is used to improve the quality of the initial solution. Most importantly, a hierarchical clustering-based strategy is adopted to promote transfer learning between similar clusters, and a dual-archive multi-objective particle swarm optimizer is used to improve the convergence and diversity of the transfer solution, thereby reducing negative transfer.

[0076] See also Figure 2 In a feasible embodiment, the urban solid waste incineration operation optimization method of the present application can be implemented based on a software program drive, wherein the step flow content of the software program can be as follows: Figure 2 That is, the modeling steps and optimization steps of the urban solid waste incineration operation optimization method of this application are the same as Figure 2 The steps of the software program shown correspond to the contents of the flow chart. In the figure, Eqs refers to multiple formulas, and Eq refers to a single formula.

[0077] Wherein, formula (1) is as follows:

[0078]

[0079] in, is the Gaussian function of the input data, is the i-th input data, c j is the center of the jth hidden layer neuron, σ j is the width of the jth hidden layer neuron.

[0080] Formula (2) is as follows:

[0081]

[0082] in, is the predicted output of the i-th input data, w j is the weight coefficient, and h is the bias weight.

[0083] Formula (3) is as follows:

[0084] Φ·w=Y r , (3)

[0085] Among them, Φ is expressed as:

[0086]

[0087] Weight w=[w1,w2,...,w J ,η] T , calculated by the following formula

[0088] w=(Φ T ·Φ) -1 ·Φ T ·Y r . (5)

[0089] Formula (6) is as follows:

[0090]

[0091] Among them, Q 1 is the evaluation model of CE, Q 2 is the evaluation model of NOx emission concentration, W1 r =1, It is the first base model and serves as the initial urban solid waste incineration optimization model.

[0092] The test root mean square error of the historical base model on the tth data block is obtained by formula (7):

[0093]

[0094] in, is the test root mean square error of the qth historical base model on the tth data block, is the predicted output of the qth historical base model corresponding to the i-th input data, Y i r is the actual output data of the i-th input data.

[0095] Formula (8) is as follows:

[0096] Θ iter+1 =Θ iter -(Q iter +λ iter I) -1 g iter , (8)

[0097] Among them, Θ is the model parameter, which represents all the parameters that need to be learned, iter is the number of learning times, I is the unit matrix, and λ is the learning rate.

[0098] The ensemble weights of the latest base model are:

[0099] W t r =0.5, (9)

[0100] Among them, W t r is the ensemble weight of the latest base model.

[0101] The ensemble weights of each selected historical base model are:

[0102]

[0103] in, The historical base model selected for the kth time The integration weight of is the test root mean square error of the historical basis model.

[0104] Finally, the developed optimization model for municipal solid waste incineration is shown in formula (11):

[0105]

[0106] Among them, Q r (X r (t)) is the output of the urban solid waste incineration optimization model, that is, the incineration optimization index.

[0107] Formula (13) is shown in the figure below:

[0108]

[0109] in, for and The angle of for The mth normalized target value of for The mth normalized target value of the source domain training dataset S his Perform hierarchical clustering into H categories, the jth category The centroid of Training dataset S for target domain cur Perform hierarchical clustering into H categories, the i-th category The centroid of

[0110] Therefore, each category in the source domain is related to the i-th category in the target domain The calculation of the matching probability is shown in formula (14):

[0111]

[0112] is the matching probability between the jth category in the source domain and the ith category in the target domain.

[0113] Transfer learning technology TrAdaboost is an integrated learning method. Each weak classifier converts the sample x Mapped to labels y(x)∈{0,1}.

[0114] The basic principle of TrAdaboost is to filter out source domain data that is different from the target domain data and make full use of source domain data that is similar to the target domain data by adjusting the sample weights. This learning process corresponds to transferring knowledge from the source domain to the target domain. Therefore, the weight of the sample is initialized as shown in formula (15) and updated as shown in formula (16):

[0115]

[0116] in, is the weight of the updated sample, J gen (x) is the weak classifier J gen At the genth iteration, the predicted label of the input data x has a maximum number of iterations of gen max , for The number of samples in for The number of samples in μ gen is the weak classifier J gen The weight coefficient is, where μ can be obtained by formula (17):

[0117]

[0118] In the genth iteration, the weak classifier J gen for The error rate is shown in formula (18):

[0119]

[0120] Among them, α gen is the error rate.

[0121] According to the error rate, the weight coefficient is calculated as shown in formula (19):

[0122] μ gen =a gen / (1-a gen ), (19)

[0123] Among them, μ gen is the weight coefficient.

[0124] Each iteration trains a weak classifier, and when the maximum number of iterations is met, the trained genmax The weak classifiers are combined through weights to output a strong classifier, where the strong classifier is shown in formula (20):

[0125]

[0126] Among them, J * is a strong classifier.

[0127] After obtaining the strong classifier J * Later, a large number of test solutions can be randomly generated and input into the strong classifier. According to the classifier results, a better solution is determined as the initial solution of the population, and a dual-file multi-objective particle swarm optimizer is used to search and obtain the final optimized solution.

[0128] A second aspect of the embodiment of the present application provides a device for optimizing the incineration of urban solid waste, comprising:

[0129] The real-time data stream acquisition module is used to execute S1: obtain the real-time data stream of urban solid waste incineration;

[0130] A neural network integrated model training module is used to execute S2: training a neural network integrated model according to the real-time data stream of urban solid waste incineration to obtain an urban solid waste incineration optimization model;

[0131] The incineration decision result acquisition module is used to execute S3: inputting the municipal solid waste incineration operation variables and related process variables at the optimization time into the municipal solid waste incineration optimization model to obtain the incineration optimization index;

[0132] A dynamic multi-objective particle swarm optimization module based on transfer learning is used to execute S4: optimizing the incineration optimization index according to the dynamic multi-objective particle swarm algorithm based on transfer learning to obtain the best incineration decision result corresponding to the inflection point;

[0133] The incineration module is used to execute S5: incinerate the urban solid waste according to the incineration decision result.

[0134] In a feasible embodiment, the solid waste incineration data stream includes continuously increasing continuous data blocks; the base model of the neural network integration model is RBFNN;

[0135] S2 performed by the neural network integrated model training module: training the neural network integrated model according to the real-time data stream of urban solid waste incineration to obtain the step of urban solid waste incineration optimization model, including:

[0136] S21: Set the maximum number of neural network ensemble base models, the learning rate and number of learning times of the neural network base model;

[0137] S22: According to the first data block, a base model based on RBFNN is established to obtain an initial incineration optimization model;

[0138] S23: according to the t-th data block, establish the t-th base model based on RBFNN, and test the performance of the t-1 historical base model according to the t-th data block;

[0139] S24: if t is less than or equal to the maximum number of integrated basis models, select all existing historical basis models; otherwise, select a historical basis model with a smaller test root mean square error for the t-th data block; and update the selected historical basis model according to the t-th data block;

[0140] S25: Determine the integration weights of each selected historical base model and the latest base model according to the test accuracy of the t-th data block to develop an optimization model for municipal solid waste incineration;

[0141] S26: Increase the value of t and jump to step S23.

[0142] In a feasible embodiment, the data block includes first data labeled as nitrogen oxide emission concentration and second data labeled as combustion efficiency.

[0143] In a feasible embodiment, the step S4 performed by the dynamic multi-objective particle swarm optimization module based on transfer learning: optimizing the incineration optimization index according to the dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain the incineration decision result corresponding to the inflection point includes:

[0144] S41: Obtaining convergence profile CA and diversity profile DA from the incineration optimization population through a dual-profile multi-objective particle swarm optimization search method;

[0145] S42: Obtaining a source domain training data set and a target domain training data set according to the convergence archive CA and the diversity archive DA;

[0146] S42: Merge the source domain training data set and the target domain training data set to construct a classifier for transfer learning, determine a high-quality solution initialization population through the classifier, use a dual-file multi-objective particle swarm optimizer to search and obtain the final optimization solution, and determine the best incineration decision result corresponding to the inflection point.

[0147] It should be noted that the urban solid waste incineration optimization device provided in the second embodiment of the present application only uses the division of the above-mentioned functional modules as an example when executing the urban solid waste incineration operation optimization method. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the equipment is divided into different functional modules to complete all or part of the functions described above. In addition, the urban solid waste incineration optimization device provided in the second embodiment of the present application and the urban solid waste incineration operation optimization method of the first embodiment of the present application belong to the same concept. The implementation process is detailed in the method embodiment, which will not be repeated here.

[0148] A third aspect of an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the urban solid waste incineration optimization method described above are implemented.

[0149] A fourth aspect of an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor implements the steps of the urban solid waste incineration operation optimization method as described above when executing the computer program.

[0150] The device embodiments described above are merely illustrative, wherein the components described as separate parts may or may not be physically separated, and the parts displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application. Ordinary technicians in this field can understand and implement it without creative work.

[0151] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0152] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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 generate 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 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the function selected in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 function selected in a box or multiple boxes.

[0153] 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. Figure 1 A process or multiple processes and / or boxes Figure 1 steps for the function selected in a box or multiple boxes.

[0154] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0155] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0156] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0157] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0158] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for optimizing the operation of municipal solid waste incineration, characterized in that: include: S1: Obtain real-time data stream of municipal solid waste incineration; S2: training a neural network integration model according to the real-time data stream of municipal solid waste incineration to obtain an optimization model for municipal solid waste incineration; S3: Inputting the municipal solid waste incineration operation variables and related process variables at the optimization time into the municipal solid waste incineration optimization model to obtain the incineration optimization index; S4: optimizing the incineration optimization index according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain an optimal incineration decision result corresponding to the inflection point; S5: Incinerate the urban solid waste according to the incineration decision result.

2. The method for optimizing the operation of municipal solid waste incineration according to claim 1, characterized in that: The municipal solid waste incineration data stream includes continuously increasing continuous data blocks; the base model of the neural network integration model is RBFNN; The step S2: training a neural network integration model according to the real-time data stream of municipal solid waste incineration to obtain a municipal solid waste incineration optimization model comprises: S21: Set the maximum number of neural network ensemble base models, the learning rate and number of learning times of the neural network base model; S22: According to the first data block, a base model based on RBFNN is established to obtain an initial incineration optimization model; S23: according to the t-th data block, establish the t-th base model based on RBFNN, and test the performance of the t-1 historical base model according to the t-th data block; S24: if t is less than or equal to the maximum number of integrated basis models, select all existing historical basis models; otherwise, select a historical basis model with a smaller test root mean square error for the t-th data block; and update the selected historical basis model according to the t-th data block; S25: Determine the integration weights of each historical base model and the latest base model according to the test accuracy of the t-th data block to develop an optimization model for urban solid waste incineration; S26: Increase the value of t and jump to step S23.

3. The method for optimizing the operation of municipal solid waste incineration according to claim 2, characterized in that: The data block includes first data labeled as nitrogen oxide emission concentration and second data labeled as combustion efficiency.

4. The method for optimizing the operation of municipal solid waste incineration according to claim 1, characterized in that: S4: The step of optimizing the incineration optimization index according to the dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain the incineration decision result corresponding to the inflection point includes: S41: Obtaining convergence profile CA and diversity profile DA from the incineration optimization population through a dual-profile multi-objective particle swarm optimization search method; S42: Obtaining a source domain training data set and a target domain training data set according to the convergence archive CA and the diversity archive DA; S42: The source domain training data set and the target domain training data set are merged to construct a classifier for transfer learning, a high-quality solution is determined by the classifier to initialize the population, a dual-file multi-objective particle swarm optimizer is used to search and obtain the final optimization solution, and the optimal incineration decision result corresponding to the inflection point is determined.

5. A municipal solid waste incineration optimization device, characterized in that: include: The real-time data stream acquisition module is used to execute S1: obtain the real-time data stream of urban solid waste incineration; A neural network integrated model training module is used to execute S2: training a neural network integrated model according to the real-time data stream of urban solid waste incineration to obtain an urban solid waste incineration optimization model; The incineration decision result acquisition module is used to execute S3: inputting the municipal solid waste incineration operation variables and related process variables at the optimization time into the municipal solid waste incineration optimization model to obtain the incineration optimization index; A dynamic multi-objective particle swarm optimization module based on transfer learning is used to execute S4: optimizing the incineration optimization index according to a dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain an optimal incineration decision result corresponding to an inflection point; The incineration module is used to execute S5: incinerate the urban solid waste according to the incineration decision result.

6. The urban solid waste incineration optimization device according to claim 5, characterized in that: The municipal solid waste incineration data stream includes continuously increasing continuous data blocks; the base model of the neural network integration model is RBFNN; S2 performed by the neural network integrated model training module: training the neural network integrated model according to the real-time data stream of solid waste incineration to obtain the step of optimizing the urban solid waste incineration model, including: S21: Set the maximum number of neural network ensemble base models, the learning rate and number of learning times of the neural network base model; S22: According to the first data block, a base model based on RBFNN is established to obtain an initial incineration optimization model; S23: according to the t-th data block, establish the t-th base model based on RBFNN, and test the performance of the t-1 historical base model according to the t-th data block; S24: if t is less than or equal to the maximum number of integrated basis models, select all existing historical basis models; otherwise, select a historical basis model with a smaller test root mean square error for the t-th data block; and update the selected historical basis model according to the t-th data block; S25: Determine the integration weights of each historical base model and the latest base model according to the test accuracy of the t-th data block to develop an optimization model for urban solid waste incineration; S26: Increase the value of t and jump to step S23.

7. The urban solid waste incineration optimization device according to claim 6, characterized in that: The data block includes first data labeled as nitrogen oxide emission concentration and second data labeled as combustion efficiency.

8. The urban solid waste incineration optimization device according to claim 5, characterized in that: S4 performed by the dynamic multi-objective particle swarm optimization module based on transfer learning: optimizing the incineration optimization index according to the dynamic multi-objective particle swarm optimization algorithm based on transfer learning to obtain the incineration decision result corresponding to the inflection point, including: S41: Obtaining convergence profile CA and diversity profile DA from the incineration optimization population through a dual-profile multi-objective particle swarm optimization search method; S42: Obtaining a source domain training data set and a target domain training data set according to the convergence archive CA and the diversity archive DA; S42: The source domain training data set and the target domain training data set are merged to construct a classifier for transfer learning, a high-quality solution is determined by the classifier to initialize the population, a dual-file multi-objective particle swarm optimizer is used to search and obtain the final optimization solution, and the optimal incineration decision result corresponding to the inflection point is determined.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the urban solid waste incineration operation optimization method as described in any one of claims 1 to 4 are implemented.

10. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable by the processor, wherein the processor implements the steps of the urban solid waste incineration operation optimization method as described in any one of claims 1 to 4 when executing the computer program.

Images