Energy-saving optimization method and optimization system for electric dust removal system of thermal power generating unit

Abstract

The application discloses a power plant electric dust removal system energy-saving optimization method and an optimization system, solves the problem that the existing optimization method only considers the operation characteristic parameter data of the unit boiler and the coal mill when constructing a coal quality neural network mathematical model, and does not introduce real-time dust parameters during the load operation of the power plant, so that the model evaluation precision is poor, and the method comprises the following steps: obtaining a power plant operation three-dimensional graph based on a power plant database, deploying a dust removal monitoring point through the power plant operation three-dimensional graph; constructing an electric dust removal optimization model, the electric dust removal optimization model calculates the load weight of the dust removal monitoring point and the dust removal load rate; the application constructs an electric dust removal optimization model based on the dust data stored in the power plant database and the operation parameters of the power plant, and then the electric dust removal optimization model accurately calculates the load weight of the dust removal monitoring point and the dust removal load rate, thereby facilitating the power plant electric dust removal system to adjust the start-stop and load of the electric dust remover.

Description

Technical Field

[0001] This invention belongs to the field of power plant dust removal technology, specifically relating to energy-saving optimization methods and optimization systems for electrostatic precipitators in thermal power units. Background Technology

[0002] Electrostatic precipitators are key equipment in coal-fired power plants for removing particulate matter from flue gas. With increasingly stringent environmental regulations, power plants need to adopt more efficient dust removal technologies and control strategies to meet ultra-low emission standards.

[0003] DCS (Distributed Control System) is an automated control system widely used in industrial production processes. It can achieve comprehensive control of equipment such as boilers, steam turbines, and generators. In power plants, optimizing the electrostatic precipitator control system is crucial for improving dust removal efficiency and reducing energy consumption.

[0004] Chinese patent CN114563993A discloses an energy-saving optimization method and system for an electrostatic precipitator system in a thermal power unit. The method includes: acquiring the operating characteristic parameter data of the unit's boiler and coal mill; establishing a coal quality neural network mathematical model based on the parameter data; establishing a coal quality prediction neural network DCS configuration based on the coal quality neural network mathematical model; and incorporating the coal quality prediction neural network DCS configuration into the thermal power unit's DCS system to obtain an optimized DCS system for the thermal power unit. However, existing optimization methods only consider the operating characteristic parameter data of the unit's boiler and coal mill when constructing the coal quality neural network mathematical model, without incorporating real-time dust parameters during the thermal power unit's load operation, resulting in poor model evaluation accuracy. To solve the above problems, we propose an energy-saving optimization method and system for an electrostatic precipitator system in a thermal power unit. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by providing an energy-saving optimization method and system for electrostatic precipitators in thermal power units. This invention solves the problem that existing optimization methods only consider the operating characteristic parameters of the boiler and coal mill when constructing the coal quality neural network mathematical model, without incorporating real-time dust parameters during the load operation of the thermal power unit, resulting in poor model evaluation accuracy.

[0006] Existing optimization methods only consider the operating characteristic parameters of the boiler and coal mill when constructing the coal quality neural network mathematical model, without incorporating real-time dust parameters during the operation of the thermal power unit under load. This results in poor model evaluation accuracy. To address this issue, we propose an energy-saving optimization method and system for the electrostatic precipitator (ESP) system of thermal power units. In short, the optimization method includes obtaining a three-dimensional diagram of the thermal power unit's operation based on a thermal power unit database, deploying dust monitoring points using the three-dimensional diagram, constructing an ESP optimization model based on the dust data stored in the thermal power unit database and the thermal power unit's operating parameters, and then executing the pre-constructed ESP optimization model using the dust data collected from the dust monitoring points and the thermal power unit's operating parameters as input. The ESP optimization model calculates the load weights of the dust monitoring points and the dust load rate. This invention constructs an electrostatic precipitator (ESP) optimization model based on dust data stored in a thermal power unit database and the operating parameters of the thermal power unit. The ESP optimization model then accurately calculates the load weight and dust removal load rate of the dust removal monitoring points, thereby facilitating the adjustment of the ESP's start-up and shutdown and load in the thermal power unit's ESP system. It also facilitates real-time monitoring of dust concentration in the thermal power unit, avoids dust accumulation during operation, reduces wear on the blades of the suction fan and booster fan, and extends the service life of the equipment. Compared with existing technologies, it offers higher dust removal accuracy and efficiency.

[0007] This invention is implemented as follows: an energy-saving optimization method for electrostatic precipitator systems in thermal power units, comprising:

[0008] A 3D map of thermal power unit operation is obtained from the thermal power unit database, and dust monitoring points are deployed using the 3D map of thermal power unit operation.

[0009] Obtain dust data collected from dust removal monitoring points, upload the dust data and thermal power unit operating parameters to the thermal power unit database, and construct an electrostatic precipitator optimization model based on the dust data and thermal power unit operating parameters stored in the thermal power unit database.

[0010] Using dust data collected from dust monitoring points and operating parameters of thermal power units as inputs, a pre-built electrostatic precipitator optimization model is executed. The electrostatic precipitator optimization model calculates the load weight and dust removal load rate of the dust monitoring points, and generates dust removal control commands based on the load weight and dust removal load rate of the dust monitoring points.

[0011] Preferably, the energy-saving optimization method for the electrostatic precipitator system of the thermal power unit further includes:

[0012] Upon receiving dust removal control commands, the electrostatic precipitator system of the thermal power unit controls the opening and closing of dust removal monitoring points and adjusts the dust removal power of the monitoring points based on the dust removal control commands.

[0013] Preferably, the method for obtaining a three-dimensional diagram of thermal power unit operation based on a thermal power unit database specifically includes:

[0014] Establish a communication connection with the thermal power unit database and retrieve the thermal power unit facility, pipeline, and valve parameter information stored in the thermal power unit database;

[0015] Load the parameter information of the thermal power unit facilities, pipelines and valves, and establish a thermal power unit architecture network based on the parameter information of the thermal power unit facilities, pipelines and valves. In the thermal power unit architecture network, the facilities and valves are set as single initial monitoring points, and the pipelines are set with multiple initial monitoring points evenly distributed. The multiple initial monitoring points are integrated and connected to form the thermal power unit architecture network.

[0016] The initial weights of dust distribution in thermal power unit facilities, pipelines, and valves are determined based on parameter information of thermal power unit facilities, pipelines, and valves.

[0017] Preferably, the method for obtaining a three-dimensional diagram of thermal power unit operation based on a thermal power unit database further includes:

[0018] The initial weights of dust distribution in thermal power unit facilities, pipelines and valves are obtained, and the dust propagation level of the initial monitoring points in the thermal power unit architecture network is determined based on the YOLO Pose algorithm and the key point detection model.

[0019] Based on the preset level threshold, determine whether the dust transmission level exceeds the preset level threshold. If it exceeds the preset level threshold, determine the current initial monitoring point as the dust removal monitoring point.

[0020] If the preset level threshold is not exceeded, adjacent initial monitoring points are merged into dust removal monitoring points based on the key point detection model;

[0021] Based on the adjusted dust monitoring points and thermal power unit architecture network, a three-dimensional diagram of thermal power unit operation was determined.

[0022] Preferably, the method for determining the dust propagation level of initial monitoring points in the thermal power unit architecture network based on the YOLO Pose algorithm and key point detection model specifically includes:

[0023] The key point detection model captures the first propagation level based on the initial weights of the dust distribution at the initial monitoring points;

[0024] The initial monitoring point corresponding to the first propagation level was determined as the priority monitoring point, and the dust propagation level of the priority monitoring point was determined to be 5. The remaining initial monitoring points were designated as ordinary monitoring points.

[0025] Calculate the dust propagation level at ordinary monitoring points based on the YOLO Pose algorithm;

[0026]

[0027] Where g is the dust propagation level of ordinary monitoring points, α is the asymmetric function constant of the YOLO Pose algorithm, f(a) is the dust propagation coefficient of ordinary monitoring points, W(a) is the weight correction function of the YOLO Pose algorithm, n is the number of ordinary monitoring points, and a is the initial weight of dust distribution.

[0028]

[0029]

[0030] Where a is the initial weight of the dust distribution, and β is the weight penalty factor of the weight correction function.

[0031] Preferably, the method for constructing an electrostatic precipitator optimization model based on dust data stored in the thermal power unit database and the operating parameters of the thermal power unit specifically includes:

[0032] Traverse the thermal power unit database and retrieve the dust data and operating parameters of the thermal power units stored in the database.

[0033] Dust data and thermal power unit operating parameters are preprocessed for dimensionality reduction, and normalized dimensionality-reduced feature vectors are output.

[0034] Load the dimensionality-reduced feature vectors and divide them into training and test sets, with a sample ratio of 3:1 between the training and test sets;

[0035] A hybrid forest model was obtained by combining a random forest model with a debugger tool, and this hybrid forest model was used as the initial model for the electrostatic precipitator optimization model.

[0036] The model is trained based on batch sampling softmax loss, with N training iterations performed on the initial model until convergence.

[0037] Load the test set, detect the optimization accuracy of the initial model based on the test set, determine whether the optimization accuracy is greater than the preset accuracy threshold, and if it is greater than the accuracy threshold, output the electrostatic precipitator optimization model.

[0038] Preferably, the method for executing the pre-built electrostatic precipitator optimization model and calculating the load weights of dust collection monitoring points and the dust collection load rate specifically includes:

[0039] The dust data collected from the dust monitoring points and the operating parameters of the thermal power units are preprocessed by the electrostatic precipitator optimization model to reduce the dimensionality of the dust data and the operating parameters of the thermal power units, and the normalized dimensionality-reduced feature vector is output.

[0040] Among them, when the electrostatic precipitator optimization model performs dimensionality reduction preprocessing on dust data and thermal power unit operating parameters, the dimensionality reduction of the data is achieved through the autoencoder of the electrostatic precipitator optimization model;

[0041]

[0042] Where, d τ For the dimensionality reduction features of the data, This indicates dimensionality reduction processing of the data. dm Describes the dimensionality reduction target parameters. x m The input data consists of dust data and operating parameters of thermal power units, where m is the number of dimensions for the dust data and operating parameters of thermal power units.

[0043]

[0044] Where m is the number of dimensions of the dust data and the operating parameters of the thermal power unit, and k is the data type of the dust data and the operating parameters of the thermal power unit.

[0045] Preferably, the method for executing the pre-built electrostatic precipitator optimization model and calculating the load weights of dust collection monitoring points and the dust collection load rate further includes:

[0046] Load the normalized dimensionality-reduced feature vector and calculate the load weight of the dust monitoring points;

[0047]

[0048] Where h(x) is the load weight value of the dust monitoring point, D is the total number of normalized dimensionality-reduced feature vectors of the input, and d τ Input the dimensionality-reduced feature vector values;

[0049] Using the load weight of the dust removal monitoring points as input, the electrostatic precipitator optimization model is executed, and the dust removal load rate of the dust removal monitoring points is output.

[0050]

[0051] Where h(d) τ ) represents the dust removal load rate of the dust removal monitoring point, β represents the scale of the correction function in the electrostatic precipitator optimization model, h(x) represents the load weight value of the dust removal monitoring point, and D represents the total number of normalized dimensionality-reduced feature vectors of the input.

[0052] On the other hand, embodiments of the present invention also provide an energy-saving optimization system for an electrostatic precipitator system of a thermal power unit, the energy-saving optimization system for an electrostatic precipitator system of a thermal power unit comprising:

[0053] The monitoring point deployment module obtains a 3D map of thermal power unit operation based on the thermal power unit database and deploys dust removal monitoring points using the 3D map of thermal power unit operation.

[0054] An optimization model building module is used to acquire dust data collected from dust monitoring points, upload the dust data and thermal power unit operating parameters to the thermal power unit database, and build an electrostatic precipitator optimization model based on the dust data and thermal power unit operating parameters stored in the thermal power unit database.

[0055] The load rate calculation module takes dust data collected from dust monitoring points and operating parameters of thermal power units as input, executes a pre-built electrostatic precipitator optimization model, calculates the load weight and dust load rate of the dust monitoring points, and generates dust control instructions based on the load weight and dust load rate of the dust monitoring points.

[0056] The optimization execution module is used to acquire dust removal control commands. The electrostatic precipitator system of the thermal power unit controls the opening and closing of dust removal monitoring points based on the dust removal control commands, and adjusts the dust removal power of the dust removal monitoring points.

[0057] Preferably, the load factor calculation module specifically includes:

[0058] The normalized dimensionality reduction unit is based on the dust data collected from the dust monitoring points and the operating parameters of the thermal power unit. The electrostatic precipitator optimization model performs dimensionality reduction preprocessing on the dust data and the operating parameters of the thermal power unit, and outputs a normalized dimensionality reduction feature vector.

[0059] A load weight determination unit is used to load a normalized dimensionality-reduced feature vector and calculate the load weight of the dust monitoring point.

[0060] The load rate determination unit takes the load weight of the dust removal monitoring point as input, executes the electrostatic precipitator optimization model, and outputs the dust removal load rate of the dust removal monitoring point.

[0061] Compared with the prior art, the embodiments of this application have the following main advantages:

[0062] This invention constructs an electrostatic precipitator (ESP) optimization model based on dust data stored in a thermal power unit database and the operating parameters of the thermal power unit. The ESP optimization model then accurately calculates the load weight and dust removal load rate of the dust removal monitoring points, thereby facilitating the adjustment of the ESP's start-up and shutdown and load in the thermal power unit's ESP system. It also facilitates real-time monitoring of dust concentration in the thermal power unit, avoids dust accumulation during operation, reduces wear on the blades of the suction fan and booster fan, and extends the service life of the equipment. Compared with existing technologies, it offers higher dust removal accuracy and efficiency.

[0063] This invention constructs a three-dimensional map of thermal power unit operation based on a thermal power unit database. Dust removal monitoring points are determined through the three-dimensional map of thermal power unit operation. At the same time, merging adjacent initial monitoring points into dust removal monitoring points based on the key point detection model can also avoid the waste of resources caused by deploying too many dust removal monitoring points.

[0064] In this embodiment of the invention, the electrostatic precipitator optimization model is trained based on batch sampling softmax loss. The training of the electrostatic precipitator optimization model is based on a combination of random forest model and debugger tool, so that the trained and constructed electrostatic precipitator optimization model can accurately evaluate and calculate the load weight of dust collection monitoring points and the dust collection load rate. Attached Figure Description

[0065] Figure 1 This is a schematic diagram illustrating the implementation process of the energy-saving optimization method for the electrostatic precipitator system of thermal power units provided by the present invention.

[0066] Figure 2 This is a schematic diagram illustrating the implementation process of the method for obtaining a three-dimensional image of a thermal power unit based on a thermal power unit database provided by the present invention.

[0067] Figure 3 This is a schematic diagram illustrating the implementation process of the method for determining the dust propagation level of initial monitoring points in a thermal power unit architecture network based on the YOLO Pose algorithm and key point detection model provided by this invention.

[0068] Figure 4 This is a schematic diagram illustrating the implementation process of the method for constructing an electrostatic precipitator optimization model based on dust data stored in a thermal power unit database and operating parameters of thermal power units, as provided by the present invention.

[0069] Figure 5 This is a schematic diagram illustrating the implementation process of the pre-built electrostatic precipitator optimization model provided by the present invention, which calculates the load weight of dust monitoring points and the dust load rate.

[0070] Figure 6 This is a schematic diagram of the energy-saving optimization system for the electrostatic precipitator system of thermal power units provided by the present invention.

[0071] Figure 7 This is a schematic diagram of the load rate calculation module provided by the present invention. Detailed Implementation

[0072] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0073] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0074] Existing optimization methods only consider the operating characteristic parameters of the boiler and coal mill when constructing the coal quality neural network mathematical model, without incorporating real-time dust parameters during the operation of the thermal power unit under load. This results in poor model evaluation accuracy. To address this issue, we propose an energy-saving optimization method and system for the electrostatic precipitator (ESP) system of thermal power units. In short, the optimization method includes obtaining a three-dimensional diagram of the thermal power unit's operation based on a thermal power unit database, deploying dust monitoring points using the three-dimensional diagram, constructing an ESP optimization model based on the dust data stored in the thermal power unit database and the thermal power unit's operating parameters, and then executing the pre-constructed ESP optimization model using the dust data collected from the dust monitoring points and the thermal power unit's operating parameters as input. The ESP optimization model calculates the load weights of the dust monitoring points and the dust load rate. This invention constructs an electrostatic precipitator (ESP) optimization model based on dust data stored in a thermal power unit database and the operating parameters of the thermal power unit. The ESP optimization model then accurately calculates the load weight and dust removal load rate of the dust removal monitoring points, thereby facilitating the adjustment of the ESP's start-up and shutdown and load in the thermal power unit's ESP system. It also facilitates real-time monitoring of dust concentration in the thermal power unit, avoids dust accumulation during operation, reduces wear on the blades of the suction fan and booster fan, and extends the service life of the equipment. Compared with existing technologies, it offers higher dust removal accuracy and efficiency.

[0075] This invention provides an energy-saving optimization method for electrostatic precipitator systems in thermal power units. Figure 1 A schematic diagram illustrating the implementation process of the energy-saving optimization method for the electrostatic precipitator system of a thermal power unit is shown. The energy-saving optimization method for the electrostatic precipitator system of a thermal power unit specifically includes:

[0076] Step S10: Obtain a 3D diagram of thermal power unit operation based on the thermal power unit database, and deploy dust monitoring points using the 3D diagram of thermal power unit operation.

[0077] Step S20: Obtain dust data collected from dust removal monitoring points; based on the dust data and thermal power unit operating parameters, upload the dust data and thermal power unit operating parameters to the thermal power unit database; and construct an electrostatic precipitator optimization model based on the dust data and thermal power unit operating parameters stored in the thermal power unit database.

[0078] Step S30: Using the dust data collected from the dust monitoring points and the operating parameters of the thermal power unit as inputs, execute the pre-built electrostatic precipitator optimization model. The electrostatic precipitator optimization model calculates the load weight and dust removal load rate of the dust monitoring points, and generates dust removal control instructions based on the load weight and dust removal load rate of the dust monitoring points.

[0079] Step S40: Obtain dust removal control instructions. The electrostatic precipitator system of the thermal power unit controls the opening and closing of dust removal monitoring points based on the dust removal control instructions, and adjusts the dust removal power of the dust removal monitoring points.

[0080] This invention constructs an electrostatic precipitator (ESP) optimization model based on dust data stored in a thermal power unit database and the operating parameters of the thermal power unit. The ESP optimization model then accurately calculates the load weight and dust removal load rate of the dust removal monitoring points, thereby facilitating the adjustment of the ESP's start-up and shutdown and load in the thermal power unit's ESP system. It also facilitates real-time monitoring of dust concentration in the thermal power unit, avoids dust accumulation during operation, reduces wear on the blades of the suction fan and booster fan, and extends the service life of the equipment. Compared with existing technologies, it offers higher dust removal accuracy and efficiency.

[0081] Currently, back-end monitoring of thermal power units is inadequate, and it is impossible to flexibly deploy monitoring points based on flue gas concentration and content. To facilitate real-time monitoring of thermal power units and improve the dust removal efficiency of electrostatic precipitators, this invention provides a method for obtaining a three-dimensional operational diagram of thermal power units based on a thermal power unit database. Figure 2 This diagram illustrates the implementation flow of the method for obtaining a 3D operational image of a thermal power unit based on a thermal power unit database. The method specifically includes:

[0082] Step S101: Establish a communication connection with the thermal power unit database and retrieve the thermal power unit facility, pipeline, and valve parameter information stored in the thermal power unit database;

[0083] It should be noted that, in this embodiment of the invention, the thermal power unit database is a pre-established database. The thermal power unit database stores information on thermal power unit facilities, pipelines, and valve parameters, and supports data access, storage, and management. The thermal power unit database can be MySQL, Oracle, or Microsoft SQL Server.

[0084] Step S102: Load the parameter information of the thermal power unit facilities, pipelines and valves, and establish a thermal power unit architecture network based on the parameter information of the thermal power unit facilities, pipelines and valves. In the thermal power unit architecture network, the facilities and valves are set as single initial monitoring points, and the pipelines are set with multiple initial monitoring points evenly distributed. The multiple initial monitoring points are integrated and connected to form the thermal power unit architecture network.

[0085] Step S103: Determine the initial weights of dust distribution in thermal power unit facilities, pipelines, and valves based on the parameter information of thermal power unit facilities, pipelines, and valves;

[0086] It should be noted that the initial weights for dust distribution in thermal power unit facilities, pipelines, and valves are determined based on the attribute characteristics of these facilities and valves. Specifically, they can be determined using expert scoring or attention mechanisms.

[0087] Step S104: Obtain the initial weights of dust distribution in thermal power unit facilities, pipelines, and valves. Determine the dust propagation level of the initial monitoring points in the thermal power unit architecture network based on the YOLO Pose algorithm and the key point detection model. In this embodiment, the dust propagation level can be 1-8.

[0088] Step S105: Determine whether the dust propagation level exceeds a preset threshold based on a preset threshold. The preset threshold can be 5.

[0089] Step S106: If the preset level threshold is exceeded, determine the current initial monitoring point as a dust removal monitoring point;

[0090] Step S107: If the preset level threshold is not exceeded, merge adjacent initial monitoring points into dust removal monitoring points based on the key point detection model.

[0091] In this embodiment, when merging adjacent initial monitoring points into dust removal monitoring points in the key point detection model, four or eight adjacent initial monitoring points can be merged into dust removal monitoring points, and dust sensors and electrostatic precipitators are deployed at the dust removal monitoring points.

[0092] Step S108: Based on the adjusted dust removal monitoring points and the thermal power unit architecture network, determine the three-dimensional diagram of thermal power unit operation.

[0093] This invention constructs a three-dimensional map of thermal power unit operation based on a thermal power unit database. Dust removal monitoring points are determined through the three-dimensional map of thermal power unit operation. At the same time, merging adjacent initial monitoring points into dust removal monitoring points based on the key point detection model can also avoid the waste of resources caused by deploying too many dust removal monitoring points.

[0094] This invention provides a method for determining the dust propagation level of initial monitoring points in a thermal power unit architecture network based on the YOLO Pose algorithm and a key point detection model. Figure 3 This diagram illustrates the implementation flow of the method for determining the dust propagation level of initial monitoring points in a thermal power unit architecture network based on the YOLO Pose algorithm and key point detection model. The method specifically includes:

[0095] Step S1041: The key point detection model captures the first propagation level based on the initial weights of the dust distribution at the initial monitoring points;

[0096] Step S1042: Determine the initial monitoring point corresponding to the first propagation level as the priority monitoring point, and determine the dust propagation level of the priority monitoring point as 5. The remaining initial monitoring points are ordinary monitoring points.

[0097] Step S1043: Calculate the dust propagation level of ordinary monitoring points based on the YOLO Pose algorithm;

[0098]

[0099] Where g is the dust propagation level of ordinary monitoring points, α is the asymmetric function constant of the YOLO Pose algorithm, and in this embodiment, the asymmetric function constant of the YOLO Pose algorithm can be 1-4, f(a) is the dust propagation coefficient of ordinary monitoring points, W(a) is the weight correction function of the YOLO Pose algorithm, n is the number of ordinary monitoring points, and a is the initial weight of dust distribution.

[0100]

[0101]

[0102] Where 'a' represents the initial weight of the dust distribution, and β represents the weight penalty factor of the weight correction function, which can be between 0.2 and 0.8.

[0103] This invention determines the dust propagation level of the initial monitoring points in the thermal power unit architecture network based on the YOLO Pose algorithm and key point detection model. This enables the electrostatic precipitator system of the thermal power unit to quickly respond to dust removal commands based on the dust propagation level. At the same time, the dust propagation level can also provide auxiliary decision support for the electrostatic precipitator optimization model to calculate the load weight of the dust removal monitoring points and the dust removal load rate.

[0104] This invention provides a method for constructing an electrostatic precipitator optimization model based on dust data stored in a thermal power unit database and the operating parameters of the thermal power units. Figure 4 The diagram illustrates the implementation flow of the method for constructing an electrostatic precipitator optimization model based on dust data stored in a thermal power unit database and the operating parameters of the thermal power units. The method specifically includes:

[0105] Step S201: Traverse the thermal power unit database and retrieve the dust data and thermal power unit operating parameters stored in the thermal power unit database.

[0106] Step S202: Perform dimensionality reduction preprocessing on the dust data and thermal power unit operating parameters, and output a normalized dimensionality reduction feature vector;

[0107] This invention embodiment preprocesses dust data and thermal power unit operating parameters by performing dimensionality reduction, and outputs a normalized dimensionality reduction feature vector that can fuse multi-source heterogeneous dust data and thermal power unit operating parameters, thereby enabling the constructed electrostatic precipitator optimization model to optimize the energy saving of the electrostatic precipitator system of thermal power units.

[0108] Step S203: Load the dimensionality reduction feature vector and divide the dimensionality reduction feature vector into a training set and a test set, wherein the sample ratio of the training set and the test set is 3:1;

[0109] Step S204: Based on the combination of random forest model and debugger tool, obtain hybrid forest model, and use hybrid forest model as the initial model of electrostatic precipitator optimization model;

[0110] Step S205: Train the model based on batch sampling softmax loss, set the number of training iterations to N, and iterate the initial model until convergence.

[0111] Step S206: Load the test set and detect the optimization accuracy of the initial model based on the test set;

[0112] Step S207: Determine whether the optimized accuracy is greater than the preset accuracy threshold.

[0113] In step S208, if the accuracy is greater than the accuracy threshold, output the electrostatic precipitator optimization model; if the accuracy is less than the accuracy threshold, return to step S203.

[0114] In this embodiment of the invention, the electrostatic precipitator optimization model is trained based on batch sampling softmax loss. The training of the electrostatic precipitator optimization model is based on a combination of random forest model and debugger tool, so that the trained and constructed electrostatic precipitator optimization model can accurately evaluate and calculate the load weight of dust collection monitoring points and the dust collection load rate.

[0115] This invention provides a method for executing a pre-built electrostatic precipitator optimization model, which calculates the load weights of dust collection monitoring points and the dust collection load rate. Figure 5 The diagram illustrates the implementation flow of the pre-built electrostatic precipitator optimization model, including the method for calculating the load weights and dust load rates of dust collection monitoring points. The specific methods for executing the pre-built electrostatic precipitator optimization model and calculating the load weights and dust load rates of dust collection monitoring points include:

[0116] Step S301: The dust data collected by the dust monitoring point and the operating parameters of the thermal power unit are used by the electrostatic precipitator optimization model to perform dimensionality reduction preprocessing on the dust data and the operating parameters of the thermal power unit, and output a normalized dimensionality reduction feature vector.

[0117] Among them, when the electrostatic precipitator optimization model performs dimensionality reduction preprocessing on dust data and thermal power unit operating parameters, the dimensionality reduction of the data is achieved through the autoencoder of the electrostatic precipitator optimization model;

[0118]

[0119] Where, d τ For the dimensionality reduction features of the data, This indicates dimensionality reduction processing of the data. dm Describes the dimensionality reduction target parameters. x m The input data consists of dust data and thermal power unit operating parameters, where m is the dimensionality of the dust data and thermal power unit operating parameters. In this embodiment, m can be 2-6.

[0120]

[0121] Where m is the dimensionality of the dust data and the operating parameters of the thermal power unit, and k is the data type of the dust data and the operating parameters of the thermal power unit. In this embodiment, k can be 2-15.

[0122] Step S302: Load the normalized dimensionality-reduced feature vector and calculate the load weight of the dust monitoring points;

[0123]

[0124] Where h(x) is the load weight value of the dust monitoring point, D is the total number of normalized dimensionality-reduced feature vectors of the input, and d τ Input the dimensionality-reduced feature vector values;

[0125] Step S303: Using the load weight of the dust removal monitoring point as input, execute the electrostatic precipitator optimization model and output the dust removal load rate of the dust removal monitoring point;

[0126]

[0127] Where h(d) τ ) represents the dust removal load rate of the dust removal monitoring point, β represents the scale of the correction function in the electrostatic precipitator optimization model, h(x) represents the load weight value of the dust removal monitoring point, and D represents the total number of normalized dimensionality-reduced feature vectors of the input.

[0128] This invention provides an energy-saving optimization system for electrostatic precipitators in thermal power units. Figure 6 The diagram shows the structure of the energy-saving optimization system for the electrostatic precipitator of the thermal power unit. The energy-saving optimization system for the electrostatic precipitator of the thermal power unit specifically includes:

[0129] The monitoring point deployment module 100 obtains a three-dimensional map of the operation of the thermal power unit based on the thermal power unit database, and deploys dust removal monitoring points through the three-dimensional map of the thermal power unit operation.

[0130] The optimization model construction module 200 is used to acquire dust data collected from dust monitoring points, upload the dust data and thermal power unit operating parameters to the thermal power unit database based on the dust data and thermal power unit operating parameters stored in the thermal power unit database, and construct an electrostatic precipitator optimization model based on the dust data and thermal power unit operating parameters stored in the thermal power unit database.

[0131] The load rate calculation module 300 takes dust data collected from dust monitoring points and thermal power unit operating parameters as input, executes a pre-built electrostatic precipitator optimization model, calculates the load weight and dust load rate of the dust monitoring points, and generates dust control instructions based on the load weight and dust load rate of the dust monitoring points.

[0132] The optimization execution module 400 is used to acquire dust removal control commands. The electrostatic precipitator system of the thermal power unit controls the opening and closing of dust removal monitoring points based on the dust removal control commands, and adjusts the dust removal power of the dust removal monitoring points.

[0133] This invention provides a load factor calculation module 300. Figure 7A schematic diagram of the load factor calculation module 300 is shown. The load factor calculation module 300 specifically includes:

[0134] The normalized dimensionality reduction unit 310 is based on the dust data collected from the dust monitoring point and the operating parameters of the thermal power unit. The electrostatic precipitator optimization model performs dimensionality reduction preprocessing on the dust data and the operating parameters of the thermal power unit, and outputs a normalized dimensionality reduction feature vector.

[0135] The load weight determination unit 320 is used to load a normalized dimensionality reduction feature vector and calculate the load weight of the dust monitoring point.

[0136] The load rate determination unit 330 takes the load weight of the dust removal monitoring point as input, executes the electrostatic precipitator optimization model, and outputs the dust removal load rate of the dust removal monitoring point.

[0137] On the other hand, embodiments of the present invention also provide a computer device, the computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, implements the method of any of the above embodiments.

[0138] In another aspect, this invention provides a computer-readable storage medium storing computer program instructions that can be executed by a processor. When executed, the computer program instructions implement the method of any of the above embodiments.

[0139] Memory, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the energy-saving optimization method for the electrostatic precipitator system of thermal power units in this application embodiment. The memory may include a program storage area and a data storage area, wherein the program storage area may store the operating system and application programs required for at least one function; the data storage area may store data created by using the energy-saving optimization method for the electrostatic precipitator system of thermal power units, etc. Furthermore, the memory may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the local module via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0140] Finally, it should be noted that the computer-readable storage medium (e.g., memory) described herein can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. By way of example, and not limitation, non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which can act as external cache memory. By way of example, and not limitation, RAM can be obtained in various forms, such as synchronous RAM (DRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The storage devices disclosed herein are intended to include, but are not limited to, these and other suitable types of memory.

[0141] The various exemplary logic blocks, modules, and circuits described herein can be implemented or performed using the following components designed to perform the functions herein: general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination of these components. A general-purpose processor may be a microprocessor, but alternatively, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors combined with a DSP, and / or any other such configuration.

[0142] In summary, this invention provides an energy-saving optimization method and system for electrostatic precipitators (ESPs) in thermal power units. The embodiments of this invention construct an ESP optimization model based on dust data stored in the thermal power unit database and the unit's operating parameters. This model accurately calculates the load weights and dust load rates at dust monitoring points, facilitating adjustments to the ESP's on / off state and load. It also facilitates real-time monitoring of dust concentration in thermal power units, prevents dust accumulation during operation, reduces wear on the blades of the suction fan and booster fan, and extends equipment lifespan. Compared to existing technologies, this method offers higher dust removal accuracy and efficiency.

[0143] It should be noted that, for the sake of simplicity, the foregoing embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to the present invention. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0144] It should be understood that the disclosed apparatus can be implemented in other ways, given the several embodiments provided in this application. For example, the apparatus embodiments described above are merely illustrative; the division of units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or communication connections shown or discussed may be through some interfaces; the indirect coupling or communication connections between devices or units may be telecommunications or other forms.

[0145] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0146] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on these embodiments, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can still combine, add, delete, or otherwise adjust the features of the various embodiments of the present invention according to the circumstances without conflict or creative effort, thereby obtaining different technical solutions that do not fundamentally depart from the concept of the present invention. These technical solutions also fall within the scope of protection of the present invention.

Claims

1. An energy-saving optimization method for electrostatic precipitator systems in thermal power units, characterized in that, The energy-saving optimization method for the electrostatic precipitator system of thermal power units includes: A 3D map of thermal power unit operation is obtained from the thermal power unit database, and dust monitoring points are deployed using the 3D map of thermal power unit operation. Obtain dust data collected from dust removal monitoring points, upload the dust data and thermal power unit operating parameters to the thermal power unit database, and construct an electrostatic precipitator optimization model based on the dust data and thermal power unit operating parameters stored in the thermal power unit database. Using dust data collected from dust monitoring points and operating parameters of thermal power units as input, a pre-built electrostatic precipitator optimization model is executed. The electrostatic precipitator optimization model calculates the load weight and dust load rate of the dust monitoring points, and generates dust control commands based on the load weight and dust load rate of the dust monitoring points. The method for obtaining a three-dimensional diagram of thermal power unit operation based on a thermal power unit database specifically includes: Establish a communication connection with the thermal power unit database and retrieve the thermal power unit facility, pipeline, and valve parameter information stored in the thermal power unit database; Load the parameter information of the thermal power unit facilities, pipelines and valves, and establish a thermal power unit architecture network based on the parameter information of the thermal power unit facilities, pipelines and valves. In the thermal power unit architecture network, the facilities and valves are set as single initial monitoring points, and the pipelines are set with multiple initial monitoring points evenly distributed. The multiple initial monitoring points are integrated and connected to form the thermal power unit architecture network. The initial weights of dust distribution in thermal power unit facilities, pipelines, and valves are determined based on parameter information of thermal power unit facilities, pipelines, and valves. The initial weights of dust distribution in thermal power unit facilities, pipelines and valves are obtained, and the dust propagation level of the initial monitoring points in the thermal power unit architecture network is determined based on the YOLO Pose algorithm and the key point detection model. Based on the preset level threshold, determine whether the dust transmission level exceeds the preset level threshold. If it exceeds the preset level threshold, determine the current initial monitoring point as the dust removal monitoring point. If the preset level threshold is not exceeded, adjacent initial monitoring points are merged into dust removal monitoring points based on the key point detection model; Based on the adjusted dust monitoring points and thermal power unit architecture network, a three-dimensional diagram of thermal power unit operation was determined.

2. The energy-saving optimization method for electrostatic precipitator systems of thermal power units as described in claim 1, characterized in that: The energy-saving optimization method for the electrostatic precipitator system of thermal power units also includes: Upon receiving dust removal control commands, the electrostatic precipitator system of the thermal power unit controls the opening and closing of dust removal monitoring points and adjusts the dust removal power of the monitoring points based on the dust removal control commands.

3. The energy-saving optimization method for electrostatic precipitator systems of thermal power units as described in claim 1, characterized in that: The method for determining the dust propagation level of initial monitoring points in a thermal power unit architecture network based on the YOLOPose algorithm and key point detection model specifically includes: The key point detection model captures the first propagation level based on the initial weights of the dust distribution at the initial monitoring points; The initial monitoring point corresponding to the first propagation level was determined as the priority monitoring point, and the dust propagation level of the priority monitoring point was determined to be 5. The remaining initial monitoring points were designated as ordinary monitoring points. Calculate the dust propagation level at ordinary monitoring points based on the YOLO Pose algorithm; （1） in, The dust transmission level is for ordinary monitoring points. For the asymmetric function constant of the YOLO Pose algorithm, The dust propagation coefficient is the standard dust propagation coefficient at a typical monitoring point. This is the weight correction function for the YOLO Pose algorithm, and The number of ordinary monitoring points, Initial weights for dust distribution; （2） （3） in, The initial weights for dust distribution are denoted as , and . is the weight penalty factor for the weight correction function.

4. The energy-saving optimization method for electrostatic precipitator systems of thermal power units as described in claim 3, characterized in that: The method for constructing an electrostatic precipitator optimization model based on dust data stored in a thermal power unit database and thermal power unit operating parameters specifically includes: Traverse the thermal power unit database and retrieve the dust data and operating parameters of the thermal power units stored in the database. Dust data and thermal power unit operating parameters are preprocessed for dimensionality reduction, and normalized dimensionality-reduced feature vectors are output. Load the dimensionality-reduced feature vectors and divide them into training and test sets, with a sample ratio of 3:1 between the training and test sets; A hybrid forest model was obtained by combining a random forest model with a debugger tool, and this hybrid forest model was used as the initial model for the electrostatic precipitator optimization model. The model is trained based on batch sampling softmax loss, with N training iterations performed on the initial model until convergence. Load the test set, detect the optimization accuracy of the initial model based on the test set, determine whether the optimization accuracy is greater than the preset accuracy threshold, and if it is greater than the accuracy threshold, output the electrostatic precipitator optimization model.

5. The energy-saving optimization method for electrostatic precipitator systems of thermal power units as described in claim 4, characterized in that: The method for executing the pre-built electrostatic precipitator optimization model and calculating the load weights and dust load rates of dust collection monitoring points specifically includes: The dust data collected from the dust monitoring points and the operating parameters of the thermal power units are preprocessed by the electrostatic precipitator optimization model to reduce the dimensionality of the dust data and the operating parameters of the thermal power units, and the normalized dimensionality-reduced feature vector is output. Among them, when the electrostatic precipitator optimization model performs dimensionality reduction preprocessing on dust data and thermal power unit operating parameters, the dimensionality reduction of the data is achieved through the autoencoder of the electrostatic precipitator optimization model; （4） in, For the dimensionality reduction features of the data, This indicates dimensionality reduction processing of the data. Describes the dimensionality reduction target parameters. The input includes dust data and operating parameters of the thermal power unit. The number of dimensionality reductions for dust data and thermal power unit operating parameters; （5） in, This refers to the dimensionality reduction of dust data and thermal power unit operating parameters. This includes data types related to dust and operating parameters of thermal power units.

6. The energy-saving optimization method for electrostatic precipitator systems of thermal power units as described in claim 5, characterized in that: The method for executing the pre-built electrostatic precipitator optimization model and calculating the load weights and dust load rates of dust collection monitoring points specifically includes: Load the normalized dimensionality-reduced feature vector and calculate the load weight of the dust monitoring points; （6） in, This represents the load weight value of the dust removal monitoring points, while The total number of normalized, dimensionality-reduced feature vectors of the input is given, while Input the dimensionality-reduced feature vector values; Using the load weight of the dust removal monitoring points as input, the electrostatic precipitator optimization model is executed, and the dust removal load rate of the dust removal monitoring points is output. （7） in, The dust removal load rate at the dust removal monitoring point, and The scale of the correction function in the electrostatic precipitator optimization model. This represents the load weight value of the dust removal monitoring points, while The total number of normalized, dimensionality-reduced feature vectors input.

7. An energy-saving optimization system for an electrostatic precipitator system of a thermal power unit, implemented using the energy-saving optimization method for an electrostatic precipitator system of a thermal power unit as described in any one of claims 1-6, characterized in that: The energy-saving optimization system for the electrostatic precipitator system of the thermal power unit includes: The monitoring point deployment module obtains a 3D map of thermal power unit operation based on the thermal power unit database and deploys dust removal monitoring points using the 3D map of thermal power unit operation. An optimization model building module is used to acquire dust data collected from dust monitoring points, upload the dust data and thermal power unit operating parameters to the thermal power unit database, and build an electrostatic precipitator optimization model based on the dust data and thermal power unit operating parameters stored in the thermal power unit database. The load rate calculation module takes dust data collected from dust monitoring points and operating parameters of thermal power units as input, executes a pre-built electrostatic precipitator optimization model, calculates the load weight and dust load rate of the dust monitoring points, and generates dust control instructions based on the load weight and dust load rate of the dust monitoring points. The optimization execution module is used to acquire dust removal control commands. The electrostatic precipitator system of the thermal power unit controls the opening and closing of dust removal monitoring points based on the dust removal control commands, and adjusts the dust removal power of the dust removal monitoring points.

8. The energy-saving optimization system for electrostatic precipitator systems of thermal power units as described in claim 7, characterized in that: The load factor calculation module specifically includes: The normalized dimensionality reduction unit is based on the dust data collected from the dust monitoring points and the operating parameters of the thermal power unit. The electrostatic precipitator optimization model performs dimensionality reduction preprocessing on the dust data and the operating parameters of the thermal power unit, and outputs a normalized dimensionality reduction feature vector. A load weight determination unit is used to load a normalized dimensionality-reduced feature vector and calculate the load weight of the dust monitoring point. The load rate determination unit takes the load weight of the dust removal monitoring point as input, executes the electrostatic precipitator optimization model, and outputs the dust removal load rate of the dust removal monitoring point.

Images