Method, device and equipment for optimizing fuel blending of thermal power unit and storage medium

Abstract

The present application belongs to the technical field of coal-fired optimization control, and discloses a thermal power unit fuel blending optimization method, device, equipment and storage medium. The method comprises the following steps: acquiring real-time operation data of a thermal power unit under different operation conditions, and constructing a degree electricity cost prediction model based on the real-time operation data; acquiring historical operation data, analyzing the historical operation data, and constructing a dynamic constraint condition according to the historical operation data; constructing a resource consumption prediction model by using a least squares support vector machine; and adjusting a damper opening degree parameter under the dynamic constraint condition based on the degree electricity cost prediction model and the resource consumption prediction model and by using a particle swarm optimization algorithm. The above-mentioned mode considers two targets of cost and resource loss, and adjusts the damper opening degree parameter of the thermal power unit by constructing respective corresponding models, thereby improving the optimization effect.

Description

Technical Field

[0001] This invention relates to the field of coal combustion optimization control technology, and in particular to a method, apparatus, equipment and storage medium for optimizing fuel blending in thermal power units. Background Technology

[0002] With the continuous increase in the proportion of renewable energy power generation, thermal power units are gradually transforming from "base load power sources" to "flexible regulation power sources," requiring frequent participation in flexible operation tasks such as grid peak shaving and frequency regulation (e.g., AGC regulation and deep peak shaving). During this process, the dynamic changes in unit operating conditions (load, damper opening, etc.) lead to significant fluctuations in the consumption of resources such as coal and liquid ammonia, as well as the cost per kilowatt-hour.

[0003] Traditional optimization methods for thermal power units typically treat environmental and economic indicators as independent multi-objective optimization problems. However, these two indicators often conflict in actual operation, leading to inaccurate optimization results.

[0004] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this invention is to provide a method, apparatus, equipment, and storage medium for optimizing fuel blending in thermal power units, aiming to solve the technical problem of inaccurate optimization results from traditional thermal power unit optimization methods.

[0006] To achieve the above objectives, the present invention provides a method for optimizing fuel blending in thermal power units, the method comprising the following steps: Acquire real-time operating data of thermal power units under different operating conditions, and build a cost-per-kilowatt-hour prediction model based on the real-time operating data; Acquire historical operational data, analyze the historical operational data, and construct dynamic constraints based on the historical operational data; A resource consumption prediction model was constructed using least squares support vector machine. The input variables of the consumption prediction model are unit load, coal feed rate and damper opening, and the output variables are coal consumption rate and liquid ammonia feed flow rate. Based on the cost-per-kilowatt-hour prediction model and resource consumption prediction model, and using the particle swarm optimization algorithm, the damper opening parameter is adjusted under dynamic constraints. The damper opening parameter is used to optimize the fuel blending of thermal power units.

[0007] In some embodiments, a cost-per-kilowatt-hour prediction model is constructed based on real-time operational data, including: Variable costs and electricity consumption are determined based on real-time operational data. The cost per kilowatt-hour is determined based on the variable costs and the amount of electricity connected to the grid. A cost-per-kilowatt-hour prediction model is constructed based on the cost per kilowatt-hour, and the model is optimized with NOx emissions and coal consumption rate as objectives.

[0008] In some embodiments, determining variable costs based on real-time operational data includes: Determine coal costs based on real-time coal consumption rate and real-time load. Determine industrial water costs based on real-time boiler feedwater flow rate; Determine the cost of liquid ammonia based on real-time liquid ammonia feed flow rate; Determine the cost of limestone slurry based on real-time limestone slurry feed flow rate; Variable costs are determined based on the costs of coal, industrial water, liquid ammonia, and limestone slurry.

[0009] In some embodiments, analyzing historical runtime data and constructing dynamic constraints based on the historical runtime data includes: Extract historical damper opening data from historical operational data; Determine the change in damper opening over a certain number of units of time based on historical damper opening data; The maximum change in damper opening is determined based on the change in damper opening over several units of time. Dynamic constraints are constructed based on the maximum change in damper opening.

[0010] In some embodiments, a resource consumption prediction model is constructed using a least-squares support vector machine, including: The input variables and target output variables are normalized to eliminate the influence of dimensional differences on the model. The input variables include unit load, coal feed rate and damper opening, and the target output variables include coal consumption rate and liquid ammonia flow rate. The input variables are used as the training dataset to train the preset model, and the current output variables are obtained. Based on the current output variable and the target output variable, the key parameters of the model are optimized through cross-validation until the error between the current output variable and the target output variable of the preset model meets the preset conditions. The key parameters include the regularization parameter and the kernel function width.

[0011] In some embodiments, adjusting the damper opening parameter based on a per-kilowatt-hour cost prediction model and a resource consumption prediction model, and using a particle swarm optimization algorithm under the dynamic constraints, includes: The optimization objective is to minimize the cost per kilowatt-hour. The objective function is solved by using the particle swarm optimization algorithm and combining it with dynamic constraints to obtain the change in damper opening corresponding to minimizing the cost per kilowatt-hour. The damper opening is adjusted based on the change in damper opening, which includes the change in secondary damper opening and the change in burnout damper opening.

[0012] Furthermore, to achieve the above objectives, the present invention also proposes a fuel blending optimization device for thermal power units, which includes: The prediction model building module is used to acquire real-time operating data of thermal power units under different operating conditions and build a cost-per-kilowatt-hour prediction model based on the real-time operating data. The constraint construction module is used to acquire historical running data, analyze historical running data, and construct dynamic constraints based on historical running data. The consumption model construction module is used to build a resource consumption prediction model using least squares support vector machine. The input variables of the consumption prediction model are unit load, coal feed rate and damper opening, and the output variables are coal consumption rate and liquid ammonia feed flow rate. The damper parameter optimization module is used to adjust the damper opening parameters under dynamic constraints based on the cost-per-kilowatt-hour prediction model and resource consumption prediction model and using the particle swarm optimization algorithm. The damper opening parameters are used to optimize the fuel blending of thermal power units.

[0013] In some embodiments, the predictive model building module is used to determine variable costs and electricity consumption based on real-time operating data; The cost per kilowatt-hour is determined based on variable costs and the amount of electricity fed into the grid. A cost-per-kilowatt-hour prediction model is constructed based on the cost per kilowatt-hour, and the model is optimized with NOx emissions and coal consumption rate as objectives.

[0014] Furthermore, to achieve the above objectives, the present invention also proposes a fuel blending optimization device for thermal power units. The fuel blending optimization device for thermal power units includes: a memory, a processor, and a fuel blending optimization program for thermal power units stored in the memory and executable on the processor. The fuel blending optimization program for thermal power units is configured to implement the steps of the fuel blending optimization method for thermal power units as described above.

[0015] In addition, to achieve the above objectives, the present invention also proposes a storage medium storing a fuel blending optimization program for thermal power units, wherein when the fuel blending optimization program for thermal power units is executed by a processor, the steps of the fuel blending optimization method for thermal power units as described above are implemented.

[0016] This invention acquires real-time operating data of thermal power units under different operating conditions and constructs a cost-per-kilowatt-hour prediction model based on the real-time operating data; acquires historical operating data, analyzes the historical operating data, and constructs dynamic constraints based on the historical operating data; constructs a resource consumption prediction model using least squares support vector machine; and adjusts the damper opening parameters under dynamic constraints based on the cost-per-kilowatt-hour prediction model and the resource consumption prediction model using particle swarm optimization algorithm. The above methods consider two objectives: cost and resource loss, and improve the optimization effect by constructing corresponding models to adjust the damper opening parameters of thermal power units. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the first embodiment of the fuel blending optimization method for thermal power units according to the present invention. Figure 2 This is a structural block diagram of the first embodiment of the fuel blending optimization device for thermal power units of the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0020] This invention provides a method for optimizing fuel blending in thermal power units, referring to... Figure 1 , Figure 1 This is a schematic flowchart of the first embodiment of the method for optimizing fuel blending in thermal power units according to the present invention.

[0021] In this embodiment, the fuel blending optimization method for thermal power units includes the following steps: Step S10: Obtain real-time operating data of thermal power units under different operating conditions, and construct a cost-per-kilowatt-hour prediction model based on the real-time operating data.

[0022] In this embodiment, the executing entity is a fuel blending optimization device for thermal power units. This fuel blending optimization device has functions such as data processing, data communication, and program execution. The fuel blending optimization device for thermal power units can be a computer terminal device or other network device, or other devices with similar functions. This embodiment does not limit the scope of the device.

[0023] It should be noted that, with the continuous increase in the proportion of renewable energy power generation, thermal power units are gradually transforming from "base load power sources" to "flexible regulation power sources," requiring frequent participation in flexible operation tasks such as grid peak shaving and frequency regulation (e.g., AGC regulation and deep peak shaving). During this process, the dynamic changes in unit operating conditions (load, damper opening, etc.) lead to significant fluctuations in resource consumption such as coal and liquid ammonia, as well as the cost per kilowatt-hour. Traditional optimization methods for thermal power units typically treat environmental and economic indicators as independent multi-objective optimization problems; however, these two often conflict in actual operation, resulting in inaccurate optimization results.

[0024] To address the aforementioned technical issues, this embodiment acquires real-time operating data of thermal power units under different operating conditions and constructs a cost-per-kilowatt-hour prediction model based on the real-time operating data; acquires historical operating data, analyzes the historical operating data, and constructs dynamic constraints based on the historical operating data; constructs a resource consumption prediction model using a least-squares support vector machine; and adjusts the damper opening parameters under the dynamic constraints based on the cost-per-kilowatt-hour prediction model and the resource consumption prediction model. This approach considers both cost and resource consumption objectives, and by constructing corresponding models to adjust the damper opening parameters of the thermal power units, it improves the optimization effect. Specifically, it can be implemented as follows.

[0025] In this specific implementation, the cost per kilowatt-hour prediction model needs to be constructed first. The construction of the cost per kilowatt-hour prediction model is based on the real-time operating data of the thermal power unit under different operating conditions. The real-time operating data includes real-time coal consumption rate, real-time load, real-time boiler feedwater flow rate, real-time liquid ammonia, and real-time limestone slurry feed flow rate.

[0026] In some embodiments, variable costs and on-grid electricity can be obtained based on the aforementioned real-time operating data. Specifically, variable costs consist of coal costs, industrial water costs, liquid ammonia costs, and limestone slurry costs. Coal costs are determined by real-time coal consumption rate and real-time load, and the calculation formula is as follows:

[0027] Where b(t) and P(t) are the real-time coal consumption rate (unit: g / (kW·h)) and the real-time load (unit: MW), respectively, and P coal The price is the standard coal price (yuan / ton).

[0028] The cost of industrial water is determined by the real-time boiler feedwater flow rate, and the calculation formula is as follows:

[0029] Where L(t) is the real-time boiler feedwater flow rate (unit: tons / hour), P water Price of industrial water (RMB / ton).

[0030] The cost of liquid ammonia is determined by the real-time liquid ammonia feed flow rate, and the calculation formula is as follows:

[0031] Where A(t) is the real-time liquid ammonia feed flow rate (unit: kg / hour), P amm Price of liquid ammonia (RMB / ton).

[0032] The cost of limestone slurry is determined by the real-time limestone slurry feed flow rate, and the calculation formula is as follows:

[0033] Where S(t) is the real-time limestone slurry feed flow rate (unit: tons / hour), P ls The price of limestone slurry is (yuan / ton). In the above calculation formulas, t... beg With t end These represent the start time and end time, respectively. The sum of all the above costs constitutes the variable cost.

[0034] Step S20: Obtain historical running data, analyze the historical running data, and construct dynamic constraints based on the historical running data.

[0035] It should be understood that in this embodiment, when performing optimization, it is necessary to first construct corresponding dynamic constraints. The construction of dynamic constraints in this embodiment relies on historical operating data. Specifically, the construction process involves extracting historical damper opening data from the historical operating data; determining the damper opening change over several unit time periods based on the historical damper opening data; determining the maximum damper opening change based on the damper opening change over several unit time periods; and constructing dynamic constraints based on the maximum damper opening change.

[0036] It should be noted that the core objects of dynamic constraints are the secondary damper opening and the burnout damper opening. The extracted historical damper opening data includes historical secondary damper openings and historical burnout damper openings. The unit time can be set to 5 seconds or 1 minute, etc., and then sampling and calculation are performed according to different unit time points. The change in damper opening per unit time can be obtained by calculating the difference between different adjacent time points. The specific calculation formula is: Δu j (t)=u j (t+Δt) u j (t), where t represents the time point, u j (t+Δt) and u j(t) represents the damper opening corresponding to t+Δt and t respectively. The change in damper opening can be calculated by the difference. j represents the j-th type of damper. In this embodiment, it refers to the opening of the secondary damper and the opening of the burnout damper. The change in the opening of the above two types of dampers can be calculated based on this formula. For all Δu j Statistical analysis was performed on the absolute value of (t) to obtain the maximum change in damper opening Δu. jmax During the optimization process, the constraint condition constructed based on the maximum damper opening is |u j (t+1)-u j (t)|≤Δu jmax u j (t) represents the current damper opening, u j (t+1) represents the optimized damper opening.

[0037] Step S30: Construct a resource consumption prediction model using least squares support vector machine.

[0038] In its implementation, the core objective of the constructed resource consumption prediction model is to predict resource consumption indicators, namely the coal consumption rate and liquid ammonia feed flow rate of thermal power units. The model employs the Least Squares Support Vector Machine (LSSVM) algorithm, with a Gaussian radial basis function (RBF) kernel. The specific training process involves normalizing the input and target output variables to eliminate the influence of dimensional differences on the model. The input variables include unit load, coal feed rate, and damper opening. The target output variables include the coal consumption rate and liquid ammonia flow rate. The input variables are then used as the training dataset to train the preset model, yielding the current output variable. Based on the current and target output variables, the key parameters of the model are optimized through cross-validation until the error between the current and target output variables meets a preset condition. These key parameters include regularization parameters and the kernel width.

[0039] It should be noted that the model can be standardized or normalized to the [0,1] interval to eliminate the impact of dimensional differences on the model. The key parameters optimized during model training are the regularization parameter and the kernel density. The regularization parameter controls the trade-off between model complexity and training error; a larger regularization parameter results in a heavier penalty for training error. The kernel width affects the locality of the RBF kernel; a smaller kernel width results in a steeper kernel, making the model more sensitive to local data. The error between the current output variable and the target output variable can be either mean squared error or mean absolute error. A preset condition can be set, for example, that the coal consumption rate prediction error is ≤1g / (kW·h). When this condition is met, the model training is considered complete. This preset condition can also be adjusted according to actual needs; this embodiment does not impose any restrictions on it.

[0040] Step S40: Based on the electricity cost prediction model and the resource consumption prediction model, and using the particle swarm optimization algorithm, adjust the damper opening parameter under the dynamic constraints.

[0041] It should be noted that the resource consumption prediction model uses the coal consumption rate and liquid ammonia feed flow rate of the thermal power unit as the core optimization objectives. The damper opening directly affects the coal consumption rate and liquid ammonia feed flow rate of the thermal power unit, and the damper opening is subject to dynamic constraints. This approach effectively improves the optimization effect. Specifically, the process involves using the energy cost prediction model as the objective function, minimizing the energy cost as the optimization objective; solving the objective function using a particle swarm optimization algorithm combined with the dynamic constraints to obtain the damper opening change corresponding to minimizing the energy cost; and adjusting the current damper opening based on the damper opening change, which includes the secondary damper opening change and the burnout damper opening change.

[0042] It should be noted that by combining the above processes, the minimum cost per kilowatt-hour (kWh) minc = F / E is minimized, where F is the variable cost and E is the energy supplied to the grid. Then, the coal cost F is calculated using the above formula. coal Cost of liquid ammonia F amm Based on the resource consumption prediction model, the damper opening parameter is determined, and finally, by combining dynamic constraints and using the particle swarm optimization algorithm, the damper opening change corresponding to the minimum cost per kilowatt-hour can be obtained, thereby completing the adjustment of the secondary damper opening and the burnout damper opening.

[0043] In this embodiment, the particle swarm optimization algorithm specifically includes initializing particle swarm parameters, such as: number of particles: set according to problem complexity (e.g., 20-50 particles); particle position: each particle's position vector X = [u_secondary damper opening, u_burnout damper opening], with initial positions randomly distributed within a reasonable range of damper openings (e.g., 0%~100%); particle velocity: initial velocity VV is randomly assigned (usually set to 10-20% of the position range); number of iterations: set the maximum number of iterations, such as 100; PSO parameters: inertia weight [0.4~0.9], cognitive coefficient, and social coefficient can be set to 2. The unit cost is the particle's fitness. Finally, updating the particle position and velocity and applying dynamic constraints allows for iterative optimization until convergence.

[0044] This embodiment acquires real-time operating data of thermal power units under different operating conditions and constructs a cost-per-kilowatt-hour prediction model based on the real-time operating data; acquires historical operating data, analyzes the historical operating data, and constructs dynamic constraints based on the historical operating data; constructs a resource consumption prediction model using least squares support vector machine; and adjusts the damper opening parameters under the dynamic constraints based on the cost-per-kilowatt-hour prediction model and the resource consumption prediction model. The above method considers both cost and resource loss objectives, and improves the optimization effect by constructing corresponding models to adjust the damper opening parameters of thermal power units.

[0045] Furthermore, this embodiment of the invention also proposes a storage medium storing a fuel blending optimization program for thermal power units. When the fuel blending optimization program for thermal power units is executed by a processor, it implements the steps of the fuel blending optimization method for thermal power units as described above.

[0046] Reference Figure 2 , Figure 2 This is a structural block diagram of the first embodiment of the fuel blending optimization device for thermal power units of the present invention.

[0047] like Figure 2 As shown, the fuel blending optimization device for thermal power units proposed in this embodiment of the invention includes: The prediction model building module 10 is used to acquire real-time operating data of thermal power units under different operating conditions, and to build a cost-per-kilowatt-hour prediction model based on the real-time operating data. The constraint construction module 20 is used to acquire historical running data, analyze the historical running data, and construct dynamic constraints based on the historical running data. The consumption model construction module 30 is used to construct a resource consumption prediction model using a least squares support vector machine. The input variables of the consumption prediction model are unit load, coal feed rate and damper opening, and the output variables are coal consumption rate and liquid ammonia feed flow rate. The damper parameter optimization module 40 is used to adjust the damper opening parameters based on the electricity cost prediction model and the resource consumption prediction model and using the particle swarm optimization algorithm under the dynamic constraints. The damper opening parameters are used to optimize the fuel blending of thermal power units.

[0048] In specific implementation, the prediction model construction module 10 first obtains real-time operating data from the unit's DCS system or the plant-level monitoring information system (SIS), calculates variable costs according to the four cost items described in claim 3, and calculates the cost per kilowatt-hour in conjunction with the electricity fed into the grid, constructing a cost per kilowatt-hour prediction model with NOx emissions and coal consumption rate as targets. The constraint condition construction module 20 extracts historical data on the opening of the secondary damper and the burning-out damper for a preset period (such as the last three months) from the historical operating database, calculates the damper opening change sequence at a set unit time (such as 1 minute), and determines the dynamic constraint boundary by statistically analyzing the maximum absolute value of each sequence. The consumption model construction module 30 uses the unit load, coal feed rate, and damper opening as input features of LSSVM, performs regression prediction on the coal consumption rate and liquid ammonia feed flow rate, and uses k-fold cross-validation to adaptively optimize the regularization parameters and RBF kernel function width until the prediction error meets the preset threshold. The damper parameter optimization module 40 takes minimizing the cost per kilowatt-hour as its optimization objective, calls the particle swarm optimization algorithm, and iteratively seeks optimization within the dynamic constraint boundary provided by the constraint construction module 20. It outputs the change in the secondary damper opening and the change in the burnout damper opening, and sends the optimized damper opening command to the unit damper actuator to complete the closed-loop control.

[0049] This embodiment acquires real-time operating data of thermal power units under different operating conditions and constructs a cost-per-kilowatt-hour prediction model based on the real-time operating data; acquires historical operating data, analyzes the historical operating data, and constructs dynamic constraints based on the historical operating data; constructs a resource consumption prediction model using least squares support vector machine; and adjusts the damper opening parameters under the dynamic constraints based on the cost-per-kilowatt-hour prediction model and the resource consumption prediction model. The above method considers both cost and resource loss objectives, and improves the optimization effect by constructing corresponding models to adjust the damper opening parameters of thermal power units.

[0050] In some embodiments, the prediction model building module 10 is used to determine variable costs and electricity consumption based on the real-time operating data; The cost per kilowatt-hour is determined based on the variable costs and the amount of electricity connected to the grid. A cost-per-kilowatt-hour prediction model is constructed based on the cost per kilowatt-hour, and the cost-per-kilowatt-hour prediction model is optimized with NOx emissions and coal consumption rate as objectives.

[0051] In some embodiments, the prediction model building module 10 is used to determine the cost of coal based on real-time coal consumption rate and real-time load; Determine industrial water costs based on real-time boiler feedwater flow rate; Determine the cost of liquid ammonia based on real-time liquid ammonia feed flow rate; Determine the cost of limestone slurry based on real-time limestone slurry feed flow rate; Variable costs are determined based on the cost of coal, the cost of industrial water, the cost of liquid ammonia, and the cost of limestone slurry.

[0052] In some embodiments, the constraint construction module 20 is used to extract historical damper opening data from the historical operating data; The change in damper opening over a certain number of unit time periods is determined based on the historical damper opening data. The maximum change in damper opening is determined based on the change in damper opening over several unit times. Dynamic constraints are constructed based on the maximum change in damper opening.

[0053] In some embodiments, the consumption model construction module 30 is used to normalize the input variables and the target output variables. The normalization process is used to eliminate the influence of dimensional differences on the model. The input variables include unit load, coal feed rate and damper opening. The target output variables include coal consumption rate and liquid ammonia flow rate. The input variables are used as the training dataset to train the preset model, and the current output variables are obtained. Based on the current output variable and the target output variable, the key parameters of the model are optimized through cross-validation until the error between the current output variable and the target output variable of the preset model meets the preset condition. The key parameters include regularization parameters and kernel function width.

[0054] In some embodiments, the damper parameter optimization module 40 is used to minimize the cost per kilowatt-hour as the optimization objective, with the cost per kilowatt-hour prediction model as the objective function. The objective function is solved by using the particle swarm optimization algorithm and combining the dynamic constraints to obtain the damper opening change corresponding to minimizing the cost per kilowatt-hour. The current damper opening is adjusted based on the change in damper opening, which includes the change in secondary damper opening and the change in burnout damper opening.

[0055] This application embodiment also provides a fuel blending optimization device for thermal power units, including a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other through the communication bus. The memory is used to store the fuel blending optimization program for thermal power units. When the processor executes the program stored in the memory, it implements the above-mentioned fuel blending optimization method for thermal power units.

[0056] The communication bus mentioned in the aforementioned fuel blending optimization equipment for thermal power units can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0057] The communication interface is used for communication between the aforementioned thermal power unit fuel blending optimization equipment and other equipment.

[0058] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0059] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0060] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0061] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0062] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0063] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0064] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0065] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0066] In addition, for technical details not described in detail in this embodiment, please refer to the fuel blending optimization method for thermal power units provided in any embodiment of the present invention, which will not be repeated here.

[0067] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0068] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0069] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0070] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

[0071] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

Claims

1. A method for optimizing fuel blending in thermal power units, characterized in that, The optimized fuel blending method for thermal power units includes: Real-time operating data of thermal power units under different operating conditions are obtained, and a cost-per-kilowatt-hour prediction model is constructed based on the real-time operating data. Acquire historical operating data, analyze the historical operating data, and construct dynamic constraints based on the historical operating data; A resource consumption prediction model is constructed using least squares support vector machine. The input variables of the resource consumption prediction model are unit load, coal feed rate and damper opening, and the output variables are coal consumption rate and liquid ammonia feed flow rate. Based on the electricity cost prediction model and the resource consumption prediction model, and using the particle swarm optimization algorithm to adjust the damper opening parameter under the dynamic constraints, the damper opening parameter is used to optimize the fuel blending of thermal power units.

2. The method for optimizing fuel blending in thermal power units as described in claim 1, characterized in that, The method for constructing a cost-per-kilowatt-hour prediction model based on the real-time operating data includes: Variable costs and electricity consumption are determined based on the real-time operating data. The cost per kilowatt-hour is determined based on the variable costs and the amount of electricity connected to the grid. A cost-per-kilowatt-hour prediction model is constructed based on the cost per kilowatt-hour, and the cost-per-kilowatt-hour prediction model is optimized with NOx emissions and coal consumption rate as objectives.

3. The method for optimizing fuel blending in thermal power units as described in claim 2, characterized in that, The step of determining variable costs based on the real-time operating data includes: Determine coal costs based on real-time coal consumption rate and real-time load. Determine industrial water costs based on real-time boiler feedwater flow rate; Determine the cost of liquid ammonia based on real-time liquid ammonia feed flow rate; Determine the cost of limestone slurry based on real-time limestone slurry feed flow rate; Variable costs are determined based on the cost of coal, the cost of industrial water, the cost of liquid ammonia, and the cost of limestone slurry.

4. The method for optimizing fuel blending in thermal power units as described in claim 1, characterized in that, The analysis of the historical operating data and the construction of dynamic constraints based on the historical operating data include: Extract historical damper opening data from the historical operating data; The change in damper opening over a certain number of unit time periods is determined based on the historical damper opening data. The maximum change in damper opening is determined based on the change in damper opening over several unit times. Dynamic constraints are constructed based on the maximum change in damper opening.

5. The method for optimizing fuel blending in thermal power units as described in claim 1, characterized in that, The method of constructing a resource consumption prediction model using least squares support vector machines includes: The input variables and target output variables are normalized to eliminate the influence of dimensional differences on the model. The input variables include unit load, coal feed rate and damper opening. The target output variables include coal consumption rate and liquid ammonia feed flow rate. The input variables are used as the training dataset to train the preset model, and the current output variables are obtained. Based on the current output variable and the target output variable, the key parameters of the model are optimized through cross-validation until the error between the current output variable and the target output variable of the preset model meets the preset condition. The key parameters include regularization parameters and kernel function width.

6. The method for optimizing fuel blending in thermal power units as described in claim 1, characterized in that, The step of adjusting the damper opening parameter under the dynamic constraints based on the electricity cost prediction model and the resource consumption prediction model, and using the particle swarm optimization algorithm, includes: The optimization objective is to minimize the cost per kilowatt-hour. The objective function is solved by using the particle swarm optimization algorithm and combining the dynamic constraints to obtain the damper opening change corresponding to minimizing the cost per kilowatt-hour. The current damper opening is adjusted based on the change in damper opening, which includes the change in secondary damper opening and the change in burnout damper opening.

7. A fuel blending optimization device for thermal power units, characterized in that, The thermal power unit fuel blending optimization device includes: The prediction model building module is used to acquire real-time operating data of thermal power units under different operating conditions, and to build a cost-per-kilowatt-hour prediction model based on the real-time operating data. The constraint construction module is used to acquire historical running data, analyze the historical running data, and construct dynamic constraints based on the historical running data. The consumption model construction module is used to construct a resource consumption prediction model using least squares support vector machine. The input variables of the consumption prediction model are unit load, coal feed rate and damper opening, and the output variables are coal consumption rate and liquid ammonia feed flow rate. The damper parameter optimization module is used to adjust the damper opening parameters based on the electricity cost prediction model and the resource consumption prediction model and using the particle swarm optimization algorithm under the dynamic constraints. The damper opening parameters are used to optimize the fuel blending of thermal power units.

8. The fuel blending optimization device for thermal power units as described in claim 7, characterized in that, The prediction model construction module is used to determine the variable cost and the amount of electricity used for internet access based on the real-time operating data; The cost per kilowatt-hour is determined based on the variable costs and the amount of electricity connected to the grid. A cost-per-kilowatt-hour prediction model is constructed based on the cost per kilowatt-hour, and the cost-per-kilowatt-hour prediction model is optimized with NOx emissions and coal consumption rate as objectives.

9. A fuel blending optimization device for thermal power units, characterized in that, The thermal power unit fuel blending optimization device includes: a memory, a processor, and a thermal power unit fuel blending optimization program stored in the memory and executable on the processor, wherein the thermal power unit fuel blending optimization program is configured to implement the steps of the thermal power unit fuel blending optimization method as described in any one of claims 1 to 6.

10. A storage medium, characterized in that, The storage medium stores a fuel blending optimization program for thermal power units, which, when executed by a processor, implements the steps of the fuel blending optimization method for thermal power units as described in any one of claims 1 to 6.

Images