Thermal power generating unit operation optimization method and system based on energy efficiency optimization

Abstract

The invention discloses a thermal power generating unit operation optimization method and system based on energy efficiency optimization, and the method comprises the steps: obtaining the operation data of a thermal power generating unit in real time, screening the operation data of an energy efficiency index of the thermal power generating unit under a steady-state working condition, and obtaining a data set; dividing the operation condition of the thermal power generating unit, dividing the data set under the steady-state condition into corresponding data subsets, and endowing the energy efficiency index with a corresponding weight value in each data subset according to the energy efficiency; training a weighted support vector machine model by using the data subset, and performing parameter optimization according to the weight value of the energy efficiency index in the training process; and inputting the new data subset into the trained weighted support vector machine model to obtain an optimal value of the energy efficiency index under the corresponding working condition, performing real-time evaluation on the energy efficiency index by taking the optimal value of the energy efficiency index as a reference, and adjusting operation of the thermal power generating unit according to an evaluation result. According to the method, the accurate optimal energy efficiency value of the thermal power generating unit can be obtained, and operation optimization of the thermal power generating unit is achieved accordingly.

Description

Technical Field

[0001] The invention relates to a generator set operation optimization technology, and in particular to a thermal power unit operation optimization method and system based on energy efficiency optimization. Background Art

[0002] With the construction of new power systems, new energy installed capacity continues to increase, and most thermal power units have been in peak operation for a long time, resulting in increased energy consumption and increased pressure to save energy and improve efficiency.

[0003] At present, for most thermal power units, operation optimization still depends entirely on the operating experience of personnel, and the operating personnel optimize and adjust the main parameters according to the unit operating conditions. However, the optimization effect of this method depends entirely on personal experience and cognitive level, and it is impossible to accurately measure the optimization effect. Some scholars use the operating data of thermal power units to self-optimize specific indicators that affect the energy efficiency of the unit (such as flue gas oxygen content, heater terminal difference), and guide the operating personnel to optimize and adjust by means of deviation warning of energy efficiency indicators, but this method does not consider the overall energy efficiency level of the unit. Some scholars also use the operating data of the unit to obtain the minimum value of the energy efficiency index in the cluster center through cluster analysis as the evaluation benchmark value, but if the unit is mostly in a non-optimal energy efficiency state during daily operation, the energy efficiency benchmark value obtained by this method has deviated from the actual optimal value. Summary of the invention

[0004] The technical problem to be solved by the present invention is as follows: In view of the above-mentioned problems in the prior art, a method and system for optimizing the operation of a thermal power unit based on energy efficiency optimization is provided, in which the optimal value of the energy efficiency of the thermal power unit is established through data analysis, and the operation of the thermal power unit is optimized accordingly to improve the energy efficiency level of the unit.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for optimizing the operation of a thermal power unit based on energy efficiency optimization comprises the following steps:

[0007] Acquire the operating data of thermal power units in real time, filter the operating data of the specified energy efficiency indicators of the thermal power units under steady-state conditions, and obtain the data set under steady-state conditions;

[0008] The operating conditions of the thermal power units are divided, and according to the division results, the data set under the steady-state condition is divided into corresponding data subsets, and in each data subset, the energy efficiency index is assigned a corresponding weight value according to the energy efficiency level;

[0009] Using the data subset to train a weighted support vector machine model, so that the trained weighted support vector machine model predicts the energy efficiency index under each working condition to obtain the optimal value and outputs it, and during the training process of the weighted support vector machine model, parameter optimization is performed according to the weight value of the energy efficiency index;

[0010] The new data subset is input into the trained weighted support vector machine model to obtain the optimal value of the energy efficiency index under the corresponding working conditions. The energy efficiency index is evaluated in real time based on the optimal value of the energy efficiency index, and the operation of the thermal power unit is adjusted according to the evaluation results.

[0011] Furthermore, when screening the operating data of the selected energy efficiency indicators of the thermal power unit under steady-state conditions, the following steps are included:

[0012] Specify at least one characterization parameter in the specified energy efficiency index;

[0013] Collect the running data in the current time window and obtain the maximum value of the running data in the current time window for each characterization parameter With minimum

[0014] Get the maximum value y of the running data in all time windows of each characterization parameter i,max With the minimum value y i,min ;

[0015] Determine the maximum value of each characterization parameter With minimum The difference compared to the maximum value y i,max With the minimum value y i,min Whether the differences are all less than or equal to a preset threshold;

[0016] If they are all less than or equal to the preset threshold, the operating data in the current time window is added to the data set under the steady-state condition, and the current time window is moved backward by a time span of a specified size, and the step of collecting the operating data in the current time window is performed again;

[0017] If at least one of the characterization parameters has a maximum value With minimum The difference compared to the maximum value y i,max With the minimum value y i,min If the difference is greater than a preset threshold, the current time window is moved back by one sampling time, and the step of collecting the operating data in the current time window is executed again.

[0018] Furthermore, the classification of the operating conditions of the thermal power units specifically refers to the classification of the operating conditions of the thermal power units based on the unit load, ambient temperature and coal quality as boundary conditions.

[0019] Furthermore, the energy efficiency index includes the energy efficiency index of each subsystem of the thermal power unit and the comprehensive energy efficiency index of the thermal power unit. The parameter vector of each data sample in the data subset includes the parameter vector jointly constituted by the energy efficiency indexes of all subsystems of the data sample and the comprehensive energy efficiency index of the thermal power unit. When assigning corresponding weight values ​​to the energy efficiency indexes according to the energy efficiency level in each data subset, it includes:

[0020] According to the order of comprehensive energy efficiency indicators from best to worst, the corresponding label value is calculated by linear interpolation algorithm, and then the parameter vector of each data sample in the data subset is assigned the corresponding label value as the weight value.

[0021] Furthermore, the weighted support vector machine model takes the parameter vector composed of the energy efficiency indicators of all subsystems as input and takes the comprehensive energy efficiency indicator as output, and the expression is as follows:

[0022]

[0023] in, is a nonlinear mapping function, ω is a weight vector, b is a bias, ξ i , is a slack variable; ε is an insensitive loss function parameter, C is a penalty coefficient; s i ,t i Respectively represent the weighting coefficients of the i-th data sample in the data subset for the parameters C and ε, 0≤s i ,t i ≤1; where s i is the weight value of the i-th data sample; N l Indicates the number of data samples contained in the data subset, x i Represents the parameter vector composed of the energy efficiency indicators of all subsystems in the i-th data sample in the data subset, y i Represents the comprehensive energy efficiency index of the i-th data sample in the data subset.

[0024] Furthermore, when the energy efficiency index is evaluated in real time based on the optimal value of the energy efficiency index, it includes:

[0025] Normalize the current operating data of the energy efficiency index of the specified working condition to obtain the current value of the energy efficiency index of the specified working condition;

[0026] The deviation and the deviation change rate between the current value and the optimal value of the energy efficiency index of the specified working condition are calculated, and the real-time quantitative evaluation value of the energy efficiency index of the specified working condition is calculated according to the deviation and the deviation change rate.

[0027] Furthermore, the real-time quantitative evaluation value expression is as follows:

[0028]

[0029] Among them, E i (k) is the deviation between the current value and the optimal value of the i-th energy efficiency indicator, is the deviation change rate between the current value and the optimal value of the i-th energy efficiency indicator.

[0030] Furthermore, when adjusting the operation of the thermal power unit according to the evaluation results, it includes: if the real-time quantitative evaluation value of the energy efficiency index of the specified operating condition is less than the specified value, adjusting the operating state of the thermal power unit according to the deviation between the optimal value and the current value of the energy efficiency index of the specified operating condition.

[0031] The present invention also proposes a thermal power unit operation optimization system based on energy efficiency optimization, comprising a microprocessor and a computer-readable storage medium connected to each other, wherein the microprocessor is programmed or configured to execute any one of the thermal power unit operation optimization methods based on energy efficiency optimization.

[0032] The present invention also proposes a computer-readable storage medium, in which a computer program is stored. The computer program is used to be programmed or configured by a microprocessor to execute any one of the methods for optimizing the operation of a thermal power unit based on energy efficiency optimization.

[0033] Compared with the prior art, the advantages of the present invention are:

[0034] The present invention performs "labeling" processing on the operating data, that is, assigning a higher weight to data with high energy efficiency and a lower weight to data with low energy efficiency, and trains a weighted support vector machine model through the processed operating data, so that the trained model can reflect the optimal energy efficiency of the unit.

[0035] The present invention takes the optimal value of the energy efficiency index output by the trained weighted support vector machine model as a benchmark, and makes a comprehensive judgment on the deviation between the normalized energy efficiency index and its optimal value and the deviation change rate, thereby realizing real-time quantitative evaluation of the energy efficiency index. The evaluation method is simple, the calculation workload is small, and it is convenient for engineering application. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The figure is a flow chart of a method according to an embodiment of the present invention.

[0037] Figure 2 Schematic diagram of the division of the normalized i-th energy efficiency index deviation and deviation change rate in an embodiment of the present invention.

[0038] Figure 3 It is the main steam pressure variation trend of the thermal power unit under a certain working condition and the energy efficiency index evaluation result in the embodiment of the present invention. DETAILED DESCRIPTION

[0039] The present invention is further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.

[0040] Embodiment 1

[0041] This embodiment proposes a method for optimizing the operation of a thermal power unit based on energy efficiency optimization. The method uses the operating data of the thermal power unit to obtain the optimal value of the energy efficiency index under various operating conditions of the unit through data analysis and nonlinear modeling methods. During the operation of the unit, the current energy efficiency index is compared with its optimal value, and the energy efficiency index of the thermal power unit is evaluated in real time according to the deviation and its deviation change rate, thereby providing guidance for the optimized operation of the thermal power unit.

[0042] like Figure 1 As shown, the method of this embodiment includes the following steps:

[0043] S101, selecting energy efficiency indicators of thermal power units, thereby determining an energy efficiency indicator set of thermal power units; obtaining operation data of thermal power units in real time, and screening operation data of selected energy efficiency indicators of thermal power units under steady-state conditions, thereby obtaining a data set Ω under steady-state conditions;

[0044] S102, dividing the operating conditions of the thermal power unit, and according to the division results, dividing the data set Ω under the steady-state condition into corresponding data subsets Ω l , assuming that it can be divided into m working conditions, for the data set Ω obtained under the steady-state working condition, it is divided into m data subsets Ω l (l=1,2,…,m); perform “labeling” processing on the data samples of each data subset. “Labeling” processing specifically refers to assigning corresponding weight values ​​to energy efficiency indicators according to energy efficiency levels;

[0045] S103, using a weighted support vector machine to establish an energy efficiency index model of the thermal power unit under various operating conditions, using a data subset to train the model, and performing parameter optimization according to the weight value of the energy efficiency index during the training process, so that the trained weighted support vector machine model predicts the energy efficiency index under various operating conditions to obtain an optimal value and outputs it, thereby obtaining the optimal value of the energy efficiency index according to the model;

[0046] S104, taking the optimal value of the energy efficiency index as a benchmark, conduct a real-time evaluation of the current value of the energy efficiency index to provide a reference for the optimized operation of the unit. Specifically, according to steps S101 and S102, the operating data of the thermal power unit under one or more operating conditions is obtained in real time, and the operating data of the energy efficiency index under stable conditions is screened out, and then one or more data subsets corresponding to the operating conditions are obtained according to the operating condition division results, and then the new data subset is input into the trained weighted support vector machine model to obtain the optimal value of the energy efficiency index under the corresponding operating condition, and the energy efficiency index in the new data subset is evaluated in real time based on the optimal value of the energy efficiency index, and the operation of the thermal power unit is adjusted according to the evaluation results.

[0047] In the following, a 660MW supercritical thermal power unit is taken as an object, and the unit operation optimization based on energy efficiency optimization is carried out according to the above steps, and each step is explained in detail.

[0048] In step S101 of this embodiment, when determining the energy efficiency index set of the thermal power unit, the thermal system of the thermal power unit is divided into the boiler and its auxiliary system, and the steam turbine and its auxiliary system. Among them, the boiler and its auxiliary system are further divided into three subsystems, namely the boiler body, the air and smoke system, and the pulverizing system; the steam turbine and its auxiliary system are further divided into three subsystems, namely the steam turbine body, the heat recovery system, and the cold end system. The energy efficiency index includes the energy efficiency index of each subsystem of the thermal power unit and the comprehensive energy efficiency index of the thermal power unit.

[0049] In this embodiment, the energy efficiency index of each subsystem is determined as shown in Appendix 1 based on the process flow of the thermal power unit and the mechanism model of the equipment; at the same time, the power supply coal consumption rate is used as a comprehensive energy efficiency index to characterize the overall energy efficiency of the unit.

[0050] Table 1 Energy efficiency indicators of each subsystem of thermal power unit

[0051]

[0052] The energy efficiency indicators in Table 1 are obtained directly by collecting the operating data of thermal power units, or by calculation according to standards such as "Power Station Boiler Performance Test Procedure" (GB / T-10184) and "Calculation Method of Technical and Economic Indicators of Thermal Power Plants" (DL / T-904).

[0053] In step S101 of this embodiment, when screening the operating data of the selected energy efficiency index of the thermal power unit in a steady-state condition, the following steps are included:

[0054] S201, determine the time span, and specify at least one characterization parameter in the selected energy efficiency index; in this embodiment, the time span t=10min is used to determine whether the thermal power unit is in a stable state; and the unit power, main steam temperature, and main steam pressure are selected as the characterization parameters;

[0055] S202, collecting the operating data in the current time window, and obtaining the maximum value of the operating data in the current time window for each characterization parameter With minimum In this embodiment, according to the real-time acquired operation data of the thermal power unit, the data starting point and end point of the time span t are determined to form a data set within the time period;

[0056] S203, obtaining the maximum value y of the operating data in all time windows of each characterization parameter i,max With the minimum value y i,min ; Determine the maximum value of each characterization parameter With minimum The difference compared to the maximum value y i,max With the minimum value y i,min Whether the difference is less than or equal to the preset threshold; In this embodiment, the following formula is used to determine whether the unit is in a stable state within a time window:

[0057]

[0058] In formula (1), and are the i-th characterization parameters y i The maximum and minimum values ​​in a time window, y i,max and i,min are the i-th characterization parameters y i The maximum and minimum values ​​of all time windows, δ is a preset threshold percentage; in this embodiment, δ = 0.5%;

[0059] S204, if they are all less than or equal to the preset threshold, add the operating data in the current time window to the data set under the steady-state condition, and move the current time window backward by a time span of a specified size, and execute the step of collecting the operating data in the current time window again; in this embodiment, if formula (1) holds, it is determined that the unit is in a stable state in the current time window, and the data set in the current time window is retained; then the start and end points of the time window are moved backward by a time span t to obtain a new time window and jump to execute step S202; otherwise, execute step S205;

[0060] S205, if at least one of the characterization parameters has a maximum value With minimum The difference compared to the maximum value y i,max With the minimum value y i,min If the difference is greater than a preset threshold, the current time window is moved back by one sampling time to obtain a new time window and the process jumps to step S202 to collect the operation data in the current time window again.

[0061] Step S102 of this embodiment specifically includes:

[0062] S301, classify the operating conditions of the thermal power unit. In this embodiment, the operating conditions of the thermal power unit are classified based on the unit load, ambient temperature and coal quality as boundary conditions. The operating condition range and classification standard are shown in Appendix 2.

[0063] Table 2 Classification of operating conditions of thermal power units

[0064]

[0065] According to Table 2, the operating conditions of the thermal power unit are divided into 132 operating conditions. Therefore, for the data set Ω under the steady-state condition obtained in step S101, it is divided into 132 data subsets Ω l (l=1,2,…,132);

[0066] S302, construct a parameter vector for each data subset. l , assuming that the number of data samples it contains is N l Assuming that the energy efficiency indexes of the boiler and its auxiliary systems, the steam turbine and its auxiliary systems together constitute the parameter vector x, and the comprehensive energy efficiency index of the thermal power unit is b, then the data subset can constitute N l The parameter vector z of the data samples i =(x i ,b i )(i=1,2,…,N l ), that is, the parameter vector z corresponding to each thermal power unit sample in the data subset i The parameter vector x is composed of the energy efficiency indicators of all subsystems of the data sample. i and the comprehensive energy efficiency index b of the data sample;

[0067] S303, in each data subset, assign corresponding weight values ​​to the energy efficiency index according to the energy efficiency level, specifically according to the comprehensive energy efficiency index b i From best to worst, the corresponding label value η is calculated using the linear interpolation algorithm. i (0≤η i ≤1.0), and then the parameter vector z for each data sample is i Assign corresponding label values ​​η respectively i as a weight value.

[0068] In this embodiment, the comprehensive energy efficiency index is divided into four categories according to the difference in quality: better, good, average, and poor. In the "better" category, the label coefficient range is 0.8-1.0; in the "good" category, the label coefficient range is 0.6-0.8; in the "average" category, the label coefficient range is 0.4-0.6; in the "poor" category, the label coefficient range is 0.0-0.4.

[0069] According to the parameter vector z i Medium comprehensive energy efficiency indexb i The parameter vector z can be determined by i The label coefficient interval of the label value is located, and then the specific label value can be obtained through a linear interpolation algorithm. The relevant calculation process is well known to those skilled in the art, and the specific calculation process is not repeated in this embodiment.

[0070] In step S103 of this embodiment, the weighted support vector machine model takes the parameter vector x composed of the energy efficiency indicators of all subsystems in the data subset as input and takes the comprehensive energy efficiency indicator b in the data subset as output, and the expression is as follows:

[0071]

[0072] in, is a nonlinear mapping function, ω is a weight vector, b is a bias, ξ i , is a slack variable; ε is an insensitive loss function parameter, C is a penalty coefficient; s i ,t i Respectively represent the weighting coefficients of the i-th data sample in the data subset for the parameters C and ε, 0≤s i ,t i ≤1; where s i Determined by the label coefficient of the data sample, it is the weight value of the i-th data sample; N l Indicates the number of data samples contained in the data subset, x i Represents the parameter vector composed of the energy efficiency indicators of all subsystems in the i-th data sample in the data subset, y i Represents the comprehensive energy efficiency index of the i-th data sample in the data subset.

[0073] In step S104 of this embodiment, when the energy efficiency index is evaluated in real time based on the optimal value of the energy efficiency index, it includes:

[0074] S501, normalizing the current operating data of the energy efficiency index of the specified working condition to obtain the current value of the energy efficiency index of the specified working condition;

[0075] S501, calculating the deviation and the deviation change rate between the current value and the optimal value of the energy efficiency index of the specified working condition, and calculating the real-time quantitative evaluation value of the energy efficiency index of the specified working condition according to the deviation and the deviation change rate.

[0076] In this embodiment, at a certain operating point, for the i-th energy efficiency index x of the new data subset i , the optimal value obtained by the trained weighted support vector machine model is At the kth moment, the deviation E between the normalized current value of the i-th energy efficiency index of the new data subset and its optimal value is i (k) and deviation change rate It can be expressed as:

[0077]

[0078] In the above formula, are the i-th energy efficiency index x i The maximum and minimum values ​​in the running data of and x i (k) represents the i-th energy efficiency index x i The optimal value of E and the normalized current value. i , The value diagram is as follows Figure 2 As shown. Figure 2 It can be seen that the deviation between the process value of the energy efficiency index and the optimal value and the deviation change rate can be uniformly attributed to the range of (-1, 1) through formula (3) and formula (4), which can not only reflect the positive and negative effects of the deviation and the deviation change rate, but also avoid the influence of different dimensions of each energy efficiency index on the evaluation results. For example, when the deviation between the optimal value and its process value is negative, but the deviation change rate is positive, it means that the deviation has a trend of being eliminated, and the evaluation score should be higher; otherwise, the deviation has a trend of being expanded, and the evaluation score should also be lower.

[0079] Correspondingly, the real-time quantitative evaluation value expression is as follows:

[0080]

[0081] Among them, E i (k) is the deviation between the current value and the optimal value of the i-th energy efficiency indicator, is the deviation change rate between the current value and the optimal value of the i-th energy efficiency indicator.

[0082] Under certain working conditions, the main steam pressure is quantitatively evaluated in real time according to the above formula. The evaluation results are shown in the attached figure. Figure 3As shown, it can be seen that when the real-time data of the main steam pressure is close to or equal to the benchmark value, the evaluation result is higher. Conversely, when the real-time data of the main steam pressure is significantly different from the benchmark value or exceeds the benchmark value by a large amount, the evaluation result is lower.

[0083] Based on the impact of the deviation between the real-time data and the optimal value of the energy efficiency indicator on the evaluation result, in this embodiment, the operation of the thermal power unit is adjusted according to the evaluation result, including: if the real-time quantitative evaluation value of the energy efficiency indicator of the specified working condition is less than the specified value, according to the deviation between the optimal value and the current value of the energy efficiency indicator of the specified working condition, the operating state of the thermal power unit is adjusted.

[0084] Specifically, if the real-time quantitative evaluation value of the energy efficiency index is greater than the specified value, it means that the actual value of the energy efficiency index under the working condition is close to the optimal value, so there is no need to adjust the operating state of the thermal power unit; on the contrary, if the real-time quantitative evaluation value of the energy efficiency index is less than the specified value, it means that there is still a large gap between the actual value of the energy efficiency index under the working condition and the optimal value, so the operating state of the thermal power unit can be adjusted according to the deviation between the two. For example, if the real-time quantitative evaluation value of the energy efficiency index of a subsystem is less than the specified value, the operating state of the corresponding subsystem of the thermal power unit is adjusted according to the deviation between the optimal value and the current value. If the real-time quantitative evaluation value of the comprehensive energy efficiency index is less than the specified value, the operating state of multiple subsystems of the thermal power unit is adjusted according to the deviation between the optimal value and the current value. How to adjust the operating state of the thermal power unit according to the deviation between the actual value and the optimal value is well known to those skilled in the art, and the specific implementation process will not be repeated in this embodiment.

[0085] Embodiment 2

[0086] This embodiment proposes a thermal power unit operation optimization system based on energy efficiency optimization, including an interconnected microprocessor and a computer-readable storage medium, wherein the microprocessor is programmed or configured to execute the thermal power unit operation optimization method based on energy efficiency optimization described in Example 1.

[0087] This embodiment further proposes a computer-readable storage medium, in which a computer program is stored. The computer program is used to be programmed or configured by a microprocessor to execute the method for optimizing the operation of a thermal power unit based on energy efficiency optimization as described in Embodiment 1.

[0088] In summary, the present invention discloses a method and system for optimizing the operation of thermal power units based on energy efficiency optimization, which performs "labeling" processing on the operation data of energy efficiency indicators, that is, assigning higher weights to data with high energy efficiency and lower weights to data with low energy efficiency, and at the same time establishing an energy efficiency indicator model by training a weighted support vector machine, so that the established energy efficiency indicator model can obtain the optimal reference value of the energy efficiency indicator under various working conditions, avoiding the drawback that the conventional clustering algorithm is affected by the "data volume" when obtaining the optimal value of the energy efficiency indicator, resulting in the judgment of the optimal energy efficiency of the unit deviating from the actual optimal value, and the obtained energy efficiency optimal value is more generalized. The present invention also conducts real-time quantitative evaluation of the deviation and deviation change rate of the normalized energy efficiency indicator from its optimal value, and the evaluation method is simple, the calculation workload is small, and it is convenient for engineering application.

[0089] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, may be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the functions in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1The steps for the functions specified in one or more boxes.

[0090] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for optimizing the operation of a thermal power unit based on energy efficiency optimization, characterized in that: The following steps are involved: Acquire the operating data of thermal power units in real time, filter the operating data of the specified energy efficiency indicators of the thermal power units under steady-state conditions, and obtain the data set under steady-state conditions; The operating conditions of the thermal power units are divided, and according to the division results, the data set under the steady-state condition is divided into corresponding data subsets, and in each data subset, the energy efficiency index is assigned a corresponding weight value according to the energy efficiency level; Using the data subset to train a weighted support vector machine model, so that the trained weighted support vector machine model predicts the energy efficiency index under each working condition to obtain the optimal value and outputs it, and during the training process of the weighted support vector machine model, parameter optimization is performed according to the weight value of the energy efficiency index; The new data subset is input into the trained weighted support vector machine model to obtain the optimal value of the energy efficiency index under the corresponding working conditions. The energy efficiency index is evaluated in real time based on the optimal value of the energy efficiency index, and the operation of the thermal power unit is adjusted according to the evaluation results.

2. The method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to claim 1, characterized in that: When screening the operating data of the selected energy efficiency indicators of the thermal power unit under steady-state conditions, the following steps are included: Specify at least one characterization parameter in the specified energy efficiency index; Collect the running data in the current time window and obtain the maximum value of the running data in the current time window for each characterization parameter With minimum Get the maximum value y of the running data in all time windows of each characterization parameter i,max With the minimum value y i,min ; Determine the maximum value of each characterization parameter With minimum The difference compared to the maximum value y i,max With the minimum value y i,min Whether the differences are all less than or equal to a preset threshold; If they are all less than or equal to the preset threshold, the operating data in the current time window is added to the data set under the steady-state condition, and the current time window is moved backward by a time span of a specified size, and the step of collecting the operating data in the current time window is performed again; If at least one of the characterization parameters has a maximum value With minimum The difference compared to the maximum value y i,max With the minimum value y i,min If the difference is greater than a preset threshold, the current time window is moved back by one sampling time, and the step of collecting the operating data in the current time window is executed again.

3. The method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to claim 1, characterized in that: The classification of the operating conditions of thermal power units specifically refers to the classification of the operating conditions of thermal power units based on unit load, ambient temperature and coal quality as boundary conditions.

4. The method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to claim 1, characterized in that: The energy efficiency index includes the energy efficiency index of each subsystem of the thermal power unit and the comprehensive energy efficiency index of the thermal power unit. The parameter vector of each data sample in the data subset includes the parameter vector composed of the energy efficiency indexes of all subsystems of the data sample and the comprehensive energy efficiency index of the thermal power unit. When the energy efficiency index is assigned a corresponding weight value according to the energy efficiency level in each data subset, it includes: According to the order of comprehensive energy efficiency indicators from best to worst, the corresponding label value is calculated by linear interpolation algorithm, and then the parameter vector of each data sample in the data subset is assigned the corresponding label value as the weight value.

5. The method for optimizing the operation of thermal power units based on energy efficiency optimization according to claim 4 is characterized in that: The weighted support vector machine model takes the parameter vector composed of the energy efficiency indicators of all subsystems as input and the comprehensive energy efficiency indicator as output, and the expression is as follows: in, is a nonlinear mapping function, ω is a weight vector, b is a bias, ξ i , is a slack variable; ε is an insensitive loss function parameter, C is a penalty coefficient; s i ,t i Respectively represent the weighting coefficients of the i-th data sample in the data subset for the parameters C and ε, where s i is the weight value of the i-th data sample; N l Indicates the number of data samples contained in the data subset, x i Represents the parameter vector composed of the energy efficiency indicators of all subsystems in the i-th data sample in the data subset, y i Represents the comprehensive energy efficiency index of the i-th data sample in the data subset.

6. The method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to claim 1, characterized in that: When the energy efficiency index is evaluated in real time based on the optimal value of the energy efficiency index, it includes: Normalize the current operating data of the energy efficiency index of the specified working condition to obtain the current value of the energy efficiency index of the specified working condition; The deviation and the deviation change rate between the current value and the optimal value of the energy efficiency index of the specified working condition are calculated, and the real-time quantitative evaluation value of the energy efficiency index of the specified working condition is calculated according to the deviation and the deviation change rate.

7. The method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to claim 6, characterized in that: The real-time quantitative evaluation value expression is as follows: Among them, E i (k) is the deviation between the current value and the optimal value of the i-th energy efficiency indicator, is the deviation change rate between the current value and the optimal value of the i-th energy efficiency indicator.

8. The method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to claim 6, characterized in that: When adjusting the operation of the thermal power unit according to the evaluation results, it includes: if the real-time quantitative evaluation value of the energy efficiency index of the specified operating condition is less than the specified value, adjusting the operating state of the thermal power unit according to the deviation between the optimal value and the current value of the energy efficiency index of the specified operating condition.

9. A thermal power unit operation optimization system based on energy efficiency optimization, characterized in that: The invention comprises a microprocessor and a computer-readable storage medium connected to each other, wherein the microprocessor is programmed or configured to execute the method for optimizing the operation of a thermal power unit based on energy efficiency optimization as claimed in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored therein, characterized in that: The computer program is used to be programmed or configured by a microprocessor to execute the method for optimizing the operation of a thermal power unit based on energy efficiency optimization according to any one of claims 1 to 8.

Images