A boiler load distribution method based on energy efficiency curve

Abstract

The present application relates to a kind of boiler load distribution methods based on energy efficiency curve, comprising the following steps: obtaining the historical data of all boilers in unit in preset time period, using specified data cleaning and processing strategy to process the historical data of each boiler, obtain the energy efficiency curve function of each boiler in unit;According to the energy efficiency curve function of each boiler in unit and the current operating state of unit, using balanced optimal strategy to carry out boiler load distribution, periodically based on current operating state adjustment the load distribution of each boiler, so that the energy efficiency of each boiler reaches optimal load in bearing range, keep the overall energy efficiency of unit in optimal state.The present application provides a kind of boiler load distribution methods based on energy efficiency curve, solves the problems of large calculation error of boiler energy efficiency in existing allocation method, poor economy of load distribution between boilers or high difficulty of calculation process, which leads to low practicality.

Description

Technical Field

[0001] This invention relates to the field of thermal power generation technology, and in particular to a boiler load allocation method based on energy efficiency curves. Background Technology

[0002] The steam pressure of the main pipe boiler is a key control parameter for unit operation, directly affecting the safe and economical operation of the unit. In the main pipe operation mode, the load increase or decrease signal needs to be sent to each boiler operating in parallel based on the deviation between the main pipe pressure and the given value. Each boiler receives the signal and adjusts the fuel quantity and air volume to quickly meet the load requirements, thereby responding to the adjustment of the main pipe pressure.

[0003] Existing boiler unit load allocation methods include allocation based on the boiler unit load ratio, allocation based on the principle of maximizing the overall unit efficiency, and allocation based on the principle of equal incremental rate of fuel consumption. These allocation methods are usually based on boiler energy efficiency data.

[0004] Boiler energy efficiency generally needs to be determined by on-site testing by an energy efficiency testing agency, or calculated based on historical operating data using the positive balance method or the reverse balance method. The positive balance method does not consider the influence of steam humidity, boiler heat loss, etc., and the calculated energy efficiency data is too affected by operating parameters and noise data. In contrast, the various heat losses in the reverse balance method require more measurements and data collection is difficult. Some losses cannot be directly measured and can only be estimated.

[0005] Existing boiler load allocation methods, based on the aforementioned methods of obtaining boiler energy efficiency, suffer from problems such as large errors in boiler energy efficiency calculation, poor economic efficiency in load allocation between boilers, or low practicality due to the complexity of the calculation process. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a boiler load allocation method based on energy efficiency curves.

[0008] (II) Technical Solution

[0009] To achieve the above objectives, the main technical solutions adopted by the present invention include:

[0010] In a first aspect, embodiments of the present invention provide a boiler load allocation method based on an energy efficiency curve, comprising the following steps:

[0011] S1, acquire historical data of all boilers in the unit within a preset time period, process the historical data of each boiler using a specified data cleaning and processing strategy, and acquire the energy efficiency curve function of each boiler in the unit.

[0012] S2, based on the energy efficiency curve function of each boiler in the unit and the current operating status of the unit, adopts the optimal balance strategy to allocate boiler load, and periodically adjusts the load allocation of each boiler based on the current operating status, so that the energy efficiency of each boiler reaches the optimal load within the bearing range, and keeps the overall energy efficiency of the unit in the optimal state.

[0013] The current operating status of the unit includes: the current operating load and rated load of all boilers in the unit, the total load to be allocated to the unit, and the minimum load adjustment; the number of boilers in the unit is greater than 1.

[0014] Optionally, S1 acquiring historical data for all boilers within a preset time period includes:

[0015] Historical boiler operating data is acquired in minutes, including timestamps, fuel consumption, boiler load, steam pressure, steam temperature, feedwater pressure, and feedwater temperature.

[0016] Obtain fuel analysis report data on a daily or shift basis, including timestamps and lower heating value of fuel;

[0017] Data on heat loss metrics are acquired in minutes, including timestamps, flue gas volume, flue gas temperature, exhaust oxygen content, pulverized coal fineness, furnace temperature, and the content of different gases in the flue gas.

[0018] Optionally, S1 includes:

[0019] S101, the historical data is horizontally stitched together using timestamps as indexes, and the operating data segment with boiler load fluctuation amplitude less than the preset threshold for stable operating conditions is selected as the stable operating condition energy efficiency dataset by sliding filter.

[0020] S102, Based on the stable energy efficiency dataset under operating conditions, the effective data segment that meets the preset threshold of the heat loss measurement index is selected as the first preprocessed dataset by filtering using the box plot method according to the heat loss measurement index.

[0021] The heat loss measurement indicators include one or more of the following: flue gas volume, flue gas temperature, exhaust oxygen content, pulverized coal fineness, furnace temperature, and the content of different gases in the flue gas.

[0022] S103, calculate the boiler energy efficiency E based on the first preprocessed dataset, and combine it with the first preprocessed data to obtain the second preprocessed dataset;

[0023] The formula for calculating boiler energy efficiency E is as follows:

[0024]

[0025] in,

[0026] h给水 =H(P 给水 ,T 给水 ),

[0027] h 蒸汽 =H(P 蒸汽 ,T 蒸汽 ),

[0028] Q represents the boiler load, and C represents the fuel consumption. net For the lower heating value of fuel, P 给水 T 给水 These represent the water supply pressure and water supply temperature, respectively. 蒸汽 T 蒸汽 These represent steam pressure and steam temperature, respectively, and H is a function for calculating the enthalpy of water and water vapor.

[0029] S104, perform outlier processing on the second preprocessed dataset to generate an energy efficiency model dataset;

[0030] S105, Fit a quadratic polynomial curve to the energy efficiency model dataset to obtain the boiler energy efficiency curve function:

[0031] E = f(Q),

[0032] Here, Q represents the boiler load and E represents the boiler energy efficiency.

[0033] Optionally, the sliding filter selects operating data segments where the boiler load fluctuation amplitude is less than a preset threshold for stable operating conditions, including:

[0034] Slide along the timestamp in a window longer than 10 minutes to select operating data segments where the boiler load fluctuation amplitude is less than the preset threshold for stable operating conditions as valid data;

[0035] The preset threshold for stable operating conditions is 1%.

[0036] Optionally, step S102 uses a box plot method for filtering, including:

[0037] If we determine the upper quartile q1 and lower quartile q3 of the heat loss measurement index, and the interquartile range iqr = q1 - q3, then the upper limit of the effective data is q1 + 1.5 * iqr, and the lower limit of the effective data is q3 - 1.5 * iqr.

[0038] Optionally, step S104 includes the following steps:

[0039] SS1 divides the boiler load Q into equal-length intervals with a preset step size, and divides the second preprocessed dataset into different intervals according to the interval distribution of the boiler load Q;

[0040] SS2 uses a BP neural network prediction model to handle outliers for each interval, generating valid data for that interval.

[0041] SS3 merges the valid data from each interval to obtain the energy efficiency model dataset;

[0042] The preset step size is 1.

[0043] Optionally, the SS2 includes:

[0044] The BP neural network prediction model uses heat loss as the independent variable and boiler energy efficiency E as the dependent variable. It includes an input layer, a hidden layer, and an output layer, and selects tanh as the neuron activation function.

[0045] The BP neural network prediction model is trained using data within the specified interval to obtain an interval prediction model;

[0046] The interval prediction model is used to predict the data within the interval, and the predicted value is obtained. The percentage of absolute error is then calculated using the following formula:

[0047]

[0048] Data with an absolute error percentage greater than the preset error threshold are removed as outliers, and the remaining valid data are saved as valid data for that range.

[0049] The range of the preset error threshold σ is as follows:

[0050] 0.1≥σ≥0.01.

[0051] Optionally, the boiler load allocation using the optimal balance strategy includes the following specific steps:

[0052] S2-1, Based on the energy efficiency curve function of each boiler, obtain the optimal load for each boiler that maximizes its energy efficiency.

[0053] S2-2, the load allocation factor for each boiler is obtained based on the optimal load, current operating load, and rated load of each boiler;

[0054] S2-3, load allocation is performed based on the load allocation factors of all boilers in the unit, as well as the total load to be allocated and the minimum load adjustment amount:

[0055] If the total load to be allocated to the unit is >0, the boiler with the smallest load allocation factor will be allocated the load with the smallest adjustment amount of load increase;

[0056] If the total load to be allocated to the unit is less than 0, the boiler with the largest load allocation factor will be allocated the load with the minimum adjustment amount of load reduction.

[0057] S2-4: If the total load to be allocated to the unit is 0, the boiler load allocation ends; otherwise, update the current operating status of the unit and repeat S2-2 and S2-3.

[0058] Optionally, S2-2 includes:

[0059] The load distribution factor is calculated using the following formula:

[0060]

[0061] Among them, κ i Let Q be the load distribution factor for boiler i. i The current operating load of boiler i, Q is the rated load of boiler i. i,best This represents the optimal load for boiler i.

[0062] In a second aspect, the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the boiler load allocation method based on the energy efficiency curve described above.

[0063] (III) Beneficial Effects

[0064] Compared with existing technologies, this invention, based on the energy efficiency curve function of each boiler within the unit and the current operating status of the unit, employs a balanced optimal strategy for boiler load allocation. It periodically adjusts the load allocation of each boiler based on the current operating status, ensuring that the energy efficiency of each boiler reaches its optimal load within its tolerance range. This maintains the overall energy efficiency of the unit at its optimal state, achieving coordinated and optimized control among multiple boilers, and providing precise support for the economical and safe operation of the unit. It improves reliability, safety, stability, efficiency, and accuracy.

[0065] Furthermore, this invention utilizes a large amount of historical boiler data and employs effective data cleaning and filtering methods, including selecting stable operating data with small boiler load fluctuations through sliding filtering and preprocessing the data using a heat loss measurement index based on a box plot method. This reduces the impact of heat loss during energy efficiency calculations while avoiding the impact of data acquisition difficulty and measurement accuracy, thereby improving the accuracy and effectiveness of the obtained energy efficiency curves.

[0066] Meanwhile, during the load allocation process of the boiler unit, this invention uses the fitted energy efficiency curve function and the current operating status of the boiler to calculate the load allocation factor. When the load demand increases, it is preferentially allocated to the boiler with the smallest load allocation factor, and when the load demand decreases, it is preferentially allocated to the boiler with the largest load allocation factor, so as to ensure that the relative distance between the operating load of each boiler and the optimal operating state is approximately the same. The allocation method described in this invention takes into account both economic efficiency and safe operation, and eliminates the limitation of the technical level of the operators. The real-time load allocation calculation results guide the realization of coordinated and optimized control among multiple boilers, providing precise support for the economic and safe operation of the unit. Attached Figure Description

[0067] Figure 1 This is a flowchart of the boiler load allocation method based on energy efficiency curves according to the present invention;

[0068] Figure 2 A box plot of heat loss measurement index provided in an embodiment of the present invention;

[0069] Figure 3 This is a boiler energy efficiency fitting curve provided in an embodiment of the present invention;

[0070] Figure 4 This is a schematic diagram of anomaly identification based on a neural network prediction model provided in an embodiment of the present invention. Detailed Implementation

[0071] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention can be understood more clearly and thoroughly, and that the scope of the present invention can be fully conveyed to those skilled in the art.

[0072] Example 1

[0073] like Figure 1 As shown, this embodiment provides a boiler load allocation method based on energy efficiency curves, including the following steps:

[0074] S1, acquire historical data of all boilers in the unit within a preset time period, process the historical data of each boiler using a specified data cleaning and processing strategy, and acquire the energy efficiency curve function of each boiler in the unit.

[0075] In practical applications, for example, historical data can be:

[0076] Historical boiler operating data is acquired in minutes, including timestamps, fuel consumption, boiler load, steam pressure, steam temperature, feedwater pressure, and feedwater temperature.

[0077] Obtain fuel analysis report data on a daily or shift basis, including timestamps and lower heating value of fuel;

[0078] Data on heat loss metrics are acquired in minutes, including timestamps, flue gas volume, flue gas temperature, exhaust oxygen content, pulverized coal fineness, furnace temperature, and the content of different gases in the flue gas.

[0079] The fuel testing report data obtained on a daily or shift basis indicates that the data remains unchanged at any time within the current operating shift or on the same day.

[0080] In practical applications, the number of boilers in a unit is greater than one.

[0081] S2, based on the energy efficiency curve function of each boiler in the unit and the current operating status of the unit, adopts the optimal balance strategy to allocate boiler load, and periodically adjusts the load allocation of each boiler based on the current operating status, so that the energy efficiency of each boiler reaches the optimal load within the bearing range, and keeps the overall energy efficiency of the unit in the optimal state.

[0082] For example, the current operating status of the unit includes: the current operating load and rated load of all boilers in the unit, the total load to be allocated to the unit, and the minimum load adjustment amount, etc.

[0083] The boiler load allocation method based on energy efficiency curves in this embodiment can adjust the load allocation of each boiler based on the current operating status, so that the energy efficiency of each boiler reaches the optimal load within the bearing range, maintain the overall energy efficiency of the unit in the optimal state, realize coordinated and optimized control among multiple boilers, and provide precise support for the economic and safe operation of the unit.

[0084] To better understand step S1 above, sub-steps A1 to A5 will be explained in detail below.

[0085] A1. The historical data is horizontally stitched together using timestamps as indexes, and the operating data segment with boiler load fluctuation amplitude less than the preset threshold for stable operating conditions is selected as the stable operating condition energy efficiency dataset by sliding filter.

[0086] For example, by sliding a time window longer than 10 minutes along the time axis, the operating data segment with a boiler load fluctuation of less than 1% is selected as valid data, and data with abnormal fluctuations in boiler operating efficiency caused by changes in operating conditions is removed.

[0087] In this embodiment, the time window is not limited to 10 minutes and can be selected according to actual needs. The smaller the granularity of the window, the more accurate the data.

[0088] A2, based on the stable energy efficiency dataset under operating conditions, uses the box plot method to filter according to the heat loss measurement index, and selects the effective data segment that meets the preset threshold of the heat loss measurement index as the first preprocessed dataset.

[0089] In practical applications, for example, the heat loss measurement indicators include one or more of the following: flue gas volume, flue gas temperature, exhaust oxygen content, pulverized coal fineness, furnace temperature, and the content of different gases in the flue gas.

[0090] For example, the parameters of a box plot can be set as follows: determine the upper quartile q1 and lower quartile q3 of the heat loss measurement index, and the interquartile range iqr = q1 - q3. Then the upper limit of the effective data is q1 + 1.5 * iqr, and the lower limit of the effective data is q3 - 1.5 * iqr.

[0091] like Figure 2 The figure shows a box plot of the flue gas oxygen content and furnace temperature, which are the heat loss measurement indicators described in this invention. The top and bottom of the box are the upper and lower quartiles of the data, respectively. The line in the middle of the box represents the median. The horizontal lines outside the box are the upper and lower edges, respectively. Points outside the upper and lower edges are outliers. The height of the box reflects the degree of data fluctuation to a certain extent. The flatter the box, the more concentrated the data.

[0092] A3. The boiler energy efficiency E is calculated based on the first preprocessed dataset and then incorporated into the first preprocessed data to obtain the second preprocessed dataset.

[0093] The formula for calculating boiler energy efficiency E is as follows:

[0094]

[0095] in,

[0096] h 给水 =H(P 给水 ,T 给水 ),

[0097] h 蒸汽 =H(P 蒸汽 ,T 蒸汽 ),

[0098] Q represents the boiler load, and C represents the fuel consumption. net For the lower heating value of fuel, P 给水 T 给水 These represent the feedwater pressure (in MPa) and feedwater temperature (in K), respectively. 蒸汽 T 蒸汽 These represent steam pressure (in MPa) and steam temperature (in K), respectively, and H is a function for calculating the enthalpy of water and water vapor.

[0099] A4, perform outlier processing on the second preprocessed dataset to generate the energy efficiency model dataset;

[0100] A5, by fitting a quadratic polynomial curve to the energy efficiency model dataset, the boiler energy efficiency curve function is obtained:

[0101] E = f(Q),

[0102] Here, Q represents the boiler load and E represents the boiler energy efficiency.

[0103] like Figure 3 The figure shows the energy efficiency fitting curve of a boiler with a rated load of 100t / h using the method described in this invention.

[0104] The method used in this embodiment utilizes a large amount of historical boiler data and employs effective data cleaning and filtering techniques, including selecting stable operating data with small boiler load fluctuations through sliding filtering and preprocessing the data using a heat loss measurement index based on a box plot method. This reduces the impact of heat loss during energy efficiency calculations while avoiding the impact of data acquisition difficulty and measurement accuracy, thereby improving the accuracy and effectiveness of the obtained energy efficiency curves.

[0105] To better understand step A4 above, sub-steps A4-1 to A4-3 will be used for detailed explanation below.

[0106] A4-1, divide the boiler load Q into equal-length intervals with a preset step size, and divide the second preprocessed dataset into different intervals according to the interval distribution of the boiler load Q;

[0107] For example, the preset step size can be 1.

[0108] A4-2, for each interval, use a BP neural network prediction model to handle outliers and generate valid interval data;

[0109] For example, in practical applications, the outlier handling using the BP neural network prediction model in this step can be done in the following way:

[0110] Using heat loss metrics such as flue gas volume, flue gas temperature, flue gas oxygen content, pulverized coal fineness, furnace temperature, and the content of different gases in the flue gas as independent variables, and boiler energy efficiency E as the dependent variable, a BP neural network prediction model was established. The neural network model structure (containing an input layer, hidden layer, and output layer, with tanh selected as the neuron activation function) was built using the Keras module in Python and trained using data within the interval to obtain the interval prediction model.

[0111] Using an interval prediction model to predict data within an interval, calculate the percentage of absolute error:

[0112]

[0113] Set an error threshold σ (0.1 ≥ σ ≥ 0.01), and remove data with an error percentage greater than the threshold σ as outliers.

[0114] like Figure 4 The diagram shown illustrates the outlier handling using the aforementioned neural network prediction model in this embodiment. Within the boiler load range of 80-81%, an error threshold σ = 0.03 is set, and data points with an absolute error percentage greater than σ are identified as outliers.

[0115] A4-3, merge the valid data from each interval to obtain the energy efficiency model dataset;

[0116] In this embodiment, the use of a BP neural network prediction model for outlier processing is characterized by high reliability and ease of operation. Furthermore, because the structure of the neural network prediction model is consistent across all intervals, with only the intervals being divided, it allows for rapid batch modeling during implementation, thereby enabling the rapid and accurate identification and processing of outliers in energy efficiency data.

[0117] To better understand step S2 above, sub-steps B1 to B4 will be used for detailed explanation below.

[0118] B1. Based on the energy efficiency curve function of each boiler, obtain the optimal load for each boiler that maximizes its energy efficiency.

[0119] For example, the energy efficiency curve function of boiler i is:

[0120] E i =f i (Q),

[0121] Among them, E i Let Q be the energy efficiency of boiler i, Q be the boiler load, and i be the boiler number (i = 1, 2, ..., n, where n is the number of boilers in the unit).

[0122] Based on this energy efficiency curve function, and using conventional mathematical methods, the energy efficiency E of boiler i can be obtained. i The optimal load Q that reaches its maximum value i,best .

[0123] B2, the load allocation factor for each boiler is obtained based on the optimal load, current operating load, and rated load of each boiler;

[0124] For example, the load distribution factor can be calculated using the following formula:

[0125]

[0126] Among them, κ i Let Q be the load distribution factor for boiler i. i The current operating load of boiler i, Q is the rated load of boiler i. i,best Let i be the optimal load for boiler i, where i is the boiler number (i = 1, 2, ..., n, where n is the number of boilers in the unit).

[0127] B3. Load allocation is performed based on the load allocation factors of all boilers within the unit, as well as the total load to be allocated and the minimum load adjustment.

[0128] If the total load to be allocated to the unit is >0, the boiler with the smallest load allocation factor will be allocated the load with the smallest adjustment amount of load increase;

[0129] If the total load to be allocated to the unit is less than 0, the boiler with the largest load allocation factor will be allocated the load with the minimum adjustment amount of load reduction.

[0130] B4. If the total load to be allocated to the unit is 0, the boiler load allocation ends; otherwise, update the current operating status of the unit and repeat B2 and B3.

[0131] In this embodiment, the above-mentioned optimal balance strategy prioritizes the boiler with the smallest load allocation factor when the load demand increases, and prioritizes the boiler with the largest load allocation factor when the load demand decreases. Through the coordinated allocation among multiple boilers, it can ensure that the relative distance between the operating load of each boiler and the optimal operating state is roughly the same, thereby maximizing the unit's energy efficiency and production efficiency.

[0132] Example 2

[0133] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of any of the boiler load allocation methods based on energy efficiency curves in Embodiment 1 above.

[0134] Since the systems / devices described in the above embodiments of the present invention are systems / devices used to implement the methods of the above embodiments of the present invention, those skilled in the art can understand the specific structure and modifications of the systems / devices based on the methods described in the above embodiments of the present invention, and therefore will not be repeated here. All systems / devices used in the methods of the above embodiments of the present invention fall within the scope of protection of the present invention.

[0135] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0137] It should be noted that any reference numerals placed between parentheses in the claims should not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In claims that enumerate several means, several of these means may be embodied by the same hardware. The use of the terms first, second, third, etc., is merely for convenience of expression and does not indicate any order. These terms can be understood as part of the component names.

[0138] Furthermore, it should be noted that in the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0139] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the claims should be interpreted to include both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0140] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, then this invention should also include these modifications and variations.

Claims

1. A method for boiler load distribution based on energy efficiency curve, characterized in that, A method for boiler unit load distribution, comprising the following steps: S1, obtaining historical data of all boilers in the unit within a preset time period, processing the historical data of each boiler using a specified data cleaning and processing strategy, and obtaining an energy efficiency curve function of each boiler in the unit; S1 comprises: S101, horizontally splicing the historical data with time stamp as index, and selecting a running data segment with a boiler load fluctuation amplitude less than a stable working condition preset threshold as a stable working condition energy efficiency data set through sliding filtering; S102, based on the stable working condition energy efficiency data set, filtering according to a heat loss measurement index using a box plot method, and selecting an effective data segment meeting a heat loss measurement index preset threshold as a first preprocessed data set; Wherein, the heat loss measurement index includes one or more of the following: flue gas volume, flue gas temperature, flue gas oxygen content, coal powder fineness, furnace temperature and content of different gases in flue gas; S103, calculating the boiler energy efficiency E according to the first preprocessed data set, and integrating the first preprocessed data to obtain a second preprocessed data set; Wherein, the boiler energy efficiency E is calculated according to the following formula: ， Wherein, ， ， Q is the boiler load, C is the fuel consumption, Q is the boiler load, C is the fuel consumption, Pw and Tw are the feedwater pressure and temperature, respectively, Pw and Tw are the feedwater pressure and temperature, respectively, H is a function of the enthalpy of water and steam. S104, performing abnormal point processing on the second preprocessed data set to generate an energy efficiency model data set; S105, fitting a quadratic polynomial curve to the energy efficiency model data set to obtain a boiler energy efficiency curve function: ， Wherein, Q is the boiler load, and E is the boiler energy efficiency; S2, according to the energy efficiency curve function of each boiler in the unit and the current operating state of the unit, using a balance optimal strategy to distribute the boiler load, periodically adjusting the load distribution of each boiler based on the current operating state, so that the energy efficiency of each boiler reaches the optimal load within the bearing range, and the overall energy efficiency of the unit is kept in the optimal state; The current operating state of the unit includes: the current operating load and rated load of all boilers in the unit, the total load to be distributed of the unit, and the minimum load adjustment amount; the number of boilers in the unit is greater than 1.

2. A method for boiler load distribution based on energy efficiency curve as claimed in claim 1 wherein, The S1 obtains the historical data of all boilers in the unit within a preset time period, which comprises: Obtain the boiler historical operation data in minutes, including time stamp, fuel consumption, boiler load, steam pressure, steam temperature, feed water pressure and feed water temperature; Obtain the fuel test report data in days or shifts, including time stamp and low calorific value of fuel; Obtain the heat loss measurement index data in minutes, including time stamp, flue gas volume, flue gas temperature, flue gas oxygen content, coal powder fineness, furnace temperature and content of different gases in flue gas.

3. A method for boiler load distribution based on energy efficiency curve as claimed in claim 1 wherein, The sliding filtering to select a running data segment with a boiler load fluctuation amplitude less than a stable working condition preset threshold comprises: Sliding along the time stamp with a window length greater than 10 minutes, and selecting a running data segment with a boiler load fluctuation amplitude less than a stable working condition preset threshold as an effective data; Wherein, the stable working condition preset threshold is 1%.

4. A method for boiler load distribution based on energy efficiency curve as claimed in claim 1 wherein, The S102 filtering using a box plot method comprises: determining a lower quartile q3 of the heat loss measure , and an upper limit of valid data is , and a lower limit of valid data is .

5. A method for boiler load distribution based on energy efficiency curve as claimed in claim 1 wherein, The S104 comprises the following steps: SS1, dividing the boiler load Q into equal length intervals at a preset step value interval, and dividing the second preprocessed data set into different intervals according to the interval distribution of the boiler load Q; SS2, for each interval, using the BP neural network prediction model to process the abnormal points, generating interval effective data; SS3, merging each interval effective data, obtaining the energy efficiency model data set; Wherein, the preset step value is 1.

6. A method for boiler load distribution based on energy efficiency curve as claimed in claim 5 wherein, The SS2 comprises: The BP neural network prediction model takes the heat loss measurement index as the independent variable, takes the boiler energy efficiency E as the dependent variable, contains the input layer, the hidden layer, the output layer, and selects tanh as the neuron activation function; The BP neural network prediction model uses the data in the interval to train and obtain the interval prediction model; Using the interval prediction model to predict the data in the interval, obtaining the predicted value, and calculating the absolute error percentage according to the following formula: ； The data with an absolute error percentage greater than the error threshold preset value is removed as an abnormal point, and the remaining effective data is saved as the interval effective data; The error threshold preset value is in a value range of 0.1 to 0.

5. The value range of the error threshold preset value is 0.1 to 0.

5. 。 7. A method for boiler load distribution based on energy efficiency curve as claimed in claim 1 wherein, The boiler load distribution is distributed by adopting a balanced optimal strategy, comprising the following specific steps: S2-1, according to the energy efficiency curve function of each boiler, obtaining the optimal load of each boiler to make the energy efficiency of each boiler reach the maximum value; S2-2, according to the optimal load, the current running load and the rated load of each boiler, obtaining the load distribution factor of each boiler; S2-3, according to the load distribution factor of all boilers in the unit, the total load to be distributed and the minimum load adjustment amount, the load is distributed: If the total load to be distributed is greater than 0, the boiler with the smallest load distribution factor is distributed with the load of the minimum load adjustment amount; If the total load to be distributed is less than 0, the boiler with the largest load distribution factor is distributed with the load of the minimum load adjustment amount; S2-4, if the total load to be distributed is 0, the boiler load distribution is ended; Otherwise, update the current running state of the unit, and repeat S2-2 and S2-3.

8. A method for boiler load distribution based on energy efficiency curve as claimed in claim 7 wherein, The S2-2 comprises: The load distribution factor is calculated according to the following formula: ， wherein, is a load distribution factor for the boiler i, is a current operating load of the boiler i, is a rated load of the boiler i, is an optimal load of the boiler i.

9. A computer apparatus, comprising: The computer program stored in the memory and executable on the processor, when the processor executes the computer program, realizes the steps of the boiler load distribution method based on the energy efficiency curve according to any one of claims 1 to 8.

Images