Thermal power plant external heating load real-time prediction method
By screening and seasonal division of the historical data of the thermal power plant, establishing a heating steam extraction forecast model, and real-time monitoring data are corrected, the problems of large amount of thermal load prediction and high sensitivity in the existing technology have been solved, and significantly improved The prediction accuracy and model adaptability are achieved to ensure the stable operation of the power system.
Patent Information
- Application Number
- CN202411208406.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-08-30
AI Technical Summary
The existing thermal load prediction methods have problems with large-scale data processing and parameter tuning, which can affect the prediction accuracy when data is insufficient or changes are large, and complex data processing processes will also affect the actual situation. Application efficiency and reliability.
A real-time prediction method for external heating load of thermal power plants is adopted, including obtaining and screening historical heating data, dividing seasons according to geographical location characteristics, selecting typical data for analysis, establishing a prediction model for heating and steam extraction, and passing daily supply Thermal load formula fitting and real-time monitoring data are modeled to improve prediction accuracy.
Significantly improve the accuracy of heating load prediction, provide reliable data support for the scheduling and operation of cogeneration units, ensure efficient and stable operation of the power system, and reduce risks caused by load fluctuations.
Smart Images

Figure CN119093351A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of thermal power plant heating load monitoring, and in particular to a real-time prediction method for external heating load of a thermal power plant. Background Art
[0002] Thermal load forecasting is a key link in the operation of thermal power plants. With the continuous growth of energy demand and the increase in grid load volatility, the role of cogeneration units in the power system is becoming increasingly important. Cogeneration systems can not only efficiently utilize waste heat for heating or industrial heat, thereby improving energy utilization efficiency, but also provide electricity.
[0003] During peak electricity consumption periods, ensuring the stability of heat supply while ensuring the reliability and flexibility of power supply has become a key issue. Therefore, heat load forecasting is particularly important. Accurate heat load forecasting can provide reliable data support for the operation mode of the heating unit. By accurately predicting the heat load, the operation of the unit can be optimized, making it more flexible and adjustable during peak electricity consumption periods. This not only helps to improve the power supply reliability of the power system, but also effectively ensures the stability of heat supply and reduces the risks caused by load fluctuations.
[0004] The existing heat load prediction methods currently have problems with large computational complexity and high sensitivity in large-scale data processing and parameter tuning. The parameter settings and models are highly complex and highly dependent on data, which affects the prediction accuracy when the data is insufficient or varies greatly. The complex data processing process will also affect the efficiency and reliability of practical applications. Summary of the invention
[0005] The purpose of the present invention is to overcome the defects of the existing technology and propose a real-time prediction method for the external heating load of a thermal power plant, which can realize the heat load prediction more accurately, provide reliable data support for the economic operation of the cogeneration unit, enable it to have stronger flexibility and adjustment ability during peak electricity consumption, improve the reliability of the heating system, and reduce the risks caused by load fluctuations.
[0006] To achieve the above object, the present invention adopts the following specific technical solutions:
[0007] The present invention provides a method for real-time prediction of external heating load of a thermal power plant, comprising the following steps:
[0008] S1. Obtain and filter historical heating data; obtain historical heating data from the operating system of a single heating unit in a power plant in the heating area, perform time series processing, average the heating data per hour, and remove data with a heating data less than zero;
[0009] S2. Divide the climate throughout the year into seasons according to the geographical location characteristics; according to the months included in each season, take the month as the first variable to achieve the association between the month and the heating steam extraction amount;
[0010] S3. Select typical data for analysis; select the heating steam extraction data of each section of the heating unit in the past two years, select three typical working days in the first ten days, middle ten days and last ten days of each month for data analysis, establish a heating steam extraction prediction model with "day" as the second variable, and realize the correlation between "day" and heating steam extraction;
[0011] S4. Fitting the daily heating load formula: Fit the average heating steam extraction data of the selected typical working day to obtain the relationship between the heating steam extraction volume and time in the heating steam extraction volume prediction model, so as to predict the trend of the heating steam extraction volume in each section at a certain time in the future;
[0012] S5. Monitor the current heating steam extraction data in real time and modify the formula to improve the accuracy of the prediction.
[0013] Furthermore, in step S4, the daily heating load formula is fitted as follows:
[0014] F cr,i (x) = a i,4 x 4 +a i,3 x 3 +a i,2 x 2 +a i,1 x+A i ;
[0015] F r,i (x) = b i,4 x 4 +b i,3 x 3 +b i,2 x 2 +b i,1 x+B i ;
[0016] F lr,i (x) = c i,4 x 4 +c i,3 x 3 +c i,2 x 2 +c i,1 x+C i ;
[0017] F zr,i (x) = d i,4 x 4 +d i,3 x 3 +di,2 x 2 +d i,1 x+D i ;
[0018] G cro =F cr,i (x);
[0019] G ro =F r,i (x);
[0020] G lro =F lr,i (x);
[0021] G zro =F zr,i (x);
[0022] Among them, x is the current time, such as a certain time, h; a i,4 is the fourth-order coefficient of the single-pump heating formula in a certain season; a i,3 is the third-order coefficient of the single-pump heating formula in a certain season; a i,2 is the second-order coefficient of the single-draw heating formula in a certain season; a i,1 A is the first-order coefficient of the heating formula for a certain season; i is the constant of the heating formula for a certain season;
[0023] b i,4 is the fourth-order coefficient of the heat reheating formula for a certain season; b i,3 is the third-order coefficient of the heat reheating formula for a certain season; b i,2 is the second-order coefficient of the heat reheating formula for a certain season; b i,1 is the first-order coefficient of the heat reheating formula for a certain season; B i is the constant of the heat reheating formula for a certain season;
[0024] c i,4 is the fourth-order coefficient of the cooling reheating formula in a certain season; c i,3 is the third-order coefficient of the cooling reheating formula in a certain season; c i,2 is the second-order coefficient of the cooling and reheating formula in a certain season; c i,1 is the first-order coefficient of the cooling reheating formula in a certain season; C i is the constant of the cooling reheating formula in a certain season;
[0025] d i,4 is the fourth-order coefficient of the heating formula in a certain season; d i,3 is the third-order coefficient of the heating formula in a certain season; d i,2 is the second-order coefficient of the heating formula in a certain season; d i,1 is the first-order coefficient of the heating formula in a certain season; D iis the constant of the heating formula in a certain season;
[0026] i is the season, spring is 1: March to June; summer is 2: July to September; autumn is 3: October to November; winter is 4: December to February of the following year; G cr0 , G r0 , G lr0 , G zr0 They are the unit’s predicted average first extraction heat supply, hot resupply heat supply, cold resupply heat supply and middle exhaust heat supply, t / h.
[0027] Furthermore, in step S5, the current heating steam extraction data is monitored in real time and the formula is corrected. The corrected formula is as follows:
[0028]
[0029] Among them, Δx is the time difference between a future moment and the current moment, h; G cr1 , G r1 , G lr1 , G zr1 They are the unit’s predicted dynamic first-extraction heating supply, hot resupply heating supply, cold resupply heating supply and intermediate exhaust heating supply, t / h.
[0030] The present invention can achieve the following technical effects:
[0031] The real-time prediction method for external heating load of a thermal power plant provided by the present invention can significantly improve the accuracy of heating load prediction through seasonal division, data selection and analysis, formula fitting, real-time monitoring and model correction, provide reliable data support for the scheduling and operation of cogeneration units, and ensure the efficient and stable operation of the power system. During the implementation process, the data selection and model correction methods can be continuously optimized in combination with actual conditions to further improve the prediction accuracy and adaptability of the model, ensuring that a high prediction accuracy can be maintained under different working conditions. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 It is a flow chart of a method for real-time prediction of external heating load of a thermal power plant provided according to an embodiment of the present invention;
[0033] Figure 2 It is a comparison curve diagram of cold reheating amount prediction data and actual heating amount obtained by using the real-time heating load prediction method for a 300MW unit in winter provided by an embodiment of the present invention;
[0034] Figure 3 It is a comparison curve chart of the heat resupply heat prediction data and the actual heat supply obtained by using the real-time heat supply load prediction method of a certain 300MW unit in spring according to an embodiment of the present invention. DETAILED DESCRIPTION
[0035] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following description, the same modules are represented by the same reference numerals. In the case of the same reference numerals, their names and functions are also the same. Therefore, the detailed description thereof will not be repeated.
[0036] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and do not constitute a limitation of the present invention.
[0037] The embodiment of the present invention provides a method for real-time prediction of external heating load of a thermal power plant, the process of which is as follows: Figure 1 As shown, first, historical heating data is obtained from the operating system of a single heating unit in a power plant in the heating area. Taking the heating steam extraction volume as a positive value as the judgment condition, the data with a heating volume less than zero is judged as invalid data and eliminated. Secondly, the heating steam extraction volume data of each section of the heating unit in the past two years are selected, and three typical working days are selected every month according to the first, middle and last ten days for data analysis. Thirdly, the whole year is simply divided into four quarters to reflect the changes in heating demand in different seasons, and then typical working days are selected from the historical database for data analysis. Finally, the relationship between the heating steam extraction volume and time is obtained by formula fitting, and on this basis, the heating steam extraction volume monitored in real time is used to dynamically correct the obtained relationship. Thereby, an accurate prediction of the heating steam extraction volume at a certain moment in the future is achieved. The present invention can realize efficient, fast-reading and accurate dynamic prediction of heating load. Reliable data support is provided for the economic operation of the cogeneration unit, thereby improving the heating reliability of the heating system and reducing the risks caused by load fluctuations.
[0038] Specifically, the present invention provides a method for real-time prediction of external heating load of a thermal power plant, comprising the following steps:
[0039] S1. Obtain and filter historical heating data.
[0040] The historical heating data is obtained from the operating system of a single heating unit in the power plant in the heating area, and the data is processed in a time series manner, and the heating amount is averaged in each hour. The invalid data obtained is deleted, and the data with a heating amount less than zero is judged as invalid data and eliminated based on the positive value of the heating steam extraction volume. This processing method can effectively avoid data distortion and improve the accuracy of prediction.
[0041] S2. Divide the climate throughout the year into seasons based on geographical location characteristics.
[0042] According to the months included in each season, the month is taken as the first variable to realize the association between the month and the heating steam extraction volume. Based on the annual climate change law and the characteristics of heating demand, the whole year is divided into four quarters, namely spring (March-June), summer (July-September), autumn (October-November) and winter (December-February of the following year). This seasonal division method is simple and easy, and can better reflect the impact of different seasons on the heating load. For example, the heating demand in winter and spring is usually higher, while in summer, due to the higher temperature, the heating demand is relatively low. Through this seasonal division, the seasonal variation law of the heating load can be better captured, laying the foundation for subsequent data analysis and model establishment.
[0043] S3. Select typical data for analysis.
[0044] The data of heating steam extraction in each section of the heating unit in the past two years were selected, and three typical working days were selected in the first ten days, the middle ten days and the last ten days of each month for data analysis. The selected typical working days should be representative and able to reflect the changes in the heating load of that month. These data include key indicators such as the unit's first extraction heat supply, hot reheat supply, cold reheat supply and middle exhaust heat supply. Through in-depth analysis of these data, the temporal variation pattern of the heating load can be revealed, and a heating steam extraction prediction model with "day" as the second variable can be established to achieve the correlation between "day" and heating steam extraction, providing basic data support for the establishment of a prediction model.
[0045] S4. Daily heating load formula fitting.
[0046] The average heating steam extraction data of the selected typical working days were fitted to obtain the relationship between the heating steam extraction volume of each section and time in the heating steam extraction volume prediction model, so as to predict the trend of the heating steam extraction volume of each section at a certain time in the future.
[0047] Based on data analysis, the daily heating load formula is fitted to establish a prediction model. The goal of the formula fitting is to establish a relationship between the average heating steam extraction volume data of a typical working day and time through mathematical methods, so as to accurately predict the changes in heating steam extraction volume at each moment in the future. The relationship between the heating steam extraction volume and time in each section of the prediction model is as follows:
[0048] F cr,i (x) = a i,4 x 4 +a i,3 x 3 +a i,2 x 2 +a i,1 x+A i ; (1)
[0049] F r,i (x) = b i,4 x4 +b i,3 x 3 +b i,2 x 2 +b i,1 x+B i ; (2)
[0050] F lr,i (x) = c i,4 x 4 +c i,3 x 3 +c i,2 x 2 +c i,1 x+C i ; (3)
[0051] F zr,i (x) = d i,4 x 4 +d i,3 x 3 +d i,2 x 2 +d i,1 x+D i ; (4)
[0052] G cro =F cr,i (x); (5)
[0053] G ro =F r,i (x); (6)
[0054] G lro =F lr,i (x); (7)
[0055] G zro =F zr,i (x); (8)
[0056] Where x is the current time, such as a certain time, h; a i,4 is the fourth-order coefficient of the single-pump heating formula in a certain season; a i,3 is the third-order coefficient of the single-pump heating formula in a certain season; a i,2 is the second-order coefficient of the single-draw heating formula in a certain season; a i,1 A is the first-order coefficient of the heating formula for a certain season; i is the constant of the heating formula for a certain season;
[0057] b i,4 is the fourth-order coefficient of the heat reheating formula for a certain season; b i,3 is the third-order coefficient of the heat reheating formula for a certain season; b i,2 is the second-order coefficient of the heat reheating formula for a certain season; bi,1 is the first-order coefficient of the heat reheating formula for a certain season; B i is the constant of the heat reheating formula for a certain season;
[0058] c i,4 is the fourth-order coefficient of the cooling reheating formula in a certain season; c i,3 is the third-order coefficient of the cooling reheating formula in a certain season; c i,2 is the second-order coefficient of the cooling and reheating formula in a certain season; c i,1 is the first-order coefficient of the cooling reheating formula in a certain season; C i is the constant of the cooling reheating formula in a certain season;
[0059] d i,4 is the fourth-order coefficient of the heating formula in a certain season; d i,3 is the third-order coefficient of the heating formula in a certain season; d i,2 is the second-order coefficient of the heating formula in a certain season; d i,1 is the first-order coefficient of the heating formula in a certain season; D i is the constant of the heating formula in a certain season;
[0060] i is the season, spring is 1: March to June; summer is 2: July to September; autumn is 3: October to November; winter is 4: December to February of the following year; G cr0 , G r0 , G lr0 , G zr0 They are the unit’s predicted average first extraction heat supply, hot resupply heat supply, cold resupply heat supply and middle exhaust heat supply, t / h.
[0061] S5. Monitor the current heating steam extraction data in real time and modify the formula to improve the accuracy of the prediction.
[0062] In order to improve the accuracy of prediction, the present invention introduces a real-time monitoring and model correction mechanism, which monitors the current heating steam extraction data in real time and uses the data to dynamically correct the obtained relationship. The corrected formula is as follows:
[0063]
[0064] Among them, Δx is the time difference between a future moment and the current moment, h; G cr1 , G r1 , G lr1 , G zr1 They are the unit’s predicted dynamic first-extraction heating supply, hot resupply heating supply, cold resupply heating supply and intermediate exhaust heating supply, t / h.
[0065] The following describes a method for real-time prediction of external heating load of a thermal power plant in conjunction with a specific embodiment.
[0066] Example 1
[0067] Step 1: Acquisition and screening of historical data;
[0068] The historical heating data is obtained from the operating system of a single heating unit in the power plant in the heating area. The data is processed in time series, and the heating amount is averaged per hour. The invalid data obtained is deleted, and the data with a heating amount less than zero is judged as invalid data and eliminated based on the positive value of the heating steam extraction volume.
[0069] Step 2: Seasonal division;
[0070] According to the geographical location characteristics of the region, the annual climate is divided into four seasons, namely spring (March-June), summer (July-September), autumn (October-November) and winter (December-February of the following year). According to the months included in each quarter, the month is used as the first variable to realize the association between the month and the heating steam extraction volume. This seasonal division method can better reflect the impact of different seasons on the heating load and better capture the seasonal changes in the heating load.
[0071] Step 3: Selection and analysis of typical data;
[0072] The data of heating steam extraction of each section of the heating unit in the past two years were selected, and three typical working days were selected in the first, middle and last ten days of each month for data analysis. A prediction model of heating steam extraction with "day" as the second variable was established to realize the association between "day" and heating steam extraction. The selected typical working days should be representative and able to reflect the changes in the heating load of the month. These data include key indicators of heating steam extraction of each section of the unit. Through in-depth analysis of these data, the law of heating load changes over time is revealed, providing basic data support for the establishment of a prediction model.
[0073] Step 4: daily heating load formula fitting;
[0074] The average heating steam extraction data of the selected typical working days were subjected to multivariate nonlinear fitting to obtain the relationship between the heating steam extraction volume and time in each section of the prediction model. The obtained relationship can better reflect the time variation law of the heating steam extraction volume and predict the trend of the heating steam extraction volume in each section at a certain time in the future.
[0075] Step 5: Real-time monitoring and model correction;
[0076] By real-time monitoring of the current heating steam extraction data and using the data to correct the obtained relationship model, the accuracy of the prediction can be improved to ensure that the prediction results are highly consistent with the actual situation.
[0077] Figure 2 and Figure 3The comparison curve between the predicted data of winter cold recovery and spring hot recovery heat supply and the actual heat supply of a thermal power unit obtained by using the real-time prediction method of heating load is shown.
[0078] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations 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 one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0079] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.
[0080] The above specific implementations of the present invention do not constitute a limitation on the protection scope of the present invention. Any other corresponding changes and modifications made based on the technical concept of the present invention should be included in the protection scope of the claims of the present invention.
Claims
1. A real-time prediction method for external heating load of a thermal power plant, characterized in that: The steps include: S1. Obtain and filter historical heating data; obtain historical heating data from the operating system of a single heating unit in a power plant in the heating area, perform time series processing, average the heating data per hour, and remove data with a heating data less than zero; S2. Divide the climate throughout the year into seasons according to the geographical location characteristics; according to the months included in each season, take the month as the first variable to achieve the association between the month and the heating steam extraction amount; S3. Select typical data for analysis; select the heating steam extraction data of each section of the heating unit in the past two years, select three typical working days in the first ten days, middle ten days and last ten days of each month for data analysis, establish a heating steam extraction prediction model with "day" as the second variable, and realize the correlation between "day" and heating steam extraction; S4. Fitting the daily heating load formula: Fit the average heating steam extraction data of the selected typical working day to obtain the relationship between the heating steam extraction volume and time in the heating steam extraction volume prediction model, so as to predict the trend of the heating steam extraction volume in each section at a certain time in the future; S5. Monitor the current heating steam extraction data in real time and modify the formula to improve the accuracy of the prediction.
2. The method for real-time prediction of external heating load of a thermal power plant according to claim 1, characterized in that: In step S4, the daily heating load formula is fitted as follows: F cr,i (x)=a i,4 x 4 +a i,3 x 3 +a i,2 x 2 +a i,1 x+A i ; F r,i (x)=b i,4 x 4 +b i,3 x 3 +b i,2 x 2 +b i,1 x+B i ; F lr,i (x)=c i,4 x 4 +c i,3 x 3 +c i,2 x 2 +c i,1 x+C i ; F zr,i (x)=d i,4 x 4 +d i,3 x 3 +d i,2 x 2 +d i,1 x+D i ; G cro =F cr,i (x); G ro =F r,i (x); G lro =F lr,i (x); G zro =F zr,i (x); Among them, x is the current time, such as a certain time, h; a i,4 is the fourth-order coefficient of the single-pump heating formula in a certain season; a i,3 is the third-order coefficient of the single-pump heating formula in a certain season; a i,2 is the second-order coefficient of the single-draw heating formula in a certain season; a i,1 A is the first-order coefficient of the heating formula for a certain season; i is the constant of the heating formula for a certain season; b i,4 is the fourth-order coefficient of the heat reheating formula for a certain season; b i,3 is the third-order coefficient of the heat reheating formula for a certain season; b i,2 is the second-order coefficient of the heat reheating formula for a certain season; b i,1 is the first-order coefficient of the heat reheating formula for a certain season; B i is the constant of the heat reheating formula for a certain season; c i,4 is the fourth-order coefficient of the cooling reheating formula in a certain season; c i,3 is the third-order coefficient of the cooling reheating formula in a certain season; c i,2 is the second-order coefficient of the cooling and reheating formula in a certain season; c i,1 is the first-order coefficient of the cooling reheating formula in a certain season; C i is the constant of the cooling reheating formula in a certain season; d i,4 is the fourth-order coefficient of the heating formula in a certain season; d i,3 is the third-order coefficient of the heating formula in a certain season; d i,2 is the second-order coefficient of the heating formula in a certain season; d i,1 is the first-order coefficient of the heating formula in a certain season; D i is the constant of the heating formula in a certain season; i is the season, spring is 1: March to June; summer is 2: July to September; autumn is 3: October to November; winter is 4: December to February of the following year; G cr0 , G r0 , G lr0 , G zr0 They are the unit’s predicted average first extraction heat supply, hot resupply heat supply, cold resupply heat supply and middle exhaust heat supply, t / h.
3. The real-time prediction method for external heating load of a thermal power plant according to claim 2 is characterized in that: In step S5, the current heating steam extraction data is monitored in real time and the formula is corrected. The corrected formula is as follows: Among them, Δx is the time difference between a future moment and the current moment, h; G cr1 , G r1 , G lr1 , G zr1 They are the unit’s predicted dynamic first-extraction heating supply, hot resupply heating supply, cold resupply heating supply and intermediate exhaust heating supply, t / h.
Citation Information
Patent Citations
Scheduling method of heat supply unit online load and system
CN103412526A
Method for predicting heating peak modulation capacity of heat supply unit
CN105202623A
AU2020104000A4
