Method for calculating power plant auxiliary power rate of a photo-thermal power station

Abstract

The present invention relates to the field of tower solar thermal power generation and discloses a method for calculating the plant power consumption rate of a generator set in a solar thermal power station, comprising the following steps: collecting meteorological data, real-time operating data of the generator set, and historical operating data of the generator set from the solar thermal power station; preprocessing the collected data; inputting the preprocessed real-time operating data of the generator set into a three-component dynamic decomposition model of the generator set's plant power consumption rate, and determining whether a trigger condition is met based on the meteorological data. When the trigger condition is met, dynamically calibrating the model coefficients using the historical operating data of the generator set, and finally outputting the plant power consumption rate. The method disclosed in the present invention decomposes the plant power consumption rate into three dynamic components: fixed loss, start-stop transient loss, and cooling loss. By capturing the transient process of the equipment through a dynamic model, the method overcomes the problem that traditional calculation methods cannot reflect the dynamic characteristics of the generator set, and improves the calculation accuracy of the plant power consumption rate of the generator set under complex working conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of tower type solar thermal power generation, and particularly relates to a method for calculating the auxiliary power rate of a generator unit of a solar thermal power station. BACKGROUND

[0002] Solar thermal power generation is gradually becoming an indispensable key component in new power systems due to its unique advantages such as large-scale energy storage, high stability of power generation, and strong peak regulation capability. However, the prediction accuracy of the power generation of a solar thermal power station still faces severe challenges, and it is difficult to meet the accuracy requirements of power grid dispatching for power generation plans. The auxiliary power rate of a generator unit, as a key parameter for evaluating the power generation efficiency and predicting the power of a solar thermal power station, directly affects the prediction results of the power generation.

[0003] In the traditional calculation model of the auxiliary power rate of a generator unit of a solar thermal power station, there are some defects that make it difficult to meet the actual needs:

[0004] 1. There is a significant negative correlation between the fixed loss of the generator unit and the output power of the generator - when the load rate decreases, the proportion of the fixed loss in the total power consumption will increase significantly. However, the traditional model does not consider this characteristic, and cannot accurately quantify the fixed loss under different loads.

[0005] 2. During the start-stop process of a solar thermal power station, there is significant transient energy consumption in the transition stage from static to stable operation of the generator unit equipment. The traditional model does not specifically model this type of transient process, resulting in a large calculation deviation under start-stop conditions.

[0006] 3. The cooling system energy consumption (cooling loss) is strongly correlated with the condenser temperature difference (the difference between the condenser saturation temperature and the circulating water outlet temperature). In the summer high temperature period, the increase in circulating water temperature will cause the cooling loss to account for more than 30%, while in the winter it is only 5% to 10%. The traditional model cannot capture this dynamic characteristic related to the ambient temperature, and the calculation accuracy decreases significantly under extreme temperature conditions.

[0007] The above defects cause the calculation error of the auxiliary power rate of the generator unit to increase significantly when the load fluctuates, the start-stop frequency is high, or the ambient temperature changes dramatically, which in turn causes the power generation prediction deviation to exceed the standard, and cannot meet the accurate dispatching requirements of the power grid for the power generation plan of the solar thermal power station. Therefore, there is an urgent need for a calculation method that can dynamically decompose the auxiliary power components of the generator unit and accurately quantify the energy consumption of each component, in order to improve the calculation accuracy of the auxiliary power rate of the generator unit of the solar thermal power station and provide reliable support for the accurate prediction of the power generation of the solar thermal power station. SUMMARY

[0008] To solve the above technical problems, the application provides a kind of photothermal power station generating unit plant power rate calculation method, realize the high accuracy calculation of photothermal power station generating unit plant power rate under complex working condition, significantly improve the photothermal power station power generation power prediction accuracy.

[0009] To achieve the above purpose, the technical scheme of the application is as follows:

[0010] A kind of photothermal power station generating unit plant power rate calculation method, comprising the following steps:

[0011] Step one, collect the meteorological data of photothermal power station, real-time operation data of generating unit and historical operation data of generating unit;

[0012] Step two, the data collected are pretreated;

[0013] Step three, the pretreated real-time operation data of generating unit are input into generating unit plant power rate three-component dynamic decomposition model, and whether the trigger condition is met is judged according to meteorological data, real-time operation data of generating unit and historical operation data of generating unit, when the trigger condition is met, the dynamic calibration of model coefficient is carried out using historical operation data of generating unit, and finally the plant power rate is output.

[0014] In the above scheme, in step one, the meteorological data includes the solar normal direct irradiance, temperature, wind speed, PM 10 Concentration, precipitation obtained from photothermal power station weather station;The real-time operation data of generating unit include the output power of generating unit, start-stop state mark, condenser end difference, circulating water temperature obtained from photothermal power station information management system;The historical operation data of generating unit include plant power rate measured value, output power of generating unit, start-stop state mark, condenser end difference, circulating water temperature under different working conditions for more than 3 years.

[0015] In the above scheme, in step three, the generating unit plant power rate three-component dynamic decomposition model is as follows:

[0016] ;

[0017] Wherein, is the final value of generating unit plant power rate; is the output power of generating unit; is start-stop state mark, 0 for stop, 1 for start; is daily cycle correction term, t is the time of day; is condenser end difference; is attenuation coefficient, expressed by piecewise function: when t<60 minutes, =0.15min -1 When t≥60 minutes, =0.05min-1 ; is a fixed loss coefficient, is a start-stop transient coefficient, is a cooling loss coefficient;

[0018] C is a coupling correction term, and the expression is:

[0019] .

[0020] In a further technical solution, the fixed loss coefficient , the start-stop transient coefficient , and the cooling loss coefficient are calibrated as follows:

[0021] First, a three-component dynamic decomposition preliminary model of the power plant auxiliary power rate is established:

[0022] ;

[0023] wherein, is the initial value of the power plant auxiliary power rate;

[0024] Then, a loss function is established:

[0025] ;

[0026] wherein, N is the number of effective samples;

[0027] is the measured value of the auxiliary power rate of the solar-thermal power plant of the i-th effective sample, which is obtained through the solar-thermal power plant information management system;

[0028] is the calculated value of the auxiliary power rate of the solar-thermal power plant of the i-th effective sample, which is calculated through the three-component dynamic decomposition preliminary model of the auxiliary power rate;

[0029] By minimizing the loss function , the preliminary model calculated value is closest to the measured value of the power plant, thereby obtaining the optimal coefficients , , .

[0030] In a further technical solution, the fixed loss coefficient is calibrated using the measured value of the auxiliary power rate of the power plant and the output power of the generator in the historical operation data of the power plant, and the following is obtained:

[0031] ;

[0032] wherein, is the number of fixed loss event effective samples; is the measured value of the power plant electricity rate of the generator set in the qth fixed loss event valid sample, is the generator output power of the qth fixed loss event valid sample, and t(q) is the time of day of the qth fixed loss event valid sample.

[0033] In a further technical solution, the measured value of the power plant electricity rate of the generator set in the historical operation data of the generator set is used to calibrate the start-stop transient coefficient :

[0034] ;

[0035] wherein, is the number of start-stop transient event valid samples, is the measured value of the power plant electricity rate of the generator set in the vth start-stop transient event valid sample, is the vth time point after the start of the start-stop process.

[0036] In a further technical solution, the measured value of the power plant electricity rate of the generator set in the historical operation data of the generator set is used to calibrate the start-stop transient coefficient :

[0037] ;

[0038] wherein, is the measured value of the power plant electricity rate of the generator set in the Wth interval; is the condenser terminal difference in the Wth interval, is the average value.

[0039] In a further technical solution, in step three, the trigger condition for calibration of the fixed loss coefficient is as follows:

[0040] When the calculation error of the power plant electricity rate under steady state conditions exceeds 5% for 3 consecutive days, or when the generator set is continuously operated for 30 days, the calibration of the fixed loss coefficient is performed; the steady state condition is and ; the calibration formula is as follows:

[0041] ;

[0042] wherein, n is the number of valid samples selected during calibration of the fixed loss coefficient ;

[0043] is the measured value of the power plant electricity rate of the generator set in the jth sample; is the generator output power of the jth sample; is the time of day of the jth sample;​ is a daily correction term, reflecting the periodic fluctuation of fixed loss with day and night alternation; V is the wind speed;

[0044] Rainfall working condition identification standard: wind speed ≥ 10 m / s and hourly rainfall ≥ 5 mm.

[0045] In the further technical solution, in step three, the start-stop transient coefficient The trigger condition for calibration is as follows:

[0046] When the condition after the cumulative occurrence of new 5 complete start-stop events is met, the automatic calibration of is carried out, and the complete start-stop event is the process from 1 to 0; the calibration formula is as follows:

[0047] ;

[0048] Among them, is the start-stop transient coefficient being used before calibration in the current step; is the duration of the start-stop process corresponding to the gth time point; m is the number of valid samples screened when the start-stop transient coefficient is calibrated.

[0049] In the further technical solution, in step three, the cooling loss coefficient The trigger condition for calibration is as follows:

[0050] When the condition of summer high temperature period is met or at the end of each quarter, the automatic calibration of is carried out; the summer high temperature period is that the average temperature for 7 consecutive days is ≥ 30℃; the calibration formula is as follows:

[0051] ;

[0052] Among them, is the cooling loss coefficient being used before calibration in the current step; is the number of years of power station operation, and the first operation ;

[0053] Dust determination condition: the direct normal solar irradiance decreases by ≥ 50% within 10 minutes and PM 10 ≥ 500 μg / m³.

[0054] Through the above technical solution, the calculation method of the auxiliary power consumption rate of the photothermal power station generator unit provided by the application has the following beneficial effects:

[0055] ​1. The application first proposes a "three-component dynamic decomposition model of auxiliary power rate", which decomposes the auxiliary power rate of a generator unit of a photothermal power station into three dynamic components of fixed loss, start-stop transient loss and cooling loss, and captures the transient process of the equipment through a dynamic model, thereby breaking through the problem that the traditional calculation method cannot reflect the dynamic characteristics of the generator unit.

[0056] 2. The application constructs an automatic dynamic coefficient calibration mechanism, designs a self-calibration strategy in multiple dimensions, and realizes dynamic calibration of the coefficients throughout the life cycle, thereby solving the problem of energy consumption characteristic drift caused by equipment aging and environmental changes of the generator unit, and maintaining the stability of the calculation accuracy of the auxiliary power rate.

[0057] 3. The application creates a coupling correction term formula to quantify the superposition effect of each component under extreme conditions (such as the coupling of fixed loss and cooling loss at high temperature), which can reduce the calculation error of the auxiliary power rate under normal and extreme conditions, and improve the calculation accuracy of the auxiliary power rate of the generator unit under complex conditions. BRIEF DESCRIPTION OF DRAWINGS

[0058] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed in the embodiment or the prior art description will be briefly introduced as follows.

[0059] Figure 1 A flowchart of a method for calculating the auxiliary power rate of a generator unit of a photothermal power station is disclosed in the embodiments of the application. DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the application will be described clearly and completely in combination with the drawings in the embodiments of the application.

[0061] The application provides a method for calculating the auxiliary power rate of a generator unit of a photothermal power station, as shown in Figure 1 The method comprises the following steps:

[0062] Step 1: Collecting meteorological data, real-time operation data of the generator unit and historical operation data of the generator unit of the photothermal power station.

[0063] The meteorological data includes data such as solar normal direct irradiance (DNI), temperature, wind speed, PM 10 concentration and precipitation, which are obtained through a meteorological station of the photothermal power station and used for dynamic calibration of key coefficients in the three-component dynamic decomposition model of the auxiliary power rate of the generator unit.

[0064] The real-time operation data of the generator unit includes data such as generator output power, start-stop state marker, condenser terminal difference and circulating water temperature, which can be obtained from an information management system of the photothermal power station and are input parameters of the three-component dynamic decomposition model of the auxiliary power rate of the generator unit.

[0065] The historical operating data of the generator set includes the measured values ​​of the generator set's plant power consumption rate under different operating conditions for more than three years, generator output power, start and stop status marks, condenser terminal difference, circulating water temperature and other data. These data can be obtained from the CSP power station information management system and used for the calibration and dynamic calibration of key coefficients in the three-component dynamic decomposition model of the generator set's plant power consumption rate.

[0066] Step 2: Preprocess the collected data.

[0067] 2.1 Time synchronization: Based on the power plant time, the timestamps of meteorological data and power plant data are unified to a 15-minute resolution to ensure temporal consistency and output a raw data set with temporal consistency.

[0068] 2.2 Outlier elimination: Use the 3σ criterion to detect and eliminate outliers in the meteorological data and generator equipment data collected in step 1. , where f is the total number of data collected; calculate its average value , calculate its standard deviation σ, if , then Determine as outliers and remove them.

[0069] 2.3 Missing value filling: For missing data, cubic spline interpolation is used to fill it. Assuming that the discrete data points are known, , construct the cubic spline function S(x) so that in each subinterval The above is a cubic polynomial, and the function value, first-order derivative, and second-order derivative at the node are continuous, so as to achieve accurate estimation of missing data points.

[0070] 2.4 Data normalization: Normalize meteorological and equipment data of different dimensions to the range [0, 1] using the minimum-maximum normalization method:

[0071] ;

[0072] in, is the original data, and are the minimum and maximum values ​​of the variable, respectively. The data are normalized.

[0073] After the above data preprocessing process, a normalized data set (meteorological data and generator equipment data after data quality processing) is generated.

[0074] Step three, input the pre-processed real-time operation data of the generator set into the three-component dynamic decomposition model of the generator set auxiliary power rate, and determine whether the trigger condition is met according to the meteorological data, real-time operation data of the generator set and historical operation data of the generator set, when the trigger condition is met, the model coefficient is dynamically calibrated by using the historical operation data of the generator set, and finally the auxiliary power rate is output.

[0075] In order to solve the defect that the traditional fixed proportion model ignores the dynamic characteristics, the real-time data of the steam turbine equipment collected is used, three components are added, including a fixed loss term negatively related to the load rate, a start-stop transient term related to the time decay characteristic, and a cooling loss term linearly related to the condenser terminal difference, the coupling effect of each component in actual operation (such as the superposition of fixed loss and cooling loss under extreme high temperature), a coupling correction term is introduced, a three-component dynamic decomposition model of the auxiliary power rate of the generator set is constructed, by quantifying the energy consumption from different sources, the high-precision calculation of the auxiliary power rate under complex working conditions is realized, and more accurate calculation basis of the auxiliary power rate of the generator set is provided for the power prediction of the solar thermal power generation.

[0076] 3.1 Construction of three-component dynamic decomposition model of auxiliary power rate of generator set

[0077] The three-component dynamic decomposition model of the auxiliary power rate of the generator set is as follows:

[0078] ;

[0079] Among them, is the final value of the auxiliary power rate of the generator set;

[0080] is the output power of the generator, unit: MW, obtained through the solar thermal power station information management system;

[0081] is the start-stop state flag, 0 for stop and 1 for start, obtained through the solar thermal power station information management system;

[0082] is the daily cycle correction term, t is the time of the day (0-24h), the coefficient 0.02 in the formula is determined based on the fitting of 3-year historical operation data of the solar thermal power station, and represents that the upper limit of the daily cycle fluctuation amplitude is ±2%;

[0083] is the condenser terminal difference, unit: ℃, obtained by subtracting the circulating water outlet temperature from the condenser saturation temperature, obtained through the solar thermal power station information management system;

[0084] is the decay coefficient, represented by a segmented function: when t<60 minutes, =0.15min -1 , when t≥60 minutes, = 0.05 min -1 ;

[0085] C is the coupling correction term, expressed as:

[0086] ;

[0087] When the condenser approach exceeds 15℃ (extreme high temperature operating condition), the decrease in cooling system efficiency will cause a significant increase in cooling loss, and the high temperature environment will cause the auxiliary equipment (such as pumps and cooling fans) in the fixed loss to consume additional energy due to the increased heat dissipation demand. At this time, the fixed loss and the cooling loss are not independent, but form a superposition effect. By correlating the core parameters of the cooling loss with the fixed loss term, the superposition effect is quantified, breaking the limitation of "fixed loss and cooling loss independent calculation", making the generator set plant power consumption rate calculation under extreme high temperature operating condition more close to the actual energy consumption characteristics, and thus improving the calculation accuracy;

[0088] is the fixed loss coefficient, is the start-stop transient coefficient, is the cooling loss coefficient, and by screening the sample data corresponding to the operating condition (fixed loss screening steady-state data, start-stop transient loss screening start-stop process data, cooling loss screening high temperature day data), the regression algorithm is used for calibration in turn.

[0089] 3.2 Model coefficient calibration

[0090] The specific calibration method is as follows:

[0091] First, a three-component dynamic decomposition preliminary model of the generator set plant power consumption rate is established:

[0092] ;

[0093] Where, is the initial value of the generator set plant power consumption rate.

[0094] Collect complete historical operation data of the solar thermal power plant, including measured values of the generator set plant power consumption rate under different operating conditions for more than 3 years, generator output power, start-stop state marker, condenser approach, circulating water temperature, etc., which can be obtained from the solar thermal power plant information management system; use the regression algorithm to fit the parameters, establish the loss function, quantify the deviation between the model predicted value and the measured value, provide a clear target for parameter optimization, and ensure that the parameters obtained by fitting can accurately reflect the true characteristics of each component of the plant power consumption rate, and thus improve the calculation accuracy of the plant power consumption rate model.

[0095] Establish the loss function :

[0096] ;

[0097] wherein N is the number of valid samples, which can be the number of valid samples of fixed loss events, the number of valid samples of start-stop transient events, or the number of valid samples of cooling loss events;

[0098] is the measured plant power consumption rate of the i-th valid sample, which is obtained through the information management system of the solar-thermal power station;

[0099] is the calculated plant power consumption rate of the i-th valid sample, which is obtained through the three-component dynamic decomposition preliminary model;

[0100] The optimal coefficient is obtained by minimizing the deviation square sum using the loss function, so that the model calculation value is closest to the measured value of the power station 、 , .

[0101] (1) Calibration of the fixed loss coefficient

[0102] Select the data set: take the steady-state data of and , including the measured plant power consumption rate of the generator set in the historical operation data of the generator set and the generator output power, at this time, the start-stop loss is 0 and the cooling loss is in a normal state. The data selection needs to ensure that the number of valid samples is not less than 3000, covering the full range of generator output power (10-100 MW) and complete daily cycle (0-24 h). In order to ensure the fitting accuracy of the function relationship and the daily cycle correction term, and reduce the calibration error of .

[0103] At this time, the calculated plant power consumption rate of the generator set in the preliminary model is only composed of the fixed loss term, and the fitting equation is:

[0104] ;

[0105] wherein is the plant power consumption rate under the fixed loss condition, which is obtained through the historical operation data of the solar-thermal power station;

[0106] Substitute the fitting equation of the fixed loss coefficient into the above loss function formula, replace in the loss function formula with to calculate, convert the loss function into a function containing only , and use the least square method to obtain . The specific process is as follows:

[0107] ; ​

[0108] where, is the calculated value of auxiliary power rate of the qth fixed loss event valid sample, is the generator output power of the qth fixed loss event valid sample, is the time of day of the qth fixed loss event valid sample.

[0109] Substitute the above fitting equation into the loss function formula. At this time, the loss function is converted into a function containing only :

[0110] ;

[0111] where, is the number of fixed loss event valid samples;

[0112] is the measured value of auxiliary power rate of the qth fixed loss event valid sample.

[0113] In order to minimize the loss function (that is, the deviation between the calculated value and the measured value is minimized), according to the least squares method, the is solved with respect to and the derivative is set to 0:

[0114] ;

[0115] The formula is arranged as:

[0116] ;

[0117] Solving gives :

[0118] ;

[0119] where, is the generator output power of the qth fixed loss event valid sample, which is obtained through the information management system of the photovoltaic thermal power station.

[0120] By minimizing the deviation between the predicted value and the measured value of the fixed loss, it is ensured that accurately reflects the negative correlation between the load rate and the fixed loss.

[0121] (2) Calibration of start-stop transient coefficient :

[0122] Select the data set: take the start-stop process data of , including the measured value of auxiliary power rate of the generator set in the historical operation data of the generator set. Data selection needs to ensure that a single start-stop process (the start-stop process of the qth fixed loss event valid sample) is included in the data set. The sample size of the preliminary model is not less than 60 (covering t=0 to t=120 minutes, with a resolution of 1 minute), and the cumulative effective start-stop event sample size is not less than 5 times, so as to accurately capture the exponential decay characteristics and reduce random interference, and reduce the calibration error of the preliminary model.

[0123] At this time, the generator set auxiliary power consumption calculated value in the preliminary model is only composed of the start-stop transient term, and the fitting equation is:

[0124] ;

[0125] wherein, is the auxiliary power consumption under the start-stop transient event;

[0126] The start-stop transient coefficient fitting equation is substituted into the above loss function formula, and the start-stop transient coefficient is replaced with the loss function formula The loss function is converted into a function containing only , and the loss function is minimized by logarithmic linear regression, and is obtained. The specific process is:

[0127] ;

[0128] wherein, is the auxiliary power consumption model calculated value of the vth start-stop transient event effective sample, is the vth time point (unit: minute) after the start of the start-stop process.

[0129] The above fitting equation is substituted into the formula of the loss function . At this time, the loss function is converted into a function containing only :

[0130] ;

[0131] wherein, is the number of effective samples of the start-stop transient event;

[0132] is the measured value of the auxiliary power consumption of the photothermal power station generator set of the vth start-stop transient event effective sample.

[0133] After linearization of the nonlinear model by logarithmic conversion, the least square method is used to solve the optimal value of ln :

[0134] ;

[0135] can be solved: :

[0136] ​​

[0137] This process ensures Accurately reflects the exponential decay characteristics of start-stop losses over time.

[0138] (3) Cooling loss coefficient Calibration:

[0139] Filter the data set: Take the data from the period of high-temperature summer days when the circulating water temperature rises by more than 10°C, including the condenser terminal difference and the measured value of the generator set power consumption rate in the historical operation data of the generator set; the circulating water temperature rise is the largest in summer, and the cooling loss accounts for more than 30% of the power consumption rate (only 5%-10% in winter), so the data during this period can be captured more accurately. The data screening should ensure that the total effective sample size is not less than 2000 (covering 15 full days with circulating water temperature rise > 10℃ in summer high temperature days), and the data should be filtered according to the condenser end difference. In every 5℃ interval, the sample size of each interval is not less than 500 to ensure and The linear relationship fitting accuracy.

[0140] At this point, the calculated value of the generator set power consumption rate in the preliminary model is only composed of the cooling loss term, and the fitting equation is:

[0141] ;

[0142] in, is the power consumption rate of the CSP plant under cooling conditions, obtained from the historical operation data of the CSP plant;

[0143] Substitute the cooling loss coefficient fitting equation into the above loss function formula and use Replace the loss function formula Calculate and transform the loss function into By using the temperature difference bucket averaging method to minimize the loss function, we can get Specific process:

[0144] Substitute the above fitting equation into the loss function At this point, the loss function is transformed into a formula containing only Function:

[0145] ;

[0146] Among them, N cool is the number of effective samples of cooling loss events; is the condenser end difference of the effective sample of the sth cooling loss event; is the measured value of the power consumption rate of the CSP power station generator set of the sth effective sample of cooling loss event.

[0147] In order to minimize the loss function, the temperature difference bucket averaging method is adopted: the effective sample N of cooling loss events is cool According to the condenser end difference Divide into W intervals and calculate the average value of the condenser end difference in each interval (That is, all The arithmetic mean of the power consumption rate of the plant (That is, all At this time, the loss function of each interval can be simplified to the "measured mean and predicted value ( ) is calculated by making the sum of squared deviations of As close as possible , which can minimize the loss of this interval and thus minimize the loss function.

[0148] Final value:

[0149] ;

[0150] in, is the measured value of the power consumption rate of the generator set in the Wth interval; is the condenser terminal difference in the Wth interval, This process ensures that Accurately reflects the linear correlation between cooling loss and condenser end difference.

[0151] What you can get here , , Fixed values ​​for the three coefficients.

[0152] 3.3 Dynamic Self-calibration of Model Coefficients

[0153] Due to factors such as equipment aging (such as pump efficiency degradation) and long-term environmental changes (such as condenser scaling leading to decreased heat exchange efficiency), energy consumption characteristics will drift. After one year of operation, the calculation error will increase by 10%-15%, and long-term accuracy cannot be guaranteed. To solve the problem of accuracy drift caused by long-term fixed coefficients, a dynamic coefficient self-calibration mechanism is introduced into the three-component dynamic decomposition model, and the sliding window is used to iteratively calibrate the coefficients. , , The specific process is as follows:

[0154] (1) Fixed loss coefficient Self-calibration

[0155] Trigger condition: When the calculation error of the power consumption rate under steady-state conditions exceeds 5% for 3 consecutive days, or when the system has been running for 30 consecutive days, calibration; the steady-state working condition is and ;

[0156] Data screening: select the latest steady-state data (sample size not less than 1000) within 30 days before the calibration time that meet and ; the data includes the measured value of the power plant auxiliary power consumption rate, the generator output power, the wind speed, and the precipitation; n is the fixed loss coefficient The number of effective samples screened at the calibration time.

[0157] Calibration method: through the "sliding window calibration" mechanism, the new fixed loss coefficient is recalculated based on the newly screened data, covering the historical fixed value, realizing dynamic calibration, dynamic optimization for the energy consumption characteristics drift caused by equipment aging and environmental changes, and adapting to the change of the fixed loss characteristics caused by equipment aging (such as the increase of the basic energy consumption of pumps). The formula of the new fixed loss coefficient obtained after self-calibration is:

[0158] ;

[0159] Among them, is the measured value of the power plant auxiliary power consumption rate of the jth sample, obtained through the information management system of the solar thermal power station;

[0160] is the generator output power (unit: MW) of the jth sample;

[0161] t(j) is the time of the jth sample (unit: h, value 0~24), used to calculate the daily period correction term;

[0162] is the daily period correction term, reflecting the periodic fluctuation of the fixed loss with the alternation of day and night;

[0163] V is the wind speed, which can be obtained through the weather station;

[0164] Precipitation working condition recognition standard: wind speed ≥ 10 m / s and hourly rainfall ≥ 5 mm, weather data can be obtained through the weather station of the solar thermal power station. By distinguishing the working condition type, the influence of extreme weather such as rainfall on the fixed loss can be corrected, so that both the long-term aging trend of the equipment and the short-term environmental mutation can be responded to, and the precision of the model under complex working conditions can be improved.

[0165] This step upgrades the fixed loss coefficient from a "static initial value" to a "dynamic tracking value" through improvements in four aspects: dynamic triggering, recent data, operating condition segmentation, and aging adaptation. This significantly improves the calculation accuracy of the coefficient throughout the equipment's life cycle, making it particularly suitable for complex scenarios such as equipment aging and environmental fluctuations during the long-term operation of solar thermal power stations.

[0166] (2) Start-stop transient coefficient Self-calibration

[0167] Trigger condition: When the conditions of 5 new complete start-stop events are met, Automatic calibration, the complete start and stop events are The process of changing from 1 to 0;

[0168] Data filtering: Select the last 5 start and stop processes Real-time data of the start and stop (the sample size is not less than 60 for each start and stop, covering t = 0 to t = 120 minutes); the data includes the measured value of the power consumption rate of the generator set; m is the start and stop transient coefficient The effective number of samples screened during calibration.

[0169] Calibration method: weighted fusion, recalculation based on new data , retain the three-component dynamic decomposition model calculated by the power consumption rate of the power generation unit 70% of the coefficient (used to smooth the sudden change in characteristics caused by equipment wear and improve calibration stability) and the newly calculated value 30% (used to quickly respond to changes in the recent start-stop characteristics of the equipment) are weighted and integrated to compensate for the transient energy consumption characteristics changes caused by wear of the equipment start-stop mechanism. The new start-stop transient coefficient obtained after self-calibration The formula is:

[0170] ;

[0171] in, This is the start-stop transient coefficient currently in use before the current calibration step. During the first calibration, its value is the start-stop transient coefficient calculated using the preliminary model of the three-component dynamic decomposition of the generator set's plant power consumption rate. During subsequent calibrations, its value is the coefficient obtained after the previous dynamic calibration. This allows for iterative calibration of the coefficient and preserves the transient energy consumption characteristics accumulated during long-term equipment operation. is the duration of the start-stop process corresponding to the g-th time point, in minutes.

[0172] This step upgrades the start-stop transient coefficient from a "static initial coefficient" to a "dynamic tracking coefficient" through three major improvements: event-driven triggering, recent data screening, and weighted fusion calculation. It effectively compensates for the transient energy consumption characteristic drift caused by wear of the equipment's start-stop mechanism, and significantly improves the calculation accuracy of the start-stop transient coefficient throughout its life cycle.

[0173] (3) Cooling loss coefficient Self-calibration

[0174] Trigger condition: When the summer high temperature period condition is met or at the end of each quarter, automatic calibration is carried out of the summer high temperature period is 7 consecutive days with an average temperature ≥ 30℃;

[0175] Data screening: Select the latest data of the summer high temperature day (circulating water temperature rise > 10℃) in the current quarter, and divide the interval according to The data includes the condenser terminal difference, the measured value of the generator set auxiliary power rate, the solar direct normal irradiance, and the PM 10 concentration.

[0176] Calibration method: Add an aging compensation term to the original calculation basis to correct the cooling system efficiency decay problem caused by condenser fouling and pipeline aging, and consider the impact of dust conditions on the condenser heat transfer surface. The formula of the new cooling loss coefficient after self-calibration is:

[0177] ;

[0178] Where, is the cooling loss coefficient being used before calibration at the current step: when calibrated for the first time, its value is the cooling loss coefficient calculated by the three-component dynamic decomposition preliminary model of the generator set auxiliary power rate; when calibrated subsequently, its value is the is the number of years since the power station was put into operation, which is incremented by 1 every full year since the first operation to quantify the impact of long-term aging of equipment on the cooling system.

[0179] Dust determination condition: 10-minute sudden drop of solar direct normal irradiance ≥ 50% and PM 10 ≥ 500 μg / m³, (dust causes condenser heat transfer efficiency to drop by 20%~30%) At this time, both equipment aging compensation and heat transfer efficiency drop compensation caused by dust need to be considered. While the conventional working condition only needs to consider equipment aging compensation. By distinguishing the working condition type, the fine dynamic calibration of the cooling loss coefficient is realized, which not only covers the gradual impact of long-term aging of equipment, but also responds to the short-term impact of sudden environmental factors such as dust, ensuring the calculation accuracy throughout the life cycle.

[0180] This step upgrades the cooling loss coefficient from "static initial coefficient" to "full life cycle dynamic coefficient" through three improvements of multi-dimensional triggering, recent data screening, and working condition subdivision compensation, effectively compensating for the cooling energy consumption characteristics drift caused by equipment aging, environmental mutations, etc., and significantly improving the calculation accuracy of the cooling loss coefficient under complex working conditions.

[0181] The dynamic calibrated , , is brought into the three-component dynamic decomposition model of the generator set auxiliary power rate, i.e., the generator set auxiliary power rate .

[0182] Specific embodiments: Taking the 3-year operation data of a 50MW tower type photothermal power station in Qinghai, China from January 1, 2022 to December 31, 2024 as an example, the specific processing process is as follows:

[0183] Data collection:

[0184] Meteorological data: Collect solar direct normal irradiance, temperature, wind speed, PM 10 concentration, and precipitation through the weather station. The data collection frequency of the weather station is 15 minutes;

[0185] Real-time operation data of the generator set: Collect the generator output power, start-stop state marker, condenser terminal difference, and circulating water temperature through the photothermal power station information management system. The data collection frequency is 15 minutes;

[0186] Historical operation data of the generator set: Collect the measured value of the generator set auxiliary power rate, generator output power, start-stop state marker, condenser terminal difference, and circulating water temperature through the photothermal power station information management system. The data collection frequency is 15 minutes.

[0187] Through data preprocessing, the collected data is subjected to outlier rejection, missing value filling, and normalization processing to generate a normalized data set.

[0188] Calibrate the model coefficient with the preprocessed data:

[0189] Fixed loss coefficient : Select 4200 steady-state samples from 2022-2024 data that meet the conditions of and , and calculate according to the formula: = 0.85;

[0190] Start-stop transient coefficient : Select 30 complete start-stop events (each ≥60 samples), and obtain: = 1.15;

[0191] Cooling loss coefficient : per Divide each 5℃ interval, each interval sample size ≥800, take the maximum ratio: =0.028.

[0192] Coefficient dynamic calibration:

[0193] Calibration: select the occurrence of 3 consecutive days of steady-state error more than 5% (average 6.2%), trigger calibration. Select the last 30 days of steady-state samples 1200, under normal working conditions: =0.84 (2.4% more than the initial value correction);

[0194] Calibration: after 5 times of new start-stop events (4-8 times), the old coefficient (1.15×70%) and the new calculated value (1.20×30%) are fused: =1.165;

[0195] Calibration: select the conditions that meet the summer high temperature period, and appear dust working conditions, calibration: =0.032 (14.3% more than the initial value correction).

[0196] Generator set auxiliary power rate calculation:

[0197] Select the measured time period 6 June 14:00 dust high temperature working condition:

[0198] =50MW, , =25℃, =14h, =0.05min -1 , =1+0.005×(25-15)=1.05.

[0199]

[0200] The calculation error of the method of the application and the traditional method (taking constant 5%) is compared, and the results are shown in Table 1.

[0201] Table 1

[0202]

[0203] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the appended claims are intended to cover all such modifications that do not depart from the true spirit and scope of the application. Therefore, the application is not limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for calculating the power consumption rate of a solar thermal power station generator set, characterized in that: The steps include: Step 1: Collect meteorological data of the CSP plant, real-time operating data of the generator sets, and historical operating data of the generator sets; Step 2: preprocess the collected data; Step 3: Input the pre-processed real-time operating data of the generator set into the three-component dynamic decomposition model of the generator set's power consumption rate, and determine whether the trigger conditions are met based on the meteorological data, the real-time operating data of the generator set, and the historical operating data of the generator set. When the trigger conditions are met, the historical operating data of the generator set is used to dynamically calibrate the model coefficients, and finally output the power consumption rate; In step 3, the three-component dynamic decomposition model of the power consumption rate of the generator set is as follows: ； in, is the final value of the power consumption rate of the generator set; is the generator output power; It is the start / stop status flag, 0 for stop and 1 for start; is the daily cycle correction term, t is the time of the day; is the condenser end difference; is the attenuation coefficient, which is expressed by a piecewise function: when t < 60 minutes, =0.15min -1 , when t≥60 minutes, =0.05min -1 ; is a fixed loss coefficient, is the start-stop transient coefficient, is the cooling loss coefficient; C is the coupling correction term, which is expressed as: ； In step 3, the fixed loss coefficient The trigger conditions for calibration are as follows: When the calculation error of the plant power consumption rate under steady-state conditions exceeds 5% for three consecutive days, or when the plant power consumption rate has been running for 30 consecutive days, Calibration of the steady-state condition is and ;The calibration formula is as follows: ； Where n is the fixed loss coefficient The number of valid samples screened during calibration; is the measured value of the power consumption rate of the generator set of the jth sample; is the generator output power of the jth sample; is the time of day of the jth sample; is the diurnal correction term, reflecting the periodic fluctuation of fixed losses with the alternation of day and night; V is the wind speed; Rainfall condition identification standard: wind speed ≥ 10m / s and hourly rainfall ≥ 5mm; In step 3, the start-stop transient coefficient The trigger conditions for calibration are as follows: When the conditions are met after a total of 5 complete start-stop events, Automatic calibration, the complete start and stop events are The process of changing from 1 to 0; the calibration formula is as follows: ； in, is the start-stop transient coefficient being used before calibration of the current step; is the duration of the start-stop process corresponding to the g-th time point; m is the start-stop transient coefficient The number of valid samples screened during calibration; In step 3, the cooling loss coefficient The trigger conditions for calibration are as follows: When the summer high temperature period conditions are met or at the end of each quarter, Automatic calibration; the high temperature period in summer is 7 consecutive days with an average temperature of ≥30℃; the calibration formula is as follows: ； in, is the cooling loss coefficient being used before calibration in the current step; The number of years the power station has been put into operation, first operation ; Dust determination conditions: direct solar irradiance drops by ≥50% within 10 minutes and PM 10 ≥500μg / m³.

2. The method for calculating the power consumption rate of a CSP power generation unit according to claim 1, characterized in that: In step 1, the meteorological data includes the solar normal direct irradiance, temperature, wind speed, PM 10 concentration, precipitation; the real-time operating data of the generator set includes the generator output power, start-stop status mark, condenser terminal difference, and circulating water temperature obtained from the information management system of the solar thermal power station; the historical operating data of the generator set includes the measured value of the plant power rate, generator output power, start-stop status mark, condenser terminal difference, and circulating water temperature under different operating conditions for more than 3 years.

3. The method for calculating the power consumption rate of a CSP power generation unit according to claim 1, characterized in that: Fixed loss factor , start-stop transient coefficient , cooling loss coefficient The calibration process is as follows: First, a preliminary model for the dynamic decomposition of the three components of the power consumption rate of the generator sets is established: ； in, is the initial value of the power consumption rate of the generator set; Then, establish the loss function : ； Where N is the number of valid samples; is the actual value of the CSP power consumption rate of the i-th valid sample, obtained from the CSP information management system; is the calculated value of the CSP power consumption rate of the i-th valid sample, which is calculated by the preliminary model of the three-component dynamic decomposition of the power consumption rate; By minimizing the loss function , so that the calculated value of the preliminary model is closest to the measured value of the power station, thus obtaining the optimal coefficient 、 , .

4. The method for calculating the power consumption rate of a CSP power generation unit according to claim 3, characterized in that: The fixed loss coefficient is calculated by using the measured value of the power consumption rate of the generator set and the output power of the generator in the historical operation data of the generator set. The calibration of , we get: ； in, is the number of effective samples of fixed loss events; is the measured value of the power consumption rate of the generator set in the effective sample of the qth fixed loss event, is the generator output power of the qth fixed loss event valid sample, and t(q) is the time of day when the qth fixed loss event valid sample is obtained.

5. The method for calculating the power consumption rate of a CSP power generation unit according to claim 3, characterized in that: The start-stop transient coefficient is calculated by using the measured value of the generator set power consumption rate in the historical operation data of the generator set. Calibration: ； in, is the number of valid samples of start-stop transient events, is the measured value of the power consumption rate of the generator set of the vth effective sample of the start-stop transient event, It is the vth time point after the start-stop process begins.

6. The method for calculating the power consumption rate of a CSP power generation unit according to claim 3, characterized in that: The cooling loss coefficient is calculated by using the condenser terminal difference and the measured value of the power consumption rate of the generator set in the historical operation data of the generator set. Calibration: ； in, is the measured value of the power consumption rate of the generator set in the Wth interval; is the condenser terminal difference in the Wth interval, To take the average value.

Images