Photovoltaic grid-connected system adequacy assessment method considering seasonal and day and night characteristics
By constructing an equivalent photovoltaic unit model and irradiance simulation model, the problem that the existing technology is difficult to reflect the irradiance change characteristics is solved, the accurate maturity evaluation and optimization of the photovoltaic grid-connected system is realized, and the operating efficiency of the system is improved. Economical.
Patent Information
- Application Number
- CN202410915142.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-09
- Publication Date
- 2024-11-26
AI Technical Summary
The irradiance modeling of existing photovoltaic power generation systems is difficult to fully reflect the impact of different weather, different seasons and day and night changes, resulting in deviations in the evaluation and planning of photovoltaic power generation systems, affecting its operating efficiency and economics.
Using the maturity evaluation method of photovoltaic grid-connected systems that take into account quarterly and day-night characteristics, the irradiance samples were collected and fitted by constructing an equivalent photovoltaic unit model, and simulated irradiance samples were generated using Monte Carlo simulation method, combined with K-means Clustering and cune-index smoothing methods, an irradiance simulation model was established and the maturity index of each generator set was calculated.
The accurate assessment of the maturity of the photovoltaic grid-connected system has been achieved, the accuracy and reliability of the evaluation results have been improved, the design and operation strategies of the photovoltaic power generation system have been optimized, and economic benefits have been improved.
Smart Images

Figure CN119026314A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of reliability assessment of photovoltaic grid-connected systems in power systems, and in particular relates to a method for assessing the adequacy of photovoltaic grid-connected systems taking into account seasonal and diurnal characteristics. Background Art
[0002] With the rapid growth of photovoltaic power generation, many technical challenges have also been brought about. One of them is that photovoltaic power output has great uncertainty and volatility, which poses a severe test to the safety and stability of the power system. As one of the key factors in photovoltaic power generation, irradiance has complex and changeable characteristics. As the sun rises and falls, irradiance shows obvious day and night changes, and there are significant differences in irradiance in different regions, seasons and even weather conditions. Therefore, accurately grasping the periodic characteristics of irradiance is crucial for the stable operation of photovoltaic power generation systems.
[0003] At present, irradiance modeling mainly focuses on two aspects: one is to use a probabilistic model to model irradiance, usually assuming that irradiance follows a Beta distribution and fitting measured data through appropriate parameter selection; the other is to consider the temporal and spatial correlation and use the irradiance relationship between adjacent regions and continuous moments to simulate its change process. However, existing modeling methods often fail to fully reflect the impact of actual irradiance in different weather, different seasons, and day-night changes, which may lead to deviations in the evaluation and planning of photovoltaic power generation systems, thereby affecting their operating efficiency and economy. Summary of the invention
[0004] In order to solve the above technical problems, the present invention provides a photovoltaic grid-connected system adequacy assessment method taking into account seasonal and diurnal characteristics, which can accurately assess the adequacy of wind and solar grid-connected systems. The application of this method in large-scale photovoltaic grid-connected adequacy assessment is of vital importance to the reliable basis for the planning and operation of photovoltaic grid-connected systems.
[0005] The technical solution adopted by the present invention is:
[0006] The reliability assessment method of photovoltaic power generation capacity taking into account the seasonal and diurnal characteristics of irradiance includes the following steps:
[0007] Step 1: Construct an equivalent photovoltaic unit model according to the series-parallel structure of photovoltaic panels;
[0008] Step 2: Collect the long-term measured irradiance samples of the meteorological observation station and determine the number of measured irradiance samples; according to the accuracy requirements of the margin assessment, set the number of simulated irradiance samples to be generated;
[0009] Step 3: Sample the number of continuous steps of each generator set in operation and shutdown state;
[0010] Step 4: Simulate the irradiance and generate simulated irradiance samples that take into account seasonal and day-night characteristics;
[0011] Step 5: Calculate the margin index of each generator set by using the simulated irradiance sample in step S4 in combination with the Monte Carlo simulation method;
[0012] Step 6: Repeat steps 3 to 5. When the stop condition is met, the loop ends and each margin index is output.
[0013] In step 1, the photovoltaic panel series-parallel structure rule is that when any photovoltaic panel in the series photovoltaic panels fails, the photovoltaic panel string where it is located will be shut down, but it will not affect the parallel photovoltaic panel string.
[0014] In step 1, the construction of the equivalent photovoltaic unit model is as follows:
[0015] The photovoltaic string obtained by connecting g photovoltaic panels in series is regarded as an equivalent photovoltaic unit, and the equivalent failure rate λ is calculated according to the failure rate and repair time of each photovoltaic panel. PV,s and the equivalent repair time γ PV,s They are:
[0016]
[0017] Among them, λ PV and γ PV are the failure rate and repair time respectively; g represents the number of photovoltaic panels connected in series; λ PV,s represents the equivalent failure rate; γ PV,s The equivalent repair time represented by λ PV,i represents the equivalent failure rate; γ PV,i Represents the equivalent repair time of the i-th photovoltaic panel.
[0018] In step 2, the sample size N of the measured irradiance samples is determined by the source meteorological observation station, that is, the number of available samples provided by the meteorological observation station is the sample size N of the irradiance samples;
[0019] According to the evaluation accuracy, the amount of simulated irradiance samples M to be generated is set, that is, the higher the evaluation accuracy, the more samples are generated;
[0020] It is also assumed that all generator sets (conventional units and equivalent photovoltaic units) are in normal operation at the initial simulation time. This assumption is more in line with the actual operation situation and is easy to calculate.
[0021] In step 3, when the nth generator set is in operation, the number of time steps TTR for the generator set to be in operation is calculated according to the failure rate λ of the generator set. nSampling is performed as shown below:
[0022]
[0023] Where: n represents the failure rate of the nth generator set; r represents a uniformly distributed random number in the range [0,1]; Indicates that the data is rounded up, and rounding down can also be used for calculation here;
[0024] When the nth generator set is in a shutdown state, the number of time steps TTF for the generator set in the shutdown state is calculated according to the repair rate μ of the generator set. n Sampling is performed as shown below:
[0025]
[0026] Where: represents the repair rate of the nth generator set; r represents a uniformly distributed random number in the range [0,1]; Indicates that the data is rounded up, and rounding down can also be used for calculation here;
[0027] The step 4 comprises the following steps:
[0028] S41: Fit N measured irradiance samples to obtain parameters a0, b0, c0, w0, and determine the quarterly period component Y t ;
[0029] S42: Divide the measured irradiance sample by the quarterly period component to obtain the annual residual component H t , that is, H t =R t / Y t , R t represents the irradiance at time t; Y t represents the quarterly period component at time t;
[0030] S43: Calculate the extraterrestrial radiation and the correlation coefficient between the extraterrestrial radiation and the surface radiation every day. The calculation formula is as follows:
[0031]
[0032] Where: C n represents the correlation coefficient of irradiance on the nth day; G i represents the i-th cluster center, i = 1, 2, 3, 4; M represents the irradiance days.
[0033] Use K-means clustering to cluster the remaining components H tThe day is roughly divided into four categories: A-extremely strong irradiation day, B-strong irradiation day, C-moderate irradiation day, and D-weak irradiation day; the daily irradiation start and end times are counted and classified, and finally a 4k type irradiance sequence is obtained. On extremely strong irradiation days, there are fewer obstructions in the sky, and the irradiance outside the earth and on the surface shows a similar change pattern. The correlation between the irradiance outside the earth and the surface irradiance is strong. On weak irradiation days, there are more interference factors in the surface irradiance, which is more different from the irradiance outside the earth and has a weaker correlation.
[0034] S44: Fitting to obtain 4k day and night characteristic components D1, D2, ..., D 4k ,D1,D2,...,D 4k represents the 1st, 2nd, …, 4kth day and night characteristic components, each of which contains values at several moments, namely D m =(D1,D2,...,D n ); D1,D2,...,D n represents the 1st, 2nd,…, 4kth day and night characteristic components;
[0035] S45: Extract trend smoothing components S1, S2, ..., S 4k ,S1,S2,...,S 4k represents the 1st, 2nd, …, 4kth trend smoothing component, where S m =(S1,S2,...,S n ), S1,S2,...,S n represents the 1, 2, …, 4kth trend smoothing component;
[0036] And determine the best smoothing coefficient α for each category, the smoothing coefficient gradually increases from 0 to 1 with a step size of 0.001, and the best smoothing coefficient α makes the decomposition of irradiance as complete as possible;
[0037] And subtract to get the residual components E1, E2, ..., E of each category 4k ,E1,E2,...,E 4k represents the 1st, 2nd, …, 4kth residual components; where E m =(E1,E2,...,E n ); E1, E2, ..., E n Represent the 1st, 2nd,…, 4kth residual components respectively;
[0038] S46: The corresponding simulation residual component They represent the 1st, 2nd, ..., 4kth simulation residual components respectively; trend smoothing components S1, S2, ..., S 4k ,S1,S2,...,S 4krepresent the 1st, 2nd, ..., 4kth trend smoothing components respectively; the day and night characteristic components D1, D2, ..., D 4k ,D1,D2,...,D 4k Represent the 1st, 2nd, ..., 4kth day and night characteristic components respectively; and the quarterly period component Y t To reorganize,
[0039] You can get M simulated irradiance samples Respectively represent the 1st, 2nd, …, Mth simulated irradiance samples;
[0040] In step 4, an irradiance simulation model taking into account seasonal and diurnal characteristics is proposed, and its expression is:
[0041] R t =Y t (D t +S t +E t )
[0042] Among them, R t represents the irradiance at time t; Y t represents the quarterly period component at time t; D t represents the day and night characteristic component at time t; S t represents the trend smoothing component at time t; E t Represents the residual component at time t.
[0043] Make the characteristics of the simulated irradiance sample consistent with the measured irradiance sample, including:
[0044] (1) It can generate simulated irradiance samples that are consistent with the variation range of the measured irradiance samples; Figure 6 As shown, Figure 6 The measured and simulated samples are in the same variation range;
[0045] (2) It can generate simulated irradiance samples that are consistent with the statistical characteristics, probability distribution characteristics, and autocorrelation characteristics of the measured irradiance samples, such as Figure 6 and Figure 7 As shown, the simulated irradiance of the present invention can maintain the probability distribution characteristics and autocorrelation characteristics consistent with the measured irradiance samples.
[0046] In step 5, the available capacity of each generator set is calculated, and the margin index is calculated in combination with the annual load of the system, which specifically includes the following steps:
[0047] S51: Accumulate the rated capacity of all conventional generating units in operation and calculate the available capacity P of all conventional generating units G , the expression is:
[0048]
[0049] Where: N G Indicates the number of conventional generator sets in operation; G n Indicates the rated capacity of the nth conventional generator set in operation;
[0050] S52: Substitute the simulated irradiance sample obtained in step 4 into the photovoltaic panel output power expression, and calculate the available capacity P of the equivalent photovoltaic unit PF ;
[0051] The output power of photovoltaic panels is expressed as:
[0052] P s =η PV A PV R t
[0053] Where: A PV Represents the area of the photovoltaic panel; R t represents the irradiance at time t; η PV Represents the energy conversion efficiency of the photovoltaic array
[0054] Calculate the available capacity P of the equivalent photovoltaic unit PF The expression is:
[0055]
[0056] Where: N PF Indicates the number of equivalent photovoltaic units in operation, represents the simulated irradiance sample at time t; P s (·) represents the power output model of the equivalent PV unit in operation.
[0057] S53: Based on the available capacity P of the conventional unit G and the available capacity P of the equivalent PV unit PF As well as the system annual load, calculate the power shortage amount LOEE (unit: MWh / yr), power shortage time LOLE (unit: hr. / yr) and power shortage frequency LOLF (unit: occ. / yr).
[0058]
[0059] Where: ENS q They represent the power shortage in the qth year; LOEE Q They represent the expected power shortage when the simulation year is Q; LLD q Respectively represent the power shortage time in the qth year; LOLE Q They represent the expected power shortage over time when the simulation year is Q; LLO qRespectively represent the missing frequency in the qth year; LOLF Q They respectively represent the expected power outage frequency when the simulation year is Q.
[0060] In step 6, the variance coefficient of the abundance index is often used as a convergence criterion for the time series Monte Carlo simulation method, and it is usually expected that the convergence speed of the variance coefficient of the power shortage will be smaller than that of the other abundance indexes. Therefore, in the present invention, when the variance coefficient of the expected power shortage is lower than 0.02, the process is stopped, and it is considered that the abundance index evaluated at this time is close enough to the real abundance index.
[0061] The present invention provides a photovoltaic grid-connected system adequacy assessment method taking into account seasonal and diurnal characteristics, and the technical effects are as follows:
[0062] 1) Step 1 of the present invention can use the same method to construct an equivalent photovoltaic unit according to the models and capacities of different photovoltaic panels, making the calculation easier;
[0063] 2) Step 2 of the present invention can determine the sample size required for simulation according to the sample size of the measured irradiance, which is more intuitive and convenient;
[0064] 3) Step 3 of the present invention can accurately simulate the operation and shutdown conditions of the unit during normal use;
[0065] 4) The irradiance simulation sample generated in step 4 of the present invention can accurately take into account the seasonal and diurnal characteristics of irradiance;
[0066] 5) The adequacy evaluation method used in step 5 of the present invention can accurately calculate each adequacy index;
[0067] 6) The stop condition set in step 6 of the present invention can ensure that the obtained simulated irradiance is maintained at a stable and accurate level;
[0068] 7) The simulated irradiance samples generated by the irradiance simulation model of steps 1 to 2 of the method of the present invention are consistent with the measured irradiance samples in terms of statistical indicators, probability distribution characteristics, and autocorrelation characteristics. Therefore, the margin indicators of the photovoltaic grid-connected system can be evaluated, making the results of the margin evaluation more accurate. BRIEF DESCRIPTION OF THE DRAWINGS
[0069] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:
[0070] Figure 1 The present invention is a flow chart of the photovoltaic grid-connected system adequacy assessment method taking into account seasonal and diurnal characteristics.
[0071] Figure 2 This is a flow chart of irradiance simulation taking into account seasonal and diurnal characteristics of the present invention.
[0072] Figure 3 This is a flow chart for evaluating the adequacy of photovoltaic grid connection.
[0073] Figure 4 This is the topological diagram of the distributed photovoltaic power generation system.
[0074] Figure 5 A schematic diagram of annual surplus, diurnal characteristics, trends and residual components of a certain type of irradiation according to an embodiment of the present invention.
[0075] Figure 6 1 is a frequency histogram of the measured irradiance and the simulated irradiance according to an embodiment of the present invention.
[0076] Figure 7 1 is a graph showing the autocorrelation coefficients of the measured irradiance and the simulated irradiance according to an embodiment of the present invention. DETAILED DESCRIPTION
[0077] A photovoltaic grid-connected system adequacy assessment method taking into account seasonal and diurnal characteristics. First, the irradiance is fitted using Fourier series to extract the annual variation trend; then, the irradiance is classified into different typical irradiation days based on the correlation between surface and extraterrestrial radiation; finally, the irradiance is further subdivided according to the different sunshine durations, and different parameters are selected for modeling. Through in-depth analysis and precise simulation, we can better understand the variation of irradiance under various influences, thereby optimizing the design and operation strategy of photovoltaic power generation systems, maximizing their economic benefits, and making positive contributions to addressing the energy crisis.
[0078] from Figure 1 It can be seen that a photovoltaic grid-connected system adequacy assessment method taking into account seasonal and diurnal characteristics includes the following steps:
[0079] S1: Based on the series-parallel structure of photovoltaic panels, the concept of equivalent photovoltaic units is constructed;
[0080] In step S1, a typical topological diagram of a distributed photovoltaic power generation system is as follows: Figure 4 As shown in the figure, the photovoltaic string obtained by connecting g photovoltaic panels in series is regarded as an equivalent photovoltaic unit, and its corresponding failure rate and repair rate are equivalent failure rate λ PV,s and the equivalent repair time γ PV,s .
[0081] S11: Equivalent failure rate λ of g photovoltaic panels connected in series PV,s and the equivalent repair time γ PV,s They are:
[0082]
[0083] Where: PV,i represents the failure rate of the i-th photovoltaic panel; γPV,i Indicates the repair time of the i-th photovoltaic panel.
[0084] S12: The output power of the photovoltaic panel is determined by the solar irradiance, the photovoltaic panel area and the energy conversion efficiency of the photovoltaic panel, and its expression is:
[0085] P s =η pv A pv R t ;
[0086] Where: A pv Represents the area of the photovoltaic panel; R t represents the irradiance at time t; η pv Represents the energy conversion efficiency of the photovoltaic array.
[0087] S13: The energy conversion efficiency of the photovoltaic array is expressed as:
[0088] η pv =η r η pc [1-β(T c -T cref )];
[0089] Where: η r Represents the reference efficiency of the photovoltaic module; η pc represents the power regulation efficiency which is set to 1 by assuming perfect maximum power point tracking conditions; β represents the efficiency temperature coefficient, which is 0.005 / ℃; T c Indicates module temperature; T cref Indicates the reference module temperature, which is 25°C.
[0090] S14: Module temperature T c The calculation formula is:
[0091] T c =T a +[(T NOCT -20) / R NOCT ]R t ;
[0092] Where: T a Indicates ambient temperature; T NOCT and R NOCT Indicates the battery temperature and solar irradiance under rated operating battery temperature conditions, with values of 47°C and 800W / m 2 .
[0093] S2: Collect long-term measured irradiance samples from meteorological observation stations and determine the number of measured irradiance samples. According to the accuracy requirements of the margin assessment, set the number of simulated irradiance samples to be generated;
[0094] In step S2, the sample size N of the measured wind speed samples is determined by the source meteorological observation station. The sample size M of the simulated wind speed to be generated is set according to the evaluation accuracy, and it is assumed that all the generator sets (conventional sets and equivalent photovoltaic sets) are in normal operation at the initial simulation time.
[0095] S3: The number of continuous steps of the state of each generator set in the running and shutdown states;
[0096] S31: When the nth generator set is in operation, the number of time steps TTR of the generator set in operation n for:
[0097]
[0098] Where: n represents the failure rate of the nth generator set, r is a random number uniformly distributed in the interval [0,1]; Indicates the operation of rounding the data upwards.
[0099] S32: When the nth generator set is in shutdown state, the number of time steps TTF of the generator set in shutdown state n for:
[0100]
[0101] Where: Represents the repair rate of the nth generator set.
[0102] from Figure 2 It can be seen that a method for irradiance simulation taking into account seasonal and diurnal characteristics includes the following steps:
[0103] S4: simulate the irradiance and generate simulated irradiance samples taking into account seasonal and diurnal characteristics;
[0104] S41: Fit N measured irradiance samples with a time step of 1h to obtain parameters a0, b0, c0, w0, and determine the quarterly period component Y t , the expression is:
[0105] Y t =a0+b0cos(w0t)+c0sin(w0t);
[0106] Where: Y t is the quarterly cycle component at time t; a0, b0, c0, w0 are the coefficients to be fitted.
[0107] S42: Based on the characteristic categories of irradiance, a decomposition structure as shown below is constructed:
[0108] R t =Y t (D t +S t +E t );
[0109] Where: R t represents the irradiance at time t; Y t represents the quarterly period component at time t; D t represents the day and night characteristic component at time t; S t represents the trend smoothing component at time t; E t Represents the residual component at time t.
[0110] S43: Calculate the extraterrestrial radiation, the expression is:
[0111] EHI=EDNI*(cosφcosδcosω+sinφsinδ);
[0112] Where: EDNI represents the extraterrestrial solar irradiance; φ represents the local latitude; δ represents the solar declination; ω represents the solar hour angle, and the calculation of the extraterrestrial solar irradiance EDNI is shown in the formula:
[0113]
[0114] Where: E0 represents the solar constant, which is 1366.1W / m 2 ; d represents the cumulative day, that is, the ordinal number of the date in the year.
[0115] The calculation formula for the solar declination δ is:
[0116]
[0117] The calculation formula for the solar hour angle ω is:
[0118] ω=15(T T -12);
[0119] Where: T T To express true solar time, the calculation is as follows:
[0120] T T =C T +E Q +L C
[0121] =C T +E Q +(L g -120) / 15;
[0122] Where: C T Indicates local standard time; EQ Indicates time difference; L C Indicates the longitude correction value, where L g Indicates longitude.
[0123] Calculate the correlation coefficient between the daily extraterrestrial irradiance and the surface irradiance, and use K-means clustering to cluster the remaining components H t The day is roughly divided into four categories: A-extremely strong irradiation day, B-strong irradiation day, C-medium irradiation day, and D-weak irradiation day. The start and end times of the irradiation are counted and classified every day, and finally a 4k type irradiance sequence is obtained;
[0124] S44: Fitting to obtain 4k day and night characteristic components D1, D2, ..., D 4k , where each category contains values at several moments, namely D m =(D1,D2,...,D n ), 4k day and night feature component expressions:
[0125]
[0126] Where: D m (t) represents the mth (m=1, 2, ..., 4k) day and night characteristic component; a m ,b m ,c m ,w m They respectively represent the coefficients to be fitted of the mth (m=1, 2, ..., 4k) type of day and night characteristic components.
[0127] S45: Extract trend smoothing components S1, S2, ..., S 4k , where S m =(S1,S2,...,S n ), and determine the optimal smoothing coefficient α for each class, which is expressed as:
[0128]
[0129] Where: α is the static smoothing coefficient and the value of α is (0,1); Z t is the remaining component after decomposing the seasonal component and the diurnal characteristic component of the irradiance at time t, that is, Z t =R t / Y t -D t ; are respectively Z at the tth moment t The first, second, and third exponential smoothing values of .
[0130] The smoothed value S at time t t for:
[0131] S t =x t +y t +z t ;
[0132] in:
[0133]
[0134] Where: Smoothing initial value Take it as 0. The smoothing coefficient α is a variable to be determined, which reflects the influence of historical data on the exponential smoothing value.
[0135] According to E t =Z t -S t Get the residual components E1, E2, ..., E of each category 4k , where E m =(E1,E2,...,E n );
[0136] S46: Calculate the mean and variance of each type of residual component in step S45, and randomly generate M simulated residual components that obey the normal distribution and have the same mean and variance. in
[0137] S47: The corresponding simulation residual component Trend smoothing components S1, S2, ..., S 4k 、Day and night characteristic components D1,D2,...,D 4k and the quarterly cycle component Y t After reorganization, M simulated irradiance samples can be obtained.
[0138] This irradiance simulation model, which takes into account seasonal and diurnal characteristics, combines Fourier series fitting, K-means clustering, and cubic exponential smoothing. It can generate simulated irradiance samples that are consistent with the variation range of the measured irradiance samples, and its statistical characteristics, probability distribution characteristics, and autocorrelation characteristics are consistent with the measured irradiance samples.
[0139] from Figure 3 As can be seen from the figure, the photovoltaic grid-connected adequacy assessment flow chart includes the following steps:
[0140] S5: using the simulated irradiance sample in step S4 and combining it with the Monte Carlo simulation method, calculate the margin index of each generator set;
[0141] S51: Calculate the available capacity P of all conventional units G :
[0142]
[0143] Where: G n Indicates the rated capacity of the nth conventional generator set in operation; N G Indicates the number of conventional generator sets in operation.
[0144] S52: Substitute the simulated irradiance sample obtained in step S3 into the output power expression of the photovoltaic panel, and calculate the available capacity P of the equivalent photovoltaic unit PF , where the output power of the photovoltaic panel is expressed as:
[0145] P s =η pv A pv R t
[0146] Where: A pv Represents the area of the photovoltaic panel; R t represents the irradiance at time t; η pv represents the energy conversion efficiency of the photovoltaic array and is given by:
[0147] η pv =η r η pc [1-β(T c -T cref )]
[0148] Where: η r Represents the reference efficiency of the photovoltaic module; η pc represents the power regulation efficiency which is set to 1 by assuming perfect maximum power point tracking conditions; β represents the efficiency temperature coefficient, which is 0.005 / ℃; T c Indicates module temperature; T cref Indicates the reference module temperature, which is 25°C.
[0149] Where, represents the module temperature T c The calculation formula is:
[0150] T c =T a +[(T NOCT -20) / R NOCT ]R t
[0151] Where: T a Indicates ambient temperature; T NOCT and R NOCT Indicates the battery temperature and solar irradiance under rated operating battery temperature conditions, with values of 47°C and 800W / m 2 .
[0152] S53: Based on the available capacity P of the conventional unit G and the available capacity P of the equivalent photovoltaic unit PF As well as the system annual load, calculate the power shortage (MWh / yr), power shortage time (hr. / yr) and power shortage frequency (occ. / yr) in the qth year:
[0153]
[0154] Where: ENS q Indicates the power shortage in the qth year, LOEE Q It represents the expected power shortage in the quarter when the simulation year is Q.
[0155]
[0156] Where: LLD q Indicates the power shortage time in the qth year, LOLE Q It represents the expected power outage time when the simulation year is Q.
[0157]
[0158] Where: LLO q Indicates the power outage frequency in the qth year, LOLF Q It represents the expected power outage frequency when the simulation year is Q.
[0159] S6: Repeat steps S3 to S5. When the variance coefficient of the expected power shortage is lower than 0.02, stop and output the margin index.
[0160] Example:
[0161] The present invention collects the open meteorological database North Dakota agricultural weather network (NDAWN) and obtains the irradiance samples of the Linton meteorological observation station with a time step of 1h from January 1, 2011 to December 31, 2020. The irradiance simulation model taking into account the seasonal and diurnal characteristics of the present invention is used to evaluate the adequacy of the photovoltaic grid-connected IEEE-RTS power generation system. It is assumed that a photovoltaic power station is built at the Linton meteorological station, and a photovoltaic panel of model CSUN310-72P is used in the photovoltaic array. The rated capacity of the photovoltaic panel is 310W, and the area of the photovoltaic panel is 1.7522m 2 The energy conversion efficiency of the photovoltaic panel is 16%, and the failure rate and repair time are 1.384×10 -4 times / year and 24h. According to the definition of equivalent photovoltaic unit in Section 3.4.1, 13 photovoltaic panels can be connected in series as an equivalent photovoltaic unit with a capacity of 4030W, and its equivalent failure rate is about 1.8×10-3 times / year, equivalent repair time is about 24h / time.
[0162] Figure 5 Shown are the annual residual component, day and night characteristic component, trend smoothing component and residual component of the irradiance on a typical day with a certain sunshine period. It can be seen from the figure that the period of the day and night characteristic component is consistent with the period of the annual residual component, accurately showing the law of diurnal variation of irradiance, the increase and decrease of the trend component and the annual residual component have a consistent trend, and the residual component fluctuates randomly around 0.
[0163] Figure 6 The frequency histogram of the measured irradiance and the simulated irradiance is shown in Fig. Figure 6 T1, T2, ..., T 10 They represent 10 intervals of [0,100), [100,200),…, [900,1000) respectively. It can be seen from the figure that the frequency histogram of the irradiance simulation sample of the present invention is close to that of the measured irradiance sample, which proves that the simulated irradiance of the present invention can maintain the probability distribution characteristics consistent with the measured irradiance sample, and the variation range is also consistent with the measured irradiance sample.
[0164] Figure 7 The figure shows the autocorrelation coefficient curve of the measured irradiance and the simulated irradiance within a lag time of 2h. Figure 7 It can be seen that the autocorrelation coefficient of the simulated irradiance of the present invention is relatively close to the autocorrelation coefficient of the measured irradiance, indicating that the irradiance simulation sample of the present invention can maintain the same autocorrelation characteristics as the measured irradiance sample.
[0165] Table 1 Linton Observatory IEEE-RBTS PV grid-connected system adequacy index
[0166]
[0167] Assuming that the Linton photovoltaic power station is connected to the IEEE-RBTS power generation system, the measured irradiance samples, the irradiance simulation samples proposed in this chapter, and the irradiance simulation samples divided by weather type according to Euclidean distance are used to evaluate the adequacy index of the IEEE-RBTS photovoltaic grid-connected system when the photovoltaic capacity is 4MW, 8MW, and 12MW respectively. The results are shown in Table 1. The brackets in the table are the percentage changes of the adequacy index of the model compared with the measured irradiance adequacy index, from which the following conclusions can be drawn:
[0168] 1) With the increase of equivalent photovoltaic unit capacity, the margin indicators LOEE, LOLE and LOLF all show a decreasing trend. Therefore, the larger the capacity of the equivalent photovoltaic unit, the higher the margin benefit of the photovoltaic grid-connected power generation system;
[0169] 2) Under the capacities of different equivalent photovoltaic units, the deviation between the model of the present invention and the margin index of the measured irradiance samples is largely smaller than the deviation between the Euclidean distance irradiance model and the margin index of the measured irradiance samples. On the whole, the margin assessment accuracy of the model of the present invention is higher than that of the Euclidean distance irradiance model.
[0170] The irradiance simulation model proposed in the present invention, which takes into account seasonal and day-night characteristics, can accurately classify the irradiance, model each one, and generate any number of simulated irradiance samples. The simulated irradiance samples of each type of sunshine duration under each typical irradiation day have similar statistical indicators to the measured irradiance samples, and the overall annual mean, standard deviation, probability distribution characteristics and autocorrelation characteristics are also consistent with the measured irradiance samples. In addition, the abundance simulation model proposed in the present invention can accurately evaluate the abundance of wind and solar grid-connected systems. The application of the present invention in large-scale photovoltaic grid-connected abundance assessment is of vital importance to the reliable basis for the planning and operation of photovoltaic grid-connected systems.
Claims
1. A method for assessing the reliability of photovoltaic power generation capacity taking into account seasonal and diurnal characteristics of irradiance, characterized in that The following steps are involved: Step 1: Construct an equivalent photovoltaic unit model according to the series-parallel structure of photovoltaic panels; Step 2: Collect the long-term measured irradiance samples of the meteorological observation station and determine the number of measured irradiance samples; according to the accuracy requirements of the margin assessment, set the number of simulated irradiance samples to be generated; Step 3: Sample the number of continuous steps of each generator set in operation and shutdown state; Step 4: Simulate the irradiance and generate simulated irradiance samples that take into account seasonal and day-night characteristics; Step 5: Calculate the margin index of each generator set by using the simulated irradiance sample in step S4 in combination with the Monte Carlo simulation method; Step 6: Repeat steps 3 to 5. When the stop condition is met, the loop ends and each margin index is output.
2. The photovoltaic power generation capacity credibility assessment method taking into account the seasonality and diurnal characteristics of irradiance according to claim 1 is characterized by: In step 1, the construction of the equivalent photovoltaic unit model is as follows: The photovoltaic string obtained by connecting g photovoltaic panels in series is regarded as an equivalent photovoltaic unit, and the equivalent failure rate λ is calculated according to the failure rate and repair time of each photovoltaic panel. PV,s and the equivalent repair time γ PV,s They are: Among them, λ PV and γ PV are the failure rate and repair time respectively; g represents the number of photovoltaic panels connected in series; λ PV,s represents the equivalent failure rate; γ PV,s The equivalent repair time represented by λ PV,i represents the equivalent failure rate; γ PV,i Represents the equivalent repair time of the i-th photovoltaic panel.
3. The photovoltaic power generation capacity credibility assessment method taking into account the seasonality and diurnal characteristics of irradiance according to claim 1 is characterized by: In step 3, when the nth generator set is in operation, the number of time steps TTR for the generator set to be in operation is calculated according to the failure rate λ of the generator set. n Sampling is performed as shown below: Where: n represents the failure rate of the nth generator set; r represents a uniformly distributed random number in the range [0,1]; Indicates that the data is rounded up; When the nth generator set is in a shutdown state, the number of time steps TTF for the generator set in the shutdown state is calculated according to the repair rate μ of the generator set. n Sampling is performed as shown below: Where: represents the repair rate of the nth generator set; r represents a uniformly distributed random number in the range [0,1]; Indicates rounding the data upwards.
4. The photovoltaic power generation capacity credibility assessment method taking into account the seasonality and diurnal characteristics of irradiance according to claim 1 is characterized by: The step 4 comprises the following steps: S41: Fit N measured irradiance samples to obtain parameters a0, b0, c0, w0, and determine the quarterly period component Y t ; S42: Divide the measured irradiance sample by the quarterly period component to obtain the annual residual component H t , that is, H t =R t / Y t , R t represents the irradiance at time t; Y t represents the quarterly period component at time t; S43: Calculate the extraterrestrial radiation and the correlation coefficient between the extraterrestrial radiation and the surface radiation every day. The calculation formula is as follows: Where: C n represents the correlation coefficient of irradiance on the nth day; G i represents the i-th cluster center, i=1,2,3,4; M represents the number of irradiance days; Use K-means clustering to cluster the remaining components H t The day is roughly divided into four categories: A-extremely strong irradiation day, B-strong irradiation day, C-medium irradiation day, and D-weak irradiation day; the daily irradiation start and end times are counted and classified, and finally a 4k type irradiance sequence is obtained; S44: Fitting to obtain 4k day and night characteristic components D1, D2, ..., D 4k ,D1,D2,...,D 4k represents the 1st, 2nd, …, 4kth day and night characteristic components, each of which contains values at several moments, namely D m =(D1,D2,...,D n ); D1,D2,...,D n represents the 1st, 2nd,…, 4kth day and night characteristic components; S45: Extract trend smoothing components S1, S2, ..., S 4k ,S1,S2,...,S 4k represents the 1st, 2nd, …, 4kth trend smoothing component, where S m =(S1,S2,...,S n ), S1,S2,...,S n represents the 1, 2, …, 4kth trend smoothing component; And determine the optimal smoothing coefficient α for each category, and subtract them to obtain the residual components E1, E2, ..., E 4k ,E1,E2,...,E 4k represents the 1st, 2nd, …, 4kth residual components; where E m =(E1,E2,...,E n ); E1, E2, ..., E n Represent the 1st, 2nd,…, 4kth residual components respectively; S46: The corresponding simulation residual component They represent the 1st, 2nd, ..., 4kth simulation residual components respectively; trend smoothing components S1, S2, ..., S 4k ,S1,S2,...,S 4k represent the 1st, 2nd, ..., 4kth trend smoothing components respectively; the day and night characteristic components D1, D2, ..., D 4k ,D1,D2,...,D 4k Respectively represent the 1st, 2nd, ..., 4kth day and night characteristic components; and the quarterly period component Y t To reorganize, You can get M simulated irradiance samples Represent the 1st, 2nd,…,Mth simulated irradiance samples respectively.
5. According to the method for evaluating the reliability of photovoltaic power generation capacity taking into account the seasonality and diurnal characteristics of irradiance as described in claim 1, Features: In step 4, an irradiance simulation model taking into account seasonal and diurnal characteristics is proposed, and its expression is: R t =Y t (D t +S t +E t ) Among them, R t represents the irradiance at time t; Y t represents the quarterly period component at time t; D t represents the day and night characteristic component at time t; S t represents the trend smoothing component at time t; E t Represents the residual component at time t; The characteristics of the simulated irradiance samples are consistent with those of the measured irradiance samples, including: (1) It can generate simulated irradiance samples that are consistent with the variation range of the measured irradiance samples; (2) It can generate simulated irradiance samples that are consistent with the statistical characteristics, probability distribution characteristics, and autocorrelation characteristics of the measured irradiance samples.
6. The photovoltaic power generation capacity credibility assessment method taking into account the seasonality and diurnal characteristics of irradiance according to claim 1, Features: In step 5, the available capacity of each generator set is calculated, and the margin index is calculated in combination with the annual load of the system, which specifically includes the following steps: S51: Accumulate the rated capacity of all conventional generating units in operation and calculate the available capacity P of all conventional generating units G , the expression is: Where: N G Indicates the number of conventional generator sets in operation; G n Indicates the rated capacity of the nth conventional generator set in operation; S52: Substitute the simulated irradiance sample obtained in step 4 into the photovoltaic panel output power expression, and calculate the available capacity P of the equivalent photovoltaic unit PF ; The output power of photovoltaic panels is expressed as: P s =the PV A PV R t Where: A PV Represents the area of the photovoltaic panel; R t represents the irradiance at time t; η PV Represents the energy conversion efficiency of the photovoltaic array Calculate the available capacity P of the equivalent photovoltaic unit PF The expression is: Where: N PF Indicates the number of equivalent photovoltaic units in operation, represents the simulated irradiance sample at time t; P s (·) represents the power output model of the equivalent PV unit in operation; S53: Based on the available capacity P of the conventional unit G and the available capacity P of the equivalent PV unit PF As well as the system annual load, calculate the power shortage amount LOEE, power shortage time LOLE and power shortage frequency LOLF; Where: ENS q They represent the power shortage in the qth year; LOEE Q They represent the expected power shortage when the simulation year is Q; LLD q Respectively represent the power shortage time in the qth year; LOLE Q They represent the expected power shortage over time when the simulation year is Q; LLO q Respectively represent the missing frequency in the qth year; LOLF Q They respectively represent the expected power outage frequency when the simulation year is Q.
7. The photovoltaic power generation capacity credibility assessment method taking into account the seasonality and diurnal characteristics of irradiance according to claim 6 is characterized by: In step 6, the variance coefficient of the adequacy index is used as the convergence criterion of the time series Monte Carlo simulation method, and the convergence speed of the variance coefficient of the expected power shortage will be lower than that of the other adequacy indexes. When the variance coefficient of the expected power shortage is lower than 0.02, the process stops, and it is considered that the adequacy index evaluated at this time is close enough to the real adequacy index.
