Photovoltaic power generation capacity credibility evaluation method based on joint probability modeling and random process

Abstract

The invention discloses a photovoltaic power generation capacity credibility evaluation method based on joint probability modeling and a random process, and the method comprises the steps: collecting historical irradiance and temperature data of a target region, extracting an hour-level mean value and a standard deviation of the historical irradiance and temperature data, and dividing a seasonal data set; a Copula function is adopted to construct a joint probability distribution model of irradiance and temperature, a Gumbel Copula function is selected in summer, and a Clayton Copula function is selected in winter; generating a random fluctuation sequence of irradiance and temperature based on a stochastic differential equation, and calculating photovoltaic real-time output by combining a photovoltaic cell characteristic equation; a double-state Markov model is adopted to simulate traditional unit faults and load fluctuation, and system loss of load expectation (LOLE) is quantified through Monte Carlo simulation; carrying out dynamic iteration by utilizing an equivalent conventional capacity method (ECP), and calculating the capacity credibility of photovoltaic power generation; according to the invention, a reliability evaluation basis is provided for high-proportion photovoltaic grid-connected planning.

Description

Technical Field

[0001] The present invention relates to the technical field of power system planning and photovoltaic reliability assessment, and particularly relates to a method for evaluating the reliability of photovoltaic power generation capacity based on joint probability modeling and stochastic processes. Background Art

[0002] As one of the most abundant and easily accessible renewable energy sources, photovoltaic power generation has low power generation costs, is environmentally friendly, and has the potential to replace fossil fuels and meet global energy demands. Accurately and objectively evaluating the reliability of photovoltaic power generation capacity is of great significance for power system operation and planning. However, photovoltaic power generation has the characteristics of intermittency and volatility, and its output is affected by multiple factors such as weather and seasonal changes, resulting in uncertainty in photovoltaic power generation power, which poses a major challenge to the stability of high-penetration power systems. Traditional reliability assessment methods for capacity usually adopt univariate probability distributions and do not fully consider the dynamic correlation between irradiance and temperature, resulting in large deviations in assessment results. In addition, traditional methods do not consider high-precision stochastic process models to simulate the time-series characteristics of photovoltaic output, and cannot obtain the impact of short-term fluctuations in photovoltaic power on system reliability. Summary of the Invention

[0003] Object of the Invention: The object of the present invention is to provide a method for evaluating the reliability of photovoltaic power generation capacity based on joint probability modeling and stochastic processes, which solves the problems of ignoring the correlation between irradiance and temperature and time-series fluctuations in traditional methods.

[0004] Technical Solution: A method for evaluating the reliability of photovoltaic power generation capacity based on joint probability modeling and stochastic processes according to the present invention includes the following steps: S1. Collect historical irradiance data G t and ambient temperature data , and extract their hourly means G trend,t , and standard deviations σ G,t , σ T,t ; S2. Divide the data according to seasons. In summer, use the Gumbel Copula function to establish a joint probability distribution model of irradiance G t and temperature , and in winter, use the Clayton Copula function to establish a joint probability distribution model of irradiance G t and temperature ; S3. Generate a random fluctuation sequence of irradiance and temperature using the Stochastic Differential Equation (SDE); S4. Calculate the real-time PV output according to the PV cell characteristic equation P pv,t , and use a two-state Markov model to simulate the faults of traditional units and load fluctuations; S5. Generate time series data of PV output, traditional unit status, and load based on Monte Carlo simulation, and calculate the Expected Load Shedding (ELS) of the system; S6. Determine the PV capacity credibility using the Equivalent Conventional Capacity Method (ECP) CC pv.

[0005] Furthermore, in step S2, the joint probability distribution model is as follows: ; where H( ) represents the joint cumulative distribution function; G t is the historical irradiance data of the target area; C θ is the Copula function; Z G is a standard normal distribution random number related to Z T Z, T is a standard normal distribution random number related to Z G ; Z T and Z G consider their correlation through the joint probability distribution model; G trend,t is the hourly average irradiance; is the hourly average of the ambient temperature data; is the ambient temperature data of the target area; σ G,t is the variance of the historical hourly solar irradiance; σ T,t is the variance of the historical hourly ambient temperature; Φ( Z G ) is Z G 's standard normal distribution; Φ( Z T ) is Z T 's standard normal distribution.

[0006] Furthermore, in step S2, the Gumbel Copula function used in summer is as follows: ; Among them, θ reflects the correlation strength, θ the closer it is to 1, the more independent it is, θ when it approaches infinity, it indicates complete correlation.

[0007] Furthermore, in step S2, the Clayton Copula function is adopted in winter as follows: ; Among them, θ the closer it is to 0, the more independent it is, θ when it approaches infinity, it indicates complete correlation.

[0008] Furthermore, in step S3, the specific process of generating the random fluctuation sequences of irradiance and temperature by using the stochastic differential equation SDE is as follows: ; Among them, G n,t is the stochastic irradiance variable modeled by SDE; is the stochastic ambient temperature variable modeled by SDE; dW t represents the differential form of the Wiener process.

[0009] Furthermore, in step S4, the photovoltaic real-time output is calculated according to the photovoltaic cell characteristic equation P pv,t The specific process is: ; is t the output power of solar photovoltaic power generation at time is the rated power (W) of solar photovoltaic power generation; T cell,t is the temperature (°C) of the solar cell; G STC = 1000 W / m 2 is the standard test irradiance; T STC is the standard test temperature; T STC = 25 °C; γ is the temperature coefficient, -0.05 < γ<- 0.003; T NOC is the nominal operating temperature; T NOC = 46 °C; is the actual electric power of solar photovoltaic power generation at time t; is the conversion efficiency of solar photovoltaic electric power; The traditional unit failure and load fluctuation are simulated using a two-state Markov model, and the formula is as follows: ; Wherein, TTF is the duration of the failure time or the power generation equipment availability; TTR is the repair time or the time when the power generation equipment is unavailable; λ is the failure rate of each power generation unit in the two-state Markov model; r is the average repair time of each power generation unit, equal to ; is the repair rate of each power generation unit in the two-state Markov model; U is a random value uniformly distributed in [0,1], Z is a random value of the standard normal distribution; is the variance of the repair time of each generator, represented by 0.1 r ; L t is t the load magnitude at time is t the predicted value of the load magnitude at time α is the load prediction error coefficient; Z is a random number conforming to the standard normal distribution, Z ~ N (0,1).

[0010] Furthermore, step S5 is specifically as follows: ; Wherein, LOLE is the system loss-of-load expectation; d ( P t < L t ) is the duration when the total system output is less than the load; T is the total simulation duration.

[0011] Furthermore, step S6 is specifically as follows: First, calculate the baseline LOLE , keep the photovoltaic installed capacity C pv and the original traditional unit capacity C conv,initial unchanged, and generate the photovoltaic output time series data based on steps S1~S5; based on step S4, simulate the traditional unit random failure states TTF and TTR to generate the unit available capacity sequence P conv,t ; then, superimpose the actual electric power of the solar photovoltaic power generation at time t and calculate the total system output Pt , where: P t =P conv,t + P e,pv,t ; Finally, calculate the reference LOLE value LOLE base ; Then, perform iterative calculations on the equivalent traditional capacity, remove the photovoltaic system, set the step size ΔC = 1 MW, and gradually increase the capacity of the traditional unit; in the k-th iteration, the capacity of the traditional unit is updated to C conv,k , ; Calculate the new LOLE value LOLE new,k when the condition ∣ LOLE new,k − LOLE base ∣≤ϵ is met, where is the tolerance threshold, = 0.09 h / year; finally record the final equivalent traditional capacity C conv，end ; Finally, calculate the credibility of the photovoltaic power generation capacity CC pv , and its formula is: .

[0012] An electronic device according to the present invention includes a memory, a processor, and a computer program stored on the memory. When the processor executes the program, the steps of any one of the above methods are implemented.

[0013] A computer-readable storage medium according to the present invention stores a computer program, and when the program is executed by a processor, the steps of any one of the above methods are implemented.

[0014] Beneficial effects: Compared with the prior art, the present invention has the following remarkable advantages: In the evaluation of the credibility of photovoltaic power generation capacity, the present invention mainly focuses on modeling the joint probability distribution of irradiance and temperature, takes into account the influence of meteorological variables such as irradiance and temperature on the volatility of photovoltaic output and system reliability, reasonably constructs a seasonal Copula function and a stochastic differential equation model, and the capacity credibility evaluation is compatible with the requirements of both system reliability analysis and output time series fluctuations, giving full play to the advantages of stochastic processes and Monte Carlo simulation in high-precision evaluation. Description of the Drawings

[0015] Figure 1This is the flowchart of the present invention. Detailed implementation manners

[0016] The technical solution of the present invention will be further described below in conjunction with the accompanying drawings.

[0017] As Figure 1 shown, an evaluation method for the reliability of photovoltaic power generation capacity based on joint probability modeling and stochastic process is provided in an embodiment of the present invention. First, historical irradiance and temperature data of the target area are collected, the hourly mean and standard deviation are extracted, and the seasonal data sets are divided; then, the Copula function is used to construct the joint probability distribution model of irradiance and temperature, where the Gumbel Copula function is selected in summer and the Clayton Copula function is selected in winter; subsequently, based on the stochastic differential equation, the stochastic fluctuation sequences of irradiance and temperature are generated, and the real-time photovoltaic output is calculated in combination with the characteristic equation of the photovoltaic cell; then, the two-state Markov model is used to simulate the faults of traditional units and load fluctuations, and the loss of load expectation (LOLE) of the system is quantified through Monte Carlo simulation; finally, the equivalent conventional capacity method (ECP) is used for dynamic iteration to calculate the reliability of the photovoltaic power generation capacity. The specific steps are as follows: S1. Collect historical irradiance data of the target area G t and ambient temperature data , and extract their hourly means G trend,t , and standard deviations σ G,t , σ T,t ; S2. Divide the data according to seasons. In summer, the Gumbel Copula function is used to establish the joint probability distribution model of irradiance G t and temperature , and in winter, the Clayton Copula function is used to establish the joint probability distribution model of irradiance G t and temperature ; among them, the joint probability distribution model is specifically as follows: ; where H( ) represents the joint cumulative distribution function; G t is the historical irradiance data of the target area; C θ is the Copula function; Z G is a standard normal distribution random number related to Z T ZT is a standard normal distribution random number related to Z G , and its correlation is considered through a joint probability distribution model; Z T is the mean value of irradiance per hour; Z G is the mean value of environmental temperature data per hour; G trend,t is the environmental temperature data of the target area; is the variance of historical hourly solar irradiance; is the variance of historical hourly environmental temperature; Φ( σ G,t ) is σ T,t is the standard normal distribution of Z G ; Φ( Z G ) is Z T is the standard normal distribution of Z T .

[0018] In summer, the Gumbel Copula function is adopted as follows: ; wherein, θ reflects the correlation strength, θ the closer it approaches 1, the more independent it is, θ when it approaches infinity, it indicates complete correlation.

[0019] In winter, the Clayton Copula function is adopted as follows: ; wherein, θ the closer it approaches 0, the more independent it is, θ when it approaches infinity, it indicates complete correlation.

[0020] S3. Use the stochastic differential equation SDE to generate a stochastic fluctuation sequence of irradiance and temperature; the specific process is as follows: ; wherein: G n,t is the stochastic irradiance variable modeled by SDE; is the stochastic environmental temperature variable modeled by SDE; dW t represents the differential form of the Wiener process.

[0021] S4. Calculate the real-time output of the photovoltaic cell according to the photovoltaic cell characteristic equationP pv,t , and a two-state Markov model is used to simulate the faults and load fluctuations of traditional units; the real-time output of the photovoltaic power is calculated according to the characteristic equation of the photovoltaic cell P pv,t The specific process is as follows: ; In the formula: is t the output power of solar photovoltaic power generation at time the rated power (W) of solar photovoltaic power generation; T cell,t is the temperature (°C) of the solar cell; G STC = 1000 W / m 2 is the standard test irradiance; T STC is the standard test temperature, T STC = 25°C; γ is the temperature coefficient, -0.05 < γ<- 0.003; T NOC is the nominal operating temperature, T NOC = 46°C; is the actual electric power of solar photovoltaic power generation at time t; is the conversion efficiency of solar photovoltaic electric power, = 98%; The formula for simulating the faults and load fluctuations of traditional units using a two-state Markov model is as follows: ; In the formula: TTF is the duration of the fault time or the availability of the power generation equipment; TTR is the repair time or the time when the power generation equipment is unavailable; λ is the failure rate of each power generation unit in the two-state Markov model; r is the average repair time of each power generation unit, equal to ; is the repair rate of each power generation unit in the two-state Markov model; U is a random value uniformly distributed in [0,1], Z is a random value of the standard normal distribution; is the variance of the repair time of each generator, represented by 0.1 r ; L t is t the load magnitude at time ist Predicted value of the momentary load magnitude α is the load prediction error coefficient Z is a random number conforming to the standard normal distribution Z ~ N (0, 1)

[0022] S5. Generate photovoltaic power output, traditional unit status, and load time series data based on Monte Carlo simulation, and calculate the expected load shedding of the system; specifically as follows: ; In the formula: LOLE is the expected load shedding of the system d ( P t < L t ) is the duration when the total system output is less than the load T is the total simulation duration

[0023] S6. Determine the photovoltaic capacity credibility CC pv using the equivalent conventional capacity method ECP. Specifically as follows: First, calculate the benchmark LOLE , keep the photovoltaic installed capacity C pv and the original traditional unit capacity C conv,initial unchanged, and generate photovoltaic power output time series data based on steps S1 - S5; based on step S4, simulate the random failure states TTF and TTR of the traditional units to generate the available capacity sequence P conv,t of unit j; superimpose the actual electric power of solar photovoltaic power generation at time t to calculate the total system output P t , where: P t = P conv,t + P e,pv,t ; Finally, calculate the benchmark LOLE value LOLE base ; Then, perform iterative calculations on the equivalent conventional capacity, remove the photovoltaic system, set the step size ΔC = 1 MW, and gradually increase the traditional unit capacity; in the k - th iteration, the traditional unit capacity is updated to C conv,k , ; Calculate the new LOLE value LOLE new,k when the condition ∣ LOLE new,k − LOLEbase | ≤ terminate the iteration when is the tolerance threshold, = 0.09 h / year; finally record the final equivalent traditional capacity C conv，end ; Finally, calculate the credibility of the photovoltaic power generation capacity CC pv , and its formula is: .

Claims

1. A photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process, characterized in that: The following steps are involved: S1. Collect historical irradiance data of the target area G t and ambient temperature data , extract its hourly mean G trend,t , and standard deviation σ G,t , σ T,t ; S2. Divide the data according to seasons and use the Gumbel Copula function to establish irradiance in summer G t With temperature The joint probability distribution model of the irradiance in winter is established using the Clayton Copula function. G t With temperature The joint probability distribution model of S3, using stochastic differential equation SDE to generate random fluctuation sequence of irradiance and temperature; S4. Calculate the real-time photovoltaic output based on the photovoltaic cell characteristic equation P pv,t , and a two-state Markov model is used to simulate traditional unit failures and load fluctuations; S5. Generate photovoltaic output, traditional unit status and load time series data based on Monte Carlo simulation, and calculate system load loss expectation; S6. Determine the reliability of photovoltaic capacity using the equivalent conventional capacity method (ECP) CC pv.

2. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: In step S2, the joint probability distribution model is as follows: ； Where H( ) represents the joint cumulative distribution function; G t is the historical irradiance data of the target area; C θ is the Copula function; Z G is a Z T The associated standard normally distributed random number, Z T is a Z G The associated standard normally distributed random numbers, Z T and Z G Consider their correlation through joint probability distribution model; G trend,t is the hourly mean irradiance; Hourly average of ambient temperature data; is the ambient temperature data of the target area; σ G,t is the variance of historical hourly solar irradiance; σ T,t is the historical hourly ambient temperature variance; Φ( Z G )yes Z G The standard normal distribution of Z T )yes Z T The standard normal distribution of .

3. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: In step S2, the Gumbel Copula function is used in summer as follows: ； in, θ Reflects the strength of correlation. θ The closer it is to 1, the more independent it is. θ Approaching infinity indicates perfect correlation.

4. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: In step S2, the Clayton Copula function used in winter is as follows: ； in, θ The closer it is to 0, the more independent it is. θ Approaching infinity indicates perfect correlation.

5. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: In step S3, the specific process of using the stochastic differential equation SDE to generate the random fluctuation sequence of irradiance and temperature is as follows: ； in, G n,t is the random irradiance variable modeled by SDE; is the random ambient temperature variable modeled by SDE; dW t Represents the differential form of the Wiener process.

6. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: In step S4, the real-time photovoltaic output is calculated according to the photovoltaic cell characteristic equation P pv,t The specific process is: ； in, for t Solar photovoltaic power generation output power at all times; is the rated power of solar photovoltaic power generation (W); T cell,t is the temperature of the solar cell (°C); G STC =1000 W / m 2 is the standard test irradiance; T STC is the standard test temperature; γ is the temperature coefficient, -0.05< γ<- 0.003; T NOC is the nominal operating temperature; is the actual electric power of solar photovoltaic power generation at time t; is the solar photovoltaic power conversion efficiency; The two-state Markov model is used to simulate the traditional unit failure and load fluctuation formula as follows: ； in, TTF is the duration of downtime or availability of generating equipment; TTR It is the time of repair or the time when the power generation equipment is unavailable; λ is the failure rate of each power generation unit in the two-state Markov model; r is the average repair time of each power generation unit, equal to ; is the repair rate of each power generation unit in the two-state Markov model; U is a random value uniformly distributed in [0,1], Z is a random value from a standard normal distribution; is the variance of the repair time of each generator, divided by 0.1 r express; L t for t The load size at any moment; for t The predicted value of load size at each moment; α is the load forecast error coefficient; Z is a random number that follows a standard normal distribution. Z ~ N (0,1).

7. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: Step S5 is specifically as follows: ； in, LOLE is the system load loss expectation; d ( P t < L t ) is the duration of time when the total system output is less than the load; T is the total simulation time.

8. The photovoltaic power generation capacity credibility assessment method based on joint probability modeling and random process according to claim 1 is characterized in that: Step S6 is as follows: First, calculate the benchmark LOLE , maintaining photovoltaic installed capacity C pv and the original traditional unit capacity C conv,initial The photovoltaic output time series data is generated based on steps S1 to S5; based on step S4, the random fault state TTF and TTR of the traditional unit are simulated to generate the unit available capacity sequence P conv,t ; Then, the actual power of solar photovoltaic power generation at time t is superimposed to calculate the total output of the system P t ,in: P t =P conv,t + P e,pv,t ; Finally, based on step S5, calculate the benchmark LOLE value LOLE base ; Next, the equivalent traditional capacity is iteratively calculated, the photovoltaic system is removed, the step size ΔC is set to 1 MW, and the capacity of the traditional unit is gradually increased; in the kth iteration, the capacity of the traditional unit is updated to C conv,k , ; Based on step S5, calculate the new LOLE value LOLE new,k , when the condition is met | LOLE new,k − LOLE base ∣≤ The iteration is terminated when is the tolerance threshold, =0.09 hours / year; finally record the final equivalent traditional capacity C conv，end ; Finally, the reliability of photovoltaic power generation capacity is calculated CC pv , the formula is: 。 9. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: When the processor executes the program, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images