Reactive power optimization method for power distribution network with high-proportion new energy access

Abstract

The invention discloses a reactive power optimization method for a power distribution network with high-proportion new energy access, and the method comprises the steps: 1, building a wind power and photovoltaic output probability distribution model based on the randomness characteristics of wind speed and illumination data; 2, performing discretization and randomness analysis on the wind power and photovoltaic output probability distribution model established in the step 1 in combination with a traditional analytical method and a probability formula; step 3, establishing a reactive power optimization model of the power distribution network containing high-proportion new energy access based on the wind and light output combination distribution data processed in the step 2; and 4, solving the power distribution network reactive power optimization model containing high-proportion new energy access established in the step 3 by adopting an ant colony algorithm ACA and a differential evolution algorithm DEA, and obtaining a final reactive power optimization result in combination with the wind and light output probability distribution characteristics in the step 2. The method is suitable for reactive power optimization of the power distribution network with high-proportion new energy access, and is of great significance to reduction of active power loss and voltage offset.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, in particular to the reactive power optimization technology of power grids, and specifically to a reactive power optimization method for a distribution network with a high proportion of new energy access. Background Art

[0002] With the rapid development of new energy sources such as wind and solar, new energy will gradually replace traditional thermal power units, prompting the traditional power grid to gradually develop into a power grid with a high proportion of wind and solar new energy access. This will also increase the reactive power regulation pressure and complexity of the power grid, and the increase in node voltage deviation and system network loss are two main problems to be solved currently. Similar to the reactive power optimization problem of traditional power grids, the reactive power optimization of power grids considering the participation of wind and solar new energy in regulation is also a complex multi-objective optimization problem with non-linearity, non-convexity, and discrete optimization variables. However, there are various deficiencies in dealing with the distribution network with a high proportion of new energy access by traditional reactive power optimization methods. Therefore, it is crucial to study a new reactive power optimization method for the distribution network with a high proportion of new energy access.

[0003] The patent document with the application publication number CN110445127A discloses a reactive power optimization method for a distribution network facing multiple stochastic uncertainties. This method aims to minimize the active power loss, and there are equality constraints and inequality constraints at the same time. It considers the randomness of the distribution network load, the randomness of the output of distributed power sources, and the randomness of reactive power compensation devices, achieving the effect of more accurately reflecting the actual situation of the distribution network. The patent document with the application publication number CN109768573A discloses a reactive power optimization method for a distribution network based on a multi-objective differential grey wolf algorithm. This method considers the time-series volatility between photovoltaic and load. By introducing a static var generator as a compensation device into the active distribution network, according to the change of the equivalent load after the time-series fluctuating photovoltaic and load are connected to the distribution network, the dynamic reactive power of the static var generator is smoothly changed. Under the condition of the minimum output reactive power compensation capacity, the active power loss and voltage deviation are maximally reduced. At the same time, this method improves the grey wolf algorithm and introduces mutation, crossover, and fast non-dominated sorting to process multiple objectives.

[0004] Neither of the above two methods simultaneously considers the impact of wind and photovoltaic distributed power sources on the distribution network after being connected to the distribution network, resulting in the optimized results above being not very applicable to the current rapid development of new energy. At the same time, the above two methods do not perform probabilistic power flow equations, which easily makes the results inaccurate due to the randomness of the output of distributed power sources. To achieve this goal, the method proposed by the present invention simultaneously considers the impact of wind and photovoltaic distributed power sources on the distribution network, establishes a probability distribution model of wind and photovoltaic output, combines the traditional analytical method and processes the probability distribution model of wind and photovoltaic output based on probability formulas, and finally makes the reactive power optimization results more persuasive. Summary of the Invention

[0005] The object of the present invention is to address the technical problems in the above-mentioned existing technologies, which do not fully consider the impact of the simultaneous access of wind power and photovoltaic distributed power sources on the distribution network, lack probabilistic power flow calculation resulting in inaccurate results, and may deviate from the optimal solution without re-optimizing the optimal solution set. A reactive power optimization method for a distribution network with a high proportion of new energy access is proposed.

[0006] To achieve the above object of the invention, the technical solutions proposed by the present invention are as follows, including the following steps:

[0007] A reactive power optimization method for a distribution network with a high proportion of new energy access, including the following steps:

[0008] Step 1: Based on the randomness characteristics of wind speed and light intensity data, establish a probability distribution model of wind power and photovoltaic output, providing initial random data input for Step 2 to further optimize the model accuracy;

[0009] Step 2: Use the probability distribution models of wind power and photovoltaic output established in Step 1, and combine traditional analytical methods and probability formulas to discretize and analyze the randomness, providing optimized combined distribution data of wind and photovoltaic output for Step 3;

[0010] Step 3: Based on the combined distribution data of wind and photovoltaic output processed in Step 2, establish a reactive power optimization model for a distribution network with a high proportion of new energy access;

[0011] Step 4: Use the Ant Colony Algorithm (ACA) and Differential Evolution Algorithm (DEA) to solve the reactive power optimization model for a distribution network with a high proportion of new energy access established in Step 3, and combine the probability distribution characteristics of wind and photovoltaic output in Step 2 to obtain the final reactive power optimization result;

[0012] The above 4 steps can well optimize the reactive power of a distribution network with a high proportion of new energy access;

[0013] In Step 1, when establishing the probability distribution model of wind power and photovoltaic output, the following steps are adopted:

[0014] Step 1-1: Establish a probability distribution model of wind power output

[0015] Given the wind speed v, the mathematical model of the wind speed and output power of the wind turbine is:

[0016]

[0017] Where: v a 、v b 、v cThey are the cut-in wind speed, rated wind speed, and cut-out wind speed respectively; t1 = 1 / [P r (v b -v c )]; t2 = -t1v c ; P r is the rated power of the wind turbine.

[0018] The two-parameter Weibull distribution model is adopted to reflect the actual change of wind speed, and the actual output power is obtained:

[0019]

[0020] In the formula: α and β are the shape parameter and scale parameter of the Weibull distribution respectively.

[0021] Step 1-2: Establish a probability distribution model for photovoltaic output

[0022] Based on the illumination data, the shape parameter and size parameter of the Beta distribution are obtained, and a probability distribution model for photovoltaic output is established based on the shape parameter m and size parameter n of the Beta distribution:

[0023]

[0024] In the formula: m and n are the shape parameter and size parameter of the Beta distribution respectively; P is the actual output power of the photovoltaic system; P max is the maximum output power of the photovoltaic system; f is the probability density function of the illumination intensity.

[0025] In step 2, when performing discretization and randomness analysis by combining the traditional analytical method and probability formula, the following steps are adopted:

[0026] Step 2-1: The Weibull distribution of wind power generation output and the Beta distribution of photovoltaic power generation output are equivalently converted into discrete distributions by discrete sampling;

[0027] Step 2-2: The equivalent discrete distributions are combined to obtain the combined discrete random distribution of all wind-solar outputs, and its randomness is described by the corresponding probability;

[0028] Step 2-3: Calculate the conditional probability flow once under each combined state of output discretization, and consider the load randomness. At this time, the wind-solar output is a definite value, and only the load that satisfies the normal distribution in the system is a random variable;

[0029] Step 2-4: According to the state variable x of an event under the discrete state a n of the output combination, calculate the expectation and standard deviation of the probability flow considering load fluctuations, which are used to represent the event under the discrete state a nNext, consider the cumulative probability results of the probabilistic power flow considering load randomness;

[0030] Step 2-5: Use the total probability formula to combine all conditional probabilistic power flow results into the probabilistic power flow of the system comprehensively considering the random fluctuations of the wind and light output.

[0031] In Step 3, based on the combined distribution data of the wind and light output processed in Step 2, establish a reactive power optimization model for the distribution network with a high proportion of new energy access. The model includes an objective function aiming at minimizing the system network loss and voltage deviation, equality constraints, and inequality constraints.

[0032] In Step 4, use the ant colony algorithm ACA and differential evolution algorithm DEA to solve the reactive power optimization model for the distribution network with a high proportion of new energy access established in Step 3. Combining the probabilistic distribution characteristics of the wind and light output in Step 2, obtain the final reactive power optimization result. The specific steps are as follows:

[0033] Step 4-1: Set the initial parameters of ACA and the parameter values of DEA, and initialize the population according to the Weibull distribution parameters k, λ and Beta distribution parameters m, n in Step 1;

[0034] Step 4-2: Initialize the ant population and start iteration;

[0035] Step 4-3: Select the nodes passing through each layer for the current ant according to the parameters, and update the information and calculate the initial flow;

[0036] Step 4-4: Introduce DEA for the second iteration, mutate and cross each group of pheromones, combine the probabilistic distribution characteristics of the wind and light output in Step 2 to update the power flow calculation, optimize the objective function of the system network loss and voltage deviation, and obtain the optimal function value;

[0037] Step 4-5: All ant groups construct their respective optimal paths according to the pheromones, and obtain the pheromones with better results through comparison;

[0038] Step 4-6: Update the pheromones of each group of ants and pass them to the next generation;

[0039] Step 4-7: Return to Step 4-5 until all ant groups have completed the calculation;

[0040] Step 4-8: Determine the current optimal path and its length;

[0041] Step 4-9: Return to Step 4-4, calculate the next generation until the termination condition is met;

[0042] Step 4-10: Under the condition of not exceeding the iteration constraint and meeting the accuracy requirement, determine the global optimal path and its length until the optimal function value is found. If the number of times exceeds the iteration constraint, output the result.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0044] 1. In view of the randomness of wind power and photovoltaic power output, the present invention proposes a wind power output probability distribution model that follows the Weibull distribution and a photovoltaic power output probability distribution model that follows the Beta distribution. Combining with the traditional analytical method and based on the probability formula, the wind-solar power output probability distribution model is processed, and finally a reactive power optimization model with the minimum active power loss and minimum voltage deviation of the power grid as the objectives is established;

[0045] 2. When solving the reactive power optimization model of the distribution network with wind-solar access, the present invention combines the ant colony algorithm ACA and the differential evolution algorithm DEA to solve the reactive power optimization model. When there are defects in the solution of the reactive power optimization model by some algorithms, it can be improved by another algorithm. Therefore, organically combining the two algorithms is a very good method. The ant colony optimization algorithm performs heuristic search based on probability selection, while the differential evolution algorithm has a fast convergence speed and strong search ability, which is a powerful supplement to the ant colony algorithm.

[0046] BRIEF DESCRIPTION OF THE DRAWINGS The following further describes the present invention in conjunction with the drawings and embodiments:

[0047] FIG. is the method flow chart of the present invention;

[0048] Figure 1 FIG. is the method flow chart for processing the wind-solar power output probability distribution model proposed by the present invention;

[0049] Figure 2 FIG. is the ACA and DEA for solving the reactive power optimization flow chart of the distribution network with high proportion of new energy access proposed by the present invention;

[0050] Figure 3 FIG. is the schematic diagram of the IEEE 39-node system in the embodiment of the present invention.

[0051] Figure 4 FIG. is the schematic diagram of the IEEE 39-node system in the embodiment of the present invention. SPECIFIC IMPLEMENTATION METHOD

[0053] As Figure 1 shown, a reactive power optimization method for a distribution network with high proportion of new energy access is as follows:

[0054] Step 1: Based on the randomness characteristics of wind speed and light data, establish a wind power and photovoltaic power output probability distribution model, and provide initial random data input for Step 2 to further optimize the model accuracy;

[0055] Step 2: Utilize the wind power and photovoltaic power output probability distribution models established in Step 1, and combine the traditional analytical method and the probability formula to discretize and analyze the randomness thereof, and provide optimized wind-solar power output combined distribution data for Step 3;

[0056] Step 3: Based on the combined distribution data of wind and light output processed in Step 2, establish a reactive power optimization model for a distribution network with a high proportion of new energy access;

[0057] Step 4: Use ACA and DEA to solve the reactive power optimization model for the distribution network with a high proportion of new energy access established in Step 3, and combine the probability distribution characteristics of wind and light output in Step 2 to obtain the final reactive power optimization result;

[0058] In Step 1, when establishing the probability distribution models of wind power and photovoltaic output, the following steps are adopted:

[0059] Step 1-1: Establish a probability distribution model of wind power output

[0060] Given the wind speed v, the mathematical model of the wind speed and output power of the wind turbine is:

[0061]

[0062] In the formula: v a , v b , v c are the cut-in wind speed, rated wind speed and cut-out wind speed respectively; t1 = 1 / [P r (v b -v c )]; t2 = -t1v c ; P r is the rated power of the wind turbine.

[0063] Adopt a two-parameter Weibull distribution model to reflect the actual change of wind speed and obtain the actual output power:

[0064]

[0065] In the formula: α and β are the shape parameter and scale parameter of the Weibull distribution respectively.

[0066] Step 1-2: Establish a probability distribution model of photovoltaic output

[0067] Obtain the shape parameter and size parameter of the Beta distribution according to the illumination data, and establish a probability distribution model of photovoltaic output based on the shape parameter m and size parameter n of the Beta distribution:

[0068]

[0069] In the formula: m and n are the shape parameter and size parameter of the Beta distribution respectively; P is the actual output power of the photovoltaic system; P max is the maximum output power of the photovoltaic system; f is the probability density function of the illumination intensity.

[0070] In Step 2, when performing discretization and randomness analysis by combining the traditional parsing method and probability formula, the following steps are adopted:

[0071] Step 2-1: Equivalent the Weibull distribution of wind power output and the Beta distribution of photovoltaic power output to discrete distributions by discrete sampling method;

[0072] Step 2-2: Combine the equivalent discrete distributions to obtain the combined discrete random distribution of all wind-solar power outputs, and describe its randomness with corresponding probabilities;

[0073] Step 2-3: Calculate the conditional probability flow once under each combined state of output discretization, and consider the load randomness. At this time, the wind-solar power output is a definite value, and only the load that satisfies the normal distribution in the system is a random variable;

[0074] Step 2-4: According to the state variable x of an event under the discrete state a of output combination n Calculate the expectation and standard deviation obtained from the probability flow considering load fluctuations, and use them to represent the cumulative probability result of the probability flow considering load randomness under the discrete state a of output combination n ;

[0075] Step 2-5: Use the total probability formula to combine all conditional probability flow results into the probability flow of the system comprehensively considering the randomness fluctuations of wind-solar power outputs.

[0076] In Step 3, based on the combined distribution data of wind-solar power outputs processed in Step 2, establish a reactive power optimization model for a distribution network with a high proportion of new energy access. The model includes an objective function, equality constraints, and inequality constraints with the minimum of system power loss and voltage deviation as the objectives. The specific model is as follows:

[0077] 1) Objective function

[0078]

[0079] Where: P loss is the active power loss; ΔU is the voltage deviation; n is the number of system nodes; G ij is the branch admittance between nodes i and j; U i and U j are the voltage amplitudes of nodes i and j respectively; θ ij is the voltage phase difference between nodes i and j; U i ′, U imax and U imin are the rated voltage of node i, the maximum node voltage, and the minimum node voltage respectively.

[0080] 2) Equality constraints

[0081]

[0082] Where: N is the node number; P Gi , P Li are the active output power of the generator and the active power of the load at node i; Q Gi , Q Li , Q ci are the reactive power input of the generator, the reactive power of the load, and the capacity of the capacitive reactive compensation device at node i, respectively; G ij and B ij are the real part and the imaginary part of the element of the nodal admittance matrix, respectively; θ ij is the phase difference between node i and node j.

[0083] 3) Inequality constraints

[0084]

[0085] Where: Q Gi is the reactive power of the generator; Q Lm is the reactive power of the branch; Q cj is the shunt capacitor compensation capacity; V Dl is the load bus voltage.

[0086] In step 4, the ant colony algorithm ACA and the differential evolution algorithm DEA are used to solve the reactive power optimization model of the distribution network with a high proportion of new energy access established in step 3. Combining with the probability distribution characteristics of the wind and light output in step 2, the final reactive power optimization result is obtained. The specific steps are as follows:

[0087] Step 4-1: Set the initial parameters of ACA and the parameter values of DEA, and initialize the population according to the Weibull distribution parameters k, λ and the Beta distribution parameters m, n in step 1;

[0088] Step 4-2: Initialize the ant population and start iteration;

[0089] Step 4-3: Select the nodes passing through each layer for the current ant according to the parameters, and update the information and calculate the initial flow;

[0090] Step 4-4: Introduce DEA for the second iteration, mutate and cross each group of pheromones, update the power flow calculation combined with the probability distribution characteristics of the wind and light output in step 2, and optimize the objective function of the system network loss and voltage deviation to obtain the optimal function value;

[0091] Step 4-5: All ant groups construct their respective optimal paths according to the pheromones, and compare to obtain the pheromones with better results;

[0092] Step 4-6: Update the pheromones of each group of ants and pass them to the next generation;

[0093] Step 4-7: Return to Step 4-5 until all ant groups have been calculated;

[0094] Step 4-8: Determine the current optimal path and its length;

[0095] Step 4-9: Return to Step 4-4 and calculate the next generation until the termination condition is met;

[0096] Step 4-10: Determine the global optimal path and its length without exceeding the iteration constraint and meeting the accuracy requirement until the optimal function value is found. If the number of times exceeds the iteration constraint, output the result. Embodiment

[0097] Based on the IEEE 39-bus distribution network model, the present invention uses the existing ant colony algorithm ACA and the algorithm ACA-DEA combining the ant colony algorithm ACA and the differential evolution algorithm DEA to solve the reactive power optimization model of the distribution network with distributed photovoltaic access, and adds wind turbines and photovoltaic generators to the system. As Figure 4 shown, the synchronous generator G2 is replaced with a wind turbine, and the synchronous generator G6 is replaced with a photovoltaic generator. The parameters of the wind turbine and the photovoltaic generator are shown in Table 1. The total system capacity is 1000 kW, and the capacities of the replaced wind and photovoltaic generators are adjusted so that the total new energy capacity accounts for 30%-50%. The two algorithms are iterated 100 times, and the results are shown in Table 2. It can be seen from Table 2 that for the IEEE 39-bus distribution system, the ACA-DEA algorithm is superior to ACA in reducing active power loss and voltage deviation, and has less average calculation time.

[0098] Table 1

[0099]

[0100] Table 1 lists the key parameters of the wind turbine and the photovoltaic generator. Among them, the rated power of the wind turbine is 50 kW, and the rated wind speed of 16 m / s is the best output point under typical operating conditions; the rated power of the photovoltaic generator is 55 kW, and the light intensity of 600 W / m 2 reflects the maximum output capacity under sunny weather. These parameters are set based on the operating data of actual new energy equipment to ensure the authenticity and reliability of the model.

[0101] Table 2

[0102]

[0103] Table 2 shows the comparison of the optimization results of the ACA and ACA-DEA algorithms in the IEEE 39-bus system. The active power loss of the ACA algorithm is 107.6 kW, the voltage deviation is 1.098 p.u., and the average calculation time is 145.68 seconds; while the ACA-DEA algorithm optimizes the search path by introducing the differential evolution mechanism, the active power loss is reduced to 91.4 kW, the voltage deviation is reduced to 1.028 p.u., and the average calculation time is shortened to 81.26 seconds. The results show that ACA-DEA is superior to the single ACA algorithm in reducing system losses, improving voltage stability and enhancing calculation efficiency.

[0104] Verified by the above embodiments, the method proposed by the present invention can effectively meet the reactive power optimization requirements of high-proportion new energy access to the distribution network. Combining the probability distribution model of wind and light output and the ACA-DEA optimization algorithm, this method significantly reduces the active power loss and voltage deviation, while improving the calculation efficiency, and has strong engineering application value, which can provide technical support for the optimization of the distribution network under the background of the rapid development of new energy.

Claims

1. A reactive power optimization method for a distribution network with a high proportion of new energy access, characterized in that, It includes the following steps: Step 1: Based on the randomness characteristics of wind speed and light data, establish a probability distribution model for wind power and photovoltaic output, providing initial random data input for Step 2; Step 2: Use the probability distribution model of wind power and photovoltaic output established in Step 1 to discretize and analyze its randomness, providing optimized combined distribution data of wind and light output for Step 3; Step 3: Based on the combined distribution data of wind and light output processed in Step 2, establish a reactive power optimization model for a distribution network with a high proportion of new energy access; Step 4: Use the ant colony algorithm ACA and differential evolution algorithm DEA to solve the reactive power optimization model for a distribution network with a high proportion of new energy access established in Step 3, and combine with the probability distribution characteristics of wind and light output in Step 2 to obtain the final reactive power optimization result.

2. The method according to claim 1, wherein: In Step 1, when establishing the probability distribution model of wind power and photovoltaic output, the following steps are adopted: Step 1-1: First, establish a probability distribution model for wind power output; Then, use a two-parameter Weibull distribution model to reflect the actual change of wind speed and obtain the actual output power; Step 1-2: Establish a probability distribution model for photovoltaic output.

3. The method according to claim 1, characterized in that: In Step 2, when performing discretization and randomness analysis, the following steps are adopted: Step 2-1: Equivalent the Weibull distribution of wind power generation output and the Beta distribution of photovoltaic power generation output to a discrete distribution by discrete sampling; Step 2-2: Combine the equivalent discrete distributions to obtain the combined discrete random distribution of all wind and light outputs, and describe its randomness with corresponding probabilities; Step 2-3: Calculate the conditional probability flow once under each combined state of output discretization, and consider the load randomness. At this time, the wind and light output is a definite value, and only the load that satisfies the normal distribution in the system is a random variable; Step 2-4: Calculate the expectation and standard deviation obtained from the probabilistic power flow considering load fluctuations based on the state variable x of a certain event, and use them to represent the cumulative probability result of the probabilistic power flow considering load randomness under the discrete output combination state a n ; n ​ Step 2-5: Use the total probability formula to combine all conditional probability flow results into the probability flow of the system considering the randomness fluctuation of wind and light output.

4. The method according to claim 1, characterized in that: In Step 3, based on the combined distribution data of wind and light output processed in Step 2, establish a reactive power optimization model for a distribution network with a high proportion of new energy access. The model includes an objective function, equality constraints, and inequality constraints with the minimum of system network loss and voltage deviation as the goal.

5. The method according to claim 1, wherein: In Step 4, use the ant colony algorithm ACA and differential evolution algorithm DEA to solve the reactive power optimization model for a distribution network with a high proportion of new energy access established in Step 3, and combine with the probability distribution characteristics of wind and light output in Step 2 to obtain the final reactive power optimization result. The specific steps are as follows: Step 4-1: Set the initial parameters of ACA and the parameter values of DEA, and initialize the population according to the Weibull distribution parameters k, λ in Step 1 and the Beta distribution parameters m, n; Step 4-2: Initialize the ant population and start iteration; Step 4-3: Select the nodes passing through each layer search for the current ant according to the parameters, and update the information and calculate the initial flow; Step 4-4: Introduce DEA for the second iteration, mutate and cross each group of pheromones, combine with the probability distribution characteristics of wind and light output in Step 2 to update the power flow calculation, optimize the objective function of system network loss and voltage deviation, and obtain the optimal function value; Step 4-5: All ant groups construct their respective optimal paths based on pheromones and obtain pheromones with better results through comparison; Step 4-6: Update the pheromones of each group of ants and pass them to the next generation; Step 4-7: Return to Step 4-5 until all ant groups have completed the calculation; Step 4-8: Determine the current optimal path and its length; Step 4-9: Return to Step 4-4 and calculate the next generation until the termination condition is met; Step 4-10: Determine the global optimal path and its length without exceeding the iteration constraint and meeting the accuracy requirement until the optimal function value is found; if the number of times exceeds the iteration constraint, output the result.

6. The method according to claim 4, characterized in that: The objective function is specifically: Where: P loss is the active power loss; ΔU is the voltage deviation; n is the number of system nodes; G ij is the branch admittance between nodes i and j; U i and U j are the voltage amplitudes of nodes i and j respectively; θ ij is the voltage phase difference between nodes i and j; U′ i , U imax and U imin are the rated voltage of node i, the maximum node voltage and the minimum node voltage respectively.

7. The method according to claim 4, wherein: The equality constraint is specifically: Where: N is the node number; P Gi and P Li are the active output power of the generator and the active power of the load at node i; Q Gi and Q Li and Q ci are respectively the reactive power input of the generator, the reactive power of the load, and the capacity of the capacitive reactive power compensation device at node i; G ij and B ij are respectively the real part and the imaginary part of the elements of the node admittance matrix; θ ij is the phase difference between node i and node j.

8. The method according to claim 4, wherein: The inequality constraint is specifically: Where: Q Gi is the reactive power of the generator; Q Lm is the branch reactive power; Q cj is the shunt capacitor compensation capacity; V Dl is the load bus voltage.

9. The method according to claim 2, wherein: In Step 1-1, the specifically established probability distribution model of wind power output is: Given the wind speed v, the mathematical model of the wind speed and output power of the wind turbine is: where: v a , v b , v c are the cut-in wind speed, rated wind speed and cut-out wind speed respectively; t1 = 1 / [P r (v b - v c )]; t2 = -t1v c ; P r is the rated power of the wind turbine; A two-parameter Weibull distribution model is used to reflect the actual change of the wind speed and obtain the actual output power: In the formula: α and β are the shape parameter and scale parameter of the Weibull distribution respectively.

10. The method according to claim 2 or 9, characterized in that: In Step 1-2, the specifically established probability distribution model of photovoltaic output is: Obtain the shape parameter and size parameter of the Beta distribution based on the illumination data, and establish a probability distribution model of photovoltaic output based on the shape parameter m and size parameter n of the Beta distribution: Where: m and n are the shape parameter and the scale parameter of the Beta distribution respectively; P is the actual output power of the photovoltaic system; P max is the maximum output power of the photovoltaic system; f is the probability density function of the light intensity.

Images