Optimized scheduling method and device of wind-solar-storage integrated system, terminal equipment and storage medium

Abstract

The invention discloses an optimal scheduling method and device for a wind-solar-storage integrated system, terminal equipment and a storage medium, and the method comprises the steps: firstly solving a constructed initial scheduling model, taking a target function value corresponding to an optimal solution of the initial scheduling model as a reference value, and obtaining an optimal solution of the initial scheduling model according to the initial scheduling model, a preference factor and an uncertainty coefficient; a target scheduling model which aims at minimizing the negative influence of the power uncertainty on the system and enabling the power offset of the system to be smaller than the target power offset is constructed, and the influence of the power uncertainty on the risk or benefit of the system is quantified by a preference factor and an uncertainty coefficient. Therefore, when the target scheduling model is solved to generate the final scheduling variable, the scheduling variable can ensure that the offset of the power is kept within the power offset limited by the preference factor, and the power uncertainty conforms to the range of the system preference requirement, so that the system power fluctuation is effectively reduced, and the system performance is improved. And the stability and safety of the power grid are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of system optimization and scheduling, and in particular to an optimization and scheduling method, device, terminal equipment and storage medium for a wind-solar-storage integrated system. Background Art

[0002] The wind, solar and energy storage integrated system refers to a new energy system that integrates wind energy, solar energy and energy storage technology. Wind energy is converted into electrical energy through wind turbines, solar energy is converted into electrical energy through solar panels, and energy storage devices (such as battery energy storage systems) are used to provide a stable power supply when wind and solar energy are insufficient to ensure the continuous operation of the system. In the wind, solar and energy storage integrated system, the coordinated scheduling of multiple devices is crucial. Wind energy and solar energy are naturally complementary, and energy storage devices play the role of "peak shaving and valley filling" in the system.

[0003] Since the output power of wind turbines and solar panels often fluctuates, the power generation of wind and solar energy is volatile and intermittent, which may lead to unstable power supply of the system. In order to deal with these volatility and intermittency problems, it is often necessary to optimize the power generation output of various energy sources and the charging and discharging power of energy storage devices through multi-device coordinated scheduling to ensure stable power supply and efficient operation of the system. However, when the existing technology performs multi-device coordinated scheduling of the system, it is often based on big data analysis of wind and solar output data over the years to predict future power generation trends, and formulate scheduling strategies based on these prediction results. Although it can reflect the cyclical law of power generation to a certain extent, it is difficult to accurately respond to sudden power fluctuations, and lacks consideration of power uncertainty, resulting in the inability to cope with voltage and frequency fluctuations during the energy scheduling and power control of the system, thereby affecting the stability and security of the power grid. Summary of the invention

[0004] The embodiments of the present invention provide an optimization scheduling method, apparatus, terminal device and storage medium for an integrated wind, solar and storage system. The preference factors and uncertainty coefficients introduced enable the target scheduling variables finally generated to effectively reduce power fluctuations and reduce the risk of instability in the system caused by excessive power deviations. This can effectively solve the problem in the prior art that it is difficult to accurately respond to sudden power fluctuations, thereby affecting the stability and security of the power grid.

[0005] An embodiment of the present invention provides an optimization scheduling method for a wind-solar-storage integrated system, including:

[0006] According to the power data of the wind-solar-storage integrated system in different scenarios and the average power values ​​in different scenarios, an initial scheduling model is constructed with the goal of minimizing the average absolute deviation of the total system output power in all scenarios; wherein the constraints of the initial scheduling model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints, and energy storage power station charge state constraints;

[0007] Solving the initial scheduling model, and outputting the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is the smallest;

[0008] Taking the objective function value as a reference value, and based on the initial scheduling model, the preference factor and the uncertainty coefficient, a target scheduling model is constructed with the goal of minimizing the negative impact of power uncertainty on the system and the power offset of the system being less than the target power offset; wherein the preference factor is used to characterize the risk tolerance of the system when facing power uncertainty; the uncertainty coefficient is used to characterize the impact of power uncertainty on the system; and the target power offset is the power offset defined by the preference factor;

[0009] Under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, the target scheduling model is solved, and when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset, the corresponding target scheduling variables are generated; wherein the target scheduling variables include: target energy storage power station charging power and target energy storage power station discharging power;

[0010] According to the target dispatching variable, the output of the energy storage power station in the wind-solar-storage integrated system is controlled.

[0011] Preferably, the target scheduling model includes: a risk aversion model; the preference factors include:

[0012] A risk aversion factor for characterizing the degree of risk aversion of the system when facing power uncertainty; the uncertainty coefficient includes: a first uncertainty coefficient for characterizing the degree of influence of power uncertainty on system risk; the target power offset includes: a maximum power offset limited by the risk aversion factor;

[0013] The construction of the risk aversion model includes:

[0014] Taking the objective function value as a reference value, and based on the initial scheduling model, the risk aversion factor and the first uncertainty coefficient, a risk aversion model is constructed with the goal of maximizing the risk impact of power uncertainty on the system and ensuring that the power offset does not exceed the maximum power offset.

[0015] Preferably, the target scheduling model includes: a risk preference model; the preference factors include:

[0016] A risk preference factor for characterizing the degree of risk preference of the system in balancing risk and benefit when facing power uncertainty; the uncertainty coefficient includes: a second uncertainty coefficient for characterizing the degree of influence of power uncertainty on system benefit; the target power offset includes: a minimum power offset defined by the risk preference factor;

[0017] The construction of the risk preference model includes:

[0018] Taking the objective function value as a reference value, and based on the initial scheduling model, the risk preference factor and the second uncertainty coefficient, a risk preference model is constructed with the goal of minimizing the impact of power uncertainty on system revenue and ensuring that the power offset is less than the minimum power offset.

[0019] Preferably, the target scheduling variable includes: a first scheduling variable, or a second scheduling variable;

[0020] The target scheduling model is solved to generate a corresponding target scheduling variable when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset, including:

[0021] When the preference factor is determined to be a risk aversion factor, the risk aversion model is solved under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and the first dispatch variable is generated when the power uncertainty has the greatest impact on the risk of the system and the power offset does not exceed the maximum power offset; wherein the first dispatch variable includes: the charging power of the first energy storage power station and the discharging power of the first energy storage power station;

[0022] When the preference factor is determined to be a risk preference factor, the risk preference model is solved under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and a second scheduling variable is generated when the impact of power uncertainty on system revenue is minimal and the power offset is less than the minimum power offset; wherein the second scheduling variable includes: the charging power of the second energy storage power station and the discharging power of the second energy storage power station.

[0023] Preferably, the risk aversion model includes:

[0024] maxα r ;

[0025] maxF 1 ≤(1+βr )F 0 ;

[0026]

[0027] Among them, β r is the risk aversion factor, F 0 is the objective function value corresponding to the initial scheduling model when the average absolute deviation of the total system output power in all scenarios is the smallest, P L,t,s is the sum of the output power of the wind-solar-storage integrated system in the sth scenario during the tth period, P L,av,s is the average power within the scheduling period T in the sth scenario, is the probability of clustering the sth scene, Ω s is the cluster center set, α r is the first uncertainty coefficient;

[0028] P equip,t is the preset total demand power corresponding to the sum of the user power consumption and the distributed photovoltaic power generation power in period t, The predicted total power demand between user power consumption and distributed photovoltaic power generation power; is the power demand model, used to characterize the predicted total power demand The power demand model is used to solve the uncertainty coefficient α in the first degree of uncertainty. r The corresponding predicted total power demand value is:

[0029] P Load,t is the power consumption of the user during period t, P dispv,t is the distributed photovoltaic power generation power in period t, P wt,t,s is the output of the wind farm in the sth scenario during the t period, P pv,t,s is the output of the photovoltaic power station in the sth scenario during the tth period, P essdis,t,s is the charging power of the energy storage station in the sth scenario during the t period, P essch,t,s is the discharge power of the energy storage power station in the sth scenario during the t period;

[0030] The risk preference model includes:

[0031] minα c ;

[0032] f 2 ≤(1-β c )F 0 ;

[0033]

[0034] Among them, β cis the risk preference factor, α c is the second uncertainty coefficient.

[0035] Preferably, the process of generating power data in different scenarios includes:

[0036] Obtain historical data on user power consumption, distributed photovoltaic power generation, wind farm output forecast deviation, and photovoltaic power station output forecast deviation in different scenarios;

[0037] Fitting each output forecast deviation historical data of the wind farm and each output forecast deviation historical data of the photovoltaic power station respectively to generate a first probability density function for characterizing the degree of output deviation of the wind farm and a second probability density function for characterizing the degree of output deviation of the photovoltaic power station;

[0038] A joint distribution probability model for characterizing the correlation between the prediction deviations of the wind farm and the photovoltaic power station is constructed according to the first probability density function, the second probability density function and the correlation coefficient; wherein the correlation coefficient is used to characterize the correlation between the predicted output deviations of the wind farm and the photovoltaic power station;

[0039] Based on the first probability density function and the second probability density function, the joint distribution probability model is solved to generate wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios;

[0040] Based on the wind farm output prediction deviation data, photovoltaic power station output prediction deviation data, user power consumption and distributed photovoltaic power generation power under different scenarios, power data corresponding to different scenarios are generated.

[0041] Preferably, the generating of wind farm output forecast deviation data corresponding to different scenarios and photovoltaic power station output forecast deviation data corresponding to different scenarios includes:

[0042] Randomly generate samples to be selected corresponding to the joint distribution probability model;

[0043] Sort the samples to be selected in ascending order to generate a corresponding column vector; wherein the first and last ends of the column vector are respectively the preset endpoint values ​​corresponding to the first probability density function and the second probability density function;

[0044] Divide the column vector into a plurality of intervals, and construct a spline polynomial corresponding to each interval based on a spline interpolation method;

[0045] Substituting each candidate sample into the corresponding spline polynomial to generate a predicted output deviation value of the wind farm and a predicted output deviation value of the photovoltaic power station;

[0046] The predicted output deviation values ​​of the wind farm and the predicted output deviation values ​​of the photovoltaic power station are clustered according to different scenarios to generate wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios.

[0047] Based on the above method embodiments, the present invention provides corresponding device embodiments.

[0048] An embodiment of the present invention provides an optimization scheduling device for a wind-solar-storage integrated system, comprising: an initial scheduling model construction module, a first solution module, a target scheduling model construction module, a second solution module, and an output control module;

[0049] The initial scheduling model construction module is used to construct an initial scheduling model with the goal of minimizing the average absolute deviation of the total system output power in all scenarios according to the power data of the wind-solar-storage integrated system in different scenarios and the power average values ​​in different scenarios; wherein the constraints of the initial scheduling model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints, and energy storage power station charge state constraints;

[0050] The first solving module is used to solve the initial scheduling model and output the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is the smallest;

[0051] The target scheduling model construction module is used to take the target function value as a reference value, and according to the initial scheduling model, the preference factor and the uncertainty coefficient, construct a target scheduling model with the goal of minimizing the negative impact of power uncertainty on the system and the power offset of the system being less than the target power offset; wherein the preference factor is used to characterize the risk tolerance of the system when facing power uncertainty; the uncertainty coefficient is used to characterize the impact of power uncertainty on the system; the target power offset is the power offset defined by the preference factor;

[0052] The second solving module is used to solve the target scheduling model under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and generate corresponding target scheduling variables when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset; wherein the target scheduling variables include: target energy storage power station charging power and target energy storage power station discharging power;

[0053] The output control module is used to control the output of the energy storage power station in the wind-solar-storage integrated system according to the target scheduling variable.

[0054] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0055] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the optimization scheduling method of the wind, solar and storage integrated system described in the above-mentioned embodiment of the invention.

[0056] Based on the above method embodiments, the present invention provides corresponding storage medium item embodiments.

[0057] Another embodiment of the present invention provides a storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute an optimization scheduling method for a wind-solar-storage integrated system as described in the above-mentioned embodiment of the invention.

[0058] The following beneficial effects are achieved by implementing the present invention:

[0059] The embodiment of the present invention provides an optimization scheduling method, device, terminal equipment and storage medium for a wind, photovoltaic and storage integrated system. The present invention first constructs an initial scheduling model corresponding to key constraints such as wind farm output, photovoltaic power station output and energy storage power station charging and discharging power based on the total power data and power average value in different scenarios, with the goal of minimizing the average absolute deviation of the total system output in all scenarios. The initial scheduling model is initially solved to obtain the objective function value corresponding to the minimum average absolute deviation of the total system output in all scenarios. Furthermore, in order to introduce power uncertainty considerations and the negative impact of power uncertainty on the system in the solution process of each output in the initial scheduling model, the present invention further uses the objective function value corresponding to the optimal solution of the initial scheduling model as a reference value, and constructs an initial scheduling model based on the impact of power uncertainty on the system according to the initial scheduling model, preference factors and uncertainty coefficients. The target scheduling model with the smallest negative impact and the power offset of the system being less than the target power offset as the target is schematically illustrated. In constructing the target scheduling model, a preference factor for characterizing the risk tolerance of the system in the face of power uncertainty and an uncertainty coefficient for characterizing the impact of power uncertainty on the system are introduced. Therefore, when solving the target scheduling model, the final scheduling variables or scheduling strategies can be automatically adjusted within a certain power uncertainty range. Moreover, the uncertainty coefficient quantifies the specific impact of power uncertainty on the system. Therefore, when solving the target scheduling model, the negative impact of power uncertainty on the system can be minimized while maximizing risk control and minimizing the negative impact on benefits. This ensures that the final scheduling variables, in which the offsets of each power are always kept within the power offset limited by the preference factor, and that the power uncertainty meets the range of the system preference requirements. Compared with the prior art, the present invention can further optimize the initial scheduling model based on the introduced preference factors and uncertainty coefficients, so that based on the formed new target scheduling model, the target scheduling variables finally generated can effectively reduce power fluctuations and reduce the instability risk of the system caused by excessive power deviation. The final scheduling scheme can effectively cope with the voltage and frequency fluctuations in the system, so that when the units of the energy storage power station are scheduled based on the target scheduling variables, the output of the energy storage power station can be reasonably scheduled to ensure the stability and safety of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flow chart of an optimization scheduling method for a wind-solar-storage integrated system provided by an embodiment of the present invention.

[0061] Figure 2 It is a structural flow chart of an operation scheduling method for a wind-solar-storage integrated system provided by another embodiment of the present invention.

[0062] Figure 3 It is a structural schematic diagram of an optimization scheduling device for a wind-solar-storage integrated system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] like Figure 1 FIG. 1 is a flow chart of an optimization scheduling method for a wind-solar-storage integrated system provided by an embodiment of the present invention. The optimization scheduling method for a wind-solar-storage integrated system includes:

[0065] Step S1: Based on the power data of the wind-solar-storage integrated system in different scenarios and the average power values ​​in different scenarios, an initial scheduling model is constructed with the goal of minimizing the average absolute deviation of the total system output power in all scenarios; wherein the constraints of the initial scheduling model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charge and discharge power constraints, and energy storage power station charge state constraints;

[0066] Step S2: Solving the initial scheduling model, and outputting the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is the smallest;

[0067] Step S3: Taking the objective function value as a reference value, and according to the initial scheduling model, the preference factor and the uncertainty coefficient, constructing a target scheduling model with the goal of minimizing the negative impact of power uncertainty on the system and the power offset of the system being less than the target power offset; wherein the preference factor is used to characterize the risk tolerance of the system in the face of power uncertainty; the uncertainty coefficient is used to characterize the impact of power uncertainty on the system; the target power offset is the power offset defined by the preference factor;

[0068] Step S4: Under the constraints of wind farm output, photovoltaic power station output, energy storage power station charge and discharge power, and energy storage power station charge state, the target scheduling model is solved, and when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset, the corresponding target scheduling variables are generated; wherein the target scheduling variables include: target energy storage power station charging power and target energy storage power station discharging power;

[0069] Step S5: Control the output of the energy storage power station in the wind-solar-storage integrated system according to the target scheduling variable.

[0070] For step S1, in a preferred embodiment, the present invention can obtain the power data of the wind, solar and storage integrated system in different scenarios and the average power in different scenarios, so as to construct an initial scheduling model based on the acquired data. The main purpose is to reduce the volatility of the total power delivered by the wind, solar and storage integrated system to the large power grid system by means of the charging and discharging characteristics of energy storage, and to reduce the impact of wind power and photovoltaic power grid connection on the power grid due to their randomness and volatility. Therefore, the initial scheduling model established in the embodiment of the present invention takes the minimum volatility of the total power delivered by the wind, solar and storage integrated system as the optimization goal, that is, the minimum average absolute deviation of the total power delivered by the system in all scenarios.

[0071] In a preferred embodiment, the power data in different scenarios include: wind farm output prediction deviation data in different scenarios, photovoltaic power station output prediction deviation data in different scenarios, and user power consumption and distributed photovoltaic power generation power in different scenarios.

[0072] Then, first obtain the historical data of user power consumption, distributed photovoltaic power generation, wind farm output forecast deviation and photovoltaic power station output forecast deviation under different scenarios;

[0073] Fitting each output forecast deviation historical data of the wind farm and each output forecast deviation historical data of the photovoltaic power station respectively to generate a first probability density function for characterizing the degree of output deviation of the wind farm and a second probability density function for characterizing the degree of output deviation of the photovoltaic power station;

[0074] A joint distribution probability model for characterizing the correlation between the prediction deviations of the wind farm and the photovoltaic power station is constructed according to the first probability density function, the second probability density function and the correlation coefficient; wherein the correlation coefficient is used to characterize the correlation between the predicted output deviations of the wind farm and the photovoltaic power station;

[0075] Based on the first probability density function and the second probability density function, the joint distribution probability model is solved to generate wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios;

[0076] Finally, based on the wind farm output prediction deviation data, photovoltaic power station output prediction deviation data, user power consumption and distributed photovoltaic power generation power under the above-mentioned different scenarios, power data corresponding to different scenarios can be generated for building models.

[0077] In a preferred embodiment, when generating wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios, it specifically includes:

[0078] Randomly generate samples to be selected corresponding to the joint distribution probability model;

[0079] Sort the samples to be selected in ascending order to generate a corresponding column vector; wherein the first and last ends of the column vector are respectively the preset endpoint values ​​corresponding to the first probability density function and the second probability density function;

[0080] Divide the column vector into a plurality of intervals, and construct a spline polynomial corresponding to each interval based on a spline interpolation method;

[0081] Substituting each candidate sample into the corresponding spline polynomial to generate a predicted output deviation value of the wind farm and a predicted output deviation value of the photovoltaic power station;

[0082] The predicted output deviation values ​​of the wind farm and the predicted output deviation values ​​of the photovoltaic power station are clustered according to different scenarios to generate wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios.

[0083] Specifically, the present invention constructs a corresponding joint distribution probability model through the wind farm output prediction deviation and the photovoltaic power station output prediction deviation, and generates the output prediction deviation data of the wind farm and photovoltaic power station under different scenarios, so as to more accurately capture the randomness and volatility of wind power and photovoltaic output. Since the correlation between wind power and photovoltaic output is taken into account, the accuracy of the overall prediction can be improved.

[0084] It is understandable that the embodiment of the present invention can ensure that the entire possible range of the joint distribution probability model is covered by randomly generating samples to be selected, thereby more comprehensively simulating the output deviation of wind farms and photovoltaic power stations under different scenarios, so that the generated deviation data can better reflect the diversity of actual conditions. By randomly generating samples to be selected and sorting them, the continuous probability distribution can be discretized into a series of ordered sample points, which not only simplifies the calculation process, but also reduces the calculation complexity. At the same time, by using the spline interpolation method to construct a polynomial in the interval, the predicted value of any given point can be efficiently calculated, which improves the calculation efficiency.

[0085] Furthermore, constructing the spline polynomial for each interval based on the spline interpolation method can flexibly adapt to the data variation characteristics of different intervals, that is, it can provide smooth and continuous prediction values, avoiding abrupt jumps between data points, thereby improving the accuracy and reliability of the prediction.

[0086] In a preferred embodiment, Figure 2The model building process and solution process shown in the figure, the present invention first generates a predicted output deviation value scenario set, adopts the K-means clustering algorithm to cluster the scenarios, and thus constructs a wind-solar-storage integrated system operation optimization model (i.e., the above-mentioned initial scheduling model) that considers multiple scenarios of wind-solar-new energy output correlation based on the scenario clustering data, and after the initial scheduling model is preliminarily solved, the optimal function value preliminarily solved is used as the reference value of the IGDT model (i.e., the target scheduling model in the above-mentioned embodiment), thereby realizing the construction of the risk aversion model and the risk preference model, and based on different risk aversion factors and risk preference factors, the scheduling strategies under the two models under different uncertainty coefficients are obtained, such as: the optimal solution value of the objective function under different risk preference factors and different deviation factors; the uncertainty degree coefficient under different risk preference factors and different deviation factors; the decision variables under different risk preference factors and different deviation factors: the power generation power of the wind farm and the power generation power of the photovoltaic power station in the s-th scenario and the time period t; the discharge power and charging power of the energy storage power station in the s-th scenario and the time period t.

[0087] Specifically, when modeling the uncertainty of wind and solar power output, each predicted output deviation value of the wind farm and each predicted output deviation value of the photovoltaic power station can be clustered by scenario, and the wind farm output prediction deviation data corresponding to different scenarios and the photovoltaic power station output prediction deviation data corresponding to different scenarios are generated, specifically:

[0088] Considering that the output of wind and solar energy is mainly related to meteorological conditions, and there is a certain correlation between the wind speed and light in the region, when first modeling the correlation between the deviation characteristics of the predicted output and the actual output of wind and solar energy, the historical data of the predicted output deviation values ​​of the wind farm and photovoltaic power station on day D can be selected, and the probability density function of the predicted output deviation of the wind farm and photovoltaic power station in each period within 24 hours can be fitted based on the non-parametric kernel density estimation method. Among them, the kernel function selects the Gaussian kernel function, and the expression is as follows:

[0089]

[0090] Where: h is the bandwidth, x wt,t and x pv,t are the predicted output deviation values ​​of the wind farm and photovoltaic power station in time period t, respectively, wt,d,t and X pv,d,t are the historical data values ​​of the predicted output deviation of the wind farm and the photovoltaic power station at the dth day and time period t, respectively. G(·) is the Gaussian kernel function. Taking the predicted output deviation of the wind farm as an example, the expression is as follows:

[0091]

[0092] Then, the first probability density function and the second probability density function are obtained by integration, which are and

[0093] The Frank-Copula function is selected, and the joint distribution function of the predicted output deviation of wind farms and photovoltaic power stations in each period is constructed based on the first probability density function, the second probability density function and the correlation coefficient:

[0094]

[0095] Where: u t and v t are the cumulative distribution functions of the predicted output deviations of wind farms and photovoltaic power stations at each time period t, θ t is the cumulative distribution function of the predicted output deviation of wind farms and photovoltaic power stations at each time period t )and The correlation coefficient in the Frank-Copula function can be obtained by the maximum likelihood estimation method.

[0096] Based on the constructed Frank-Copula function, for each period t, generate n random number samples that conform to the joint distribution of the established Frank-Copula function

[0097] will u t and v t The respective random number samples are sorted into column vectors and sorted in ascending order. The interval endpoints 0 and 1 of the cumulative probability function are added to the beginning and end of the column vectors. The expression is as follows:

[0098]

[0099] Where: and are the nth random number samples after sorting.

[0100] From the above formula, we can see that and Each of them contains n+2 random numbers; then n+1 small intervals can be formed, and the endpoints of each interval are and A sample of The samples represented by each interval endpoint are the values ​​of the cumulative distribution function of the predicted output deviation of the wind farm and the photovoltaic power station, which can be obtained through the cumulative distribution function and Obtain the predicted output deviation value x of the wind farm and photovoltaic power station wt,t and x pv,t .

[0101] Furthermore, for each interval, the cubic spline interpolation method is used to construct a cubic spline polynomial on the interval:

[0102]

[0103] Where: a i 、b i 、c i d i are the coefficients of the cubic spline polynomial, a i ′、b i ′、c i ′、d i ' Similarly, the equation system can be constructed and solved by cubic spline interpolation method. and They are respectively the endpoints of the n+1 cells constructed above.

[0104] After obtaining the coefficients of the cubic spline polynomials on the above intervals, the u extracted according to the established Frank-Copula function joint distribution is t and v t Substituting the random number samples into the polynomial equation, the predicted output deviation value x of the wind farm and photovoltaic power station in each period can be obtained by reverse deduction. wt,t and x pv,t Scene collection.

[0105] According to the predicted output deviation value x of the wind farm and photovoltaic power station wt,t and x pv,t The scene set uses the K-means clustering algorithm to cluster the scenes. After clustering, multiple cluster centers and the number of scenes belonging to each cluster center category can be obtained. The proportion of each category is calculated based on the number of scenes, which is used as the probability of the cluster center appearing. In this way, the predicted output deviation values ​​of wind farms and photovoltaic power stations are reduced in multiple scenarios. The results after reduction are expressed as cluster centers and their probabilities, so that the predicted output deviation data of wind farms and photovoltaic power stations under different scenarios can be obtained.

[0106] Indicatively, the optimization objective of the initial scheduling model can be:

[0107]

[0108] Among them, F 0 is the function value of the initial scheduling model, P L,t,s is the sum of the output power of the wind-solar-storage integrated system in the sth scenario during the tth period, P L,av,s is the average power within the scheduling period T in the sth scenario, is the probability of clustering the sth scene, Ω s is the set of cluster center points.

[0109] The constraints of the initial dispatch model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints, and energy storage power station charge state constraints;

[0110] Then, the wind farm output constraints include:

[0111]

[0112] Among them, P wt,t,s is the output of the wind farm in the sth scenario during the t period, is the preset output of the wind farm corresponding to the t period in the sth scenario, x wt,t,s is the predicted deviation value of wind farm output in period t under the sth scenario;

[0113] The photovoltaic power station output constraints include:

[0114]

[0115] Among them, P pv,t,s is the output of the photovoltaic power station in the sth scenario during the t period, is the preset output of the photovoltaic power station corresponding to the t period in the sth scenario, x pv,t,s The predicted deviation value of the photovoltaic power station in the t period under the sth scenario;

[0116] The energy storage power station charging and discharging power constraints include:

[0117]

[0118] Among them, μ essdis,t,s and μ essch,t,s are the Boolean variables of the discharge state and the charging state of the energy storage power station in the t period under the sth scenario; P essdis,min is the minimum power value of the energy storage station discharge, P essdis,max is the maximum power value of the energy storage station discharge, P essch,min The minimum power value for charging the energy storage station, P essch,max The maximum power value for charging the energy storage station;

[0119] The state of charge constraints of the energy storage power station include:

[0120]

[0121] Among them, E ess,t,s is the preset power of the energy storage power station in the t period under the sth scenario, η ess is the charge and discharge efficiency coefficient; SOC max and SOC minThey are the upper and lower limits of the state of charge of the energy storage power station; N battery The preset total capacity of the energy storage power station; and They are respectively the charge state reference values ​​corresponding to the energy storage power station at different times.

[0122] The initial scheduling model can be solved based on the above constraints to ensure that the energy storage equipment works under safe and stable conditions and prevent equipment damage or safety accidents caused by overcharging, over-discharging or abnormal charge state.

[0123] The above constraints provide clear boundary conditions for the system's dispatch, allowing the dispatch strategy to be flexibly adjusted under the premise of meeting these conditions. This helps to optimize the use of renewable energy, balance the supply and demand of the power grid, and improve the overall operating efficiency of the system.

[0124] For step S2, the present invention can solve the initial scheduling model based on the wind farm output constraint, the photovoltaic power station output constraint, the energy storage power station charging and discharging power constraint, and the energy storage power station charge state constraint, and output the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is minimized; schematically, the objective function value is the function value corresponding to the optimal solution of the initial scheduling model, and the optimal solution of the initial scheduling model at this time is the initial energy storage power station charging power and the initial energy storage power station discharging power corresponding to the initial scheduling model under the objective function value.

[0125] It is understandable that after obtaining the preliminary objective function value F 0 After obtaining the current optimal solution (such as the initial energy storage power station charging power and the initial energy storage power station discharging power), in order to further consider the correlation between the output of the wind farm and the photovoltaic power station, the uncertainty of the load and distributed photovoltaic, the current optimal solution can be further optimized and adjusted, that is, the objective function value F corresponding to the optimal solution of the initial scheduling model can be 0 As a benchmark value, a target scheduling model, namely a risk aversion model and a risk preference model, is further constructed, so that a control strategy that effectively reduces the volatility of the power delivered by the wind-solar-storage integrated system can be obtained.

[0126] For step S3, in a preferred embodiment, in order to further enable the system scheduling to consider and optimize power uncertainty, the embodiment of the present invention can obtain the optimal function value F of the initial scheduling model. 0 After that, the initial scheduling model is continued as the basic function, and then the preference factor and the uncertainty coefficient are combined to construct a target scheduling model with the goal of minimizing the negative impact of power uncertainty on the system and making the system power offset less than the target power offset.

[0127] Since there is uncertainty in the output of wind farms and photovoltaic power stations, and there is also uncertainty in the power demand of user loads, this uncertainty will directly affect the system's output power and scheduling strategy. Therefore, system scheduling needs to pay attention not only to the average or expected value of power, but also to consider the volatility and uncertainty of power.

[0128] In order to achieve a trade-off between risk and benefit, the target scheduling model should focus on reducing the negative impact of power uncertainty on the system and ensuring the stability and safety of system operation. The target scheduling model can automatically adjust the final scheduling variables or scheduling strategies within a certain range of power uncertainty, so that the target scheduling model can minimize the negative impact of power uncertainty on the system, maximize risk control, and minimize the negative impact on benefits, ensuring that the final scheduling variables, the offset of each power is always kept within the power offset defined by the preference factor, and the power uncertainty meets the range of system preference requirements.

[0129] In order to meet the risk preferences of different decision makers or to cope with different market environments, when constructing the target scheduling model, the preference factor can be set to a risk aversion factor that characterizes the risk aversion degree of the system in the face of power uncertainty, thereby limiting the power offset and ensuring that the system can maintain stable operation in the face of power uncertainty. Alternatively, the preference factor can be set to a risk preference factor that characterizes the risk preference degree of the system in balancing risk and benefit in the face of power uncertainty, so that the target scheduling model can focus on the trade-off between risk and benefit and obtain greater system benefits by accepting a certain amount of power uncertainty.

[0130] Therefore, the present invention can set preference factors with different definitions so that when constructing a target scheduling model, a risk avoidance model can be constructed with the goal of maximizing the risk impact of power uncertainty on the system and the power offset not exceeding the maximum power offset; or a risk preference model can be constructed with the goal of minimizing the impact of power uncertainty on system benefits and the power offset being less than the minimum power offset.

[0131] It is understandable that by building a risk aversion model, it is possible to limit the power offset and ensure that the system can maintain stable operation in the face of power uncertainty, thereby reducing system failures and downtime and improving system reliability and availability. By building a risk preference model, it is possible to accept a certain amount of power uncertainty and allow the system to adopt a more active scheduling strategy in certain situations, thereby making full use of renewable energy and improving the system's power generation efficiency and economic benefits.

[0132] In a preferred embodiment, the target scheduling model may be a risk aversion model. The preference factor in the target scheduling model is a risk aversion factor used to characterize the risk aversion degree of the system when facing power uncertainty; the uncertainty coefficient is a first uncertainty coefficient used to characterize the degree of influence of power uncertainty on system risk; the target power offset is the maximum power offset limited by the risk aversion factor;

[0133] Then, the construction of the target scheduling model includes the construction of a risk aversion model, specifically:

[0134] Taking the objective function value as a reference value, and based on the initial scheduling model, the risk aversion factor and the first uncertainty coefficient, a risk aversion model is constructed with the goal of maximizing the risk impact of power uncertainty on the system and ensuring that the power offset does not exceed the maximum power offset.

[0135] Illustratively, in another preferred embodiment, the target scheduling model may be a risk preference model; the preference factor in the target scheduling model is a risk preference factor used to characterize the degree of risk preference of the system in balancing risk and benefit when facing power uncertainty; the uncertainty coefficient is a second uncertainty coefficient used to characterize the degree of influence of power uncertainty on system benefits; the target power offset is the minimum power offset defined by the risk preference factor;

[0136] Then, the construction of the target scheduling model includes the construction of the risk preference model, and the specific process is as follows:

[0137] Taking the objective function value as a reference value, and based on the initial scheduling model, the risk preference factor and the second uncertainty coefficient, a risk preference model is constructed with the goal of minimizing the impact of power uncertainty on system revenue and ensuring that the power offset is less than the minimum power offset.

[0138] Specifically, for the risk aversion model, the risk aversion factor can be defined as the maximum power offset corresponding to a power uncertainty range, that is, the risk aversion factor reflects the tolerance for power uncertainty and the risk aversion attitude. By adjusting the risk aversion factor, the model can find a balance between safety and economy.

[0139] The first uncertainty coefficient is used to define the impact of power uncertainty on system risk; the uncertainty coefficient is used to quantify the impact of power uncertainty on system risk, which takes into account factors such as the size and distribution of power uncertainty and the sensitivity of the system to uncertainty. By introducing this coefficient, the model can more accurately evaluate the impact of power uncertainty on system risk.

[0140] Therefore, the risk aversion model can maximize the risk impact of power uncertainty on the system (quantified by the first uncertainty coefficient), while ensuring that the power offset does not exceed the maximum power offset (quantified by the risk aversion factor). Its constraints are the constraints of the original initial scheduling model, such as wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints and charge state constraints, as well as additional constraints defined by the risk aversion factor and the first uncertainty coefficient.

[0141] Schematically, the risk aversion model includes:

[0142] maxα r ;

[0143] maxF 1 ≤(1+β r )F 0 ;

[0144]

[0145] Among them, β r is the risk aversion factor, F 0 is the objective function value corresponding to the initial scheduling model when the average absolute deviation of the total system output power in all scenarios is the smallest, P L,t,s is the sum of the output power of the wind-solar-storage integrated system in the sth scenario during the tth period, P L,av,s is the average power within the scheduling period T in the sth scenario, is the probability of clustering the sth scene, Ω s is the cluster center set, α r is the first uncertainty coefficient;

[0146] P equip,t is the preset total demand power corresponding to the sum of the user power consumption and the distributed photovoltaic power generation power in period t, The predicted total power demand between user power consumption and distributed photovoltaic power generation power; is the power demand model, used to characterize the predicted total power demand The power demand model is used to solve the uncertainty coefficient α in the first degree of uncertainty. r The corresponding predicted total power demand value is:

[0147] P Load,t is the power consumption of the user during period t, P dispv,t is the distributed photovoltaic power generation power in period t, P wt,t,s is the output of the wind farm in the sth scenario during the t period, P pv,t,s is the output of the photovoltaic power station in the sth scenario during the tth period, Pessdis,t,s is the charging power of the energy storage station in the sth scenario during the t period, P essch,t,s is the discharge power of the energy storage power station in period t under the sth scenario.

[0148] It can be understood that the present invention can describe the randomness of distributed photovoltaic output and the randomness of user load, that is, the fractional uncertainty model is used to describe the total power demand P equip,t uncertainty.

[0149] By solving the risk aversion model, we can obtain a set of optimal scheduling variables (i.e., target scheduling variables), which not only meet the system scheduling requirements, but also take into account the impact of power uncertainty on system risk. These optimal scheduling variables can be used as a reference for actual scheduling strategies, thereby improving the safety and economy of the system.

[0150] Furthermore, the risk preference model includes:

[0151] minα c ;

[0152] f 2 ≤(1-β c )F 0 ;

[0153]

[0154]

[0155] Among them, β c is the risk preference factor, α c is the second uncertainty coefficient.

[0156] In schematic form, the core goal of the risk preference model is to minimize the impact of power uncertainty on system revenue while ensuring that the power offset is minimal. This means that during the solution process, the risk preference model will comprehensively consider the balance between system revenue and power fluctuations to generate a set of optimal scheduling variables. These optimal scheduling variables can maximize the economic benefits of the system while meeting the system stability and reliability requirements.

[0157] It can be understood that the risk aversion model represents the total power demand P equip,t Within the uncertainty range, any decision-making scheme can make the objective function F 2 Not higher than the maximum acceptable offset value, solve for P equip,t The maximum uncertainty range is determined and the optimal control strategy is solved.

[0158] The risk preference model represents the total power demand P equip,tWithin the uncertainty range, there is at least one decision solution that makes the objective function F 3 A smaller offset value can be achieved, solving P equip,t Minimize the uncertainty range and solve the optimal control strategy.

[0159] For step S4, in a preferred embodiment, the embodiment of the present invention can solve the target scheduling model under the constraints of wind farm output, photovoltaic power station output, energy storage power station charge and discharge power, and energy storage power station charge state, so as to generate corresponding target scheduling variables when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset; wherein the target scheduling variables corresponding to the optimal solution include the target energy storage power station charging power and the target energy storage power station discharging power that can be used to control the energy storage power station. Schematically, it can also include the target wind farm output and the target photovoltaic power station output that can be used to control the system equipment.

[0160] In the embodiment of the present invention, the initially solved function value can be used as a benchmark value, thereby providing a clear reference standard for solving the target scheduling model. It can also ensure consistency during the solving process, ensuring that the generated scheduling solution fluctuates within an acceptable range from the initial solution.

[0161] Moreover, by using the initially solved function value as the benchmark value, the target scheduling model can ensure that the final scheduling plan is implemented with the minimum impact of power uncertainty on system risk, and can prevent system crashes or failures caused by power fluctuations. At the same time, a stable scheduling plan can also improve the reliability and safety of the system and ensure the continuity and stability of power supply.

[0162] In a preferred embodiment, the target scheduling variable includes: a first scheduling variable, or a second scheduling variable;

[0163] The target scheduling model is solved to generate a corresponding target scheduling variable when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset, including:

[0164] When the preference factor is determined to be a risk aversion factor, the risk aversion model is solved under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and the first dispatch variable is generated when the power uncertainty has the greatest impact on the risk of the system and the power offset does not exceed the maximum power offset; wherein the first dispatch variable includes: the charging power of the first energy storage power station and the discharging power of the first energy storage power station;

[0165] When the preference factor is determined to be a risk preference factor, the risk preference model is solved under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and a second scheduling variable is generated when the impact of power uncertainty on system revenue is minimal and the power offset is less than the minimum power offset; wherein the second scheduling variable includes: the charging power of the second energy storage power station and the discharging power of the second energy storage power station.

[0166] It is understandable that, whether solving the risk aversion model or the risk preference model, the physical constraints and power uncertainty of the system are taken into account, and the optimal solution with the best benefit is obtained through the optimization algorithm. Therefore, a more flexible scheduling solution can be provided in actual operation to adapt to different operating environments and demand changes.

[0167] Specifically, the present invention can construct a risk aversion model and a risk preference model based on the IGDT method. Specifically, based on the optimal solution value in the initial scheduling model, it can be used as the IGDT model benchmark value, and F 0 express.

[0168] Since the risk aversion model and the risk preference model are double-layer optimization problems of max-max and min-min respectively, the embodiment of the present invention can define the outer optimization problem as the main problem and the inner optimization problem as the sub-problem.

[0169] For the risk aversion model, based on a set risk aversion factor β r , the solution process is:

[0170] 1) Set α r The initial value of is 0, and the subproblem is optimized and solved. r = 0, the optimal solution of the subproblem is the IGDT model reference value F 0 ; 2) Let Δα r =0.1, Δα r is α r The change amount of each process setting is set to the end of the solution mark ε = 0.0001; 3) Let α r =α r +Δα r , solve the sub-problem; 4) If the sub-problem has a solution, go back to step 3); if the sub-problem has no solution, first let Δα r =0.1Δα r , judge Δα r Is it equal to ε? If so, jump directly to step 5), otherwise go back to step 3); 5) Output α rAnd the optimal solution of the sub-problem. The optimal solution of the sub-problem is the target wind farm output, target photovoltaic power station output, target energy storage power station charging power and target energy storage power station discharging power in the target scheduling variables.

[0171] Furthermore, for the risk preference model, a risk preference factor β is set c , the solution process is:

[0172] 1) Set α c The initial value of is 1, and the subproblem is optimized and solved; 2) Let Δα c =0.1, Δα c is α c The change amount of each process setting is set to the end of the solution mark ε = 0.0001; 3) Let α c =α c -Δα c , solve the sub-problem; 4) If the sub-problem has a solution, go back to step 3); if the sub-problem has no solution, first let Δα c =0.1Δα c , judge Δα c Is it equal to ε? If so, jump directly to step 5), otherwise go back to step 3); 5) Output α c And the optimal solution of the sub-problem. The optimal solution of the sub-problem is the wind farm output, photovoltaic power station output, energy storage power station charging power and energy storage power station discharging power in the updated dispatching variables.

[0173] In the embodiment of the present invention, the operation dispatch control strategy results of the wind-solar-storage integrated system under different risk preference factors or different risk aversion factors can be obtained. The control strategy results may include:

[0174] 1) The optimal solution of the function under different risk preference factors or different deviation factors; 2) The uncertainty coefficient α under different risk preference factors and different deviation factors r With α c ; 3) Decision variables under different risk preference factors or different deviation factors, that is, the optimal dispatch variable value, such as: the output P of the wind farm in the sth scenario during the t period wt,t,s , the output P of the photovoltaic power station in the t period under the sth scenario pv,t,s , the charging power P of the energy storage station in the sth scenario during the t period essdis,t,s , the discharge power P of the energy storage station in the sth scenario during the t period essch,t,s .

[0175] In a preferred embodiment, all uncertain scenarios can be integrated to calculate the total power generation of the wind farm and photovoltaic power station in time period t by weighted sum of multiple scenarios, as shown in the following formula:

[0176]

[0177] The total discharge power and total charging power of the energy storage power station in time period t can also be calculated by weighted sum of multiple scenarios, as shown in the following formula:

[0178]

[0179] The embodiment of the present invention can output control strategy solutions under different risk preference factors and different deviation factors. By constructing a risk aversion model and a risk preference model, the present invention provides two decision-making solutions under different risk preferences for operation decision-makers, so that the most appropriate scheduling strategy can be selected according to different operating environments and demand changes. At the same time, the model also takes into account the uncertainty of the total power demand, which can enable the system to better cope with various uncertain scenarios and improve adaptability.

[0180] For step S5, in a preferred embodiment, the present invention can control the output of the energy storage power station in the wind-solar-storage integrated system based on the solved target scheduling variable, thereby ensuring that the output power of the system matches the total power demand, thereby maintaining stable operation of the system and reducing system fluctuations and failures caused by power mismatch.

[0181] Moreover, according to different operating environments and demand changes, different risk preference factors and control schemes under different deviation factors can be selected to meet various complex operating requirements.

[0182] Therefore, the present invention takes into account the correlation between the output of wind farms and photovoltaic power stations, and constructs an operation scheduling optimization model (i.e., an initial scheduling model) of a wind-solar-storage integrated system that takes into account the correlation between wind and solar output based on the Frank-Copula method. The optimal solution of the initial scheduling model is used as the benchmark value of the IGDT method, thereby constructing a risk aversion model and a risk preference model that take into account the correlation between wind and solar output and are based on the IGDT method, so that the optimal scheduling strategy finally solved can comprehensively consider the correlation between the output of wind farms and photovoltaic power stations, the uncertainty of load and distributed photovoltaics, and obtain a control strategy that effectively reduces the volatility of the power delivered by the wind-solar-storage integrated system.

[0183] like Figure 3 As shown, based on the above-mentioned various embodiments of the optimization scheduling method of the wind-solar-storage integrated system, the present invention provides a corresponding device embodiment;

[0184] An embodiment of the present invention provides an optimization scheduling device for a wind-solar-storage integrated system, comprising: an initial scheduling model construction module, a first solution module, a target scheduling model construction module, a second solution module, and an output control module;

[0185] The initial scheduling model construction module is used to construct an initial scheduling model with the goal of minimizing the average absolute deviation of the total system output power in all scenarios according to the power data of the wind-solar-storage integrated system in different scenarios and the power average values ​​in different scenarios; wherein the constraints of the initial scheduling model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints, and energy storage power station charge state constraints;

[0186] The first solving module is used to solve the initial scheduling model and output the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is the smallest;

[0187] The target scheduling model construction module is used to take the target function value as a reference value, and according to the initial scheduling model, the preference factor and the uncertainty coefficient, construct a target scheduling model with the goal of minimizing the negative impact of power uncertainty on the system and the power offset of the system being less than the target power offset; wherein the preference factor is used to characterize the risk tolerance of the system when facing power uncertainty; the uncertainty coefficient is used to characterize the impact of power uncertainty on the system; the target power offset is the power offset defined by the preference factor;

[0188] The second solving module is used to solve the target scheduling model under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and generate corresponding target scheduling variables when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset; wherein the target scheduling variables include: target energy storage power station charging power and target energy storage power station discharging power;

[0189] The output control module is used to control the output of the energy storage power station in the wind-solar-storage integrated system according to the target scheduling variable.

[0190] It should be noted that the device embodiments described above are merely schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without paying creative labor.

[0191] Those skilled in the art can clearly understand that, for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0192] Based on the above-mentioned various embodiments of the optimization scheduling method of the wind-solar-storage integrated system, the present invention provides corresponding terminal equipment embodiments.

[0193] An embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, an optimization scheduling method for a wind, solar and storage integrated system as described in any method embodiment of the present invention is implemented.

[0194] The terminal device may be a computing terminal device such as a desktop computer, a notebook, a palm computer, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0195] Based on the above-mentioned various embodiments of the optimization scheduling method of the wind-solar-storage integrated system, the present invention provides corresponding storage medium item embodiments.

[0196] An embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute an optimization scheduling method for a wind-solar-storage integrated system described in any method embodiment of the present invention.

[0197] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. An optimization scheduling method for a wind-solar-storage integrated system, characterized in that: include: According to the power data of the wind-solar-storage integrated system in different scenarios and the average power values ​​in different scenarios, an initial scheduling model is constructed with the goal of minimizing the average absolute deviation of the total system output power in all scenarios; wherein the constraints of the initial scheduling model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints, and energy storage power station charge state constraints; Solving the initial scheduling model, and outputting the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is the smallest; Taking the objective function value as a reference value, and based on the initial scheduling model, the preference factor and the uncertainty coefficient, a target scheduling model is constructed with the goal of minimizing the negative impact of power uncertainty on the system and the power offset of the system being less than the target power offset; wherein the preference factor is used to characterize the risk tolerance of the system when facing power uncertainty; the uncertainty coefficient is used to characterize the impact of power uncertainty on the system; and the target power offset is the power offset defined by the preference factor; Under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, the target scheduling model is solved, and when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset, the corresponding target scheduling variables are generated; wherein the target scheduling variables include: target energy storage power station charging power and target energy storage power station discharging power; According to the target dispatching variable, the output of the energy storage power station in the wind-solar-storage integrated system is controlled.

2. The optimization scheduling method of the wind-solar-storage integrated system according to claim 1, characterized in that: The target scheduling model includes: a risk aversion model; the preference factor includes: a risk aversion factor for characterizing the degree of risk aversion of the system when facing power uncertainty; the uncertainty coefficient includes: a first uncertainty coefficient for characterizing the degree of influence of power uncertainty on system risk; The target power offset includes: a maximum power offset limited by a risk aversion factor; The construction of the risk aversion model includes: Taking the objective function value as a reference value, and based on the initial scheduling model, the risk aversion factor and the first uncertainty coefficient, a risk aversion model is constructed with the goal of maximizing the risk impact of power uncertainty on the system and ensuring that the power offset does not exceed the maximum power offset.

3. The optimization scheduling method of the wind-solar-storage integrated system according to claim 2, characterized in that: The target scheduling model includes: a risk preference model; the preference factor includes: a risk preference factor for characterizing the degree of risk preference of the system in balancing risk and benefit when facing power uncertainty; the uncertainty coefficient includes: a second uncertainty coefficient for characterizing the degree of influence of power uncertainty on system benefit; the target power offset includes: a minimum power offset defined by the risk preference factor; The construction of the risk preference model includes: Taking the objective function value as a reference value, and based on the initial scheduling model, the risk preference factor and the second uncertainty coefficient, a risk preference model is constructed with the goal of minimizing the impact of power uncertainty on system revenue and ensuring that the power offset is less than the minimum power offset.

4. The optimization scheduling method of the wind-solar-storage integrated system according to claim 3, characterized in that: The target scheduling variable includes: a first scheduling variable or a second scheduling variable; The target scheduling model is solved to generate a corresponding target scheduling variable when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset, including: When the preference factor is determined to be a risk aversion factor, the risk aversion model is solved under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and the first dispatch variable is generated when the power uncertainty has the greatest impact on the risk of the system and the power offset does not exceed the maximum power offset; wherein the first dispatch variable includes: the charging power of the first energy storage power station and the discharging power of the first energy storage power station; When the preference factor is determined to be a risk preference factor, the risk preference model is solved under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and a second scheduling variable is generated when the impact of power uncertainty on system revenue is minimal and the power offset is less than the minimum power offset; wherein the second scheduling variable includes: the charging power of the second energy storage power station and the discharging power of the second energy storage power station.

5. The optimization scheduling method of the wind-solar-storage integrated system according to claim 4, characterized in that: The risk aversion model comprises: maxα r ; maxF1≤(1+β r )F0; Among them, β r is the risk aversion factor, F0 is the objective function value corresponding to the initial scheduling model when the average absolute deviation of the total system output power in all scenarios is the smallest, and P L,t,s is the sum of the output power of the wind-solar-storage integrated system in the sth scenario during the tth period, P L,av,s is the average power within the scheduling period T in the sth scenario, is the probability of clustering the sth scene, Ω s is the cluster center set, α r is the first uncertainty coefficient; P equip,t is the preset total demand power corresponding to the sum of the user power consumption and the distributed photovoltaic power generation power in period t, The predicted total power demand between user power consumption and distributed photovoltaic power generation; is the power demand model, used to characterize the predicted total power demand The power demand model is used to solve the uncertainty coefficient α in the first degree of uncertainty. r The corresponding predicted total power demand value is: P Load,t is the power consumption of the user during period t, P dispv,t is the distributed photovoltaic power generation power in period t, P wt,t,s is the output of the wind farm in the sth scenario during the t period, P pv,t,s is the output of the photovoltaic power station in the sth scenario during the tth period, P essdis,t,s is the charging power of the energy storage station in the sth scenario during the t period, P essch,t,s is the discharge power of the energy storage power station in the sth scenario during the t period; The risk preference model includes: minα c ; minF2≤(1-β c )F0; Among them, β c is the risk preference factor, α c is the second uncertainty coefficient.

6. The optimization scheduling method of the wind-solar-storage integrated system according to claim 1, characterized in that: The process of generating power data in different scenarios includes: Obtain historical data on user power consumption, distributed photovoltaic power generation, wind farm output forecast deviation, and photovoltaic power station output forecast deviation in different scenarios; Fitting each output forecast deviation historical data of the wind farm and each output forecast deviation historical data of the photovoltaic power station respectively to generate a first probability density function for characterizing the degree of output deviation of the wind farm and a second probability density function for characterizing the degree of output deviation of the photovoltaic power station; A joint distribution probability model for characterizing the correlation between the prediction deviations of the wind farm and the photovoltaic power station is constructed according to the first probability density function, the second probability density function and the correlation coefficient; wherein the correlation coefficient is used to characterize the correlation between the predicted output deviations of the wind farm and the photovoltaic power station; Based on the first probability density function and the second probability density function, the joint distribution probability model is solved to generate wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios; Based on the wind farm output prediction deviation data, photovoltaic power station output prediction deviation data, user power consumption and distributed photovoltaic power generation power under different scenarios, power data corresponding to different scenarios are generated.

7. The optimization scheduling method of the wind-solar-storage integrated system according to claim 6, characterized in that: The generating of wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios includes: Randomly generate samples to be selected corresponding to the joint distribution probability model; Sort the samples to be selected in ascending order to generate a corresponding column vector; wherein the first and last ends of the column vector are respectively the preset endpoint values ​​corresponding to the first probability density function and the second probability density function; Divide the column vector into a plurality of intervals, and construct a spline polynomial corresponding to each interval based on a spline interpolation method; Substituting each candidate sample into the corresponding spline polynomial to generate a predicted output deviation value of the wind farm and a predicted output deviation value of the photovoltaic power station; The predicted output deviation values ​​of the wind farm and the predicted output deviation values ​​of the photovoltaic power station are clustered according to different scenarios to generate wind farm output prediction deviation data corresponding to different scenarios and photovoltaic power station output prediction deviation data corresponding to different scenarios.

8. An optimization and scheduling device for a wind, solar and storage integrated system, characterized in that: include: An initial scheduling model building module, a first solving module, a target scheduling model building module, a second solving module and an output control module; The initial scheduling model construction module is used to construct an initial scheduling model with the goal of minimizing the average absolute deviation of the total system output power in all scenarios according to the power data of the wind-solar-storage integrated system in different scenarios and the power average values ​​in different scenarios; wherein the constraints of the initial scheduling model include: wind farm output constraints, photovoltaic power station output constraints, energy storage power station charging and discharging power constraints, and energy storage power station charge state constraints; The first solving module is used to solve the initial scheduling model and output the objective function value currently corresponding to the initial scheduling model when the average absolute deviation corresponding to the total system output power in all scenarios is the smallest; The target scheduling model construction module is used to take the target function value as a reference value, and according to the initial scheduling model, the preference factor and the uncertainty coefficient, construct a target scheduling model with the goal of minimizing the negative impact of power uncertainty on the system and the power offset of the system being less than the target power offset; wherein the preference factor is used to characterize the risk tolerance of the system when facing power uncertainty; the uncertainty coefficient is used to characterize the impact of power uncertainty on the system; the target power offset is the power offset defined by the preference factor; The second solving module is used to solve the target scheduling model under the constraints of wind farm output, photovoltaic power station output, energy storage power station charging and discharging power, and energy storage power station charge state, and generate corresponding target scheduling variables when the negative impact of power uncertainty on the system is minimal and the power offset of the system is less than the target power offset; wherein the target scheduling variables include: target energy storage power station charging power and target energy storage power station discharging power; The output control module is used to control the output of the energy storage power station in the wind-solar-storage integrated system according to the target scheduling variable.

9. A terminal device, characterized in that: It comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements an optimization scheduling method for a wind-solar-storage integrated system as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the optimization scheduling method for the wind-solar-storage integrated system as described in any one of claims 1 to 7.

Images