Generation method and device of output regulation and control strategy of novel power system, and control system

Abstract

The invention discloses a novel power system output regulation and control strategy generation method and device and a control system, and belongs to the technical field of power system regulation and control. The output regulation and control strategy generation method comprises the following steps: representing net load historical data corresponding to net load climbing demand uncertainty characteristics by adopting a Gaussian mixture model; the uncertainty of the net load climbing requirement is reduced; inputting the net load day-ahead prediction data corresponding to the net load historical data and the net load uncertainty scene set into a two-stage risk avoidance model based on the lowest operation cost and the worst condition risk probability assessment theory; the two-stage risk avoidance model divides the whole strategy generation process into day-ahead optimization scheduling plan generation and intra-day real-time scheduling plan generation; the day-ahead optimal scheduling plan is optimized by using the obtained intra-day real-time scheduling plan to obtain a safe and economic output scheduling strategy; and furthermore, the novel power system is controlled in different scenes to realize efficient, stable and economical operation.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system regulation and control, and more specifically, relates to a novel power system output regulation strategy generation method, device and control system. Background Art

[0002] Under the background of dual carbon goals, the integration of new energy into the power system can help reduce dependence on fossil fuels, reduce greenhouse gas emissions, and improve the sustainability of the energy system. The introduction of new energy can diversify the energy mix of the power system and avoid over-reliance on a single or a few energy resources. The diversified energy mix can improve the resilience and stability of the power system and reduce the risk of energy price fluctuations. However, the large-scale grid connection of new energy has led to increasing randomness in the new power system. In addition, a large number of electric vehicles will be introduced in the future, and the uncertainty of net load fluctuations will increase, bringing more challenges to the safe, reliable, stable and economical operation of the power system.

[0003] In the new power system, with the increasing use of renewable energy (such as wind energy and solar energy) in the energy structure, the dispatch process of the power system has become more and more complicated. Net load is the difference between conventional net load and renewable energy output. Due to the uncertainty and randomness of renewable energy output, the characterization and prediction of net load is more difficult than conventional net load, which also brings challenges to the safe, stable and economic operation of the power system.

[0004] Therefore, it is very important for the safe and stable operation of the power system to build a more accurate model in the new power system to characterize the uncertain fluctuations of the net load and establish a method for generating an output control strategy that can effectively avoid the risk of net load uncertainty. Summary of the invention

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a method, device and control system for generating an output control strategy for a new power system, which aims to solve the technical problem that the output control strategy generated in the new power system is difficult to effectively avoid the net load uncertainty risk.

[0006] To achieve the above object, according to one aspect of the present invention, a method for generating an output control strategy of a novel power system is provided, comprising:

[0007] The net load history data of the new power system is represented as a multi-component Gaussian mixture model;

[0008] Obtaining net load day-ahead forecast data and a net load uncertainty scenario set using a multi-component Gaussian mixture model corresponding to the net load historical data;

[0009] The net load day-ahead forecast data and the net load uncertainty scenario set are input into a two-stage risk aversion model; the two-stage risk aversion model includes a day-ahead optimization dispatch model and an intraday real-time dispatch model; wherein the day-ahead optimization dispatch model aims to minimize the operating cost of the new power system, and is used to input the net load day-ahead forecast data to obtain the day-ahead optimal dispatch plan; the intraday real-time dispatch model is constructed based on the worst-case risk probability assessment theory, and is used to input the net load uncertainty scenario set to obtain the intraday real-time dispatch plan;

[0010] The real-time dispatch plan within the day is used to optimize the optimal dispatch plan on the day before, and a target output control strategy is obtained.

[0011] In one embodiment, the objective function of the two-stage risk aversion model is:

[0012]

[0013] in, is the objective function of the day-ahead optimization scheduling model, is the objective function of the intraday real-time dispatch model; the decision variable x is the required power of the net load, F 1 is the total cost of the day-ahead operation, Δx is the decision variable of the second stage, τ sm is the net load forecast value, ρ is the risk aversion factor, F 1,m (Δx|τ sm ) represents the cost incurred during real-time power balancing, Φ β,m (Δx|τ sm ) represents the threshold value corresponding to the economic risk caused by the error between the net load forecast value and the actual value.

[0014] In one embodiment, the objective function of the day-ahead optimization scheduling model is expressed as: Among them, F I is the cost of thermal power, F V2G is the charging and discharging cost of electric vehicles, F cut is the load shedding cost, F car is the carbon emission cost, F gc The transaction cost of green certificates.

[0015] In one of the embodiments, the method of obtaining the net load day-ahead forecast data and the net load uncertainty scenario set using the multi-component Gaussian mixture model corresponding to the net load historical data includes:

[0016] Performing component reduction on the multi-component Gaussian mixture model;

[0017] The multi-component Gaussian mixture model after reduction is sampled to obtain the net load day-ahead forecast data;

[0018] Smoothing the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a modified Gaussian mixture model;

[0019] The modified Gaussian mixture model is used to generate a net load uncertainty scenario set.

[0020] In one of the embodiments, the component reduction of the multi-component Gaussian mixture model includes: performing component reduction on the multi-component Gaussian mixture model based on component-retaining hierarchical density reduction to obtain the reduced multi-component Gaussian mixture model.

[0021] In one embodiment, the smoothing of the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a revised Gaussian mixture model includes: using a covariance matrix correction method to smooth the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a revised Gaussian mixture model.

[0022] In one of the embodiments, the generating of the net load uncertainty scenario set by using the modified Gaussian mixture model includes: processing the modified Gaussian mixture model by using a scenario generation method to obtain the net load uncertainty scenario set.

[0023] According to another aspect of the present invention, a device for generating output control strategy of a new type of power system is provided, including: a characterization module, an acquisition module, an input module and an optimization module. The characterization module is used to characterize the net load historical data of the new type of power system as a multi-component Gaussian mixture model; the acquisition module is used to obtain the net load day-ahead forecast data and the net load uncertainty scenario set using the multi-component Gaussian mixture model corresponding to the net load historical data; the input module is used to input the net load day-ahead forecast data and the net load uncertainty scenario set into a two-stage risk aversion model; the two-stage risk aversion model includes a day-ahead optimization scheduling model and an intraday real-time scheduling model; the day-ahead optimization scheduling model takes minimizing the operating cost of the new type of power system as the goal, and is used to input the net load day-ahead forecast data to obtain the optimal scheduling plan for the day; the intraday real-time scheduling model is constructed based on the worst-case risk probability assessment theory, and is used to input the net load uncertainty scenario set to obtain the intraday real-time scheduling plan; the optimization module is used to optimize the optimal scheduling plan for the day-ahead using the intraday real-time scheduling plan to obtain the target output control strategy.

[0024] According to another aspect of the present invention, a control system of a novel power system is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0025] According to another aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0026] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects compared with the prior art:

[0027] (1) The present invention provides a method for generating an output control strategy for a new type of power system. First, a Gaussian mixture model with a consistent number of components and a sample number is used to characterize the net load historical data, and the net load uncertainty characterization model is subjected to component reduction; the net load day-ahead forecast data and the net load uncertainty scenario set corresponding to the net load historical data are input into a two-stage risk aversion model based on the minimum operating cost theory and the worst-case risk probability assessment theory; the two-stage risk aversion model divides the entire planning process into day-ahead optimization scheduling plan generation and intra-day real-time scheduling plan generation; finally, the intra-day real-time scheduling plan is used to optimize the day-ahead optimal scheduling plan to obtain a target output control strategy, which can provide a safe and economical target output scheduling strategy for decision makers of the new type of power system, and further can achieve efficient, stable and economical power system operation under different scenarios.

[0028] (2) Compared with the prior art, the objective function of the two-stage risk aversion model described in this solution takes into account the economic risks that may be encountered under the worst conditions, which can minimize the maximum risk cost and achieve an overall reduction in the economic risk cost.

[0029] (3) The objective function of the day-ahead optimization dispatch model described in this scheme involves thermal power costs, electric vehicle charging and discharging costs, load shedding costs, carbon emission costs, and green certificate trading costs. The orderly charging and discharging of electric vehicles and energy storage are introduced into the objective function as ramping resources in the new power system. In particular, green certificate trading and carbon emissions are taken into consideration, which can adjust the uncertainties of conventional thermal power units and various flexibility resources.

[0030] (4) This scheme samples the reduced multi-component Gaussian mixture model to obtain the net load day-ahead forecast data, taking into account the correlation between different load modes, which can effectively reduce the forecast error, achieve more accurate load demand forecasting, and enhance the decision-making support capabilities of power dispatching. Then, the modified Gaussian mixture model is used to generate a net load uncertainty scenario set, taking into account load volatility, weather factors, and the changing trend of historical load data, which can truly reflect the uncertainty of the net load and achieve more accurate risk assessment and optimized dispatching.

[0031] (5) This scheme performs component reduction on the multi-component Gaussian mixture model based on component-preserving hierarchical density reduction, taking into account the balance between model complexity and computational efficiency. It can significantly reduce the amount of computation and memory consumption while maintaining prediction accuracy, thereby achieving efficient processing and real-time application of large-scale data sets.

[0032] (6) This scheme uses a covariance matrix correction method to smooth the fitting curve corresponding to the multi-component Gaussian mixture model after the reduction. Compared with the traditional net load characterization method, the multi-bit Gaussian mixture model based on covariance correction and component reduction proposed in the present invention has better fitting accuracy and can better reflect the multi-peak characteristics of the sample.

[0033] (7) This solution uses a scenario generation method to process the modified Gaussian mixture model to obtain the net load uncertainty scenario set, and uses the errors between net load prediction samples and real-time samples to bring about a multi-scenario set of uncertainty risks, taking into account the risks that the system may encounter under the worst conditions and minimizing its economic risk costs, thereby reducing the overall economic risks brought about by risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of a method for generating an output control strategy for a new power system provided in Example 1 of the present invention;

[0035] Figure 2 is a schematic diagram of the working environment of the novel power system provided by Embodiment 1 of the present invention;

[0036] Figure 3 Schematic diagram of the workflow of the two-stage risk avoidance model provided in Example 1 of the present invention;

[0037] Figure 4 It is a schematic diagram of component reduction of a multi-component Gaussian mixture model based on component-preserving layered density reduction provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0039] Example 1

[0040] like Figure 1As shown, the present invention provides a method for generating an output regulation strategy for a new type of power system, comprising: S1: characterizing the net load historical data of the new type of power system as a multi-component Gaussian mixture model; S2: using the multi-component Gaussian mixture model corresponding to the net load historical data to obtain the net load day-ahead forecast data and the net load uncertainty scenario set; S3: inputting the net load day-ahead forecast data and the net load uncertainty scenario set into a two-stage risk aversion model; the two-stage risk aversion model includes a day-ahead optimization scheduling model and an intraday real-time scheduling model; wherein, the day-ahead optimization scheduling model aims to minimize the operating cost of the new type of power system, and is used to input the net load day-ahead forecast data to obtain the optimal scheduling plan for the day-ahead; the intraday real-time scheduling model is constructed based on the worst condition risk probability assessment theory, and is used to input the net load uncertainty scenario set to obtain the intraday real-time scheduling plan; S4: using the intraday real-time scheduling plan to optimize the day-ahead optimal scheduling plan to obtain the target output regulation strategy.

[0041] For example, the working environment of the new power system is Figure 2 As shown, the new power system supply and demand balance, thermal power unit constraints, electric vehicle orderly charging and discharging constraints, energy storage constraints, green certificate trading constraints, etc. are as follows:

[0042]

[0043] u ev,c +u ev,d =1;

[0044] SOC min ≤SOC t ≤SOC max ;

[0045]

[0046]

[0047] Q gc ≥γ(E total -E re );

[0048]

[0049] Where: p G,t is the output of thermal power units in the new power system during period t; p pv,t is the output of the photovoltaic unit of the new power system in period t; p w,t is the wind turbine output of the new power system in period t; p x,t is the penalty power of the new power system in period t; p base,t is the basic load of the new power system in period t; p bat,tis the energy storage output of the new power system in period t; p car,t is the required load of electric vehicles in the new power system during period t; p cut,t is the load shedding of the new power system in period t. i,max and P i,min is the maximum and minimum power of the i-th unit; and is the up-ramp power and down-ramp power of the i-th unit. and The maximum charge and discharge power of energy storage per unit time; and is the charging and discharging power of the energy storage during period t; It represents the state of charge of the energy storage in the period t; η c and η d is the charging and discharging efficiency of energy storage; and It is the maximum and minimum state of charge of energy storage. i,t is the state of charge of the ith electric vehicle at time period t; ε ev is the self-discharge rate of the battery of the electric vehicle; u ev,c and u ev,d is a 01 variable, which is the charging and discharging state of the electric vehicle; C ev and U ev is the rated capacitance and voltage of the electric vehicle battery; η ev,c and η ev,d is the charging and discharging efficiency of the electric vehicle; and is the charging and discharging power of the ith electric vehicle in period t, both of which are positive values. SOC max and SOC min Indicates the maximum and minimum state of charge of electric vehicle batteries; SOC set Indicates the SOC expected by electric vehicle owners out The value of t end -t is the charging time required to reach the charging amount set by the electric car owner; P dis_max and P ch_max The purpose is to achieve the maximum discharge and charging power that electric vehicles can achieve; represents the absolute value of the maximum charging power of the i-th electric vehicle in period t; J EV_ch,t To determine whether an electric vehicle is overcharged, E total is the total power consumption of the power system, E re is the power generated by renewable energy, Q gc The number of green certificate transactions.

[0050] It should be noted that the worst-case risk probability assessment theory is as follows:

[0051]

[0052] Wlq β (x) = supCVaR(x);

[0053]

[0054] Where η is the VaR value at a given confidence level β, [f(x,y)-η] + max{f(x,y)-η,0}. Where: N ζ is the number of scenes in the data set. The probability density function p(y) of y belongs to a set of probability distributions ξ, and some information about ξ is known.

[0055] As an optional implementation, Figure 3 This is the workflow of the two-stage risk aversion model; the objective function of the two-stage risk aversion model is:

[0056]

[0057] in, is the objective function of the day-ahead optimization scheduling model,

[0058] is the objective function of the intraday real-time dispatch model; the decision variable x is the required power of the net load, F 1 is the total cost of the day-ahead operation, Δx is the decision variable of the second stage, τ sm is the net load forecast value, ρ is the risk aversion factor, F 1,m (Δx|τ sm ) represents the cost incurred during real-time power balancing, Φ β,m (Δx|τ sm ) represents the threshold value corresponding to the economic risk caused by the error between the net load forecast value and the actual value.

[0059] Among them, the intraday real-time scheduling model, objective function and corresponding real-time control constraints are as follows:

[0060]

[0061] Where: and It is the scene sm The increase and decrease of the power of the i-th thermal power unit in the t period caused by the error between the predicted value and the actual value of the net load on the day before; and is the unit adjustment cost and reduction cost of the i-th unit; and For the scene smThe charging and discharging power adjustment of the i-th electric vehicle in period t is generated by the error between the day-ahead net load forecast value and the real-time value; Battery loss cost caused by adjusting the output of energy storage; sm For the scene sm The corresponding probability.

[0062] The corresponding real-time control constraints are as follows:

[0063]

[0064]

[0065] u ev,c +u ev,d =1;

[0066]

[0067]

[0068] As an optional implementation, the objective function of the day-ahead optimization scheduling model is expressed as: Among them, F I is the cost of thermal power, F V2G is the charging and discharging cost of electric vehicles, F cut is the load shedding cost, F car is the carbon emission cost, F gc is the transaction cost of green certificates. The specific expressions of each cost can be recorded as follows:

[0069]

[0070] F gc =p gc Q gc ;

[0071] Where: F I is the thermal power cost, T represents the power generation time, generally 24 hours, P i I is the active power output of the ith generator; c i is the price per kilowatt of electricity generated by the i-th generator; and is the up-ramp and down-ramp power of the i-th unit in period t; and is the unit power cost of ramping up and ramping down of the i-th thermal power unit. cut is the load shedding cost; The load shedding power at time t, C cut High electricity prices due to load shedding, Abandoned power. is the battery charge and discharge loss cost of the i-th electric vehicle at time t, Γ i,t is the cycle charge and discharge amount of the i-th electric vehicle at time t, is the battery capacity of i electric vehicles, C change The cost of replacing a battery. V2G is the cost of charging and discharging electric vehicles, All electric vehicle charging and discharging costs, is the charging power of the ith electric vehicle at time t, is the discharge power of the ith electric vehicle at time t, and is the charging / discharging price of the ith electric vehicle at time t, is the service fee charged by the i-th electric vehicle at time t. is the charge and discharge loss cost of the ith battery at time t, is the cycle charge and discharge capacity of the energy storage battery at time t, is the battery capacity, F is the cost of replacing a battery. car is the carbon emission cost, P car Unit carbon emission price, δ gt is the carbon emission per unit power of thermal power units, F gc is the transaction cost of green certificates, p gc is the unit green certificate transaction cost.

[0072] As an optional implementation, step S2 includes:

[0073] S21: Component reduction of multi-component Gaussian mixture models;

[0074] S22: sampling the reduced multi-component Gaussian mixture model to obtain the net load day-ahead forecast data;

[0075] S23: smoothing the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a modified Gaussian mixture model;

[0076] S24: Generate a set of net load uncertainty scenarios using the modified Gaussian mixture model.

[0077] As an optional implementation, Figure 4 As shown, step S21 includes: performing component reduction on the multi-component Gaussian mixture model based on component-retaining hierarchical density reduction to obtain a reduced multi-component Gaussian mixture model.

[0078] Among them, the multi-component Gaussian mixture model is expressed as: and

[0079] The process of reducing components of a multi-component Gaussian mixture model based on component-preserving hierarchical density reduction is as follows:

[0080] The reduced multi-component Gaussian mixture model is expressed as: The corresponding constraints are: Further,

[0081] The expectation step and maximization step corresponding to the reduced multi-component Gaussian mixture model are as follows:

[0082] Expectation steps:

[0083]

[0084] Maximization step:

[0085]

[0086] Where: f (b) (x) is the probability density function of data x under GMM1; the Gaussian component number K of GMM1 b is equal to n, is the weight of the i-th Gaussian component of GMM1; μ (b) and Σ (b) are the mean vector and covariance matrix respectively. (m) (x) is the probability distribution model of x after component simplification, K m is the number of Gaussian components after reduction. i,j is the variational parameter, subscript i∈[1,K b ]; subscript l, j∈[1,K m ]; tr is the trace of the matrix; the number of virtual samples M is generally 10 times K b ; K m <K b .

[0087] As an optional implementation, step S23 includes: using a covariance matrix correction method to smooth the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a corrected Gaussian mixture model.

[0088] The modified Gaussian mixture model is expressed as follows: and The corresponding Expectation step and Maximization step are as follows:

[0089] Modified Expectation steps:

[0090]

[0091] Modified Maximization step:

[0092]

[0093] Where: κ is the correction coefficient of the covariance matrix, is the modified covariance matrix of the mth Gaussian component, is the main diagonal element matrix of the covariance matrix of the mth Gaussian component before correction, is the non-diagonal element matrix of the covariance matrix of the mth Gaussian component. is the modified variational parameter, subscript i∈[1,K b ]; is the modified mathematical expectation; and are the modified weight, mean and covariance matrix respectively.

[0094] As an optional implementation, step S24 includes: using a scenario generation method to process the modified Gaussian mixture model to obtain a net load uncertainty scenario set.

[0095] Example 2

[0096] This embodiment provides a device for generating output control strategy of a new type of power system, including: a characterization module, an acquisition module, an input module and an optimization module. Among them, the characterization module is used to characterize the net load historical data of the new type of power system as a multi-component Gaussian mixture model; the acquisition module is used to obtain the net load day-ahead forecast data and the net load uncertainty scenario set using the multi-component Gaussian mixture model corresponding to the net load historical data; the input module is used to input the net load day-ahead forecast data and the net load uncertainty scenario set into a two-stage risk aversion model; the two-stage risk aversion model includes a day-ahead optimization scheduling model and an intraday real-time scheduling model; among them, the day-ahead optimization scheduling model aims to minimize the operating cost of the new type of power system, and is used to input the net load day-ahead forecast data to obtain the optimal scheduling plan for the day-ahead; the intraday real-time scheduling model is constructed based on the worst-case risk probability assessment theory, and is used to input the net load uncertainty scenario set to obtain the intraday real-time scheduling plan; the optimization module is used to optimize the optimal scheduling plan for the day-ahead using the intraday real-time scheduling plan to obtain the target output control strategy.

[0097] Furthermore, the objective function of the two-stage risk aversion model is:

[0098]

[0099] in, is the objective function of the day-ahead optimization scheduling model, is the objective function of the intraday real-time dispatch model; the decision variable x is the required power of the net load, F 1 is the total cost of the day-ahead operation, Δx is the decision variable of the second stage, τ sm is the net load forecast value, ρ is the risk aversion factor, F 1,m (Δx|τ sm ) represents the cost incurred during real-time power balancing, Φ β,m (Δx|τ sm ) represents the threshold value corresponding to the economic risk caused by the error between the net load forecast value and the actual value.

[0100] Furthermore, the objective function of the day-ahead optimization scheduling model is expressed as: Among them, F I is the cost of thermal power, F V2G is the charging and discharging cost of electric vehicles, F cut is the load shedding cost, F car is the carbon emission cost, F gc The transaction cost of green certificates.

[0101] Furthermore, the acquisition module is used to perform the following steps:

[0102] Perform component reduction on multi-component Gaussian mixture models;

[0103] The multi-component Gaussian mixture model after reduction is sampled to obtain the net load day-ahead forecast data;

[0104] Smoothing the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a modified Gaussian mixture model;

[0105] The modified Gaussian mixture model is used to generate a set of net load uncertainty scenarios.

[0106] Furthermore, the multi-component Gaussian mixture model is subjected to component reduction, including: performing component reduction on the multi-component Gaussian mixture model based on component-retaining hierarchical density reduction to obtain a reduced multi-component Gaussian mixture model.

[0107] Furthermore, the fitting curve corresponding to the reduced multi-component Gaussian mixture model is smoothed to obtain a modified Gaussian mixture model, including:

[0108] The covariance matrix correction method is used to smooth the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain the corrected Gaussian mixture model.

[0109] Furthermore, the modified Gaussian mixture model is used to generate a net load uncertainty scenario set, including: using a scenario generation method to process the modified Gaussian mixture model to obtain the net load uncertainty scenario set.

[0110] Example 3

[0111] This embodiment provides a control system for a new type of power system. The new type of power system includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of the output control strategy generation method provided in Example 1 are implemented.

[0112] Example 4

[0113] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the output control strategy generation method provided in Embodiment 1 are implemented.

[0114] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A new method for generating an output control strategy for a power system, characterized in that: include: The net load history data of the new power system is represented as a multi-component Gaussian mixture model; Obtaining net load day-ahead forecast data and a net load uncertainty scenario set using a multi-component Gaussian mixture model corresponding to the net load historical data; The net load day-ahead forecast data and the net load uncertainty scenario set are input into a two-stage risk aversion model; the two-stage risk aversion model includes a day-ahead optimization dispatch model and an intraday real-time dispatch model; wherein the day-ahead optimization dispatch model aims to minimize the operating cost of the new power system, and is used to input the net load day-ahead forecast data to obtain the day-ahead optimal dispatch plan; the intraday real-time dispatch model is constructed based on the worst-case risk probability assessment theory, and is used to input the net load uncertainty scenario set to obtain the intraday real-time dispatch plan; The real-time dispatch plan within the day is used to optimize the optimal dispatch plan on the day before, and a target output control strategy is obtained.

2. The method for generating an output control strategy for a new power system according to claim 1, characterized in that: The objective function of the two-stage risk aversion model is: in, is the objective function of the day-ahead optimization scheduling model, is the objective function of the intraday real-time scheduling model; the decision variable x is the required power of the net load, F1 is the total cost of the day-ahead operation, Δx is the second-stage decision variable, τ sm is the net load forecast value, ρ is the risk aversion factor, F 1,m (Δx|τ sm ) represents the cost incurred during real-time power balancing, Φ β,m (Δx|τ sm ) represents the threshold value corresponding to the economic risk caused by the error between the net load forecast value and the actual value.

3. The method for generating an output control strategy for a new power system according to claim 2, characterized in that: The objective function of the day-ahead optimization scheduling model is expressed as: Among them, F I is the cost of thermal power, F V2G is the charging and discharging cost of electric vehicles, F cut is the load shedding cost, F car is the carbon emission cost, F gc The transaction cost of green certificates.

4. The method for generating an output control strategy for a new power system according to claim 1, characterized in that: The method of using a multi-component Gaussian mixture model corresponding to the net load historical data to obtain net load day-ahead forecast data and a net load uncertainty scenario set includes: Performing component reduction on the multi-component Gaussian mixture model; The multi-component Gaussian mixture model after reduction is sampled to obtain the net load day-ahead forecast data; Smoothing the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a modified Gaussian mixture model; The modified Gaussian mixture model is used to generate a net load uncertainty scenario set.

5. The method for generating an output control strategy for a new power system according to claim 4, characterized in that: The step of performing component reduction on the multi-component Gaussian mixture model comprises: The multi-component Gaussian mixture model is subjected to component reduction based on component-preserving hierarchical density reduction to obtain the reduced multi-component Gaussian mixture model.

6. The method for generating an output control strategy for a new power system according to claim 4, characterized in that: The step of smoothing the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a modified Gaussian mixture model includes: The covariance matrix correction method is used to smooth the fitting curve corresponding to the reduced multi-component Gaussian mixture model to obtain a corrected Gaussian mixture model.

7. The method for generating an output control strategy for a new power system according to claim 6, characterized in that: The method of generating a net load uncertainty scenario set by using the modified Gaussian mixture model includes: The modified Gaussian mixture model is processed using a scenario generation method to obtain the net load uncertainty scenario set.

8. A novel device for generating output control strategy of a power system, characterized in that: include: A characterization module, used to characterize the net load history data of the new power system as a multi-component Gaussian mixture model; An acquisition module, used for acquiring net load day-ahead forecast data and a net load uncertainty scenario set using a multi-component Gaussian mixture model corresponding to the net load historical data; An input module is used to input the net load day-ahead forecast data and the net load uncertainty scenario set into a two-stage risk aversion model; the two-stage risk aversion model includes a day-ahead optimization dispatch model and an intraday real-time dispatch model; wherein the day-ahead optimization dispatch model aims to minimize the operating cost of the new power system, and is used to input the net load day-ahead forecast data to obtain the day-ahead optimal dispatch plan; the intraday real-time dispatch model is constructed based on the worst-case risk probability assessment theory, and is used to input the net load uncertainty scenario set to obtain the intraday real-time dispatch plan; The optimization module is used to optimize the optimal scheduling plan of the previous day by using the real-time scheduling plan within the day to obtain a target output control strategy.

9. A control system for a new type of power system, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method for generating an output regulation strategy for a new power system according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for generating an output regulation strategy for a new power system according to any one of claims 1 to 7 are implemented.

Images