Novel electric power system day-ahead optimization scheduling method based on evolution of virtual net load

Abstract

The invention discloses a novel electric power system day-ahead optimization scheduling method based on evolution of virtual net load, and belongs to the technical field of electric power system day-ahead optimization scheduling. Performing scene generation on the time sequence of the wind and light output and the load by using an Informer model; an evolution virtual net load is defined on the basis of the virtual net load, and the evolution virtual net load of the power system is calculated; the method comprises the following steps: considering the operation characteristics of a system, taking supportability to purchase renewable energy electric quantity as a basis, taking minimization of the total cost of the system including the operation cost and the carbon emission cost of each unit as a target, and reflecting the demand of source load uncertainty by a system flexibility constraint condition; establishing a novel electric power system day-ahead optimization scheduling model based on evolution of the virtual net load; and solving the model to obtain a novel power system day-ahead optimization scheduling plan based on evolution of the virtual net load. According to the method, the capacity of the system for coping with the uncertain fluctuation of the source load can be effectively improved, the carbon emission cost is reduced, and the consumption of the on-grid electric quantity of a renewable energy power generation project is promoted.

Description

Technical Field

[0001] The invention belongs to the field of power system day-ahead optimization dispatching in power system planning and operation, and in particular relates to a novel power system day-ahead optimization dispatching method based on evolving virtual net load. Background Art

[0002] Faced with the problem of optimizing the dispatch of new power systems with a high proportion of renewable energy under the condition of considering uncertain factors, the early dispatch method research considering uncertain factors mainly focused on the opportunity constraint method, robust optimization method and traditional scenario generation method. Although certain effects have been achieved, the complexity of the model has been increased and the calculation is relatively cumbersome. At present, the problem of renewable energy power consumption in power systems has been studied in depth, but most of the research is based on the "Measures for the Supervision of the Full Purchase of Renewable Energy Power by Power Grid Enterprises" (hereinafter referred to as the "Old Measures") to formulate dispatch methods and power consumption plans. These methods and plans are no longer applicable after the "Measures for the Supervision of the Full Guaranteed Purchase of Renewable Energy Power" (hereinafter referred to as the "Measures") were issued and implemented.

[0003] "Net load" refers to the difference between the total load and wind and solar power generation in the power system during a certain period of time. The introduction of this concept provides convenience for the study of power system planning and operation, grid stability analysis and demand side management. "Virtual load" is a multi-source heterogeneous controllable integrated load. By applying advanced communication technology and software architecture, the diversified load resources distributed in different locations are centralized and integrated to achieve the identification, complementarity and coordinated management of these loads. The proposal of the virtual load management model aims to promote the comprehensive development of smart grids, ensure the efficient use of power resources and the stable and safe operation of the power grid. "Virtual net load" refers to the load after the minimum technical output of thermal power and the forced output of hydropower are removed from the net load. It unifies the research on power supply operation constraints and source-load uncertainty, making grid dispatch clear and simple. However, the virtual net load defined in the "Old Measures" lost its theoretical basis after the "Old Measures" were abolished and its dispatch method is no longer applicable. In order to clarify the distinction between "net load", "virtual load" and "virtual net load" and implement the Measures, the present invention defines "evolved virtual net load" as follows: in a power system containing wind, solar, hydro, thermal and storage, the total load in a certain period of time minus the power corresponding to the guaranteed purchase of electricity in renewable energy (wind, solar and hydro) power generation and the minimum technical output of thermal power, the load obtained is called the evolved virtual net load of that period.

[0004] Therefore, there is an urgent need in the existing technology for a new power system day-ahead optimization dispatching method that comprehensively considers wind and solar power generation and load power forecasting and evolves virtual net load. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a new power system day-ahead optimization scheduling method based on evolving virtual net load, which is used to solve the problems in the prior art that the uncertainty of source and load of the new power system has an adverse effect on the formulation of scheduling plans, and there is a lack of a method for fully guaranteeing the purchase and consumption of renewable energy electricity.

[0006] The technical solution adopted by the present invention is to provide a new power system day-ahead optimal dispatching method based on evolving virtual net load, comprising the following steps:

[0007] 1) Use the Informer model to generate scenarios for the time series of wind and solar power output and load, and establish a wind and solar power generation and load power prediction model based on the Informer model;

[0008] 2) defining the evolving virtual net load based on the virtual net load and calculating the evolving virtual net load of the power system;

[0009] 3) Considering the operating characteristics of the system, based on the guaranteed purchase of renewable energy electricity, with the goal of minimizing the total system cost including the operating cost of each unit and the carbon emission cost, and with the system flexibility constraint condition reflecting the demand for source and load uncertainty, a day-ahead optimal scheduling model based on evolving virtual net load is established;

[0010] 4) Solve the model and obtain a new power system day-ahead optimal dispatch plan based on evolving virtual net load.

[0011] The specific method of establishing the wind-solar power generation and load power prediction model based on the Informer model in step 1) is:

[0012] ① The historical data of wind and solar power generation and load power are normalized respectively, and then the normalized data sets are divided into training set, test set and validation set. p and timestamp feature w time Reconstruct and combine to get W E ={w p , w time}, and with W E As input data for prediction models;

[0013] ② The multi-headed probabilistic sparse self-attention layer of the encoder in the prediction model is used to mine the long-term dependencies between data to reduce the time complexity and extract the characteristics of the input wind and solar power generation and load power sequence;

[0014] ③ Through the self-attention distillation layer of the encoder in the prediction model, the output of each probabilistic sparse self-attention layer is convolved and pooled, and the output of each layer is halved while retaining the key information of the data, thereby reducing the amount of network calculation;

[0015] ④ The w composed of some historical power data of the encoder and future power occupancy data p ′ and its corresponding timestamp feature w time ′ is reconstructed and combined to obtain W D ={w p ′,w time ′} is used as the input of the decoder in the prediction model, and the combined sequence of input historical power and zero value is paid attention to through the masked multi-head probabilistic sparse self-attention layer. In this process, the masking mechanism can be introduced to prevent the model from falling into autoregression, and the combined sequence features can be extracted;

[0016] ⑤ Input the feature vector processed by the masked multi-head probabilistic sparse self-attention layer in the decoder and the feature vector output by the encoder into the multi-head attention layer in the decoder for multi-head attention calculation, and output the calculation results using a generative method to achieve one-time generation of the output sequence, thereby shortening the dynamic decoding time and reducing memory usage;

[0017] ⑥Set up a fully connected layer to convert the output dimension of the decoder into the output dimension required for prediction, and finally perform denormalization on the output result to obtain the predicted value of wind and solar power generation and and the predicted value of load power P L .

[0018] The calculation method of the evolved virtual net load in step 2) is formula (1),

[0019]

[0020] Where P VL,t is the evolving virtual net load of the system during period t; P L,t is the total load of the system during period t; β 1,t is the guaranteed purchase coefficient of wind and solar power in period t; β 2,t is the hydropower security acquisition coefficient for period t; and are the maximum power generation that wind power, photovoltaic power and hydropower can achieve during period t; N h 、N g are the total number of hydropower units and thermal power units respectively; is the minimum technical output of thermal power unit i in period t.

[0021] The objective function of the day-ahead optimization scheduling model based on the evolving virtual net load in step 3) is formula (2):

[0022] min C=C g +C h +C w +C pv+C ps +C carbon (2)

[0023] Where, C is the total system cost; C g is the operating cost of thermal power units; C h is the operating cost of the hydropower unit; C w is the wind farm operating cost; C pv is the operating cost of the photovoltaic power station; C ps is the start-up and shutdown cost of the pumped storage unit; C carbon The cost of carbon emissions.

[0024] The operating cost of the thermal power unit is C g 、Operation cost of hydropower unit C h , Wind farm operating costs C w and the operating cost of the photovoltaic power station C pv , and the start-up and shutdown cost C of the pumped storage unit ps and carbon emission cost C carbon The calculation method is as follows:

[0025] aThe operating cost of thermal power unit C g See formula (3),

[0026]

[0027] Where T is the total number of time periods in the scheduling cycle; a i 、b i and c i are the cost coefficients of thermal power unit i; P g,i,t is the power generation power of thermal power unit i during period t;

[0028] b. Operation cost of hydropower unit C h See formula (4),

[0029]

[0030] In the formula, k h P is the unit power generation cost coefficient of the hydropower unit; h,i,t is the power generation of hydropower unit i during period t;

[0031] c Wind farm operation cost w See formula (5),

[0032] C w =k w P w,t (5)

[0033] In the formula, k w is the power generation cost coefficient of the wind farm; P w,tis the power generation of the wind farm during period t;

[0034] d Photovoltaic power station operation cost C pv See formula (6),

[0035] C pv =k pv P pv,t (6)

[0036] In the formula, k pv is the power generation cost coefficient of photovoltaic power generation; P pv,t is the power generation of the photovoltaic power station during period t;

[0037] eStart-up and shutdown cost of pumped storage unit C ps See formula (7),

[0038]

[0039] Where N ps is the total number of pumped storage units; They are the power generation and pumping state variables of pumped storage unit i in period t, which are 1 when working and 0 when stopped; are the power generation and pumping costs of pumped storage unit i, respectively;

[0040] fCarbon emission cost C carbon See formula (8),

[0041]

[0042] In the formula, C carbon is the carbon emission cost; η is the carbon trading price; E P is the total carbon emissions during the scheduling period; E L is the total carbon emission quota in a scheduling cycle; δ i is the carbon emission of thermal power unit i; ε is the unit electricity quota coefficient of system carbon emission.

[0043] The constraints of the day-ahead optimization scheduling model based on the evolving virtual net load in step 3) include:

[0044] a The operating constraints of thermal power units are shown in formula (9).

[0045]

[0046] In the formula, are the upper and lower limits of the output of thermal power unit i; P g,i,t-1 is the dispatch output of thermal power unit i during period t-1; are the upper and lower limits of the ramp rate of thermal power unit i respectively;

[0047] b The operating constraints of the hydropower unit are shown in formula (10),

[0048]

[0049] In the formula, the hydropower security acquisition coefficient β2 is 0.4. is the power corresponding to the market transaction electricity of hydropower unit i in period t; are the upper and lower limits of the output of hydropower unit i; P h,i,t-1 is the dispatching output of hydropower unit i during period t-1; are the upper and lower limits of the climbing slope of hydropower unit i respectively;

[0050] c Wind farm constraints are shown in formula (11),

[0051]

[0052] In the formula, the non-hydropower guaranteed acquisition coefficient β1 is 0.3. is the power corresponding to the market transaction electricity of the wind farm during period t;

[0053] d The constraints of the photovoltaic power station are shown in formula (12),

[0054]

[0055] In the formula, the non-hydropower guaranteed acquisition coefficient β1 is 0.3. The power corresponding to the market transaction electricity of the photovoltaic power station during period t;

[0056] The operating constraints of the pumped storage unit are shown in formula (13).

[0057]

[0058] Where P ps,i,t , P pm,i,t are the power generation and pumping power of pumped storage unit i during period t, respectively; are the upper limits of power generation and pumping power of pumped storage unit i, respectively; are the capacities of the upper and lower reservoirs in period t, respectively; are the upper and lower reservoir capacity limits, respectively; are the lower limits of the upper and lower reservoir capacities, respectively; is the upper reservoir capacity during period t-1; are the water consumption and pumping capacity of pumped storage unit i during period t; η gen , η pm are the hydropower conversion coefficients for power generation and pumping of pumped storage unit i, respectively;

[0059] The power balance constraint f is shown in formula (14),

[0060]

[0061] In the formula, is the peak load of thermal power during period t;

[0062] gFlexibility constraints include flexibility demand constraints (see formula (15), flexibility supply constraints (see formula (16)-(17) and flexibility margin constraints (see formula (18).

[0063]

[0064] In the formula, △P VL,t is the variation of the evolving virtual net load between the next period and period t; are the upward and downward flexibility requirements of the system in period t respectively; P w,t+1 , P pv,t+1 , P L，t+1 Forecast output for wind power, photovoltaic power and load in period t+1; ω u and ω d , γ u and γ d , u and λ d are the upward and downward flexibility demand coefficients of wind power, photovoltaic power, and load caused by the forecast error; λ e The upward flexibility demand coefficient for unit failure maintenance;

[0065]

[0066] In the formula, for t The upward flexibility supply of the time slot system; for t Upward flexibility supply of thermal power, hydropower and pumped storage units during the period;

[0067]

[0068] In the formula, for t Downward flexibility provision of the time slot system; for t Downward flexibility supply of thermal power, hydropower and pumped storage units during the time period;

[0069]

[0070] In the formula, Respectively t The upward and downward flexibility margin possessed by the time period system.

[0071] Through the above design scheme, the present invention can bring the following beneficial effects:

[0072] a) In the wind and solar power generation and load power forecasting links, the present invention can deeply learn the long-term dependencies and complex nonlinear relationships in time series, and has a good effect on the forecasting objects driven by multiple factors and with complex dynamic characteristics.

[0073] b) For renewable energy power generation companies, the present invention can well implement the guaranteed purchase mechanism and fully absorb the guaranteed purchase electricity of renewable energy grid-connected electricity, so that they can obtain stable income; at the same time, it can encourage power generation companies to make full use of market bidding opportunities, reduce power generation costs through technological progress, strive for more trading opportunities, and obtain greater benefits.

[0074] c) For the dispatching department of the power system, the present invention takes into account constraints such as power supply flexibility while tracking the evolution of the virtual net load. This method can not only ensure the safe operation of the system, but also effectively reduce the power generation cost of each unit and increase carbon revenue, thereby improving the overall economic benefits of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a schematic diagram of the basic principle of a novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0076] Figure 2 This is a wind and solar power generation and load power prediction model diagram based on the Informer model of the novel power system day-ahead optimization dispatching method based on evolving virtual net load of the present invention;

[0077] Figure 3 It is a wind power prediction curve diagram of the whole province based on the novel power system day-ahead optimization dispatching method of the present invention based on evolving virtual net load;

[0078] Figure 4 It is a photovoltaic power prediction curve diagram of the whole province based on the novel power system day-ahead optimization dispatching method of the present invention based on evolving virtual net load;

[0079] Figure 5 It is a load forecast curve diagram of the whole province of the novel power system day-ahead optimization dispatching method based on the evolving virtual net load of the present invention;

[0080] Figure 6 It is a province-wide virtual net load curve and an evolving virtual net load curve diagram of a novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0081] Figure 7 It is an output curve diagram of each unit after the day-ahead optimization dispatching of a new type of power system based on evolving virtual net load of the present invention;

[0082] Figure 8 It is the output curve diagram of each unit after dispatching of method 1 of the novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0083] Fig. 9 It is the output curve diagram of each unit after dispatching according to method 2 of the novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0084] Fig.10 It is the output curve diagram of each unit after dispatching of method 3 of the novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0085] Fig.11 It is the output curve diagram of each unit after dispatching of method 4 of the novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0086] Fig.12 It is an upward flexibility margin curve diagram of the novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention;

[0087] Fig.13 It is a downward flexibility margin curve diagram of the novel power system day-ahead optimal dispatching method based on evolving virtual net load of the present invention. DETAILED DESCRIPTION

[0088] The present invention is further described below with reference to the accompanying drawings and specific embodiments:

[0089] Specific embodiment: A new power system day-ahead optimal dispatching method based on evolving virtual net load, the basic principle diagram is as follows Figure 1 It includes: using the Informer model to build a prediction model for wind and solar power generation and load power, calculating the evolving virtual net load, establishing a new power system day-ahead optimal dispatch model based on the evolving virtual net load, and solving the new power system day-ahead optimal dispatch model based on the evolving virtual net load. The specific contents are:

[0090] 1) Parameter setting and scenario prediction of wind and solar power and load

[0091] The parameters of the hydropower units and other parameters are shown in Tables 1 and 2;

[0092] Table 1 Hydropower unit parameters

[0093]

[0094] Table 2 Other parameters

[0095]

[0096] Figure 2This is a wind and solar power generation and load power prediction model diagram based on the Informer model. Using the Informer model to build a wind and solar power generation and load power prediction model can overcome the shortcomings of the Transformer model such as high time complexity, high memory occupancy and inherent limitations of the encoder-decoder architecture, and is more effective in processing and predicting data with complex time dependencies.

[0097] Figure 3 , Figure 4 and Figure 5 The wind-solar power generation and load power curves are obtained by respectively using the prediction model of the present invention and the auto-regressive integrated moving average (ARIMA) model to perform scenario prediction on wind-solar power generation and load.

[0098] Figure 6 It is the virtual net load curve and the evolving virtual net load curve of the whole province, which intuitively reflects the dynamic changes of the virtual net load and evolving virtual net load of the whole province on that day. When the evolving virtual net load curve is in the positive interval, it indicates that the guaranteed purchase electricity of wind power and photovoltaic power generation in the province has been fully absorbed; if the curve is in the negative interval, it indicates that there is a lot of renewable energy power generation in the province. At this time, pumped storage units should be fully mobilized to adjust the excess renewable energy power generation to ensure that the guaranteed purchase electricity of renewable energy is fully absorbed.

[0099] 2) Solve a new power system day-ahead optimal dispatch model based on evolving virtual net load

[0100] Four different scheduling methods were designed for comparative analysis. The design methods are as follows:

[0101] Method 1 is to use the prediction model of the present invention to predict wind and solar power output and load power scenarios. On the basis of satisfying the operating constraints of each unit and the system power balance constraints, the rotating reserve capacity is set according to the constant proportion of the total load of the system and optimized and scheduled.

[0102] Method 2 is based on Method 1 and ensures that when the wind and solar power output and load power deviate from the expected maximum, the system can still meet the spinning reserve constraints (that is, the uncertainty of wind and solar power output and load power is considered in the spinning reserve constraints) and perform optimized scheduling.

[0103] Method 3 uses a method based on the ARIMA model to predict wind and solar output and load power scenarios, and uses the scheduling method of Method 1 for optimized scheduling.

[0104] Method 4 is to use the method based on ARIMA model to predict wind and solar scenarios and load power scenarios, and use the scheduling method of method 2 for optimized scheduling.

[0105] Figure 7-Figure 11 After optimizing the power system using the new power system day-ahead optimal dispatching method based on evolving virtual net load, method 1, method 2, method 3 and method 4, the output curves of each unit are obtained.

[0106] Fig.12 is the upward flexibility margin of different scheduling methods. All five scheduling methods have positive upward flexibility margins, which can ensure the upward adjustment capacity required for stable operation of the system. However, too high flexibility margins may lead to rising costs and waste of resources, thus affecting the economic efficiency of system operation. Therefore, under the condition of positive margin, a lower upward flexibility margin is more conducive to improving economic efficiency. By analyzing the upward flexibility margin curves of each method, it can be seen that the proposed scheduling method can reduce resource waste on the basis of ensuring the necessary flexibility of the system, and has more advantages in terms of economy.

[0107] Fig.13 is the downward flexibility margin of different dispatching methods. The smaller the downward flexibility margin, the better the economic efficiency of the system operation. However, if the downward flexibility margin is lower than zero, it may mean that the system is insufficient in dealing with load fluctuations and uncertainty in renewable energy generation, making it difficult to meet power balance requirements and affecting system safety. By analyzing the downward flexibility margin curves of each method, it can be seen that the proposed dispatching method can achieve economic benefits while ensuring safe and stable operation of the system.

[0108] The daily operating cost of the system after optimized scheduling using different scheduling methods is shown in Table 3. Compared with other scheduling methods, the scheduling method proposed in the present invention has relatively small operating costs for each unit and higher carbon benefits, thereby reducing the overall power generation cost of the system and achieving better economy.

[0109] Table 3 System daily operation cost under different scheduling methods

[0110]

[0111] By comparing and analyzing the system's flexibility margin curve and daily power generation cost, we can see that:

[0112] Method 1 focuses on maintaining sufficient spinning reserve to ensure the safety of system operation, but does not fully consider the uncertainty of wind, solar and load, resulting in excessive flexibility margin. In addition, the operating cost of each unit in this method is high, affecting the overall economy and safety of the system.

[0113] Method 2 takes the uncertainty of wind and solar power generation into consideration during scheduling to ensure the safety of the system under extreme conditions. Although it can reduce the excess flexibility margin compared to Method 1, this method is still conservative in reducing unit operating costs and promoting the consumption of renewable energy. Its scheduling results undermine economic efficiency due to excessive focus on safety.

[0114] Method 3 and Method 4 use the ARIMA model to predict the scenario of wind and solar power generation and load power. The model only relies on historical power data for prediction without additional variables. However, the model requires that the input historical data is close to the Gaussian distribution, and when the non-stationary series is differentiated to achieve stationarity, excessive differentiation may lose key data information, thereby affecting the accuracy of the scenario prediction. Compared with the prediction method proposed in the present invention, Method 3 and Method 4 perform poorly in terms of the economic efficiency of scheduling.

[0115] The scheduling method proposed in the present invention adopts the wind and solar power generation and load power prediction model based on the Informer model for scenario prediction. Through its advanced attention mechanism, it can better extract time-dependent features from the data, mine and learn the complex nonlinear relationships in the power data, so as to predict wind and solar power generation and load power almost accurately. The optimization scheduling method proposed in the present invention can not only ensure the stable operation of the power system, but also effectively avoid redundant consumption of resources by reducing unnecessary flexibility margins. At the same time, the method can complete the consumption of renewable energy grid-connected electricity by optimizing load curve operation. Compared with the other four methods, the scheduling method proposed in the present invention has more significant advantages in improving economic benefits and ensuring safe and stable operation of the system.

[0116] The implementation methods of the present invention are not limited to the above-mentioned embodiments, and any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A new power system day-ahead optimal dispatching method based on evolving virtual net load, characterized by: The following steps are involved: 1) Use the Informer model to generate scenarios for the time series of wind and solar power output and load, and establish a wind and solar power generation and load power prediction model based on the Informer model; 2) defining the evolving virtual net load based on the virtual net load and calculating the evolving virtual net load of the power system; 3) Considering the operating characteristics of the system, based on the guaranteed purchase of renewable energy electricity, with the goal of minimizing the total system cost including the operating cost of each unit and the carbon emission cost, and with the system flexibility constraint condition reflecting the demand for source and load uncertainty, a day-ahead optimal scheduling model based on evolving virtual net load is established; 4) Solve the model and obtain a new power system day-ahead optimal dispatch plan based on evolving virtual net load.

2. The novel power system day-ahead optimal dispatching method based on evolving virtual net load according to claim 1 is characterized by: The specific method of establishing the wind-solar power generation and load power prediction model based on the Informer model in step 1) is: ① The historical data of wind and solar power generation and load power are normalized respectively, and then the normalized data sets are divided into training set, test set and validation set. p and timestamp feature w time Reconstruct and combine to get W E ={w p , w time }, and with W E As input data for prediction models; ② The multi-headed probabilistic sparse self-attention layer of the encoder in the prediction model is used to mine the long-term dependencies between data to reduce the time complexity and extract the characteristics of the input wind and solar power generation and load power sequence; ③ Through the self-attention distillation layer of the encoder in the prediction model, the output of each probabilistic sparse self-attention layer is convolved and pooled, and the output of each layer is halved while retaining the key information of the data, thereby reducing the amount of network calculation; ④ The w composed of some historical power data of the encoder and future power occupancy data p ′ and its corresponding timestamp feature w time ′ is reconstructed and combined to obtain W D ={w p ′,w time ′} is used as the input of the decoder in the prediction model, and the combined sequence of input historical power and zero value is calculated through the masked multi-head probabilistic sparse self-attention layer. In this process, the mask mechanism can be introduced to prevent the model from falling into autoregression, and the combined sequence features can be extracted; ⑤ Input the feature vector processed by the masked multi-head probabilistic sparse self-attention layer in the decoder and the feature vector output by the encoder into the multi-head attention layer in the decoder for multi-head attention calculation, and output the calculation results using a generative method to achieve one-time generation of the output sequence, thereby shortening the dynamic decoding time and reducing memory usage; ⑥Set up a fully connected layer to convert the output dimension of the decoder into the output dimension required for prediction, and finally perform denormalization on the output result to obtain the predicted value of wind and solar power generation and and the predicted value of load power P L .

3. The novel power system day-ahead optimal dispatching method based on evolving virtual net load according to claim 1 is characterized by: The calculation method of the evolved virtual net load in step 2) is formula (1), Where P VL,t is the evolving virtual net load of the system during period t; P L,t is the total load of the system during period t; β 1,t is the guaranteed purchase coefficient of wind and solar power in period t; β 2,t is the hydropower security acquisition coefficient for period t; and are the maximum power generation that wind power, photovoltaic power and hydropower can achieve during period t; N h 、N g are the total number of hydropower units and thermal power units respectively; is the minimum technical output of thermal power unit i in period t.

4. The novel power system day-ahead optimal dispatching method based on evolving virtual net load according to claim 1 is characterized by: The objective function of the day-ahead optimization scheduling model based on the evolving virtual net load in step 3) is formula (2): minC=C g +C h +C w +C pv +C ps +C carbon (2) Where, C is the total system cost; C g is the operating cost of thermal power units; C h is the operating cost of the hydropower unit; C w is the wind farm operating cost; C pv is the operating cost of the photovoltaic power station; C ps is the start-up and shutdown cost of the pumped storage unit; C carbon The cost of carbon emissions.

5. The novel power system day-ahead optimal dispatching method based on evolving virtual net load according to claim 4 is characterized by: The operating cost of the thermal power unit is C g 、Operation cost of hydropower unit C h , Wind farm operating costs C w and the operating cost of the photovoltaic power station C pv , and the start-up and shutdown cost C of the pumped storage unit ps and carbon emission cost C carbon The calculation method is as follows: aThe operating cost of thermal power unit C g See formula (3), Where T is the total number of time periods in the scheduling cycle; a i , b i and c i are the cost coefficients of thermal power unit i respectively; P g,i,t is the power generation of thermal power unit i during period t; b. Operation cost of hydropower unit C h See formula (4), In the formula, k h P is the unit power generation cost coefficient of the hydropower unit; h,i,t is the power generation of hydropower unit i during period t; c Wind farm operation cost w See formula (5), C w =k w P w,t (5) In the formula, k w is the power generation cost coefficient of the wind farm; P w,t is the power generation of the wind farm during period t; d Photovoltaic power station operation cost C pv See formula (6), C pv =k pv P pv,t (6) In the formula, k pv is the power generation cost coefficient of photovoltaic power generation; P pv,t is the power generation of the photovoltaic power station during period t; eStart-up and shutdown cost of pumped storage unit C ps See formula (7), Where N ps is the total number of pumped storage units; They are the power generation and pumping state variables of pumped storage unit i in period t, which are 1 when working and 0 when stopped; are the power generation and pumping costs of pumped storage unit i, respectively; fCarbon emission cost C carbon See formula (8), In the formula, C carbon is the carbon emission cost; η is the carbon trading price; E P is the total carbon emissions during the scheduling period; E L is the total carbon emission quota in a scheduling cycle; δ i is the carbon emission of thermal power unit i; ε is the unit electricity quota coefficient of system carbon emission.

6. The novel power system day-ahead optimal dispatching method based on evolving virtual net load according to claim 1 is characterized by: The constraints of the day-ahead optimization scheduling model based on the evolving virtual net load in step 3) include: a The operation constraints of thermal power units are shown in formula (9). In the formula, are the upper and lower limits of the output of thermal power unit i; P g,i,t-1 is the dispatch output of thermal power unit i during period t-1; are the upper and lower limits of the ramp rate of thermal power unit i respectively; b The operating constraints of the hydropower unit are shown in formula (10), In the formula, the hydropower security acquisition coefficient β2 is 0.

4. is the power corresponding to the market transaction electricity of hydropower unit i in period t; are the upper and lower limits of the output of hydropower unit i; P h,i,t-1 is the dispatching output of hydropower unit i during period t-1; are the upper and lower limits of the climbing slope of hydropower unit i respectively; c Wind farm constraints are shown in formula (11), In the formula, the non-hydropower guaranteed acquisition coefficient β1 is 0.

3. is the power corresponding to the market transaction electricity of the wind farm during period t; d The constraints of the photovoltaic power station are shown in formula (12), In the formula, the non-hydropower guaranteed acquisition coefficient β1 is 0.

3. The power corresponding to the market transaction electricity of the photovoltaic power station during period t; The operating constraints of the pumped storage unit are shown in formula (13). Where P ps,i,t , P pm,i,t are the power generation and pumping power of pumped storage unit i during period t, respectively; are the upper limits of power generation and pumping power of pumped storage unit i, respectively; are the capacities of the upper and lower reservoirs in period t, respectively; are the upper and lower reservoir capacity limits, respectively; are the lower limits of the upper and lower reservoir capacities, respectively; is the upper reservoir capacity during period t-1; are the water consumption and pumping capacity of pumped storage unit i during period t; η gen , η pm are the hydropower conversion coefficients for power generation and pumping of pumped storage unit i, respectively; The power balance constraint f is shown in formula (14), In the formula, is the peak load of thermal power during period t; gFlexibility constraints include flexibility demand constraints (see formula (15), flexibility supply constraints (see formula (16)-(17) and flexibility margin constraints (see formula (18). In the formula, △P VL,t is the evolution virtual net load change between the next period and period t; F t de,up 、F t de,dn are the upward and downward flexibility requirements of the system in period t respectively; P w,t+1 , P pv,t+1 , P L，t+1 Forecast output for wind power, photovoltaic power and load in period t+1; ω u and ω d , γ u and γ d , u and λ d are the upward and downward flexibility demand coefficients of wind power, photovoltaic power, and load caused by the forecast error; λ e The upward flexibility demand coefficient for unit failure maintenance; In the formula, F t su,up The upward flexibility supply for the system during period t; It is the upward flexibility supply for thermal power, hydropower and pumped storage units during period t; In the formula, F t su,dn The downward flexibility supply for the system during period t; It is the downward flexibility supply for thermal power, hydropower and pumped storage units during period t; In the formula, F t up 、F t dn They respectively represent the upward and downward flexibility margins that the system has during period t.

Images