Method for optimal power selling plan of new energy station based on conditional value at risk

Abstract

The application discloses a new energy station optimal power selling plan method based on a conditional value at risk, establishes a new energy station optimal power selling model, and then uses simulation software to solve the model. The application can effectively guide the plan power of the new energy station in the inter-provincial power spot transaction in the day-ahead-day-in stage, and improve the enthusiasm of the new energy station in participating in cross-regional power transmission.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of inter-provincial power spot trading, and particularly relates to a new energy station optimal power selling plan method based on conditional value at risk. BACKGROUND

[0002] In recent years, new energy in China has developed rapidly. As of the end of 2021, the annual power generation of new energy in China has exceeded 1 trillion kilowatt-hours. However, the wind and solar energy resources in China are mainly concentrated in the three northern regions, which are inversely distributed with the load in the central and eastern regions, resulting in different degrees of wind / photovoltaic curtailment in new energy bases. In order to solve the problem of new energy consumption, China has introduced various energy policies, and marketization is one of the important ways. In the future, China will gradually form a provincial-inter-provincial two-level power market mode of 'unified market and two-level operation'.

[0003] The introduction of the two-level power market operation mode will make the coupling between inter-provincial and intra-provincial power trading closer. According to the operation idea of the current inter-provincial-intra-provincial two-level power market mode, the inter-provincial power trading result will be used as the boundary condition for the intra-provincial power trading of each province, so the inter-provincial power trading will have a certain influence on the intra-provincial trading. On the one hand, the power of new energy stations in inter-provincial trading will inevitably lower the trading space of the intra-provincial power market; on the other hand, the uncertainty of the new energy output in the province makes there be a certain deviation between the inter-provincial power plan and the actual trading power in the day-ahead and intraday, and the deviation part will bring certain market risk to the new energy station income. Considering that the intra-provincial and inter-provincial market trading has different market modes and trading rules, how does the sending end regional dispatching department comprehensively decide the inter-provincial power spot trading power for the new energy stations under its jurisdiction to improve the utilization rate of new energy and promote the optimal allocation of energy resources in a large range, while appropriately avoiding the risk brought by the uncertainty of new energy and load, is a problem to be solved at present. SUMMARY

[0004] The purpose of the application is to provide a new energy station optimal power selling plan method based on conditional value at risk, which can guide the new energy station to participate in the inter-provincial power spot trading plan in the day-ahead-intraday stage and improve the enthusiasm of the new energy station in participating in the cross-regional power transmission.

[0005] The technical scheme adopted by the application is: a new energy station optimal power selling plan method based on conditional value at risk, comprising the following steps:

[0006] Step 1, predicting the next day's power curve of the new energy station in the sending end province, obtaining the power generation bidding information of other power generation enterprises in the sending end province, the load demand curve and the system peak regulation demand information, and the inter-provincial tie line information;

[0007] Step 2, constructing an upper layer model, taking the power generation bidding information of other power generation enterprises in the sending province, the load demand curve and the system peak shaving demand information obtained in step 1 as input parameters of the upper layer model, taking the minimum total coal consumption cost of the sending province as an optimization target, taking the planned output value of each time thermal power unit as a decision variable, considering the system power balance constraint, the sending province peak shaving demand balance constraint, the peak shaving capacity constraint, the thermal power unit output constraint, the thermal power unit climbing rate constraint, and solving the planned output value optimization result of the thermal power unit at each time;

[0008] Step 3, constructing a lower layer model, taking the inter-provincial tie line information obtained in step 1 as an input parameter, and according to the planned output value optimization result of the thermal power unit at each time obtained in step 2, taking the maximum comprehensive utility of the new energy station as a target, taking the planned power of the new energy station participating in the inter-provincial power spot transaction as a decision variable, considering the new energy output value constraint, the new energy station participating in the inter-provincial transaction capacity constraint, the inter-provincial tie line capacity constraint and the conditional value at risk related constraint;

[0009] Step 4, solving the upper layer model obtained in step 2 and the lower layer model obtained in step 3 through a simulation software to obtain the optimal planned power of each new energy station participating in the inter-provincial power spot transaction.

[0010] The application also has the characteristics that,

[0011] The objective function established in step 2 with the optimization target of the sending province participating party minimizing the total coal consumption cost is:

[0012]

[0013]

[0014] In formula (1), C coal,t is the total coal consumption cost of the thermal power unit in the sending province at time t; F i is the consumption characteristic function of each thermal power unit; and p coal is the coal price. i In formula (2), F Thermal,i,t (P Thermal,i,t ) is the coal consumption of the thermal power unit i at time t when the unit runs at the planned output value P i , b i , and c i are the consumption characteristic parameters of the unit i.

[0015] The system power balance constraint in step 2 is:

[0016]

[0017] In formula (3), P new,t n,GPn(t) is the power of the new energy station n participating in the market transaction of the sending-end province at time t t Etc Pi(t) is the output value of the unit other than the new energy and thermal power unit in the sending-end province at time t t load Pd(t) is the load demand of the sending-end province at time t

[0018] The peak shaving demand balance constraint of the sending-end province is:

[0019]

[0020] In formula (4), P i,l,t Deep Ui(t) is the l-th step peak shaving power of the thermal power unit i at time t t Psys(t) is the system peak shaving demand at time t

[0021] The peak shaving capacity constraint is:

[0022] The total peak shaving power of the thermal power unit at each time should be less than the maximum peak shaving capacity, that is,

[0023]

[0024] In formula (5), P i,t,max Deep Umax(i) is the maximum deep peak shaving capacity of the thermal power unit i at time t

[0025] The output constraint of the thermal power unit is:

[0026]

[0027] In formula (6), P i,t min Pmin(i) is the minimum output value of the unit i allowed to run at time t i,t max Pmax(i) is the maximum output value of the unit i allowed to run at time t

[0028] The ramp rate constraint of the thermal power unit is:

[0029]

[0030] In formula (7), P i up Pup(i) is the maximum upward ramp rate of the thermal power unit i in the system i down Pdown(i) is the maximum downward ramp rate of the thermal power unit i in the system; T is the length of the statistical period.

[0031] The objective function established in step 3 with the maximum comprehensive utility of the new energy station as the target is:

[0032]

[0033]

[0034]

[0035] In formula (8), U new n is the comprehensive utility; O new n is the income utility; R new n is the risk utility; in formula (9), I new n is the expected income of the provincial market of the sending province; G new n is the expected income of the inter-provincial market; in formula (10), δ CVaR is the conditional risk value; the risk preference coefficient β is the degree of preference of the new energy station to the total income risk under the two-level market.

[0036] The expected income I of the provincial market of the sending province of the new energy station new n is represented as:

[0037]

[0038]

[0039]

[0040] In formula (11), κ w is the probability of scenario w; E new,t n is the income of the provincial market of the sending province of the new energy station at time t; C new,t n,Deep is the peak shaving auxiliary service cost that needs to be allocated by the new energy station n at time t, that is, the real-time deep regulation cost prediction information; γ t n is the on-grid price of the new energy station n in the provincial market of the sending province at time t; λ t n is the total peak shaving cost proportion coefficient of the new energy station n at time t; the planned output value of each thermal power unit at each time is obtained by the upper model output; ρ l,t is the actual clearing price of the thermal power unit i at time t; ζ is the peak shaving cost correction coefficient; in formula (12) and formula (13), C i is the capacity of the thermal power unit i;

[0041] The electricity P of the new energy station n participating in the provincial market transaction of the sending province new,t n,G and the total peak shaving cost proportion coefficient λt n are respectively represented as follows:

[0042]

[0043]

[0044]

[0045] In formula (14), μ t n is the participation index of the new energy station n in the inter-provincial spot market at time t; P new,t n is the total output value of the new energy station n at time t; in formula (15), P new,i,t M , P thermal,t M , P wind,t M , P pv,t M and P nuclear,t M respectively correspond to the corrected electricity of the new energy station, the thermal power unit, the wind power unit, the photovoltaic power station and the nuclear power station whose load rate is higher than the compensated peak regulation benchmark at time t, and the real-time deep peak regulation compensated service compensation fee is shared by the thermal power plant, the wind power plant, the photovoltaic power station and the nuclear power plant in the sending end province whose load rate is higher than the compensated peak regulation benchmark; in formula (16), P Thermal,k,i,t is the actual power generation of the thermal power unit i at the kth level at time t; β k is the power correction coefficient of the thermal power plant; P wind,i,t , P pv,i,t , P nuclear,i,t are respectively the actual on-grid power of the wind power unit, the photovoltaic power station and the nuclear power station at time t; d is the correction coefficient; the values of p, q and z are related to whether the wind power plant and the photovoltaic power station are subsidized and the utilization hours of the last year.

[0046] The expected income G new,t n in the inter-provincial market is represented as:

[0047]

[0048] In formula (17), π t n is the marginal clearing price of the cross-regional spot system at time t; P new,t n,IP is the optimization result of the inter-provincial power spot transaction of the new energy unit n at time t.

[0049] The new energy output value constraint in step 3 is:

[0050]

[0051] In formula (18), P new,t n is the output value of new energy unit n at time t, P new,t n,max is the maximum output value of new energy unit n at time t;

[0052] The capacity constraint of the new energy station participating in inter-provincial trading is:

[0053]

[0054] In formula (19), P new,t n,IP,max , P new,t n,IP,min are the upper and lower limits of the capacity of new energy unit n participating in inter-provincial power spot trading;

[0055] The capacity constraint of the inter-provincial tie-line is:

[0056]

[0057] In formula (20), P f,min and P f,max are the maximum and minimum stable limits that the section f can withstand, respectively;

[0058] The conditional value-at-risk related constraint is:

[0059]

[0060]

[0061] m w ≥ 0 (23)

[0062]

[0063] In formula (21), the optimization result of variable δ VaR is defined as the VaR value of the total expected income of the intra-provincial-inter-provincial market; ε is the confidence level; m w is a non-negative auxiliary variable introduced, which is the amount of the total income risk of the intra-provincial-inter-provincial market exceeding δ VaR in each prediction scenario.

[0064] In step 4, the YALMIP toolbox is used for modeling, and the CPLEX commercial software is called for solving the upper model and the lower model, to obtain the inter-provincial power spot trading optimization result P new,t n,IP of new energy unit n at time t.

[0065] The beneficial effects of the present application are: the new energy station optimal power selling plan method based on conditional risk value can guide the planned power of new energy station in inter-provincial power spot trading in day-ahead-intra-day stage, and improve the enthusiasm of new energy station in cross-regional power transmission through constructing new energy station optimal power selling model and solving. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 is a flowchart of the new energy station optimal power selling plan method based on conditional risk value of the present application;

[0067] Figure 2 is a typical day-ahead type of unit output and load demand prediction curve chart of the present application embodiment;

[0068] Figure 3 is a thermal power unit planned output and deep regulation cost prediction information chart of the present application embodiment;

[0069] Figure 4 is a new energy inter-provincial and intra-provincial trading power optimization result chart of the present application embodiment;

[0070] Figure 5 is a power optimization result chart of each scene under different risk preferences of the present application embodiment;

[0071] Figure 6 is a market income chart of each scene under different risk preferences of the present application embodiment;

[0072] Figure 7 is a cross-regional spot market participation optimization result chart of new energy station in typical day A and B of the present application embodiment. DETAILED DESCRIPTION

[0073] The present application will be described in detail below in combination with the drawings and specific embodiments.

[0074] The present application provides a new energy station optimal power selling plan method based on conditional risk value, according to the transaction status of new energy station in inter-provincial and intra-provincial two-level market, considering the uncertainty of new energy output and load power, such as Figure 1 as shown, a new energy station optimal power selling model based on conditional risk value is constructed, and the optimal decision of new energy station power selling plan is realized through solving the model, the risk brought by uncertain factors is effectively controlled, the power selling enthusiasm of new energy station is improved, which is conducive to playing the role of inter-provincial mutual aid of power transmission channel, and promoting new energy consumption and energy saving and emission reduction. The specific implementation is as follows:

[0075] Step 1, predict the next day's new energy station power curve in the sending end province, obtain the power generation bidding information of other power generation enterprises in the sending end province, the load demand curve and system peak regulation demand information, and the inter-provincial tie line information;

[0076] In step 1, the next day's new energy station power curve and load demand curve are obtained by forecasting the new energy station power curve and load demand curve in the previous day. Other power generation enterprises report the power generation bidding information participating in the market, and various types of power wholesale users and power selling companies in the province report the power load demand information to the provincial dispatch center to obtain the inter-provincial tie line information.

[0077] In step 2, the upper model takes the minimum total coal consumption cost of the provincial participants as the optimization objective, considers the system power balance constraint, the system peak regulation constraint and the thermal power unit operation constraint, determines the planned output value of each thermal power unit at each time and transmits it to the lower layer. Specifically:

[0078] The objective function established by the upper model taking the minimum total coal consumption cost of the provincial participants as the optimization objective is:

[0079]

[0080]

[0081] Where, C coal,t is the total coal consumption cost of the provincial thermal power units at t time; F i is the consumption characteristic function of each thermal power unit; p coal is the coal price. F i (P Thermal,i,t ) is the coal consumption of the thermal power unit i at t time when it operates at the planned output value P Thermal,i,t ; a i , b i , c i are the consumption characteristic parameters of the unit i.

[0082] The constraint conditions of the upper model include the system power balance constraint, the provincial peak regulation demand balance constraint, the peak regulation capacity constraint, the thermal power unit output constraint, and the thermal power unit climbing rate constraint.

[0083] The system power balance constraint is:

[0084]

[0085] Where, P new,t n,G is the power of the new energy station n participating in the market transaction in the sending province at t time, P t Etc is the output value of other units in the province except new energy and thermal power units at t time; P t load is the load demand in the province at t time.

[0086] The provincial peak regulation demand balance constraint is:

[0087]

[0088] Ptiis the lth step of the i th thermal power unit at time t, U i,l,t Deep Ptiis the lth step of the i th thermal power unit at time t, U t Ptiis the lth step of the i th thermal power unit at time t, U

[0089] The peak regulation capacity constraint is:

[0090] The total peak regulation capacity of the thermal power unit at each time should be less than the maximum peak regulation capacity, i.e.

[0091]

[0092] Ptiis the lth step of the i th thermal power unit at time t, U i,t,max Deep Ptiis the maximum deep peak regulation capacity of the i th thermal power unit at time t;

[0093] The thermal power unit output constraint is:

[0094]

[0095] Ptiis the lth step of the i th thermal power unit at time t, U i,t min Ptiis the minimum output value of the i th thermal power unit allowed to run at time t; P i,t max Ptiis the maximum output value of the i th thermal power unit allowed to run at time t.

[0096] The thermal power unit climbing rate constraint is

[0097]

[0098] Ptiis the maximum upward climbing rate of the i th thermal power unit in the system; P i up Ptiis the maximum upward climbing rate of the i th thermal power unit in the system; P i down Ptiis the maximum downward climbing rate of the i th thermal power unit in the system; T is the statistical period length, here T = 15 min.

[0099] Step 3, the lower layer model takes the maximum comprehensive utility of the new energy station as the target, considers the new energy output constraint, the new energy participation in inter-provincial transaction capacity constraint, the inter-provincial tie line capacity constraint and the conditional risk value related constraint. At the same time, the planned output value of the thermal power unit solved by the upper layer model is taken as the input value of the lower layer model to solve the lower layer model.

[0100] The objective function of the lower layer model taking the maximum comprehensive utility of the new energy station as the target is:

[0101]

[0102]

[0103]

[0104] wherein, U new n is the comprehensive utility; O new n is the income utility; R new n is the risk utility; I new n is the expected income of the intraprovincial market; G new n is the expected income of the interprovincial market; δ CVaR is the conditional risk value. The risk preference coefficient β is the degree to which the new energy station is inclined to the total income risk under the two-level market, and its value range is β≥0. When β=0, the new energy station operator is risk-neutral, that is, the operator does not care about the income loss risk caused by the uncertainty factor, neither avoiding risk nor actively pursuing risk; when β>0, the new energy station operator is risk-averse, that is, a more conservative operation strategy is adopted to avoid income risk, and the greater the value of β, the higher the degree of risk aversion, and the more conservative the strategy adopted.

[0105] The expected income of the intraprovincial market I new n is represented as:

[0106]

[0107]

[0108]

[0109] wherein, κ w is the probability of scenario w; E new,t n is the intraprovincial market electricity income of the new energy station at time t; C new,t n,Deep is the peak shaving auxiliary service cost that the new energy station n needs to share at time t, that is, the real-time deep regulation cost prediction information; γt n is the intraprovincial on-grid price of the new energy station n at time t; λt n is the total peak regulation cost proportion coefficient of the new energy station n at time t; the planned output value of each time of the thermal power unit output by the upper model is obtained; ρ l,t is the actual clearing price of the thermal power unit i at time t; ζ is a peak regulation cost correction coefficient, which is 1 in the heating period and 0.5 in the non-heating period; C i is the capacity of the thermal power unit i.

[0110] The electricity P traded by the new energy station n in the intraprovincial marketnew,t n,G Coefficient λ of total peak-shaving cost t n They are represented as follows:

[0111]

[0112]

[0113]

[0114] Where, μt n % represents the participation rate of new energy power stations in cross-regional inter-provincial spot trading at time t; P new,t n Let P be the total output of the renewable energy power station n at time t. new,i,t M P thermal,t M P wind,t M P pv,t M and P nuclear,t M These correspond to the corrected electricity volumes for renewable energy plants, thermal power units, wind power units, photovoltaic power plants, and nuclear power plants at time t whose load factors exceed the paid peak-shaving benchmark. The compensation cost for real-time deep peak-shaving ancillary services is shared by the thermal power plants, wind farms, photovoltaic power plants, and nuclear power plants within the province whose load factors exceed the paid peak-shaving benchmark. P Thermal,k,i,t β represents the actual power generation of thermal power unit i at time t in step k; k P is the power generation correction factor for thermal power plants; 1 for load factor less than 70%, 1.5 for load factor between 70% and 80%, and 2 for load factor greater than 80%. wind,i,t P pv,i,t P nuclear,i,t The values ​​represent the actual grid-connected electricity of wind turbines, photovoltaic power plants, and nuclear power plants at time t, respectively; the correction coefficient d is 2 during the heating season and 1 during the non-heating season; the values ​​of p, q, and z are related to whether the wind farm and photovoltaic power plant receive subsidies and the number of utilization hours in the previous year, and the values ​​range from 0 to 1.

[0115] Expected returns in inter-provincial markets G new,t n Represented as:

[0116]

[0117] Where, π t n Let P be the marginal clearing electricity price of the cross-regional spot system at time t. new,t n,IP The optimization results for inter-provincial electricity spot trading for new energy generating unit n at time t.

[0118] The constraints of the lower-level model include constraints on the output value of new energy sources, the capacity constraints of new energy power stations participating in inter-provincial transactions, the capacity constraints of inter-provincial connecting lines, and constraints related to conditional value at risk.

[0119] The power output constraint for new energy sources is:

[0120]

[0121] Among them, P new,t n is the output value of the new energy unit n at time t, P new,t n,max Let n be the maximum output value of the new energy unit n at time t.

[0122] The capacity constraints for new energy power plants participating in inter-provincial transactions are as follows:

[0123]

[0124] Among them, P new,t n,IP,max P new,t n,IP,min This refers to the upper and lower limits of the capacity of new energy generating units (n) participating in inter-provincial electricity spot trading.

[0125] The capacity constraint for inter-provincial connecting lines is:

[0126]

[0127] Among them, P f,min and P f,max These are the maximum and minimum stability limits that section f can withstand, respectively.

[0128] The constraints related to conditional value at risk are:

[0129]

[0130]

[0131] m w ≥0 (23)

[0132]

[0133] Wherein, variable δ VaR The optimization result is defined as the VaR value of the total expected return in the intra-provincial and inter-provincial markets; ε is the confidence level; m w The non-negative auxiliary variable introduced is used to determine whether the risk of total market return exceeding δ in each prediction scenario is within the province or between provinces. VaR The amount.

[0134] Step 4, the YALMIP toolbox is used for modeling, and the CPLEX commercial software is called for model solving, and the optimal planning power of each new energy station participating in the inter-provincial power spot transaction is obtained.

[0135] Embodiment

[0136] The day-ahead 15-minute interval new energy output prediction curve, the thermal power planning output curve and the load demand prediction curve of a certain place on a typical day are shown in Figure 2 .

[0137] It is assumed that the prediction error of new energy output and load demand satisfies the normal distribution with mean value of 0 and variance of 0.2, a large number of scenarios are formed by Monte Carlo, and then the scenario reduction technology is used to reduce them to 10 scenarios, and the typical scenarios and their probability values are shown in Table 1.

[0138] Table 1 Typical prediction scenarios and probabilities

[0139]

[0140] As can be seen from Table 1, the probability of scenario 4 is the largest. The power curve of the in-province thermal power unit and the peak shaving cost under scenario 4 are shown in Figure 3 , and the results of the new energy participating in the in-province market transaction and the inter-province market optimization corresponding thereto are shown in Figure 4 .

[0141] As can be seen from Figure 3 : the size of the peak shaving cost reflects the degree of the peak shaving auxiliary service provided by the thermal power unit: when the total output of the thermal power unit increases, the peak shaving cost generally shows a downward trend; and when the total output of the thermal power unit decreases, the peak shaving cost generally shows an upward trend. The peak shaving cost of the new energy station is mainly determined by the size of the load and the new energy output capacity: in the period of low load, the load value is small and the new energy output capacity is often large, so in order to fully accommodate the new energy, the thermal power unit needs to provide peak shaving auxiliary service for the new energy unit. In addition, the deep peak shaving cost is zero in the period of 68-76, and the thermal power unit does not provide peak shaving auxiliary service, because this period is in the peak load period and the new energy output capacity is low, as a adjustable power source, the thermal power unit must increase the output value in order to maintain the system power balance, when the output value of all thermal power units is higher than the paid peak shaving reference, the moment of peak shaving cost of 0 appears.

[0142] From Figure 4The data shows that the optimal outcome for new energy participation in the inter-regional spot market generally occurs when peak-shaving costs are high and new energy generation capacity is high. When thermal power units face difficulties in peak shaving, new energy power plants within the province transmit this portion of their power across regions. When the penetration rate of new energy is high, the system utilizes the peak-shaving capacity of thermal power units to reduce their output, resulting in a significant increase in peak-shaving costs. This indicates that more peak-shaving costs are required to mitigate the impact of the anti-peak-shaving characteristics of new energy. If all thermal power units are reduced to their minimum economic output, and the peak-shaving demand of renewable energy is still insufficient, wind and solar curtailment will occur. The optimal outcome occurs when thermal power output is low, i.e., when the grid has to reduce renewable energy output to ensure safe operation. New energy power plants rationally allocate a portion of their power to participate in the inter-provincial electricity spot market. This reduces the deep peak-shaving costs of thermal power units, reflecting the market's role in optimizing resource allocation, and also reduces the probability of wind and solar curtailment, improving renewable energy consumption. Intra-provincial new energy power plants participate in the inter-provincial electricity spot market during the 6-12, 15-16, and 46-53 time periods. The periods during which wind farms participate in the inter-provincial market are generally continuous. The discontinuity in the 13-14 period is due to a sudden drop in renewable energy generation capacity and deep-shifting costs during this period, making participation in the intra-provincial market more beneficial. Furthermore, although wind power generation capacity is not significantly different between the 2-4 and 46-53 periods, wind farms do not participate in the inter-provincial electricity spot market during these periods. This demonstrates that a large overall wind power generation capacity is not the sole factor in determining whether to participate in the inter-provincial market. Instead, a comprehensive consideration of load demand and the operational flexibility of thermal power units should be taken into account to rationally allocate electricity for participation in both intra- and inter-provincial markets, maximizing the benefits for renewable energy plants.

[0143] Figures 5-6 The market returns and power optimization results for each scenario are presented under different risk preference coefficients.

[0144] from Figure 5 The results show that as risk aversion increases, market returns in risky scenarios increase significantly, while those in non-risky scenarios decrease. Specifically, the difference in market returns between scenario 7 (maximum return) and scenario 1 (minimum return) increases from 343,000 yuan when β=0 to 157,000 yuan when β=1 and 29,000 yuan when β=30. This demonstrates that the conditional value at risk approach effectively improves the ability to assess the impact of uncertainty in the control of renewable energy power plants on their electricity sales decisions.

[0145] from Figure 6As can be seen, when the risk appetite coefficient β increases from 0 to 30, the level of control over uncertainty by renewable energy power plants continuously improves, and the traded electricity in the inter-provincial electricity spot market also increases. Between times 4-20 and 44-58, the inter-provincial traded electricity volume increases with the increase in the risk aversion level of inter-provincial traders. Taking time 11 as an example, the inter-provincial traded electricity volume increases from 602MW when β=0 to 675MW when β=1, and then to 690MW when β=30. This demonstrates that the risk appetite coefficient has a significant impact on the electricity sales decisions of renewable energy power plants and affects the planned electricity in the inter-provincial market.

[0146] A specific analysis is conducted using data from typical days A and B in a certain location. The capacities of wind farms I and II, and the photovoltaic power station are 99 MW, 200 MW, and 22 MW, respectively. The stochastic optimization method is compared with this method, assuming that the constraints of the stochastic optimization method and this method are the same.

[0147] Figure 7 The graph shows the optimized results of cross-regional spot market participation of new energy power plants on typical days A and B.

[0148] from Figure 7 The analysis shows that, from a single-day perspective, wind farms I and II, and photovoltaic power plants, choose roughly the same times for power transmission under both methods: on typical day A, power transmission is concentrated between 6:00-16:00 and 47:00-53:00; on typical day B, power transmission is concentrated between 11:00-23:00 and 95:00-96:00. However, the participation decision results of the two methods differ in both timing and magnitude, becoming the fundamental reason for the difference in profitability.

[0149] The revenue of a new energy power station on a typical day is shown in Table 2.

[0150] Table 2. Revenue of New Energy Power Stations on a Typical Day

[0151]

[0152] Compared to the stochastic optimization method, although the photovoltaic power plant's revenue on a typical day (A) is slightly lower under the method of this invention, it is significantly higher under other conditions. Overall, the two-tiered electricity market operation mode is more conducive to the revenue of power plants and the development of clean energy in the era without subsidies. The stochastic optimization method, because it relies entirely on probabilistic prediction to judge random scenarios, results in highly volatile revenue, making it difficult to apply with current prediction capabilities. This invention, by adjusting the risk preference coefficient of power plant staff, fully leverages the coupling characteristics of the electricity market across multiple spatial scales and the existing level of electricity prediction, effectively promoting the consumption of clean energy.

Claims

1. A method for optimal electricity sales planning at renewable energy power plants based on conditional value-at-risk, characterized in that, Includes the following steps: Step 1: Predict the power curve of new energy power plants in the sending province for the next day, and obtain the power generation bidding information, load demand curve and system peak-shaving demand information of other power generation companies in the sending province, as well as inter-provincial interconnection line information. Step 2: Construct the upper-level model. Use the power generation bidding information of other power generation enterprises in the sending-end province, the load demand curve, and the system peak-shaving demand information obtained in Step 1 as the input parameters of the upper-level model. Take minimizing the total coal consumption cost of the participants in the sending-end province as the optimization objective, and take the planned output value of thermal power units at each time as the decision variable. Consider the system power balance constraints, peak-shaving demand balance constraints in the sending-end province, peak-shaving capacity constraints, thermal power unit output constraints, and thermal power unit ramp rate constraints to solve for the optimization results of the planned output value of thermal power units at each time. Step 3: Construct the lower-level model. Using the inter-provincial tie-line information obtained in Step 1 as input parameters, and based on the optimization results of the planned output values ​​of thermal power units at each time point obtained in Step 2, the objective is to maximize the comprehensive utility of new energy power plants. The decision variable is the planned power of each new energy power plant participating in inter-provincial electricity spot trading. Constraints are considered on new energy output value, capacity constraints for new energy power plants participating in inter-provincial trading, inter-provincial tie-line capacity constraints, and conditional risk value-related constraints. The objective function established with the goal of maximizing the comprehensive utility of new energy power plants is: （8） （9） （10） In equation (8), U new n For overall utility; O new n For the benefit of utility; R new n For risk utility; in equation (9), I new n Expected revenue from the provincial market; G new n For the expected returns in the inter-provincial market; in formula (10), δ CVaR Conditional risk value; Risk preference coefficient β This indicates the degree of preference for total return and risk in both the primary and secondary markets for new energy power plants. The constraints related to conditional value at risk are: （21） （22） （23） （24） In equation (21), the variable δ VaR The optimization result is defined as the VaR value of the total expected return in the intra-provincial and inter-provincial markets; ε Confidence level; κ w For the scene w The probability of; m w The non-negative auxiliary variable introduced is used to determine whether the risk of total market return exceeding the threshold in each prediction scenario is within the province or between provinces. δ VaR The amount; γ t n for t Shike New Energy Power Station n On-grid electricity price within the sending province; P new,t n,IP For new energy units n exist t Optimization results of inter-provincial electricity spot trading at any time; π t n for t Marginal clearing electricity prices in the cross-regional spot market system; P new,t n,G for t Shike New Energy Power Station n Electricity volume participating in intra-provincial market transactions in the sending province; ρ l,t for t thermal power units i exist l Actual clearing electricity price; P i,l,t Deep for t thermal power units i No. l Peak-shaving power consumption; ζ This is a correction factor for peak shaving costs; Step 4: Solve the upper-level model obtained in Step 2 and the lower-level model obtained in Step 3 together using simulation software to obtain the optimal planned power for each new energy power station to participate in inter-provincial electricity spot trading.

2. The optimal electricity sales plan method for new energy power plants based on conditional value-of-risk as described in claim 1, characterized in that, The objective function established in step 2, with the goal of minimizing the total coal consumption cost among participants in the sending province, is as follows: （1） （2） In equation (1), C coal,t for t The total coal consumption cost of thermal power units within the province is sent at any time. F i For each thermal power unit, the consumption characteristic function is given. ρ coal The price of coal; in equation (2), F i ( P Thermal,i,t ) for thermal power units i exist t Always running at the planned output value P Thermal,i,t Coal consumption per hour; a i 、b i 、c i For the unit i Consumption characteristic parameters.

3. The optimal electricity sales plan method for new energy power plants based on conditional value at risk as described in claim 2, characterized in that, The system power balance constraint in step 2 is: （3） In equation (3), P new,t n,G for t Shike New Energy Power Station n Electricity traded in the intra-provincial market of the sending province. P t Etc for t The power output of all generating units in the province other than new energy and thermal power units is sent to the end point at all times. P t load for t The system is constantly sending data to the load demand within the province. The constraint for balancing peak-shaving demand within the sending province is: （4） In equation (4), P i,l,t Deep for t thermal power units i No. l Peak-shaving power consumption U t for t The system allocates peak-shaving demand at all times; Peak shaving capacity constraints are: The total peak-shaving power of thermal power units at any given time should be less than the maximum peak-shaving capacity, that is... （5） In equation (5), P i,t,max Deep for t thermal power units i Maximum depth peak shaving capability; The output constraints of thermal power units are: （6） In equation (6), P i,t min for t Time crew i The minimum output value allowed for operation; P i,t max for t Time crew i The maximum output value allowed to run; The ramp rate constraint for thermal power units is: （7） In equation (7), P i up For thermal power units within the system i Maximum uphill climbing rate; P i down For thermal power units within the system i The maximum downhill climbing rate; T is the duration of the statistical period.

4. The optimal electricity sales plan method for new energy power plants based on conditional value at risk as described in claim 3, characterized in that, The expected revenue in the province where the delivery point is located. I new n Represented as: （11） （12） （13） In equation (11), κ w For the scene w The probability of; E new,t n for t Revenue from electricity sold by new energy power plants within the province; C new,t n,Deep for t Shike New Energy Power Station n The cost of peak shaving ancillary services that need to be shared, i.e., real-time deep shaving cost forecast information; γ t n for t Shike New Energy Power Station n On-grid electricity price within the sending province; λ t n for t Shike New Energy Power Station n Total peak-shaving cost ratio coefficient; obtained from the planned output value of thermal power units at each time point output by the upper-level model; ρ l,t for t thermal power units i exist l Actual clearing electricity price; ζ For peak shaving cost correction coefficients; in equations (12) and (13) C i For thermal power units i The capacity; New energy power stations n Electricity traded in the provincial market of the sending province P new,t n,G Coefficient of proportion of total peak shaving cost λ t n They are represented as follows: （14） （15） （16） In equation (14), μ t n % for t Shike New Energy Power Station n Indicators of participation in cross-regional inter-provincial spot trading; P new,t n for t Shike New Energy Power Station n Total output value; in equation (15), P new,i,t M , P thermal,t M , P wind,t M , P pv,t M and P nuclear,t M Corresponding to t For new energy power plants, thermal power units, wind power units, photovoltaic power plants, and nuclear power plants whose load factor is higher than the paid peak shaving benchmark, the corrected electricity volume shall be shared by the paid ancillary service compensation fee for real-time deep peak shaving, which shall be shared by the thermal power plants, wind farms, photovoltaic power plants, and nuclear power plants in the sending province whose load factor is higher than the paid peak shaving benchmark; in formula (16), P Thermal,k,i,t for t thermal power units i exist k Actual power generation; β k This is a correction factor for the power generation of thermal power plants; P wind,i,t , P pv,i,t , P nuclear,i,t They are respectively t The actual electricity generated by wind turbines, photovoltaic power plants, and nuclear power plants at any given time; d This is a correction factor; p , q and z The value is related to whether the wind farm and photovoltaic power station receive subsidies and the number of utilization hours in the previous year.

5. The optimal electricity sales plan method for new energy power plants based on conditional value-at-risk as described in claim 4, characterized in that, Expected returns in the inter-provincial market G new,t n Represented as: （17） In equation (17), π t n for t Marginal clearing electricity prices in the cross-regional spot market system; P new,t n,IP For new energy units n exist t Optimization results of inter-provincial electricity spot trading.

6. The optimal electricity sales plan method for new energy power plants based on conditional value at risk as described in claim 5, characterized in that, The constraint on the new energy output value in step 3 is as follows: （18） In equation (18), P new,t n for t Shike New Energy Unit n The output value, P new,t n,max for t Shike New Energy Unit n Maximum output value; The capacity constraints for new energy power plants participating in inter-provincial transactions are as follows: （19） In equation (19), P new,t n,IP,max , P new,t n,IP,min For new energy units n Upper and lower limits on the capacity for participation in inter-provincial electricity spot trading; The capacity constraint for inter-provincial connecting lines is: （20） In equation (20), P f,min and P f,max Cross-sections f The maximum and minimum stability limits that can be tolerated.

7. The optimal electricity sales plan method for new energy power plants based on conditional value-of-risk as described in claim 6, characterized in that, In step 4, the YALMIP toolkit is used for modeling, and the commercial software CPLEX is called to solve the upper and lower level models to obtain the new energy unit. n exist t Optimization results of inter-provincial electricity spot trading P new,t n,IP .

Images