A method for calling reserve capacity to cope with the downward ramp of wind power

Abstract

The present invention provides a method for calling spare capacity to cope with the downward ramping of wind power, belonging to the technical field of wind power dispatching. The method includes: obtaining wind power output sample data, and calculating the downward ramping power data of wind power by using the ramping power formula; calculating the marginal distribution functions of the wind power output sample data and the downward ramping power data of wind power, and constructing a joint distribution model by using the Copula function; calculating the conditional probability expression of the wind power ramping under the condition of the predicted wind power, and drawing the wind power output - downward ramping boundary curve of wind power; designing the ramping reserve service including the time scale; obtaining the predicted value of the wind power ramping reserve demand by using the boundary curve, constructing a ramping reserve model and obtaining the ramping reserve demand, and obtaining the optimal call of the spare capacity for the downward ramping of wind power. The present invention provides the call of the spare capacity for the downward ramping of wind power on a longer time scale, uses cold reserve resources to participate in power optimization dispatching, reduces wind curtailment, and improves economy.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power dispatching, and particularly relates to a method for calling reserve capacity to cope with the downward ramp of wind power. Background Art

[0002] The output of wind power has strong volatility. The output of wind power fluctuates little on the second - level time scale, and the possible fluctuations gradually increase as the time scale increases. The existing technologies generally focus on the relationship between the predicted output of wind power in conventional scenarios and the output fluctuations on a short - time scale, lacking the consideration of the downward ramp of wind power on a longer time scale. However, in the scenario of high penetration of new energy, the continuous downward ramp of wind power output on a longer time scale will cause a large amount of power deficit, and it is necessary to pay attention to the downward ramp of wind power on a longer time scale to alleviate the risk brought by the ramp event to the power balance of the power system.

[0003] In the scenario of high penetration of new energy, most conventional units are in cold reserve state, and the flexible regulation resources with a response speed within 1 hour are tense. However, the current power optimization dispatching models generally only consider calling flexible regulation resources with fast response speeds such as energy storage and thermal reserve, and rarely consider the participation of cold reserve resources. To alleviate the problem of tense reserve resources, a large amount of wind power curtailment needs to be arranged, which affects the economy. Summary of the Invention

[0004] Aiming at the above - mentioned deficiencies in the prior art, a method for calling reserve capacity to cope with the downward ramp of wind power provided by the present invention solves the problems that the prior art lacks the attention to the downward ramp of wind power on a longer time scale, the participation of cold reserve resources in the power optimization dispatching model is small, and a large amount of wind power curtailment is arranged to alleviate the tense reserve resources.

[0005] In order to achieve the above - mentioned purpose, the technical solution adopted by the present invention is: a method for calling reserve capacity to cope with the downward ramp of wind power, including the following steps:

[0006] S1. Obtain the sample data of wind power output, and calculate the downward ramp power data of wind power according to the sample data of wind power output;

[0007] S2. Calculate the marginal distribution functions of the sample data of wind power output and the downward ramp power data of wind power respectively, and use the Copula function to construct the joint distribution model of wind power output and the downward ramp of wind power;

[0008] S3. According to the joint distribution model of wind power output and the downward ramp of wind power, calculate the conditional probability equation of the wind power ramp under the condition of the predicted wind power, and input the historical data of wind power output and the downward ramp of wind power into the conditional probability equation of the wind power ramp to obtain the wind power output - wind power downward ramp boundary curve;

[0009] S4. Design the ramp reserve service including the time scale according to the response time of the cold - reserve gas units;

[0010] S5. Substitute the predicted wind power value into the wind power output - wind power down-ramping boundary curve to obtain the predicted value of the wind power ramping reserve demand, and combine the spinning reserve and the ramping reserve to construct a ramping reserve model; and use the ramping reserve model to obtain the ramping reserve demand, and use the ramping reserve service to obtain the optimal dispatch of the reserve capacity for wind power down-ramping.

[0011] The beneficial effects of the present invention are as follows: The present invention uses the mutual cooperation of the ramping reserve and the spinning reserve to cope with the down-ramping under wind power uncertainty, increases the wind power consumption in the high wind power penetration scenario, reduces the operation cost of the power system, and ensures the power balance of the power system; the present invention uses the Copula function to construct a joint distribution model, finds out the correlation between the wind power output and the wind power down-ramping power, and determines the ramping reserve demand based on the predicted wind power output value, which improves the economy of the present invention on the basis of ensuring the adequacy of power and electricity when combined with the ramping reserve model.

[0012] Further, the S2 includes the following steps:

[0013] S201. Calculate the marginal distribution functions of the wind power output sample data and the wind power down-ramping power data respectively;

[0014] S202. Determine the parameters of the Copula function using the maximum likelihood estimation according to the marginal distribution functions of the wind power output sample data and the wind power down-ramping power data;

[0015] S203. Evaluate and screen the Copula function using the Akaike information criterion according to the goodness of fit and the number of parameters of the Copula function;

[0016] S204. Use the screened Copula function to construct a joint distribution model of the wind power output and the wind power down-ramping.

[0017] The beneficial effects of the above further solution are as follows: The present invention determines the parameters of the Copula function using the maximum likelihood estimation and constructs the best joint distribution model of the wind power output and the wind power down-ramping in combination with the Akaike information criterion, which improves the fitting degree and accuracy of the ramping reserve model.

[0018] Still further, the S5 includes the following steps:

[0019] S501. Substitute the predicted wind power value into the wind power output - wind power down-ramping boundary curve and calculate to obtain the predicted value of the wind power ramping reserve demand;

[0020] S502. Use the spinning reserve and the ramping reserve to provide reserve capacity;

[0021] S503. Construct a ramping reserve model by using the predicted value of wind power ramping reserve demand, load fluctuations, spinning reserve capacity, and ramping reserve capacity, and reserve adjustment capacity by using spinning reserve constraints;

[0022] S504. Combine the predicted value of wind power ramping reserve demand, obtain the ramping reserve demand by using the ramping reserve model, and obtain the optimal dispatch of the reserve capacity for the downward ramping of wind power by using the constraints related to ramping reserve services and ancillary services including time scales.

[0023] Furthermore, the expression of the ramping reserve model is as follows:

[0024] ;

[0025] ;

[0026] ;

[0027] where represents t the ramping reserve demand with a time scale of x in the period, t represents x the predicted value of wind power ramping reserve demand with a time scale of in the period, represents t the load fluctuation in the period, t represents x the ramping reserve capacity with a time scale of in the period, represents t the spinning reserve capacity in the period, RMx represents the maximum dispatching period of the ramping service

[0028] The beneficial effects of the above further solution are as follows: The present invention uses ramping reserve and spinning reserve to provide reserve capacity to cope with the downward ramping power deficit of wind power, the uncertainty of wind power output, and load fluctuations. The spinning reserve copes with small new energy or load fluctuations, ensuring power balance on a short time scale. The ramping reserve copes with a large downward ramping power deficit of wind power, ensuring sufficient capacity of the power system on a long time scale.

[0029] Furthermore, the constraints related to ancillary services include: ramping reserve constraints, positive and negative spinning reserve constraints, and ramping reserve service duration constraints;

[0030] The expression of the climbing reserve constraint is as follows:

[0031] ;

[0032] ;

[0033] ;

[0034] where, represents the set of units providing climbing reserve service with a time scale of 1, represents the set of units providing climbing reserve service with a time scale of 3, represents the set of units providing climbing reserve service with a time scale of 8, represents the set of pumped-storage units, represents the set of conventional units, represents at t the state of the unit with a time scale of 1 participating in the climbing reserve service during the period, i represents at the state of the unit with a time scale of 1 participating in the climbing reserve service during the period, the state of the unit with a time scale of 1 participating in the climbing reserve service during the period, i represents at the state of the unit with a time scale of 1 participating in the climbing reserve service during the period, the state of the unit with a time scale of 1 participating in the climbing reserve service during the period, i represents at the state of the unit with a time scale of 3 participating in the climbing reserve service during the period, t the state of the unit with a time scale of 3 participating in the climbing reserve service during the period, i represents at the state of the unit with a time scale of 8 participating in the climbing reserve service during the period, t the state of the unit with a time scale of 8 participating in the climbing reserve service during the period, i represents at the response capacity of the unit with a time scale of 1 participating in the climbing reserve service, i represents at the response capacity of the unit with a time scale of 3 participating in the climbing reserve service, i represents at the response capacity of the unit with a time scale of 8 participating in the climbing reserve service, i represents at the maximum output of the pumped-storage unit, i represents the power generation of the pumped-storage unit at during the period, i the power generation of the pumped-storage unit, represents the power generation of the pumped-storage unit at i during the period, represents the power generation of the pumped-storage unit at i during the period, represents the pumping power of the pumped-storage unit during the i period, represents the pumping power of the pumped-storage unit during the i period, represents the pumping power of the pumped-storage unit during the i period, represents the positive spinning reserve capacity provided by the unit during the i period, represents the positive spinning reserve capacity provided by the unit during the i period, represents the positive spinning reserve capacity provided by the unit during the i period, represents t the ramp reserve demand with a time scale of 1 during the period, t represents T the ramp reserve demand with a time scale of 3 during the

[0035] Furthermore, the expressions of the positive and negative spinning reserve constraints are as follows:

[0036] ;

[0037] ;

[0038] wherein, represents t the positive spinning reserve capacity provided by the unit during the i period, represents t the generating power of the pumped-storage unit during the i period, represents t the pumping power of the pumped-storage unit during the i period, represents t the load fluctuation during the period, t represents i the negative spinning reserve capacity provided by the unit during the period, t represents i the output power of the pumped-storage unit during the period, t represents i the predicted output of the wind turbine during the period, t the wind turbine during thei Abandoned wind power

[0039] The expression of the ramp reserve service duration constraint is as follows:

[0040] ;

[0041] Wherein, represents the minimum duration for the unit i to provide ramp reserve service, h represents the current time period, represents at t The time scale of the time period is x The state of the unit i participating in the ramp reserve service, represents at The time scale of the time period is x The state of the unit i participating in the ramp reserve service, x represents the time scale, T represents the maximum scheduling time period.

[0042] The beneficial effects of the above further solution are as follows: The present invention adopts ramp reserve constraints to make the ancillary services at different scales cooperate with each other, sets the ramp reserve duration constraint to enable the continuous provision of ancillary services, improves the enthusiasm of ancillary service providers, and ensures the reliability of power system operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is the method flow chart of the present invention.

[0044] Figure 2 is the ramp reserve model diagram in this embodiment.

[0045] Figure 3 is the IEEE24RTS system topology diagram in this embodiment.

[0046] Figure 4 is the conventional unit parameter diagram in this embodiment.

[0047] Figure 5 is the wind power predicted output and actual output curve diagram in this embodiment.

[0048] Figure 6 is the 8hr wind power down-ramp scenario Copula fitting result diagram in this embodiment.

[0049] Figure 7 is the power and energy balance index diagram of Scheme I, Scheme II, Scheme III and Scheme IV in this embodiment.

[0050] Figure 8This is the graph of the reserve capacity situation of Scenario I, Scenario II, Scenario III, and Scenario IV in this embodiment.

[0051] Figure 9 This is the graph of the power and energy balance indicators of Scenario IV, Scenario V, and Scenario VI in this embodiment.

[0052] Figure 10 This is the graph of the 8-hour wind power down-ramp demand prediction situation of Scenario IV, Scenario V, Scenario VI, and the actual wind power ramping down in this embodiment. Detailed implementation manners

[0053] The following describes the detailed implementation manners of the present invention to facilitate those skilled in the art of this technical field to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the detailed implementation manners. For those of ordinary skill in this technical field, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions created using the concept of the present invention are within the scope of protection.

[0054] Before describing this embodiment, the following terms are first explained:

[0055] Copula function: Link function;

[0056] Gaussian-Copula: Gaussian link function;

[0057] t-Copula: t-distribution link function;

[0058] AIC: Akaike information criterion.

[0059] Embodiment 1

[0060] As Figure 1 shown, the present invention provides a method for calling reserve capacity to cope with wind power down-ramping, and its implementation method is as follows:

[0061] S1. Obtain the wind power output sample data, and calculate the wind power down-ramping power data according to the wind power output sample data.

[0062] In this embodiment, in a power system with a high proportion of wind power, the risk brought by wind power down-ramping events is much greater than that brought by wind power up-ramping. Therefore, the present invention focuses on how to cope with wind power down-ramping events. In a scenario with large wind speed variability (such as extreme weather), the wind power output often continuously decreases within a certain period of time and causes a large power deficit. It can be considered to analyze the relationship between wind power output and wind power down-ramping and purchase ancillary service capacity to avoid the power and energy imbalance that may be caused by the wind power output deficit.

[0063] In this embodiment, after obtaining the wind power output sample data, the down-ramp power data of the wind power is calculated using the ramp power formula, and the expression of the ramp power formula is as follows:

[0064] ;

[0065] wherein, represents the down-ramp power of the wind power output in the t time period within x hours, represents the output of the wind power in the t time period in the wind power output sample data, represents the output of the wind power in the time period in the wind power output sample data.

[0066] S2. Calculate the marginal distribution functions of the wind power output sample data and the wind power down-ramp power data respectively, and use the Copula function to construct the joint distribution model of the wind power output and the wind power down-ramp. The specific steps are as follows:

[0067] S201. Calculate the marginal distribution functions of the wind power output sample data and the wind power down-ramp power data respectively;

[0068] S202. Determine the parameters of the Copula function using maximum likelihood estimation according to the marginal distribution function of the wind power output sample data and the marginal distribution function of the wind power down-ramp power data;

[0069] S203. Evaluate and screen the Copula function using the Akaike information criterion according to the goodness of fit and the number of parameters of the Copula function;

[0070] S204. Use the screened Copula function to construct the joint distribution model of the wind power output and the wind power down-ramp.

[0071] In this embodiment, the marginal distribution functions of the wind power output sample data and the wind power down-ramp data are calculated respectively, and the Copula function is used to construct the joint distribution model of the wind power output and the wind power down-ramp:

[0072] ;

[0073] wherein, and respectively represent the marginal distribution functions of the wind power output and the wind power down-ramp, represents the down-ramp power of the wind power output in the x hours in the wind power output sample data, represents the output of the wind power in the wind power output sample data;

[0074] Determine the parameters of the Copula function using maximum likelihood estimation, and determine the type of Copula function to be used. The expression of the maximum likelihood estimation is as follows:

[0075] ;

[0076] where represents the total number of samples, represents the parameter vector of the Copula function, represents the maximum likelihood estimation of the parameter vector of the Copula function;

[0077] Evaluate the Copula function from two dimensions of goodness of fit and number of parameters using the Akaike information criterion, and select a suitable Copula function. The expression of the AIC is as follows:

[0078] ;

[0079] where represents the number of parameters in the Copula function, reflecting the complexity of the model fitted by the Copula function, represents the maximum likelihood function value of the Copula function, reflecting the degree of fit between the model fitted by the Copula function and the actual data, A represents the Akaike information criterion, and a smaller A indicates that the overall goodness of fit of the model fitted by the Copula function is better.

[0080] S3. According to the joint distribution model of wind power output and wind power down-ramp, calculate the conditional probability equation of wind power ramp under the condition of wind power prediction power, and input the historical data of wind power output and wind power down-ramp into the conditional probability equation of wind power ramp to obtain the wind power output - wind power down-ramp boundary curve.

[0081] In this embodiment, according to the joint distribution model of wind power output and wind power down-ramp established by the Copula function, calculate the conditional probability expression of wind power ramp under the condition of wind power prediction power. The conditional probability expression is as follows:

[0082] ;

[0083] ;

[0084] where, from the conditional probability expression, the wind power output W is when the confidence level is of the wind power down-ramp boundary , and use the minimum value formula to ensure the finally output wind power down-ramp power Not exceeding the wind power output, represents the possible wind power ramp rate;

[0085] And input the historical data of wind power output and wind power down-ramp into the conditional probability expression of wind power ramp, to obtain the wind power output - wind power down-ramp boundary curve.

[0086] S4. According to the response time of cold standby gas units, design the ramp reserve service including time scales.

[0087] In this embodiment, as shown in Table 1 (Table 1 is the ramp reserve service table), according to the response time of cold standby gas units, design the ramp reserve service with different time scales. The design is mainly through refining the start-up time of gas units to ensure that the cold standby capacity of gas units can be called in time within the specified time, alleviating the peak shaving pressure of the power system. The reserve services include: spinning reserve, RM 1, RM 3 and RM 8. For RM 1, the units participating in RM 1 need to start within 1 hour after receiving the instruction to call the capacity and ensure that they can maintain the established power operation for 2 hours. RM 1 and RM 3 can participate in regulation during the idle period that RM 8 can respond to, to ensure the power balance during this period. Among them, the auxiliary service cost is negatively correlated with the unit response time, and the auxiliary service cost with a long response time is low.

[0088] Table 1

[0089]

[0090] S5. Substitute the wind power prediction value into the wind power output - wind power down-ramp boundary curve to obtain the predicted value of wind power ramp reserve demand, and combine spinning reserve and ramp reserve to construct a ramp reserve model; and use the ramp reserve model to obtain the ramp reserve demand, and use the ramp reserve service to obtain the optimal call of the standby capacity for wind power down-ramp. The specific steps are as follows:

[0091] S501. Substitute the wind power prediction value into the wind power output - wind power down-ramp boundary curve, and calculate to obtain the predicted value of wind power ramp reserve demand;

[0092] S502. Use spinning reserve and ramp reserve to provide standby capacity;

[0093] S503. Use the predicted value of wind power ramp reserve demand, load fluctuation, spinning reserve capacity and ramp reserve capacity to construct a ramp reserve model, and use the spinning reserve constraint to reserve the regulation capacity;

[0094] S504. Combine the predicted value of wind power ramp reserve demand, use the ramp reserve model to obtain the ramp reserve demand, and use the constraints related to ramp reserve service and ancillary service including time scale to obtain the optimal dispatch of the reserve capacity for the downward ramp of wind power.

[0095] In this embodiment, the predicted value of wind power is brought into the wind power output - downward ramp boundary curve of wind power to calculate the predicted value of wind power ramp reserve demand. And in the optimal dispatch, spinning reserve and ramp reserve are used to provide reserve capacity to cope with the uncertainty of wind power output and load fluctuations. Since the response speed of ramp reserve service is relatively slow, in order to ensure that the power system can reserve sufficient flexible regulation capacity, the spinning reserve constraint is added. Using the predicted value of wind power ramp reserve demand, load fluctuations, spinning reserve capacity and ramp reserve capacity, as Figure 2 shown, construct a ramp reserve model, and use the spinning reserve constraint to reserve flexible regulation capacity; combine the predicted value of wind power ramp reserve demand, use the ramp reserve model to obtain the ramp reserve demand, and use the constraints related to ramp reserve service and ancillary service to obtain the optimal dispatch of the reserve capacity for the downward ramp of wind power, and complete the dispatch of the reserve capacity for the downward ramp of wind power;

[0096] The expression of the ramp reserve model is as follows:

[0097] ;

[0098] ;

[0099] ;

[0100] Wherein, represents t the ramp reserve demand with a time scale of x in the period, t represents x the predicted value of wind power ramp reserve demand with a time scale of in the period, represents t the load fluctuation in the period, t represents x the ramp reserve capacity with a time scale of in the period, represents t the spinning reserve capacity in the t period, that is, the unit in the offline state in the t period. To reserve sufficient response time for the cold reserve capacity providing ramp reserve service (i.e., the unit in the offline state in the The ramping reserve capacity required during the time period to ensure that there is sufficient reserve capacity during the time period to cope with the changes in wind power output. Since the response speed of ramping reserve service is relatively slow, adding spinning reserve constraints ensures that the power system can reserve sufficient flexible regulation capacity. represents the maximum scheduling time period of the ramping service RMx .

[0101] In this embodiment, ramping reserve and spinning reserve are provided simultaneously to cope with the power deficit of the downward ramping of wind power. Among them, spinning reserve can cope with small new energy or load fluctuations and ensure power balance on a short time scale; ramping reserve mainly ensures sufficient capacity in the power system on a long time scale and can cope with a large power deficit of the downward ramping of wind power.

[0102] Embodiment 2

[0103] In this embodiment, as Figure 3 shown, an improved IEEE 24RTS (Institute of Electrical and Electronics Engineers 24-bus request-to-send protocol power network model) example is used for example analysis. As shown in Table 2 - Table 5 and Figure 4 shown, Table 2 shows the participation of each unit in reserve service, Figure 4 are the parameters of conventional units, Table 3 are the parameters of hydropower units, Table 4 are the parameters of pumped-storage units, Table 5 is the ramping reserve cost ratio. At this time, the downward ramping amplitude of wind power exceeds 30% of the installed capacity within 6 hours. The curtailment penalty and load shedding penalty are set at 700 yuan / MW and 2100 yuan / MW respectively. The scheduling period is set at 24 hr, the scheduling interval is set at 1 hr, the confidence level of wind power output fluctuation is 95%, and the load fluctuation is set at 10% of the load in this time period. This embodiment is written using MATLAB R2021a, and the optimization scheduling model is solved using GUROBI 9.5.2.

[0104] Table 2

[0105]

[0106] Table 3

[0107]

[0108] Table 4

[0109]

[0110] Table 5

[0111]

[0112] In this embodiment, as Figure 5As shown, a wind power output dataset of a certain country's power grid from 2020 to 2021 is selected to establish a sample set for wind power down-ramp prediction. The obtained wind power down-ramp prediction results are distributed to the wind farm nodes according to the load ratio of the example. The 6-hour down-ramp power of the selected wind farm output scenario exceeds 30% of the wind power installed capacity.

[0113] In this embodiment, it is considered to minimize the operating cost of the power system, including the operating cost of conventional units, the operating cost of pumped-storage units, the auxiliary service cost, the curtailment cost, and the load shedding cost. The units in the power system include conventional units, pumped-storage units, and wind turbines; the conventional units include: coal-fired units, large gas units, peaking units, hydroelectric units, and nuclear power units;

[0114] The operating cost of the power system is expressed as follows:

[0115] ;

[0116] ;

[0117] ;

[0118] ;

[0119] ;

[0120] ;

[0121] Among them, the cost: represents the operating cost of conventional units, represents the operating cost of pumped-storage units, represents the auxiliary service cost, represents the curtailment cost, represents the load shedding cost;

[0122] Set: represents the set of conventional units, represents the set of pumped-storage units, represents the set of units providing spinning reserve, represents the set of units providing ramping reserve service, represents the set of wind turbines, represents the set of nodes of hydroelectric units, represents the set of load nodes, represents the set of response times of ramping reserve products;

[0123] Variable: represents t the conventional unit in the iOutput power, and respectively represent t the start-up decision variable and shutdown decision variable of the conventional unit in the time period, and both are i variables, and and respectively represent t the output power and pumping power of the pumped storage unit in the time period, i and and respectively represent t the positive spinning reserve capacity and negative spinning reserve capacity provided by the unit in the time period, i represents the capacity provided by the unit in the time period t for the ramping reserve service i , RMx represents the curtailed power of the wind turbine in the time period t , i represents the curtailed power of the wind turbine in the time period t , i represents the load shed in the time period t , represents RMx the maximum scheduling time period of the ramping service T represents the maximum scheduling time.

[0124] Coefficients: , and respectively represent the power generation cost coefficient, start-up cost coefficient and shutdown cost coefficient of the controllable conventional unit i , and respectively represent the output cost coefficient and pumping cost coefficient of the pumped storage unit i , and respectively represent the cost coefficients of the unit i providing positive spinning reserve and negative spinning reserve, represents i the power generation cost coefficient of the unit RMx providing the ramping reserve service and respectively represent the curtailment cost coefficient and water curtailment cost coefficient, represents i the cost coefficient of the load shed at the cut-off node.

[0125] In this embodiment, a DC power flow model is used to model the power system, and its expression is as follows:

[0126] ;

[0127] ;

[0128] ;

[0129] Among them, represents t the load at the time period node i , and respectively represent t the voltage phase angle of the time period line l flowing into the bus and the voltage phase angle flowing out of the bus, represents the admittance of the line l , represents the maximum capacity of the line from node m to node n .

[0130] In this embodiment, the expression of the conventional unit constraint is as follows:

[0131] The expression of the unit capacity constraint is as follows:

[0132] ;

[0133] ;

[0134] ;

[0135] ;

[0136] The expression of the unit ramp rate constraint is as follows:

[0137] ;

[0138] The expression of the maximum start-stop times constraint of the conventional unit is as follows:

[0139] ;

[0140] The expressions of the unit state transition equation and the start-stop time constraint are as follows:

[0141] ;

[0142] ;

[0143] ;

[0144] Among them, represents t the conventional unit in the time periodi Whether it is in the running state and is a variable, indicating the conventional unit in a time period i Whether it is in the running state, and respectively indicate the i minimum output power and the maximum output power of the unit, and respectively indicate the i maximum up and down ramp rates, and respectively indicate t the conventional unit in a time period i start decision variable and stop decision variable, T indicates the maximum scheduling time. For the safety of the equipment, the coal-fired units and large gas turbines need to limit the number of starts and stops on the same day, indicates the i maximum number of starts of the unit, and respectively indicate the i minimum on-time and minimum off-time of the unit, while the nuclear power unit remains in the on state.

[0145] In this embodiment, for the wind turbines, it is required that the wind power output does not exceed the wind power prediction value. Therefore, the expression of the wind turbine constraint is as follows:

[0146] ;

[0147] ;

[0148] ;

[0149] Among them, indicates t the wind power curtailment of the wind turbines in a time period i and and respectively indicate t the scheduled output and the predicted output of the wind turbines in a time period i ;

[0150] In this embodiment, for the hydropower units, considering the water volume limitation, it is necessary to add the daily hydropower distribution power constraint. Therefore, the expressions of the unit power generation capacity constraint and the unit pumping state output constraint are as follows:

[0151] ;

[0152] ;

[0153] Among them,T represents the maximum scheduling time, represents t the output of the hydropower unit i during a certain period, represents t the curtailment power of the wind power unit i during a certain period, represents the daily allocated power i of the hydropower unit.

[0154] In this embodiment, for the pumped-storage unit, considering the reservoir capacity limit and the conversion between pumping and generating states, it is necessary to add a reservoir capacity constraint and a water balance constraint. The expression of the reservoir capacity constraint is as follows:

[0155] ;

[0156] ;

[0157] The expression of the water balance constraint is as follows:

[0158] ;

[0159] Among them, and represent t the generating state and pumping state of the pumped-storage unit i during a certain period, and are variables, and represent the minimum output and maximum output i of the pumped-storage unit, represents the initial reservoir capacity i of the reservoir where the energy storage unit is located, and respectively represent the minimum reservoir capacity and maximum reservoir capacity i of the reservoir where the energy storage unit is located, represents the reservoir capacity-electricity conversion coefficient, represents t the cumulative power generation i of the pumped-storage unit during a certain period, t represents the total water consumption of the pumped-storage unit during i a certain period,

[0160] In this embodiment, auxiliary service-related constraints are also constructed. The auxiliary service-related constraints include: a ramping reserve constraint, a positive and negative spinning reserve constraint, and a ramping reserve service duration constraint;

[0161] The expression of the ramping reserve constraint is as follows:

[0162] ;

[0163] ;

[0164] ;

[0165] Among them, represents the set of units providing uphill reserve services, represents the set of pumped-storage units, represents the set of conventional units, represents at t The unit at the time scale of x is in the state of participating in uphill reserve services, i represents the response capacity of the unit at the time scale of represents the time scale of x The unit of i participates in the response capacity of uphill reserve services, represents the maximum output of the pumped-storage unit i of represents During the period, the pumped-storage unit i of the generated power, represents During the period, the pumped-storage unit i of the pumping power, represents During the period, the unit i provides positive spinning reserve capacity, represents t During the period, the uphill reserve demand with a time scale of 1, represents t During the period, the uphill reserve demand with a time scale of 3.

[0166] The expressions of the positive and negative spinning reserve constraints are as follows:

[0167] ;

[0168] ;

[0169] Among them, represents the set of conventional units, represents the set of pumped-storage units, represents t During the period, the unit i provides positive spinning reserve capacity, represents the pumped-storage unit i of the maximum output, represents t During the period, the pumped-storage unit iThe power generation represents t the pumping power of the pumped-storage unit during i a certain period; represents t the load fluctuation during a certain period; t represents i the negative spinning reserve capacity provided by the unit during a certain period; t represents i the output power of the pumped-storage unit during a certain period; t represents i the predicted output of the wind turbine during a certain period; t represents i the curtailed power of the wind turbine during

[0170] The expression of the ramp reserve service duration constraint is as follows:

[0171] ;

[0172] wherein, represents i the minimum duration for the unit to provide ramp reserve service, h represents the current period, represents t at the time scale of x for the unit during i the state of participating in the ramp reserve service, represents at the time scale of x for the unit during i the state of participating in the ramp reserve service, x represents the time scale, T represents the maximum scheduling period.

[0173] Embodiment 3

[0174] In this embodiment, the Copula function is used to model the correlation between wind power output and wind power down-ramp. For the wind power down-ramp scenarios with time scales of 1hr, 3hr, and 8hr, the AIC calculation results of the Copula function are shown in Table 6. It can be seen from the AIC index that Clayton-Copula has the best goodness of fit in the three time scale scenarios. Therefore, Clayton-Copula is used for wind power down-ramp power prediction in the joint distribution model of wind power output and wind power down-ramp.

[0175] Table 6

[0176]

[0177] As Figure 6 shown, as the wind power output increases, the possible downward ramping power of wind power also gradually increases, which is consistent with actual experience. In this embodiment, the downward ramping prediction fitted by the joint distribution model of wind power output and wind power downward ramping can cover most historical wind power downward ramping scenarios. Therefore, it can prove the effectiveness of the wind power downward ramping prediction result of the joint distribution model of wind power output and wind power downward ramping, and reserve spare capacity based on this prediction result can cope with the vast majority of wind power downward ramping events.

[0178] In this embodiment, four schemes are set as follows:

[0179] Scheme I: Deterministic optimization + spinning reserve;

[0180] Scheme II: Robust optimization + spinning reserve;

[0181] Scheme III: Ramping reserve considering output-downward ramping correlation + coupling of cold and hot reserves;

[0182] Scheme IV: Considering output-downward ramping correlation + ramping reserve considering only cold reserve + spinning reserve (the method of this embodiment);

[0183] Among them, as shown in Table 7, Table 7 is the evaluation index of power and energy balance. Considering the influence of wind power uncertainty, the practical net load is calculated when calculating the evaluation index of power and energy balance; to simulate the actual dispatching situation, the actual wind power output data is used for adequacy test; and the evaluation indexes of power and energy balance are standardized for convenient comparison of the performance of different evaluation indexes of power and energy balance.

[0184] Table 7

[0185]

[0186] In this embodiment, as shown in Table 8, Table 8 compares the day-ahead planning costs of the four schemes. It can be obtained that the cost of Scheme IV is higher than that of Scheme II, lower than that of Scheme I, and equal to that of Scheme III. Scheme IV and Scheme III have more sources of ancillary service capacity, so they are more inclined to purchase ancillary services to meet the requirements of power and energy adequacy; while the available spare capacity of Scheme II is limited. In scenarios with high wind power uncertainty, in order to meet the requirements of safe operation, it is necessary to reduce wind power output and arrange negative reserve to ensure that the spare capacity can cover the uncertainty of wind power output. Therefore, its economy is low; Scheme I chooses to maximize the consumption of wind power to avoid curtailment penalties. Although its economy is the highest, due to its failure to consider the influence of uncertainty factors, and without purchasing sufficient ancillary service capacity or performing a certain amount of wind curtailment, it is difficult to cope with the net load fluctuation, and Scheme I undertakes part of the upward spare capacity due to low-cost hydropower and pumped storage, so the spinning reserve cost is less.

[0187] Table 8

[0188]

[0189] In this embodiment, as Figure 7 shown, the power and electricity balance indicators of four scenarios are presented. In terms of power and electricity adequacy, Scenario II, Scenario III, and Scenario IV consider the adverse effects that the uncertainty of wind power output may bring to the power system, enabling the power system to maintain a relatively high level of power and electricity adequacy. However, Scenario I does not consider the uncertainty of wind power output, resulting in a problem of insufficient adequacy during actual operation. In terms of power supply balance, Scenario I and Scenario II use wind curtailment and thermal reserve capacity to cope with the uncertainty of wind power, and can flexibly adjust the reserve capacity to make the adequacy more balanced in each time period. Scenario III and Scenario IV introduce cold reserve capacity. Since considering a relatively long response time and duration is required when calling cold reserve capacity, the flexibility is limited, so the balance is relatively poor. However, Scenario IV adds constraints related to thermal reserve capacity, enabling the power system to retain more thermal reserve capacity and increasing the flexible adjustment ability of the power system. Therefore, the power supply balance index of Scenario IV is better than that of Scenario III.

[0190] In this embodiment, as Figure 8 shown, the reserve capacity situations of four scenarios are presented. It can be seen that Scenario I does not consider the uncertainty of wind power and purchases less upward reserve capacity; Scenario II chooses to increase wind curtailment and purchase a large amount of downward reserve capacity to ensure that the overall reserve capacity can cover the entire uncertain interval of wind power. The overall reserve capacities of Scenario III and Scenario IV are close, but Scenario III does not limit the thermal reserve capacity and is more inclined to purchase low-cost cold reserve capacity to increase economic efficiency; while Scenario IV adds constraints on thermal reserve capacity and purchases more thermal reserve resources, enabling a more sufficient spinning reserve capacity to be directly called when there are small fluctuations in wind power or load during daily operation, making the power system more flexible.

[0191] In this embodiment, to verify the application effect of the uncertain demand prediction model, Scenario V and Scenario VI are added:

[0192] Scenario V: Interval optimization + ramp reserve + spinning reserve;

[0193] Scenario VI: Do not consider the correlation between down-ramp and wind power output + ramp reserve + spinning reserve;

[0194] Among them, the method for generating the uncertain interval of wind power output in interval optimization is the same as that of robust optimization, and the ramp reserve demand in interval optimization is taken as the maximum distance between two time periods in the power generation interval. Not considering the correlation between wind power output and down-ramp means only giving the corresponding quantile data based on historical down-ramp conditions without considering the relationship between wind power ramp and wind power output.

[0195] In this embodiment, as shown in Table 9, the table shows the day-ahead planned costs of Scenario IV, Scenario V, and Scenario VI. Among them, the cost of Scenario V increased by 5.5% and 7.4% compared to the costs of Scenario IV and Scenario VI respectively. The increased costs mainly come from the ancillary service cost and the operation cost. As Figure 9 shown, the power and energy balance indicators of Scenario IV, Scenario V, and Scenario VI were evaluated. The other indicators of the three scenarios are relatively close. The peak shaving available capacity of Scenario VI is the highest, and the peak shaving available capacities of the other two scenarios are close. Further analysis is needed on the reasons for the situation.

[0196] Table 9

[0197]

[0198] As Figure 10 shown, the downward ramp demand prediction situation with a time scale of 8 hours was selected for analysis. In the scenario of wind power change, Scenario IV and Scenario V can predict the change trend of the actual downward ramp demand. However, Scenario V only calculates based on the uncertain interval of the downward ramp demand and does not relax based on the historical output-downward ramp situation constraint. Therefore, the calculated reserve demand is too conservative and the economy is poor. Scenario VI does not consider the relationship between wind power output and wind power downward ramp during the scheduling period and cannot adjust the reserve capacity for the possible change trend of the downward ramp demand. Although Figure 9 judging from the evaluation indicators, the peak shaving available capacity of Scenario VI is relatively large, but when the wind power output is small, Scenario VI still maintains a relatively high reserve demand, far exceeding the actual reserve demand. Therefore, the economy of Scenario VI is poor.

[0199] In this embodiment, to analyze the feasibility of the method in this embodiment under different new energy penetrations, Scenario II, Scenario IV, and Scenario VI were selected for comparison, and 3 scenarios were set for experiments;

[0200] Scenario I: The total installed capacity of wind power is 20% of the total installed capacity;

[0201] Scenario II: The total installed capacity of wind power is 25% of the total installed capacity;

[0202] Scenario III: The total installed capacity of wind power is 30% of the total installed capacity;

[0203] As shown in Table 10, the table presents the operating costs of the power system under different wind power penetration rates. As the wind power penetration rate continues to rise, on the one hand, the power system can save generation costs by accommodating a large amount of wind power. On the other hand, the cost of ancillary services required to cope with the uncertainty of wind power output gradually increases. Compared with Plan IV, Plan II does not have abundant reserve resources to support a higher level of wind power accommodation. Therefore, Plan II increases the planned curtailment of wind power and purchases a large amount of downward spinning reserve. While the ancillary service cost increases, the generation cost saved is limited, and the total operating cost of Plan II increases. Therefore, Plan IV introduces a ramping reserve service and uses cold reserve resources to cope with the uncertainty of wind power, which can alleviate the problem of difficult wind power accommodation in a high-proportion new energy system and save the operating cost of the power system. When the wind power penetration rate rises, the cost of Plan VI decreases slightly. However, since Plan VI does not conduct a more refined modeling of wind power output and downward ramping power, there is more redundant reserve capacity when the wind power output is small. Therefore, the economy of Plan VI is poor.

[0204] Table 10

[0205]

Claims

1. A method for calling reserve capacity to cope with wind power ramp down, applied to a power system, characterized in that: The following steps are involved: S1. Obtain wind power output sample data, and calculate wind power down-slope power data based on the wind power output sample data; S2. Calculate the marginal distribution functions of wind power output sample data and wind power ramp-down power data respectively, and use the Copula function to construct a joint distribution model of wind power output and wind power ramp-down, specifically: S201, respectively calculating the marginal distribution functions of wind power output sample data and wind power down-ramp power data; S202, determining the parameters of the Copula function using maximum likelihood estimation according to the marginal distribution function of the wind power output sample data and the marginal distribution function of the wind power down-ramp power data; S203, evaluating and screening the Copula function using the Akaike Information Criterion according to the goodness of fit and the number of parameters of the Copula function; S204, using the selected Copula function, constructing a joint distribution model of wind power output and wind power ramp down; S3. According to the joint distribution model of wind power output and wind power ramp-down, the conditional probability equation of wind power ramp-down under the wind power forecast power condition is calculated, and the historical data of wind power output and wind power ramp-down are input into the conditional probability equation of wind power ramp-down to obtain the wind power output-wind power ramp-down boundary curve; S4. Design a ramping standby service with a time scale based on the response time of the cold standby gas group; S5. Use the wind power prediction value to bring into the wind power output-wind power ramp boundary curve to obtain the wind power ramp reserve demand prediction value, and combine the rotating reserve and the ramp reserve to build a ramp reserve model; and use the ramp reserve model to obtain the ramp reserve demand, and use the ramp reserve service to obtain the optimal call of the wind power ramp reserve capacity, specifically: S501, bringing the wind power prediction value into the wind power output-wind power down-ramp boundary curve, and calculating the wind power ramp reserve demand prediction value; S502, using spinning reserve and ramping reserve to provide reserve capacity; S503, using the wind power ramp reserve demand forecast value, load fluctuation, spinning reserve capacity and ramp reserve capacity to build a ramp reserve model, and using spinning reserve constraints to reserve adjustment capacity; S504. In combination with the predicted value of the wind power ramping reserve demand, the ramping reserve demand is obtained using the ramping reserve model, and the optimal call of the wind power ramping reserve capacity is obtained using the ramping reserve service and auxiliary service related constraints including the time scale.

2. The method for calling the reserve capacity for wind power down-slope according to claim 1 is characterized in that: The expression of the ramp-up standby model is as follows: ； ； ； in, express t The time scale of the period is x The climbing reserve demand, express t The time scale of the period is x The wind power ramp reserve demand forecast value, express Load fluctuations during the period, express t Load fluctuations during the period, express t The time scale of the period is x The climbing reserve capacity, express Spinning reserve capacity for the period, express t Spinning reserve capacity for the period, Indicates climbing service RMx The maximum scheduling period.

3. The method for calling the reserve capacity for wind power down-slope according to claim 1 is characterized in that: The auxiliary service related constraints include: climbing reserve constraints, positive and negative spinning reserve constraints, and climbing reserve service duration constraints; The expression of the climbing reserve constraint is as follows: ； ； ； in, represents the set of units that provide ramp reserve service with a time scale of 1, represents the set of units that provide ramp reserve service with a time scale of 3, represents the set of units that provide ramp reserve service with a time scale of 8, represents the collection of pumped storage units, Represents a collection of conventional units, Indicated in t Units with a time scale of 1 i The status of participating in the ramp standby service, Indicated in Units with a time scale of 1 i The status of participating in the ramp standby service, Indicated in Units with a time scale of 1 i The status of participating in the ramp standby service, Indicated in t Units with a time scale of 3 i The status of participating in the ramp standby service, Indicated in t Units with a time scale of 8 i The status of participating in the ramp standby service, Represents a unit with a time scale of 1 i The response capacity of the participating ramp-up reserve service, Represents a unit with a time scale of 3 i The response capacity of participating in the ramping reserve service, Represents a unit with a time scale of 8 i The response capacity of participating in the ramping reserve service, Pumped storage unit i The maximum output, express Period pumped storage unit i The power generation capacity, express Period pumped storage unit i The power generation capacity, express Period pumped storage unit i The power generation capacity, express Period pumped storage unit i The pumping power, express Period pumped storage unit i The pumping power, express Period pumped storage unit i The pumping power, express Time unit i The positive spinning reserve capacity provided, express Time unit i The positive spinning reserve capacity provided, express Time unit i The positive spinning reserve capacity provided, express t The ramp reserve demand with a time scale of 1, express t The ramp reserve demand with a time scale of 3, T Indicates the maximum scheduling period.

4. The method for calling the reserve capacity for wind power down-slope according to claim 3 is characterized in that: The expressions of the positive and negative spinning reserve constraints are as follows: ； ； in, express t Time unit i The positive spinning reserve capacity provided, express t Period pumped storage unit i The power generation capacity, express t Period pumped storage unit i The pumping power, express t Load fluctuations during the period, express t Time unit i Negative spinning reserve capacity provided, express t Period pumped storage unit i The output power, express t Wind turbines i The predicted output of express t Wind turbines i of wind power abandoned; The expression of the ramp standby service duration constraint is as follows: ； in, Indicates the unit i The minimum duration for providing ramp backup service, h Indicates the current period. Indicated in t The time scale of the period is x The crew i The status of participating in the ramp standby service, Indicated in The time scale of the period is x The crew i The status of participating in the ramp standby service, x Indicates the time scale.

Images