Optimization method and system for regulating pumped storage units to promote new energy consumption

Abstract

The present invention discloses an optimization method and system for promoting the consumption of new energy by regulating a pumped-storage unit. The method takes the minimum net load variance and the lowest comprehensive operating cost as optimization goals, considers the constraint regulation inside the power system and the pumped-storage unit, proposes indicators for the consumption of new energy by the pumped-storage unit, establishes a model for promoting the consumption of new energy by fine regulation of pumped-storage power stations, adopts a clustering method of a Gaussian mixture model, divides the daily load into three categories according to the characteristics of the daily load change, and establishes a daily load time series characteristic and time period classification model; according to the characteristics of the daily load time period classification model, three schemes for consuming new energy by multiple pumped-storage units that effectively track the changes in the daily load rate curve of the power grid are established, and the fine control of the pumped-storage units in the variable time period is realized to promote the consumption of new energy; and the fine regulation of multiple pumped-storage units in the multi-energy complementary joint system in time and section is realized to provide an accurate and feasible solution for the efficient consumption of renewable energy and grid connection in the new power system.

Description

Technical Field

[0001] The present invention belongs to the field of coordinated optimization of pumped storage power stations for new energy, and specifically relates to an optimization method and system for regulating pumped storage units to promote the consumption of new energy. Background Art

[0002] The problem of new energy consumption such as wind and solar abandonment in new power systems has always plagued the healthy development of wind power and photovoltaic power. In fact, the problem of new energy consumption is strongly correlated with key factors such as load scale, power supply regulation, and grid interconnection. Especially in the current stage of rapid growth of new energy, the problems of mismatch between power supply growth and load growth, insufficient peak-shaving capacity (flexibility) of the power system, and poor transmission channels for new energy are becoming increasingly prominent, which leads to greater challenges for the consumption of new energy in power systems with a high proportion of renewable energy. At present, the solution to this dilemma is to redistribute energy in time and space with combined energy storage technology (such as pumped storage or electrochemical energy storage, etc.), so as to provide guarantees for achieving high consumption of new energy and enhancing the operational stability of the power system. Among energy storage technologies, pumped storage power stations are widely used because of their large energy storage capacity, low energy storage cost, and few controllable restrictions on energy storage cycles. In addition, pumped storage units also have efficient two-way response capabilities. On the power supply side, pumped storage power stations can smooth the effect of new energy output and ensure stable power transmission; on the grid side, pumped storage power stations can alleviate the peak-to-valley difference of load and be used to consume surplus electricity during peak and valley periods. Therefore, for the new power system based on new energy, further utilizing pumped-storage power stations to promote the consumption of new energy and ensure power supply is of great significance to promoting the green and low-carbon transformation of energy.

[0003] In addition, with the high proportion of wind power and photovoltaic power connected to the grid, the impact of uncontrollable factors such as environment and climate has increased, and the intermittent, volatility and uncertainty of wind and solar power output have become increasingly prominent. In addition, it is difficult to accurately predict wind and solar power output, making it difficult to match the output power and load demand when large-scale wind and solar power are connected to the grid. In the traditional power system optimization strategy, the potential of pumped storage power stations has not been fully tapped. The load and start and stop of the units are still arranged only by the energy consumption of the units to adjust the volatility of wind and solar power output. The optimization strategy goal usually only considers the total power generation cost of the system to achieve the optimal operation economy. This strategy often leads to frequent start and stop of the units and cannot meet the installed capacity requirements of new energy. The frequent start and stop of pumped storage units will seriously threaten the safety and economy of the power grid, which is not conducive to low-carbon economic operation. At the same time, traditional optimization strategies are difficult to achieve precise control of pumped storage power stations when the daily load fluctuations of the power system are large. Therefore, how to give full play to the advantages of flexible and precise deployment of pumped storage in the grid connection of a high proportion of new energy, propose new optimization configuration strategies, make full use of the characteristics of fast start-up speed and strong peak-shaving capacity of pumped storage units, alleviate the pressure on the grid facing the access of large-scale new energy, and realize the refined regulation of pumped storage to promote the consumption of new energy. This is a difficult problem that urgently needs to be broken through for the low-carbon economic operation of the current new power system. Summary of the invention

[0004] The purpose of the present invention is to provide an optimization configuration strategy based on the refined regulation of pumped storage power stations to promote the consumption of new energy, give full play to the advantages of flexible and accurate calling of pumped storage in the grid connection of high proportion of new energy, propose a new optimization control strategy, make full use of the characteristics of fast starting speed and strong peak-shaving capacity of pumped storage units, alleviate the pressure on the power grid to access large-scale new energy, and realize the refined regulation of pumped storage to promote the consumption of new energy.

[0005] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is: an optimization method for regulating a pumped storage unit to promote the consumption of new energy, comprising the following steps:

[0006] Taking the objective function of improving the grid load optimally by pumped storage power stations, i.e. minimizing the net load variance within the dispatching cycle and minimizing the comprehensive operating cost, and considering the operating constraints of each unit in the power system, a model for promoting the consumption of new energy by fine-tuning pumped storage power stations is established.

[0007] According to the establishment of a model for promoting the consumption of new energy through refined regulation of pumped storage power stations, the peak regulation capacity of pumped storage units under pumping conditions is extended to the consumption capacity of pumped storage units for new energy, and the utilization of new energy by pumped storage units is divided into four working conditions;

[0008] The clustering method of Gaussian mixture model is used for daily load curve, and the daily load time series characteristics and time period classification model are established for the daily load curve;

[0009] Based on the four working conditions, three strategies are formulated for pumped storage power station regulation involving pumped storage units in the consumption of new energy;

[0010] According to the daily load time series characteristics and time period classification model, for the low-precision time periods, medium-precision time periods and high-precision time periods of the daily load curve, based on the predicted output of new energy and the daily load characteristic curve, under the premise of meeting the reservoir capacity and water level constraints, the three strategies involving pumped-storage units in the consumption of new energy are used to adjust the pumped-storage units to achieve the consumption of new energy through variable time period control.

[0011] Furthermore, the model of promoting the consumption of new energy through refined regulation of pumped storage power stations is a two-layer model.

[0012] Inner model: It is necessary to meet the operating constraints of the pumped storage power station and adjust the power of the pumped storage power station through demand response assistance;

[0013] Outer model: The optimization goal is to improve the peak-to-valley fluctuation of net load by pumped-storage units and minimize the comprehensive operating cost of the joint operation system. Three strategies for three types of pumped-storage units to participate in regulating new energy consumption are formulated for the new energy consumption capacity model.

[0014] Furthermore, three strategies involving pumped storage units in the consumption of new energy are:

[0015] Full tracking strategy, the output of the pumped storage unit fully tracks the daily load rate characteristic curve of the power grid;

[0016] Coordinated tracking strategy, dividing the daily load curve of the power grid into three periods: peak, flat and valley. Pumped storage units assist thermal power units in storing combined unit output to track the daily load curve of the power grid.

[0017] In the incomplete tracking strategy, the pumping capacity and power generation capacity of the pumped storage unit are both at their full capacity. The pumped storage unit generates electricity when the grid load is peak and pumps water when the load is low. The daily pumping power and total power generation are balanced.

[0018] Furthermore, the clustering method of Gaussian mixture model is used for the daily load curve to establish the daily load curve. The daily load time series characteristics and time period classification model include:

[0019] Use Gaussian mixture model for clustering. Assume that the input samples obey k Gaussian distributions with different parameters. Fit k mixed Gaussian distributions by maximum expectation method to obtain k cluster distributions of all samples and their distribution means and covariances. Use the obtained means and covariances to cluster samples based on the maximum probability γ(a,b).

[0020] According to the time series characteristics of daily load, time period clustering is carried out to conduct static analysis between each time period, and the average change rate S of the daily load time period is used.b,ave and standard deviation S b,std Characterizes the magnitude and severity of daily load changes in time series, S b,ave and S b,std The larger the value, the greater the change in daily load and the more drastic the degree. The average value f of the frequency of daily load changes in the direction within adjacent time periods is used. b,ave and standard deviation f b,std Characterizes the frequency and trend of daily load changes in direction, where f b,ave and f b,std The larger the value is, the less obvious the change trend of daily load is. That is, the input sample of the period characteristic of daily load curve is:

[0021] x b =[S b,ave ,S b,std ,f b,ave ,f b,std ] (8)

[0022] Set the daily load change rate mean, daily load change frequency mean, daily load change rate standard deviation judgment threshold, and daily load change frequency standard deviation judgment threshold to That is, each parameter in the sample has two situations: greater than or less than the threshold. If x b Clustering generates 16 types, which are then simplified into 4 types. The change amplitude of daily load limits the influence of its change frequency on the stability of daily load. Under the premise, regardless of f b,ave ,f b,std Whether it is greater than the threshold, the daily load change degree is low, and then three types of daily load are obtained: low-precision period (T l ), medium precision period (T m ) and high-precision time period (T h ).

[0023] Furthermore, the four working conditions are:

[0024] ① The pumped storage unit operates at rated power under pumping conditions, and the pumped storage power station fully utilizes its capacity to absorb new energy;

[0025] ② The pumped storage units operate at less than rated power under pumping conditions, the system does not abandon wind and solar power, the pumped storage power station does not put all units into operation to pump water, and the pumped storage power station has fully utilized its capacity to absorb new energy;

[0026] ③ When the pumped storage unit operates at rated power under pumping conditions, there is wind and solar power abandonment in the system, and the amount of wind and solar power abandonment in the system is less than the remaining pumpable capacity of the pumped storage power station. The capacity of the pumped storage power station to absorb new energy is:

[0027]

[0028] ④ The pumped storage power station operates at less than rated power under pumping conditions, and there is wind and solar power abandonment in the system. The amount of wind and solar power abandonment in the system is greater than the remaining pumpable capacity of the pumped storage power station. The capacity of the pumped storage power station to absorb new energy is:

[0029]

[0030] P w,t is the predicted power of the system wind power and P wr,t is the wind power consumption of the system; P pv,t is the predicted power of the system photovoltaic and P pvr,t is the photovoltaic power consumption of the system.

[0031] Furthermore, the objective function includes the variance of the net load variance within a day and the comprehensive operating cost, as follows:

[0032] Objective function I: Minimize the variance of net load variance within a day

[0033]

[0034] Objective function II: Minimize the operating cost of the joint operation system

[0035]

[0036] Where: P L,t is the load value in period t; The power generation capacity of the pumped storage power station; Pumping power of pumped storage power station; F SO F is the operating cost of thermal power units; PS is the operating cost of the pumped storage power station; j , b j and c j They correspond to the coal consumption coefficient of thermal power units; P G,j,t is the output value of the thermal power unit; S jt is the start-up and shutdown cost of thermal power unit j; u j,t is the start and stop state of the thermal power unit; where: and are Boolean variables indicating that the unit is in power generation and pumping conditions respectively; C h The loss cost of starting and shutting down the unit.

[0037] Furthermore, the power system constraints include power balance constraints, thermal power unit output constraints, thermal power unit climbing constraints, thermal power unit minimum start and stop time constraints, pumped storage unit power constraints, pumped storage unit and single operating condition constraints, pumped storage unit working unit number constraints, pumped storage power station water level and capacity constraints and unit maximum start and stop times constraints.

[0038] Furthermore, the daily load time series characteristics and time period classification model are as follows:

[0039] ① When the mean and standard deviation of the daily load curve change rate are both less than the set judgment threshold, that is, The daily load curve shows a stable trend, with a low variation range and low complexity, and is defined as the low-precision period T l , the day-ahead scheduling plan takes 1h as the time scale;

[0040] ② The daily load curve change rate is greater than or equal to the judgment threshold, and the change frequency is less than or equal to the judgment threshold, that is, The daily load curve of the medium-precision period T is defined as the daily load curve of the medium-precision period T. m , the day-ahead scheduling plan uses 30 minutes as the time scale;

[0041] ③ The mean and standard deviation of the daily load curve change rate and frequency are both greater than or equal to the judgment threshold, that is, The daily load curve changes frequently in direction, with a large change range and an unclear trend, and is highly complex. It is defined as a high-precision period T h The day-ahead dispatch plan adopts a 15-minute time scale, which can more accurately fit the daily load characteristics.

[0042] Furthermore, T l During this period, the daily load fluctuation event is smoothed by the pumped storage power station, and the thermal power plant adopts a constant power generation mode, as shown in the following formula:

[0043]

[0044] Where: P G,i,t is the output of thermal power unit i at time t; end Indicates the end time point of this period;

[0045] T m During this period, the daily load demand changes greatly and shows an obvious increase or decrease trend. At this time, the system needs to have a large capacity climbing capability. The thermal power unit adopts the conventional mode and operates in conjunction with the pumped storage power station to meet the load demand and climbing demand. The pumped storage unit performs a part of the sharing. The output of the thermal power unit is:

[0046]

[0047] T h During this period, although the daily load changes frequently in direction and with a large amplitude, there is no obvious increase or decrease trend, and the system does not need to have a large-capacity climbing capability. The fluctuation of daily load is smoothed by the pumped storage power station, and the thermal power unit adopts a constant power generation mode;

[0048] When the storage capacity cannot meet S SOPs When the constraint is met, the pumped storage unit cannot work. During this period, the pumped storage unit cannot work. In order to meet the daily load demand, the system needs to abandon wind and solar power or cut load.

[0049] Based on the above time-division control, set the SOPs exist The interval is the adjustable storage capacity allowed by the reservoir, among which and is the minimum and maximum value of the reservoir water level that can be adjusted, and controls S SOPs Maintain within normal water storage capacity.

[0050] Based on the same inventive concept, the present invention provides an optimization configuration system for promoting the consumption of new energy based on the regulation of pumped storage units, including a consumption model construction module, a classification module, a clustering module, a strategy acquisition module and a curve acquisition module:

[0051] The consumption model construction module is used to take the pumped storage power station to improve the optimal grid load, that is, the minimum net load variance within the dispatch cycle and the lowest comprehensive operating cost as the objective function, consider the constraints of the operation of each unit in the power system, and establish a pumped storage power station fine regulation to promote the consumption of new energy models;

[0052] The classification module is used to establish a model for fine-tuning pumped storage power stations to promote the consumption of new energy, extend the peak-shaving capacity of pumped storage units under pumping conditions to the consumption capacity of pumped storage units for new energy, and divide the use of new energy by pumped storage units into four conditions;

[0053] The clustering module is used to cluster the daily load curve using the Gaussian mixture model to establish a daily load time series characteristic and time period classification model for the daily load curve;

[0054] The strategy acquisition module is used to formulate three strategies for the pumped storage power station to participate in the consumption of new energy sources based on the four working conditions;

[0055] The curve acquisition module is used to adjust the variable time period control of the pumped storage unit to realize the absorption of new energy according to the predicted output of new energy and the daily load characteristic curve of the low-precision period, medium-precision period and high-precision period of the daily load curve based on the daily load time series characteristics and time period classification model, and under the premise of meeting the reservoir capacity and water level constraints, according to the three strategies involving pumped storage units in the absorption of new energy.

[0056] Compared with the prior art, the present invention has at least the following beneficial effects: the present invention takes the optimization of daily load peak-valley fluctuation improvement by pumped storage and the minimum comprehensive operation cost of the joint operation system as the objective function, comprehensively considers the constraints of the units in the system, and establishes a pumped storage power station fine regulation model to promote new energy consumption in view of the ability of pumped storage power stations to consume new energy and serve the power grid, and formulates three strategies involving pumped storage power stations in participating in new energy consumption. The clustering method of the Gaussian mixture model is used to divide the daily load curve into three categories, and the daily load time series characteristics and time period classification model are established, realizing the fine control of the variable time period of the pumped storage unit, and providing a feasible, economical and accurate operation optimization configuration strategy for the fine regulation of pumped storage power stations to promote the consumption of new energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 Functional diagram of the pumped storage unit's ability to serve the grid.

[0058] Figure 2 Optimal control strategy for pumped storage units participating in new energy consumption. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings.

[0060] Reference Figure 1 Pumped storage is a storage technology that uses water as an energy carrier. It pumps water to absorb surplus electricity when the system load is low, and releases water to generate electricity when the system load is peak. It is an excellent peak-shaving power source in the system. At the same time, due to its flexible operation and rapid adjustment, it can also undertake important functions such as frequency regulation, phase regulation, emergency backup and black start in the power system.

[0061] Reference Figure 2 According to the daily load curve of the power grid and the predicted output of wind and solar power, three strategies for pumped storage power stations to absorb new energy are formulated. The inner layer formulates the pumped storage absorption strategy: the daily load curve is divided into three periods of low, medium and high precision, with the pumped storage power station improving the optimal net load peak-valley fluctuation and the lowest comprehensive operating cost as the optimization goal; the outer layer sets the output and storage capacity constraints of the pumped storage power station: based on the divided daily load curve, it is necessary to meet the operating constraints of the pumped storage power station and adjust the power of the pumped storage power station through demand response assistance. For the specific processes of the inner and outer layers, please refer to the attached manual Figure 2 ,

[0062] Step 1, input the predicted values ​​of system wind power, photovoltaic power, load and;

[0063] Step 2: Gaussian mixture clustering is used to divide the daily load curve into low, medium and high precision periods according to the mean and standard deviation of the change rate and frequency, with the optimization goal of improving the peak-to-valley fluctuation of net load and minimizing the joint operation cost as (2a).

[0064] Step 3, under the premise of satisfying the reservoir capacity water level (SOPs) constraint, the optimization control strategy based on the regulation of pumped storage units to promote the consumption of new energy is as follows: taking the improvement of the net load peak-valley fluctuation to be the best and the lowest joint operation cost as the optimization goal, recorded as (2a), setting the number of pumped storage power station units according to the predicted output of wind power and photovoltaic power generation, and taking the reference initial value of the output of each unit as (2b);

[0065] Step 4: Control the pumped storage machine in a refined variable time period to promote the consumption of new energy. That is, according to the predicted output of new energy and the three time periods of daily load, calculate the pumping power P under the reference initial value of the predicted output. ps At the same time, the capacity of pumped storage is calculated to determine whether the capacity of pumped storage exceeds the capacity constraint of the pumped storage unit and whether the reservoir water level is above the warning value. If so, the output of the pumped storage unit and the consumption of wind power and photovoltaic power in this period are output. Otherwise, the demand response auxiliary function is used to return to step (2b).

[0066] The present invention aims at the wind-solar-thermal-pumped storage joint operation system, and the role of pumped storage in cooperating with renewable energy operation to improve the level of renewable energy consumption is highlighted. An indicator of the new energy consumption capacity of pumped storage units is proposed, and a model for the refined regulation of pumped storage power stations to promote the new energy consumption capacity is established. When the pumped storage unit operates in pumping conditions, it absorbs the active power of the power grid, which provides space for the consumption of new energy. Therefore, the pumped storage unit can be used as the ability to consume new energy when operating under water pump conditions.

[0067] According to the output characteristics of the pumped storage unit under pumping conditions and the prediction of new energy in the system, the utilization of new energy consumed by the pumped storage unit is divided into the following conditions:

[0068] Condition I: The pumped storage power station operates at rated power under pumping conditions (P h,t =P hN )

[0069] The wind and solar power abandonment rates of pumped storage power stations that are not put into operation are:

[0070]

[0071]

[0072] Where: P w,t is the predicted power of wind power in the system; P wr,tis the wind power consumption of the system; P pv,t is the predicted power of the system photovoltaic; P pvr,t is the photovoltaic power consumption of the system.

[0073] If the pumped storage power station is put into operation at this time, the wind and solar power abandonment rates of the system are:

[0074]

[0075]

[0076] In this case, it is considered that the pumped storage power station has fully utilized its capacity to absorb new energy, that is:

[0077] N N,t =100% (19)

[0078] Working condition II: When the pumped storage unit is running at less than rated power under pumping conditions (P h,t ≤P hN ) The system has wind and solar power curtailment (P w,t ≥P wr,t ,P pv,t ≥P pvr,t ) and the amount of wind and solar power abandoned in the system is less than the remaining pumpable capacity of the pumped storage power station: (P w,t -P wr,t +P pv,t -P pvr,t )≤(P hN -P h,t )

[0079] If the pumped storage power station is put into operation at this time, the wind and solar power abandonment rates of the system are:

[0080]

[0081]

[0082] At this time, the system has no abandoned wind and solar power. Although the pumped storage power station has not put all the units into use to pump water, it is still considered that the pumped storage power station has fully utilized its capacity to absorb new energy, that is:

[0083] N N,t =100% (22)

[0084] In this case, the capacity of the pumped storage power station to absorb new energy is:

[0085]

[0086] Condition III: The pumped storage power station operates at less than rated capacity under pumping conditions (P h,t ≤P hN ) The system has wind and solar power curtailment (P w,t ≥Pwr,t ,P pv,t ≥P pvr,t ) and the amount of wind and solar power abandoned in the system is greater than the remaining capacity of the pumped storage power station that can be pumped out of the hydropower station (P w,t -P wr,t +P pv,t -P pvr,t )≥(P hN -P h,t ), the calculation formula for wind and solar power abandonment rates of the pumped storage power station in this case is the same as that of working condition II. At this time, the capacity of the pumped storage power station to absorb new energy is:

[0087]

[0088] Objective function I: Minimize the variance of net load variance within a day

[0089] In order to maximize the role of the power generation / pumping conditions of the pumped storage unit in "peak shaving and valley filling", the objective function is set to minimize the variance of the net load variance within a day, as shown in the following formula:

[0090]

[0091] Where: P L,t is the load value in period t; The power generation capacity of the pumped storage power station; Pumping power of a pumped-storage power station.

[0092] Objective function II: operating cost of the joint operation system

[0093] The operating costs of the joint operation system include two parts:

[0094] minF=F SO +F PS (26)

[0095] Where: F SO F is the operating cost of thermal power units; PS Operating costs of pumped storage power stations.

[0096] (1) The operating cost of a thermal power unit is:

[0097]

[0098] a j , b j and c j They correspond to the coal consumption coefficient of thermal power units; P G,j,t is the output value of the thermal power unit; S jt is the start-up and shutdown cost of thermal power unit j; u j,t It is the start and stop status of the thermal power unit.

[0099] (2) Operation costs of pumped storage power stations

[0100] The electricity revenue generated by the unit during the daily operation cycle of the pumped storage power station must be greater than the cost of the unit's pumping electricity consumption. The pumping cost of the power station during the period T is:

[0101]

[0102] Start-up and shutdown loss cost of pumped storage units:

[0103]

[0104] Where: and are Boolean variables indicating that the unit is in power generation and pumping conditions respectively; C h The loss cost of starting and shutting down the unit.

[0105] The operating cost of a pumped storage power station is:

[0106]

[0107] Constraints of the units in the system model:

[0108] (1) Power balance constraints

[0109]

[0110] Where: P G,i,t for t The output of thermal power unit i at the moment; P wr,t is the wind power consumption of the system; P pvr,t is the power consumption of the photovoltaic system;

[0111] (2) Constraints on thermal power units

[0112] ① Output constraints of thermal power units:

[0113]

[0114] ② Thermal power unit climbing constraints:

[0115] -r d,j Δt≤P G,j,t -P G,j,t-1 ≤r u,j Δt (33)

[0116] ③ Minimum start and stop time constraints for thermal power units:

[0117]

[0118] (3) Constraints on the operating characteristics of pumped storage power stations

[0119] ① Unit power constraints:

[0120]

[0121]

[0122] Where: P hg,min ,P hg,max are the upper and lower limits of the power generation capacity of the pumped storage unit respectively; P hp,min ,P hp,max They are the upper and lower limits of the pumping power of the pumped storage unit respectively.

[0123] ② Constraints on single operating conditions of units and power stations:

[0124]

[0125] Y G +Y P ≤1 (38)

[0126] Where: The sum of the two is less than or equal to 1, indicating that unit k can only be in pumping and generating conditions at the same time; Y G ,Y P They represent the power generation and pumping conditions of the pumped-storage power station respectively.

[0127] ③ Constraints on the number of units:

[0128]

[0129]

[0130] ④ Water level and capacity constraints of pumped storage power stations

[0131] In actual operation, the pumped storage unit needs to constrain the reservoir water level for safety reasons. The reservoir water level is controlled at the highest water level and the lowest dead water level of the reservoir, that is, SOPs is recorded as S SOPs To ensure the continuous output of the pumped storage power station, it is necessary to control its S SOPs Keep it good. As follows:

[0132]

[0133]

[0134] Where: and S SOPs The upper and lower limits of E pss,t Indicates the size of the pumped storage capacity at time t; Indicates the adjustable capacity of pumped storage.

[0135] ⑤ Constraints on the maximum number of start and stop times of the unit:

[0136]

[0137] Where: M h It is the maximum number of starts and stops of a single unit in period T.

[0138] Daily load curve time series characteristics and time period clustering:

[0139] (1) Clustering method of Gaussian mixture model

[0140] When using the Gaussian mixture model for clustering, assume that the input samples obey k Gaussian distributions with different parameters, and samples that obey the same parameter distribution are classified into one category. Use the maximum expectation algorithm to fit the k mixed Gaussian distributions to obtain k cluster distributions that traverse all samples and their distribution means and covariances. The specific steps for iterative fitting using the maximum expectation algorithm are as follows:

[0141] ① Input sample x b (b=1,2,…,n), set the parameters μ of k multivariate Gaussian distributions a and ε a , where n is the total number of samples.

[0142] ② Calculate the probability γ(a,b) that the sample point belongs to that Gaussian distribution:

[0143]

[0144] Where: T is the scheduling period; p(x b |ξ b =a) represents x b Category b is the probability function of a; d is the sample dimension.

[0145] ③ According to the maximum likelihood estimation, μ of each Gaussian distribution can be obtained a and μ a The replacement value μ′ a and ε′ a as follows:

[0146]

[0147] ④ Repeat the above steps until all parameters converge, and use the iteratively obtained Gaussian parameters to cluster the samples based on the maximum probability γ(a,b).

[0148] (2) Establishing daily load time series characteristics and time period classification model

[0149] According to the time series characteristics of daily load, time period clustering is carried out to conduct static analysis between each time period, and the average change rate S of the daily load time period is used. b,ave and standard deviation S b,std Characterizes the magnitude and severity of daily load changes in time series, S b,ave and S b,std The larger the value, the greater the change in daily load and the more drastic the degree. The average value f of the frequency of daily load changes in the direction within adjacent time periods is used. b,ave and standard deviation f b,std Characterizes the frequency and trend of daily load changes in direction, where f b,ave and f b,std The larger the value is, the less obvious the change trend of daily load is. That is, the input sample of the period characteristic of daily load curve is:

[0150] x b =[S b,ave ,S b,std ,f b,ave ,f b,std ] (46)

[0151] Set the daily load change rate mean, daily load change frequency mean, daily load change rate standard deviation judgment threshold, and daily load change frequency standard deviation judgment threshold to That is, each parameter in the sample has two situations: greater than or less than the threshold. If x in formula (39) is b Clustering can generate 16 types, but too many types will increase the complexity of the model, so the model is simplified. b,ave The magnitude of the change represented by the parameter represents the severity of the change to a certain extent. b,ave When S is greater than the threshold b,std Usually it is also greater than the threshold, and f b,ave and f b,std It also describes the changing trend of daily load from different angles. When analyzing the daily load characteristic curve, the two are also combined. Therefore, the present invention simplifies 16 types into 4 types, among which the change amplitude of daily load limits the influence of its change frequency on the stability of daily load. Under the premise, regardless of f b,ave ,f b,std Whether it is greater than the threshold, the daily load change degree is low. Finally, three types of daily load are obtained: low-precision period (T l ), medium precision period (T m ) and high-precision time period (T h ) are shown in the following table.

[0152] Table 1 The basis for the division of daily load curves according to the rate of change and frequency of change

[0153]

[0154] Pumped storage power stations participate in the new energy consumption strategy

[0155] In order to give full play to the role of pumped-storage units in promoting the consumption of renewable energy in new power systems and ensure that the average output rate of the wind-solar-thermal-pumped-storage combined operation system can effectively track the changes in the daily load rate of the power grid, a two-layer model of the strategy for accommodating renewable energy by pumped-storage units is established based on the clustering results of daily load curves.

[0156] Inner model: It is necessary to meet the operating constraints of the pumped storage power station and adjust the power of the pumped storage power station through demand response assistance;

[0157] Outer model: The optimization goal is to improve the peak-to-valley fluctuation of net load by pumped-storage units and minimize the comprehensive operating cost of the joint operation system. Three strategies for three types of pumped-storage units to participate in regulating new energy consumption are formulated for the new energy consumption capacity model.

[0158] (1) Full tracking strategy

[0159] The output power of the pumped storage unit completely tracks the daily load rate characteristic curve of the power grid:

[0160]

[0161] Where: P N is the operating capacity of the thermal power unit; P s (i) is the output of the pumped storage unit at time i; P s1 The minimum capacity of the pumped storage unit. p(i) is the system load rate at time i.

[0162] (2) Coordinated tracking strategy

[0163] The daily load curve of the power grid is divided into three periods: peak, flat and valley. The pumped storage unit assists the thermal power unit in tracking the daily load curve of the power grid:

[0164]

[0165] Where: p f ,p p ,p g are the average load rates of the three periods of peak, flat and valley respectively; P sf ,P sp ,P sg They are the pumped storage output in three periods: peak, flat and valley.

[0166] (3) Incomplete tracking strategy

[0167] The pumping capacity and power generation capacity of the pumped storage unit are both at their full capacity. The pumped storage unit generates electricity when the grid load is peak and pumps water when the load is low. The daily pumping power and total power generation are balanced:

[0168]

[0169] Where: the right side is the daily average output rate of the thermal-storage combined unit; γ is the equivalent daily average load rate of the power grid; t f ,t p They are the daily power generation hours and daily pumping hours of the pumped storage unit, respectively.

[0170] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. An optimization method for regulating pumped storage units to promote the consumption of new energy, characterized in that: The following steps are involved: Taking the objective function of improving the grid load optimally by pumped storage power stations, i.e. minimizing the net load variance within the dispatching cycle and minimizing the comprehensive operating cost, and considering the operating constraints of each unit in the power system, a model for promoting the consumption of new energy by fine-tuning pumped storage power stations is established. According to the establishment of a model for promoting the consumption of new energy through refined regulation of pumped storage power stations, the peak regulation capacity of pumped storage units under pumping conditions is extended to the consumption capacity of pumped storage units for new energy, and the utilization of new energy by pumped storage units is divided into four working conditions; The clustering method of Gaussian mixture model is used for daily load curve to establish the daily load time series characteristics and time period classification model; including: Use Gaussian mixture model for clustering, assuming that the input samples obey k Gaussian distributions with different parameters are used to obtain k A mixed Gaussian distribution is fitted to obtain the k The mean and covariance of the cluster distribution and its distribution are used to calculate the probability The samples are clustered based on the maximum value; According to the time series characteristics of daily load, time periods are clustered and the load of each time period is analyzed. The average rate of change within the daily load period is used to calculate the load. and standard deviation Characterizes the magnitude and severity of daily load changes in time series. and The larger the value, the greater the change in daily load and the more drastic the degree; the average value of the frequency of daily load changes in the direction within adjacent time periods. and standard deviation Characterizes the frequency and trend of daily load changes in direction, where and The larger the value, the less obvious the change trend of daily load is. The input sample of the period characteristics of daily load curve is: (1) Set the daily load change rate mean judgment threshold, daily load change frequency mean judgment threshold, daily load change rate standard deviation judgment threshold, and daily load change frequency standard deviation judgment threshold to Each parameter in the sample has two conditions: greater than or equal to or less than the threshold. Clustering generates 16 types, which are then simplified into 4 types. The change amplitude of daily load limits the influence of its change frequency on the stability of daily load. Under the premise, regardless of Whether it is greater than the threshold, the daily load change degree is low, and then three types of daily load are obtained: low-precision period , medium precision period and high-precision time periods , the daily load time series characteristics and time period classification model are as follows: ① When the mean and standard deviation of the daily load curve change rate are both less than the set judgment threshold, that is, The daily load curve shows a stable trend, with a low variation range and low complexity, which is defined as a low-precision period. , the day-ahead scheduling plan takes 1h as the time scale; ② The daily load curve change rate is greater than or equal to the judgment threshold, and the change frequency is less than the judgment threshold, that is, The daily load curve of the medium-precision period has a moderate complexity and does not change frequently in direction, but shows a significant upward or downward trend. , the day-ahead scheduling plan uses 30 minutes as the time scale; ③ The mean and standard deviation of the daily load curve change rate and frequency are both greater than or equal to the judgment threshold, that is, The daily load curve changes frequently in direction, with large changes and unclear trends, and is highly complex, so it is defined as a high-precision period. ,The day-ahead dispatch plan adopts 15 minutes as the time scale, which can more accurately fit the daily load characteristics; During this period, the daily load fluctuation event is smoothed by the pumped storage power station, and the thermal power plant adopts a constant power generation mode, as shown in the following formula: (6) Where: for Moment Output of thermal power units; Indicates the end time point of this period; During this period, the daily load demand changes greatly and shows an obvious increase or decrease trend. The system needs to have a large capacity climbing capability at this time. The thermal power unit adopts the conventional mode and operates in conjunction with the pumped storage power station to meet the load demand and climbing demand. The pumped storage unit performs a part of the sharing. The output of the thermal power unit is: (7) During this period, although the daily load changes frequently in direction and with a large amplitude, there is no obvious increase or decrease trend, and the system does not need to have a large-capacity climbing capability. The fluctuation of daily load is smoothed by the pumped storage power station, and the thermal power unit adopts a constant power generation mode; When the storage capacity cannot meet When the constraint is met, the pumped storage unit cannot work. During this period, the pumped storage unit cannot work. In order to meet the daily load demand, the system needs to abandon wind and solar power or cut load. Based on the above time-division control, set exist The interval is the adjustable storage capacity allowed by the reservoir, among which and The minimum and maximum values ​​of the reservoir water level can be adjusted. Maintain the water level within normal storage capacity; Based on the four working conditions, three strategies are formulated for pumped storage power station regulation involving pumped storage units in the consumption of new energy; According to the daily load time series characteristics and time period classification model, for the low-precision time periods, medium-precision time periods and high-precision time periods of the daily load curve, based on the predicted output of new energy and the daily load characteristic curve, under the premise of meeting the reservoir capacity and water level constraints, the three strategies involving pumped-storage units in the consumption of new energy are used to adjust the pumped-storage units to achieve the consumption of new energy through variable time period control.

2. The optimization method for regulating pumped storage units to promote new energy consumption as claimed in claim 1, characterized in that: The model of promoting new energy consumption through refined regulation of pumped storage power stations is a two-layer model. Inner model: It is necessary to meet the operating constraints of the pumped storage power station and adjust the power of the pumped storage power station through demand response assistance; Outer model: The optimization goal is to improve the peak-to-valley fluctuation of net load by pumped-storage units and minimize the comprehensive operating cost of the joint operation system. Three strategies for pumped-storage units to participate in regulating new energy consumption are formulated for the new energy consumption capacity model.

3. The optimization method for regulating pumped storage units to promote new energy consumption as claimed in claim 2, characterized in that: The three strategies involving pumped storage units in the consumption of new energy are as follows: ① Full tracking strategy: The output power of the pumped storage unit fully tracks the daily load curve of the power grid; ② Coordinated tracking strategy: Divide the daily load curve of the power grid into three periods: peak, flat and valley. Pumped storage units assist thermal power units in storing combined units to track the daily load curve of the power grid. ③ Incomplete tracking strategy: The pumped storage unit generates electricity when the grid load is peak and pumps water when the load is low, so that the daily pumping power and total power generation are balanced.

4. The optimization method for regulating pumped storage units to promote new energy consumption as claimed in claim 1, characterized in that: The four working conditions are: ① The pumped storage unit operates at rated power under pumping conditions, and the pumped storage power station fully utilizes its capacity to absorb new energy; ② The pumped storage units operate at less than rated power under pumping conditions, the system does not abandon wind and solar power, the pumped storage power station does not put all units into operation to pump water, and the pumped storage power station has fully utilized its capacity to absorb new energy; ③ When the pumped storage unit operates at rated power under pumping conditions, there is wind and solar power abandonment in the system, and the amount of wind and solar power abandonment in the system is less than the remaining pumpable capacity of the pumped storage power station. The capacity of the pumped storage power station to absorb new energy is: (2) ④ The pumped storage power station operates at less than rated power under pumping conditions, and there is wind and solar power abandonment in the system. The amount of wind and solar power abandonment in the system is greater than the remaining pumpable capacity of the pumped storage power station. The capacity of the pumped storage power station to absorb new energy is: (3) is the predicted power of the system wind power and is the wind power consumption of the system; is the predicted power of the system photovoltaic and is the photovoltaic power consumption of the system.

5. The optimization method for regulating pumped storage units to promote new energy consumption as claimed in claim 1, characterized in that: The objective function includes the variance of net load variance within a day and the comprehensive operating cost, as follows: Objective function I: Minimize the variance of net load variance within a day (4) Objective function II: Minimize the operating cost of the joint operation system (5) in: For joint operation system t Net load value of the time period; The power generation capacity of the pumped storage power station; is the pumping power of the pumped storage power station; The operating cost of thermal power units; The operating costs of pumped storage power stations; and They correspond to the coal consumption coefficients of thermal power units respectively; is the output value of the thermal power unit; For thermal power units The start-stop costs; It is the start and stop status of the thermal power unit; and They are Boolean variables indicating that the unit is in power generation condition and pumping condition respectively; The loss cost of starting and shutting down the unit.

6. The optimization method for regulating pumped storage units to promote new energy consumption as claimed in claim 1, characterized in that: The power system constraints include power balance constraints, thermal power unit output constraints, thermal power unit climbing constraints, thermal power unit minimum start and stop time constraints, pumped storage unit power constraints, pumped storage unit and single operating condition constraints, pumped storage unit working unit number constraints, pumped storage power station water level and capacity constraints and unit maximum start and stop times constraints.

7. An optimization configuration system based on pumped storage unit regulation to promote new energy consumption, characterized in that: The optimization method for realizing the regulation of pumped storage units to promote the consumption of new energy as described in any one of claims 1 to 6 comprises a consumption model construction module, a classification module, a clustering module, a strategy acquisition module and a curve acquisition module: The consumption model construction module is used to take the pumped storage power station to improve the optimal grid load, that is, the minimum net load variance within the dispatch cycle and the lowest comprehensive operating cost as the objective function, consider the constraints of the operation of each unit in the power system, and establish a pumped storage power station fine regulation to promote the consumption of new energy models; The classification module is used to establish a model for fine-tuning pumped storage power stations to promote the consumption of new energy, extend the peak-shaving capacity of pumped storage units under pumping conditions to the consumption capacity of pumped storage units for new energy, and divide the use of new energy by pumped storage units into four conditions; The clustering module is used to cluster the daily load curve using the Gaussian mixture model to establish a daily load time series characteristic and time period classification model; The strategy acquisition module is used to formulate three strategies for the pumped storage power station to participate in the consumption of new energy sources based on the four working conditions; The curve acquisition module is used to adjust the pumped storage units for variable time period control based on the predicted output of new energy and the daily load characteristic curve during the low-precision, medium-precision and high-precision periods of the daily load curve according to the daily load time series characteristics and time period classification model, and on the premise of meeting the reservoir capacity and water level constraints, according to the three strategies involving pumped storage units in the consumption of new energy, so as to realize the consumption of new energy.

Images