Inertia characteristic-considered energy management method for wind storage combined system

Abstract

A wind storage combined system energy management method considering inertia characteristics comprises the following steps: establishing a hybrid energy storage system model composed of a battery and a flywheel energy storage system, and calculating the power of each energy storage system; according to the operation limitation of the energy storage system, constraint conditions are given; predicting the wind power of the next day based on a multi-layer perceptron neural network to obtain a mean value and a standard deviation; generating a wind power uncertainty model to obtain a predicted wind power scene, and reducing the predicted wind power scene by using a clustering algorithm to obtain a mean value and a standard deviation of wind power prediction power under the reduced scene; obtaining the next-day delivery power and the charging and discharging power of the hybrid energy storage system by using the day-ahead market model; and the charging and discharging power is adjusted. The method can reduce the influence of randomness and uncertainty of traditional wind power generation on the power market, improves the demand response of a wind power generation system to the power market such as the day-ahead market, and improves the profitability of a wind power plant in the power market.

Description

Technical Field

[0001] The method relates to a management method for a wind power generation energy storage system, in particular to an energy management method for a wind power generation energy storage combined system taking inertia characteristics into consideration. Background Art

[0002] With the development of the times, the large-scale use of fossil energy, environmental pollution, climate change and other problems have become increasingly prominent. In addition, due to the continuous growth of energy demand, the demand for electricity has also been increasing. The use of renewable energy such as wind and solar energy and its integration into the power grid is crucial to solving environmental and energy problems. In areas with abundant wind resources, large-scale wind power generation systems require less area than other renewable energy systems, such as photovoltaic systems. Compared with other renewable energy generation technologies, wind power generation has technical and economic advantages. Therefore, wind power generation systems have been widely used, and wind power generation output plays an important role in the power market. However, wind speed is intermittent and uncertain, and the output power of wind turbines fluctuates randomly. With the increase in wind power penetration and the development of large-scale wind turbines, this random and uncertain energy grid connection will bring many problems, such as frequency changes, voltage flicker, causing power system instability and affecting grid reliability. Therefore, it is necessary to control the energy output of renewable energy.

[0003] In order to control the energy output of renewable energy and reduce the difficulty of its grid connection, wind power plants need to participate in the day-ahead (DA) market and the real-time (RT) market in the electricity market. Due to the intermittent nature of wind speed, wind power plants face errors in power generation forecasts when participating in the electricity market, especially over a longer period of time. In this case, system operators may penalize them. Therefore, the accuracy of wind power forecasts will affect the profits of wind power plants. Power plants need to purchase excess electricity from other participants in regulation markets such as the real-time (RT) market as an upward adjustment, and sell excess electricity to other participants as a downward adjustment to compensate for the deviation between published output and actual output.

[0004] As the application of renewable energy in the power grid increases, the demand for ancillary services will also increase to ensure the reliability and security of the power grid. Existing studies have provided solutions, such as more accurate wind speed or power forecasts, system operators re-dispatching controllable generators or flexible loads, and using energy storage systems (ESS) to reduce uncertainties in wind power forecasts. Through these methods, the operating costs of wind power plants can be reduced and the economic benefits of wind power plants can be improved. In wind power systems, in addition to smoothing and compensating for wind power fluctuations, imbalances caused by uncertainty and prediction errors in wind output power can also be compensated by charging and discharging energy storage devices. Therefore, some studies have considered the operating strategies and power bidding of wind power systems equipped with energy storage systems (ESS) in the power market.

[0005] Existing studies have considered the day-ahead dispatch (DA) and other power markets involving renewable energy sources, including wind power. One of the main concerns of power systems with intermittent renewable energy sources is how to deal with energy deviations in the balancing market. In existing studies, the uncertainty of renewable energy sources is taken into account, and optimization methods such as stochastic programming or mixed integer linear programming are used to dispatch renewable energy sources on the day-ahead. However, when modeling day-ahead dispatch energy management, power exchange with the real-time (RT) market is not considered. In addition, some studies consider both the DA market and the RT market involving renewable energy sources, propose a two-stage optimal bidding, and use stochastic programming methods to deal with the uncertainty of the day-ahead market. Other studies consider the DA and RT markets separately.

[0006] Disadvantages of existing methods:

[0007] 1) Due to its inherent randomness and uncertainty, traditional wind power generation faces challenges in market forecasting, which in turn leads to forecasting errors. This error directly affects the economic benefits of wind power projects and reduces their profitability.

[0008] 2) There is little research on energy storage systems in wind power plants, which results in wind power generation weakening the peak-shaving capacity of the power system and failing to compensate for market prices in a timely manner. This ultimately leads to high dispatch costs and high market volatility risks. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide an energy management method for a wind-storage combined system taking into account inertia characteristics, which can reduce the impact of the randomness and uncertainty of traditional wind power generation on the electricity market, improve the demand response of the wind power generation system to the electricity market such as the day-ahead market, and improve the profitability of wind power plants in the electricity market.

[0010] The present invention relates to an energy management method for a wind-storage combined system taking into account inertia characteristics, comprising:

[0011] S1. Establish a hybrid energy storage system model consisting of a battery energy storage system and a flywheel energy storage system for wind power, and calculate the real-time power of the battery energy storage system in the i-th hour Real-time power of the flywheel energy storage system in the first hour Real-time charging power of the battery energy storage system in the i-th hour and the discharge power in the i-th hour The power lacking in the battery energy storage system is supplemented by the flywheel energy storage system;

[0012] S2. According to the operating limitations of the battery energy storage system and the flywheel energy storage system, the constraints of the state of charge, energy value and charge and discharge power of the battery energy storage system and the flywheel energy storage system are given;

[0013] S3. Predict the wind power generated the next day based on the multi-layer perceptron neural network to obtain the wind power data MLP model, use the model to obtain the predicted wind power data set, and fit the normal distribution function to obtain the mean and standard deviation of the predicted wind power; and use the same method to obtain the predicted day-ahead market price data;

[0014] S4. Generate a wind power uncertainty model based on the mean and standard deviation of the predicted wind power obtained in S3, obtain predicted wind power scenarios under different probabilities, and use a clustering algorithm to reduce the predicted wind power scenarios to obtain the mean and standard deviation of the predicted wind power under the reduced scenarios;

[0015] S5. Using the day-ahead market model, the model input is the predicted day-ahead market price data and the predicted wind power under the curtailment scenario, as well as the constraints, and the model output is the power delivered by the hybrid energy storage system to the operator the next day and the charging and discharging power of the battery energy storage system; and adjusting the charging power and discharging power of the battery energy storage system obtained in S1 according to the output of the model;

[0016] S6. Based on the profit obtained in the day-ahead market calculated based on the power delivered to the operator, a mixed integer linear programming model is constructed with the goal of maximizing the day-ahead market profit to determine the day-ahead optimal dispatch of the battery energy storage system.

[0017] As a further preferred embodiment, the hourly power of the battery energy storage system in S1 is The calculation formula is as follows:

[0018]

[0019] In the above formula, To limit wind power, it is determined by formula (2); is the power delivered to the operator, which can be determined by formula (3);

[0020] The wind power limit The formula is as follows:

[0021]

[0022] The power delivered to the operator The formula is as follows:

[0023]

[0024] In the above formula, μ i is the mean of the predicted wind power per hour, σ i is the standard deviation of the hourly forecast wind power.

[0025] As a further preferred embodiment, the i-th hour power of the flywheel energy storage system in S1 The calculation formula is as follows:

[0026]

[0027] In the above formula: represents the maximum wind power generation limit when the battery energy storage system is not charged, represents the minimum wind power generation limit for charging the battery energy storage system, Indicates the wind power generation per hour.

[0028] As a further preferred embodiment, the charging power of the battery energy storage system in the i-th hour in S1 is and the discharge power in the i-th hour The calculation formula is as follows:

[0029]

[0030] In the above formula: ph i is the charging or discharging efficiency of the battery energy storage system in the i-th hour.

[0031] As a further preferred embodiment, the energy value stored by the battery energy storage system and the flywheel energy storage system in S2 in the i-th hour is determined by the following formula:

[0032]

[0033] In the above formula: is the energy value stored by energy storage system j at the beginning of the i-th hour; is the remaining energy value of energy storage system j at hour i; is the charging power of energy storage system j at the i-th hour; is the discharge power of energy storage system j at the i-th hour; η cj is the charging efficiency of energy storage system j; η dj is the discharge efficiency of energy storage system j; Δt is the discrete sampling time length, which is determined by the type of energy storage system used by the user and the energy value sampling time length of the energy storage system within one hour; T d is the scheduling interval, which is 1 hour; i represents the number of hours; is the energy value of energy storage system j at hour i-1, is the energy value of energy storage system j at hour i. These two energy values ​​are both 0 at the beginning and end of the day, while in other time periods of the day, the energy value of energy storage system j is determined by the charging and discharging power of the energy storage system obtained by the day-ahead market model. The energy storage system j is a battery energy storage system or a flywheel energy storage system. When j=1, it indicates a battery energy storage system; when j=2, it indicates a flywheel energy storage system.

[0034] As a further preferred embodiment, the charge state of the battery energy storage system and the flywheel energy storage system in S2 Determined by the following formula:

[0035]

[0036] In the above formula: Indicates the initial state of charge of the battery energy storage system or flywheel energy storage system, which is determined by the user; is the nominal energy of the battery or flywheel energy storage system, determined by the type of energy storage system selected by the user; is the energy value of energy storage system j at the beginning of the i-th hour, determined by formula (6).

[0037] As a further preference, the constraint condition of the state of charge of the battery or flywheel energy storage system in S2 should be maintained between the maximum state of charge and the minimum state of charge, and the formula is:

[0038]

[0039] in, is the minimum state of charge of energy storage system j, is the maximum state of charge of energy storage system j, is the state of charge of energy storage system j at the i-th hour; energy storage system j is a battery energy storage system or a flywheel energy storage system. When j=1, it represents a battery energy storage system; when j=2, it represents a flywheel energy storage system;

[0040] The energy value constraint condition of the battery or flywheel energy storage system in S2 is that the energy of energy storage system j is not less than its nominal energy, and the formula is:

[0041]

[0042] Where E Sj is the nominal energy of energy storage system j;

[0043] The constraint condition of the charging power of the battery or flywheel energy storage system in S2 is that its charging power is not less than its nominal power, and the formula is:

[0044]

[0045] Where P Sj is the nominal power of energy storage system j;

[0046] The discharge power constraint condition of the battery or flywheel energy storage system in S2 is that its discharge power should not be higher than its nominal power, and the formula is:

[0047]

[0048] This method has the following advantages:

[0049] 1) The present invention relates to an energy management method for a wind-storage combined system taking into account inertia characteristics, which can reduce the impact of the randomness and uncertainty of traditional wind power generation on the power market, improve the demand response of the wind power generation system to the power market such as the day-ahead market, and manage and dispatch the hybrid energy storage system model from the perspective of maximizing the day-ahead market profit of the wind power plant, thereby improving the profitability of the wind power plant in the power market.

[0050] 2) This method can improve the profit of the hybrid energy storage system and ensure the stability of the energy storage system. The two designed energy storage systems can dispatch the battery energy storage system to sell electricity in the day-ahead market, and can stabilize the power of the battery energy storage system, that is, guarantee the power of the units with long-term contracts to the day-ahead market suppliers, so that the wind power plant can obtain better economic benefits.

[0051] 3) This method can reduce the demand for wind power generation on the peak-shaving capacity of the power system, improve the output stability of wind turbines, reduce the dispatching cost of wind power, and thus reduce the impact on the power market. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flow chart of the method.

[0053] Figure 2 It is a power and energy variation diagram of the battery energy storage system described in this method.

[0054] Figure 3 It is the fitted wind power scenario diagram.

[0055] Figure 4 is a flow chart of the day-ahead market planning model used in this method.

[0056] Figure 5 It is a graph showing the charge state of the energy storage system and the power delivered to the operator during day-ahead dispatch.

[0057] Figure 6 It is a comparison chart of the power delivered to operators with and without the introduction of the present invention. Specific implementation methods

[0059] like Figure 1 As shown, the present invention relates to a wind-storage combined system energy management method considering inertia characteristics, comprising the following steps:

[0060] S1. Establish a hybrid energy storage system model consisting of a battery energy storage system and a flywheel energy storage system for wind power, and calculate the real-time power of the battery energy storage system in the i-th hour Real-time power of the flywheel energy storage system in the first hour Real-time charging power of the battery energy storage system in the i-th hour and the discharge power in the i-th hour The power lacking in the battery energy storage system is supplemented by the flywheel energy storage system; the complementary outputs of the battery energy storage system and the flywheel energy storage system are used to manage wind power to balance wind power generation, report electricity quantity, participate in the day-ahead market, and improve the utilization rate of wind power and the profit of wind power plants.

[0061] The i-th hour power of the battery energy storage system The calculation formula is as follows:

[0062]

[0063] In the above formula, To limit wind power, it is determined by formula (2); is the power that the wind power plant is scheduled to deliver to the operator at the i-th hour, which can be determined by formula (3);

[0064] The wind power limit The formula is as follows:

[0065]

[0066] The power delivered to the operator The formula is as follows:

[0067]

[0068] In the above formula, μ i is the mean of the predicted wind power per hour, σ i is the standard deviation of the hourly forecast wind power.

[0069] The i-th hour power of the flywheel energy storage system The calculation formula is as follows:

[0070]

[0071] In the above formula: represents the maximum wind power generation limit when the battery energy storage system is not charged, represents the minimum wind power generation limit for charging the battery energy storage system, Indicates the wind power generation per hour.

[0072] The i-th hour charging power of the battery energy storage system and the discharge power in the i-th hour The calculation formula is as follows:

[0073]

[0074] In the above formula: ph iis the charging or discharging efficiency of the battery energy storage system in the i-th hour.

[0075] In actual use, the battery energy storage system can select a lithium battery energy storage system. The complementary flywheel energy storage system needs to have the characteristics of fast response and large discharge power, so as to achieve energy complementarity with the battery energy storage system in a short time, so a flywheel energy storage system or a supercapacitor is used. In the application process of the present invention, the flywheel energy storage system is regarded as a general energy storage system, which is the same as the model of the battery energy storage system, so as to simplify the process of solving the calculation.

[0076] In the present invention, according to the charging or discharging strategy of the battery energy storage system, power bidding is conducted in the day-ahead market. By predicting the electricity price in the day-ahead market, the charge and discharge cycle of the battery energy storage system and its initial SoC are optimally scheduled to maximize profits in the day-ahead market. The discharge time of the battery energy storage system is scheduled during the peak price period, which not only utilizes the arbitrage ability of the energy storage system to compensate for the imbalance in the day-ahead power dispatching, but also utilizes the compensation ability of the energy storage system. In addition, in real-time operation, if the battery energy storage system is fully charged or discharged before the scheduled time, the flywheel energy storage system will compensate for the imbalance caused by the prediction error. Therefore, no error will occur in the power delivery plan announced by the day-ahead market and the charge and discharge cycle predetermined by the battery energy storage system. In the real-time operation strategy, if necessary and within a limited time, power will be exchanged with the hourly market and the real-time market to operate the flywheel energy storage system to compensate for the prediction error. Taking the first 6 hours of a day as an example, the power value of the battery energy storage system calculated by the above formula (5) is shown in the following table. Since the battery energy storage system is dispatched once an hour, its charging or discharging power is 0 at the beginning or end of each hour. In the first 6 hours, the power of the battery energy storage system is negative and is in a charging state. At this time, the actual limited wind power is greater than the power delivered to the operator, and the excess power is absorbed by the battery energy storage system.

[0077]

[0078] S2. According to the operating limitations of the battery energy storage system and the flywheel energy storage system, the constraints on the state of charge, energy value and charging and discharging power of the battery energy storage system and the flywheel energy storage system are given.

[0079]

[0080] In the above formula: is the energy value stored by energy storage system j at the beginning of the i-th hour; is the remaining energy value of energy storage system j at hour i; is the charging power of energy storage system j at the i-th hour; is the discharge power of energy storage system j at the i-th hour; ηcj is the charging efficiency of energy storage system j; η dj is the discharge efficiency of energy storage system j; Δt is the discrete sampling time length, which is determined by the type of energy storage system used by the user and the energy value sampling time length of the energy storage system within one hour; T d is the scheduling interval, which is 1 hour; i represents the number of hours; is the energy value of energy storage system j at hour i-1, is the energy value of energy storage system j at hour i. These two energy values ​​are both 0 at the beginning and end of the day, while in other time periods of the day, the energy value of energy storage system j is determined by the charging and discharging power of the energy storage system obtained by the day-ahead market model. The energy storage system j is a battery energy storage system or a flywheel energy storage system. When j=1, it indicates a battery energy storage system; when j=2, it indicates a flywheel energy storage system.

[0081] Taking some moments of a day as an example, the energy value changes of the battery energy storage system calculated by equations (6) and (7) are shown in the following table. In the first 6 hours, the absolute value of the energy of the battery energy storage system gradually increases, indicating that the battery energy storage system is in a charging state, which is the same as the change in its power. At this time, the battery energy storage system compensates for the difference between the power delivered to the operator and the actual limited wind power.

[0082]

[0083] The state of charge of the battery energy storage system and the flywheel energy storage system Determined by the following formula:

[0084]

[0085] In the above formula: Indicates the initial state of charge of the battery energy storage system or flywheel energy storage system, which is determined by the user; is the nominal energy of the battery or flywheel energy storage system, determined by the type of energy storage system selected by the user; is the energy value of energy storage system j at the beginning of the i-th hour, determined by formula (6).

[0086] The state of charge constraint of the battery or flywheel energy storage system should be maintained between the maximum state of charge and the minimum state of charge, and the formula is:

[0087]

[0088] in, is the minimum state of charge of energy storage system j, is the maximum state of charge of energy storage system j, is the state of charge of energy storage system j at the i-th hour; energy storage system j is a battery energy storage system or a flywheel energy storage system. When j=1, it represents a battery energy storage system; when j=2, it represents a flywheel energy storage system;

[0089] The energy value constraint condition of the battery or flywheel energy storage system is that the energy of energy storage system j is not less than its nominal energy, and its formula is:

[0090]

[0091] Where E Sj is the nominal energy of energy storage system j;

[0092] The constraint condition of the charging power of the battery or flywheel energy storage system in S2 is that its charging power is not less than its nominal power, and its formula is:

[0093]

[0094] Where P Sj is the nominal power of energy storage system j;

[0095] The discharge power constraint condition of the battery or flywheel energy storage system is that its discharge power should not be higher than its nominal power, and its formula is:

[0096]

[0097] S3. Based on the multi-layer perceptron neural network, the wind power generated on the next day is predicted to obtain the MLP model of wind power data. The model is used to obtain the predicted wind power data set, and the normal distribution function is fitted to obtain the mean and standard deviation of the wind power prediction. The same method is used to obtain the predicted day-ahead market price data. The specific steps are:

[0098] First, input the wind power data set into the multi-layer perceptron neural network, then preprocess the wind power data set and delete the outliers in the data set. Then, construct the MLP model of wind power data, and compile and train the constructed MLP model. Next, use the trained wind power data MLP model to predict the wind power generated the next day to obtain the predicted wind power data set. Finally, using the model evaluation function in the neural network, the loaded wind power data set and the predicted wind power data set generated by the wind power data MLP model can be evaluated and analyzed, and the generated predicted wind power data set can be fitted with the normal distribution function to obtain the mean μ and standard deviation σ of the wind power prediction.

[0099] S4. Generate a wind power uncertainty model based on the mean μ and standard deviation σ obtained in S3, obtain predicted wind power scenarios under different probabilities, and use a clustering algorithm to reduce the predicted wind power scenarios to obtain the mean μ and standard deviation σ of the predicted wind power under the reduction scenario. The specific operations are:

[0100] First, the probability density function (PDF) of the predicted wind power data fitted by S3 is discretized and divided into several intervals with a length equal to the standard deviation value of the prediction error. The area of ​​each interval represents the probability of all error values ​​occurring in the interval. In order to simplify the calculation, a seven-step distribution (0, ±σ, ±2σ, ±3σ) is usually used, so that the uncertainty model used can cover 99% of the prediction uncertainty. Secondly, a roulette mechanism is used to select an interval from the seven intervals of the probability density function for modeling. The present invention uses the Monte Carlo method to generate a random value of the prediction error in the range of 0-1 in an interval randomly selected by the roulette mechanism. Therefore, according to the interval selected by the roulette wheel, the generated wind power prediction error is determined with a certain normalized probability. The wind power uncertainty model is generated by using different normalized probabilities to generate predicted wind power scenarios.

[0101] Then, in order to reduce the computational complexity, a clustering algorithm is used to reduce the scenarios, and the clustering algorithm can select scenarios with obvious features and a high probability of occurrence from the original scenarios. The clustering algorithm uses the K-center point algorithm to reduce the scenarios as an example.

[0102] The specific steps of using clustering algorithm to reduce scenarios are as follows: (1) Randomly select r scenarios from the original scenario samples as cluster centers, which can be recorded as (2) According to the principle of the shortest distance between scenarios, the original scenarios except the cluster center are divided into various categories. (3) According to the above formula, a new cluster center is found to replace the current cluster center. (4) Determine whether it converges. If it does not converge, return to step (2) and recalculate. (5) After completing the scenario clustering, the cluster center J is obtained. 1 ,J 2 ,…,J r That is, the reduced scenario, and the probability corresponding to the reduced scenario is the ratio of the number of scenarios in its class to the number of original scenarios. Through the above steps, the reduced scenario can be obtained.

[0103] The wind power scenario fitted after reduction is as follows Figure 3 As shown, the mean μ of wind power forecast is 574 and the standard deviation σ is -0.114. Here we get the mean μ of wind power forecast for one hour. i and standard deviation σ i, when the actual wind power output is greater than or less than this value, the limited wind power is calculated by formula (2). The power delivered to the operator every hour is determined by formula (3). The difference between the power delivered to the operator and the actual limited wind power is compensated by the battery energy storage system. The flywheel energy storage system absorbs or generates excess or insufficient wind power relative to the limited wind power to compensate for the wind power value outside the defined range.

[0104] S5. Using a day-ahead market model, the model inputs are the predicted day-ahead market price data and the predicted wind power under the curtailment scenario, as well as the constraints; the model outputs are the power delivered by the hybrid energy storage system to the operator the next day and the charging and discharging power of the battery energy storage system; and the charging and discharging power of the battery energy storage system obtained in S1 are adjusted according to the output of the model.

[0105] The predicted wind power data set and the day-ahead market price data set will be used in a day-ahead market planning model, the input of which is the predicted day-ahead market price data and wind power forecast data as well as the constraints; the output of which is the power delivered by the hybrid energy storage system to the operator the next day and the charging power and discharging power of the battery energy storage system proposed in step S1.

[0106] The day-ahead market model may be selected by the user according to actual conditions. Figure 4 This is a flow chart of the method using the day-ahead market model. Based on this day-ahead market model, the state of charge of the battery and flywheel energy storage system in the day-ahead dispatch and the power delivered to the operator change as shown in the figure. Figure 5 As shown. The two straight lines in the figure represent the constraint values ​​of the state of charge of the two energy storage systems. The minimum state of charge constrained is 0.2, and the maximum state of charge is 0.9. The present invention stipulates that the power and energy of the two energy storage systems when charging are negative, and vice versa. Figure 3 As shown in the figure, when the energy reduction of the battery energy storage system is negative with its power, the corresponding Figure 6 The state of charge of the battery energy storage system increases, and the battery energy storage system is in a charging state; from formula (3), it can be seen that the power delivered by the operator at this time is equal to the minimum value of the limited wind power in the same hour, and the flywheel energy storage system absorbs the actual wind power that is higher than the predicted power, and its power is calculated by formula (4). The battery energy storage system compensates for the deviation in the power delivered by the wind power plant to the operator due to the prediction deviation, and its power is calculated by formula (5). The same applies to other situations.

[0107] The strategy for adjusting the charging power and discharging power of the battery energy storage system obtained in S1 according to the output of the model is as follows: Consider the optimal result of the day-ahead dispatch phase and use it as the input of the hourly market operation strategy. If there is an error in the wind power generation forecast, follow the result of the day-ahead dispatch phase. The flywheel energy storage system compensates for the imbalance in the hourly market according to the following operation strategy.

[0108] During the battery energy storage system charging period, if the battery energy storage system's state of charge reaches The power delivered to the operator the next day and the charging and discharging power of the battery energy storage system output in the day-ahead market model are still implemented. In this case, the flywheel energy storage system will also store wind power that exceeds the power delivered to the operator. The additional power stored by the flywheel energy storage system is sold to the real-time market at a reduced price. Its discharge power and time need to be calculated to meet the constraints of its energy and state of charge values. If the state of charge of the battery energy storage system does not reach the maximum state of charge This method is used to calculate the real-time charging and discharging power of the battery energy storage system and the flywheel energy storage system. During the discharge period of the battery energy storage system, if the state of charge of the battery energy storage system reaches the minimum state of charge The dispatch method of the day-ahead market model is still implemented. Again, the flywheel energy storage system discharges and compensates for the part of the wind power that is lower than the power delivered to the operator. This discharged power is purchased once an hour in the real-time market. When purchasing energy, it is ensured that the state of charge value of the flywheel energy storage system is restored to the initial state of charge SoC 0 To meet the operating restrictions of the flywheel energy storage system. If the state of charge of the battery energy storage system does not reach the minimum state of charge The present invention is used to calculate the real-time charging and discharging power of the battery energy storage system and the flywheel energy storage system.

[0109] To compare the effectiveness of the present invention, the day-ahead market model only inputs the predicted wind power data and the day-ahead market price data, and outputs the power delivered to the operator. Figure 6 As shown, through Figure 6 It can be seen that in the process of scheduling once an hour a day, compared with the wind farm that has not introduced the present invention, the management method of the present invention can increase the power delivered by the wind farm to the operator in most time periods, can more effectively stabilize the power output of the wind farm, reduce the occurrence of wind abandonment, and thus improve the economic benefits of the wind farm.

[0110] Based on the day-ahead market model output, the power delivered by the hybrid energy storage system to the operator and the charging and discharging power of the battery energy storage system, the hourly power and energy changes of the battery energy storage system are calculated as follows: Figure 2As shown. The nominal energy of the selected battery energy storage system model is 700kWh and the nominal power is 500kW, while the nominal energy of the flywheel energy storage system is 200kWh and the nominal power is 400kW. In the process of using the model, this method can ensure that the power and energy of the battery energy storage system vary within the nominal value range, which shows the feasibility of this method in practical application.

[0111] S6. Based on the profit obtained in the day-ahead market calculated based on the power delivered to the operator, a mixed integer linear programming model is constructed to determine the day-ahead optimal dispatch of the battery energy storage system with the goal of maximizing the day-ahead market profit.

[0112] The day-ahead market model is used to obtain the power of the hybrid energy storage system in the day-ahead market, as well as the charging and discharging power of the battery energy storage system, actual wind power, and day-ahead market data. In the day-ahead dispatching stage, the mixed integer linear programming (MILP) model is solved to determine the day-ahead optimal dispatching of the battery energy storage system. The profit obtained in the day-ahead market is calculated as follows (13).

[0113]

[0114] Profit da is the random profit obtained by the market on the day before, π sw represents the probability of scenario sw occurring, the subscript sw represents the scenario index related to wind power forecast error, π sp It represents the probability of scenario sp occurring, and the subscript sp represents the scenario index related to the day-ahead market price forecast error. represents the power delivered to the operator by scenario sw in the i-th hour, represents the day-ahead market price of scenario sp in hour i. NS represents the number of scenarios, and n is the number of hours in a day.

[0115] During the day-ahead dispatch phase, the battery energy storage system is allowed to undergo two complete charge and discharge cycles to improve battery life. i As a binary variable, the number of charge and discharge cycles c is used as an integer variable to construct a mixed integer linear programming (MILP) model. The model maximizes equation (13) as the objective function, and under the constraints of equations (9)-(12), the proposed day-ahead scheduling will be optimal. The mixed integer linear programming model can be solved using MATLAB programs and codes. In addition, commercial solvers such as Gurobi and CPLEX can also be used to solve the proposed MILP optimization model.

[0116] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A wind-storage combined system energy management method considering inertia characteristics, characterized in that Here are the steps: S1. Establish a hybrid energy storage system model consisting of a battery energy storage system and a flywheel energy storage system for wind power, and calculate the real-time power of the battery energy storage system in the i-th hour Real-time power of the flywheel energy storage system in the first hour Real-time charging power of the battery energy storage system in the i-th hour and the discharge power in the i-th hour The power lacking in the battery energy storage system is supplemented by the flywheel energy storage system; S2. According to the operating limitations of the battery energy storage system and the flywheel energy storage system, the constraints of the state of charge, energy value and charge and discharge power of the battery energy storage system and the flywheel energy storage system are given; S3. Predict the wind power generated the next day based on the multi-layer perceptron neural network to obtain the wind power data MLP model, use the model to obtain the predicted wind power data set, and fit the normal distribution function to obtain the mean and standard deviation of the predicted wind power; and use the same method to obtain the predicted day-ahead market price data; S4. Generate a wind power uncertainty model based on the mean and standard deviation of the predicted wind power obtained in S3, obtain predicted wind power scenarios under different probabilities, and use a clustering algorithm to reduce the predicted wind power scenarios to obtain the mean and standard deviation of the predicted wind power under the reduced scenarios; S5. Using the day-ahead market model, the model input is the predicted day-ahead market price data and the predicted wind power under the curtailment scenario, as well as the constraints, and the model output is the power delivered by the hybrid energy storage system to the operator the next day and the charging and discharging power of the battery energy storage system; and adjusting the charging power and discharging power of the battery energy storage system obtained in S1 according to the output of the model; S6. Based on the profit obtained in the day-ahead market calculated based on the power delivered to the operator, a mixed integer linear programming model is constructed with the goal of maximizing the day-ahead market profit to determine the day-ahead optimal dispatch of the battery energy storage system.

2. The energy management method of a wind-storage combined system considering inertia characteristics according to claim 1 is characterized by: The i-th hour power of the battery energy storage system in S1 The calculation formula is as follows: In the above formula, To limit wind power, it is determined by formula (2); is the power delivered to the operator, which can be determined by formula (3); The wind power limit The formula is as follows: The power delivered to the operator The formula is as follows: In the above formula, μ i is the mean of the predicted wind power per hour, σ i is the standard deviation of the hourly forecast wind power.

3. The energy management method of a wind-storage combined system considering inertia characteristics according to claim 1 is characterized in that: The i-th hour power of the flywheel energy storage system in S1 The calculation formula is as follows: In the above formula: represents the maximum wind power generation limit when the battery energy storage system is not charged, represents the minimum wind power generation limit for charging the battery energy storage system, Indicates the wind power generation per hour.

4. The energy management method of a wind-storage combined system considering inertia characteristics according to claim 1 is characterized in that: The charging power of the battery energy storage system in S1 at the i-th hour and the discharge power in the i-th hour The calculation formula is as follows: In the above formula: ph i is the charging or discharging efficiency of the battery energy storage system in the i-th hour.

5. The energy management method of a wind-storage combined system considering inertia characteristics according to claim 1 is characterized in that: The energy value stored by the battery energy storage system and the flywheel energy storage system in S2 in the i-th hour is determined by the following formula: In the above formula: is the energy value stored by energy storage system j at the beginning of the i-th hour; is the remaining energy value of energy storage system j at hour i; is the charging power of energy storage system j at the i-th hour; is the discharge power of energy storage system j at the i-th hour; η cj is the charging efficiency of energy storage system j; η dj is the discharge efficiency of energy storage system j; Δt is the discrete sampling time length, which is determined by the type of energy storage system used by the user and the energy value sampling time length of the energy storage system within one hour; T d is the scheduling interval, which is 1 hour; i represents the number of hours; is the energy value of energy storage system j at hour i-1, is the energy value of energy storage system j at hour i. These two energy values ​​are both 0 at the beginning and end of the day, while in other time periods of the day, the energy value of energy storage system j is determined by the charging and discharging power of the energy storage system obtained by the day-ahead market model. The energy storage system j is a battery energy storage system or a flywheel energy storage system. When j=1, it indicates a battery energy storage system; when j=2, it indicates a flywheel energy storage system.

6. The energy management method of a wind-storage combined system considering inertia characteristics according to claim 1 is characterized by: The state of charge of the battery energy storage system and the flywheel energy storage system in S2 Determined by the following formula: In the above formula: Indicates the initial state of charge of the battery energy storage system or flywheel energy storage system, which is determined by the user; is the nominal energy of the battery or flywheel energy storage system, determined by the type of energy storage system selected by the user; is the energy value of energy storage system j at the beginning of the i-th hour, determined by formula (6).

7. The energy management method of a wind-storage combined system considering inertia characteristics according to claim 1 is characterized by: The constraint condition of the state of charge of the battery or flywheel energy storage system in S2 should be maintained between the maximum state of charge and the minimum state of charge, and the formula is: in, is the minimum state of charge of energy storage system j, is the maximum state of charge of energy storage system j, is the state of charge of energy storage system j at the i-th hour; energy storage system j is a battery energy storage system or a flywheel energy storage system. When j=1, it represents a battery energy storage system; when j=2, it represents a flywheel energy storage system; The energy value constraint condition of the battery or flywheel energy storage system in S2 is that the energy of energy storage system j is not less than its nominal energy, and the formula is: Where E Sj is the nominal energy of energy storage system j; The constraint condition of the charging power of the battery or flywheel energy storage system in S2 is that its charging power is not less than its nominal power, and the formula is: Where P Sj is the nominal power of energy storage system j; The discharge power constraint condition of the battery or flywheel energy storage system in S2 is that its discharge power should not be higher than its nominal power, and the formula is:

Images