Efficient active power allocation method for wind farms taking into account wind turbine fatigue damage

Abstract

The present invention belongs to the field of wind farm grid-connected control, and specifically relates to a method for efficiently distributing active power of a wind farm taking into account the degree of fatigue damage of wind turbines, aiming to solve the problem that the traditional method evenly distributes the dispatching command power to all wind turbines without considering the influence of power regulation on fatigue damage of wind turbines, which leads to an increase in the probability of wind turbine failure. The present invention includes: wind turbine load estimation based on force analysis and energy conservation, low-complexity fatigue damage quantification based on improved energy method, and a multi-objective optimization model based on Gulin method, entropy weight method and improved matter-element extension model with the goal of minimizing the total fatigue load of the wind farm and the most uniform distribution of fatigue loads between wind turbines. The present invention realizes fast and accurate power distribution, enables wind turbines to operate in a suitable state, reduces the accumulation of fatigue damage and the risk of failure, and improves the overall operating efficiency and economic benefits of the wind farm.

Description

Background Art

[0002] With the continuous advancement of technology and the gradual reduction of production costs, wind power generation has rapidly become an important part of the growth of new energy scale. Among them, large-capacity wind turbines have gradually become the mainstream choice for wind power installation due to their high efficiency, excellent low wind speed power generation capacity and small footprint. These large wind turbines can not only make more effective use of wind energy resources and improve the overall power generation efficiency of wind farms, but also significantly reduce the occupation of land resources. In addition, the large-scale application of large-capacity wind turbines has promoted the maturity of the industry, reduced the unit power generation cost, and improved the overall economic benefits. With the introduction of new technologies such as intelligent control, wind turbine design optimization and material technology progress, the operating efficiency and reliability of wind power systems have been continuously improved.

[0003] However, during operation, large-capacity wind turbines are easily affected by wind loads and their own weight due to the strong flexibility of their mechanical components, which leads to the accumulation of fatigue damage. This damage mainly comes from the cyclic stress and strain that key components such as wind turbine blades and towers are subjected to during long-term operation. The accumulation of fatigue damage not only increases operation and maintenance costs and reduces power generation efficiency, but may also cause structural safety risks and even lead to wind turbine failures or accidents. These problems will not only affect the normal operation of wind farms and lead to frequent shutdowns and maintenance, but may also pose a threat to the surrounding environment and personal safety. Therefore, reducing fatigue damage of large-capacity wind turbines is of great significance to ensuring the stable operation of wind farms, extending the service life of wind turbines, reducing operation and maintenance costs, and improving the economic benefits of wind power projects.

[0004] After receiving the dispatch command, the wind farm will respond to these instructions at a frequency of 1 time per second through its own automatic generation control (AGC) system. The core function of the AGC system is to calculate in real time and distribute the dispatch command to each wind turbine in the wind farm. Each wind turbine adjusts its mechanical parameters according to the allocated power, so as to achieve accurate tracking and response to the allocated power. The traditional power allocation scheme is usually an average distribution, that is, the dispatch command power is evenly distributed to all wind turbines. Although this simple and intuitive method can effectively meet the basic power generation needs in the initial application, it has significant limitations. The main problem is that this method does not take into account the power of each wind turbine. The actual operating status, environmental conditions and accumulated fatigue damage level mean that some wind turbines will be forced to operate at overload for a long time under unsuitable conditions, which will accelerate the accumulation of fatigue damage and increase the risk of failure. At the same time, this simple allocation method fails to effectively control the fatigue damage of wind turbines, limits the overall optimization of wind farms, and leads to waste of resources and reduced economic benefits. Specifically, when some wind turbines operate under overload, problems such as decreased efficiency and increased failure rate may occur, which will affect the power generation capacity of the entire wind farm. In addition, frequent failures and repairs will also lead to increased operating and maintenance costs, further weakening the economic benefits of wind farms.

[0005] In summary, the degree of wind turbine fatigue damage is an important indicator affecting the operating efficiency and economic benefits of wind farms. Wind farms should take this into consideration when allocating active power so as to optimize the active power of wind farms and reduce the impact of damage on wind turbine life. However, there is still a lack of feasible methods to quickly quantify the cumulative fatigue damage degree of wind turbines and to quickly solve the optimization scheduling model. Summary of the invention

[0006] In order to solve the above-mentioned problem in the prior art, that is, the traditional method of evenly distributing the dispatching command power to all wind turbines without considering the influence of power regulation on wind turbine fatigue damage, which leads to an increased probability of wind turbine failure, the first aspect of the present invention proposes a method for efficiently distributing active power of a wind farm taking into account the degree of wind turbine fatigue damage, the method comprising the following steps:

[0007] Step S1, obtaining fan parameter data, and estimating the fan load through force analysis and energy conservation based on the fan parameter data to obtain a fan load value;

[0008] Step S2, combining the load value of the fan, performing quantitative calculation of the fatigue damage degree with low complexity by using an improved energy method, and obtaining a cumulative fatigue value of the fan;

[0009] Step S3, constructing a wind farm active power multi-objective optimization scheduling model based on the wind turbine load value and the wind turbine accumulated fatigue value, and then obtaining the optimized active power of each unit in the wind farm;

[0010] The method for constructing a multi-objective optimization dispatch model for active power of a wind farm is:

[0011] Based on Gulin's method, entropy weight method and improved matter-element extension model, the objective function of optimization solution is constructed, including the construction of objective function and the establishment of objective function weight;

[0012] The optimized power reference value constraint conditions are constructed under the conditions that the sum of the active power reference values ​​after optimization is equal to the grid dispatching instruction value, the active power reference value after optimization is not greater than the rated output power, the output power of each wind turbine is less than the rated power, the generator torque is subject to upper and lower limit constraints, and the rotor speed is subject to upper and lower limit constraints;

[0013] In combination with the objective function and the optimized power reference value constraint condition, a large-scale wind farm active power multi-objective optimization scheduling model is constructed.

[0014] Beneficial effects of the present invention:

[0015] (1) The present invention uses a wind turbine load estimation model based on force analysis and energy conservation to perform force analysis on the wind turbine tower, and combines pitch angle control and transmission system analysis to construct a wind turbine mechanical power model to achieve rapid estimation of the wind turbine load; (2) The present invention uses a low-complexity fatigue damage quantification method based on an improved energy method to read wind turbine load data, calculate the load amplitude in combination with the rain flow method, and calculate the accumulated fatigue damage through an improved energy method; (3) The present invention constructs a multi-objective optimization model with the goal of minimizing the total fatigue load of the wind farm and maximizing the uniformity of fatigue load distribution among wind turbines, thereby achieving rapid and accurate power allocation, allowing the wind turbine to operate in a suitable state, reducing the accumulation of fatigue damage and the risk of failure, and improving the overall operating efficiency and economic benefits of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0017] Figure 1 It is a flow chart of a method for efficiently distributing active power in a wind farm taking into account the degree of fatigue damage of wind turbines according to an embodiment of the present invention;

[0018] Figure 2 is a flow chart of fan load estimation of the present invention;

[0019] Figure 3 It is a flow chart for quantifying the fatigue damage degree of a low-complexity fan of the present invention;

[0020] Figure 4 This is a flow chart of the optimization dispatch of active power of a large-scale wind farm according to the present invention;

[0021] Figure 5 It is a graph showing the changes in weights during the solution process of the present invention;

[0022] Figure 6 It is a diagram of the initial power distribution value and the accumulated fatigue damage value of the fan of the present invention;

[0023] Figure 7 It is a graph of power distribution value and accumulated fatigue damage value of the fan in the process of the present invention;

[0024] Figure 8 It is the power distribution diagram of the present invention at the time of 10s;

[0025] Fig. 9 It is a graph of cumulative fatigue damage value of the thrust of the wind turbine 14 tower before and after the optimization of the present invention;

[0026] Fig.10 It is a diagram of the cumulative fatigue damage value of the wind turbine 10 tower thrust before and after the optimization of the present invention. DETAILED DESCRIPTION

[0027] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.

[0028] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0029] In order to more clearly explain the wind farm active power efficient distribution method taking into account the fatigue damage degree of the wind turbine according to the present invention, the following is combined with Figures 1 to 10 Each step in the embodiment of the present invention is described in detail.

[0030] A first embodiment of the present invention provides a method for efficiently distributing active power in a wind farm taking into account the degree of fatigue damage of wind turbines. The method comprises the following steps:

[0031] Step S1, obtaining wind turbine parameter data, and based on the wind turbine parameter data, estimating the wind turbine load based on force analysis and energy conservation to obtain the wind turbine load value, including wind turbine main shaft torque estimation and wind turbine tower thrust estimation. The main process includes the construction of the main shaft torque state equation, the construction of the tower thrust state equation, and the joint solution of the wind turbine component load. The method is as follows:

[0032] Step S11: Obtain wind turbine parameter data, including wind speed , reference value of each fan power, wind speed The wind speed at the location of the collected wind farm is usually obtained directly from the wind turbine measurement device;

[0033] Step S12: Based on the wind speed , build a pitch angle control model to determine the pitch angle of the wind turbine blades β , build a tower thrust model and calculate the thrust on the wind turbine tower ;

[0034] In this embodiment, the specific process of calculating the wind turbine tower thrust is as follows:

[0035] Step S12A: Establishing the generator speed and rated speed The pitch angle control model is used to adjust the deviation of the pitch angle. β :

[0036] ;

[0037] In the formula, is the proportional coefficient of the PI controller, is the integral coefficient of the PI controller, S is the Laplace operator;

[0038] Step S12B: Constructing tower thrust model:

[0039] ;

[0040] In the formula, is the tower thrust; is the tower thrust coefficient, which can be obtained by looking up the common standards for wind turbines; is the air density; R is the blade radius; is the wind speed. The three commonly used standards for wind turbines mentioned in the article can be found in: GRUNNET JD, SOLTANI M, KNUDSEN T, et al. Aeolus toolbox for dynamics wind farmmodel, simulation and control[J]. European Wind Energy Conference and Exhibition 2010, EWEC 2010, 2010, 4: 3119-3129. ;

[0041] When the wind speed is 0, the tower thrust is:

[0042] ;

[0043] In the formula, is the tower thrust, is the tower thrust under the current wind force, The tower thrust at rated wind speed;

[0044] Step S13: Based on the power reference values ​​of each fan, the power reference values ​​are used as input to establish a main shaft torque state equation, and the low-speed shaft speed is used as a state variable to obtain a calculation result of the main shaft torque;

[0045] In this embodiment, the specific process of obtaining the main shaft torque is as follows:

[0046] Step S13A: Calculate the mechanical power of the wind turbine generator , that is, to calculate the wind energy captured by the wind turbine:

[0047] ;

[0048] In the formula, is the mechanical power, is the air density, R is the blade radius, is the tip speed ratio, is the pitch angle, is the wind speed, is the power factor, by finding the input and The commonly used standards for fans are obtained;

[0049] Step S13B: Using formula (4), a single mass model of the low-speed shaft motion equation is used to describe the wind turbine transmission system, that is:

[0050] ;

[0051] In the formula, and are the rotational inertia of the blade wind wheel and the generator, and Concentrated into an equivalent mass , is the mechanical torque, is the generator torque, is the generator speed, is the gearbox transmission ratio, is the low speed shaft speed;

[0052] Step S13C: construct the electric torque equation as:

[0053] ;

[0054] In the formula, is the power reference command, The electric torque is the electric torque, is the generator efficiency, assuming that the wind turbine can fully respond to the dispatch command, and Output power for the generator.

[0055] Step S13D: Calculate the low-speed shaft torque. The transmission shaft torque is the low-speed shaft torque. , Construct the low-speed shaft torque motion equation, and the low-speed shaft torque calculation formula is:

[0056] ;

[0057] Assuming that the main shaft torque is 0 at rated wind speed, the main shaft torque can be defined as:

[0058] ;

[0059] In the formula, is the spindle torque, is the low-speed shaft torque under the current wind force, is the low speed shaft torque at rated wind speed;

[0060] Step S14: construct a joint solution method for main shaft torque and tower to complete a rapid estimation of the wind turbine load value;

[0061] In this embodiment, the specific process of the rapid estimation of the fan load value is as follows:

[0062] right Linearization can be obtained:

[0063] ;

[0064] Linearize the tower thrust calculation formula and get:

[0065] ;

[0066] in, , , is the parameter defined after linearization, and its value is:

[0067] ;

[0068] To simplify the calculation, we searched for common standards for fans and took , , is a constant;

[0069] General The linearization of the low-speed shaft motion equation yields:

[0070] ;

[0071] In the formula, P m It can be directly calculated according to the current time conditions;

[0072] General The linearization of the electric torque equation yields:

[0073] ;

[0074] General The linearization of the low-speed shaft torque equation yields:

[0075] ;

[0076] In the formula, are the parameters defined after linearization, ;

[0077] In summary, , , , , , , and the state equation is obtained:

[0078] ;

[0079] Using the forward Euler method, the above equation is discretized to obtain:

[0080] ;

[0081] Based on the above formula, the fan load can be quickly estimated and the fan load value can be obtained.

[0082] Step S2, obtaining the load value of the wind turbine, quantifying the degree of fatigue damage with low complexity based on the improved energy method, and obtaining the cumulative fatigue value of the wind turbine;

[0083] In this embodiment, see Figure 3 The specific process of rapid estimation of the fan load value is as follows:

[0084] S21, calculate the total energy loss accumulated by the material during fatigue life:

[0085] ;

[0086] In the formula, is the total energy loss accumulated by the material during fatigue life; is the fatigue life, i.e. the maximum number of cycles at a specific stress level; The energy loss for each cycle is:

[0087] Step S21 A, calculating the energy loss per unit time:

[0088] Calculate the corresponding stress :

[0089] ;

[0090] In the formula, For the Seconds of load, is the cross-sectional area under load;

[0091] Calculate strain :

[0092] ;

[0093] In the formula, E is the elastic modulus, with a value of N;

[0094] Calculation of strain energy density :

[0095] ;

[0096] Step S21B, calculating the accumulated energy loss:

[0097] ;

[0098] Step S22, calculating the cumulative fatigue damage :

[0099] ;

[0100] In the formula, is the total energy loss accumulated when fatigue failure occurs in the material. ; is the cumulative fatigue damage coefficient;

[0101] Step S23, ,Mode Bring-in The final calculation formula for fan fatigue damage is:

[0102] ;

[0103] In the formula, It is the maximum load value of the material at tensile fracture;

[0104] Step S24, the fatigue cumulative value calculation method is linearized:

[0105] General Bring-in The total accumulated energy loss can be obtained as:

[0106] ;

[0107] General Substitute into the cumulative fatigue damage formula and ,have to:

[0108] ;

[0109] when When Linearization gives and formula:

[0110] ;

[0111] Similarly, we get and formula:

[0112] ;

[0113] Similarly, we get and formula:

[0114] ;

[0115] Step S3, constructing a wind farm active power multi-objective optimization scheduling model based on the wind turbine load value and the wind turbine accumulated fatigue value, and then obtaining the optimized active power of each unit in the wind farm;

[0116] The method for constructing a multi-objective optimization dispatch model for active power of a wind farm is:

[0117] Based on Gulin's method, entropy weight method and improved matter-element extension model, the objective function of optimization solution is constructed, including the construction of objective function and the establishment of objective function weight;

[0118] The optimized power reference value constraint conditions are constructed under the conditions that the sum of the active power reference values ​​after optimization is equal to the grid dispatching instruction value, the active power reference value after optimization is not greater than the rated output power, the output power of each wind turbine is less than the rated power, the generator torque is subject to upper and lower limit constraints, and the rotor speed is subject to upper and lower limit constraints;

[0119] In combination with the objective function and the optimized power reference value constraint condition, a large-scale wind farm active power multi-objective optimization scheduling model is constructed.

[0120] In an embodiment of the present invention, the objective function includes objective function construction and objective function weight establishment, which is based on Gulin method, entropy weight method and improved matter-element extension model, and the method is:

[0121] Step S31A, the objective function includes the minimum cumulative fatigue load of the wind farm as a whole and the more uniform distribution of the cumulative fatigue load of each wind turbine in the wind farm. Considering the active power tracking of each wind turbine, the optimal objective function is constructed with the sum of the cumulative fatigue load of the wind farm as a whole and the sum of the averaged cumulative fatigue load index of the wind turbines in the wind farm. :

[0122] ;

[0123] In the formula, For the i The increase in the cumulative fatigue damage degree of the tower thrust of the typhoon during the sampling period is ; For the i The increase in the cumulative fatigue damage degree of the main shaft torque of the typhoon turbine during the sampling period is ; For the current moment i The cumulative fatigue value of the typhoon tower thrust; For the current moment i The cumulative fatigue value of the main shaft torque of the typhoon turbine; For each unit i The normalized fatigue growth penalty coefficient of tower thrust and main shaft torque is calculated as follows: , where the values ​​are: x Take it as the cumulative fatigue value of the previous moment, and m is taken as M, which is preferably 10 here. The principle is that when D grows rapidly, The value of will increase exponentially, thus increasing the fatigue penalty index, forcing the wind farm to select a unit with a smaller D to undertake the power allocation task; is the weight coefficient and satisfies ;

[0124] Step S31B, the establishment of the objective function weight includes four parts, and the method is as follows:

[0125] Step a: First, use the Gulin method to determine the value of the subjective weight. Given the cost of all objective functions, set the cost of the first objective function to 1, and then subjectively determine the ratio of the importance of the next one to the previous one. A j ,Right now:

[0126] ;

[0127] Step b, secondly, determine the objective weight by entropy weight method: solve the objective weight based on historical data, assuming that the historical data has m The objective function is n The decision matrix can be constructed by X as follows:

[0128] ;

[0129] In the formula Indicates i The evaluation object is j The value of each indicator;

[0130] Standardization is used to normalize the original data to the interval [0.01, 0.99]:

[0131] ;

[0132] In the formula is the value after normalization;

[0133] According to the entropy value of each indicator, calculate its corresponding objective weight , the weight is inversely proportional to the entropy value, and the specific formula is:

[0134] ;

[0135] In the formula, Indicates j The entropy value of each indicator;

[0136] Step c: sum and integrate subjective weights , objective weight , the sum of weights is obtained by summing , as shown below:

[0137] ;

[0138] In the formula, is the objective function j The fusion weight of is the fusion weight;

[0139] The weight is normalized and the sum of the weights is 1, that is:

[0140] ;

[0141] Step d, iteratively correct the normalized weights, and use the improved matter-element extension model to dynamically adjust the coefficients. The improved matter-element extension model is based on the traditional matter-element model and combines the extension theory to dynamically adjust the weight changes in the timing optimization. The method is as follows:

[0142] Step d1, construct the formula The object-element model matrix shown is used to calculate the correlation function based on the object-element model matrix. K ij :

[0143] ;

[0144] Step d2: In order to improve the accuracy of the model, the formula for modifying the weight can be adjusted according to the correlation value of each indicator, using the correlation function K ij Corrected weight, corrected weight It can be defined as:

[0145] ;

[0146] Calculate the maximum value of the increase in the cumulative fatigue damage degree of the tower thrust , the minimum increase in the cumulative fatigue damage degree of the tower thrust , the maximum increase in the degree of cumulative fatigue damage of the main shaft torque , the minimum increase in the degree of cumulative fatigue damage of the spindle torque , , , and , where: the maximum value of the fatigue growth penalty coefficient of the tower thrust multiplied by the cumulative fatigue value of the wind turbine tower thrust , the fatigue growth penalty coefficient of the tower thrust multiplied by the minimum value of the cumulative fatigue value of the wind turbine tower thrust , the maximum value of the fatigue growth penalty coefficient of the spindle torque multiplied by the cumulative fatigue value of the spindle torque , the fatigue growth penalty coefficient of the spindle torque multiplied by the minimum value of the accumulated fatigue value of the spindle torque , used to average the load fatigue damage values ​​of each fan, , , , , that is, the maximum and minimum values ​​of each component, and use the formula Normalized, that is:

[0147] ;

[0148] In the formula, z Represents the various components in the objective function;

[0149] In the embodiment of the present invention, the optimization power reference value constraint condition is specifically performed as follows:

[0150] The sum of the active power reference values ​​after the optimization allocation is equal to the grid dispatching instruction, constructing the optimization power reference value constraint condition;

[0151] ;

[0152] In the formula, is the total regulation demand of the wind farm, i.e., the dispatching instruction, t is the current moment, and the optimization model is solved for a certain time section. To simplify the expression, the subscript t is omitted in the subsequent formulas;

[0153] The output power of each wind turbine is limited by the rated output power, and the output power constraint of the wind turbine is constructed:

[0154] ;

[0155] In the formula, is the output power of unit i, is the rated power of unit i, where the rated power of the generator set is its maximum value;

[0156] The reference value of the active power allocated after optimization is not greater than the rated output power, and the reference value constraint of the active power allocated after optimization is constructed:

[0157] ;

[0158] The results are as follows: The difference does not exceed the power difference ,Right now:

[0159] ;

[0160] The generator torque is subject to upper and lower limit constraints, and the generator torque constraint is constructed:

[0161] ;

[0162] In the formula, For the crew i The generator torque, For the crew i The upper limit of the generator torque;

[0163] The rotor speed constraint is constructed by constraining the rotor speed by the upper and lower limits:

[0164] ;

[0165] In the formula, For the crew i The rotor speed, For the crew i The upper limit of the rotor speed, is the rotor speed at the cut-in wind speed;

[0166] In the embodiment of the present invention, the multi-objective optimization scheduling model for active power of a large-scale wind farm is constructed, and the multi-objective optimization model is:

[0167] ;

[0168] Passing ,Mode The linearization process of the multi-objective optimization model can be converted into a single-objective linear programming problem, which is solved by using the commercial large-scale linear optimization solver cplex or the python linear solving library. The input data and settings of the solving process are not the content to be protected by this application and will not be described in detail here.

[0169] In addition, in order to further understand the method of the present invention, the measured data of a certain wind farm is used as an example to perform the optimal scheduling of the active power of the wind farm. It is assumed that the wind farm has a total of 100 wind turbines, and the measured data is a certain 50s data, including the input wind speed and low-speed shaft speed of the wind turbine. The data of the first five wind turbines are shown in the following table: Table 1 is a wind speed table of some wind turbines, Table 2 is a pitch angle table of some wind turbines, and Table 3 is a low-speed shaft speed table of some wind turbines;

[0170] Table 1

[0171] ;

[0172] ;

[0173] Table 2

[0174] ; ;

[0175] Table 3

[0176] ;

[0177] ;

[0178] First, after receiving the dispatch command, the wind farm will respond to these instructions at a frequency of 1 time per second through its own automatic generation control (AGC) system. According to the formula, the wind speed, pitch angle and low-speed shaft speed of each wind turbine are used as inputs, and the power reference value of each wind turbine (that is, the decision variable to be optimized) is used as a known quantity to obtain the tower thrust, main shaft torque and wind turbine load estimation functions of each wind turbine with respect to the decision variables; then, the formula is used to take the tower thrust and main shaft torque of each wind turbine as inputs to obtain the cumulative fatigue damage value of each wind turbine and the fatigue damage at each moment. The net increase of damage; finally, based on the formula, the wind farm active power optimization scheduling model considering the degree of fatigue damage is constructed. The optimal active power reference value (allocated value) of each wind turbine can be obtained by solving the model, and it is judged whether the optimization cycle has been reached. If it is, the automatic generation control (AGC) will send the scheduling instruction, that is, the optimal active power reference value (allocated value) at each moment of the output to each wind turbine in the wind farm. Each wind turbine adjusts its mechanical parameters according to the allocated power, so as to achieve accurate tracking and response to the allocated power. If not, the input data will be updated and the output data of the previous moment will be stored.

[0179] The calculation of the values ​​of each coefficient in the objective function is as follows: first, based on human subjective design, subjective weights are formed, and the subjective relative coefficients are defined as [1, 0.5, 1] ​​based on the Gulin method, that is, the importance of the first two objective functions is twice that of the last two;

[0180] Then, the input data is read, the proportion of each evaluation object on each indicator is calculated, and the entropy value of each indicator and the weight index of each object on each objective function are further obtained. The objective weight of the objective function calculated by formula (39) is [0.086, 0.198, 0.241, 0.214]. This weight indicates that the information entropy of the latter two objective functions is small, which is more important in the optimization solution process. Let , the fusion weight can be calculated to be [0.315, 0.291, 0.178, 0.215], which is the weight at time 1 and is also the initial value of the improved matter-element extension model iteration;

[0181] The active power multi-objective optimization scheduling model considering the cumulative fatigue characteristics of wind turbine load constructed by the present invention is adopted, assuming that the wind farm performs power allocation once per second, and the power reference value of each wind turbine at each time point is calculated;

[0182] See also Figure 5 , Figure 5 It is shown in the figure that in this case, the fatigue value is divided into two parts, the unit output of the part with small fatigue value is larger, and the unit output of the part with large fatigue value is smaller;

[0183] A time section is selected at the beginning and during the optimization process to obtain the active power allocation scheme of this time section, and the cumulative fatigue damage value of each wind turbine load at the previous moment is given. Figure 6 and Figure 7 are the power allocation schemes for 5s and 95s and the accumulated fatigue damage value at the previous moment;

[0184] Figure 6 In order to optimize the comparison chart between the initial power distribution value and the fatigue damage value of the fan, since the cumulative fatigue damage value of the fan at the initial moment is 0, the fatigue damage value of each fan is small in the initial process, and the output of each fan is greatly affected by the fatigue damage value;

[0185] Figure 7 This is a comparison chart between the power distribution value and the fan fatigue damage value during the optimization process. As the fatigue damage value accumulates, the decisive influence of the fatigue damage value on the fan power distribution gradually decreases. The fan output is affected by the accumulated fatigue value at the previous moment, and the two are negatively correlated.

[0186] From the figure Figure 7 It can be seen that the optimized wind turbine power reference value at this moment is divided into two. The specific reasons are as follows Figure 8 Similar; the figure marks the power that each unit should generate according to the equal distribution method. It can be seen that the reference value of the wind turbine power after optimization by the optimization model meets the constraint conditions, that is, the reference power is not greater than the wind turbine output and the output is within plus or minus 1MW of the equal distribution mode; the sum of the cumulative fatigue damage values ​​of the loads of 100 wind turbines before and after optimization is shown in Table 4. It can be seen that the fatigue damage values ​​of the wind turbine loads before and after optimization are significantly reduced. Fig. 9 and Fig.10 It can be seen that when the optimization model is added, the load cumulative fatigue damage value of the fan will be significantly reduced.

[0187] Table 4

[0188] ;

[0189] Although the various steps in the above embodiment are described in the above-mentioned order, those skilled in the art can understand that in order to achieve the effect of this embodiment, different steps do not have to be executed in such an order. They can be executed simultaneously (in parallel) or in a reverse order. These simple changes are within the scope of protection of the present invention.

[0190] A second embodiment of the present invention is a wind farm active power efficient distribution system taking into account the fatigue damage degree of wind turbines, the system comprising: a wind turbine load estimation module, configured to obtain wind turbine parameter data; based on the wind turbine parameter data, the wind turbine load is estimated by force analysis and energy conservation to obtain the wind turbine load value; a wind turbine fatigue damage quantification module, configured to obtain the wind turbine load value, and perform low-complexity fatigue damage degree quantification based on an improved energy method to obtain the accumulated fatigue value of the wind turbine; a wind farm active power multi-objective optimization scheduling module, configured to construct a wind farm active power multi-objective optimization scheduling model based on the wind turbine load value and the accumulated fatigue value of the wind turbine, thereby obtaining the optimized active power of each unit in the wind farm;

[0191] The method for constructing a multi-objective optimization dispatch model for active power of a wind farm is:

[0192] Based on Gulin's method, entropy weight method and improved matter-element extension model, an objective function for optimization solution is constructed, including the construction of the objective function and the establishment of the objective function weight; under the conditions that the sum of the active power reference values ​​after optimization is equal to the grid dispatch instruction value, the active power reference value allocated after optimization is not greater than the rated output power, the output power of each wind turbine unit is less than the rated power, the generator torque is subject to upper and lower limit constraints, and the rotor speed is subject to upper and lower limit constraints, an optimized power reference value constraint condition is constructed; combined with the objective function and the optimized power reference value constraint condition, a large-scale wind farm active power multi-objective optimization scheduling model is constructed.

[0193] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0194] An electronic device according to the third embodiment of the present invention comprises: at least one processor; and a memory communicatively connected to at least one of the processors; wherein the memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the method for efficiently allocating active power in a wind farm taking into account the degree of fatigue damage of the wind turbine.

[0195] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the electronic device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0196] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for efficiently distributing active power of a wind farm taking into account the degree of fatigue damage of wind turbines, characterized in that: The method comprises the following steps: Step S1, obtaining fan parameter data, and estimating the fan load based on the fan parameter data by force analysis and energy conservation to obtain the fan load value; the method for obtaining the fan parameter data, estimating the fan load based on force analysis and energy conservation to obtain the fan load value is as follows: Step S11: Obtain wind turbine parameter data, including wind speed , power reference value of each fan; Step S12: Based on the , build a pitch angle control model to determine the pitch angle of the wind turbine blades Based on the ， Construct tower thrust model to calculate wind turbine tower thrust ; Step S13: constructing a main shaft torque state equation according to the power reference values ​​of each wind turbine; using the low-speed shaft speed as a state variable, inputting the main shaft torque state equation, and obtaining a calculation result of the main shaft torque; Step S14: combining the calculation results of the main shaft torque and the tower thrust, constructing a joint solution method for the main shaft torque and the tower thrust, and completing a rapid estimation of the wind turbine load value; Step S2, combining the load value of the fan, using an improved energy method to perform quantitative calculation of the fatigue damage degree with low complexity, and obtaining the cumulative fatigue value of the fan, the method is: S21, calculate the total energy loss accumulated by the material during fatigue life: ； In the formula, is the total energy loss accumulated by the material during fatigue life; is the fatigue life, i.e. the maximum number of cycles at a specific stress level; is the energy loss for each cycle; Step S21 A, calculating the energy loss per unit time: Calculate the corresponding stress : ； In the formula, For the Seconds of load, is the cross-sectional area under load; Calculate strain : ； In the formula, E is the elastic modulus, with a value of N; Calculation of strain energy density : ； Step S21B, calculating the accumulated energy loss: ； Step S22, calculating the cumulative fatigue damage : ； In the formula, is the total energy loss accumulated when fatigue failure occurs in the material. ; is the cumulative fatigue damage coefficient; Step S23, ,Mode Bring-in The final calculation formula for fan fatigue damage is: ； In the formula, It is the maximum load value of the material at tensile fracture; Step S24, the fatigue cumulative value calculation method is linearized: General Bring-in The total accumulated energy loss can be obtained as: ； General Substitute into the cumulative fatigue damage formula and ,have to: ； when =3, linearize the equation and get and formula: ； Step S3, constructing a wind farm active power multi-objective optimization scheduling model based on the wind turbine load value and the wind turbine accumulated fatigue value, and then obtaining the optimized active power of each unit in the wind farm; The method for constructing a multi-objective optimization dispatch model for active power of a wind farm is: The objective function of optimization solution is constructed based on Gulin method, entropy weight method and improved matter-element extension model, including objective function construction and establishment of objective function weight. The method for establishing the weight coefficient of the objective function is: Step a: Determine subjective weights using the Gulin method ; Step b: Determine the objective weight by entropy weight method ; Step c, , Perform weighted fusion to obtain the sum of weights , and Normalize and get the normalized weight ; Step d, Iterative correction is performed, and the coefficient is dynamically adjusted using the improved matter-element extension model. The method is as follows: Construct a matter-element model matrix and calculate the correlation function based on the matter-element model matrix K ij , using the correlation function K ij Corrected weight, corrected weight: （35）； ； in, is the corrected weight, is the weight before correction, Indicates i The evaluation object is j The value of the indicator, , , They represent the maximum cost, minimum cost, maximum feature, and minimum feature of the evaluation object under the jth indicator respectively; The optimized power reference value constraint conditions are constructed under the conditions that the sum of the active power reference values ​​after optimization is equal to the grid dispatching instruction value, the active power reference value after optimization is not greater than the rated output power, the output power of each wind turbine is less than the rated power, the generator torque is subject to upper and lower limit constraints, and the rotor speed is subject to upper and lower limit constraints; In combination with the objective function and the optimized power reference value constraint condition, a large-scale wind farm active power multi-objective optimization scheduling model is constructed.

2. The method for efficiently distributing active power of a wind farm taking into account the degree of fatigue damage of wind turbines according to claim 1 is characterized in that: The wind turbine tower thrust is calculated , the method is: Step S12A: Establishing the generator speed and rated speed The pitch angle control model is adjusted according to the deviation: ； In the formula, is the proportional coefficient of the PI controller, is the integral coefficient of the PI controller, S is the Laplace operator; Step S12B: Constructing tower thrust model: ； In the formula, is the tower thrust; is the tower thrust coefficient; is the air density; R is the blade radius; is the wind speed; When the wind speed is 0, the tower thrust is: ； In the formula, is the tower thrust, is the tower thrust under the current wind force, The tower thrust at rated wind speed.

3. The method for efficiently distributing active power of a wind farm taking into account the degree of fatigue damage of wind turbines according to claim 1 is characterized in that: The method for calculating the spindle torque is as follows: Step S13A: Calculate the mechanical power of the wind turbine generator : ； In the formula, is the mechanical power, is the air density, R is the blade radius, is the tip speed ratio, is the pitch angle, is the wind speed, is the power coefficient; Step S13B: Use the single mass model of the low-speed shaft motion equation to describe the wind turbine transmission system, that is: ； In the formula, and are the rotational inertia of the blade wind wheel and the generator, and Concentrated into an equivalent mass , is the mechanical torque, is the generator torque, is the generator speed, is the gearbox transmission ratio, is the low speed shaft speed; Step S13C: construct the electric torque equation as: ； In the formula, is the power reference command, The electric torque is the electric torque, is the generator efficiency, assuming that the wind turbine can fully respond to the dispatch command, and Output power for the generator; Step S13D: Calculate the low speed shaft torque: ； Assuming that the main shaft torque is 0 at rated wind speed, the main shaft torque can be defined as: ； In the formula, is the main shaft torque, is the low-speed shaft torque under the current wind force, is the low speed shaft torque at rated wind speed.

4. The method for efficiently distributing active power of a wind farm taking into account the degree of fatigue damage of wind turbines according to claim 1 is characterized in that: The method for jointly solving the main shaft torque and tower thrust is as follows: Linearization is performed, and we get: ； In the formula, is the initial value of the pitch angle, is the initial value of the generator speed; Linearize the tower thrust calculation formula and get: ； in, is the initial value of tower thrust, is the initial value of wind speed, is the initial value of the low-speed shaft speed, is the initial value of the pitch angle; , , is the parameter defined after linearization, and its value is: ； To simplify the calculation, take , , is a constant; General The linearization of the low-speed shaft motion equation yields: ； In the formula, is the initial value of the low-speed shaft speed, is the initial value of the low-speed shaft torque, is the initial value of the generator torque, P m Calculate directly; General The electric torque equation is linearized to obtain: ； In the formula, is the initial value of the generator torque, is the fan power, is the initial value of the fan power, is the initial value of the generator speed, is the initial value of the low-speed shaft speed; General The low-speed shaft torque is linearized to obtain: ； In the formula, Initial value of spindle torque, is the initial value of the generator torque, is the initial value of the low-speed shaft torque, is the generator efficiency, is the initial value of tower thrust, is the initial value of wind speed, are the parameters defined after linearization, ; make , , , , , , and the state equation is obtained: ； In the formula, is the initial value of the low-speed shaft speed, is the proportional coefficient of the PI controller, is the integral coefficient of the PI controller; Using the forward Euler method, the above equation is discretized to obtain: ； In the formula, is the initial value of the low-speed shaft speed; The fan load value can be obtained based on the above formula.

5. The method for efficiently distributing active power of a wind farm taking into account the degree of fatigue damage of wind turbines according to claim 1, characterized in that: The multi-objective optimization scheduling model for active power of a large-scale wind farm is constructed by combining the objective function and the optimized power reference value constraint condition as follows: （43）； in, is the objective function; For the i The increase in the cumulative fatigue damage of the tower thrust of the typhoon during the sampling period is ; For the i The increase in the cumulative fatigue damage degree of the main shaft torque of the typhoon turbine during the sampling period is ; is the cumulative fatigue value of the tower thrust of the i-th wind turbine at the current moment; is the accumulated fatigue value of the main shaft torque of the i-th fan at the current moment; are the fatigue growth penalty coefficients of the normalized tower thrust and main shaft torque of unit i, , x is taken as the accumulated fatigue value of the previous moment, and m is taken as M; is the weight coefficient and satisfies , is the total regulation demand of the wind farm, i.e., the dispatching instruction, and t is the current time; As constraints, Optimize power reference value, The total regulation demand for wind farms; is the output power of unit i, is the rated power of unit i; is the average power distribution value; is the generator torque of unit i, is the upper limit of the generator torque of unit i, is the rotor speed of unit i, is the upper limit of the rotor speed of unit i, is the rotor speed at the cut-in wind speed.

Images