CN106786807B - A kind of wind power station active power control method based on Model Predictive Control - Google Patents

Abstract

The invention discloses a kind of wind power station active power control methods based on Model Predictive Control, the wind speed according to locating for Wind turbines classifies Wind turbines, and the local linearization state-space model of every class unit is established using equivalent modeling method and is based on this predictive controller that designs a model, consider the dynamic response characteristic of inhomogeneity unit, dispatching of power netwoks value is distributed into every class unit by certain priority, every obtained active reference value of class unit is revised by model predictive controller again, is revised value and is proportionately distributed to all units in such.The present invention considers the operation difference between unit using active power of wind power field control as target, and is based on Model Predictive Control, makes it quickly to track dispatching of power netwoks value.

Description

A wind farm active power control method based on model predictive control

technical field

The invention belongs to the technical field of active power control of wind farms, and more particularly relates to a method for controlling active power of wind farms based on model predictive control.

Background technique

As the most important form of renewable energy utilization, wind power has developed rapidly. The traditional wind farm control method is an autonomous active power control method, which allows each wind turbine in the wind farm to generate electricity as much as possible according to the change of wind energy. As the penetration level of wind power increases gradually, the fluctuation of active power of wind farms caused by the uncertainty of wind energy brings great challenges to the safe operation of power grids. The realization of controllable operation of wind farms will gradually become the development trend of grid-connected operation of large-scale wind farms. The key technology for the controllable operation of wind farms is active power control of wind farms.

Wind farm active power control means that the wind farm can do its best to track the grid dispatch value. In order to stabilize the fluctuation of active power output of wind farms, a common method is to use large-scale energy storage equipment, but this method has high equipment cost, technical cost and maintenance cost. Another more economical method is the cooperative control method of wind turbines. This method makes use of the collective effect of the wind farm, assigns a power reference value to each wind turbine through the control unit of the wind farm layer, treats each wind turbine as an actuator, and the sum of the active power output of each unit is the total output of the wind farm. Active power.

Scholars at home and abroad have carried out a series of discussions on the active power cooperative control method of wind turbines. The most commonly used method is to adopt the proportional-integral control method according to the deviation between the grid dispatch value and the active power output of the wind farm measured at the public link point, and then use The proportional distribution method distributes the active reference value to each unit. Because the wind farm is a multi-variable and strongly coupled system with constraints, it is very suitable to be controlled by the model predictive control method. Recently, some scholars have proposed to apply the model predictive control method to wind farm control, and there are two schemes: centralized model predictive control and distributed model predictive control. Since the wind farm model is a multi-input and multi-output system, with the increase of the number of wind turbines, the order increases significantly. Compared with the centralized model predictive control scheme, the distributed model prediction scheme can reduce the amount of calculation. In general, these current studies all control wind farms in a unified manner, and do not take into account the differences in wind conditions at different locations of the units. However, if the differences between units are not considered, resulting in improper power distribution, it will inevitably lead to poor tracking of total active power. And the current research ignores the response time of the unit, and believes that the active power output of the unit can reach the reference value instantly, and instead focuses on the fan load control under the power traceability. Therefore, there are few studies aimed at fast-tracking grid dispatching.

Therefore, an active power control method based on the model predictive control method, which classifies the wind turbines under different wind conditions and reasonably allocates the active power adjustment, so that the wind farm can quickly track the grid dispatch command is of great significance.

SUMMARY OF THE INVENTION

The purpose of the present invention is to overcome the deficiencies of the prior art and provide a wind farm active power control method based on model predictive control, so that the wind farm responds quickly, avoids unnecessary actions of the units, and realizes the tracking of the grid dispatch value.

In order to achieve the above purpose of the invention, the method for controlling the active power of a wind farm based on model predictive control of the present invention is characterized in that it includes the following steps:

(1) According to the factory parameters of the wind turbine, classify the wind conditions of the wind turbine from the cut-in wind speed v _in to the rated wind speed v _rated and then to the cut-out wind speed v _out , and then according to the classification results, the active dispatch power from the grid dispatch center and real-time active power output of wind farms measured from public link points Allocate the active power reference value at the wind farm level, so that each type of wind turbine can be assigned to the active power reference value Among them, i=1,2,...,n, n represents the total number of classifications of wind turbines;

(2) Carry out equivalent modeling for each type of wind turbine after classification according to the classification category, and then design the model predictive controller of each type of wind turbine on the basis of the equivalent modeling, and finally use the model predictive controller to Make corrections to get the active power value

(3), the active power value According to the output ratio of a single wind turbine in this type of wind turbine, it is allocated to a single wind turbine, so that each wind turbine is assigned to the output power value m _i represents the total number of wind turbines of the i-th category;

(4), each wind turbine according to the assigned output power value Output, so as to complete the active power control of the entire wind farm.

Wherein, in the step (1), each type of wind turbine is assigned to an active power reference value The specific steps are:

(2.1), power preprocessing

Calculate the active power output of each type of wind turbine

in, represents the output of the jth wind turbine in the i-th type of wind turbine, and m _i represents the total number of the i-th type of wind turbine;

Calculate the active output of the entire wind farm

in, is the active power output of the i-th type of wind turbines, and n represents the total number of classifications of wind turbines;

Calculate the amount to be regulated ΔP of the wind farm:

in, is the active dispatch power from the grid dispatch center;

(2.2) Allocate the active power reference value of each type of wind turbine

(2.2.1), when ΔP>0, it means that Then the active power reference value of each type of wind turbine The allocation steps are:

1) Calculate the power-up capability of each type of wind turbine:

2) According to the preset regulation sequence, the power-up capability of each type of wind turbine is accumulated in turn. , according to the following formula for distribution to obtain the active reference value of each type of wind turbine

in, Represents the capacity of the i-th type wind turbine to increase the power, Represents the maximum value of active power output of the i-th type of wind turbines, and t means that it can be accumulated to the t-th type of wind turbines in turn to meet the dispatching target;

(2.2.2), when ΔP<0, it means that Then the active power reference value of each type of wind turbine The allocation steps are:

1) Calculate the power reduction capability of each type of wind turbine:

2) According to the preset regulation sequence, the power reduction capability of each type of wind turbine is accumulated in turn. , according to the following formula for distribution to obtain the active reference value of each type of wind turbine

in, Indicates the ability of the i-th type unit to reduce power, Indicates the minimum value of the active output of the i-th type unit.

Further, in the step (2), the method of performing equivalent modeling for each type of wind turbine according to the classification category is:

(3.1), suppose that the wind conditions of the wind turbines in step (1) are divided into n categories, among which, the units in the low wind speed area occupy category k-1, which are expressed as T ₁ , T ₂ ,...T _k-1 ; near the rated wind speed The units in the high wind speed area occupy one class, denoted as T _k ; the units in the high wind speed area occupy the nk class, denoted as T _k+1 , T _k+2 ,…T _n ;

(3.2) Equivalent value for all types of wind turbines

(3.2.1), on the premise that the capacity is equivalent, carry out the equivalent value for each type of wind turbine;

1), perform the equivalent value for the unit in the low wind speed area:

1.1), wind speed, etc.

Calculate the power of each wind turbine in the i-th type of wind turbine according to the wind speed-power function: P _ij =F(vi _j )

Calculate the equivalent active power of the i-th wind turbine

Calculate the equivalent wind speed of the i-th type wind turbine by using the backward method

Among them, j represents the jth wind turbine in the i-th type of wind turbines, m _i represents the total number of the i-th type of wind turbines, F(v _ij ) is the wind speed-power function, v _ij represents the j-th type of wind turbines in the i-th type of wind turbines The wind speed of the typhoon;

1.2), torque equivalent

in, is the equivalent torque of the i-type wind turbine, and its value is equivalent to the torque of a specific unit j ^* in the i-type;

2) Equivalent values for the units near the rated wind speed and those in the high wind speed area;

2.1), wind speed equivalent:

2.2), torque equivalent:

(3.3), establish a state space model according to the equivalent value

(3.3.1), the mechanical torque of the fan The approximation of the first-order Taylor expansion at a given stable operating point yields:

where δ represents the deviation of the variable from its stable operating point, ρ is the air density, R is the rotor radius, v is the wind speed, C _p (λ, β) is the power utilization coefficient, β is the pitch angle, and λ is the tip speed Compare, ω _t is the fan speed, T _t can be regarded as a function of the fan speed ω _t , the pitch angle β and the wind speed v, the overline represents its steady-state operating point;

(3.3.2), establish a state space model

y=δP _g =Cx

in, d=δv

where ω _g is the motor speed, T _g is the electromagnetic torque, The superscript * represents the given value, K _s and B _s represent the equivalent elastic coefficient and equivalent damping coefficient of the rotating shaft, respectively, J _t and J _g represent the moment of inertia of the fan and motor, respectively, τ _g and τ represent the electromechanical time constant and pitch angle time constant, η is the motor efficiency.

Further, in the step (2), the design process of the model predictive controller is:

(4.1), discretize the state space model to get:

x(k+1)=A'x(k)+B _u 'u(k)+B _d 'd(k)

y(k+1)=C′x(k)

(4.2), design optimization objective function and constraints

The optimization objective function is:

The constraints are:

Among them, n _c , n _p represent the control time domain and prediction time domain, respectively, _QP , _QR , _QS represent the weight coefficients, respectively, respectively represent the maximum value of electromagnetic torque and pitch angle of the i-th type wind turbine, and Represents the given value of electromagnetic torque and the given value of pitch angle of the i-th type wind turbine, and the sampling time and change rate constraints are known Multiplying the two can calculate Represent the minimum and maximum value of the electromagnetic torque change rate and the minimum and maximum value of the pitch angle change rate of the i-th type of wind turbine.

The purpose of the invention of the present invention is achieved in this way:

The present invention is a wind farm active power control method based on model predictive control. The wind turbines are classified according to the wind speed where the wind turbines are located, and the equivalent modeling method is used to establish a local linearized state space model of each type of turbines and predict based on the design model. The controller, considering the dynamic response characteristics of different types of units, assigns the grid dispatch value to each type of unit according to a certain priority, and the active power reference value obtained by each type of unit is then revised by the model prediction controller. Proportionately allocated to all units in the class. The invention takes the active power control of the wind farm as the goal, considers the operation difference between the units, and based on the model prediction control, so that the grid dispatch value can be quickly tracked.

Meanwhile, the wind farm active power control method based on the model predictive control of the present invention also has the following beneficial effects:

(1) Compared with the existing PI control method, the wind turbines are classified according to the different wind speeds where they are located, so that the power is distributed according to the operating differences of the units;

(2) Compared with the existing PI control method, the power distribution has priority, which generally reduces the unnecessary frequent action of the pitch angle of the unit (especially the pitch angle action of the unit near the rated wind speed area), The response speed is accelerated, and it is beneficial to prolong the life of the unit;

(3) The essence of the model predictive controller used is to optimize the control quantity on the basis of predicting the future behavior of the process; the prediction does not have strict requirements on the accuracy of the model based on it, which gives the wind farm this kind of High-order nonlinear complex system modeling has a certain fault tolerance space; optimization is to determine the future control effect through the optimization of a certain performance index in a limited period of time in the future. The optimization is repeated online, which is different from the traditional sense. The model predictive control is a closed-loop control algorithm, which makes full use of the actual output error for feedback correction, so it can get a good control effect.

Description of drawings

Figure 1 is a system block diagram of a wind farm active power control method based on model predictive control;

Fig. 2 is the system block diagram of the existing PI control method;

Figure 3 is a diagram of the wind speed in the wind farm used for simulation verification;

Figure 4 is a comparison diagram of the effect of tracking the dispatch value of the power grid;

Figure 5 is a comparison diagram of the pitch angle change of a unit near the rated wind speed area;

Figure 6 is a comparison diagram of the change of the active power output of the unit.

Detailed ways

The specific embodiments of the present invention are described below with reference to the accompanying drawings, so that those skilled in the art can better understand the present invention. It should be noted that, in the following description, when the detailed description of known functions and designs may dilute the main content of the present invention, these descriptions will be omitted here.

Example

Figure 1 is a system block diagram of a wind farm active power control method based on model predictive control.

In this embodiment, as shown in FIG. 1 , the system architecture of the wind farm active power control method based on model predictive control of the present invention mainly includes: 1. allocation of active power reference value at the wind farm level; 2. reference value based on model predictive control 3. Allocation of active power reference revision value among units. The three parts are described in detail below.

1. Distribution of active power reference value at the wind farm level

The assignment of active power reference values at the wind farm level is based on the classification of units. Because even in a small wind farm, due to factors such as geographical location, the wind speed of each wind turbine cannot be exactly the same. Wind turbines are classified into several sub-categories, but in general they can be divided into three categories according to the control method:

Low wind speed area unit: This type of wind turbine operates between the cut-in wind speed and the rated wind speed, the pitch angle is kept at 0°, and the wind energy utilization coefficient is changed by controlling the fan torque to adjust the speed to achieve active power control. The inertia time constant is small, and the response high speed.

Units near the rated wind speed area: This type of unit operates in a small range near the rated wind speed. Because it is near the inflection point of the wind speed-power curve, its working state may change greatly due to small wind speed disturbances, which will cause propellers The constant change of the pitch angle leads to unnecessary mechanical actions of the unit and increases the response time, so it is generally not included in the adjustment, and a fixed active reference value is always assigned to it.

High wind speed area unit: This type of wind turbine operates between the rated wind speed and the cut-out wind speed, and is in the state of constant power generation at rated speed. Active power control is realized by adjusting the pitch angle, and the change of the pitch angle is a mechanical action, and the inertia time constant The higher the wind speed, the slower the response speed; the higher the wind speed of the unit in the high wind speed area, the more sensitive the output power to the change of the pitch angle, that is, the higher the wind speed, the higher the wind speed. The greater the power change.

In this embodiment, the wind conditions in which the wind turbines are located are divided into n categories, wherein the units in the low wind speed area occupy category k-1, which are represented as T ₁ , T ₂ , . . . T _k-1 ; the units in the adjacent rated wind speed area occupy One class, represented as T _k ; high wind speed area units occupy nk class, represented as T _k+1 , T _k+2 ,...T _n ;

When the wind farm is receiving active dispatch power from the grid dispatch center and real-time active power output of wind farms measured from public link points Then, the deviation ΔP between the grid dispatch value and the real-time active power output of the wind farm is obtained, that is, the amount to be regulated by the wind farm, and it is judged whether the wind farm should be controlled to increase or decrease the power. Allocated to each type of unit, so as to obtain the active power reference value of each type of unit In order to speed up the response speed of the wind farm, it is obvious that the pitch angle should be adjusted as little as possible. The priority order of each type of units involved in the adjustment is that the units in the low wind speed area first participate in the adjustment, and the units in the lowest wind speed area are followed by the wind speed increase direction. , units in the next low wind speed area, until a certain type of units in the low wind speed area whose wind speed is closest to the rated wind speed are included in the active power adjustment. If the active power control target has not been reached, the units in the high wind speed area will participate in the adjustment again. The units in the highest wind speed area, the second highest wind speed area, and some types of units whose wind speed is closest to the rated wind speed in the high wind speed area are included in the active power regulation. Among them, after a certain type of unit participates in the adjustment and can meet the grid scheduling requirements, the active power output of the following units will maintain the original value according to the priority order.

The active power reference value is assigned to each type of wind turbine as follows The specific steps are described in detail, as follows:

(1) Calculate the active power output of each type of wind turbine

in, represents the output of the jth wind turbine in the i-th type of wind turbine, and m _i represents the total number of the i-th type of wind turbine;

Calculate the active output of the entire wind farm

in, is the active power output of the i-th type of wind turbines, and n represents the total number of classifications of wind turbines;

Calculate the amount to be regulated ΔP of the wind farm:

in, is the active dispatch power from the grid dispatch center;

(2) Allocate the active power reference value of each type of wind turbine

(2.1), when ΔP>0, it means that Indicates that in order to track the grid dispatch value, the wind farm needs to increase the active power output at this time, then the active power reference value of each type of wind turbine The allocation steps are:

1) Calculate the power-up capability of each type of wind turbine:

2) According to the preset regulation sequence, the power-up capability of each type of wind turbine is accumulated in turn. , according to the following formula for distribution to obtain the active reference value of each type of wind turbine

in, Represents the capacity of the i-th type wind turbine to increase the power, Represents the maximum value of active power output of the i-th type of wind turbines, and t means that it can be accumulated to the t-th type of wind turbines in turn to meet the dispatching target;

(2.2), when ΔP<0, it means that Indicates that in order to track the grid dispatch value, the wind farm needs to reduce the active power output at this time, then the active power reference value of each type of wind turbine The allocation steps are:

1) Calculate the power reduction capability of each type of wind turbine:

2) According to the preset regulation sequence, the power reduction capability of each type of wind turbine is accumulated in turn. , according to the following formula for distribution to obtain the active reference value of each type of wind turbine

in, Indicates the ability of the i-th type unit to reduce power, Represents the minimum value of the active power output of the i-th type of wind turbines, and t means that it can be accumulated to the t-th type of wind turbines in turn to meet the dispatching target.

2. Reference value revision based on model predictive control

In order to design the model predictive controller, it is necessary to carry out equivalent modeling for each type of wind turbine. The following is based on the premise that the capacity is equivalent, and the single-machine equivalent of each type of wind turbine is carried out respectively. The equivalent method is divided into two situations:

1), perform the equivalent value for the unit in the low wind speed area:

1.1), wind speed, etc.

Calculate the power P _ij of each wind turbine in the i-th wind turbine according to the wind speed-power function: P _ij =F(v _ij )

Calculate the equivalent active power of the i-th wind turbine

Calculate the equivalent wind speed of the i-th type wind turbine by using the backward method

Among them, j represents the jth wind turbine in the i-th type of wind turbines, m _i represents the total number of the i-th type of wind turbines, F(v _ij ) is the wind speed-power function, v _ij represents the j-th type of wind turbines in the i-th type of wind turbines The wind speed of the typhoon;

For example, if two units with wind speeds of 6m/s and 8m/s respectively perform wind speed equivalence, the wind speed-power function is Air density ρ=1.2231kg/m ³ , rotor radius R=63m, since the unit is located in the low wind speed area, the power utilization coefficient takes the maximum value C _p =C _pmax =0.482, and v is the wind speed. First, according to the wind speed-power function, the powers of the two units are calculated respectively, 0.793872MW and 1.881772MW. Calculate the sum, and then inversely solve the wind speed-power function to obtain the equivalent wind speed of about 9m/s.

1.2), torque equivalent

in, is the equivalent torque of the i-type wind turbine, and its value is equivalent to the torque of a specific unit j ^* in the i-type;

2) Equivalent the units in the adjacent rated wind speed area and the units in the high wind speed area;

2.1), wind speed equivalent:

2.2), torque equivalent:

3), establish a state space model according to the equivalent value

3.1) Since the model predictive control does not have high requirements on the accuracy of the predictive model, the mechanical torque of the fan is directly The first-order Taylor expansion is performed at a given stable operating point, and the partial linearization of T _t near the stable operating point is approximated to obtain:

where δ represents the deviation of the variable from its stable operating point, ρ is the air density, R is the rotor radius, v is the wind speed, C _p (λ, β) is the power utilization coefficient, β is the pitch angle, and λ is the tip speed Compare, ω _t is the fan speed, T _t can be regarded as a function of the fan speed ω _t , the pitch angle β and the wind speed v, the overline represents its steady-state operating point;

3.2), establish a state space model

y=δP _g =Cx

in, d=δv

where ω _g is the motor speed, T _g is the electromagnetic torque, The superscript * represents the given value, K _s and B _s represent the equivalent elastic coefficient and equivalent damping coefficient of the rotating shaft, respectively, J _t and J _g represent the moment of inertia of the fan and motor, respectively, τ _g and τ represent the electromechanical time constant and pitch angle time constant, η is the motor efficiency.

4), and then design the model predictive controller according to the state space model, and its design process is:

4.1), discretize the state space model to get:

x(k+1)=A'x(k)+B _u 'u(k)+B _d 'd(k)

y(k+1)=C′x(k)

4.2), design optimization objective function and constraints

The optimization objective function is:

The constraints are:

Among them, n _c , n _p represent the control time domain and prediction time domain, respectively, _QP , _QR , _QS represent the weight coefficients, respectively, respectively represent the maximum value of electromagnetic torque and pitch angle of the i-th type wind turbine, and respectively represent the difference between the given value of electromagnetic torque and the given value of pitch angle of the i-th type of wind turbine at two sampling times, Respectively represent the given value of electromagnetic torque and given value of pitch angle of the i-th type wind turbine, and the sampling time and change rate constraints are known Multiplying the two can calculate Represent the minimum and maximum value of the electromagnetic torque change rate and the minimum and maximum value of the pitch angle change rate of the i-th type of wind turbine.

3. Allocation of active power reference correction value among units

In order to ensure the fairness of allocation among units with similar operating states, the revised value of the active power reference value of each type of units obtained by the model predictive controller is used. According to the output ratio of a single wind turbine in this type of wind turbine, it is allocated to a single wind turbine, so that each wind turbine is assigned to the output power value

FIG. 2 is a system block diagram of a conventional PI control method.

In this method, the proportional-integral control method is adopted to control only the deviation between the grid dispatch value and the active power output of the wind farm measured at the public link point, and then the active power reference value is distributed to each unit using the proportional distribution method. This method does not consider the differences between units, and there is no priority in the adjustment process. The active power reference value of all units will fluctuate with the fluctuation of the grid dispatch value, which will lead to frequent pitch angles of some units, especially units near the rated wind speed area. The action will inevitably prolong the response time, which is not conducive to the fast tracking of the grid dispatch value of the wind farm, but also damages the life of the unit and increases the cost of the wind farm.

Example

In order to illustrate the technical effect of the present invention, the active power control method of the present invention is applied to a wind farm with a capacity of 5MW each and 14 wind turbines for simulation verification. And assuming that the wind speed measured by the wind farm is shown in Figure 3, according to the fan parameters used in the simulation (cut-in wind speed v _in = 3m/s, rated wind speed v _rated = 11.4m/s, cut-out wind speed v _out = 25m/ s) Divide the wind speed into 5 sub-categories, as shown in Table 1. The 14 units in the wind farm are classified into 5 sub-categories according to the initial wind speed. The classification of units is shown in Table 2. The effectiveness of the method of the present invention is illustrated by taking the grid dispatch value variation of 50MW-42MW-36MW-40MW-45MW-52MW as an example.

Table 1 is divided into 5 sub-categories by wind speed

Table 2 Classification of units in wind farms

Figure 4 is a comparison diagram of the effect of tracking the dispatch value of the power grid. It can be seen from FIG. 4 that the wind farm performs active power control according to the method of the present invention, and the response speed is faster than that of the existing PI control method. The time for the active power of the wind farm to reach and then stabilize within the error band of the grid dispatch value ±5% is defined as the adjustment time, the index of the present invention is 1.27s, and the index of the PI control method is 2.4s. Compared with the existing method, the speed is improved by 47.2%.

Figure 5 is a comparison diagram of the pitch angle changes of the units in the near-rated wind speed area. It can be seen from Figure 5 that the active power control of the wind farm is carried out according to the method of the present invention, and the active power reference value allocated to each unit in the adjacent rated wind speed area is always 5MW, so that its pitch angle is always kept at 0° , which is conducive to the rapid response of the wind farm; while the PI control method does not consider the sensitivity characteristics of such units to wind speed disturbances, and constantly assigns power reference values to them, resulting in frequent changes in the pitch angle, which is not conducive to the rapid response of the wind farm.

Figure 6 is a comparison diagram of the change of the active power output of the unit. From Figure 6, it can be seen that the power variation trend of the five types of units under the two methods. It is described by its power value at six sampling instants of stable operation. The six moments are 5s, 15s, 25s, 35s, 45s, 55s, 65s. The first three bar charts of each type describe the output of each type of units when the wind farm tracks the gradually decreasing grid dispatch power (the grid dispatch value suddenly changes from the initial 50MW to 42MW to 36MW). When the dispatch value suddenly changes from the initial 50MW to 42MW, the present invention is used to control the active power, and the first and second types of units in the low wind speed area take the lead to reduce to the minimum value, because the power reduction amount has not yet been met, and the high wind speed area is at the highest The fifth category of wind speed units participated in the adjustment, but the fifth category of units still could not meet the adjustment amount after the minimum power was reduced, and then the fourth category of units with the second highest wind speed in the high wind speed area also reduced their output and participated in the adjustment. This process It can be seen from the first to second bar graphs of each type of unit; when the grid dispatch value suddenly changes from 42MW to 36MW, since the first, second and fifth types of units have all dropped to the set minimum power value, only The fourth type of unit still has an adjustment margin. The fourth type of unit continues to reduce output and completes the tracking target. This process can be seen from the second to third bar graphs of each type of unit. Similar to power reduction, in the process of power increase, active power control is performed according to the method of the present invention, and each type of unit still participates in the adjustment according to the preset priority order, that is, the first type, the second type, the fifth type, the first type. The four types of units act in sequence, which is in line with the expected effect. However, the active power output of each type of unit using the PI control method will change continuously with the change of the grid dispatch value, and the system stability is not good.

Although the illustrative specific embodiments of the present invention have been described above to facilitate the understanding of the present invention by those skilled in the art, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, As long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the inventive concept are included in the protection list.

Claims (3)

1. a wind farm active power control method based on model predictive control, is characterized in that, comprises the following steps: (1) According to the factory parameters of the wind turbine, classify the wind conditions of the wind turbine from the cut-in wind speed v _in to the rated wind speed v _rated and then to the cut-out wind speed v _out , and then according to the classification results, the active dispatch power from the grid dispatch center and real-time active power output of wind farms measured from public link points Allocate the active power reference value at the wind farm level, so that each type of wind turbine can be assigned to the active power reference value Among them, i=1,2,...,n, n represents the total number of classifications of wind turbines; (2) Carry out equivalent modeling for each type of wind turbine after classification according to the classification category, and then design the model predictive controller of each type of wind turbine on the basis of the equivalent modeling, and finally use the model predictive controller to Make corrections to obtain the active power value P _i ^ref ; (3) Allocate the active power value P _i ^ref to the single wind turbine according to the output ratio of the single wind turbine in this type of wind turbine, so that each wind turbine is assigned to the output power value j=1,2,...,m _i , where m _i represents the total number of wind turbines of the i-th type; (4), each wind turbine according to the assigned output power value output to complete the active power control of the entire wind farm; Among them, each type of wind turbine is assigned to the active power reference value The specific steps are: (2.1), power preprocessing Calculate the active power output P _i ^out of each type of wind turbine: in, represents the output of the jth wind turbine in the i-th type of wind turbine, and m _i represents the total number of the i-th type of wind turbine; Calculate the active output of the entire wind farm Among them, P _i ^out is the active power output of the i-th type of wind turbines, and n represents the total number of classifications of wind turbines; Calculate the amount to be regulated ΔP of the wind farm: in, is the active dispatch power from the grid dispatch center; (2.2) Allocate the active power reference value of each type of wind turbine (2.2.1), when ΔP>0, it means that Then the active power reference value of each type of wind turbine The allocation steps are: 1) Calculate the power-up capability of each type of wind turbine: ΔP _i ^up =P _i ^max -P _i ^out 2) According to the preset regulation sequence, the power-up capability of each type of wind turbine is accumulated in turn. , according to the following formula for distribution to obtain the active reference value of each type of wind turbine Among them, ΔP _i ^up represents the ability of the i-th type of wind turbines to increase the power, P _i ^max represents the maximum active power output of the i-th type of wind turbines, and t means that it can be accumulated to the t-th type of wind turbines in turn to meet the dispatch target; (2.2.2), when ΔP<0, it means that Then the active power reference value of each type of wind turbine The allocation steps are: 1) Calculate the power reduction capability of each type of wind turbine: ΔP _i ^down =P _i ^out -P _i ^min 2) According to the preset regulation sequence, the power reduction capability of each type of wind turbine is accumulated in turn. , according to the following formula for distribution to obtain the active reference value of each type of wind turbine Among them, ΔP _i ^down represents the power reduction capability of the i-th type unit, and P _i ^min represents the minimum value of the active output of the i-th type unit. 2 . The wind farm active power control method based on model predictive control according to claim 1 , wherein in the step (2), the method for performing equivalent modeling for each type of wind turbines according to the classification category is: 2 . : (3.1), suppose that the wind conditions of the wind turbines in step (1) are divided into n categories, among which, the units in the low wind speed area occupy category k-1, which are expressed as T ₁ , T ₂ ,...T _k-1 ; near the rated wind speed Units occupy one class, denoted as T _k ; units in high wind speed area occupy class nk , denoted as T _k+1 , T _k+2 ,...T _n ; (3.2) Equivalent value for all types of wind turbines (3.2.1), on the premise that the capacity is equivalent, carry out the equivalent value for each type of wind turbine; 1), perform the equivalent value for the unit in the low wind speed area: 1.1), wind speed, etc. Calculate the power of each wind turbine in the i-th wind turbine according to the wind speed-power function: P _ij =F(v _ij ) Calculate the equivalent active power P _i ^eq of the ith wind turbine: Calculate the equivalent wind speed of the i-th type wind turbine by using the backward method Among them, j represents the jth wind turbine in the i-th type of wind turbines, m _i represents the total number of the i-th type of wind turbines, F(v _ij ) is the wind speed-power function, v _ij represents the j-th type of wind turbines in the i-th type of wind turbines The wind speed of the typhoon; 1.2), torque equivalent in, is the equivalent torque of the i-type wind turbine, and its value is equivalent to the torque of a specific unit j ^* in the i-type; 2) Equivalent values for the units near the rated wind speed and those in the high wind speed area; 2.1), wind speed equivalent: 2.2), torque equivalent: (3.3), establish a state space model according to the equivalent value (3.3.1), the mechanical torque of the fan A first-order Taylor expansion approximation at a given stable operating point yields: where δ represents the deviation of the variable from its stable operating point, ρ is the air density, R is the rotor radius, v is the wind speed, C _p (λ, β) is the power utilization coefficient, β is the pitch angle, and λ is the tip speed Compare, ω _t is the fan speed, T _t is regarded as a function of the fan speed ω _t , the pitch angle β and the wind speed v, and the overline - represents its steady-state operating point; K _ω , K _v and K _β are respectively related to the fan speed , the coefficient of pitch angle and wind speed; (3.3.2), establish a state space model y=δP _g =Cx in, represents the derivative of x with respect to time t; Among them, P _g is the generator power, ω _g is the motor speed, T _g is the electromagnetic torque, The superscript * represents a given value, K _s and B _s represent the equivalent elastic coefficient and equivalent damping coefficient of the rotating shaft, respectively, J _t and J _g represent the moment of inertia of the fan and motor, respectively, τ _g and τ represent the electromechanical time constant and pitch angle time constant, η is the motor efficiency. 3. The wind farm active power control method based on model predictive control according to claim 2, is characterized in that, in described step (2), the design process of model predictive controller is: (4.1), discretize the state space model to get: x(k+1)=A'x(k)+B _u 'u(k)+B _d 'd(k) y(k+1)=C′x(k) (4.2), design optimization objective function and constraints The optimization objective function is: The constraints are: where |||| represents the norm, n _c , n _p represent the control time domain and prediction time domain, respectively, Q _P , Q _R , Q _S represent the weight coefficients, respectively, respectively represent the maximum value of electromagnetic torque and pitch angle of the i-th type wind turbine, and respectively represent the difference between the given value of electromagnetic torque and the given value of pitch angle of the i-th type wind turbine at two sampling times, Respectively represent the electromagnetic torque given value of the i-th type of wind turbine, and the sampling time and change rate constraints are known Multiply the two to get Respectively represent the minimum and maximum values of electromagnetic torque and the minimum and maximum values of the pitch angle of the i-type wind turbine; P _i ^max represents the maximum active output of the i-type wind turbine, and P _i ^min represents the active power of the i-type wind turbine. Output the minimum value.