Summary of the invention
The invention provides a kind of wind driven generator unit yaw system control performance optimization method and system, be used for solving now
Frequent starting yaw system in technology is had to reduce the service life of relevant device, wind power generating set, the difference to different model
Flow wind speed and take identical optimisation strategy, not there is specific aim, control the problem that effect is undesirable, generated energy lifting is low.
In order to solve above-mentioned technical problem, a technical scheme of the present invention is for providing a kind of wind driven generator unit yaw system
Control performance optimization method, including:
Within a predetermined period of time, flowing of a wind generating set engine room front to be optimized is obtained every Fixed Time Interval
Wind data, wherein, flows wind data and includes flowing wind speed, flowing true wind direction and flow the wind direction in relative cabin;
Described true wind direction interval by a predetermined angle of flowing is carried out a point sector, by the described wind speed that flows by predetermined wind speed district
Between carry out segmentation;
Carry out being grouped for the first time to the described wind data that flows according to point sector and segmentation;
Calculate, according to often organizing the wind direction flowing relative cabin described in data, the yaw error that correspondence is organized, obtain yaw error
Optimized model;
When described wind driven generator unit yaw system to be optimized optimizes, by actual measurement flow wind speed and to flow true wind direction defeated
Enter in described yaw error Optimized model, the yaw error that coupling is corresponding, the yaw error of this correspondence is adapted to described driftage
In the input of control system.
Another technical scheme of the present invention optimizes system, bag for providing a kind of wind driven generator unit yaw system control performance
Include:
Sampling module, within a predetermined period of time, obtains a wind power generating set to be optimized every Fixed Time Interval
Front, cabin flow wind data, wherein, flow wind data and include flowing wind speed, flowing true wind direction and flow relative cabin
Wind direction;
Divide module, for described true wind direction interval by a predetermined angle of flowing is carried out a point sector, flow wind by described
Speed carries out segmentation by predetermined wind speed interval;
First grouping module, for carrying out being grouped for the first time to the described wind data that flows according to point sector and segmentation;
Yaw error seismic responses calculated module, for calculating according to the wind direction often flowing relative cabin described in group data
The yaw error of corresponding group obtains the Optimized model of yaw error;
Yaw error optimizes module, for when described wind driven generator unit yaw system to be optimized optimizes, by actual measurement
Flowing wind speed and flow in the true wind direction described yaw error Optimized model of input, the yaw error that coupling is corresponding, by this correspondence
Yaw error be adapted in the input of described yaw control system.
The present invention wind power generating set of different model can be in difference to flow take under wind conditions different excellent
Change strategy, it is possible to optimize the yaw error of yaw system targetedly, improve the optimization precision of yaw error, it is possible to reach to carry
Rise the purpose of generated energy.
When being embodied as, predetermined amount of time is such as 30 days.Depending on predetermined amount of time also can be according to optimizing precision, the present invention couple
Its concrete value does not limits.
Fixed Time Interval is such as 10min.Depending on Fixed Time Interval can be according to the degree of stability of data, it is had by the present invention
Body value does not limits.
Flow each variable data in wind data be sampling in Fixed Time Interval (sample frequency as 1s) obtain should
The meansigma methods of variable instantaneous value.Further, in order to ensure that data are useful, flowing wind data is in wind power generating set to be optimized
Gather under normal operating condition, the data recorded should be rejected under the abnormal conditions such as unit fault, maintenance.
Specifically, flow wind speed and to flow the wind direction in relative cabin be by being installed on wind generating set engine room to be optimized
On anemometer record, to flow true wind direction be that wind power generating set is according to wind direction and the nacelle position meter flowing relative cabin
Obtain.
Step 102, carries out a point sector by described true wind direction interval by a predetermined angle of flowing, by the described wind speed that flows by advance
Determine wind speed interval and carry out segmentation.
When dividing sector, Sector Range should be reduced as far as possible, make each sector be affected by landform, barrier etc. identical, right
In the unit that landform, barrier situation are especially complex, it is proposed that the angular range of packet should be less than 15 °.Predetermined angular can be according to treating
Depending on optimizing landform, barrier situation residing for wind power generating set, the value of predetermined angular is not limited by the present invention.
Predetermined wind speed interval can be mean allocation, as 6m/s.Predetermined wind speed interval can be also low wind speed (threshold wind velocity-
6m/s), middle wind speed (7m/s-rated wind speed) and high wind speed section (more than rated wind speed).
Step 103, carries out being grouped for the first time to the described wind data that flows according to point sector and point wind speed.Data after packet
Can store in the form of a list.
Step 104, calculates, according to often organizing the wind direction flowing relative cabin described in data, the yaw error that correspondence is organized, obtains
Yaw error Optimized model.
The yaw error that this step obtains is that yaw system is pacified by wind vane (wind vane in wind power generating set) zero-bit
The synthetic error that dress azimuth, cabin wind vane are caused by impeller wake effect or control strategy the deficient validity.
Concrete, as in figure 2 it is shown, the detailed process of step 104 is:
Step 1041: add up and flow the probability that the wind direction in relative cabin occurs in every group, by probability from big to small suitable
The wind direction flowing relative cabin in ordered pair often group is ranked up.
Step 1042: the meansigma methods calculating the often wind direction that group top n flows relative cabin obtains the yaw error of corresponding group.
Step 1043: the yaw error often organizing correspondence is gathered and obtains described yaw error Optimized model.
Step 105, when described wind driven generator unit yaw system to be optimized optimizes, flowing actual measurement wind speed and flowing
True wind direction inputs in described yaw error Optimized model, mates flowing wind speed and flowing corresponding inclined of true wind direction of this actual measurement
Boat error, is adapted to the yaw error of this correspondence in the input of described yaw control system.
This enforcement uses the mode of comprehensive yaw error, in the case of not considering to cause the concrete reason of yaw error,
Directly compensate, improve enforceability and its uncertainty of reduction that driftage optimizes.
In order to the yaw error Optimized model of the wind power generating set to be optimized by expand to other same models treat excellent
Change wind power generating set, in one embodiment of the invention, also include after obtaining yaw error Optimized model:
Determine effective sector of described wind power generating set to be optimized;When another wind power generating set blade normal (i.e.
Cabin centrage) when effective sector, described yaw error Optimized model is applicable to another wind power generating set described, can root
According to yaw error Optimized model, another wind power generating set described is carried out yaw error optimization;Wherein, another electromotor described
Organize identical with described wind power generating set model to be optimized.
During enforcement, effective sector of wind power generating set to be optimized can be achieved by the prior art, and here is omitted.
In order to verify the accuracy that yaw error optimizes, in one embodiment of the invention, as it is shown on figure 3, by the following method
Checking yaw error Optimized model is the most effective:
Step 301: the yaw system of wind power generating set to be optimized the most also includes that the power before statistic op-timization is special
Linearity curve fraction.
Step 302: also include the power characteristic fraction after statistic op-timization after described yaw system optimization.
Described power characteristic fraction is calculated by equation below:
The power characteristic fraction of K: wind power generating set;
N: the interval number of statistics, according to wind speed range be 0.5m/s be a statistics interval, interval center is 0.5m/s
Integral multiple;
Pi: wind power generating set is in active power and exports under control model for maximum, and the interval inner blower of i-th statistics is real
The average active power value of border output, unit is kW;
Pi': under standard atmosphere density conditions, the active power in the corresponding i-th statistics interval that producer ensures, unit is
kW;
The frequency that i-th statistics is interval;
Ni: wind speed falls into the data amount check that i-th statistics is interval;
The total quantity of N: air speed data.
Step 303: the power characteristic fraction before comparing the power characteristic fraction after optimization and optimizing, as
Power characteristic fraction after fruit optimizes is more than the power characteristic fraction before optimizing, the most described yaw error optimization
Model is effective.
If power characteristic fraction is less than power characteristic fraction before optimizing after You Huaing, then need re-optimization inclined
Boat error, and the rotating speed of wind power generating set is carried out frequency analysis so that rotating speed controls steadily before and after yaw error compensates,
Avoid the occurrence of bigger rotating speed shake and the abnormal vibrations of unit.
In one embodiment of the invention, bent also by the power features after comparing the power features curve before optimization and optimizing
Line, when power features curve entirety offsets to the right, then yaw error Optimized model is effective.Concrete, the power of unit to be optimized
Characteristic curve can use existing method to obtain, obtain as described in wind generating set engine room front to be optimized flow wind data
While also obtain ambient temperature, air pressure and unit to be optimized output electrical power, combine flow wind data statistics machine to be optimized
The power characteristic of group.
In one embodiment of the invention, as shown in Figure 4, wind driven generator unit yaw system control performance optimization method also wraps
Include:
Step 401: carry out being grouped for the second time to carrying out flow data by low wind speed section, middle wind speed section and high wind speed section;Wherein, low
Wind speed range is V0< V≤V1, middle wind speed range is V1< V≤V2, high wind speed scope is V > V2, V2For rated wind speed, V0For opening
Wind symptom speed, V1For low wind speed section threshold value, as 6m/s, V for flowing wind speed.
Concrete, the segmentation of high, medium and low wind speed section is according to controlling according to the different operation characteristics of unit, the different of driftage
System strategy or propeller pitch angle, generating unit speed statistical analysis carry out segmentation.As it is shown in figure 5, according to wheel speed and propeller pitch angle at unit
Different characteristics under the different operation phase carries out the example of segmentation, and first paragraph be " blade rotational speed rising, propeller pitch angle constant " stage,
Second segment is " it is critical that wheel speed and the propeller pitch angle change " stage, and the 3rd section is " wheel speed is constant, and propeller pitch angle raises " rank
Section.According to segmentation shown in Fig. 5, determine the wind speed of each segmentation critical point, thus obtain V1And V2。
Step 402: the often group of second time packet is set different driftage startup angle and optimizing criterions, obtains starting angle and optimizing
Model.
As started angle and optimizing criterion it is:
It is D that low wind speed starts the Optimality Criteria at anglel±Cl, wherein, DlAngle, C is started for the driftage of low wind speedlFor adjustment amount;
It is D that middle wind speed starts the Optimality Criteria at anglem-Cm, wherein, DmAngle, C is started for the driftage of middle wind speedmFor adjustment amount;
It is D that high wind speed starts the Optimality Criteria at angleh±Ch, wherein, DhAngle, C is started for high wind speed driftagehFor adjustment amount.
During enforcement, also can be according to optimizing precision adjusting and optimizing criterion, Optimality Criteria is not specifically limited by the present invention.
Step 403: when described yaw system optimizes, inputs described startup angle and optimizing model by the wind speed that flows of described actual measurement
In, driftage corresponding to wind speed of flowing obtaining described actual measurement starts angle, and the to be surveyed wind direction flowing relative cabin reaches described
When corresponding driftage starts angle, start described yaw system action.
In order to verify the accuracy starting angle and optimizing, in one embodiment of the invention, as shown in Figure 6, test by the following method
It is the most effective that card starts angle and optimizing model:
Step 601: also include the power characteristic fraction before statistic op-timization before described yaw system optimization, utilizes bag
Acquisition is flowed the wind direction in relative cabin and flows wind speed and carry out statistical analysis by network method or probability distribution method, obtains optimizing front control
Dead band processed length.
Concrete, envelope method calculates the process of controlling dead error and is: sets and flows wind speed as abscissa, flows relative cabin
Wind direction is vertical coordinate, and wind data is rehearsed according to the size carrying out wind velocity in the future, utilizes envelope method flowing statistics relatively
The data edges in cabin carries out envelope, and coenvelope deducts lower envelope and i.e. obtains controlling dead error length.
Probability distribution method calculates the process of controlling dead error: according to wind speed order demarcation interval section from small to large, need to
The to be calculated wind direction flowing relative cabin is assigned in each interval of wind speed calculate, and for each wind speed interval section, calculates respectively
The probability that the wind direction in each relative cabin occurs, is then carried out centered by the wind direction in the highest relative cabin of probability of occurrence to the left and right
Recursive form wind direction is interval, and calculates the probability that wind direction is interval, when wind direction interval probability is more than fixation probability, as 95%
Time, this wind direction interval difference is controlling dead error length.
Step 602: also include during described yaw system optimization adding up described wind power generating set to be optimized and a mark post
The driftage of blower fan controls number of times, and described mark post blower fan is any one wind-driven generator around described wind power generating set to be optimized
Group.
Step 603: also include the power characteristic fraction after statistic op-timization after described yaw system optimization, statistics the
Secondary divides the controlling dead error length after the optimization of wind speed section.
The computational methods of controlling dead error see step 601, and here is omitted.
Step 604: the controlling dead error length after optimizing compares with controlling dead error length before optimization respectively, after optimizing
Power characteristic fraction compares, by described wind power generating set to be optimized with the power characteristic fraction before optimization
Driftage controls number of times and compares with the driftage control number of times of described mark post blower fan.
Step 605: if low wind speed section optimize after controlling dead error length with optimization before controlling dead error length difference little
In the first predetermined threshold, the controlling dead error length after middle wind speed section optimizes is less than the controlling dead error length before optimizing, high wind speed section
Controlling dead error length after optimization and the controlling dead error length difference before optimization are less than the second predetermined threshold;Power after optimization is special
Linearity curve fraction is more than or equal to the power characteristic fraction before optimizing;The driftage of described wind power generating set to be optimized
Control number of times and control number of times less than or equal to the driftage of described mark post unit, then start angle and optimizing model effective.
When being embodied as, the first predetermined threshold and the second predetermined threshold specifically can determine according to unit.High wind speed section starts
During angle and optimizing, whether unit to be paid close attention to produces Types of Abnormal Vibration Appearances, and the second set predetermined threshold need to can guarantee that unit does not goes out
Existing abnormal vibrations.
Yaw error, startup angle that operation wind generating set yaw can be controlled by the present invention carry out individual character optimization,
Improve the power characteristic of unit while improving yaw system reliability, improve the generated energy of unit.
As it is shown in fig. 7, Fig. 7 is the knot that one embodiment of the invention wind driven generator unit yaw system control performance optimizes system
Composition.Concrete, this system 700 includes:
Sampling module 701, within a predetermined period of time, obtains a wind-driven generator to be optimized every Fixed Time Interval
That organizes front, cabin flows wind data, wherein, flows wind data and includes flowing wind speed, flowing true wind direction and flow relative cabin
Wind direction.
Divide module 702, for described true wind direction interval by a predetermined angle of flowing is carried out a point sector, flow described
Wind speed carries out segmentation by predetermined wind speed interval.
First grouping module 703, for carrying out being grouped for the first time to the described wind data that flows according to point sector and segmentation.
Yaw error seismic responses calculated module 704, for according to the wind direction often flowing relative cabin described in group data
The yaw error calculating correspondence group obtains yaw error Optimized model.
Yaw error optimizes module 705, for when described wind driven generator unit yaw system to be optimized optimizes, will survey
Flow wind speed and flow true wind direction and input in described yaw error Optimized model, the yaw error that coupling is corresponding, this is right
The yaw error answered is adapted in the input of described yaw control system.
In another embodiment of the present invention, as described in Figure 8, described system also includes,
Second grouping module 706, for carrying out second time by low wind speed section, middle wind speed section and high wind speed section to carrying out flow data
Packet;Wherein, low wind speed range is V0< V≤V1, middle wind speed range is V1< V≤V2, high wind speed scope is V > V2, V2For specified
Wind speed, V0For threshold wind velocity, V1For low wind speed section threshold value, as 6m/s, V for flowing wind speed.Start angle and optimizing model computation module
707, start angle and optimizing criterion for the often group of second time packet is set different driftages, obtain starting angle and optimizing model;
Start angle and optimizing module 708, when described yaw system optimizes, flowing described actual measurement described in wind speed input
Starting in angle and optimizing model, driftage corresponding to wind speed of flowing obtaining described actual measurement starts angle, and to be surveyed flows relative cabin
Wind direction reach described correspondence driftage start angle time, start described yaw system action.
The present invention wind power generating set of different model can be in difference to flow take under wind conditions different excellent
Change strategy, it is possible to the yaw error, the driftage that control operation wind generating set yaw start angle and carry out individual character optimization, improve
Yaw error and the optimization precision at startup angle, improve the power characteristic of unit while improving yaw system reliability,
Improve the generated energy of unit.
The above is merely to illustrate the technical scheme of the application, any those of ordinary skill in the art all can without prejudice to
Under the spirit and the scope of the present invention, above-described embodiment is modified and changes.Therefore, the scope of the present invention should regard
Right is as the criterion.