CN111350638A - Method and device for calculating power generation loss of wind generating set - Google Patents

Method and device for calculating power generation loss of wind generating set Download PDF

Info

Publication number
CN111350638A
CN111350638A CN201811571791.8A CN201811571791A CN111350638A CN 111350638 A CN111350638 A CN 111350638A CN 201811571791 A CN201811571791 A CN 201811571791A CN 111350638 A CN111350638 A CN 111350638A
Authority
CN
China
Prior art keywords
wind
power
wind speed
power value
calculating
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201811571791.8A
Other languages
Chinese (zh)
Inventor
李向楠
毕占磊
石峰毓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Goldwind Science and Creation Windpower Equipment Co Ltd
Original Assignee
Beijing Goldwind Science and Creation Windpower Equipment Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Goldwind Science and Creation Windpower Equipment Co Ltd filed Critical Beijing Goldwind Science and Creation Windpower Equipment Co Ltd
Priority to CN201811571791.8A priority Critical patent/CN111350638A/en
Publication of CN111350638A publication Critical patent/CN111350638A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2260/00—Function
    • F05B2260/82—Forecasts
    • F05B2260/821—Parameter estimation or prediction

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Water Supply & Treatment (AREA)
  • Marketing (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Sustainable Energy (AREA)
  • Sustainable Development (AREA)
  • Public Health (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Combustion & Propulsion (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Control Of Eletrric Generators (AREA)
  • Wind Motors (AREA)

Abstract

The disclosure provides a method and a device for calculating the power generation loss of a wind generating set. The method comprises the following steps: calculating a first power curve and a second power curve of the wind generating set at different wind speed sections according to the state data of the wind generating set; calculating a predicted power value of the wind power plant using a first power curve and a second power curve of the wind power plant according to fault status data of the wind power plant; calculating a power generation loss of the wind turbine generator set based on the predicted power value. By using the method and the device disclosed by the invention, the accuracy of data and the accuracy of power generation amount statistics can be improved.

Description

Method and device for calculating power generation loss of wind generating set
Technical Field
The invention relates to the technical field of wind power generation, in particular to a method and a device for calculating the power generation loss of a wind generating set.
Background
The generated energy loss statistics is one of important quantitative indexes for evaluating the economy of the wind power plant. At present, the conventional method for estimating the power generation loss is to estimate the power generation loss in a fault state by combining ten-minute data of a wind generating set with a specific power curve of the wind generating set.
However, in the conventional power generation amount loss estimation method, since ten-minute data, that is, a ten-minute power mean value and a wind speed mean value representing the entire ten-minute power value and wind speed value, respectively, are used, a large statistical error of the data has been caused and the wind speed value used is a wind speed value converted by a transfer function of a main control program rather than an actual wind speed value obtained by a wind turbine generator set, calculating the power generation amount loss using the compressed data increases the statistical error of the power generation amount loss.
In addition, the conventional power generation amount loss estimation method generally uses only one specific power curve to calculate the power loss value, but due to the randomness of data, an under-fit condition occurs, so that all wind generating sets of the model are not evaluated by only one power curve when the power loss value is calculated, and considering that the condition of power limitation exists in the operation process of a wind power plant, if the power loss value is calculated by only the power curve, the calculated power value may not meet the real condition, so that the power generation amount loss estimation needs to be performed on the target wind generating set by referring to other wind generating set data in the same power plant.
Disclosure of Invention
Exemplary embodiments of the present invention provide a method of calculating a power generation loss of a wind turbine generator system and an apparatus thereof, which solve at least the above technical problems and other technical problems not mentioned above and provide the following advantageous effects.
An aspect of the present invention provides a method of calculating a loss of power generation of a wind turbine generator system, which may include: calculating a first power curve and a second power curve of the wind generating set at different wind speed sections according to the state data of the wind generating set; calculating a predicted power value of the wind power plant using a first power curve and a second power curve of the wind power plant according to fault status data of the wind power plant; and calculating the power generation loss of the wind turbine generator set based on the predicted power value.
The step of calculating a first power curve and a second power curve of the wind park from the status data of the wind park may further comprise: acquiring state data of a wind generating set in a wind power plant, wherein the type of the wind generating set is the same as that of the wind generating set; calculating a similar output unit weight matrix for the wind power generator units using the status data of the same type of wind power generator units, wherein the step of calculating the predicted power value for the wind power generator units may comprise calculating the predicted power value for the wind power generator units using one of the similar output unit weight matrix, the first power curve and the second power curve of the wind power generator units.
The step of calculating the first and second power curves of the wind park may comprise: determining a wind speed critical point according to the wind speed value and the power value in the state data; performing data binning based on the wind speed values in the state data to calculate a wind speed value and a first power value of each wind speed interval of the wind generating set; dividing the calculated wind speed value into a first wind speed section and a second wind speed section according to the determined wind speed critical point; and calculating a first power curve and a second power curve of the wind park, respectively, based on the wind speed value in the first wind speed segment, the wind speed value in the second wind speed segment and the first power value.
The step of determining a wind speed critical point may comprise: obtaining a third power curve according to the calculated wind speed value and the first power value; and a wind speed value having a large deviation from the third power curve among the calculated wind speed values is taken as a wind speed critical point.
The step of calculating the output similar unit weight matrix of the wind generating set may comprise: calculating, for each wind park, a second power value within the respective wind speed interval using the first power curve and the second power curve of the respective wind park based on the wind speed value within each wind speed interval; calculating a first ratio of a second power value in each wind speed interval to a second power value of the wind generating set in the corresponding wind speed interval in the same state aiming at the wind generating set; when the first ratio is within a preset range, keeping the first ratio at an original value, and when the first ratio is not within the preset range, setting the first ratio to be zero; and forming a similar output unit weight matrix of the wind generating set according to the set first ratio.
The step of calculating the predicted power value of the wind park may comprise: selecting part of normal state data of the wind generating set, and making a normal state in the normal state data into a fault state; calculating a third power value of the wind power plant using one of the first power curve, the second power curve, and the similar-output plant weight matrix of the wind power plant based on the wind speed value in the normal state data; calculating a first error based on the third power value and a corresponding actual power value in the normal state data; it is determined whether the first error satisfies the prediction accuracy. When the first error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using one of a first power curve, a second power curve and an output similar set weight matrix of the wind generating set.
The method may further comprise: selecting a fourth power value calculated using the first power curve and the second power curve from the third power values when the first error does not satisfy the prediction accuracy; calculating a second ratio of the fourth power value to a corresponding actual power value in the normal state data; determining an optimal parameter factor according to the second ratio; a fifth power value is calculated based on the fourth power value and the optimal parameter factor. Calculating a second error based on the fifth power value and a corresponding actual power value in the normal state data; it is determined whether the second error satisfies the prediction accuracy. When the second error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the first power curve, the second power curve, the optimal parameter factor and the output similarity set weight matrix of the wind generating set.
The method may further comprise: calculating a sixth power value by a machine learning algorithm based on the third power value when the second error does not satisfy the prediction accuracy; calculating a third error based on the sixth power value and the corresponding actual power value in the normal state data; determining whether the third error satisfies the prediction accuracy; when the third error does not satisfy the prediction accuracy, continuing to calculate the sixth power value using the machine learning algorithm until the third error satisfies the prediction accuracy. When the third error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the power value of the wind generating set calculated by one of the first power curve, the second power curve and the output similar set weight matrix of the wind generating set in combination with a machine learning algorithm.
The step of acquiring fault state data of the wind generating set may comprise: and performing up-sampling processing on the state data of the wind generating set, and selecting fault state data of the wind generating set from the state data subjected to up-sampling processing.
The step of up-sampling the state data of the wind generating set may include up-sampling interpolation fitting of the state data of the wind generating set.
The step of calculating the predicted power value of the wind park may comprise: acquiring the state data of the wind generating set of which the instantaneous wind speed in the fault state data of the wind generating set is within a wind speed interval in a similar set weight matrix; determining the number M of non-fault wind generating sets in a non-fault state in the wind generating sets; and determining whether to use the first and second power curves or the similar output unit weight matrix to calculate the predicted power value of the wind turbine generator unit according to whether the number M is greater than 0.
The step of determining whether to use the first and second power curves or the similar-contribution group weight matrix to calculate the predicted power value of the wind park according to whether the number M is greater than 0 may comprise: when the number M is greater than 0, determining the number Q of wind generating sets of which the instantaneous wind speed of the non-fault wind generating sets is within the wind speed interval, and calculating a predicted power value of the wind generating sets by using the first and second power curves or the output similar set weight matrix according to whether the number Q is equal to 0 or not; and when the number M is equal to 0, calculating a predicted power value for the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
The step of calculating a predicted power value of the wind park using the first and second power curves or the similar-contribution park weight matrix depending on whether the quantity Q is equal to 0 may comprise: when the number Q is larger than 0, calculating a predicted power value of the wind generating set by using the ratio in the wind speed interval in the output similar set weight matrix of the wind generating set and the instantaneous power value of the wind generating set of which the instantaneous wind speed of the non-fault wind generating set is in the wind speed interval; and when the quantity Q is equal to 0, calculating a predicted power value of the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
In calculating the predicted power value, the method may also perform power-limited correction on the predicted power value.
When the first power curve, the second power curve and the output similar unit weight matrix of the wind generating set are calculated, the method can also count the state data according to seasons.
Another aspect of the present invention is to provide an apparatus for calculating a power generation amount loss of a wind turbine generator system, which may include a data processing module and a power generation amount calculation module. The data processing module can calculate a first power curve and a second power curve of the wind generating set at different wind speed sections according to state data of the wind generating set; and the power generation amount calculation module may calculate a predicted power value of the wind turbine generator set using the first power curve and the second power curve of the wind turbine generator set according to the fault state data of the wind turbine generator set, and calculate the power generation amount loss of the wind turbine generator set based on the predicted power value.
The data processing module can acquire state data of a wind generating set of the same type as the wind generating set in the wind power plant; and calculating a similar output unit weight matrix of the wind generating sets by using the state data of the wind generating sets of the same type, wherein the power generation amount calculating module can calculate the predicted power value of the wind generating sets by using one of the similar output unit weight matrix, the first power curve and the second power curve of the wind generating sets.
When calculating the first and second power curves of different wind speed sections of the wind generating set, the data processing module may determine a wind speed critical point according to the wind speed value and the power value in the state data; performing data binning based on the wind speed values in the state data to calculate a wind speed value and a first power value of each wind speed interval of the wind generating set; dividing the calculated wind speed value into a first wind speed section and a second wind speed section according to the determined wind speed critical point; and calculating a first power curve and a second power curve of the wind park, respectively, based on the wind speed value in the first wind speed segment, the wind speed value in the second wind speed segment and the first power value.
The data processing module can also obtain a third power curve according to the calculated wind speed value and the first power value; and the wind speed value with larger deviation from the third power curve in the state data is taken as a wind speed critical point.
When calculating the output similar unit weight matrix of the wind generating sets, the data processing module may further calculate, for each wind generating set, a second power value within the corresponding wind speed interval using the first power curve and the second power curve of the corresponding wind generating set based on the wind speed value within each wind speed interval; calculating a first ratio of a second power value in each wind speed interval to a second power value of the wind generating set in the corresponding wind speed interval in the same state aiming at the wind generating set; when the first ratio is within a preset range, keeping the first ratio at an original value, and when the first ratio is not within the preset range, setting the first ratio to be zero; and forming a similar output unit weight matrix of the wind generating set according to the set first ratio.
In calculating the predicted power value, the power generation amount calculation module may perform the following operations: selecting part of normal state data of the wind generating set, and making a normal state in the normal state data into a fault state; calculating a third power value of the wind power plant using one of the first power curve, the second power curve, and the similar-output plant weight matrix of the wind power plant based on the wind speed value in the normal state data; calculating a first error based on the third power value and a corresponding actual power value in the normal state data; determining whether the first error satisfies a prediction accuracy; when the first error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using one of a first power curve, a second power curve and an output similar set weight matrix of the wind generating set.
In calculating the predicted power value, the power generation amount calculation module may further perform the following operations: selecting a fourth power value calculated using the first power curve and the second power curve from the third power values when the first error does not satisfy the prediction accuracy; calculating a second ratio of the fourth power value to a corresponding actual power value in the normal state data; determining an optimal parameter factor according to the second ratio; a fifth power value is calculated based on the fourth power value and the optimal parameter factor. Calculating a second error based on the fifth power value and a corresponding actual power value in the normal state data; determining whether the second error satisfies the prediction accuracy; when the second error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the first power curve, the second power curve, the optimal parameter factor and the output similarity set weight matrix of the wind generating set.
In calculating the predicted power value, the power generation amount calculation module may further perform the following operations: calculating a sixth power value by a machine learning algorithm based on the third power value when the second error does not satisfy the prediction accuracy; calculating a third error based on the sixth power value and the corresponding actual power value in the normal state data; determining whether the third error satisfies the prediction accuracy; when the third error does not meet the prediction accuracy, continuing to use the machine learning algorithm to calculate the sixth power value until the third error meets the prediction accuracy; when the third error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the power value of the wind generating set calculated by one of the first power curve, the second power curve and the output similar set weight matrix of the wind generating set in combination with a machine learning algorithm.
The data processing module can also perform up-sampling processing on the state data of the wind generating set, and select the fault state data of the wind generating set from the state data after the up-sampling processing.
The data processing module can also perform up-sampling interpolation fitting on the state data of the wind generating set.
In calculating the predicted power value, the power generation amount calculation module may perform the following operations: acquiring the state data of the wind generating set of which the instantaneous wind speed in the fault state data of the wind generating set is within a wind speed interval in a similar set weight matrix; determining the number M of non-fault wind generating sets in a non-fault state in the wind generating sets; and determining whether to use the first and second power curves or the similar output unit weight matrix to calculate the predicted power value of the wind turbine generator unit according to whether the number M is greater than 0.
In calculating the predicted power value, the power generation amount calculation module may further perform the following operations: when the number M is greater than 0, determining the number Q of wind generating sets of which the instantaneous wind speed of the non-fault wind generating sets is within the wind speed interval, and calculating a predicted power value of the wind generating sets by using the first and second power curves or the output similar set weight matrix according to whether the number Q is equal to 0 or not; and when the number M is equal to 0, calculating a predicted power value for the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
In calculating the predicted power value, the power generation amount calculation module may further perform the following operations: when the number Q is larger than 0, calculating a predicted power value of the wind generating set by using the ratio in the wind speed interval in the output similar set weight matrix of the wind generating set and the instantaneous power value of the wind generating set of which the instantaneous wind speed of the non-fault wind generating set is in the wind speed interval; and when the number Q is equal to 0, calculating a predicted power value for the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
The power generation amount calculation module can also carry out power limit correction on the predicted power value.
The data processing module can also count the state data according to seasons.
An aspect of the present invention is to provide a computer-readable storage medium storing a program, characterized in that the program may include instructions for executing the above-described method of calculating a loss in power generation amount of a wind turbine generator set.
An aspect of the invention provides a computer comprising a readable medium having a computer program stored thereon, characterized in that the computer program comprises instructions for executing the above-mentioned method of calculating a loss of power production of a wind park.
Based on the method and the device for calculating the power generation amount loss of the wind generating set, the two power curves of the wind generating set can be calculated according to the wind speed in different seasons, the weight matrix of the output similar set is calculated at the same time, the two power curves are decided together to calculate the power loss value of the wind generating set, and the estimation precision of the power loss is improved. By performing up-sampling processing on the state data of the wind generating set and performing interpolation fitting on the basis of keeping the original real power value, the continuity of power change is met, and the statistical accuracy of the power generation loss is improved. In addition, according to the state conditions of each wind field and the wind generating set, various power generation loss calculation methods are provided, a more appropriate method is selected for power value filling, the problem of high estimation error caused by only using one calculation method is solved, and the estimation accuracy of the power generation loss is improved.
Drawings
The above features and other objects, features and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flow chart of a method for calculating a loss of power generation in accordance with an exemplary embodiment of the present disclosure;
FIG. 2 is a flow chart of a method for calculating a loss of power generation in accordance with another exemplary embodiment of the present disclosure;
FIG. 3 is a flow chart of a method for calculating first and second power curves according to an exemplary embodiment of the present disclosure;
FIG. 4 is a flowchart of a method for calculating a contribution similarity group weight matrix according to an exemplary embodiment of the present disclosure;
FIG. 5 is a flow chart of a method of calculating a power value using one of a first power curve, a second power curve, and a similar-contribution crew weight matrix, according to an exemplary embodiment of the present disclosure;
FIG. 6 is a block diagram of an apparatus for calculating power generation loss according to an exemplary embodiment of the present disclosure;
FIG. 7 is a diagram of a third power curve, according to an example embodiment of the present disclosure;
fig. 8 and 9 are diagrams of a first power curve and a second power curve according to an exemplary embodiment of the present disclosure.
Detailed Description
Reference will now be made in detail to the exemplary embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. The embodiments are described below in order to explain the present invention by referring to the figures. It is to be understood that the described embodiments are merely a subset of the disclosed embodiments and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present application.
In the present disclosure, terms including ordinal numbers such as "first", "second", etc., may be used to describe various elements, but these elements should not be construed as being limited to only these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and vice-versa, without departing from the scope of the present disclosure.
Before setting forth the inventive concepts of the present disclosure, a related description is made of terms employed in the present disclosure.
The state data is instantaneous state data of the wind turbine generator system, is a time series data source, and is usually 7s data (10 s, 20s data, etc.), that is, the interval between two records is approximately 7s, each record includes instantaneous quantities of each state of the wind turbine generator system, for example, analog quantities including wind speed, rotation speed, power, temperature, etc., and digital quantities including left yaw motion, right yaw motion, power limit flag, etc.
The power curve refers to a corresponding curve of the output power and the wind speed of the wind generating set and is used for describing a functional relation between the power output and the wind speed of the wind generating set.
The unit output similarity refers to the output similarity degree of two wind generating sets in power in a certain wind speed interval, the measurement standard is the difference absolute value of the power of the two wind generating sets in the same wind speed interval, and the smaller the difference absolute value is, the higher the output similarity of the two wind generating sets is.
The output similar unit weight matrix refers to the output similar unit weight matrix of the specific wind generating set, which is formed by the ratio of the corresponding power value of the power curve of the specific wind generating set in each wind speed interval to the power value of the power curves of other wind generating sets in the wind farm in the same wind speed interval.
Fig. 1 is a flowchart of a method for calculating a loss in power generation according to an exemplary embodiment of the present disclosure.
Referring to fig. 1, in step S101, a first power curve and a second power curve of a wind turbine generator set at different wind speed sections are calculated according to state data of the wind turbine generator set. The state data for the 7s interval may be used as a data source for subsequent calculations. Here, the first power curve and the second power curve are divided according to the magnitude of the wind speed value. Specifically, a wind speed critical point is determined according to a wind speed value and a power value in state data, data binning is performed on the basis of the wind speed value in the state data to calculate a wind speed value and a first power value of each wind speed interval of the wind generating set, the calculated wind speed value is divided into a first wind speed section and a second wind speed section according to the determined wind speed critical point, and a first power curve and a second power curve of the wind generating set are respectively calculated on the basis of the wind speed value in the first wind speed section, the wind speed value in the second wind speed section and the first power value. The method of calculating the first power curve and the second power curve will be described in detail below with reference to fig. 3.
In step S102, a predicted power value of the wind park is calculated using the first power curve and the second power curve of the wind park according to fault status data of the wind park. After obtaining the first power curve and the second power curve, a predicted power value of the wind turbine generator set is calculated using the instantaneous wind speed in the fault state data and the first and second power curves of the wind turbine generator set. For example, when the instantaneous wind speed in the fault status data is less than the critical point, the first power curve (the small wind speed power curve) may be used to calculate the predicted power value at this instantaneous wind speed.
In step S103, the power generation amount loss of the wind turbine generator set is calculated based on the predicted power value. After the predicted power value is calculated, power limit correction may be performed on the predicted power value, and the corrected predicted power value may be used to calculate the power generation amount loss. Next, a detailed flowchart of a method for calculating a loss of power generation of a wind turbine generator set will be described in detail with reference to fig. 2.
Fig. 2 is a detailed flowchart of a method for calculating a loss in power generation according to an exemplary embodiment of the present disclosure.
Referring to fig. 2, in step S201, status data of a specific wind park in the wind farm and a wind park of the same type as the specific wind park is acquired.
In step S202, the acquired state data of the wind turbine generator system is screened. In the process of screening the state data, the data can be screened based on the method specified by the IEC61400-12-1 standard, such as deleting fault data, data during operation and maintenance, and keeping normal data. For example, unhealthy status data is deleted according to the power generation status and the data availability status of the wind turbine generator set to retain healthy status data.
In step S203, a first power curve and a second power curve of each wind generating set in different wind speed sections are calculated according to the screened state data of the wind generating sets.
How to calculate the first and second power curves of the wind park will be described in detail below with reference to fig. 3. Fig. 3 is a flowchart of a method for calculating first and second power curves according to an exemplary embodiment of the present disclosure. In the present disclosure, the method flow of fig. 3 may be employed to calculate first and second power curves for each wind generating set in a wind farm.
Referring to fig. 3, in step S301, data binning is performed based on the wind speed values in the status data. And grouping the screened state data according to the wind speed data in the state data according to the length of the preset interval to obtain a plurality of wind speed intervals. For example, the screened state data may be grouped according to the wind speed values in the state data with 0.5m/s as a predetermined section length, thereby obtaining a plurality of wind speed sections with a section length of 0.5 m/s. However, the above examples are merely exemplary, and the present disclosure is not limited thereto.
In step S302, a wind speed value and a first power value of the wind turbine generator set in each wind speed interval are calculated according to the screened state data of the wind turbine generator set. For example, for each wind speed interval divided in step S302, a wind speed mean value and a power mean value within each wind speed interval are calculated using the state data of the wind turbine generator set. Alternatively, a plurality of original wind speed values and a plurality of percentile of the original power values included in each wind speed interval may be used as the wind speed value and the first power value in the corresponding wind speed interval, respectively.
In step S303, a wind speed critical point is determined according to the wind speed value and the power value in the status data. The wind speed critical point may be determined by comparing a curve fitted using the wind speed values and the first power value within each wind speed interval with discrete points formed using the wind speed values and the first power value within each wind speed interval. Specifically, a data fitting method is used for fitting a power curve of the wind generating set according to the calculated wind speed value and the first power value in each wind speed interval. For example, a 10 th order polynomial fitting method may be used to perform a curve fitting on the wind speed value and the first power value in each wind speed interval of a certain wind turbine generator system, and a fitted polynomial curve is drawn, as shown in fig. 7, wherein the abscissa represents the wind speed and the ordinate represents the power. The curve in fig. 7 is a power curve fitted with the wind speed values and the first power values in each wind speed interval of a certain wind turbine generator set, and the discrete points in fig. 7 are points directly plotted with the wind speed values and the first power values in each wind speed interval, respectively. As can be seen from fig. 7, the deviation of the fitted function curve from the true value is large due to the presence of the outlier (discrete point having a large deviation from the fitted power curve), and therefore, the outlier may be taken as the wind speed critical point, for example, 12.25m/s may be taken as the wind speed critical point.
In step S304, the calculated wind speed value is divided into a first wind speed segment and a second wind speed segment according to the determined wind speed critical point. Specifically, a wind speed value at which the calculated wind speed value is less than or equal to the wind speed critical point is classified into a first wind speed section, and a wind speed value at which the calculated wind speed value is greater than the wind speed critical point is classified into a second wind speed section. For example, in FIG. 7, 12.25m/s is taken as the wind speed critical point, a wind speed value less than or equal to 12.25m/s is taken as a small wind speed section (i.e., a first wind speed section), and a wind speed value greater than 12.25m/s is taken as a large wind speed section (i.e., a second wind speed section).
In step S305, a first power curve of the wind park is calculated based on the wind speed values in the first wind speed segment and the corresponding first power values. Specifically, after classifying the calculated wind speeds, a first power curve is calculated using a polynomial fitting method (e.g., 10 th order polynomial fitting method) based on a wind speed value in a first wind speed section and a first power value in a wind speed interval corresponding to the wind speed value, as shown in fig. 8, where fig. 8 shows the first power curve (power curve corresponding to a small wind speed) with the abscissa representing the wind speed and the ordinate representing the power.
In step S306, a second power curve of the wind park is calculated based on the wind speed values in the second wind speed segment and the respective first power values. Specifically, after classifying the calculated wind speeds, a polynomial fitting method (e.g., 10 th order polynomial fitting method) is used to calculate a second power curve based on the wind speed values in the second wind speed segment and the first power values in the wind speed interval corresponding to the wind speed values, as shown in fig. 9, where fig. 9 shows the second power curve (power curve corresponding to a large wind speed) with the abscissa representing the wind speed and the ordinate representing the power. As can be seen from fig. 8 and 9, the first power curve and the second power curve fit well, so that the error of the value calculated from the fitted power curve function from the true value is also small.
In the present disclosure, when calculating the first and second power curves of the wind park, the status data may first be counted quarterly, and then the first power curve and the second power curve are calculated based on the status data of each quarterly, i.e. the first and second power curves for the wind park for each quarterly are calculated, respectively.
In step S204, the filtered state data is used to calculate a similar output unit weight matrix for the specific wind generating set. How to calculate the contribution similarity group weight matrix for a particular wind park will be described in detail below with reference to fig. 4.
FIG. 4 is a flowchart of a method for calculating a contribution similarity group weight matrix according to an exemplary embodiment of the present disclosure. In the present disclosure, the method flow of fig. 4 may be employed to calculate a similar output unit weight matrix for each wind generating set in the wind farm.
Referring to fig. 4, in step S401, a second power value within each wind speed interval is calculated for each wind park using the first and second power curves of the respective wind park based on the wind speed value within each wind speed interval. For example, taking wind park a as an example, a second power value within each wind speed interval is calculated from the status data of each wind speed interval of wind park a in combination with the respective first and second power curves.
In step S402, a first ratio of the second power value of the specific wind turbine generator set in each wind speed interval to the second power value of the wind turbine generator set of the same type operating in the same state in the corresponding wind speed interval is calculated. For example, assuming that when calculating the first ratio for the first wind speed interval of the wind park a, the first ratio (i.e., weight) of the wind parks of the same type as the wind park a may be calculated according to equation (1), respectively:
Wi＝PA/Pi(1)
wherein, WiRepresenting a first ratio, P, of the ith wind park within a first interval of wind speedsARepresenting a second power value, P, of the wind energy plant A in the first interval of wind speedsiAnd indicating a second power value of the ith wind generating set in the first wind speed interval.
In step S403, it is determined whether the calculated first ratio is within a preset range. The value of the preset range may be set according to the experience of a designer. For example, the range of the weight coefficient may be set to be between 0.9 and 1.1. When the calculated first ratio is within the preset range, the step S404 is entered, and the original first ratio is maintained; when the calculated first ratio is not within the preset range, the process proceeds to step S405, and the first ratio is set to a zero value, i.e., a null value.
In step S406, an output similar unit weight matrix of the wind generating set is formed according to the set first ratio. A similar contribution unit weight matrix for the particular wind generating set may be obtained based on the first ratio in accordance with the wind speed interval.
As an example, it is assumed that at a certain moment, the model of the wind turbine generator set a is 2300, in the wind speed interval 4.25m/s to 4.75m/s, in the no-fault state, in the infinite power state and in the power value 425KW, the model of the wind turbine generator set T is 2300, in the wind speed interval 4.25m/s to 4.75m/s, in the no-fault state, in the infinite power state and in the power value 435KW, and thus it can be seen that the wind turbine generator set a and the wind turbine generator set T belong to the same model, in the same operating state, and the power value ratio therebetween is 0.97, which is between 0.9 and 1.1 (i.e., in the preset range), and therefore, the weighting factor of the wind turbine generator set a with respect to the wind turbine generator set T is 0.97. If the following condition is not satisfied: and when the wind speed ranges are the same, and the weight coefficient (first ratio) is within a preset range, setting the first ratio as a null value. And forming a similar output unit weight matrix of the wind generating set by using the first ratio calculated according to the method.
According to the embodiment of the disclosure, when the weight matrix of the output similarity unit is calculated, second-level up-sampling interpolation fitting can be carried out on the screened state data. Specifically, parameters such as time, wind speed, power, temperature, and fault state are extracted from the state data, and up-sampling processing with a time dimension of second is performed on the parameters. Here, parameters such as wind speed, power, and temperature are analog parameters, and when the up-sampling process is performed, the 7s state data may be up-sampled into second-order state data using a linear interpolation fitting method. However, the fault state in the state data is a digital quantity, and the up-sampled fault state can be set according to the following manner. For example, if the fault state of a specific wind generating set is 0 at 10:20:30 and 1 at 10:20:37, the fault state of 10:20:30-10:20:33 is set to 0 and the fault state of 10:20:34-10:20:37 is set to 1 after the up-sampling process. By performing up-sampling processing on the 7s state data, the continuity of power change is better met, and the accuracy of the data is improved.
Alternatively, the up-sampling process may be performed on the state data acquired in step S201 or S202. In the up-sampling process using the directly acquired state data, first, time, wind speed, power, temperature, and failure state in the 7s state data are up-sampled in seconds, and the process is the same as the up-sampling process described above. That is to say, in the process of calculating the first power curve, the second power curve and the weight matrix of the output similar unit, the filtered state data may be directly used for calculation, or the obtained state data may be subjected to up-sampling processing, and then the state data subjected to up-sampling processing is used for calculation.
In step S205, a part of the normal state data of the specific wind turbine generator system is selected, and the normal state in the normal state data is made into a fault state. The method for replacing the normal state of a specific wind turbine generator set with the fault state can adopt a random method. For example, a fault-free state data of a specific wind turbine generator set within 6 continuous hours is selected, 5 segments of continuous data of 10 to 20 minutes are randomly selected to replace the normal state of the data with the fault state, a random fault is simulated by using the mode, and a power value is estimated according to the selected normal state data to be compared with an actual power value.
In step S206, a third power value of the particular wind park is calculated using one of the first power curve, the second power curve and the similar park weight matrix of the particular wind park based on the wind speed in the normal state data. How to determine whether to use the first power curve, the second power curve or the similar-contribution unit weight matrix to calculate the power value will be described below with reference to fig. 5.
Fig. 5 is a flowchart of a method of calculating a power value using one of a first power curve, a second power curve, and a similar-contribution crew weight matrix, according to an exemplary embodiment of the present disclosure. The method flow shown in fig. 5 is applicable to calculating the third power value as well as the predicted power value.
Referring to fig. 5, in step S501, status data of a wind turbine generator set in which an instantaneous wind speed in fault status data of a specific wind turbine generator set is within a wind speed interval in a similar-output-set weight matrix thereof is acquired. It should be noted that, when the third power value is calculated, the normal state in the normal state data is made into the fault state, so the normal state data made into the fault state is the fault state data of the specific wind turbine generator set in step S501, and when the predicted power value is calculated, the real fault state data of the specific wind turbine generator set is used.
In step S502, the number M of non-faulty wind power generator sets in a non-faulty state among the wind power generator sets is determined. Specifically, after obtaining the state data of the wind generating sets with similar output at each fault moment of a specific wind generating set, it is first determined whether the wind generating sets exist in a non-fault state, that is, the number M of the non-fault wind generating sets in the non-fault state.
In step S503, it is determined whether the number M of non-faulty wind turbine generators in the non-faulty state is greater than zero. When the number M is larger than 0, it proceeds to step S504. In step S504, the number Q of wind park sets for which the instantaneous wind speed of the non-faulty wind park set is within the respective wind speed interval is determined. In particular, after determining the non-faulty wind energy installations, it needs to be further determined whether the instantaneous wind speeds of these non-faulty wind energy installations are within the wind speed interval to be calculated. For example, when calculating the predicted power value within the first wind speed interval for which a particular wind park is in a fault condition, it is necessary to determine whether the instantaneous wind speed of the non-faulty wind park is within the first wind speed interval.
When the number M is equal to 0, proceeding to step S507, a predicted power value for the particular wind park is calculated using the instantaneous wind speed and the corresponding first or second power curve. Specifically, when the number M is 0, i.e., when there is no status data of the non-faulty wind turbine generator set, the predicted power value of the specific wind turbine generator set, i.e., the power value lost by the specific wind turbine generator set, may be calculated using the instantaneous wind speed in the fault status of the specific wind turbine generator set and the first power curve or the second power curve corresponding to the time. Here, when the instantaneous wind speed at that time belongs to the first wind speed section (i.e., the wind speed at that time belongs to the small wind speed section), the predicted power value at that instantaneous wind speed is calculated using the first power curve. When the instantaneous wind speed at that time belongs to the second wind speed segment (i.e., the wind speed at that time belongs to the large wind speed segment), the predicted power value at that instantaneous wind speed is calculated using the second power curve.
It should be noted that when the number M of non-faulty wind park in non-faulty state is 0, it is not necessary to use the contribution similarity park weight matrix to calculate the power loss of a particular wind park.
After determining the number Q, it is determined whether the number Q is greater than zero in step S505. When the number Q is larger than 0, it proceeds to step S506. In step S506, the power value of the specific wind turbine generator set is calculated by using the ratio in the corresponding wind speed interval in the output similar set weight matrix of the specific wind turbine generator set and the instantaneous power value of the wind turbine generator set of which the instantaneous wind speed of the non-fault wind turbine generator set is in the wind speed interval. Specifically, the state data of the wind generating sets with the instantaneous wind speed of the non-fault wind generating set within the wind speed interval, that is, the state data is taken to be the state data of the wind generating sets with the same type of wind generating sets conforming to the non-fault state, being in the power limiting state and having the instantaneous wind speed of the non-fault wind generating set within the wind speed interval in which the instantaneous wind speed of the specific wind generating set in the fault state is located (the condition may be referred to as a first trigger condition), after the state data is taken, the power loss value of each moment of the specific wind generating set may be calculated according to equation (2):
Figure BDA0001915733040000151
wherein, PlossRepresenting the power loss value, w, at a certain moment of the particular wind energy plantiRepresenting the weight value (i.e., the first ratio), P, within the corresponding wind speed interval in the output-like unit weight matrix corresponding to that momentiAnd representing the instantaneous power value corresponding to the instantaneous wind speed of the ith wind generating set meeting the first triggering condition.
When the number Q is equal to 0, proceeding to step S507, a predicted power value for the particular wind park is calculated using the instantaneous wind speed and the corresponding first or second power curve. For example, when the first trigger condition is not met, the power loss of the specific wind turbine generator set can be calculated by using a first or second power curve of the specific wind turbine generator set corresponding to a certain moment and substituting the instantaneous wind speed of the certain moment into the corresponding first or second power curve.
Here, if there is a similar output wind power generation unit that meets the first trigger condition among the similar output wind power generation units in the non-fault state, the power loss of the specific wind power generation unit is calculated using the similar output unit weight matrix. Different methods are used according to different conditions to calculate the power loss of a specific wind park at each fault moment.
Referring back to fig. 2, in step S207, a first error is calculated based on the calculated third power value and the corresponding actual power value in the normal state data. For example, the first error may be calculated using equation (3) below:
Figure BDA0001915733040000152
wherein, PrealiRepresenting the actual power value P of a specific wind generating set in the ith wind speed intervalestiAnd representing the predicted power value of the specific wind generating set in the ith wind speed interval.
In step S208, the calculated first error is compared with the prediction accuracy to determine whether the first error satisfies the prediction accuracy. The prediction accuracy may be set according to the working experience of the designer. For example, the prediction accuracy may be a confidence interval that 90% of the probability falls within plus or minus 10%.
When the first error satisfies the prediction accuracy, proceeding to step S209, a predicted power value of the particular wind park is calculated using one of the first power curve, the second power curve, and the similar-output park weight matrix of the particular wind park based on the fault status data of the particular wind park. The calculation process of step S209 may be calculated by referring to the method of fig. 5, and is not described herein again.
When the first error does not satisfy the prediction accuracy, proceeding to step S210, a fourth power value calculated using the first power curve and the second power curve is selected from the third power values. When the third power value is calculated, firstly, the weight matrix of the output similarity unit is used for calculating the predicted power value, and if the first trigger condition is not met, the first power curve and the second power curve are used for calculating the predicted power value. The fourth power value is a predicted power value calculated from the third power value using the first and second power curves.
In step S211, a second ratio of the fourth power value to the corresponding actual power value in the normal state data is calculated. I.e. the quotient of the actual power value and the predicted power value at each time instant is calculated using the predicted power values obtained from the first and second power curves, which may be named parameter factor herein.
In step S212, an optimal parameter factor is determined according to the second ratio. For example, the calculated parameter factors are traversed, and a certain quantile of 1 to 100 of the parameter factors at each moment is selected as the optimal parameter factor.
In step S213, a fifth power value is calculated based on the fourth power value and the optimum parameter factor. And after the optimal parameter factors are determined, multiplying the fourth power values by the optimal parameter factors respectively to obtain fifth power values. In addition, when calculating the fifth power value, the noise reduction process may be performed on the calculated fifth power value by traversing the normalized cutoff frequency to determine an optimal normalized cutoff frequency.
In step S214, a second error is calculated based on the fifth power value and the corresponding actual power value in the normal state data. Specifically, the second error is calculated using equation (3) based on the predicted power value calculated using the similar-to-output unit weight matrix in the fifth power value and the third power value and the corresponding actual power value in the normal-state data.
In step S215, it is determined whether the calculated second error satisfies the prediction accuracy. Here, the prediction accuracy is the same as that in step S208. When the second error satisfies the prediction accuracy, proceeding to step S216, a predicted power value of the specific wind turbine generator set is calculated based on the fault state data of the specific wind turbine generator set using the first power curve, the second power curve, the output similar set weight matrix, and the optimal parameter factor of the specific wind turbine generator set. Specifically, the predicted power value is first calculated using the method shown in fig. 5, and then the predicted power value calculated using the first and second power curves is multiplied by the optimal parameter factor to obtain the final predicted power value, while the predicted power value calculated using the output similar unit weight matrix is not required to be multiplied by the optimal parameter factor.
When the second error does not satisfy the prediction accuracy, it proceeds to step S217. In step S217, a sixth power value is calculated by a machine learning algorithm based on the third power value. For example, a lightgbm library in python is called, sampling data of a sample stored previously is used as a training set, state data is sampled 100 times again to be used as a test set, the mean value, the standard deviation, the minimum value, the maximum value, the 0.3 quantile, the 0.6 quantile and the 0.9 quantile of the wind speed and the predicted power value of the wind generating set are used as characteristic quantities, the quotient (namely parameter factor) of the actual power value and the predicted power value is used as a target quantity, model training is carried out by combining the parameters or increasing the sample quantity, and a trained parameter factor is used for correcting (such as multiplying) a third power value to obtain a sixth power value.
In step S218, a third error is calculated based on the sixth power value and the corresponding actual power value in the normal state data. Here, the third error may be calculated using equation (3).
In step S219, it is determined whether the calculated third error satisfies the prediction accuracy. Here, the prediction accuracy is the same as that in step S208. When the third error satisfies the prediction accuracy, proceeding to step S220, a power value of the specific wind turbine generator set calculated using one of the first power curve, the second power curve, and the output-similar-unit weight matrix of the specific wind turbine generator set based on the fault state data of the specific wind turbine generator set is combined with a machine learning algorithm to calculate a predicted power value of the specific wind turbine generator set. Specifically, a machine learning method is adopted, previously stored sample data is used as a training set, re-extracted state data is used as a test set, the mean value, the standard deviation, the minimum value, the maximum value, the 0.3 quantile, the 0.6 quantile, the 0.9 quantile and the like of the wind speed and the predicted power value of the wind generating set are used as characteristic quantities, a parameter factor (namely the quotient of the actual power value and the predicted power value) is used as a target quantity to train the parameter factor, and the trained parameter factor is used for correcting the predicted power value calculated by using one of the first power curve, the second power curve and the output similar set weight matrix.
When the third error does not satisfy the prediction accuracy, returning to step S217, the machine learning algorithm continues to be used to train the parameter factor until the third error between the predicted power value and the actual power value corrected using the trained parameter factor satisfies the prediction accuracy.
In step S221, after the predicted power value is calculated, the predicted power value is subjected to power limit correction. Specifically, if there is a power limit condition for a particular wind turbine generator set in a fault condition, the calculated predicted power value needs to be corrected by the value of the power limit. The state data comprises a limited power identifier and a limited power value, when the state data is read, the limited power identifier and the limited power value need to be read, if the two variables do not exist in the state data, the limited power identifier needs to be added manually, the limited power identifier is set to be 0 to represent the situation that the limited power does not exist, the limited power value is set to be 1.1 times of the power of the machine set type, and the limited power value is 1.1 times of the power of the specific wind generating set type if the power is limited.
In addition, when the limited power is corrected, a specific wind generating set has the rated power of the specific wind generating set, if the calculated predicted power value is more than 1.1 times of the power value of the machine type, the predicted power value is obviously unreasonable, and at the moment, the correction according to 1.1 times of the power of the machine type can be carried out.
In step S222, the corrected predicted power value of the specific wind turbine generator set is used to calculate the power generation amount loss of the specific wind turbine generator set. For example, after calculating the power loss value of a specific wind turbine generator set at all the fault times, the power generation amount loss of the specific wind turbine generator set can be calculated using equation (4):
Figure BDA0001915733040000181
wherein E islossRepresenting the loss of power production, P, of that particular wind power plantlIndicating the power loss value of the specific wind generating set at the ith fault moment.
The above examples are merely illustrative of the method for calculating the power generation loss of a certain wind generating set in a wind farm, and in the present disclosure, the power generation loss of each of the different wind generating sets in the wind farm may be used to obtain the power generation loss of the whole wind farm.
Fig. 6 is a block diagram of an apparatus for calculating a loss in power generation amount according to an exemplary embodiment of the present disclosure.
Referring to fig. 6, the apparatus 600 for calculating the power generation amount loss may include a data processing module 601 and a power generation amount calculation module 602. Each module in the apparatus 600 according to the present disclosure may be implemented by one or more modules, and names of the corresponding modules may vary according to types of apparatuses. In various embodiments, some modules in apparatus 600 may be omitted, or additional modules may also be included. Furthermore, modules according to various embodiments of the present disclosure may be combined to form a single entity, and thus the functions of the respective modules before combination may be equivalently performed.
The data processing module 601 may calculate a first power curve and a second power curve of the wind generating set at different wind speed sections according to the state data of the wind generating set. Specifically, the data processing module 601 may determine a wind speed critical point according to a wind speed value and a power value in the status data, perform data binning to calculate a wind speed value and a first power value for each wind speed interval of the wind turbine generator set based on the wind speed value in the status data, divide the calculated wind speed value into a first wind speed segment and a second wind speed segment according to the determined wind speed critical point, and calculate a first power curve and a second power curve of the wind turbine generator set based on the wind speed value in the first wind speed segment, the wind speed value in the second wind speed segment, and the first power value, respectively. The process of calculating the first and second power curves by the data processing module 601 is the same as the method shown in fig. 4, and will not be described in detail here.
After obtaining the first and second power curves of the wind turbine generator set, the power generation amount calculation module 602 may calculate a predicted power value of the wind turbine generator set using the first power curve and the second power curve of the wind turbine generator set according to the fault state data of the wind turbine generator set, and calculate a power generation amount loss of the wind turbine generator set based on the predicted power value.
Furthermore, the data processing module 601 may obtain status data of wind generating sets of the same type as the specific wind generating set in the wind farm and calculate a similar output set weight matrix for the specific wind generating set using the status data of the wind generating sets of the same type. Specifically, the data processing module 601 may calculate, for each wind turbine generator set, a second power value in each wind speed interval using the first power curve and the second power curve of the corresponding wind turbine generator set based on the wind speed value in each wind speed interval, and calculate, for a specific wind turbine generator set, a first ratio of the second power value in each wind speed interval to a second power value in the corresponding wind speed interval of the wind turbine generator set operating in the same state. And when the first ratio is not in the preset range, setting the first ratio to be zero, and finally forming a similar output unit weight matrix of the wind generating set according to the set first ratio. The process of calculating the weight matrix of the similar-output unit by the data processing module 601 is the same as the method shown in fig. 5, and detailed description thereof is omitted here.
According to an embodiment of the present disclosure, the power generation amount calculation module 602 may calculate the predicted power value of the particular wind generating set using one of the similar-to-output set weight matrix, the first power curve, and the second power curve of the particular wind generating set. For example, the power generation calculation module 602 preferably uses the similar contribution unit weight matrix to calculate the predicted power value, and uses the first and second power curves to calculate the predicted power value when the obtained other wind power generation units do not satisfy the first trigger condition.
To improve the accuracy of the power generation amount loss estimation, different calculation methods may be used to calculate the predicted power value. The normal state data of the wind generating set can be selected firstly, the normal state in the normal state data is made to be a fault state, and the prediction power value is determined by using which calculation method to calculate the prediction power value according to the error between the prediction power value and the actual power value.
Specifically, when the predicted power value is calculated, the power generation amount calculation module 602 selects a part of normal state data of the specific wind turbine generator set, makes a normal state in the normal state data into a fault state, and calculates a third power value of the specific wind turbine generator set by using one of the first power curve, the second power curve and the output similar set weight matrix of the specific wind turbine generator set based on the wind speed value in the normal state data.
Specifically, the power generation amount calculation module 602 acquires state data of the wind turbine generator set in which the instantaneous wind speed in the normal state data of the specific wind turbine generator set, which is made into the fault state, is within a wind speed interval in the output similar set weight matrix; determining the number M of non-fault wind generating sets in a non-fault state in the wind generating sets; and determining whether to use the first and second power curves or the similar output unit weight matrix to calculate the predicted power value of the specific wind turbine generator unit according to whether the number M is greater than 0.
When the number M is greater than 0, determining the number Q of the wind generating sets of which the corresponding instantaneous wind speed of the specific non-fault wind generating set is within the corresponding wind speed interval, and calculating the predicted power value of the specific wind generating set by using the first and second power curves or the output similar set weight matrix according to whether the number Q is equal to 0 or not; and when the number M is equal to 0, calculating a predicted power value for the particular wind park using the respective instantaneous wind speed and the first and second power curves for the particular wind park.
When the quantity Q is larger than 0, calculating a predicted power value of the specific wind generating set by using the ratio of the output similar set weight matrix of the specific wind generating set in the corresponding wind speed interval and the instantaneous power value of the wind generating set of which the corresponding instantaneous wind speed of the non-fault wind generating set is in the corresponding wind speed interval; and when the number Q is equal to 0, calculating a predicted power value for the particular wind park using the respective instantaneous wind speed and the first and second power curves for the particular wind park. The process of calculating the third power value is the same as the method shown in fig. 5, and will not be described in detail here.
Next, the power generation amount calculation module 602 calculates a first error based on the third power value and the corresponding actual power value in the normal state data, and determines whether the first error satisfies the prediction accuracy. When the first error satisfies the prediction accuracy, the power generation amount calculation module 602 calculates a predicted power value of the specific wind turbine generator set using one of the first power curve, the second power curve, and the output-similar-unit weight matrix of the specific wind turbine generator set based on the fault state data of the specific wind turbine generator set. Specifically, the power generation amount calculation module 602 acquires the state data of the wind generating set, in the fault state data of the specific wind generating set, of which the instantaneous wind speed is within the wind speed interval in the output similar set weight matrix; determining the number M of non-fault wind generating sets in a non-fault state in the wind generating sets; and determining whether to use the first and second power curves or the similar output unit weight matrix to calculate the predicted power value of the specific wind turbine generator unit according to whether the number M is greater than 0.
When the number M is greater than 0, determining the number Q of the wind generating sets of which the corresponding instantaneous wind speed of the specific non-fault wind generating set is within the corresponding wind speed interval, and calculating the predicted power value of the specific wind generating set by using the first and second power curves or the output similar set weight matrix according to whether the number Q is equal to 0 or not; and when the number M is equal to 0, calculating a predicted power value for the particular wind park using the respective instantaneous wind speed and the first and second power curves for the particular wind park.
When the quantity Q is larger than 0, calculating a predicted power value of the specific wind generating set by using the ratio of the output similar set weight matrix of the specific wind generating set in the corresponding wind speed interval and the instantaneous power value of the wind generating set of which the corresponding instantaneous wind speed of the non-fault wind generating set is in the corresponding wind speed interval; and when the number Q is equal to 0, calculating a predicted power value for the particular wind park using the respective instantaneous wind speed and the first and second power curves for the particular wind park. The process of calculating the predicted power value is the same as the method shown in fig. 5, and will not be described in detail here.
When the first error does not satisfy the prediction accuracy, the power generation amount calculation module 602 selects a fourth power value calculated using the first power curve and the second power curve from the third power values, calculates a second ratio of the fourth power value to a corresponding actual power value in the normal state data, determines an optimal parameter factor according to the second ratio, and calculates a fifth power value based on the fourth power value and the optimal parameter factor. The power generation amount calculation module 602 then calculates a second error based on the fifth power value and the corresponding actual power value in the normal state data, and determines whether the second error satisfies the prediction accuracy. When the second error satisfies the prediction accuracy, the power generation amount calculation module 602 calculates a predicted power value of the specific wind turbine generator set using the first power curve, the second power curve, the optimal parameter factor, and the output similarity set weight matrix of the specific wind turbine generator set based on the fault state data of the specific wind turbine generator set. This process is the same as step S216, and is not described here again.
When the second error does not satisfy the prediction accuracy, the power generation amount calculation module 602 calculates a sixth power value by a machine learning algorithm based on the third power value, calculates a third error based on the sixth power value and a corresponding actual power value in the normal state data, and determines whether the third error satisfies the prediction accuracy. When the third error does not satisfy the prediction accuracy, the power generation amount calculation module 602 continues to calculate the sixth power value using the machine learning algorithm until the third error satisfies the prediction accuracy, and when the third error satisfies the prediction accuracy, the power generation amount calculation module 602 calculates the predicted power value of the wind turbine generator set using the power value of the wind turbine generator set calculated by one of the first power curve, the second power curve, and the output-similar set weight matrix of the specific wind turbine generator set in combination with the machine learning algorithm based on the fault state data of the specific wind turbine generator set. This process is the same as step S220, and is not described here again.
According to the embodiment of the disclosure, the data processing module 601 may further perform up-sampling processing on the state data of the wind generating set, and select fault state data of a specific wind generating set from the up-sampled state data. For example, the data processing module 601 may perform an up-sampling interpolation fit on the state data of the wind turbine generator set. As an example, when calculating the first power curve, the second power curve, or the weight matrix of the similar output unit, the obtained state data may be up-sampled, and the weight matrix of the similar output unit may be calculated using the up-sampled state data.
After the predicted power value is calculated, the power generation amount calculation module 602 performs power limit correction on the predicted power value. Specifically, if there is a power limit condition for a particular wind generating set under a fault condition, the power generation amount calculation module 602 needs to correct the calculated predicted power value by the value of the power limit. In addition, when power is limited to be corrected, a specific wind generating set has a rated power of the specific wind generating set, if the calculated predicted power value is larger than 1.1 times of the power value of the model, the predicted power value is obviously unreasonable, and at this time, the power generation amount calculation module 602 can correct the power according to 1.1 times of the power of the model.
In the present disclosure, the data processing module 601 may count the status data by quarters. Specifically, when the first power curve, the second power curve and the similar output unit weight matrix are calculated, the first power curve, the second power curve and the similar output unit weight matrix of each quarter can be calculated by using the state data of each quarter, and the predicted power value is calculated by using the first power curve, the second power curve and the similar output unit weight matrix of the corresponding quarter according to which quarter the fault state data of the wind generating set is located, so that the accuracy of the generated energy loss statistics is improved.
The method of calculating the loss of power generation of a wind turbine generator set according to an exemplary embodiment of the present disclosure may be implemented as computer-readable instructions on a computer-readable recording medium or may be transmitted through a transmission medium. The computer readable recording medium is any data storage device that can store data which can be thereafter read by a computer system. Examples of the computer readable recording medium include, but are not limited to, read-only memory (ROM), random-access memory (RAM), compact discs (CD-ROMs), Digital Versatile Discs (DVDs), magnetic tapes, floppy disks, and optical data storage devices. The transmission medium may include a carrier wave transmitted over a network or various types of communication channels. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable instructions are stored and executed in a distributed fashion.
Based on the method and the device for calculating the power generation loss of the wind generating set, the invention introduces two power curves and the weight matrix of the output similar set, so that the multi-decision combined action can be realized during power estimation, and the estimation error caused by the decision only using a single power curve is reduced. In addition, the invention provides a plurality of methods for calculating the power generation loss, reduces the estimation error caused by only using a single method, and improves the statistical accuracy of the power generation loss.
While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the following claims.

Claims (32)

1. A method of calculating a loss in power generation of a wind turbine generator system, the method comprising:
calculating a first power curve and a second power curve of the wind generating set at different wind speed sections according to the state data of the wind generating set;
calculating a predicted power value of the wind power plant using a first power curve and a second power curve of the wind power plant according to fault status data of the wind power plant;
calculating a power generation loss of the wind turbine generator set based on the predicted power value.
2. The method of claim 1, wherein the step of calculating a first power curve and a second power curve of the wind park from the status data of the wind park further comprises:
acquiring state data of a wind generating set in a wind power plant, wherein the type of the wind generating set is the same as that of the wind generating set;
calculating a similar contribution unit weight matrix for the wind power plant using the status data of the same type of wind power plant,
wherein the step of calculating the predicted power value of the wind turbine generator set comprises:
calculating a predicted power value for the wind park using one of the similar-to-output park weight matrix, the first power curve, and the second power curve of the wind park.
3. The method of claim 1, wherein the step of calculating the first and second power curves of the wind park comprises:
determining a wind speed critical point according to the wind speed value and the power value in the state data;
performing data binning based on the wind speed values in the state data to calculate a wind speed value and a first power value of each wind speed interval of the wind generating set;
dividing the calculated wind speed value into a first wind speed section and a second wind speed section according to the determined wind speed critical point;
calculating a first power curve and a second power curve of the wind turbine generator set based on the wind speed value in the first wind speed segment, the wind speed value in the second wind speed segment and the first power value, respectively.
4. The method of claim 3, wherein the step of determining a wind speed critical point comprises:
obtaining a third power curve according to the calculated wind speed value and the first power value;
and taking the wind speed value with larger deviation from the third power curve in the calculated wind speed values as a wind speed critical point.
5. The method of claim 3, wherein the step of calculating the similar-to-output unit weight matrix for the wind turbine generator set comprises:
calculating, for each wind park, a second power value within the respective wind speed interval using the first power curve and the second power curve of the respective wind park based on the wind speed value within each wind speed interval;
calculating a first ratio of a second power value in each wind speed interval to a second power value of the wind generating set in the corresponding wind speed interval in the same state aiming at the wind generating set;
when the first ratio is within a preset range, keeping the first ratio at an original value, and when the first ratio is not within the preset range, setting the first ratio to be zero;
and forming a similar output unit weight matrix of the wind generating set according to the set first ratio.
6. The method of claim 2, wherein the step of calculating the predicted power value for the wind turbine generator set comprises:
selecting part of normal state data of the wind generating set, and making a normal state in the normal state data into a fault state;
calculating a third power value of the wind power plant using one of the first power curve, the second power curve, and the similar-output plant weight matrix of the wind power plant based on the wind speed value in the normal state data;
calculating a first error based on the third power value and a corresponding actual power value in the normal state data;
determining whether the first error satisfies a prediction accuracy;
when the first error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using one of a first power curve, a second power curve and an output similar set weight matrix of the wind generating set.
7. The method of claim 6, wherein the method further comprises:
selecting a fourth power value calculated using the first power curve and the second power curve from the third power values when the first error does not satisfy the prediction accuracy;
calculating a second ratio of the fourth power value to a corresponding actual power value in the normal state data;
determining an optimal parameter factor according to the second ratio;
calculating a fifth power value based on the fourth power value and the optimal parameter factor;
calculating a second error based on the fifth power value and a corresponding actual power value in the normal state data;
determining whether the second error satisfies the prediction accuracy;
when the second error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the first power curve, the second power curve, the optimal parameter factor and the output similarity set weight matrix of the wind generating set.
8. The method of claim 7, wherein the method further comprises:
calculating a sixth power value by a machine learning algorithm based on the third power value when the second error does not satisfy the prediction accuracy;
calculating a third error based on the sixth power value and the corresponding actual power value in the normal state data;
determining whether the third error satisfies the prediction accuracy;
when the third error does not meet the prediction accuracy, continuing to use the machine learning algorithm to calculate the sixth power value until the third error meets the prediction accuracy;
when the third error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the power value of the wind generating set calculated by one of the first power curve, the second power curve and the output similar set weight matrix of the wind generating set in combination with a machine learning algorithm.
9. The method of claim 1, wherein the step of fault status data acquisition of the wind turbine generator set comprises: and performing up-sampling processing on the state data of the wind generating set, and selecting fault state data of the wind generating set from the state data subjected to up-sampling processing.
10. The method of claim 9, wherein the step of upsampling the wind turbine generator system status data comprises upsampling an interpolated fit to the wind turbine generator system status data.
11. The method of claim 2, wherein the step of calculating the predicted power value for the wind turbine generator set comprises:
acquiring the state data of the wind generating set of which the instantaneous wind speed in the fault state data of the wind generating set is within a wind speed interval in a similar set weight matrix;
determining the number M of non-fault wind generating sets in a non-fault state in the wind generating sets;
determining whether to use the first and second power curves or the similar output unit weight matrix to calculate the predicted power value of the wind turbine generator set according to whether the number M is greater than 0.
12. The method of claim 11, wherein the step of determining whether to use the first and second power curves or the similar-output unit weight matrix to calculate the predicted power value for the wind generating set based on whether the number M is greater than 0 comprises:
when the number M is greater than 0, determining the number Q of wind generating sets of which the instantaneous wind speed of the non-fault wind generating sets is within the wind speed interval, and calculating a predicted power value of the wind generating sets by using the first and second power curves or the output similar set weight matrix according to whether the number Q is equal to 0 or not;
when the number M is equal to 0, calculating a predicted power value for the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
13. The method of claim 12, wherein the step of calculating the predicted power value of the wind park using the first and second power curves or the similar-contribution park weight matrix depending on whether the quantity Q is equal to 0 comprises:
when the number Q is larger than 0, calculating a predicted power value of the wind generating set by using the ratio in the wind speed interval in the output similar set weight matrix of the wind generating set and the instantaneous power value of the wind generating set of which the instantaneous wind speed of the non-fault wind generating set is in the wind speed interval;
when the quantity Q is equal to 0, calculating a predicted power value of the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
14. The method of any of claims 1 to 13, further comprising power-limited correction of the predicted power value.
15. The method of any one of claims 1 to 13, further comprising performing statistics on the status data quarterly.
16. An apparatus for calculating a loss of power generation of a wind turbine generator system, the apparatus comprising:
the data processing module is used for calculating a first power curve and a second power curve of the wind generating set at different wind speed sections according to the state data of the wind generating set;
and the power generation amount calculation module is used for calculating a predicted power value of the wind generating set by using the first power curve and the second power curve of the wind generating set according to the fault state data of the wind generating set and calculating the power generation amount loss of the wind generating set based on the predicted power value.
17. The apparatus of claim 16, wherein the data processing module is to:
acquiring state data of a wind generating set in a wind power plant, wherein the type of the wind generating set is the same as that of the wind generating set;
calculating a similar contribution unit weight matrix for the wind power plant using the status data of the same type of wind power plant,
wherein the power generation amount calculation module calculates the predicted power value of the wind turbine generator set using one of the similar output set weight matrix, the first power curve, and the second power curve of the wind turbine generator set.
18. The apparatus of claim 16, wherein the data processing module is to:
determining a wind speed critical point according to the wind speed value and the power value in the state data;
performing data binning based on the wind speed values in the state data to calculate a wind speed value and a first power value of each wind speed interval of the wind generating set;
dividing the calculated wind speed value into a first wind speed section and a second wind speed section according to the determined wind speed critical point;
calculating a first power curve and a second power curve of the wind turbine generator set based on the wind speed value in the first wind speed segment, the wind speed value in the second wind speed segment and the first power value, respectively.
19. The apparatus of claim 18, wherein the data processing module is further to:
obtaining a third power curve according to the calculated wind speed value and the first power value;
and taking the wind speed value with larger deviation from the third power curve in the calculated wind speed values as a wind speed critical point.
20. The apparatus of claim 18, wherein the data processing module is further to:
calculating, for each wind park, a second power value within the respective wind speed interval using the first power curve and the second power curve of the respective wind park based on the wind speed value within each wind speed interval;
calculating a first ratio of a second power value in each wind speed interval to a second power value of the wind generating set in the corresponding wind speed interval in the same state aiming at the wind generating set;
when the first ratio is within a preset range, keeping the first ratio at an original value, and when the first ratio is not within the preset range, setting the first ratio to be zero;
and forming a similar output unit weight matrix of the wind generating set according to the set first ratio.
21. The apparatus of claim 17, wherein the power generation amount calculation module is to:
selecting part of normal state data of the wind generating set, and making a normal state in the normal state data into a fault state;
calculating a third power value of the wind power plant using one of the first power curve, the second power curve, and the similar-output plant weight matrix of the wind power plant based on the wind speed value in the normal state data;
calculating a first error based on the third power value and a corresponding actual power value in the normal state data;
determining whether the first error satisfies a prediction accuracy;
when the first error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using one of a first power curve, a second power curve and an output similar set weight matrix of the wind generating set.
22. The apparatus of claim 21, wherein the power generation calculation module is further configured to:
selecting a fourth power value calculated using the first power curve and the second power curve from the third power values when the first error does not satisfy the prediction accuracy;
calculating a second ratio of the fourth power value to a corresponding actual power value in the normal state data;
determining an optimal parameter factor according to the second ratio;
calculating a fifth power value based on the fourth power value and the optimal parameter factor;
calculating a second error based on the fifth power value and a corresponding actual power value in the normal state data;
determining whether the second error satisfies the prediction accuracy;
when the second error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the first power curve, the second power curve, the optimal parameter factor and the output similarity set weight matrix of the wind generating set.
23. The apparatus of claim 22, wherein the power generation calculation module is further configured to:
calculating a sixth power value by a machine learning algorithm based on the third power value when the second error does not satisfy the prediction accuracy;
calculating a third error based on the sixth power value and the corresponding actual power value in the normal state data;
determining whether the third error satisfies the prediction accuracy;
when the third error does not meet the prediction accuracy, continuing to use the machine learning algorithm to calculate the sixth power value until the third error meets the prediction accuracy;
when the third error meets the prediction accuracy, calculating a predicted power value of the wind generating set based on the fault state data of the wind generating set by using the power value of the wind generating set calculated by one of the first power curve, the second power curve and the output similar set weight matrix of the wind generating set in combination with a machine learning algorithm.
24. The apparatus of claim 16, wherein the data processing module is to:
and performing up-sampling processing on the state data of the wind generating set, and selecting fault state data of the wind generating set from the state data subjected to up-sampling processing.
25. The apparatus of claim 24, wherein the data processing module is further configured to perform an upsampled, interpolated fit on the wind turbine generator system state data.
26. The apparatus of claim 17, wherein the power generation amount calculation module is to:
acquiring the state data of the wind generating set of which the instantaneous wind speed in the fault state data of the wind generating set is within a wind speed interval in a similar set weight matrix;
determining the number M of non-fault wind generating sets in a non-fault state in the wind generating sets;
determining whether to use the first and second power curves or the similar output unit weight matrix to calculate the predicted power value of the wind turbine generator set according to whether the number M is greater than 0.
27. The apparatus of claim 26, wherein the power generation calculation module is further configured to:
when the number M is greater than 0, determining the number Q of wind generating sets of which the instantaneous wind speed of the non-fault wind generating sets is within the wind speed interval, and calculating a predicted power value of the wind generating sets by using the first and second power curves or the output similar set weight matrix according to whether the number Q is equal to 0 or not;
when the number M is equal to 0, calculating a predicted power value for the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
28. The apparatus of claim 27, wherein the power generation calculation module is further configured to:
when the number Q is larger than 0, calculating a predicted power value of the wind generating set by using the ratio in the wind speed interval in the output similar set weight matrix of the wind generating set and the instantaneous power value of the wind generating set of which the instantaneous wind speed of the non-fault wind generating set is in the wind speed interval;
when the quantity Q is equal to 0, calculating a predicted power value of the wind park using the instantaneous wind speed and the first and second power curves of the wind park.
29. The apparatus of any one of claims 16 to 28, wherein the power generation calculation module is configured to perform power limited correction of the predicted power value.
30. Apparatus according to any of claims 16 to 28, wherein the data processing module is arranged to count the status data quarterly.
31. A computer-readable storage medium storing a program, the program comprising instructions for performing the method of any one of claims 1-15.
32. A computer comprising a readable medium having a computer program stored thereon, wherein the computer program comprises instructions for performing the method according to any one of claims 1-15.
CN201811571791.8A 2018-12-21 2018-12-21 Method and device for calculating power generation loss of wind generating set Pending CN111350638A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811571791.8A CN111350638A (en) 2018-12-21 2018-12-21 Method and device for calculating power generation loss of wind generating set

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811571791.8A CN111350638A (en) 2018-12-21 2018-12-21 Method and device for calculating power generation loss of wind generating set

Publications (1)

Publication Number Publication Date
CN111350638A true CN111350638A (en) 2020-06-30

Family

ID=71195643

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811571791.8A Pending CN111350638A (en) 2018-12-21 2018-12-21 Method and device for calculating power generation loss of wind generating set

Country Status (1)

Country Link
CN (1) CN111350638A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112381185A (en) * 2021-01-15 2021-02-19 北京工业大数据创新中心有限公司 Industrial equipment characteristic curve similarity obtaining method and device
US20230213560A1 (en) * 2021-12-30 2023-07-06 SparkCognition, Inc. Calculating energy loss during an outage

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112381185A (en) * 2021-01-15 2021-02-19 北京工业大数据创新中心有限公司 Industrial equipment characteristic curve similarity obtaining method and device
US20230213560A1 (en) * 2021-12-30 2023-07-06 SparkCognition, Inc. Calculating energy loss during an outage
US12066472B2 (en) * 2021-12-30 2024-08-20 SparkCognition, Inc. Calculating energy loss during an outage

Similar Documents

Publication Publication Date Title
CN112765560B (en) Equipment health state evaluation method, device, terminal equipment and storage medium
CN106570790B (en) Wind power plant output data restoration method considering wind speed data segmentation characteristics
CN112734130B (en) Fault early warning method for double-fed fan main shaft
CN111414703B (en) Method and device for predicting residual life of rolling bearing
CN105469216B (en) Method and system for evaluating wind power plant operation risk by combining weather and wind speed
CN111881617A (en) Data processing method, and performance evaluation method and system of wind generating set
EP4311934A1 (en) Computer-implemented method for optimizing the operation of a drivetrain of a wind turbine
CN111709569B (en) Wind power plant output power prediction correction method and device
CN111861023A (en) Statistical-based hybrid wind power prediction method and device
CN111350638A (en) Method and device for calculating power generation loss of wind generating set
CN116484175A (en) Power battery usage data repairing system and method
CN115578084A (en) Wind turbine generator set frequency converter fault early warning method based on deep convolution self-encoder
Cevasco et al. O&M cost-Based FMECA: Identification and ranking of the most critical components for 2-4 MW geared offshore wind turbines
CN109066651A (en) The calculation method of wind-powered electricity generation-load scenarios limit transmitted power
CN117708547B (en) Method and system for processing vibration signals of steam turbine unit
CN104834816A (en) Short-term wind speed prediction method
CN108665376B (en) Nuclear density estimation method for determining maximum value of cellular load in space load prediction
CN114004405A (en) Photovoltaic power prediction method and system based on Elman neural network and satellite cloud picture
CN112347655A (en) Wind power plant theoretical power calculation method based on unit operation performance evaluation
CN117977583B (en) Energy storage auxiliary black start wind speed prediction method based on double-order error compensation
CN112653129B (en) Transient power angle stability margin estimation method, device and system
CN117972363B (en) Wind speed prediction method and system for energy storage auxiliary black start based on stability
CN118311692B (en) Energy storage auxiliary black start wind speed prediction method based on two-dimensional error compensation method
Dou et al. An Improved Fault Detection Method based on Canonical Variate Analysis for Tricky Faults of Wind Turbine
CN118013236B (en) Energy storage auxiliary black start wind speed prediction method and system based on error compensation method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20200630

WD01 Invention patent application deemed withdrawn after publication