A wind power prediction data processing and analysis system and method
By dividing the wind farm and collecting multi-angle data, combining seawater surface feature information, multiple prediction models are established to predict wind speed, which solves the error problem of traditional wind power prediction methods under diversified conditions and improves wind power prediction accuracy and efficiency.
Patent Information
- Application Number
- CN202410780378.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-18
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-06-18
AI Technical Summary
The traditional wind power power prediction method is relatively single in the analysis of diversity conditions, resulting in large errors in the prediction results, reducing the efficiency and accuracy of the prediction work.
By dividing the target wind power detection locations in areas, collecting and analyzing various historical and current characteristic data such as wind speed, wind direction, temperature, and air pressure, establishing multiple prediction models, combining factors such as seawater specific heat capacity, sea surface waves and light intensity, Multi-angle prediction and comprehensive analysis generate multiple wind speed prediction results, and then conduct comprehensive wind power prediction.
It improves the data processing accuracy and working efficiency of wind power prediction, reduces the problems of data confusion and prediction singularity, enhances the comprehensive processing of diversified characteristic data, and improves the diversity and accuracy of prediction results.
Smart Images

Figure CN118349920B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of wind power prediction, and in particular to a wind power prediction data processing and analysis system and method. Background Art
[0002] The main forms of new energy power generation include wind power, photovoltaics, nuclear power, and hydropower. As traditional fossil energy becomes increasingly depleted, renewable energy represented by wind power is beginning to become a trend and trend in world energy development.
[0003] Traditional wind power prediction methods generally analyze the changes in historical characteristic information data of wind farms (including wind power, wind speed, temperature, air pressure, etc.), and predict wind power based on the above-collected characteristic information by establishing a prediction model. However, the analysis of the diversity of the final wind power prediction results is relatively simple, and a more comprehensive analysis is not conducted, resulting in large errors in the final prediction results, leading to low efficiency of the entire prediction work. Summary of the invention
[0004] In order to overcome the above-mentioned deficiencies of the prior art, the present application provides a wind power prediction data processing and analysis system and method.
[0005] In a first aspect, the present application provides a method for processing and analyzing wind power prediction data, the method comprising:
[0006] Obtaining a target wind power detection position, dividing the target wind power detection position and collecting characteristic data of wind speed, wind direction, temperature and air pressure in the divided area, and outputting four historical main characteristic data sets;
[0007] Four historical directional target areas are obtained by regionalizing the target wind power detection position, and offshore characteristic information in each of the four historical directional target areas is processed to obtain four historical influencing factor characteristic data sets;
[0008] Establishing four prediction models, inputting the four historical main feature data sets and the four historical influencing factor feature data sets into the four prediction models for training to obtain four wind change prediction models;
[0009] Obtaining four current main feature data sets and four current influencing factor feature data sets, and inputting the four current main feature data sets and the four current influencing factor feature data sets into four wind change prediction models for testing to obtain four wind speed direction prediction results;
[0010] Analyze the time points, maintenance durations, and angles formed by the four wind directions at which the four initial wind speeds in the four historical main feature data sets arrive at the target wind power detection position, and output four arrival time points, four maintenance durations, and wind direction angle data sets;
[0011] If the four arrival time points are the same, comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction are performed according to the four maintenance durations and wind direction angle data sets, and a first type of wind power prediction result is output;
[0012] If two or three of the four arrival time points are the same, then comprehensive processing of the four wind speed direction prediction results and comprehensive wind power prediction are performed according to the four maintenance durations and wind direction angle data sets, and a second type of wind power prediction result is output;
[0013] If the four arrival time points are all different, the four wind speed prediction results are comprehensively processed and the comprehensive wind power is predicted according to the four maintenance durations, and a third type of wind power prediction result is output.
[0014] By adopting the above technical scheme, the target wind power detection position is divided into a fixed area, and the initially divided fixed area is further divided into four directional target areas, so as to realize the collection and analysis of a plurality of characteristic data information in the four directional target areas respectively, and four prediction models are trained respectively to predict the wind speed direction prediction results under four different situations, and the four wind speed direction prediction results are comprehensively analyzed under multiple time conditions to obtain multiple final comprehensive prediction results, so as to improve the accuracy of data processing in the entire prediction process and the efficiency of prediction work, and avoid the data confusion and the singleness of prediction situations caused by the unified processing of multiple characteristic data information in the unified area of traditional technology.
[0015] Preferably, a circular detection area is obtained by dividing the target wind power detection position as the center, and the target circular detection area is output;
[0016] Divide the target circular detection area into areas of equal area in four historical directions to obtain four historical direction target areas;
[0017] Based on the four historical directional target areas, four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures of the four historical directional target areas are obtained;
[0018] The four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures are combined into four historical main feature data sets.
[0019] By adopting the above technical scheme, through the fixed area division of the target wind power detection position and the more detailed equal area block division, the more comprehensive consideration of the collection and processing of diverse characteristic data information is expanded, and the wind speed, wind direction, temperature and air pressure data information in the target areas in four directions are detected respectively, so as to facilitate the subsequent prediction and processing of four results under four different regional conditions.
[0020] Preferably, four types of seawater specific heat capacity data and four types of sea surface wave amplitude values are collected in the target areas of the four historical directions;
[0021] Based on the four sea surface wave amplitude values, four initial water mist densities of target areas in four historical directions are obtained;
[0022] Obtain four types of illumination intensities in the target area in four historical directions, evaluate four types of initial water mist densities based on four types of seawater specific heat capacity data, four types of sea surface wave amplitude values and four types of illumination intensities, and obtain four types of final water mist densities;
[0023] The four types of seawater specific heat capacity data, four types of sea surface wave amplitude values and four types of final water mist densities are combined into four historical influencing factor characteristic data sets.
[0024] Preferably, according to the comprehensive effect of the four regional temperatures, the four regional air pressures and the four influencing factor characteristic data sets, four arrival time points of the four initial wind speeds at the target wind power detection position are estimated;
[0025] Based on the four wind speed direction prediction results, a wind speed standard detection threshold is set;
[0026] Based on the wind speed standard detection threshold, four wind speed maintenance times that reach the wind speed standard detection threshold in the four wind speed direction prediction results are predicted to obtain four maintenance times;
[0027] The four wind directions in the four wind speed direction prediction results are randomly and freely combined to intersect and form an angle, and a wind direction angle data set is output.
[0028] By adopting the above technical scheme, the buoyancy of water mist on the sea surface is predicted by taking into account some characteristic information of the sea surface, such as sea surface roughness, solar intensity, etc., because these factors will be restrained by each other when the initial wind speed and wind direction pass through the sea surface, resulting in a certain change in the wind speed direction, affecting the prediction of the final wind power, and all the characteristic data information collected above are correspondingly transmitted to the four prediction models for model training, thereby enhancing the efficiency of large-scale data information processing work, so as to obtain four trained wind change prediction models, so as to predict the current wind speed direction, and individually count the wind speed characteristic information in the four wind speed direction prediction results, that is, the time points when the four wind speeds arrive at the target wind power detection position, as well as the duration that can be maintained and the angles formed between them, so as to facilitate the subsequent comprehensive analysis and processing of the four wind speed direction prediction results, and improve the accuracy of the prediction results.
[0029] Preferably, if the four arrival time points are all the same, a first type of multiple groups of angle data formed between the four wind directions are extracted from the wind direction angle data set;
[0030] Determine whether there is a maximum value among the four maintenance durations. If there is no maximum value among the four maintenance durations, comprehensively process the four wind speed direction prediction results based on the first type of multiple groups of angle data and output the four wind speed direction comprehensive processing results;
[0031] If there is a maximum value among the four maintenance durations, a wind speed direction prediction result corresponding to the maximum value among the four maintenance durations is superimposed on the four wind speed direction comprehensive processing results to output an additional wind speed direction comprehensive processing result;
[0032] The four wind speed direction comprehensive processing results and the additional wind speed direction comprehensive processing result are combined into a first type of wind speed direction comprehensive prediction result;
[0033] Based on the first type of wind speed comprehensive prediction results, the final wind power detected at the target wind power detection position is estimated to obtain the first type of wind power prediction results.
[0034] Preferably, if two or three of the four arrival time points are the same, a second type of multiple groups of angle data formed by two or three different wind directions are extracted from the wind direction angle data set;
[0035] Based on the second type of multiple groups of angle data, when two of the four wind speed direction prediction results have the same arrival time points, comprehensive processing is performed to obtain the first two wind speed direction processing results; when three of the four wind speed direction prediction results have the same arrival time points, comprehensive processing is performed to obtain the first three wind speed direction processing results;
[0036] The wind speed end time point is obtained by integrating the corresponding maintenance time of the two wind speeds with different arrival time points to the smaller arrival time point in the prediction results;
[0037] When the wind speed end time point is less than or equal to the larger arrival time point of the two different arrival time points, the two wind speed direction prediction results of the different arrival time points are sequentially superimposed on the first two wind speed direction processing results to output the second two wind speed direction comprehensive processing results;
[0038] Get the ending time point of the wind speed in the three wind speed prediction results at the same arrival time point;
[0039] When the arrival time point of a wind speed direction prediction result with different arrival time points is greater than or equal to the end time point, the wind speed direction prediction result with different arrival time points is sequentially superimposed on the first three wind speed direction processing results to output the second three wind speed direction comprehensive processing results;
[0040] The second two wind speed direction comprehensive processing results and the second third wind speed direction comprehensive processing results are combined into a second type of wind speed direction comprehensive prediction result;
[0041] Based on the second-type wind speed comprehensive prediction result, the final wind power detected at the target wind power detection position is estimated to obtain the second-type wind power prediction result.
[0042] Preferably, if the four arrival time points are all different, the four wind speed end time points are statistically analyzed for the four wind speed direction prediction results according to the four maintenance durations to obtain four wind speed end time points;
[0043] Sorting the four wind speed end time points in ascending order to obtain a wind speed end time point sorting result, and extracting the three wind speed end time points ranked in the top three positions from the wind speed end time point sorting result;
[0044] Obtain the ascending results of the four arrival time points after the wind speed end time point sorting results are in ascending order, and extract the three arrival time points ranked in the last three places from the four arrival time point ascending results;
[0045] The three wind speed end time points of the first three positions are compared with the three arrival time points of the last three positions in sequence to obtain the comparison results. When the comparison results are all less than or equal to, the four wind speed direction prediction results are superimposed to obtain the third type of wind speed direction comprehensive prediction results.
[0046] Based on the third type of wind speed comprehensive prediction result, the final wind power detected at the target wind power detection position is estimated to obtain the third type of wind power prediction result.
[0047] By adopting the above technical scheme, the wind power prediction analysis under different conditions is carried out through the relationship between the time points when the four wind speeds arrive at the target wind power detection position. Due to the different arrival time points, an angle may be formed between the four wind speeds, or an angle may not be formed. When an angle is formed, there will be a situation of wind speed strengthening or weakening between the wind speeds in different prediction results. If no angle is formed, the wind speeds can be directly superimposed, which expands the diversity and comprehensiveness of the comprehensive prediction of wind speeds in various situations, makes the final prediction results more diverse, and reduces the error of the data prediction results.
[0048] In a second aspect, a wind power prediction data processing and analysis system includes:
[0049] A target area processing unit is used to obtain a target wind power detection position, divide the target wind power detection position and collect characteristic data of wind speed, wind direction, temperature and air pressure in the divided area, and output four historical main characteristic data sets;
[0050] A characteristic information analysis unit is used to divide the target wind power detection position into four historical directional target areas, and process the offshore characteristic information in each of the four historical directional target areas to obtain four historical influencing factor characteristic data sets;
[0051] A model building unit is used to build four prediction models, and input four historical main feature data sets and four historical influencing factor feature data sets into the four prediction models for training to obtain four trained wind change prediction models;
[0052] A wind speed direction prediction unit is used to obtain the current four main feature data sets and the current four influencing factor feature data sets, and input the current four main feature data sets and the current four influencing factor feature data sets into four wind change prediction models for testing to obtain four wind speed direction prediction results;
[0053] The wind speed direction feature analysis unit is used to analyze the time points, maintenance durations and angles formed by the four wind directions when the four initial wind speeds in the four historical main feature data sets arrive at the target wind power detection position, and output four arrival time points, four maintenance durations and wind direction angle data sets;
[0054] The first type result prediction unit is used to perform comprehensive processing of four wind speed prediction results and comprehensive wind power prediction according to four maintenance durations and wind direction angle data sets if the four arrival time points are the same, and output the first type wind power prediction result;
[0055] The second type result prediction unit is used for, if two or three of the four arrival time points are the same, performing comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction according to the four maintenance durations and wind direction angle data sets, and outputting the second type wind power prediction result;
[0056] The third type result prediction unit is used to perform comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction according to the four maintenance durations if the four arrival time points are all different, and output the third type wind power prediction result.
[0057] Compared with the prior art, the present invention has the following characteristics and beneficial effects:
[0058] By dividing the fixed area around the target wind power detection position into four equal-area areas, that is, segmenting into four visible directions, it is convenient to distinguish and detect the characteristic information of various influencing factors of wind power in the target areas in the four directions, so as to realize the prediction of the comprehensive wind speed direction results in the target areas in the four directions, that is, by establishing four trained prediction models, the characteristic information possessed by the target areas in the four directions is predicted respectively to obtain four wind speed direction prediction results, which expands the judgment of wind speed direction in multiple spatial directions and predicts multiple results, making the analysis of the factors affecting wind power more comprehensive and the prediction work more efficient.
[0059] By predicting the four arrival time points of the four wind speed direction prediction results at the target wind power detection position and performing statistics on the maintenance duration and the angles formed between the four wind speed directions, a comprehensive prediction of the final wind power in three cases based on the four maintenance durations and angle data sets is made according to the comparison results of whether the values between the four arrival time points obtained above are the same, that is, the first type of wind power prediction results, the second type of wind power prediction results and the third type of wind power prediction results are obtained, thereby realizing the prediction of wind power in various situations and improving the more comprehensive analysis of the prediction results in the prediction process and the accuracy of data analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] Figure 1 It is a flowchart of a method for processing and analyzing wind power prediction data mainly embodied in this embodiment.
[0061] Figure 2 It is a step flowchart of the S1 sub-step mainly embodied in this embodiment.
[0062] Figure 3 It is a step flowchart of the S2 sub-step mainly embodied in this embodiment.
[0063] Figure 4 It is a step flowchart of the S5 sub-step mainly embodied in this embodiment.
[0064] Figure 5 It is a step flowchart of the S6 sub-step mainly embodied in this embodiment.
[0065] Figure 6 It is a step flowchart of the S7 sub-step mainly embodied in this embodiment.
[0066] Figure 7 It is a step flowchart of the S8 sub-step mainly embodied in this embodiment.
[0067] Figure 8 It is a structural block diagram of a wind power prediction data processing and analysis system mainly embodied in this embodiment.
[0068] Explanation of the accompanying drawings: 1. Target area processing unit; 2. Feature information analysis unit; 3. Model building unit; 4. Wind speed direction prediction unit; 5. Wind speed direction feature analysis unit; 6. First type result prediction unit; 7. Second type result prediction unit; 8. Third type result prediction unit. DETAILED DESCRIPTION
[0069] The present invention is further described in detail below in conjunction with the following examples.
[0070] Reference Figure 1 , a wind power prediction data processing and analysis system and method, the method comprising the following steps:
[0071] S1. Obtain the target wind power detection position, divide the target wind power detection position and collect the characteristic data of wind speed, wind direction, temperature and air pressure in the divided area, and output four historical main characteristic data sets.
[0072] S2. The target wind power detection position is divided into four historical directional target areas, and the offshore characteristic information in each of the four historical directional target areas is processed to obtain four historical influencing factor characteristic data sets.
[0073] S3. Four prediction models are established, and four historical main feature data sets and four historical influencing factor feature data sets are input into the four prediction models for training to obtain four wind change prediction models.
[0074] S4. Obtain the current four main feature data sets and the current four influencing factor feature data sets, and input the current four main feature data sets and the current four influencing factor feature data sets into four wind change prediction models for testing to obtain four wind speed direction prediction results.
[0075] S5. Analyze the time points, maintenance durations, and angles formed by the four wind directions when the four initial wind speeds in the four historical main feature data sets arrive at the target wind power detection position, and output four arrival time points, four maintenance durations, and wind direction angle data sets.
[0076] S6. If the four arrival time points are the same, comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction are performed according to the four maintenance durations and wind direction angle data sets, and a first type of wind power prediction result is output.
[0077] S7. If two or three of the four arrival time points are the same, the four wind speed prediction results are comprehensively processed and the comprehensive wind power is predicted based on the four maintenance durations and wind direction angle data sets, and the second type of wind power prediction result is output.
[0078] S8. If the four arrival time points are all different, then the four wind speed prediction results are comprehensively processed and the comprehensive wind power is predicted according to the four maintenance durations, and the third type of wind power prediction result is output.
[0079] Specifically, the target wind power detection position is first determined, and the fixed area of the target wind power detection position is equally divided into four equal-area areas, so as to make it easier to carry out the large error of the characteristic information analysis of various influencing factors in the traditional unified overall spatial wind power prediction and the low accuracy of data analysis. By dividing the target areas into four historical directions, the characteristic information in each area is collected and analyzed and judged separately, and four prediction models are established, and all the four collected data are input into the established four prediction models for training to obtain four wind change prediction models after training, and the current four directions are used for target areas. The wind speed direction results of the target area are predicted by predicting the respective time points at which the corresponding four wind speed direction prediction results predicted in the target area in the current four directions arrive at the target wind power detection position, and the wind speed direction of each prediction result is estimated. The maintenance time and the angle formed by the four wind speed directions are statistically estimated. Subsequently, according to whether the values between the four arrival time points are the same and the same number, a comprehensive prediction of wind power based on the four maintenance durations and angle data sets under different situations is carried out. Finally, the prediction results under various different situations are obtained, which improves the accuracy of data processing and the comprehensiveness of information processing in the overall prediction process.
[0080] Reference Figure 2 , the specific step S1 includes the following sub-steps:
[0081] S1001. Divide a circular detection area with the target wind power detection position as the center, and output the target circular detection area.
[0082] S1002. Divide the target circular detection area into equal-area areas in four historical directions to obtain four historical directional target areas.
[0083] S1003. Based on the four historical directional target areas, four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures of the four historical directional target areas are obtained.
[0084] S1004. Four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures are combined into four historical main feature data sets.
[0085] Specifically, if the coordinates of the target wind power detection position are set to (0,0), then (0,0) is used as the center point of the circle. If the target wind power detection position can detect a distance of 5 kilometers, then a circle area with a radius of 5 kilometers is divided with (0,0) as the center point. In order to make the analysis results of the various feature information required for subsequent detection more accurate, the divided target circular detection area (with a radius of 5 kilometers) is divided into four equal blocks again, and the four equal blocks are equal, and the distribution direction positions are exactly the four directions of "east, south, west and north". According to the four historical direction target areas after division, four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures in these four historical direction target areas are performed by existing technologies. Because the wind speed, direction, temperature and air pressure in different areas at sea may be different, the individual detection and analysis of the feature information in multiple regional directions are performed.
[0086] Reference Figure 3 , the specific step S2 includes the following sub-steps:
[0087] S2001. Collect four types of seawater specific heat capacity data and four types of sea surface wave amplitude values in four historical direction target areas.
[0088] S2002. Based on the four sea surface wave amplitude values, four initial water mist densities of the target areas in four historical directions are obtained.
[0089] S2003. Obtain four types of illumination intensities in the target area in four historical directions, and evaluate four types of initial water mist densities according to four types of seawater specific heat capacity data, four types of sea surface wave amplitude values, and four types of illumination intensities to obtain four types of final water mist densities.
[0090] S2004. Four types of seawater specific heat capacity data, four types of sea surface wave amplitude values and four types of final water mist density are combined into four historical influencing factor characteristic data sets.
[0091] Specifically, the specific heat capacity of seawater refers to the amount of heat that needs to be absorbed or released for a unit mass of water to change its temperature by 1°C. The larger the specific heat capacity of seawater, the stronger its ability to absorb and release heat. Therefore, the temperature of seawater rises and falls relatively slowly. The sea surface wave amplitude value, i.e., the sea surface roughness, refers to the geometric structure of the sea surface to a certain depth, which is affected by wind, waves and other environmental factors. For example, a calm sea surface can be regarded as a flat surface with low backscattering, while a windy and choppy sea surface is a rough surface with significantly enhanced scattered echoes. The floating water mist phenomenon on the sea surface, usually called sea fog, is formed by the evaporation of seawater and the condensation of water vapor in the atmosphere. Because the water mist density has a certain restraint on the subsequent wind speed direction to be detected, and under the influence of four types of seawater specific heat capacity data in different regions, four types of sea surface wave amplitude values and four types of light intensities (such as the data comparison table between the specific heat capacity of seawater and the water mist buoyancy, the data comparison table between the sea surface wave amplitude value and the water mist density, and the light intensity). The water mist density finally formed in different areas is also different. Finally, the three water mist densities evaluated by the three data comparison tables are integrated to obtain four final water mist densities. According to all the above collected data information, they are input into the four established prediction models for training to obtain four wind change prediction models, and the same processing steps are repeated to collect all the current data information, and the final wind speed direction is predicted through the four trained wind change prediction models, that is, four wind speed direction prediction results (i.e. wind speed size and wind speed direction) are obtained.
[0092] Reference Figure 4 , the specific step S5 includes the following sub-steps:
[0093] S5001. Based on the comprehensive effects of four regional temperatures, four regional air pressures and four influencing factor characteristic data sets, four arrival time points of the four initial wind speeds at the target wind power detection position are estimated.
[0094] S5002. Based on the prediction results of the four wind speed directions, set the wind speed standard detection threshold.
[0095] S5003. Based on the wind speed standard detection threshold, the maintenance time of four wind speeds reaching the wind speed standard detection threshold in the four wind speed direction prediction results is predicted to obtain four maintenance time.
[0096] S5004. Randomly and freely combine the four wind directions in the four wind speed direction prediction results to intersect and form an angle, and output a wind direction angle data set.
[0097] Specifically, for example, in a one-variable multi-time function equation, y=(a+b+c)x, y is the dependent variable (the time point when the wind speed reaches the target wind power detection position, and the straight-line distance (i.e., 5 kilometers) for the four wind speeds to reach the target wind power detection position is the same), a refers to the temperature of the four regions, b refers to the air pressure of the four regions, c refers to the characteristic data sets of the four influencing factors, and x refers to the four initial wind speeds. The wind speed standard detection threshold is set (because the wind speed will gradually decrease with the increase of time, it is necessary to timely detect the wind speed that reaches a certain required value, so that the subsequent wind power The result of rate prediction is more accurate and the prediction work is efficient). Furthermore, according to the wind speed values in the prediction results of the four wind speed directions predicted above, the wind speed maintenance time is predicted, that is, four maintenance time lengths are obtained. Since the wind speed direction arriving at the target wind power detection position is the prediction result of the target area in the four directions, the four wind directions are very likely to be different. In order to more accurately judge whether the relationship between the four wind speed directions is mutually reinforcing or weakening, it is necessary to statistically analyze the angles formed between the four wind directions, so as to facilitate the subsequent judgment of the comprehensive relationship of the wind speed based on the relationship between the angles.
[0098] Reference Figure 5 , the specific step S6 includes the following sub-steps:
[0099] S6001. If the four arrival time points are the same, extract the first type of multiple groups of angle data formed between the four wind directions from the wind direction angle data set.
[0100] S6002. Determine whether there is a maximum value among the four maintenance durations. If there is no maximum value among the four maintenance durations, based on the first type of multiple groups of angle data, comprehensively process the four wind speed direction prediction results and output the comprehensive processing results of the four wind speed directions.
[0101] S6003. If there is a maximum value among the four maintenance durations, a wind speed direction prediction result corresponding to the maximum value among the four maintenance durations is superimposed on the comprehensive processing results of the four wind speed directions to output an additional wind speed direction comprehensive processing result.
[0102] S6004. The comprehensive processing results of the four wind speed directions and the comprehensive processing results of the additional wind speed directions are combined into the first type of wind speed direction comprehensive prediction results.
[0103] S6005. Based on the first type of wind speed comprehensive prediction result, the final wind power detected at the target wind power detection position is estimated to obtain the first type of wind power prediction result.
[0104] Specifically, if it is determined that the four arrival time points are the same, then it can be determined with certainty that an angle must be formed between the four wind speed directions, and there is a relationship of wind speed strengthening or weakening between them. The first type of multiple groups of angle data formed between the four wind directions are extracted from the above-stated wind direction angle data set. When the wind directions are different: the larger the angle value formed, the smaller the degree of weakening between the wind speeds, and vice versa, the smaller the angle value formed, the greater the degree of weakening between the wind speeds. When the wind directions are the same: the larger the angle value formed, the greater the degree of weakening between the wind speeds, and vice versa, the smaller the angle value formed, the smaller the degree of weakening between the wind speeds. To determine whether there is a maximum value among the four maintenance time lengths, that is, to determine whether there are additional separately superimposed wind speeds in the prediction results of the four wind speed directions, according to the first type of multiple groups of angle data obtained above and the four maintenance time lengths (the four wind speed direction predictions can be determined based on the maintenance time lengths), Whether to directly perform wind speed integration prediction between measurement results or to perform additional wind speed superposition when one or several wind speeds still exist beyond one or several maintenance times), comprehensively process the four wind speed direction prediction results (in this case, the four wind speed direction prediction results arrive at the target wind power detection position at the same time and the maintenance time is the same), and according to the first type of wind speed direction comprehensive prediction result finally obtained, such as: taking two wind speed direction prediction results as an example, y=a+bx, y is the prediction result of the final comprehensive wind speed, a and b are respectively one of the two wind speed direction prediction results, x is the angle formed between the two wind speed direction prediction results, when the wind directions are the same, y=a+bx is still used for calculation, when the wind directions are different, y=a-bx is used for calculation, and the final wind power detected at the target wind power detection position is budgeted according to the first type of wind speed direction comprehensive prediction result to obtain the first type of wind power prediction result.
[0105] Reference Figure 6 , the specific step S7 includes the following sub-steps:
[0106] S7001. If two or three of the four arrival time points are the same, extract a second type of multiple groups of angle data formed by two or three different wind directions from the wind direction angle data set.
[0107] S7002. Based on the second type of multiple groups of angle data, when two of the four wind speed direction prediction results have the same arrival time points, comprehensive processing is performed to obtain the first two wind speed direction processing results; when three of the four wind speed direction prediction results have the same arrival time points, comprehensive processing is performed to obtain the first three wind speed direction processing results.
[0108] S7003. Perform time integration of the corresponding maintenance time of the two wind speeds at different time points in the prediction results at the smaller time point to obtain the wind speed end time point.
[0109] S7004. When the wind speed end time point is less than or equal to the larger time point of the two wind speed direction prediction results at different time points, the two wind speed direction prediction results at different time points are sequentially superimposed on the first two wind speed direction processing results to output the second two wind speed direction comprehensive processing results.
[0110] S7005. Obtain the end time point of the wind speed in the three wind speed direction prediction results at the same time point.
[0111] S7006. When the arrival time point of a wind speed direction prediction result at a different time point is greater than or equal to the end time point, the wind speed direction prediction result at a different time point is sequentially superimposed on the first three wind speed direction processing results to output the second three wind speed direction comprehensive processing results.
[0112] S7007. The second two wind speed direction comprehensive processing results and the second three wind speed direction comprehensive processing results are combined into the second type of wind speed direction comprehensive prediction results.
[0113] S7008. Based on the second type of wind speed comprehensive prediction result, the final wind power detected at the target wind power detection position is estimated to obtain the second type of wind power prediction result.
[0114] Specifically, if it is determined that there are two or three of the four arrival time points that are the same, then an angle will be formed between the two or three wind speeds with the same arrival time points, and a comprehensive wind speed enhancement or weakening prediction is performed between these two or three wind speeds. The same as the above processing steps, the arrival time point size is judged for the two wind speed direction prediction results with different arrival times. For example, when the smaller time point is based on the corresponding maintenance time, the wind speed end time point can be known. For example, the two wind speed direction prediction results with different time points (A and B, and the arrival time point of A is 7 o'clock, the corresponding maintenance time is 20 minutes, and the arrival time point of B is 7:30, then the wind speed of A The end time point is 7:20, then the wind speeds between A and B do not form an angle, and the wind speed superposition processing can be performed directly). Similarly, a wind speed direction prediction result with a different time point is subjected to the same time value comparison and judgment processing as above, and the other two wind speed direction predictions are superimposed (such as y=a+ or -bx+c+d, at this time the other two or one wind speed is still maintained after the above two or three wind speeds are maintained) to obtain the second type of wind speed direction comprehensive prediction result. According to the second type of wind speed direction comprehensive prediction result, the final wind power detected at the target wind power detection position is budgeted to obtain the second type of wind power prediction result.
[0115] Reference Figure 7 , the specific step S8 includes the following sub-steps:
[0116] S8001. If the four arrival time points are all different, the four wind speed end time points are statistically analyzed for the four wind speed direction prediction results according to the four maintenance durations to obtain the four wind speed end time points.
[0117] S8002. Sort the four wind speed end time points in ascending order to obtain a wind speed end time point sorting result, and extract the three wind speed end time points ranked in the top three positions from the wind speed end time point sorting result.
[0118] S8003. Obtain the ascending results of the four arrival time points after the wind speed end time point sorting results are in ascending order, and extract the three arrival time points ranked in the last three places from the four arrival time point ascending results.
[0119] S8004. Compare the three wind speed end time points of the first three positions with the three arrival time points of the last three positions in sequence to obtain comparison results. When the comparison results are all less than or equal to, the four wind speed direction prediction results are superimposed to obtain the third type of wind speed direction comprehensive prediction result.
[0120] S8005. Based on the third-category wind speed comprehensive prediction result, the final wind power detected at the target wind power detection position is estimated to obtain the third-category wind power prediction result.
[0121] Specifically, if it is determined that the four arrival time points are not the same, at this time, the four wind speeds form an angle with each other, and the next wind speed arrives at the target wind power detection position after the previous wind speed ends within the maintenance time, then y=a+b+c+d, that is, the four wind speed direction prediction results are superimposed to obtain the third type of wind speed direction comprehensive prediction result, and according to the third type of wind speed direction comprehensive prediction result, the final wind power detected at the target wind power detection position is budgeted to obtain the third type of wind power prediction result. For example, the four wind speed direction prediction results are abcd respectively, and the arrival time point of a is 6 o'clock and the maintenance time is 10 minutes, then the wind speed end time point of a is 6:10, the arrival time point of b is 6:10 and the maintenance time is 20 minutes, then the wind speed end time point of b is 6:30, the arrival time point of c is 6:40 and the maintenance time is 20 minutes, then the wind The wind speed end time point is 7:00, the arrival time point of d is 7:00, and the maintenance time is 10 minutes. Then the wind speed end time point of a is 7:10. The four wind speed end time points are sorted in ascending order to obtain the wind speed end time point sorting results (6:10, 6:30, 7:00, 7:10). The three wind speed end time points (6:10, 6:30, 7:00) ranked in the top three are extracted from the wind speed end time point sorting results, and the three comparison arrival time points (6:10, 6:40, 7:00) are compared with the three comparison arrival time points in sequence to obtain the comparison results. When the comparison results are all less than or equal to (that is, the wind speed end time point of a - 6:10 is equal to the arrival time point of b - 6:10, then the two wind speed prediction results can be directly superimposed, and so on).
[0122] A wind power prediction data processing and analysis system, by applying a wind power prediction data processing and analysis method as described above, comprises a target area processing unit 1, a feature information analysis unit 2, a model building unit 3, a wind speed direction prediction unit 4, a wind speed direction feature analysis unit 5, a first type result prediction unit 6, a second type result prediction unit 7 and a third type result prediction unit 8, with reference to Figure 8, the target wind power detection position is obtained through the target area processing unit 1, the target wind power detection position is divided and the wind speed, wind direction, temperature and air pressure characteristic data in the divided area are collected, and four historical main characteristic data sets are output; the target wind power detection position is divided into four historical direction target areas through the characteristic information analysis unit 2, and the offshore characteristic information in each direction target area of the four historical direction target areas is processed to obtain four historical influencing factor characteristic data sets; four prediction models are established through the model establishment unit 3, and the four historical main characteristic data sets and the four historical influencing factor characteristic data sets are input into the four prediction models for training to obtain four trained wind change prediction models; the current four main characteristic data sets and the current four influencing factor characteristic data sets are obtained through the wind speed direction prediction unit 4, and the current four main characteristic data sets and the current four influencing factor characteristic data sets are input into the four wind change prediction models for testing to obtain four wind speed direction prediction results; the wind speed direction characteristic data sets are used to establish four prediction models, and ... The feature analysis unit 5 analyzes the time points, maintenance durations and angles formed by the four wind directions when the four initial wind speeds in the four historical main feature data sets arrive at the target wind power detection position, and outputs four arrival time points, four maintenance durations and wind direction angle data sets; through the first type of result prediction unit 6, if the four arrival time points are the same, the four wind speed direction prediction results are comprehensively processed and the comprehensive wind power is predicted according to the four maintenance durations and wind direction angle data sets, and the first type of wind power prediction result is output; through the second type of result prediction unit 7, if two or three of the four arrival time points are the same, the four wind speed direction prediction results are comprehensively processed and the comprehensive wind power is predicted according to the four maintenance durations and wind direction angle data sets, and the second type of wind power prediction result is output; through the third type of result prediction unit 8, if the four arrival time points are different, the four wind speed direction prediction results are comprehensively processed and the comprehensive wind power is predicted according to the four maintenance durations, and the third type of wind power prediction result is output.
[0123] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.
Claims
1. A method for processing and analyzing wind power prediction data, It is characterized in that The following steps are involved: Obtaining a target wind power detection position, dividing the target wind power detection position and collecting characteristic data of wind speed, wind direction, temperature and air pressure in the divided area, and outputting four historical main characteristic data sets; A circular detection area is obtained with the target wind power detection position as the center, and the target circular detection area is output. The target circular detection area is divided into four historical direction equal area ranges to obtain four historical direction target areas, where the four directions are exactly the four directions of southeast, northwest, and northeast. The offshore feature information in each of the four historical direction target areas is processed to obtain four historical influencing factor feature data sets; Establishing four prediction models, inputting the four historical main feature data sets and the four historical influencing factor feature data sets into the four prediction models for training to obtain four wind change prediction models; Obtaining four current main feature data sets and four current influencing factor feature data sets, and inputting the four current main feature data sets and the four current influencing factor feature data sets into four wind change prediction models for testing to obtain four wind speed direction prediction results; Analyze the time points, maintenance durations, and angles formed by the four wind directions at which the four initial wind speeds in the four historical main feature data sets arrive at the target wind power detection position, and output four arrival time points, four maintenance durations, and wind direction angle data sets; If the four arrival time points are the same, comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction are performed according to the four maintenance durations and wind direction angle data sets, and a first type of wind power prediction result is output; If the four arrival time points are all the same, extracting a first type of multiple groups of angle data formed between the four wind directions from the wind direction angle data set; Determine whether there is a maximum value among the four maintenance durations. If there is no maximum value among the four maintenance durations, comprehensively process the four wind speed direction prediction results based on the first type of multiple groups of angle data and output the four wind speed direction comprehensive processing results; If there is a maximum value among the four maintenance durations, a wind speed direction prediction result corresponding to the maximum value among the four maintenance durations is superimposed on the four wind speed direction comprehensive processing results to output an additional wind speed direction comprehensive processing result; The four wind speed direction comprehensive processing results and the additional wind speed direction comprehensive processing result are combined into a first type of wind speed direction comprehensive prediction result; Based on the first type of wind speed comprehensive prediction results, predicting the final wind power detected at the target wind power detection position to obtain a first type of wind power prediction result; If two or three of the four arrival time points are the same, then comprehensive processing of the four wind speed direction prediction results and comprehensive wind power prediction are performed according to the four maintenance durations and wind direction angle data sets, and a second type of wind power prediction result is output; If two or three of the four arrival time points are the same, extracting a second type of multiple groups of angle data formed by two or three different wind directions from the wind direction angle data set; Based on the second type of multiple groups of angle data, when two of the four wind speed direction prediction results have the same arrival time points, comprehensive processing is performed to obtain the first two wind speed direction processing results; when three of the four wind speed direction prediction results have the same arrival time points, comprehensive processing is performed to obtain the first three wind speed direction processing results; The wind speed end time point is obtained by integrating the corresponding maintenance time of the two wind speeds with different arrival time points to the smaller arrival time point in the prediction results; When the wind speed end time point is less than or equal to the larger arrival time point of the two different arrival time points, the two wind speed direction prediction results of the different arrival time points are sequentially superimposed on the first two wind speed direction processing results to output the second two wind speed direction comprehensive processing results; Get the ending time point of the wind speed in the three wind speed prediction results at the same arrival time point; When the arrival time point of a wind speed direction prediction result with different arrival time points is greater than or equal to the end time point, the wind speed direction prediction result with different arrival time points is sequentially superimposed on the first three wind speed direction processing results to output the second three wind speed direction comprehensive processing results; The second two wind speed direction comprehensive processing results and the second third wind speed direction comprehensive processing results are combined into a second type of wind speed direction comprehensive prediction result; Based on the second-type wind speed direction comprehensive prediction result, the final wind power detected at the target wind power detection position is predicted to obtain the second-type wind power prediction result; If the four arrival time points are all different, then the four wind speed direction prediction results are comprehensively processed and the comprehensive wind power is predicted according to the four maintenance durations, and a third type of wind power prediction result is output; If the four arrival time points are all different, the four wind speed end time points are statistically analyzed for the four wind speed direction prediction results according to the four maintenance durations to obtain four wind speed end time points; Sorting the four wind speed end time points in ascending order to obtain a wind speed end time point sorting result, and extracting the three wind speed end time points ranked in the top three positions from the wind speed end time point sorting result; Obtain the ascending results of the four arrival time points after the wind speed end time point sorting results are in ascending order, and extract the three arrival time points ranked in the last three places from the four arrival time point ascending results; The three wind speed end time points of the first three positions are compared with the three arrival time points of the last three positions in sequence to obtain the comparison results. When the comparison results are all less than or equal to, the four wind speed direction prediction results are superimposed to obtain the third type of wind speed direction comprehensive prediction results. Based on the third type of wind speed comprehensive prediction result, the final wind power detected at the target wind power detection position is predicted to obtain the third type of wind power prediction result.
2. A method for processing and analyzing wind power prediction data according to claim 1, It is characterized in that The steps of obtaining a target wind power detection position, dividing the target wind power detection position and collecting characteristic data of wind speed, wind direction, temperature and air pressure in the divided area, and outputting four historical main characteristic data sets are specifically as follows: A circular detection area is obtained by dividing the target wind power detection position as the center, and the target circular detection area is output; Divide the target circular detection area into areas of equal area in four historical directions to obtain four historical direction target areas; Based on the four historical directional target areas, four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures of the four historical directional target areas are obtained; The four initial wind speeds, four initial wind directions, four regional temperatures and four regional air pressures are combined into four historical main feature data sets.
3. A method for processing and analyzing wind power prediction data according to claim 2, It is characterized in that The circular detection area is divided with the target wind power detection position as the center, and the target circular detection area is output. The target circular detection area is divided into four historical direction equal area ranges to obtain four historical direction target areas, where the four directions are exactly the four directions of southeast, northwest, and northeast. The offshore feature information in each of the four historical direction target areas is processed to obtain four historical influencing factor feature data sets, specifically: Collecting four types of seawater specific heat capacity data and four types of sea surface wave amplitude values in the target areas of the four historical directions; Based on the four sea surface wave amplitude values, four initial water mist densities of target areas in four historical directions are obtained; Obtain four types of illumination intensities in the target area in four historical directions, evaluate four types of initial water mist densities based on four types of seawater specific heat capacity data, four types of sea surface wave amplitude values and four types of illumination intensities, and obtain four types of final water mist densities; The four types of seawater specific heat capacity data, four types of sea surface wave amplitude values and four types of final water mist densities are combined into four historical influencing factor characteristic data sets.
4. A method for processing and analyzing wind power prediction data according to claim 2, It is characterized in that The steps of analyzing the time points at which the four initial wind speeds in the four historical main feature data sets arrive at the target wind power detection position, the maintenance duration, and the angles formed by the four wind directions, and outputting four arrival time points, four maintenance durations, and wind direction angle data sets are specifically as follows: According to the comprehensive effect of the four regional temperatures, the four regional air pressures and the four influencing factor characteristic data sets, the four initial wind speeds are predicted to arrive at the target wind power detection position at four arrival time points; Based on the four wind speed direction prediction results, a wind speed standard detection threshold is set; Based on the wind speed standard detection threshold, four wind speed maintenance times that reach the wind speed standard detection threshold in the four wind speed direction prediction results are predicted to obtain four maintenance times; The four wind directions in the four wind speed direction prediction results are randomly and freely combined to intersect and form an angle, and a wind direction angle data set is output.
5. A system for processing and analyzing wind power prediction data, It is characterized in that The system is used to implement a wind power prediction data processing and analysis method as described in any one of claims 1 to 4, comprising: A target area processing unit is used to obtain a target wind power detection position, divide the target wind power detection position and collect characteristic data of wind speed, wind direction, temperature and air pressure in the divided area, and output four historical main characteristic data sets; A characteristic information analysis unit is used to divide the target wind power detection position into a circular detection area with the target wind power detection position as the center, output the target circular detection area, divide the target circular detection area into four historical directions with equal area ranges to obtain four historical direction target areas, where the four directions are exactly the four directions of southeast, northwest, and northeast, and process the marine characteristic information in each of the four historical direction target areas to obtain four historical influencing factor characteristic data sets; A model building unit is used to build four prediction models, and the four historical main feature data sets and the four historical influencing factor feature data sets are input into the four prediction models for training to obtain four wind change prediction models; A wind speed direction prediction unit is used to obtain four current main feature data sets and four current influencing factor feature data sets, and input the four current main feature data sets and the four current influencing factor feature data sets into four wind change prediction models for testing to obtain four wind speed direction prediction results; A wind speed direction feature analysis unit, used to analyze the time points, maintenance durations, and angles formed by the four wind directions of the four historical main feature data sets when the four initial wind speeds arrive at the target wind power detection position, and output four arrival time points, four maintenance durations, and wind direction angle data sets; The first type result prediction unit is used to perform comprehensive processing of four wind speed prediction results and comprehensive wind power prediction according to four maintenance durations and wind direction angle data sets if the four arrival time points are the same, and output the first type wind power prediction result; The second type result prediction unit is used for, if two or three of the four arrival time points are the same, performing comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction according to the four maintenance durations and wind direction angle data sets, and outputting the second type wind power prediction result; The third type result prediction unit is used to perform comprehensive processing of the four wind speed prediction results and comprehensive wind power prediction according to the four maintenance durations if the four arrival time points are all different, and output the third type wind power prediction result.
