Port micro-grid wind energy electric field power stabilizing method, storage medium and equipment

Abstract

The invention discloses a port micro-grid wind energy electric field power stabilizing method, a storage medium and equipment, and the method comprises the steps: firstly generating the actual power generation power of a wind driven generator through a wind energy conversion model, then extracting typical daily data through employing a k-means clustering algorithm, and on this basis, carrying out the calculation of the typical daily data; the improved sliding filtering algorithm and the amplitude limiting filtering algorithm are used for filtering the generated power, so that the power fluctuation ratio of wind energy is remarkably reduced, the filtering algorithm is optimized, a theoretical basis and practical guidance are provided for reliable grid connection of renewable energy power generation equipment, and remarkable environmental protection benefits and engineering application value are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of renewable energy in the power system, and mainly relates to a method for suppressing the power of a wind energy farm in a port microgrid, a storage medium, and a device. Background Art

[0002] In the current field of renewable energy in the power system, renewable energy power generation has volatility and intermittency, which poses a huge challenge to the stable operation of the power system. Especially with the large-scale application of renewable energy, how to effectively control the power volatility of renewable energy power generation to achieve reliable grid connection has become a technical bottleneck that the power industry urgently needs to break through.

[0003] Traditional filtering algorithms are mainly based on algorithms such as sliding filtering. Although these methods can meet the filtering requirements in a timely manner, they cannot effectively achieve moderate suppression. Excessive suppression will cause the energy storage to act excessively, bringing a great burden to the energy storage. How to combine the characteristics of renewable energy power generation to effectively suppress the power generation remains a technical problem that urgently needs to be solved in the field of power systems. Summary of the Invention

[0004] In view of the problems existing in the prior art, the present invention proposes a method for suppressing the power of a wind energy farm in a port microgrid, a storage medium, and a device. First, the actual power generation of a wind turbine is generated through a wind energy conversion model, and then the k-means clustering algorithm is used to extract typical day data. On this basis, an improved sliding filtering algorithm and a clipping filtering algorithm are used to filter the power generation, thereby significantly reducing the power volatility of the wind energy. The method of the present invention provides a theoretical basis and practical guidance for the reliable grid connection of renewable energy power generation equipment by optimizing the filtering algorithm, and has significant environmental benefits and engineering application value.

[0005] To achieve the above object, the technical solution adopted by the present invention is: A method for suppressing the power of a wind energy farm in a port microgrid, comprising the following steps:

[0006] S1. Data generation: Generate the actual power generation of the wind turbine through a wind energy conversion model. The wind energy conversion model calculates the wind energy output based on the wind speed value at the height of the wind turbine shaft and outputs the output power of the wind turbine;

[0007] S2. Selection of typical days: Use the k-means clustering algorithm to select typical day data. The k-means clustering algorithm is specifically: Use the Euclidean distance as the evaluation index of sample similarity to divide the samples, and update the centroid of each cluster to the average value of all data points in the cluster through iterative calculation until the centroid position no longer changes or reaches a predetermined number of iterations, and the sum of squared error criterion function between data points in the cluster is the smallest;

[0008] S3. Filtering: Suppress power fluctuations by improving the sliding filtering algorithm and the amplitude-limiting filtering algorithm;

[0009] The specific improvement of the sliding filtering algorithm is as follows: Each newly sampled power signal is stored in a window array. At the same time, the data at the end of the array is removed, and the arithmetic mean of the array is calculated as the power smoothing grid-connected power at the current moment. After each filtering is completed, the window number is dynamically updated and the change of the window number is restricted;

[0010] The specific amplitude-limiting filtering algorithm is as follows: Each time a new generated power is sampled and compared with the previous sampled value. If the power change exceeds the set threshold, the change value is set to the maximum allowable value, and the original output value is replaced with this value.

[0011] As an improvement of the present invention, in the step S1, the relationship between the output power of the wind turbine and the actual wind speed is:

[0012]

[0013] In the formula, P wind represents the output power of the wind turbine; P w_rate represents the rated power of the wind turbine; v s represents the cut-in wind speed; v r represents the rated wind speed; v c represents the cut-out wind speed; v represents the wind speed at the height of the wind turbine shaft.

[0014] As another improvement of the present invention, the conversion relationship between the wind speed value at the height of the wind turbine shaft and the measured wind speed value is specifically:

[0015]

[0016] In the formula, h represents the height of the wind turbine shaft; h0 is the height of the wind speed measurement position; v0 is the wind speed at the wind speed measurement position; μ is the friction coefficient.

[0017] As yet another improvement of the present invention, the square error criterion function in the step S2 is specifically:

[0018]

[0019] In the formula, J is the error criterion function; K1 is the number of clusters; R j represents the jth cluster; c j is the centroid of R j ; m is the sample index; M is the number of samples; x m represents the mth sample; d mj represents whether the mth sample belongs to R j .

[0020] As another improvement of the present invention, in step S3, the grid-connected power based on the sliding filtering algorithm is expressed as:

[0021]

[0022] In the formula, L is the sliding window width, and P G (t) is the grid-connected power smoothed at time t;

[0023] The expression of the grid-connected power based on the amplitude-limiting filtering algorithm is:

[0024]

[0025] In the formula, ΔP pos represents the positive limit value of the power generation power fluctuation, and ΔP neg represents the negative limit value of the power generation power fluctuation, and P lim (t) represents the grid-connected power after amplitude-limiting filtering at time t.

[0026] To achieve the above object, the technical solution adopted by the present invention is also: a non-transitory machine-readable storage medium, on which executable code is stored. When the executable code is executed by a processor of an electronic device, the processor is caused to execute a method for suppressing the power of a wind energy power plant in a port microgrid as described in any one of claims 1-5 above.

[0027] To achieve the above object, the technical solution adopted by the present invention is also: a computer device, including:

[0028] A memory, on which executable code is stored;

[0029] A processor, configured to execute the executable code, so that the computer device executes the operations of a method for suppressing the power of a wind energy power plant in a port microgrid as described in any one of claims 1-5.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention discloses a method, a storage medium and a device for suppressing the power of a wind energy power plant in a port microgrid. The wind power generation power is generated based on a wind energy conversion model to achieve accurate power generation; based on the k-means clustering algorithm, clustering is realized and typical days are selected; based on the improved sliding filtering algorithm and amplitude-limiting filtering algorithm, effective power suppression is realized. The method of the present invention provides a theoretical basis and practical guidance for the reliable grid connection of renewable energy power generation equipment by optimizing the filtering algorithm, has significant environmental benefits and engineering application value, and also provides an innovative technical solution for the stable grid connection of renewable energy in a port microgrid. Description of the Drawings

[0031] Figure 1It is the flow chart of the steps of the method of the present invention;

[0032] Figure 2 It is the actual wind power numerical value diagram output by step S1 of Embodiment 2 of the present invention;

[0033] Figure 3 It is the schematic diagram of the dynamic window number of the filtering algorithm adopted in Embodiment 2 of the present invention;

[0034] Figure 4 It is the comparison diagram of the volatility of the original and filtered wind power in Embodiment 2 of the present invention. Detailed implementation manners

[0035] The present invention will be further clarified below in conjunction with the accompanying drawings and specific implementation manners. It should be understood that the following specific implementation manners are only used to illustrate the present invention and not to limit the scope of the present invention.

[0036] Embodiment 1

[0037] A method for suppressing the power of a wind energy farm in a port microgrid based on an improved sliding filtering algorithm and a clipping filtering algorithm, as Figure 1 shown, includes the following steps:

[0038] Step S1, data generation: Generate actual power generation through a wind energy conversion model.

[0039] The output of a wind turbine depends on the wind speed at the height of the turbine shaft and the output characteristics of the wind turbine. When calculating the wind energy output, it is necessary to convert the measured wind speed value into the wind speed value at the height of the turbine shaft according to the variation law of the wind speed with height. The corresponding conversion formula is as follows:

[0040]

[0041] In the formula, h represents the height of the wind turbine shaft; h0 is the height of the wind speed measurement position; v is the wind speed at the height of the wind turbine shaft; v0 is the wind speed at the wind speed measurement position; μ is the friction coefficient.

[0042] The relationship between the output power of the wind turbine and the actual wind speed can be expressed as:

[0043]

[0044] In the formula, P wind represents the output power of the wind turbine; P w_rate represents the rated power of the wind turbine; v s represents the cut-in wind speed; v r represents the rated wind speed; v c represents the cut-out wind speed.

[0045] Step S2: Select typical day data using the k-means clustering algorithm.

[0046] The k-means algorithm is an iterative clustering analysis method based on partitioning. Its core objective is to divide n data points into several clusters, and according to the similarity between data points, assign each data point to the cluster corresponding to its nearest centroid, so that the squared error criterion function within the cluster is stabilized at the minimum value. Usually, the Euclidean distance is used as the evaluation index of sample similarity to divide the samples, and the centroid of each cluster is updated to the average value of all data points within the cluster through iterative calculation until the centroid position no longer changes or reaches the predetermined number of iterations.

[0047] The squared error criterion function is the sum of the squared errors, and its formula is as follows:

[0048]

[0049] In the formula, J is the error criterion function; K1 is the number of clusters; R j represents the jth cluster; c j is the centroid of R j ; m is the sample index; M is the number of samples; x m represents the mth sample, which is a vector composed of relevant factors of the factors to be clustered; d mj represents whether the mth sample belongs to R j .

[0050] Step S3: Filtering: Suppress power fluctuations through an improved sliding filtering algorithm and a clipping filtering algorithm.

[0051] In view of the problem of over-suppression or under-suppression caused by the inability to adjust the window number of the sliding filtering algorithm in a timely manner in the present invention, the algorithm structure of the sliding filtering algorithm is improved, and the window number is dynamically adjusted and the change of the window number is restricted after each filtering. Considering that the clipping filtering algorithm can limit the power change, this is used to make up for the problem of too large window number of the improved sliding filtering algorithm. The improved sliding filtering algorithm and the clipping filtering algorithm are combined to form a new filtering algorithm.

[0052] There are certain requirements for the grid-connected power fluctuations of new energy power generation. The definition of the grid-connected volatility rate is as follows:

[0053]

[0054] In the formula, P(t) is the power at time t; is the grid-connected power volatility rate at time t; P rate is the rated installed capacity of the power station; Δt is the sampling time interval.

[0055] The moving average filtering algorithm is a classical algorithm for suppressing power fluctuations. It uses the current sampling point and the average power within a time interval before the sampling point to obtain a smoothing filtering effect. The implementation process is as follows: Each new sampled power signal is stored in a window array. At the same time, the data at the end of the array is removed, and the arithmetic mean of the array is calculated as the grid-connected power of the current moment after power smoothing. The grid-connected power based on the moving filtering algorithm is expressed as:

[0056]

[0057] In the formula, L is the width of the moving window; P G (t) is the grid-connected power after smoothing at time t.

[0058] Considering that the moving average filtering algorithm has certain requirements for the time window, if the window is too large, even if the grid-connected fluctuation requirements are met, a large phase lag will occur. Based on the limiting filtering algorithm, the difference in wind power between adjacent sampling periods is limited and smoothed. The specific implementation process is as follows: Each time a new generated power is sampled and compared with the previous sampling value. If the power change exceeds the set threshold, the change value is set to the maximum allowable value, and the original output value is replaced with this value.

[0059] The expression of the grid-connected power based on the limiting filtering algorithm is:

[0060]

[0061] In the formula, ΔP pos represents the positive limit value of the generated power fluctuation; ΔP neg represents the negative limit value of the generated power fluctuation; P lim (t) represents the grid-connected power after limiting filtering at time t.

[0062] In this embodiment, the specific operation steps are as follows:

[0063] 1) The total length of the sampled power generation data points is N, where k represents the kth sampling point; k = 1 is the starting point of the data, directly connected to the power grid; initialize k = 2 and the maximum window number L max ;

[0064] 2) Each time k is updated, reset the window number L = 1 and jump to step 3);

[0065] 3) Determine whether the window number is less than the sampling point k and the maximum window number L max : If the condition is satisfied, go to step 4); otherwise, run the limiting filtering algorithm and output the result, then go to step 5).

[0066] 4) Determine whether the grid connection fluctuation meets the requirements: If it meets the requirements, output the result, and then go to step 5); otherwise, L = L + 1, and go to step 3);

[0067] 5) Determine whether k exceeds the total length of the sampling points: If so, end the algorithm process; otherwise, perform the filtering calculation for the next sampling point, k = k + 1, and jump to step 2).

[0068] Embodiment 2

[0069] In this embodiment, specific values are substituted to highlight the effectiveness and reliability of the method of the present invention. A method for suppressing the power fluctuation of a wind energy power plant in a port microgrid based on an improved sliding filtering algorithm and a clipping filtering algorithm, as Figure 1 shown, includes the following steps:

[0070] Step S1, data generation: Generate the actual generated power through a wind energy conversion model.

[0071] The output of a wind turbine depends on the wind speed at the height of the turbine shaft and the output characteristics of the wind turbine. When calculating the wind energy output, it is necessary to convert the measured wind speed value into the wind speed value at the height of the turbine shaft according to the variation law of the wind speed with height. The corresponding conversion formula is as follows:

[0072]

[0073] In the formula, h represents the height of the wind turbine shaft, set to 100 m; h0 is the height of the wind speed measurement position, set to 10 m; v is the wind speed at the height of the wind turbine shaft; v0 is the wind speed at the wind speed measurement position; μ is the friction coefficient, set to 0.1 in the open sea area.

[0074] The relationship between the output power of the wind turbine and the actual wind speed can be expressed as:

[0075]

[0076] In the formula, P wind represents the output power of the wind turbine; P w_rate represents the rated power of the wind turbine, set to 1200 kW; v s represents the cut-in wind speed, set to 3 m / s; v r represents the rated wind speed, set to 20 m / s; v c represents the cut-out wind speed, set to 25 m / s.

[0077] As Figure 2As shown, based on the collected sea surface wind speed data, wind power output data with a rated capacity of 1200 kW was generated through a wind turbine power conversion model, and the power output range of this model is strictly limited to 0 - 1200 kW.

[0078] Step S2: Use the k-means clustering algorithm to select typical day data.

[0079] The k-means algorithm is an iterative clustering analysis method based on partitioning. Its core goal is to divide n data points into several clusters, and according to the similarity between data points, assign each data point to the cluster corresponding to its nearest centroid, so that the squared error criterion function among the data points within the cluster is stabilized at the minimum value. Usually, the Euclidean distance is used as the evaluation index of sample similarity to divide the samples, and the centroid of each cluster is updated to the average value of all data points within the cluster through iterative calculation until the centroid position no longer changes or reaches the predetermined number of iterations.

[0080] The squared error criterion function is the sum of the sum of squared errors, and its formula is as follows:

[0081]

[0082] In the formula, J is the error criterion function; K1 is the number of clusters, set to 5; R j represents the jth cluster; c j is the centroid of R j ; m is the sample index; M is the number of samples; x m represents the mth sample, which is a vector composed of relevant factors of the factors to be clustered; d mj represents whether the mth sample belongs to R j .

[0083] Table 1 shows the result values of the k-means algorithm clustering:

[0084] cluster 1 2 3 4 5 typical day 121 67 58 97 22

[0085] As shown in Table 1, the k-means clustering algorithm is used to cluster the power generation data for 365 days throughout the year. When the number of clusters K1 = 5, the wind power generation data is divided into 5 clusters, and below each cluster is the number of typical days of the wind power generation data.

[0086] Step S3: Filtering: Suppress power fluctuations through an improved sliding filtering algorithm and a limit filtering algorithm.

[0087] There are certain requirements for the grid-connected power fluctuations of new energy power generation. The definition of the grid-connected volatility rate is as follows:

[0088]

[0089] Wherein, P(t) is the power at time t; is the grid-connected power volatility at time t, set to 0.05; P rate is the rated installed capacity of the power plant; Δt is the sampling time interval, set to 15 min.

[0090] The moving average filtering algorithm is a classic algorithm for suppressing power fluctuations. By using the average power value within the current sampling point and the time interval before the sampling point, a smooth filtering effect is obtained. The implementation process is as follows: Each newly sampled power signal is stored in a window array. At the same time, the data at the end of the array is removed, and the arithmetic mean of the array is calculated as the smooth grid-connected power at the current moment. The grid-connected power based on the moving filtering algorithm is expressed as:

[0091]

[0092] Wherein, L is the width of the moving window; P G (t) is the smoothed grid-connected power at time t.

[0093] Considering that the moving average filtering algorithm has certain requirements for the time window, if the window is too large, even if the grid connection fluctuation requirements are met, a large phase lag will be generated. Based on the amplitude-limiting filtering algorithm, the wind power difference between adjacent sampling periods is limited and smoothed. The specific implementation process is as follows: Each time a new generated power is sampled and compared with the previous sampling value. If the power change exceeds the set threshold, the change value is set to the maximum allowable value, and the original output value is replaced with this value.

[0094] The expression of the grid-connected power based on the amplitude-limiting filtering algorithm is:

[0095]

[0096] Wherein, ΔP pos represents the positive limit value of the generated power fluctuation; ΔP neg represents the negative limit value of the generated power fluctuation; P lim (t) represents the grid-connected power after amplitude-limiting filtering at time t.

[0097] In this embodiment, the specific operation steps are as follows:

[0098] 1) The total length of the sampled power generation data points is N, where k represents the kth sampling point; k = 1 is the starting point of the data, directly connected to the grid; initialize k = 2 and the maximum window number L max , set to 3;

[0099] 2) Each time k is updated, reset the window number L = 1 and jump to step 3);

[0100] 3) Determine whether the number of windows is less than the sampling point k and the maximum number of windows L max : If the condition is satisfied, go to step 4); otherwise, run the amplitude-limiting filtering algorithm, output the result, and then go to step 5);

[0101] 4) Determine whether the grid connection fluctuation meets the requirements: If it meets the requirements, output the result and then go to step 5); otherwise, L = L + 1, and go to step 3);

[0102] 5) Determine whether k exceeds the total length of the sampling points: If so, end the algorithm process; otherwise, perform the filtering calculation for the next sampling point, k = k + 1, and jump to step 2).

[0103] In this embodiment, for the single-day wind power time series, an improved filtering algorithm is used for processing Figure 3 shows the dynamic window number of the adopted filtering algorithm, and the window width is adjusted in real time according to the dynamic threshold condition, and its maximum value is constrained to 3. Figure 4 shows the comparison of the volatility between the original and filtered wind power Figure 4 It can be seen from that by setting the volatility threshold of 5% and applying the improved filtering algorithm proposed by the present invention, the system can realize the real-time suppression of the fluctuation component and ensure that the output power volatility always meets the constraint requirement of not exceeding 5%.

[0104] In summary, the method of the present invention improves the original filtering algorithm in view of the disadvantage that the original sliding filtering algorithm cannot dynamically adjust the number of windows; in view of the large delay effect caused by too large a number of windows in the sliding filtering algorithm, the amplitude-limiting filtering algorithm and the improved sliding filtering algorithm are combined to limit the excessive smoothing of the sliding filtering algorithm. The present invention generates wind power generation power based on the wind energy conversion model to realize the accurate generation of power; based on the k-means clustering algorithm, clustering is realized and typical days are selected; based on the improved sliding filtering algorithm and the amplitude-limiting filtering algorithm, the effective smoothing of power is realized, solving the technical problems that urgently need to be overcome in the field of power systems, and providing an effective guarantee for the reliable grid connection of renewable energy power generation equipment.

[0105] It should be noted that the above content only illustrates the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. For those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and retouches can be made, and these improvements and retouches all fall within the protection scope of the claims of the present invention.

Claims

1. A method for suppressing the power fluctuation of a wind farm in a port microgrid, characterized in that It includes the following steps: S1. Data generation: Generate the actual power generation of the wind turbine through a wind energy conversion model. The wind energy conversion model calculates the wind energy output based on the wind speed value at the height of the wind turbine shaft and outputs the output power of the wind turbine; S2. Selection of typical days: Use the k-means clustering algorithm to select typical day data. The specific k-means clustering algorithm is as follows: Use the Euclidean distance as the evaluation index of sample similarity to divide the samples, and update the centroid of each cluster to the average value of all data points in the cluster through iterative calculation until the centroid position no longer changes or reaches a predetermined number of iterations. The sum of squared errors criterion function between data points within the cluster is minimized; S3. Filtering: Suppress power fluctuations through an improved sliding filter algorithm and a limit filtering algorithm; The specific improved sliding filter algorithm is as follows: Store each newly sampled power signal in a window array, and at the same time remove the data at the end of the array. Calculate the arithmetic mean of the array as the power smoothing grid-connected power at the current moment; After each filtering ends, dynamically update the window number and limit the change of the window number; The specific limit filtering algorithm is as follows: Sample the new power generation each time and compare it with the previous sampled value. If the power change exceeds the set threshold, set the change value to the maximum allowable value and replace the original output value with this value.

2. The power smoothing method for a port microgrid wind energy farm according to claim 1, characterized in that: In the step S1, the relationship between the output power of the wind turbine and the actual wind speed is: Wherein, P wind represents the output power of the wind turbine; P w_rate represents the rated power of the wind turbine; v s represents the cut-in wind speed; v r represents the rated wind speed; v c represents the cut-out wind speed; v represents the wind speed at the height of the wind turbine shaft.

3. A method for suppressing the power fluctuations of a wind farm in a port microgrid according to claim 2, characterized in that: In the step S1, the conversion relationship between the wind speed value at the height of the wind turbine shaft and the measured wind speed value is specifically: In the formula, h represents the height of the wind turbine shaft; h0 is the height of the wind speed measurement position; v0 is the wind speed at the wind speed measurement position; μ is the friction coefficient.

4. A method for suppressing the power fluctuations of a wind farm in a port microgrid according to claim 1, characterized in that: The sum of squared errors criterion function in the step S2 is specifically: Where J is the error criterion function; K1 is the number of clusters; R j represents the j-th cluster; c j is the centroid of R j ; m is the sample index; M is the number of samples; x m represents the m-th sample; d mj indicates whether the m-th sample belongs to R j .

5. A method for suppressing the power fluctuation of a wind farm in a port microgrid according to claim 1, characterized in that: In the step S3, the grid-connected power based on the sliding filter algorithm is expressed as: where L is the sliding window width, and P G (t) is the grid-connected power after smoothing at time t; The expression of the grid-connected power based on the limit filtering algorithm is: where, ΔP pos represents the positive limit value of the power generation power fluctuation, and ΔP neg represents the negative limit value of the power generation power fluctuation, and P lim (t) represents the grid-connected power after amplitude limiting filtering at time t.

6. A non-transitory machine-readable storage medium, characterized in that: It stores executable code, and when the executable code is executed by the processor of the electronic device, the processor is caused to execute a method for suppressing the power of a wind energy power station in a port microgrid as described in any one of claims 1-5 above.

7. A computer device, characterized in that: It includes: A memory, on which executable code is stored; A processor, configured to execute the executable code, so that the computer device executes the operations of a method for suppressing the power of a wind energy power station in a port microgrid as described in any one of claims 1-5.

Images