PCC voltage out-of-limit adjustment method based on photovoltaic power station output capability difference

Abstract

The invention provides a photovoltaic power station output capability difference-based PCC voltage out-of-limit adjustment method, which comprises the following steps of: dividing a photovoltaic power station sub-cluster, and outputting a sub-cluster division result; based on the sub-cluster division result, photovoltaic power generation cluster power prediction is carried out through an LSTM network, and a cluster total power prediction result is output; performing PCC voltage out-of-limit risk prediction based on the cluster total power prediction result, and outputting an out-of-limit risk judgment result; if the judgment result is that the out-of-limit risk exists, reactive power adjustment based on inverter coherence is executed, and PCC voltage is controlled to be within a safe range. The method is simple in principle, can realize preventive adjustment of the grid-connected point voltage out-of-limit risk, and provides guarantee for stable operation of a novel power system.

Description

Technical Field

[0001] This invention relates to the field of power system stability control, specifically a PCC voltage over-limit adjustment method based on the difference in output capacity of photovoltaic power plants. Background Technology

[0002] With the continuous advancement of the construction of new power systems, distributed photovoltaic (PV) power is exhibiting a high penetration rate in distribution networks, improving the flexibility of the power system. However, the uncertainty and volatility of its output bring about problems such as voltage fluctuations and voltage limits exceeding limits. During periods of high output power from PV power plant clusters, a large amount of power is injected into the distribution network system, increasing the risk of voltage limits exceeding the grid connection point. Power forecasting methods can be used to predict the grid connection point voltage level over future periods, thereby enabling the development of preventative adjustment strategies for voltage limit exceeding scenarios.

[0003] To address voltage exceeding limits, one approach is to regulate voltage through on-load tap changer on transformers or in combination with switching capacitor banks. However, this method lacks specificity in terms of location and cannot provide continuous, smooth adjustment, making it impractical and limiting its application scenarios. Another approach is to configure an energy storage system. However, the energy storage model used in simulation studies is a simplified model, thus failing to accurately assess the full lifecycle utilization value of energy storage. Furthermore, adding hardware inevitably increases costs. Summary of the Invention

[0004] To prevent the risk of grid connection point voltage exceeding limits in new power systems, this invention proposes a method for adjusting the voltage exceeding limits at the point of common coupling (PCC) based on the difference in output capacity of photovoltaic power plants. The method is based on the division of photovoltaic sub-clusters and then predicts the power of the photovoltaic clusters. The predicted power is used to obtain the PCC voltage situation for future time periods. Under the premise of judging that there is a risk of voltage exceeding limits, the reactive power is adjusted using a multi-inverter synchronization model to suppress the PCC voltage exceeding limits.

[0005] The technical solution adopted in this invention is:

[0006] A method for PCC voltage over-limit adjustment based on the difference in output capacity of photovoltaic power plants includes the following steps:

[0007] Divide the photovoltaic power station into sub-clusters and output the sub-cluster division results;

[0008] Based on the sub-cluster partitioning results, the photovoltaic power generation cluster power is predicted using an LSTM network, and the total cluster power prediction result is output.

[0009] Based on the total power prediction results of the cluster, the PCC voltage over-limit risk is predicted, and the over-limit risk judgment result is output.

[0010] If the judgment result indicates that there is a risk of exceeding the limit, reactive power adjustment based on inverter synchronization is performed to control the PCC voltage within a safe range.

[0011] Furthermore, the sub-cluster division result is as follows: based on the power grid planning, geographical distribution and the degree of influence of climate conditions, the photovoltaic cluster is divided into multiple sub-clusters, so that the power stations within the same sub-cluster have similar output characteristics.

[0012] Furthermore, the process of predicting the photovoltaic power generation cluster power using an LSTM network based on the sub-cluster partitioning results, and outputting the total cluster power prediction result, specifically includes:

[0013] The actual output data of each sub-cluster over the past month was used as the input to the LSTM neural network prediction model. The resolution is 15 minutes;

[0014] Output power forecast data for each sub-cluster for the next day. ;

[0015] The total power prediction result of the cluster is obtained by summing the prediction data of each sub-cluster:

[0016] (1);

[0017] In the formula: Let be the total power of the cluster at time t. Number of sub-clusters For the first Power prediction data for each sub-cluster at time t.

[0018] Furthermore, the prediction of PCC voltage exceedance risk based on the total power prediction result of the cluster, and the output of the exceedance risk judgment result, specifically includes:

[0019] Based on cluster total power prediction results Calculate the grid connection point voltage by considering the distribution network voltage level, the equivalent total impedance of lines and transformers, and the local load. :

[0020] (2);

[0021] (3);

[0022] In the formula: For distribution network voltage levels; The equivalent total impedance of the lines and transformers between the photovoltaic power station access point and the distribution network; This represents the voltage at the photovoltaic grid connection point in the distribution network at the current moment. For local load; The remaining power after the current photovoltaic power station has been used to meet the local load;

[0023] Calculate the grid connection point voltage With preset safety threshold contrast:

[0024] (4);

[0025] If equation (4) holds true, then it is determined that the PCC voltage will exceed the limit at time t.

[0026] Furthermore, if the judgment result indicates a risk of exceeding limits, reactive power adjustment based on inverter synchronization is performed to control the PCC voltage within a safe range, specifically including:

[0027] Based on the judgment result of step (3) that the over-limit phenomenon will occur, each sub-cluster inverter is equivalent to the maximum power output unit in MPPT mode.

[0028] Based on the available regulation capacity of the equivalent inverters of each sub-cluster, the reactive power adjustment is allocated according to the linear proportional principle.

[0029] Each sub-cluster uses the remaining capacity of the inverter to adjust the reactive power output according to the allocated reactive power adjustment amount, so that the PCC voltage is limited within a safe range.

[0030] Furthermore, the allocation of reactive power adjustment based on the available regulation capability of the equivalent inverters of each sub-cluster, using a linear proportional principle, includes:

[0031] (5);

[0032] (6);

[0033] In the formula: For the first The available allocated power of the equivalent inverter of each sub-cluster at time t; For the first The predicted power of the equivalent inverter of each sub-cluster at time t.

[0034] Furthermore, each sub-cluster utilizes the remaining capacity of the inverter to regulate reactive power output according to the allocated reactive power adjustment amount, thereby limiting the PCC voltage within a safe range, including:

[0035] Predicted power of equivalent inverters for each sub-cluster based on predictions The inverter's active and reactive power outputs are controlled to regulate voltage. Based on PCC over-limit conditions, the inverter's voltage regulation state is divided into two types:

[0036] (1) State I: The PCC voltage does not exceed the limit. The reactive power output is as small as possible without voltage regulation. The reactive power of the inverter occupies the capacity. =0;

[0037] (2) State II: The PCC voltage exceeds the limit. The remaining power of the inverter is used for voltage regulation, and the remaining capacity of the inverter is used first. Reactive power regulation is performed, during which active power is increased. Still output in MPPT format;

[0038] According to the order in which the control functions are implemented, the voltage regulation states I and II of the inverter at the next moment are screened step by step. When the photovoltaic cluster operates in maximum power point tracking mode at the next moment, the PCC voltage is adjusted to the upper limit value. At that time, the reactive power regulation of the inverter's remaining capacity is calculated using equation (7). :

[0039] (7)

[0040] In the formula: This indicates the maximum active power that the photovoltaic cluster can generate in the next moment;

[0041] When calculating When, explain To reach the upper limit When the inverter needs to release reactive power for voltage regulation, it will operate in state I when the inverter satisfies equation (8). At this time, the active and reactive power output values ​​of the inverter are as shown in equation (9):

[0042] (8)

[0043] (9)

[0044] In the formula: This is the reference value for the active power output of the inverter. This is the reference value for the reactive power output of the inverter.

[0045] When calculating When this occurs, it indicates that the inverter needs to absorb reactive power and perform a voltage reduction operation in the next moment. When the inverter satisfies equation (10), it will operate in state II. After adjustment, the active and reactive power output values ​​of the inverter are as shown in equation (11):

[0046] .

[0047] The beneficial effects of this invention are as follows:

[0048] (1) In response to the changes in power flow and voltage distribution caused by large-scale photovoltaic access, which further leads to voltage over-limit situations, this paper proposes to pre-judge the risk of PCC voltage over-limit based on cluster power prediction, and proposes reactive power adjustment measures for over-limit risks. This can effectively prevent and adjust voltage over-limit situations and improve the operational reliability of the new power system.

[0049] (2) When the PCC voltage exceeds the limit in the system, the scheme utilizes the remaining capacity of the inverter for reactive power regulation while distributing the power of the inverter according to the linear proportional principle, so that inverters with different output capabilities can operate at a reasonable output level. Under the condition of overvoltage suppression, the inverter capacity is reasonably applied. The principle is simple, the regulation effect is good, and it is easy to implement in engineering. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the power output data of a power plant cluster on a certain day during the spring and winter seasons, according to an embodiment of the present invention.

[0051] Figure 2 The image shows the power prediction effect.

[0052] Figure 3 This is a graph showing the error between the actual and predicted values.

[0053] Figure 4 This is a flowchart of a PCC voltage over-limit adjustment method based on the difference in output capacity of a photovoltaic power station, according to an embodiment of the present invention.

[0054] Figure 5 This is a schematic diagram of the PCC voltage deviation when the inverter is not operating.

[0055] Figure 6 This is a schematic diagram of the PCC voltage deviation when the inverter is operating. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Please see Figure 4 This invention provides a PCC voltage over-limit adjustment method based on the difference in output capacity of photovoltaic power plants;

[0058] Step 1: Photovoltaic Power Plant Sub-cluster Division. Photovoltaic sub-cluster division involves dividing photovoltaic (PV) clusters into different sub-clusters based on different grid planning, geographical distribution, and the degree of influence from climate conditions. The division criteria are: I. Ensuring that power plants within the same sub-cluster belong to the same grid area. II. Ensuring that the prediction accuracy is not significantly affected during sub-cluster division. III. Considering the geographical conditions and degree of climate influence among different power plants, striving to ensure that PV power plants within the same sub-cluster have similar output characteristics.

[0059] Step 2: Power prediction for the photovoltaic cluster using an LSTM network. The LSTM network is an optimized version of the Recurrent Neural Network (RNN), capable of efficiently handling the gradient explosion problem. The LSTM is trained on a training set based on historical data, and a prediction model is built using the LSTM's input gate, output gate, and forget gate for prediction. First, the sample data is preprocessed, and all historical data is normalized. The processed historical data is divided into a training set and a test set. The training set is used to train the model, and the test set is used to evaluate the model's error. Based on the processed data, the cluster power is predicted using an LSTM network. LSTM network units receive... Input at any time The state of the unit group at the previous time step and the state of the neuron at the previous moment. Internal information is input into each gate, where calculations are performed to determine whether to activate the neuron, thus establishing the neuron's state. .final The output of the LSTM is formed through nonlinear function operations and dynamic control of the output gate. The expression for an LSTM network is shown below:

[0060] ;

[0061] In the formula: , , , These are the weight values ​​at the input end; , , , The output weight value of the previous time; , , , The bias parameter corresponding to the gate; , , The activation value for each gate.

[0062] In this invention, the actual power output data of each sub-cluster in the power plant over the past month is used as the input to the LSTM neural network prediction model. With a resolution of 15 minutes, the predicted power for the next day is obtained as follows: .

[0063] Step 3: Obtain the total power of the photovoltaic cluster using the cluster accumulation method. Then, obtain the PCC voltage at the predicted time and compare it with the allowable operating voltage range at the grid connection point to determine if there is a risk of exceeding the limit. Based on the prediction of the power of each sub-cluster using the LSTM network model, the total power of the entire photovoltaic power station cluster is obtained according to equation (1).

[0064] (1)

[0065] In the formula: Let be the total power of the cluster at time t. Number of sub-clusters For the first Predicted output data for each sub-cluster at time t.

[0066] The photovoltaic cluster output prediction results obtained based on equation (1) The remaining power after satisfying the local load is obtained using equation (2), and the grid connection point voltage at the predicted time is calculated using equation (3). .

[0067] (2)

[0068] (3)

[0069] In the formula: For distribution network voltage levels, The equivalent total impedance of the lines and transformers between the photovoltaic power station access point and the distribution network; This represents the voltage at the photovoltaic grid connection point in the distribution network at the current moment. For local load; The remaining power after the current photovoltaic power station has supplied enough power to meet the local load.

[0070] As shown in equation (3), when the power of a photovoltaic power station exceeds a certain value, it will inevitably cause the PCC voltage to exceed the limit. To ensure the stable operation of the power grid, according to the regulations on the photovoltaic grid connection point voltage, the PCC voltage should be maintained at... The following uses equations (3) and (4) to determine whether the grid connection point of the photovoltaic power station will experience voltage over-limit risk within the predicted time period, i.e.:

[0071] (4)

[0072] If equation (4) holds true, it can be determined that the PCC voltage will exceed the limit at time t.

[0073] Step 5: Based on the prediction that the PCC voltage will exceed the limit at time t within the predicted time, the sub-cluster inverters are equivalent and their power is adjusted. The sub-cluster inverters are equivalent according to the maximum power output of the inverters in MPPT control mode. For periods where the grid connection point voltage exceeds the limit, the power of each equivalent sub-cluster inverter is allocated according to the power adjustability of each equivalent sub-cluster, following a linear proportional principle, as shown in equation (5):

[0074] (5)

[0075] (6)

[0076] In the formula: For the first The available allocated power of the equivalent inverter of each sub-cluster at time t; For the first The predicted power of the equivalent inverter of each sub-cluster at time t, and at the same time This also represents the current upper limit of the output power of the sub-cluster photovoltaic power station.

[0077] Inverter power based on predictions for each sub-cluster The risk of PCC voltage exceeding the limit is determined by equation (4), and a reactive power control command is issued. After receiving the reactive power command, the photovoltaic power station cluster regulation reference value can be obtained by combining the cluster's predicted total power. Then, the reference value is allocated to each sub-cluster equivalent inverter according to equation (6). Each equivalent inverter uses its remaining capacity to regulate the reactive power output according to the allocated value to control the cluster's PCC voltage.

[0078] (1) State I: The PCC voltage does not exceed the limit. The reactive power output is as small as possible without voltage regulation. The reactive power of the inverter occupies the capacity. It is 0.

[0079] (2) State II: PCC voltage exceeds the limit. Voltage regulation is performed using the inverter's remaining power; in this case, the inverter's remaining capacity is utilized first. Reactive power regulation is performed, during which active power is increased. Still output in MPPT format.

[0080] According to the priority order of control function implementation, the inverter's next state I and II are screened step by step. When the active power of the photovoltaic cluster is output according to MPPT in the next moment, the PCC voltage is adjusted to the upper limit value. At that time, the reactive power regulation of the inverter's remaining capacity can be calculated using equation (7). :

[0081] (7)

[0082] In the formula This indicates the maximum active power that the photovoltaic cluster can generate at the next moment.

[0083] When calculating At that time, due to The magnitude of this is directly proportional to the reactive power output of the inverter. If at this time... =0, It will not exceed the limit. When the inverter satisfies equation (8), it will operate in state I. At this time, the active and reactive power output values ​​of the inverter are as shown in equation (10):

[0084] (8)

[0085] (9)

[0086] When calculating When, that is, when the inverter satisfies equation (10), Exceeding the limit indicates that the inverter will need to absorb reactive power and perform a voltage reduction operation in the next moment, and will operate in state II. Adjustments are made. At this time, the active and reactive power output values ​​of the inverter are as shown in equation (11):

[0087] ;

[0088] Example verification:

[0089] Six months of historical data from various aspects of a specific power plant cluster were selected. The power plants were then divided into sub-clusters according to sub-clustering criteria, resulting in historical data for each sub-cluster. Based on the regulations governing the safe operating range of grid connection voltage, with a voltage level of 220kV, the permissible voltage deviation for the PCC is [not specified]. To increase the safety margin, the voltage is set to +9.5% in this example. To improve the accuracy of the prediction data, this invention selects output data for a full month (as of the current date), with a resolution of 15 minutes. Partial data from the photovoltaic power station cluster is shown below. Figure 1 As shown.

[0090] Simulation verification and power prediction were performed using MATLAB. An LSTM network structure was built, with the power output data of each sub-cluster of the power plant within a certain period as input and the power output data for a future period as output.

[0091] Predictive capability verification: Based on the LSTM network built according to this invention, the power output data of each power plant sub-cluster in April was used as input to predict and sum the power generation data of each sub-cluster on May 3, so as to obtain the predicted power generation data of the entire cluster. Some predicted and summed data are shown in Table 1. Figure 2As shown, the predicted trend matches the actual output characteristics of the photovoltaic power station cluster. The cluster begins outputting power around 6:00 AM, with the maximum power occurring between 11:00 AM and 2:00 PM. Afterward, the power output decreases, ceasing operation around 6:00 PM, gradually dropping to zero. Figure 2 It can be seen that the predicted value curve obtained from the LSTM network is highly similar to the actual value curve on May 3rd, and the prediction error curve is... Figure 3 Provided.

[0092] Table 1. Statistics of Simulation Results

[0093]

[0094] Over-limit risk prediction and voltage regulation capability verification: based on Figure 2 Based on the data in Table 1 and Equation (4), a prediction is made regarding whether the PCC voltage will exceed its limit during the predicted time period, as shown in Table 1. Figure 5 The chart shows the voltage conditions at the grid connection point of the photovoltaic cluster during the day, indicating that when the inverter does not make any voltage adjustments, the cluster experiences significant voltage exceedance issues between 9:00 and 16:00. Regarding... Figure 5 In scenarios exceeding limits, the voltage is adjusted using the multi-inverter synchronous reactive power adjustment strategy proposed in this invention. The adjusted PCC voltage bias state is as follows: Figure 6 As shown, after adjustment, the over-limit phenomenon occurred between 9:00 and 14:00, and during the period of over-limit phenomenon, the voltage was controlled within the standard value, indicating that the strategy effectively controlled the over-limit phenomenon and over-limit time, and verified the effectiveness of the method proposed in this invention.

[0095] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for PCC voltage over-limit adjustment based on the difference in output capacity of photovoltaic power plants, characterized in that, Includes the following steps: Divide the photovoltaic power station into sub-clusters and output the sub-cluster division results; Based on the sub-cluster partitioning results, the photovoltaic power generation cluster power is predicted using an LSTM network, and the total cluster power prediction result is output. Based on the total power prediction results of the cluster, the PCC voltage over-limit risk is predicted, and the over-limit risk judgment result is output. If the judgment result indicates that there is a risk of exceeding the limit, reactive power adjustment based on inverter synchronization is performed to control the PCC voltage within a safe range.

2. The method according to claim 1, characterized in that, The specific sub-cluster division result is as follows: based on the power grid planning, geographical distribution and the degree of influence of climate conditions, the photovoltaic cluster is divided into multiple sub-clusters, so that the power stations within the same sub-cluster have similar output characteristics.

3. The method according to claim 1, characterized in that, The process of predicting the power of a photovoltaic power generation cluster using an LSTM network based on the sub-cluster partitioning results, and outputting the total cluster power prediction result, specifically includes: The actual output data of each sub-cluster over the past month was used as the input to the LSTM neural network prediction model. The resolution is 15 minutes. Output power forecast data for each sub-cluster for the next day. ; The total power prediction result of the cluster is obtained by summing the prediction data of each sub-cluster: （1）； In the formula: Let be the total power of the cluster at time t. Number of sub-clusters For the first Power prediction data for each sub-cluster at time t.

4. The method according to claim 1, characterized in that, The process of predicting PCC voltage exceedance risk based on the total power prediction result of the cluster, and outputting the exceedance risk judgment result, specifically includes: Based on cluster total power prediction results Calculate the grid connection point voltage by considering the distribution network voltage level, the equivalent total impedance of lines and transformers, and the local load. : （2）； （3）； In the formula: For distribution network voltage levels; The equivalent total impedance of the lines and transformers between the photovoltaic power station access point and the distribution network; This represents the voltage at the photovoltaic grid connection point in the distribution network at the current moment. For local load; The remaining power after the current photovoltaic power station has been used to meet the local load; Calculate the grid connection point voltage With preset safety threshold contrast: （4）； If equation (4) holds true, then it is determined that the PCC voltage will exceed the limit at time t.

5. The method according to claim 1, characterized in that, If the judgment result indicates a risk of exceeding limits, reactive power adjustment based on inverter synchronization is performed to control the PCC voltage within a safe range, specifically including: Based on the judgment that over-limit phenomena will occur, each sub-cluster inverter is equivalent to the maximum power output unit in MPPT mode; Based on the available regulation capacity of the equivalent inverters of each sub-cluster, the reactive power adjustment is allocated according to the linear proportional principle. Each sub-cluster uses the remaining capacity of the inverter to adjust the reactive power output according to the allocated reactive power adjustment amount, so that the PCC voltage is limited within a safe range.

6. The method according to claim 5, characterized in that, The allocation of reactive power adjustment based on the available regulation capacity of the equivalent inverters in each sub-cluster, using a linear proportional principle, includes: （5）； （6）； In the formula: For the first The available allocated power of the equivalent inverter of each sub-cluster at time t; For the first The predicted power of the equivalent inverter of each sub-cluster at time t.

7. The method according to claim 6, characterized in that, Each sub-cluster adjusts its reactive power output using the remaining capacity of the inverter according to the allocated reactive power adjustment amount, thereby limiting the PCC voltage within a safe range, including: Predicted power of equivalent inverters for each sub-cluster based on predictions The inverter's active and reactive power outputs are controlled to regulate voltage. Based on PCC over-limit conditions, the inverter's voltage regulation state is divided into two types: (1) State I: The PCC voltage does not exceed the limit. The reactive power output is as small as possible without voltage regulation. The reactive power of the inverter occupies the capacity. =0; (2) State II: The PCC voltage exceeds the limit. The remaining power of the inverter is used for voltage regulation, and the remaining capacity of the inverter is used first. Reactive power regulation is performed, during which active power is increased. Still output in MPPT format; According to the order in which the control functions are implemented, the voltage regulation states I and II of the inverter at the next moment are screened step by step. When the photovoltaic cluster operates in maximum power point tracking mode at the next moment, the PCC voltage is adjusted to the upper limit value. At that time, the reactive power regulation of the inverter's remaining capacity is calculated using equation (7). : （7）； In the formula: This indicates the maximum active power that the photovoltaic cluster can generate in the next moment; When calculating When, explain To reach the upper limit When the inverter needs to release reactive power for voltage regulation, it will operate in state I when the inverter satisfies equation (8). At this time, the active and reactive power output values ​​of the inverter are as shown in equation (9): （8）； （9）； In the formula: This is the reference value for the active power output of the inverter. This is the reference value for the reactive power output of the inverter. When calculating When this occurs, it indicates that the inverter needs to absorb reactive power and perform a voltage reduction operation in the next moment. When the inverter satisfies equation (10), it will operate in state II. After adjustment, the active and reactive power output values ​​of the inverter are as shown in equation (11): 。

Images