Photovoltaic grid-connected side energy storage cluster control method and device, terminal equipment and computer readable storage medium

Abstract

The invention discloses a photovoltaic grid-connected side energy storage cluster control method and device, terminal equipment and a computer readable storage medium. The method comprises the following steps: acquiring load characteristic data of a photovoltaic power station at the current moment; inputting the load characteristic data into a trained load prediction model, and generating a predicted load value of the photovoltaic power station at a selected moment in the future; acquiring the illumination intensity and the photovoltaic cell temperature at the current moment, and calculating the actual photovoltaic output at the current moment; calculating target energy storage power of each energy storage unit in the energy storage cluster according to the predicted load value and the actual photovoltaic output; the actual photovoltaic output power, the actual load and the actual energy storage power at each moment between the current moment and the selected moment are obtained in real time, and the power deviation is calculated; and adjusting the target energy storage power according to the power deviation, and enabling each energy storage unit in the energy storage cluster to store energy according to the adjusted energy storage power. By implementing the method, the demand change of the power system can be smoothly responded.

Description

Technical Field

[0001] The present invention relates to the field of energy storage control, and in particular to a photovoltaic grid-connected side energy storage cluster control method, device, terminal equipment and computer-readable storage medium. Background Art

[0002] As the world pays more attention to environmental protection and sustainable development, energy transformation is accelerating. As a clean and renewable energy utilization method, solar photovoltaic power generation accounts for an increasing proportion in the power system. However, since photovoltaic output depends on the intensity of solar radiation, it is affected by factors such as weather changes (such as sunny, cloudy, cloudy, etc.) and day and night alternation, and has significant intermittent and volatile characteristics. This unstable output characteristic brings many challenges to the safe and stable operation of photovoltaic grid-connected power, and may cause photovoltaic grid-connected voltage fluctuation problems. Summary of the invention

[0003] The embodiments of the present invention provide a photovoltaic grid-connected side energy storage cluster control method, device, terminal equipment and computer-readable storage medium, which can smoothly respond to demand changes of the power system and significantly enhance the stability and reliability of photovoltaic grid connection.

[0004] An embodiment of the present invention provides a photovoltaic grid-connected energy storage cluster control method, comprising:

[0005] Obtain the load characteristic data of the photovoltaic power station at the current moment; the load characteristic data includes: historical load value, light intensity, ambient temperature and the time period to which it belongs;

[0006] Input the load characteristic data into the trained load forecasting model to generate the predicted load value of the photovoltaic power station at a selected time in the future;

[0007] Obtain the current light intensity and photovoltaic cell temperature, and calculate the actual photovoltaic output at the current moment based on the light intensity and photovoltaic cell temperature;

[0008] Calculate the target energy storage power of each energy storage unit in the energy storage cluster based on the predicted load value and actual photovoltaic output;

[0009] Real-time acquisition of the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment;

[0010] Whenever the actual photovoltaic output power, actual load and actual energy storage power are obtained at a certain moment, the power deviation is calculated according to the actual photovoltaic output power, actual load and actual energy storage power; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is required to store energy according to the adjusted energy storage power.

[0011] Furthermore, the target energy storage power of each energy storage unit in the energy storage cluster includes: the dischargeable power of the energy storage unit to be discharged and the chargeable power of the energy storage unit to be charged;

[0012] According to the predicted load value and actual photovoltaic output, the target energy storage power of each energy storage unit in the energy storage cluster is calculated, including:

[0013] Calculate the target total energy storage power of the energy storage cluster based on the predicted load value and actual photovoltaic output;

[0014] Obtain the state of charge and rated capacity of each energy storage unit in the energy storage cluster;

[0015] When the target total energy storage power is greater than 0, the energy storage units with a state of charge greater than the preset minimum state of charge are regarded as energy storage units to be discharged, and the dischargeable power of the energy storage units to be discharged is calculated by the following formula:

[0016]

[0017] When the target total energy storage power is less than 0, the energy storage units with a state of charge less than the preset maximum state of charge are taken as energy storage units to be charged, and the rechargeable power of the energy storage units to be charged is calculated by the following formula:

[0018]

[0019] Among them, P d,i (t) represents the dischargeable power of the i-th energy storage unit to be discharged, P es (t) represents the target total energy storage power of the energy storage cluster, SOC d,i (t) represents the state of charge of the i-th energy storage unit to be discharged, SOC min Indicates the preset minimum state of charge, E d,i represents the rated capacity of the i-th energy storage unit to be discharged, D represents the set of energy storage units to be discharged; P c,i (t) represents the rechargeable power of the i-th energy storage unit to be charged, SOC c,i (t) represents the state of charge of the i-th energy storage unit to be charged, SOC max Indicates the preset maximum state of charge, E c,i represents the rated capacity of the i-th energy storage unit to be charged, and C represents the set of energy storage units to be charged.

[0020] Furthermore, the power deviation is calculated based on the actual photovoltaic output power, the actual load and the actual energy storage power, including:

[0021] The power deviation is obtained by subtracting the sum of the actual load and the actual energy storage power from the actual PV output power.

[0022] Furthermore, according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, including:

[0023] According to the power deviation, the target energy storage power is adjusted by the following formula to obtain the adjusted energy storage power:

[0024]

[0025] Where u(t) represents the adjusted energy storage power, K p Represents the proportionality coefficient, K i Indicates the integral coefficient, K d represents the differential coefficient, ΔP(t) represents the power deviation, and κ represents the future κ time.

[0026] Furthermore, the load forecasting model is trained in the following way:

[0027] Obtaining a number of training samples; the training samples include: load characteristic data of the photovoltaic power station at a historical moment, and the historical actual load value of the photovoltaic power station at the corresponding historical selected moment;

[0028] Perform outlier processing and normalization processing on each training sample to obtain a preprocessed training sample;

[0029] A number of preprocessed training samples are input into the load forecasting model to be trained for iterative training until the loss function converges or reaches a preset number of training rounds, thereby obtaining a trained load forecasting model; wherein, in each iterative training, the preprocessed load characteristic data of the photovoltaic power station at a historical moment is used as input, the historical predicted load value is used as output, and the loss function is calculated based on the historical predicted load value and the historical actual load value.

[0030] Furthermore, outlier processing and normalization processing are performed on each training sample to obtain preprocessed training samples, including:

[0031] Compare each data value in each training sample with a preset abnormal threshold, and regard the data value exceeding the preset abnormal threshold as an abnormal value;

[0032] Use linear interpolation or mean substitution to process outliers and obtain cleaned training samples;

[0033] The cleaned training samples are normalized to obtain preprocessed training samples.

[0034] Furthermore, the actual photovoltaic output at the current moment is calculated based on the light intensity and the photovoltaic cell temperature, including:

[0035] According to the light intensity and the temperature of the photovoltaic cell, the photovoltaic output calculated based on the physical model is calculated by the following formula:

[0036]

[0037] According to the light intensity and the temperature of the photovoltaic cell, the photovoltaic output based on the statistical model is calculated by the following formula:

[0038] P pv2 (t)=β0+β1G(t)+β2T(t)+ε(t);

[0039] The photovoltaic output calculated based on the physical model and the photovoltaic output calculated based on the statistical model are weightedly added to obtain the actual photovoltaic output at the current moment;

[0040] Among them, P pv1 (t) represents the photovoltaic output calculated based on the physical model, P STC represents the photovoltaic power under standard test conditions, G(t) represents the light intensity at the current moment, and G STC represents the light intensity under standard test conditions, k represents the temperature coefficient, T(t) represents the photovoltaic cell temperature at the current moment, T STC Indicates the photovoltaic cell temperature under standard test conditions, P pv2 (t) represents the PV output calculated based on the statistical model, β0, β1 and β2 represent regression coefficients, and ε(t) represents the error term.

[0041] Based on the above method embodiment, the present invention provides a corresponding device embodiment, including: a load characteristic data acquisition module, a load value prediction module, a photovoltaic output calculation module, an energy storage unit power calculation module, an actual data monitoring module and an energy storage power adjustment module;

[0042] The load characteristic data acquisition module is used to obtain the load characteristic data of the photovoltaic power station at the current moment; the load characteristic data includes: historical load value, light intensity, ambient temperature and the time period to which it belongs;

[0043] A load value prediction module is used to input load characteristic data into a trained load prediction model to generate a predicted load value of the photovoltaic power station at a selected time in the future;

[0044] The photovoltaic output calculation module is used to obtain the light intensity and photovoltaic cell temperature at the current moment, and calculate the actual photovoltaic output at the current moment based on the light intensity and photovoltaic cell temperature;

[0045] The energy storage unit power calculation module is used to calculate the target energy storage power of each energy storage unit in the energy storage cluster based on the predicted load value and the actual photovoltaic output;

[0046] The actual data monitoring module is used to obtain the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment in real time;

[0047] The energy storage power adjustment module is used to calculate the power deviation according to the actual photovoltaic output power, actual load and actual energy storage power every time the actual photovoltaic output power, actual load and actual energy storage power are obtained at a certain moment; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is required to store energy according to the adjusted energy storage power.

[0048] Based on the above method embodiment, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the steps of the photovoltaic grid-connected side energy storage cluster control method as described in the present invention are implemented.

[0049] Based on the above method embodiment, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which controls the device where the computer-readable storage medium is located to execute the steps of the photovoltaic grid-connected side energy storage cluster control method as described in the present invention when the computer program is running.

[0050] Compared with the prior art, the beneficial effects of the embodiment of this solution are:

[0051] The present invention obtains the load characteristic data of the photovoltaic power station at the current moment, wherein the load characteristic data includes historical load values, light intensity, ambient temperature and the time period to which it belongs, and then inputs the load characteristic data into a trained load prediction model to generate a predicted load value of the photovoltaic power station at a selected time in the future. The load prediction can predict the power load demand at the future moment in advance, and then obtains the light intensity and photovoltaic cell temperature at the current moment, and calculates the actual photovoltaic output at the current moment based on the light intensity and the photovoltaic cell temperature; calculates the target energy storage power of each energy storage unit in the energy storage cluster based on the predicted load value and the actual photovoltaic output, and determines the energy storage state that the energy storage unit should reach in the future time period, so as to balance the supply and demand relationship and ensure the power system stable operation; then, the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment are obtained in real time; when the actual photovoltaic output power, actual load and actual energy storage power at each moment are obtained, the power deviation is calculated according to the actual photovoltaic output power, actual load and actual energy storage power, where the power deviation represents the difference between the actual photovoltaic output power, actual load and actual energy storage power, reflecting the balance of photovoltaic grid connection; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is adjusted according to the adjusted energy storage power, so as to perform scheduling control more smoothly and enhance the stability and reliability of photovoltaic grid connection.

[0052] In summary, the present invention can smoothly respond to changes in power system demand through load forecasting, power scheduling of energy storage units, and real-time monitoring of photovoltaic grid-connected balance, thereby significantly enhancing the stability and reliability of photovoltaic grid-connected balance. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flow chart of a photovoltaic grid-connected side energy storage cluster control method provided by an embodiment of the present invention;

[0054] Figure 2 It is a flowchart of a load forecasting model training process provided by an embodiment of the present invention;

[0055] Figure 3 is a comprehensive diagram of photovoltaic-energy storage system performance analysis provided by an embodiment of the present invention; wherein, Figure 3 (a) is a time line chart of load forecast and photovoltaic power generation. Figure 3 (b) is the time line graph of the energy storage system status. Figure 3 (c) is the time histogram of the difference between PV power generation and load;

[0056] Figure 4 It is a structural schematic diagram of a photovoltaic grid-connected side energy storage cluster control device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] like Figure 1 As shown, an embodiment of the present invention provides a photovoltaic grid-connected side energy storage cluster control method, the method comprising at least the following steps:

[0059] Step S1: Obtain the load characteristic data of the photovoltaic power station at the current moment; the load characteristic data includes: historical load value, light intensity, ambient temperature and time period;

[0060] For step S1, a series of load characteristic data of the photovoltaic power station at the current time t is collected, wherein the load characteristic data includes historical load values, light intensity, ambient temperature and the time period to which it belongs. Specifically, the historical load value refers to the load value at the last H moments, for example, the load value per minute in the past hour, or the load value per hour in the past 24 hours, etc. The value of H needs to be determined according to actual needs. These load values ​​are presented in the form of time series, revealing the trend of load changes. In addition, in the operation of the photovoltaic power station, the change of load often shows certain periodicity and seasonal laws, which are closely related to the time period. Therefore, the present invention identifies the specific time range corresponding to the current time by the time period to which it belongs. In this embodiment, a 24-hour timing system is used to clarify the time period.

[0061] In order to facilitate subsequent data processing and analysis, the above load characteristic data are presented in the form of vectors.

[0062] Step S2: inputting the load characteristic data into the trained load prediction model to generate a predicted load value of the photovoltaic power station at a selected time in the future;

[0063] For step S2, in actual applications, the load characteristic data collected in step S1 often have outliers, missing values, and inconsistencies in different characteristic dimensions. These problems will directly affect the accuracy and stability of the load forecasting model. Therefore, before the load characteristic data is input into the load forecasting model, the load characteristic data vector of step S1 is processed for outliers and normalized to remove outliers, fill in missing values, and normalize data, so as to ensure data quality and improve the accuracy of the model. Then, the preprocessed load characteristic data vector x = (x1, x2, ..., x n) is input into the trained load forecasting model, which is built based on machine learning or deep learning algorithms. It can predict the predicted load value of the photovoltaic power station at the future time t+κ based on the load characteristic data vector. In this embodiment, the predicted load value of the photovoltaic power station after 3 moments is selected, so κ=3.

[0064] The following is a detailed description of the training process of the load forecasting model:

[0065] like Figure 2 As shown in Figure 1, the training process of the load forecasting model includes the following steps:

[0066] Step S201: obtaining a number of training samples; the training samples include: load characteristic data of the photovoltaic power station at a historical moment, and the historical actual load value of the photovoltaic power station at the corresponding historical selected moment;

[0067] For step S201, before training the load prediction model, a series of training samples need to be obtained first. These training samples consist of two parts: the load characteristic data of the photovoltaic power station at the historical moment, and the actual load value of the photovoltaic power station at the historical selected moment corresponding to these characteristic data. Among them, the load characteristic data of the photovoltaic power station at the historical moment includes the historical load value at the historical moment, the light intensity at the historical moment, the ambient temperature at the historical moment, and the time period at the historical moment.

[0068] It should be noted that the time span of historical load values ​​between training samples should be long enough to cover the impact of factors such as different seasons, weather and date types on the load, which can help the model learn more comprehensive load change patterns.

[0069] Step S202: performing outlier processing and normalization processing on each training sample to obtain a preprocessed training sample;

[0070] In a preferred embodiment, outlier processing and normalization processing are performed on each training sample to obtain a preprocessed training sample, including:

[0071] Compare each data value in each training sample with a preset abnormal threshold, and regard the data value exceeding the preset abnormal threshold as an abnormal value;

[0072] Use linear interpolation or mean substitution to process outliers and obtain cleaned training samples;

[0073] The cleaned training samples are normalized to obtain preprocessed training samples.

[0074] For step S202, since the collected training samples may have various problems, such as outliers, inconsistent data ranges, etc., preprocessing is required. The present invention performs outlier processing and normalization processing on each training sample. Specifically, an abnormal threshold is pre-set to determine whether a data value is abnormal. For example, if the mean of a feature is 10 and the standard deviation is 2, then a value exceeding the mean plus twice the standard deviation (i.e., 14) may be considered abnormal. For each data value in the training sample, it will be compared with the preset abnormal threshold. If the data value exceeds this threshold, it is considered to be an outlier. Then, the outliers are processed by linear interpolation or mean substitution to obtain cleaned training samples, which ensures the quality and consistency of the training data, reduces the impact of outliers on the training of the load forecasting model, and thus improves the accuracy and robustness of the load forecasting model.

[0075] Then, the cleaned training samples are normalized using the minimum-maximum normalization formula:

[0076]

[0077] Among them, x norm represents the data after normalization, x represents the original data before normalization, and x min Represents the minimum value in the original data set, x max Represents the maximum value in the original data set.

[0078] Step S203: Input a number of preprocessed training samples into the load forecasting model to be trained for iterative training until the loss function converges or reaches a preset number of training rounds to obtain a trained load forecasting model; wherein, in each iterative training, the preprocessed load characteristic data of the photovoltaic power station at a historical moment is used as input, the historical predicted load value is used as output, and the loss function is calculated based on the historical predicted load value and the historical actual load value.

[0079] For step S203, the preprocessed training samples are divided into a training set, a validation set, and a test set according to different proportions. In this embodiment, the specific division ratio is 60% of the data is used to train the model, 20% of the data is used to verify the performance of the model, and the remaining 20% ​​of the data is used to finally test the accuracy of the model.

[0080] Construct the structure of the load prediction model, which includes the number of input layer nodes n, the number of hidden layer nodes m, the number of hidden layer layers, and the number of output layer nodes l. Then, set these parameters according to the training samples used for load prediction model training. First, based on the processed historical load characteristic data, determine the number of input layer nodes n, which depends on the number of selected influencing factors. In this embodiment, the load characteristic data of the photovoltaic power station at the historical moment in the training sample includes the historical load value (the load value of the last three moments), the light intensity at the historical moment, the ambient temperature at the historical moment, and the time period at the historical moment. Then the number of input layer nodes n=3+1+1+1+1=7; the model finally outputs the predicted value of the load, so the number of output layer nodes l is set to 1; then, the number of hidden layer nodes m is determined by the following formula:

[0081]

[0082] Among them, m represents the number of hidden layer nodes, n represents the number of input layer nodes, l represents the number of output layer nodes, and a represents a constant. a is a constant between 1 and 10; finally, the number of hidden layers is set according to the complexity of the problem and the performance requirements of the model.

[0083] Randomly initialize the weights w of the neural network connecting the input layer and the hidden layer ij , the weight w connecting the hidden layer and the output layer jk , hidden layer bias b j And the output layer bias b k , the initialization values ​​of weights and biases are usually in a small range, such as [-0.1, 0.1].

[0084] The training sample x preprocessed in step S202 i =(x i1 ,x i2 ,...,x in ) are input into the neural network one by one. For each input sample, the neural network calculates the input z of the neuron in the hidden layer. j and output a j :

[0085]

[0086] a j =f(z j )

[0087] Wherein, f(·) represents an activation function. In this embodiment, a ReLU activation function f(x)=max(0,x) is used;

[0088] At the output layer, the predicted value is calculated

[0089]

[0090] Wherein, g(·) represents a linear function. In this embodiment, g(x)=x;

[0091] The mean square error (MSE) is used as the loss function to calculate the error between the predicted value and the actual value:

[0092]

[0093] Where N represents the number of training samples, y i Indicates the actual load value. Indicates the predicted load value.

[0094] Then, the load forecasting model is back-propagated to update the weights and biases. Specifically, the gradient of the loss function with respect to the weights and biases is calculated. First, the gradient of the output layer is calculated:

[0095]

[0096] Then, calculate the gradient of the hidden layer:

[0097]

[0098] Update the weights and biases using the gradient descent algorithm:

[0099]

[0100] Where a represents the learning rate, which is used to control the step size of each update and is usually selected between [0.001, 0.1].

[0101] After each training iteration (epoch), the validation set is used to evaluate the model performance and the MSE of the validation set is calculated. When the MSE of the validation set no longer decreases (or the decrease is less than the preset threshold) or the maximum number of training rounds is reached, the training is stopped and the trained load forecasting model is obtained.

[0102] Step S3: obtaining the light intensity and photovoltaic cell temperature at the current moment, and calculating the actual photovoltaic output at the current moment according to the light intensity and photovoltaic cell temperature;

[0103] For step S3, the light intensity G(t) and the photovoltaic cell temperature T(t) at the current moment are obtained, and the photovoltaic output is calculated using a weighted physical model-statistical model.

[0104] Preferably, the actual photovoltaic output at the current moment is calculated according to the light intensity and the photovoltaic cell temperature, including:

[0105] According to the light intensity and the temperature of the photovoltaic cell, the photovoltaic output calculated based on the physical model is calculated by the following formula:

[0106]

[0107] According to the light intensity and the temperature of the photovoltaic cell, the photovoltaic output based on the statistical model is calculated by the following formula:

[0108] P pv2 (t)=β0+β1G(t)+β2T(t)+ε(t);

[0109] The photovoltaic output calculated based on the physical model and the photovoltaic output calculated based on the statistical model are weightedly added to obtain the actual photovoltaic output at the current moment;

[0110] Among them, P pv1 (t) represents the photovoltaic output calculated based on the physical model, P STC represents the photovoltaic power under standard test conditions, G(t) represents the light intensity at the current moment, and G STC represents the light intensity under standard test conditions, k represents the temperature coefficient, T(t) represents the photovoltaic cell temperature at the current moment, T STC Indicates the photovoltaic cell temperature under standard test conditions, P pv2 (t) represents the PV output calculated based on the statistical model, β0, β1 and β2 represent regression coefficients, and ε(t) represents the error term.

[0111] Specifically, the photovoltaic output P calculated based on the physical model pv1 (t) Depends on the physical characteristics of photovoltaic cells and the photovoltaic output P calculated based on statistical models pv2 (t) Use historical data and regression analysis to establish the relationship between light intensity, photovoltaic cell temperature and photovoltaic output. In order to combine the advantages of the two models and improve the accuracy and robustness of the prediction, the photovoltaic output calculated based on the physical model and the photovoltaic output calculated based on the statistical model are weighted and added. The mathematical formula is as follows:

[0112] P pv (t) = w1P pv1 (t)+w2P pv2 (t)

[0113] Among them, P pv (t) represents the actual photovoltaic output at the current moment, w1 represents the weight of the photovoltaic output calculated based on the physical model, and w2 represents the weight of the photovoltaic output calculated based on the statistical model.

[0114] Step S4: Calculate the target energy storage power of each energy storage unit in the energy storage cluster according to the predicted load value and the actual photovoltaic output;

[0115] Preferably, the target energy storage power of each energy storage unit in the energy storage cluster includes: the dischargeable power of the energy storage unit to be discharged and the chargeable power of the energy storage unit to be charged;

[0116] According to the predicted load value and actual photovoltaic output, the target energy storage power of each energy storage unit in the energy storage cluster is calculated, including:

[0117] Calculate the target total energy storage power of the energy storage cluster based on the predicted load value and actual photovoltaic output;

[0118] Obtain the state of charge and rated capacity of each energy storage unit in the energy storage cluster;

[0119] When the target total energy storage power is greater than 0, the energy storage units with a state of charge greater than the preset minimum state of charge are regarded as energy storage units to be discharged, and the dischargeable power of the energy storage units to be discharged is calculated by the following formula:

[0120]

[0121] When the target total energy storage power is less than 0, the energy storage units with a state of charge less than the preset maximum state of charge are taken as energy storage units to be charged, and the rechargeable power of the energy storage units to be charged is calculated by the following formula:

[0122]

[0123] Among them, P d,i (t) represents the dischargeable power of the i-th energy storage unit to be discharged, P es (t) represents the target total energy storage power of the energy storage cluster, SOC d,i (t) represents the state of charge of the i-th energy storage unit to be discharged, SOC min Indicates the preset minimum state of charge, E d,i represents the rated capacity of the i-th energy storage unit to be discharged, D represents the set of energy storage units to be discharged; P c,i (t) represents the rechargeable power of the i-th energy storage unit to be charged, SOC c,i (t) represents the state of charge of the i-th energy storage unit to be charged, SOC max Indicates the preset maximum state of charge, E c,i represents the rated capacity of the i-th energy storage unit to be charged, and C represents the set of energy storage units to be charged.

[0124] For step S4, according to the predicted load value and the actual photovoltaic output, the target total energy storage power of the energy storage cluster is calculated by the following formula:

[0125]

[0126] Real-time monitoring of the state of charge (SOC) and rated capacity (E) of each energy storage unit in the energy storage cluster i , let the SOC of the i-th energy storage unit be SOC i (t), and set the upper and lower limits of SOC min and SOC max , to prevent the energy storage unit from being damaged by overcharging or over-discharging;

[0127] The total energy storage power P calculated according to the power balance target es (t), and is distributed in the energy storage cluster. Specifically:

[0128] When P es When (t)>0, it means that the energy storage unit needs to discharge to supplement the power supply. In this case, first check whether the SOC of each energy storage unit is greater than SOC min Only the energy storage units that meet this condition, namely the energy storage units to be discharged, are eligible to participate in the discharge process. Subsequently, the discharge power is allocated according to the proportion of the remaining dischargeable capacity of the energy storage units to be discharged, and the formula is as follows:

[0129]

[0130] When P es When (t)<0, it means that there is excess power that needs to be charged by the energy storage unit. In this case, check whether the SOC of each energy storage unit is less than SOC max Only energy storage units that meet this condition, namely energy storage units to be charged, are eligible to participate in the charging process. Subsequently, the charging power is allocated according to the proportion of the remaining rechargeable capacity of the energy storage units to be charged, and the formula is as follows:

[0131]

[0132] Among them, P d,i (t) represents the dischargeable power of the i-th energy storage unit to be discharged, P es (t) represents the target total energy storage power of the energy storage cluster, SOC d,i (t) represents the state of charge of the i-th energy storage unit to be discharged, SOC min Indicates the preset minimum state of charge, E d,i represents the rated capacity of the i-th energy storage unit to be discharged, D represents the set of energy storage units to be discharged; P c,i (t) represents the rechargeable power of the i-th energy storage unit to be charged, SOC c,i (t) represents the state of charge of the i-th energy storage unit to be charged, SOC max Indicates the preset maximum state of charge, E c,irepresents the rated capacity of the i-th energy storage unit to be charged, and C represents the set of energy storage units to be charged.

[0133] It should be noted that by scheduling SOC during discharge min The above energy storage units are discharged and the SOC is scheduled during charging. max Charging the following energy storage units can ensure that each energy storage unit operates within its optimal operating range, improve the utilization rate of the energy storage unit, and reduce energy loss caused by overcharging and discharging.

[0134] In addition, the dispatching cost can be monitored to adjust the charging and discharging power, further optimizing the economic efficiency of the energy storage system. To achieve this goal, a cost function C(t) is introduced to describe the dispatching cost, which comprehensively considers the cost of energy storage charging and discharging:

[0135] C(t)=C c (t)P c (t)+C d (t)P d (t)

[0136] Among them, C c (t) represents the energy storage charging cost coefficient, C d (t) represents the energy storage discharge cost coefficient, P c (t) represents the energy storage charging power, P d (t) represents the energy storage discharge power.

[0137] The scheduling cost is the integral of the cost function over a certain time range, which reflects the total cost of the system over the entire time period:

[0138]

[0139] Among them, J(t) represents the scheduling cost from time t to time t+κ.

[0140] In practical applications, the charging and discharging power of the energy storage system can be adjusted according to changes in electricity market prices to reduce the total dispatch cost. For example, the charging power can be increased when the electricity market price is low to store more electricity and discharge it when the electricity market price is higher in the future, and the discharging power can be increased when the electricity market price is high, thereby reducing the total dispatch cost.

[0141] Step S5: Real-time acquisition of the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment;

[0142] For step S5, between time t and t+k, the photovoltaic output power, actual load and energy storage power data of the photovoltaic power station are collected in real time by monitoring equipment. In addition, appropriate data processing tools and algorithms can be used to clean and normalize the data to ensure the accuracy and completeness of the data.

[0143] Step S6: whenever the actual photovoltaic output power, actual load and actual energy storage power at a certain moment are obtained, the power deviation is calculated according to the actual photovoltaic output power, actual load and actual energy storage power; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is required to store energy according to the adjusted energy storage power.

[0144] In a preferred embodiment, the power deviation is calculated according to the actual photovoltaic output power, the actual load and the actual energy storage power, including:

[0145] The power deviation is obtained by subtracting the sum of the actual load and the actual energy storage power from the actual PV output power.

[0146] For step S6, when both photovoltaic power generation and load consumption fluctuate and change, in order to ensure that the system can operate smoothly and effectively utilize photovoltaic power generation resources, it is necessary to dynamically adjust the charging and discharging power of the energy storage system according to the actual situation to smooth power fluctuations and reduce the impact of power fluctuations on grid stability. Specifically, at each moment between t and t+k, the actual photovoltaic output power is subtracted from the sum of the actual load and the actual energy storage power to calculate the power deviation, which is mathematically expressed as follows:

[0147]

[0148] Among them, ΔP(t) represents the power deviation at time t, L act (t) represents the actual load at time t, represents the actual photovoltaic output power at time t, Represents the actual energy storage power at time t.

[0149] Next, the charging and discharging power of the energy storage is adjusted according to the power deviation. The proportional-integral-derivative (PID) controller method can be used to dynamically adjust the energy storage power to achieve precise control of the system operation.

[0150] Preferably, adjusting the target energy storage power according to the power deviation to obtain the adjusted energy storage power includes:

[0151] According to the power deviation, the target energy storage power is adjusted by the following formula to obtain the adjusted energy storage power:

[0152]

[0153] Where u(t) represents the adjusted energy storage power, K p Represents the proportionality coefficient, K i Indicates the integral coefficient, K d represents the differential coefficient, ΔP(t) represents the power deviation, and k represents the kth time in the future.

[0154] Specifically, the proportional, integral and differential coefficients in the PID controller correspond to the proportional control, integral control and differential control of the system respectively. By comprehensively utilizing the control effects of these three parts, it is possible to better respond to system changes and achieve accurate control.

[0155] Finally, the adjusted power instructions are redistributed to each energy storage unit in the energy storage cluster and executed according to the adjusted energy storage power allocation strategy. The adjusted energy storage power can better adapt to real-time photovoltaic power generation and load fluctuations, effectively smooth power fluctuations, reduce dependence on traditional power grids, and improve the utilization efficiency of clean energy.

[0156] In order to verify the effect of the present invention, a complex photovoltaic-energy storage system architecture was constructed in MATLAB, and the photovoltaic grid-connected side energy storage cluster control method proposed in the present invention was executed for testing.

[0157] like Figure 3 (a) shows a time line graph of load forecast and photovoltaic power generation, indicating the fluctuations of load forecast and photovoltaic power generation during the day. The figure shows obvious fluctuations between peaks and troughs, with a significant increase in load during the day and higher output of photovoltaic power generation during the peak period of the day.

[0158] like Figure 3 (b) shows a time line graph of the energy storage system status, which shows the fluctuation of the energy storage system status. Due to the large fluctuation of load, the energy storage system charges when there is excess power and discharges when there is insufficient power, resulting in significant fluctuations in the energy storage status.

[0159] like Figure 3 (c) shows the time bar graph of the difference between photovoltaic power generation and load, which shows the power difference between photovoltaic power generation and load. The frequent alternation of positive and negative values ​​indicates that the system needs to perform charging and discharging operations frequently, resulting in increased fluctuations in the energy storage state.

[0160] like Figure 4 As shown, based on the above method embodiment, a corresponding device embodiment is provided;

[0161] An embodiment of the present invention provides a photovoltaic grid-connected energy storage cluster control device, including: a load characteristic data acquisition module, a load value prediction module, a photovoltaic output calculation module, an energy storage unit power calculation module, an actual data monitoring module, and an energy storage power adjustment module;

[0162] The load characteristic data acquisition module is used to obtain the load characteristic data of the photovoltaic power station at the current moment; the load characteristic data includes: historical load value, light intensity, ambient temperature and the time period to which it belongs;

[0163] A load value prediction module is used to input load characteristic data into a trained load prediction model to generate a predicted load value of the photovoltaic power station at a selected time in the future;

[0164] The photovoltaic output calculation module is used to obtain the light intensity and photovoltaic cell temperature at the current moment, and calculate the actual photovoltaic output at the current moment based on the light intensity and photovoltaic cell temperature;

[0165] The energy storage unit power calculation module is used to calculate the target energy storage power of each energy storage unit in the energy storage cluster based on the predicted load value and the actual photovoltaic output;

[0166] The actual data monitoring module is used to obtain the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment in real time;

[0167] The energy storage power adjustment module is used to calculate the power deviation according to the actual photovoltaic output power, actual load and actual energy storage power every time the actual photovoltaic output power, actual load and actual energy storage power are obtained at a certain moment; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is required to store energy according to the adjusted energy storage power.

[0168] It can be understood that the above-mentioned device item embodiments correspond to the method item embodiments of the present invention, and can implement the photovoltaic grid-connected side energy storage cluster control method provided by any of the above-mentioned method item embodiments of the present invention.

[0169] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art may understand and implement the present invention without creative work.

[0170] Based on the above-mentioned embodiment of the photovoltaic grid-connected side energy storage cluster control method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the photovoltaic grid-connected side energy storage cluster control method based on any embodiment of the present invention is implemented.

[0171] Exemplarily, in this embodiment, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more module elements may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the terminal device.

[0172] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0173] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0174] Based on the above method embodiment, another embodiment is provided: another embodiment of the present invention provides a computer-readable storage medium, including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the photovoltaic grid-connected side energy storage cluster control method described in any one of the above method embodiments of the present invention.

[0175] Wherein, the module / unit based on the photovoltaic grid-connected energy storage cluster control device / terminal device integration, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0176] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A photovoltaic grid-connected energy storage cluster control method, characterized in that: include: Obtain the load characteristic data of the photovoltaic power station at the current moment; The load characteristic data includes: historical load value, light intensity, ambient temperature and the time period to which it belongs; Inputting the load characteristic data into a trained load prediction model to generate a predicted load value of the photovoltaic power station at a selected time in the future; Obtaining the light intensity and photovoltaic cell temperature at the current moment, and calculating the actual photovoltaic output at the current moment based on the light intensity and photovoltaic cell temperature; Calculating the target energy storage power of each energy storage unit in the energy storage cluster according to the predicted load value and the actual photovoltaic output; Real-time acquisition of the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment; Whenever the actual photovoltaic output power, actual load and actual energy storage power are obtained at a certain moment, the power deviation is calculated according to the actual photovoltaic output power, actual load and actual energy storage power; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is required to store energy according to the adjusted energy storage power.

2. The photovoltaic grid-connected energy storage cluster control method according to claim 1, characterized in that: The target energy storage power of each energy storage unit in the energy storage cluster includes: the dischargeable power of the energy storage unit to be discharged and the chargeable power of the energy storage unit to be charged; The calculating the target energy storage power of each energy storage unit in the energy storage cluster according to the predicted load value and the actual photovoltaic output includes: Calculating a target total energy storage power of the energy storage cluster according to the predicted load value and the actual photovoltaic output; Obtain the state of charge and rated capacity of each energy storage unit in the energy storage cluster; When the target total energy storage power is greater than 0, the energy storage unit with a state of charge greater than the preset minimum state of charge is used as the energy storage unit to be discharged, and the dischargeable power of the energy storage unit to be discharged is calculated by the following formula: When the target total energy storage power is less than 0, the energy storage unit whose state of charge is less than the preset maximum state of charge is taken as the energy storage unit to be charged, and the chargeable power of the energy storage unit to be charged is calculated by the following formula: Among them, P d，i (t) represents the dischargeable power of the i-th energy storage unit to be discharged, P es (t) represents the target total energy storage power of the energy storage cluster, SOC d，i (t) represents the state of charge of the i-th energy storage unit to be discharged, SOC min Indicates the preset minimum state of charge, E d，i represents the rated capacity of the i-th energy storage unit to be discharged, D represents the set of energy storage units to be discharged; P c，i (t) represents the rechargeable power of the i-th energy storage unit to be charged, SOC c，i (t) represents the state of charge of the i-th energy storage unit to be charged, SOC max Indicates the preset maximum state of charge, E c，i represents the rated capacity of the i-th energy storage unit to be charged, and C represents the set of energy storage units to be charged.

3. The photovoltaic grid-connected energy storage cluster control method according to claim 2, characterized in that: Calculate the power deviation based on the actual PV output power, actual load and actual energy storage power, including: The power deviation is obtained by subtracting the sum of the actual load and the actual energy storage power from the actual PV output power.

4. The photovoltaic grid-connected energy storage cluster control method according to claim 3 is characterized in that: According to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, including: According to the power deviation, the target energy storage power is adjusted by the following formula to obtain the adjusted energy storage power: Where u(t) represents the adjusted energy storage power, K p Represents the proportionality coefficient, K i Indicates the integral coefficient, K d represents the differential coefficient, ΔP(t) represents the power deviation, and κ represents the future κ time.

5. The photovoltaic grid-connected energy storage cluster control method according to claim 1, characterized in that: The load forecasting model is trained in the following way: Acquire a number of training samples; the training samples include: load characteristic data of the photovoltaic power station at a historical moment, and the historical actual load value of the photovoltaic power station at a corresponding historical selected moment; Perform outlier processing and normalization processing on each training sample to obtain a preprocessed training sample; A number of preprocessed training samples are input into the load forecasting model to be trained for iterative training until the loss function converges or reaches a preset number of training rounds, thereby obtaining a trained load forecasting model; wherein, in each iterative training, the preprocessed load characteristic data of the photovoltaic power station at a historical moment is used as input, the historical predicted load value is used as output, and the loss function is calculated based on the historical predicted load value and the historical actual load value.

6. The photovoltaic grid-connected energy storage cluster control method according to claim 5, characterized in that: Perform outlier processing and normalization on each training sample to obtain preprocessed training samples, including: Compare each data value in each training sample with a preset abnormal threshold, and regard the data value exceeding the preset abnormal threshold as an abnormal value; Use linear interpolation or mean substitution to process outliers and obtain cleaned training samples; The cleaned training samples are normalized to obtain preprocessed training samples.

7. The photovoltaic grid-connected energy storage cluster control method according to claim 1, characterized in that: According to the light intensity and the photovoltaic cell temperature, the actual photovoltaic output at the current moment is calculated, including: According to the light intensity and the temperature of the photovoltaic cell, the photovoltaic output calculated based on the physical model is calculated by the following formula: According to the light intensity and the photovoltaic cell temperature, the photovoltaic output calculated based on the statistical model is calculated by the following formula: P pv2 (t)=β0+β1G(t)+β2T(t)+ε(t); Performing weighted addition of the photovoltaic output calculated based on the physical model and the photovoltaic output calculated based on the statistical model to obtain the actual photovoltaic output at the current moment; Among them, P pv1 (t) represents the photovoltaic output calculated based on the physical model, P STC represents the photovoltaic power under standard test conditions, G(t) represents the light intensity at the current moment, and G STC represents the light intensity under standard test conditions, k represents the temperature coefficient, T(t) represents the photovoltaic cell temperature at the current moment, T STC Indicates the photovoltaic cell temperature under standard test conditions, P pv2 (t) represents the PV output calculated based on the statistical model, β0, β1 and β2 represent regression coefficients, and ε(t) represents the error term.

8. A photovoltaic grid-connected energy storage cluster control device, characterized in that: include: Load characteristic data acquisition module, load value prediction module, photovoltaic output calculation module, energy storage unit power calculation module, actual data monitoring module and energy storage power adjustment module; The load characteristic data acquisition module is used to acquire the load characteristic data of the photovoltaic power station at the current moment; the load characteristic data includes: historical load value, light intensity, ambient temperature and the time period to which it belongs; The load value prediction module is used to input the load characteristic data into the trained load prediction model to generate a predicted load value of the photovoltaic power station at a selected time in the future; The photovoltaic output calculation module is used to obtain the light intensity and photovoltaic cell temperature at the current moment, and calculate the actual photovoltaic output at the current moment according to the light intensity and photovoltaic cell temperature; The energy storage unit power calculation module is used to calculate the target energy storage power of each energy storage unit in the energy storage cluster according to the predicted load value and the actual photovoltaic output; The actual data monitoring module is used to obtain the actual photovoltaic output power, actual load and actual energy storage power at each moment between the current moment and the selected moment in real time; The energy storage power adjustment module is used to calculate the power deviation according to the actual photovoltaic output power, the actual load and the actual energy storage power each time the actual photovoltaic output power, the actual load and the actual energy storage power are obtained at a certain moment; according to the power deviation, the target energy storage power is adjusted to obtain the adjusted energy storage power, and each energy storage unit in the energy storage cluster is allowed to store energy according to the adjusted energy storage power.

9. A terminal device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the photovoltaic grid-connected side energy storage cluster control method as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: include: A stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the photovoltaic grid-connected side energy storage cluster control method according to any one of claims 1 to 7.

Images