Power grid frequency modulation method, system and equipment based on probability density, product and medium

Abstract

The invention relates to the technical field of energy management, and provides a power grid frequency modulation method, system, equipment, product and medium based on probability density, comprising the following steps: calculating a system load deviation component, a tie line deviation component and a unit deviation component through operation data; obtaining a new energy output prediction model to predict new energy power generation equipment, calculating a probability density function to obtain a new energy prediction deviation, and calculating a new energy deviation component; calculating the net load power and the net load power standard deviation of the target power grid, inputting a plurality of parameters into the deep neural network, obtaining a deviation prediction result, and calculating a loss function; iterating the loss function to obtain a target loss function, and extracting a target predicted value from the target loss function; and establishing an evaluation parameter screening condition, calculating an evaluation parameter, carrying out iteration on the target prediction value to obtain a prediction result, and carrying out frequency modulation on the target power grid. The frequency modulation efficiency of the power grid is improved.

Description

Technical Field

[0001] The present invention relates to the field of energy management technology, and in particular to a power grid frequency modulation method, system, equipment, product and medium based on probability density. Background Art

[0002] With the development of new power systems, new energy sources such as wind power and photovoltaic power are continuously connected to the grid for large-scale power generation. The randomness and volatility of their power generation seriously affect the stability of the power system frequency. In order to cope with the large-scale grid connection of new energy sources such as wind power and photovoltaic power, the power system needs to provide enough AGC automatic generation control (Automatic Generation Control) frequency modulation capacity, but too much frequency modulation capacity cannot meet the economic operation of the system. Reasonable prediction of the AGC frequency modulation capacity required for system operation is of great significance to the stability of the power grid frequency and the economic operation of the system.

[0003] For regional power grids, AGC frequency regulation capacity is mainly used to balance regional minute-level power fluctuations, and to balance the power difference between actual power load and power generation output in the region with regional control deviation as a reference. At present, the calculation methods of power system frequency regulation capacity mainly include: frequency regulation capacity calculation method based on percentage. This type of method is mainly determined based on dispatching operation experience, and the system's demand for frequency regulation capacity is determined by the percentage of the maximum load of the power system. For example, the Beijing-Tianjin-Tangshan control area of ​​the North China Power Grid determines the frequency regulation capacity demand by 10% of the weekly predicted maximum load. The Shanxi Power Grid uses 5% to 15% of the maximum direct dispatch power generation demand throughout the day as the system frequency regulation capacity demand. This type of method is simple and practical, but it relies too much on dispatching experience, does not consider the volatility of new energy power generation such as wind power and photovoltaic power, and has a large error.

[0004] Frequency regulation capacity calculation method based on load deviation coverage. This method takes into account the volatility of load and unit output and calculates the future frequency regulation capacity based on operating data. This method takes into account the impact of system power change volatility factors, but only relies on operating data and lacks consideration of the actual situation of the power grid, so its applicability is low. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the related art. To this end, the present invention provides a power grid frequency regulation method, system, device, product and medium based on probability density to achieve high confidence prediction of the frequency regulation capacity of the power grid frequency regulation system.

[0006] The present invention provides a power grid frequency modulation method based on probability density, comprising: S1: acquiring operation data of a target power grid, and calculating a system load deviation component, a tie line deviation component and a unit deviation component of the target power grid through the operation data; S2: Obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model to obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; S3: Calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; S4: iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; S5: Establishing evaluation parameter screening conditions, calculating the evaluation parameters of the target prediction value, screening the target prediction value by the evaluation parameters and the evaluation parameter screening conditions, obtaining a prediction result, and frequency-modulating the target power grid by the prediction result.

[0007] According to the probability density-based power grid frequency modulation method provided by the present invention, step S1 further comprises: S11: determining the target power grid, acquiring the operating data of the target power grid, calculating the system load forecast deviation and the system load fluctuation deviation of the target power grid through the operating data, and calculating the system load deviation component through the system load forecast deviation and the system load fluctuation deviation; S12: acquiring a planned exchange value after a time period and a planned exchange value in the current time period of the tie line from the operation data, and obtaining the tie line deviation component through the planned exchange value after a time period and the planned exchange value in the current time period; S13: acquiring a planned output value after a period of time and a planned output value for the current period of time of the generator set from the operation data, and calculating the deviation component of the generator set by the planned output value after a period of time and the planned output value for the current period of time.

[0008] According to the probability density-based power grid frequency modulation method provided by the present invention, S21: determining the prediction time and the new energy power generation equipment in the target power grid, obtaining a new energy output prediction model, the new energy output prediction model predicts the power generation power of the new energy power generation equipment multiple times according to the prediction time, and obtains a prediction result; S22: determining a confidence interval of the prediction result, constructing a probability density function of the prediction result through the probability distribution of the prediction result, and obtaining the new energy prediction deviation through the confidence interval and the probability density function; S23: obtaining a new energy output prediction value of the new energy power generation equipment through a probability density function of the prediction result, and obtaining the new energy deviation component through the new energy prediction deviation and the new energy output prediction value.

[0009] According to the probability density-based power grid frequency modulation method provided by the present invention, step S3 further includes: S31: Calculating the net load power and the net load power standard deviation of the target power grid according to the new energy output forecast value, the power grid load forecast value, the planned output value for this period, and the planned exchange value for this period; S32: Obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the fully connected layer and the activation function layer of the deep neural network to obtain the deviation prediction result; S33: constructing a likelihood function of the deviation prediction result, and taking the logarithm of the likelihood function to obtain the loss function of the deviation prediction result.

[0010] According to the probability density-based power grid frequency regulation method provided by the present invention, when iterating the loss function, an optimizer is obtained, and the parameters of the deep neural network are adjusted by the optimizer. A plurality of the deviation prediction results are obtained by the deep neural network, and the loss function of each of the deviation prediction results is calculated. The loss function with the minimum loss is found in the loss function of each of the deviation prediction results, and is used as the target loss function.

[0011] According to the probability density-based power grid frequency modulation method provided by the present invention, in step S5, when the evaluation parameter meets the evaluation parameter screening condition, the target prediction value is used as the prediction result; When the evaluation parameter does not meet the evaluation parameter screening condition, the confidence interval of the prediction result is changed and the target prediction value is recalculated until the target prediction value meets the evaluation parameter screening condition, and the target prediction value at this time is used as the prediction result.

[0012] The present invention also provides a power grid frequency regulation system based on probability density, comprising: Deviation component module: used for acquiring the operation data of the target power grid, and calculating the system load deviation component, the tie line deviation component and the unit deviation component of the target power grid through the operation data; New energy deviation component module: used to obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model, obtain a prediction result, construct a probability density function of the prediction result and obtain the new energy prediction deviation of the new energy power generation equipment, and calculate the new energy deviation component through the new energy prediction deviation; Loss function module: used to calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; Target prediction value module: used for iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; The power grid frequency modulation module is used to establish evaluation parameter screening conditions, calculate the evaluation parameters of the target prediction value, screen the target prediction value through the evaluation parameters and the evaluation parameter screening conditions, obtain the prediction result, and modulate the frequency of the target power grid through the prediction result.

[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of any of the probability density-based grid frequency modulation methods described above are implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the probability density-based power grid frequency regulation methods described above.

[0015] The present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the steps of any of the probability density-based power grid frequency regulation methods described above.

[0016] The above one or more technical solutions in the embodiments of the present invention have at least one of the following technical effects: The power grid frequency modulation method, system, equipment, product and medium based on probability density provided by the present invention comprehensively considers the influence of system load, new energy, interconnection line, unit and system net load on AGC frequency modulation capacity, avoids excessive reliance on dispatching operation experience or only considers the influence of single factor of load fluctuation in traditional frequency modulation capacity calculation, and is more adaptable to the frequency modulation capacity demand of new power system in high proportion of new energy environment. In addition, the present invention combines the probability density function with the deep neural network, and evaluates the target prediction value in combination with the first evaluation parameter and the second evaluation parameter to ensure that the final prediction result meets the requirements, thereby providing the power grid with a more accurate prediction result of frequency modulation capacity, and modulating the power grid through the prediction result of frequency modulation capacity.

[0017] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a flow chart of the probability density-based power grid frequency modulation method provided by the present invention.

[0020] Figure 2 It is a structural schematic diagram of the power grid frequency modulation system based on probability density provided by the present invention.

[0021] Figure 3 It is a structural schematic diagram of the power grid frequency modulation device based on probability density provided by the present invention.

[0022] Reference numerals: 100, deviation component module; 200, new energy deviation component module; 300, loss function module; 400, target prediction value module; 500, power grid frequency modulation module; 810, processor; 820, communication interface; 830, memory; 840, communication bus. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme in the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present invention. The following embodiments are used to illustrate the present invention, but cannot be used to limit the scope of the present invention.

[0024] In the description of the embodiments of the present invention, it should be noted that the terms “first”, “second” and “third” are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0025] In the description of the embodiments of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the embodiments of the present invention can be understood according to specific circumstances.

[0026] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0027] Combine the following Figures 1 to 3 Describe embodiments of the present invention, Figure 1 This is a flow chart of the power grid frequency regulation method based on probability density. As shown in the figure, the operating data of the target power grid is first obtained, and then the deviation components of the system load, interconnection line, and unit are calculated. The new energy deviation component is obtained through the new energy output prediction model, and then the loss function is obtained through the deep neural network. The loss function is iterated to obtain the target prediction value, and finally the first evaluation parameter and the second evaluation parameter are calculated to determine whether they meet the evaluation parameter screening conditions. If they meet the conditions, the target result is obtained and the target power grid frequency is modulated. Otherwise, the confidence interval needs to be changed and the new energy deviation component needs to be recalculated.

[0028] The present invention provides a power grid frequency modulation method based on probability density, comprising: S1: acquiring operation data of a target power grid, and calculating a system load deviation component, a tie line deviation component and a unit deviation component of the target power grid through the operation data; Furthermore, the purpose of this stage is to calculate the system load deviation component, tie line deviation component and unit deviation component of the target power grid according to the operation data of the target power grid, so as to provide data for subsequent deep neural network prediction. Among them, step S1 further includes: S11: determining the target power grid, acquiring the operating data of the target power grid, calculating the system load forecast deviation and the system load fluctuation deviation of the target power grid through the operating data, and calculating the system load deviation component through the system load forecast deviation and the system load fluctuation deviation; S12: acquiring a planned exchange value after a time period and a planned exchange value in the current time period of the tie line from the operation data, and obtaining the tie line deviation component through the planned exchange value after a time period and the planned exchange value in the current time period; S13: acquiring a planned output value after a period of time and a planned output value for the current period of time of the generator set from the operation data, and calculating the deviation component of the generator set by the planned output value after a period of time and the planned output value for the current period of time.

[0029] With respect to the above steps, the specific implementation methods in this embodiment are as follows: First, determine the target power grid that needs frequency modulation and obtain the target power grid's operating data. The operating data includes the target power grid's current and past operating data obtained by the target power grid's dispatching system, as well as the dispatching system's prediction of the target power grid's load level, power generation of generators, and exchange volume of tie lines in the future. Then, calculate the target power grid's system load forecast deviation in the i-th period through the operating data. and the system load fluctuation deviation of the target power grid in the i-th period : Wherein, avg() means to find the average value of the values ​​in the brackets, which is divided by the length of the i-th period in this embodiment. represents the planned load value of the power grid in the ith period, represents the predicted value of power grid load in the ith period, represents the maximum value of the grid load in the ith period of the operating data, It represents the minimum value of the grid load in the i-th period in the operating data. The above data are all included in the operating data.

[0030] Then the system load deviation component is calculated by the system load forecast deviation and the system load fluctuation deviation. : .

[0031] Then obtain the i-th tie line from the operation data The post-period planned exchange value for the period and the planned exchange value of the i-th period , the tie line deviation component of the i-th period is obtained by the planned exchange value after the period and the planned exchange value of the current period : Here, since the target power grid is connected to other power grids, other power grids will also exchange energy with the target power grid, and the energy exchange between power grids is carried out through interconnection lines. If the target power grid is an isolated system, the interconnection line deviation component is 0.

[0032] Finally, the i-th generator set is obtained from the operating data. The planned output value after each period and the planned output value of the i-th period , thus calculating the deviation component of the group : .

[0033] S2: Obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model to obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; Furthermore, the purpose of this stage is to obtain a new energy output prediction model and predict the new energy power generation equipment through the new energy output model to obtain a prediction structure and calculate the new energy prediction deviation, and finally calculate the new energy deviation component, wherein step S2 further includes: S21: Determine the prediction time and the new energy power generation equipment in the target power grid, obtain a new energy output prediction model, and the new energy output prediction model predicts the power generation power of the new energy power generation equipment multiple times according to the prediction time to obtain a prediction result; S22: determining a confidence interval of the prediction result, constructing a probability density function of the prediction result through the probability distribution of the prediction result, and obtaining the new energy prediction deviation through the confidence interval and the probability density function; S23: obtaining a new energy output prediction value of the new energy power generation equipment through a probability density function of the prediction result, and obtaining the new energy deviation component through the new energy prediction deviation and the new energy output prediction value.

[0034] With respect to the above steps, the specific implementation methods in this embodiment are as follows: First, determine the prediction time. In this embodiment, the prediction time is the i-th period and the i-th period. time periods. It is also necessary to determine the new energy power generation equipment in the target power grid. In this embodiment, the new energy power generation equipment is solar power generation equipment and wind power generation equipment. Since the power generation capacity of new energy power generation equipment is greatly affected by external factors such as weather, it is necessary to obtain a new energy output prediction model. The new energy output prediction model can obtain data on external factors such as weather, and make multiple predictions on the power generation power of the new energy power generation equipment based on the data of these external factors and the status of the new energy power generation equipment, thereby obtaining a prediction result. Due to the characteristics of the new energy output prediction model, the predicted results output by the new energy output prediction model will present a probability distribution, that is, it can output multiple results and there is a probability of outputting a certain result among these results, so the prediction result also presents a probability distribution.

[0035] Then, the confidence interval of the prediction result is determined. In this embodiment, the confidence interval is selected as 90%. At the same time, the probability density function of the prediction result can be constructed according to the probability distribution of the prediction structure. The new energy prediction deviation can be obtained through the confidence interval and the probability density function. The new energy prediction deviation includes the wind power prediction deviation value of the ith period under the 90% confidence interval. And the photovoltaic prediction deviation value of the i-th period under the 90% confidence interval .

[0036] Then, the result with the highest probability in the probability density function of the prediction results is taken as the new energy output prediction value of the new energy power generation equipment, including the photovoltaic output prediction value of the i-th period , i-th The photovoltaic output forecast value for each period , wind power output forecast value in the i-th period and the i Wind power output forecast value for each period .

[0037] Then, the photovoltaic output forecast value of the i-th period and the i-th period are used to calculate the photovoltaic power output of the i-th period. The photovoltaic output forecast value of each period is used to calculate the photovoltaic fluctuation deviation of the i-th period : Similarly, the wind power output forecast value of the i-th period and the i-th period The wind power output forecast value of each period is used to calculate the wind power fluctuation deviation of the ith period. : Then, the wind power comprehensive forecast deviation of the i-th period is obtained by using the wind power fluctuation deviation and wind power forecast deviation value: The photovoltaic comprehensive prediction deviation of the i-th period is obtained by using the photovoltaic fluctuation deviation and photovoltaic prediction deviation value. : Finally, the photovoltaic comprehensive forecast deviation and wind power comprehensive forecast deviation are combined to obtain the new energy deviation component of the i-th period: : .

[0038] S3: Calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; Furthermore, the purpose of this stage is to predict through a deep neural network, obtain a deviation prediction result, and then construct a likelihood function of the deviation prediction result, and finally obtain a loss function of the deviation prediction result. Among them, step S3 further includes: S31: Calculating the net load power and the net load power standard deviation of the target power grid according to the new energy output forecast value, the power grid load forecast value, the planned output value for this period, and the planned exchange value for this period; S32: Obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the fully connected layer and the activation function layer of the deep neural network to obtain the deviation prediction result; S33: constructing a likelihood function of the deviation prediction result, and taking the logarithm of the likelihood function to obtain the loss function of the deviation prediction result.

[0039] With respect to the above steps, the specific implementation methods in this embodiment are as follows: First, calculate the net load power of the target power grid in the i-th period and the standard deviation of net load power in the i-th period : in, is the predicted value of the new energy output in the i-th period, in this embodiment, it is the sum of the predicted value of the photovoltaic output in the i-th period and the predicted value of the wind power output in the i-th period, and N is the number of periods participating in the prediction.

[0040] Then, the deep neural network is obtained, and the system load deviation component, tie line deviation component, unit deviation component, new energy deviation component, net load power and net load power standard deviation of the i-th period are all input into the fully connected layer and activation function layer of the deep neural network. The deep neural network is used for calculation to obtain the parameter matrix of the Gaussian distribution of the frequency modulation capacity of the target power grid as the deviation prediction result. ,in, is the power grid frequency regulation capacity in the i-th period, is the expected value of the grid frequency regulation capacity, is the standard deviation of the grid frequency regulation capacity, and f() represents the parameter matrix of Gaussian distribution.

[0041] Then construct the likelihood function of the deviation prediction result : in, represents the maximum likelihood function of the parameters in brackets, () means taking the maximum value of the parameters in the brackets.

[0042] Then take the logarithm of both sides of the likelihood function to get the loss function of the deviation prediction result : Among them, loss() represents the loss function of the parameters in the brackets, and exp() represents the exponential function in the brackets.

[0043] S4: iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; Furthermore, the purpose of this stage is to iterate the loss function to obtain a target loss function, thereby obtaining a target prediction value. When iterating the loss function, an optimizer is obtained, and the parameters of the deep neural network are adjusted by the optimizer, and a plurality of the deviation prediction results are obtained by the deep neural network, and the loss function of each deviation prediction result is calculated, and the loss function with the minimum loss is found in the loss function of each deviation prediction result, and it is used as the target loss function.

[0044] With respect to the above steps, the specific implementation methods in this embodiment are as follows: First, multiple predictions are made through the deep neural network, and an optimizer is obtained at the same time. The optimizer selected in this embodiment is the Adam optimizer. The optimizer can adjust the internal parameters of the deep neural network according to the output of the deep neural network during the deep neural network prediction process, so as to quickly output the minimum loss function. After obtaining multiple deviation prediction results, the loss function of each deviation prediction result is calculated, and the loss function with the minimum loss is used as the target loss function. Here, the minimum loss refers to the minimum value of the loss function. The target prediction value of the i-th time period can be obtained from the target loss function. S5: Establishing evaluation parameter screening conditions, calculating the evaluation parameters of the target prediction value, screening the target prediction value by the evaluation parameters and the evaluation parameter screening conditions, obtaining a prediction result, and frequency-modulating the target power grid by the prediction result.

[0045] Furthermore, the purpose of this stage is to calculate the evaluation parameters, and determine whether the target prediction value meets the demand through the evaluation parameter screening conditions, take the target prediction value that meets the demand as the prediction result, and adjust the frequency of the target power grid through the prediction result.

[0046] Wherein, when the evaluation parameter meets the evaluation parameter screening condition, the target prediction value is used as the prediction result; When the evaluation parameter does not meet the evaluation parameter screening condition, the confidence interval of the prediction result is changed and the target prediction value is recalculated until the target prediction value meets the evaluation parameter screening condition, and the target prediction value at this time is used as the prediction result.

[0047] With respect to the above steps, the specific implementation methods in this embodiment are as follows: First, determine the evaluation parameter screening condition. In this embodiment, the evaluation parameter screening condition is that the evaluation parameter needs to be higher than the evaluation parameter threshold determined based on experience. Next, the evaluation parameters include the first evaluation parameter and the second evaluation parameter. The first evaluation parameter after the target prediction value of the i-th period is calculated and applied to the target power grid is : in, is the average value of the grid frequency deviation in the i-th period, K is the regional frequency response coefficient of the target grid, is the regional frequency RMS control target value of the target power grid. In this embodiment, the length of each time period is 1 minute.

[0048] In addition, the target prediction value also needs to meet the needs of the target power grid in a longer period of time, so it is also necessary to calculate the second evaluation parameter with a time length of multiple periods. Assuming that the length of each period is t, in this embodiment, the test time m=Ft, and the target prediction value at this time is used as the prediction result, and the target prediction value is evaluated in F periods. How many periods can make the fluctuation of the target power grid within the preset range, and the second evaluation parameter after the target prediction value of the i-th period is applied to the target power grid is calculated. : Among them, G is the number of qualified time periods, indicating how many time periods among the F time periods can make the fluctuation of the target power grid within the preset range.

[0049] When the evaluation parameters, i.e., the first evaluation parameter and the second evaluation parameter, meet the evaluation parameter screening conditions, the target prediction value is used as the prediction result; when the evaluation parameters, i.e., the first evaluation parameter and the second evaluation parameter, do not meet the evaluation parameter screening conditions, it is necessary to change the confidence interval of the prediction result in S22 and recalculate the target prediction value until the target prediction value meets the evaluation parameter screening conditions, and the target prediction value at this time is used as the prediction result. The prediction result is the target power grid frequency regulation capacity of the i-th time period, and the demand for frequency regulation capacity is determined according to the target power grid frequency regulation capacity and the target power grid is frequency regulated.

[0050] The method provided by the present invention realizes accurate prediction of the frequency regulation capacity of the power grid through probability density function and confidence interval, thereby improving the accuracy and efficiency of frequency regulation of the power grid.

[0051] The power grid frequency regulation device based on probability density provided by the present invention is described below. The power grid frequency regulation device based on probability density described below and the power grid frequency regulation method based on probability density described above can be referred to each other.

[0052] Figure 2 The structural diagram of the power grid frequency regulation system based on probability density is shown as an example. Figure 2 As shown, the method for executing the power grid frequency regulation method based on probability density as described above includes: Deviation component module 100: used to obtain operation data of the target power grid, and calculate the system load deviation component, tie line deviation component and unit deviation component of the target power grid through the operation data; A new energy deviation component module 200 is used to obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model, obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; The loss function module 300 is used to calculate the net load power and the standard deviation of the net load power of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the standard deviation of the net load power into the deep neural network, obtain a deviation prediction result, and calculate a loss function of the deviation prediction result; Target prediction value module 400: used to iterate the loss function to obtain a target loss function, and extract a target prediction value from the target loss function; The power grid frequency modulation module 500 is used to establish evaluation parameter screening conditions, calculate the evaluation parameters of the target prediction value, screen the target prediction value through the evaluation parameters and the evaluation parameter screening conditions, obtain the prediction result, and modulate the frequency of the target power grid through the prediction result.

[0053] on the other hand, Figure 3 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 3 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to execute the power grid frequency modulation method based on probability density, and the method includes: S1: acquiring operation data of a target power grid, and calculating a system load deviation component, a tie line deviation component and a unit deviation component of the target power grid through the operation data; S2: Obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model to obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; S3: Calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; S4: iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; S5: Establishing evaluation parameter screening conditions, calculating the evaluation parameters of the target prediction value, screening the target prediction value by the evaluation parameters and the evaluation parameter screening conditions, obtaining a prediction result, and frequency-modulating the target power grid by the prediction result.

[0054] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0055] On the other hand, the present invention further provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, when the program instructions are executed by a computer, the computer can execute the power grid frequency modulation method based on probability density provided by the above methods, the method comprising: S1: acquiring operation data of a target power grid, and calculating a system load deviation component, a tie line deviation component and a unit deviation component of the target power grid through the operation data; S2: Obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model to obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; S3: Calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; S4: iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; S5: Establishing evaluation parameter screening conditions, calculating the evaluation parameters of the target prediction value, screening the target prediction value by the evaluation parameters and the evaluation parameter screening conditions, obtaining a prediction result, and frequency-modulating the target power grid by the prediction result.

[0056] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for power grid frequency regulation based on probability density provided by the above methods is implemented, and the method comprises: S1: acquiring operation data of a target power grid, and calculating a system load deviation component, a tie line deviation component and a unit deviation component of the target power grid through the operation data; S2: Obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model to obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; S3: Calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; S4: iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; S5: Establishing evaluation parameter screening conditions, calculating the evaluation parameters of the target prediction value, screening the target prediction value by the evaluation parameters and the evaluation parameter screening conditions, obtaining a prediction result, and frequency-modulating the target power grid by the prediction result.

[0057] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0058] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A power grid frequency modulation method based on probability density, characterized in that: include: S1: acquiring operation data of a target power grid, and calculating a system load deviation component, a tie line deviation component and a unit deviation component of the target power grid through the operation data; S2: Obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model to obtain a prediction result, construct a probability density function of the prediction result and obtain a new energy prediction deviation of the new energy power generation equipment, and calculate a new energy deviation component through the new energy prediction deviation; S3: Calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; S4: iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; S5: Establishing evaluation parameter screening conditions, calculating the evaluation parameters of the target prediction value, screening the target prediction value by the evaluation parameters and the evaluation parameter screening conditions, obtaining a prediction result, and frequency-modulating the target power grid by the prediction result.

2. The method for power grid frequency modulation based on probability density according to claim 1, characterized in that: Step S1 further comprises: S11: determining the target power grid, acquiring the operating data of the target power grid, calculating the system load forecast deviation and the system load fluctuation deviation of the target power grid through the operating data, and calculating the system load deviation component through the system load forecast deviation and the system load fluctuation deviation; S12: acquiring a planned exchange value after a time period and a planned exchange value in the current time period of the tie line from the operation data, and obtaining the tie line deviation component through the planned exchange value after a time period and the planned exchange value in the current time period; S13: acquiring a planned output value after a period of time and a planned output value for the current period of time of the generator set from the operation data, and calculating the deviation component of the generator set by the planned output value after a period of time and the planned output value for the current period of time.

3. The method for power grid frequency modulation based on probability density according to claim 1, characterized in that: Step S2 further comprises: S21: Determine the prediction time and the new energy power generation equipment in the target power grid, obtain a new energy output prediction model, and the new energy output prediction model predicts the power generation power of the new energy power generation equipment multiple times according to the prediction time to obtain a prediction result; S22: determining a confidence interval of the prediction result, constructing a probability density function of the prediction result through the probability distribution of the prediction result, and obtaining the new energy prediction deviation through the confidence interval and the probability density function; S23: obtaining a new energy output prediction value of the new energy power generation equipment through a probability density function of the prediction result, and obtaining the new energy deviation component through the new energy prediction deviation and the new energy output prediction value.

4. The method for power grid frequency modulation based on probability density according to claim 1, characterized in that: Step S3 further comprises: S31: Calculating the net load power and the net load power standard deviation of the target power grid according to the new energy output forecast value, the power grid load forecast value, the planned output value for this period, and the planned exchange value for this period; S32: Obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the fully connected layer and the activation function layer of the deep neural network to obtain the deviation prediction result; S33: constructing a likelihood function of the deviation prediction result, and taking the logarithm of the likelihood function to obtain the loss function of the deviation prediction result.

5. The method for power grid frequency modulation based on probability density according to claim 1, characterized in that: In step S4, when iterating the loss function, an optimizer is obtained, and the parameters of the deep neural network are adjusted by the optimizer. A plurality of the deviation prediction results are obtained by the deep neural network, and the loss function of each of the deviation prediction results is calculated. The loss function with the minimum loss is found in the loss function of each of the deviation prediction results, and is used as the target loss function.

6. The method for power grid frequency modulation based on probability density according to claim 1, characterized in that: In step S5, when the evaluation parameter meets the evaluation parameter screening condition, the target prediction value is used as the prediction result; When the evaluation parameter does not meet the evaluation parameter screening condition, the confidence interval of the prediction result is changed and the target prediction value is recalculated until the target prediction value meets the evaluation parameter screening condition, and the target prediction value at this time is used as the prediction result.

7. A power grid frequency modulation method system based on probability density, used to execute the power grid frequency modulation method based on probability density according to any one of claims 1 to 6, characterized in that: include: Deviation component module: used for acquiring the operation data of the target power grid, and calculating the system load deviation component, the tie line deviation component and the unit deviation component of the target power grid through the operation data; New energy deviation component module: used to obtain a new energy output prediction model, predict the new energy power generation equipment in the target power grid through the new energy output prediction model, obtain a prediction result, construct a probability density function of the prediction result and obtain the new energy prediction deviation of the new energy power generation equipment, and calculate the new energy deviation component through the new energy prediction deviation; Loss function module: used to calculate the net load power and the net load power standard deviation of the target power grid, obtain a deep neural network, input the system load deviation component, the tie line deviation component, the unit deviation component, the new energy deviation component, the net load power and the net load power standard deviation into the deep neural network, obtain a deviation prediction result, and calculate a loss function for the deviation prediction result; Target prediction value module: used for iterating the loss function to obtain a target loss function, and extracting a target prediction value from the target loss function; The power grid frequency modulation module is used to establish evaluation parameter screening conditions, calculate the evaluation parameters of the target prediction value, screen the target prediction value through the evaluation parameters and the evaluation parameter screening conditions, obtain the prediction result, and modulate the frequency of the target power grid through the prediction result.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the power grid frequency regulation method based on probability density as described in any one of claims 1 to 6 are implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the power grid frequency regulation method based on probability density as described in any one of claims 1 to 6 are implemented.

10. A computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, characterized in that: When the program instructions are executed by a computer, the computer can execute the steps of the probability density-based power grid frequency regulation method as described in any one of claims 1 to 6.

Images