A microgrid intelligent operation optimization method and system

Abstract

The present invention provides a microgrid intelligent operation optimization method and system, which relates to the field of microgrid technology. The method includes: obtaining historical power generation, historical power consumption and historical energy replenishment power, and determining historical power efficiency; training a power consumption prediction model; obtaining external condition parameters at the current moment, and selecting a historical target moment, and then determining the predicted power consumption; according to the predicted power consumption and the trained power consumption prediction model, determining the predicted power generation and the predicted energy replenishment power of the external power grid, and based on this, controlling the power generation equipment and the connection between the microgrid and the external power grid. According to the present invention, the relationship between the power consumption of each power consumption equipment and the power consumption efficiency of the microgrid can be determined, and the historical target moment can be screened based on the external condition parameters to obtain the predicted power consumption, improve the prediction accuracy, and accurately plan the power generation of the power generation equipment and the energy replenishment power of the external power grid to improve the overall power consumption efficiency of the microgrid.

Description

Technical Field

[0001] The present invention relates to the field of microgrid technology, and in particular to a microgrid intelligent operation optimization method and system. Background Art

[0002] In the related art, CN118761519A discloses an optimization method and system for a smart microgrid, which belongs to the field of smart microgrid technology. The method includes: obtaining power parameters that affect the operating efficiency of the smart microgrid; standardizing or normalizing the power parameters and operating efficiency to obtain multiple data points; randomly selecting K data points as initial cluster centers; calculating the distance from each data point to all cluster centers, and assigning the data points to the clusters where the nearest cluster centers are located, recalculating the positions of the cluster centers according to the data points in the clusters, and repeating the steps until the positions of the cluster centers no longer change or reach a preset number of iterations, and obtaining clustering results; analyzing the clustering results, finding the cluster with the best operating efficiency, and extracting the average or median of the power parameters in the cluster as the parameter value corresponding to the best operating efficiency; applying the parameter value corresponding to the best operating efficiency to the smart microgrid. This solution can effectively improve the operating efficiency.

[0003] CN118469154B discloses an intelligent microgrid evaluation system and method based on distributed energy consumption, including collecting energy production, consumption and storage of each node in the microgrid and key data information in the operation of the microgrid through intelligent sensing devices and sensors and preprocessing; classifying according to the scale of the microgrid, defining and calculating the first energy consumption efficiency; obtaining the adjustment parameters of the energy consumption efficiency through the load demand and energy loss conditions; and adjusting the energy cost appropriately; comparing the adjustment parameters of the energy consumption efficiency with the preset adjustment parameters of the energy consumption efficiency to obtain the adjusted energy consumption efficiency; and then establishing a microgrid evaluation model in combination with relevant data information. The scheme realizes real-time evaluation and optimization regulation of the microgrid, and improves the energy consumption efficiency and reliability of the microgrid.

[0004] Therefore, although the relevant technology can improve the operating efficiency of the microgrid by calculating a variety of power data, different power consumption conditions of different power-consuming equipment are different, and different power generation equipment are also different. For example, some power-consuming equipment needs to be transformed, rectified and other processes before they can be used. During these processes, certain losses will be caused. The electric energy generated by some power generation equipment also needs to be transformed and rectified before it can be input into the microgrid for use, which will also cause certain losses. Therefore, under the influence of different power generation equipment and power-consuming equipment, the overall power efficiency of the microgrid is different, and the relevant technology cannot analyze the power efficiency under different circumstances, and thus cannot reasonably plan the power generation power of the power generation equipment and the energy replenishment power of the external power grid based on the power demand of multiple power-consuming equipment and the overall power efficiency.

[0005] The information disclosed in the background technology section of this application is only intended to deepen the understanding of the general background technology of this application, and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to those skilled in the art. Summary of the invention

[0006] The present invention provides a microgrid intelligent operation optimization method and system, which can solve the problem that the related technology cannot analyze the power efficiency under different situations, and thus cannot reasonably plan the power generation power of the power generation equipment and the energy supplement power of the external power grid based on the power demand of various power-consuming equipment and the overall power efficiency.

[0007] According to a first aspect of the present invention, a microgrid intelligent operation optimization method is provided, comprising:

[0008] At multiple historical moments, obtain and record the historical power generation of multiple power generation equipment, the historical power consumption of multiple power consumption equipment, and the historical energy replenishment power of the external power grid;

[0009] Determining historical power efficiency at multiple historical moments according to the historical generated power, the historical power consumption and the historical energy replenishment power;

[0010] According to the historical generated power, the historical power consumption, the historical energy replenishment power and the historical power consumption efficiency, an electric power consumption prediction model is trained to obtain a trained electric power consumption prediction model;

[0011] Acquire the current time and multiple external condition parameters at the current time, wherein the external condition parameters include temperature parameters, wind parameters, rainfall parameters and light parameters;

[0012] Determining a historical target moment corresponding to the current moment from among multiple historical dates according to the current moment and the multiple external condition parameters;

[0013] Determine predicted power consumption of multiple electrical devices at multiple times within a preset time period after a current moment according to first historical power consumption of multiple electrical devices at multiple times within a preset time period after the historical target moment;

[0014] Determine the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model;

[0015] A plurality of power generation devices are controlled according to the predicted power generation power, and when the predicted energy supplement power is not all zero, the microgrid is connected to an external power grid.

[0016] According to a second aspect of the present invention, there is provided a microgrid intelligent operation optimization system, comprising:

[0017] A recording module is used to obtain and record the historical power generation of multiple power generation equipment, the historical power consumption of multiple power consumption equipment, and the historical energy replenishment power of the external power grid at multiple historical moments;

[0018] A historical electricity efficiency module, used to determine the historical electricity efficiency at multiple historical moments according to the historical generated power, the historical electricity power and the historical energy replenishment power;

[0019] A training module, used to train the power consumption prediction model according to the historical generated power, the historical power consumption, the historical energy replenishment power and the historical power consumption efficiency, so as to obtain the trained power consumption prediction model;

[0020] An external condition parameter module is used to obtain the current moment and multiple external condition parameters at the current moment, wherein the external condition parameters include temperature parameters, wind parameters, rainfall parameters and light parameters;

[0021] A historical target moment module, used to determine a historical target moment corresponding to the current moment from a plurality of historical dates according to the current moment and the plurality of external condition parameters;

[0022] A power consumption prediction module, configured to determine the predicted power consumption of multiple power consumption devices at multiple moments in a preset time period after a current moment according to the first historical power consumption of multiple power consumption devices at multiple moments in a preset time period after the historical target moment;

[0023] A prediction module, used to determine the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model;

[0024] The control module is used to control multiple power generation devices according to the predicted power generation power, and connect the microgrid to the external power grid when the predicted energy replenishment power is not all zero.

[0025] Technical effect: According to the present invention, the power consumption prediction model can be trained based on historical data of multiple historical moments, so as to determine the relationship between the power consumption of each power-consuming device and the power efficiency of the microgrid, and provide a data basis for optimizing the power generation power of the power generation equipment and the energy replenishment power of the external power grid. The historical target moments can be screened based on multiple external condition parameters to obtain the predicted power consumption, improve the prediction accuracy, and then accurately plan the power generation power of the power generation equipment and the energy replenishment power of the external power grid based on the predicted power consumption and the trained power consumption prediction model, thereby improving the overall power efficiency of the microgrid. When training the power consumption prediction model, the loss function of the power consumption prediction model can be determined by the gap between the historical training power consumption and the historical power consumption, and the gap between the training power consumption efficiency and the measured historical power consumption efficiency, thereby improving the calculation accuracy of the power consumption of each power-consuming device under a specific power consumption efficiency during the training process, and improving the training effect of the power consumption prediction model. When determining the constraint conditions, the constraint conditions can be used to limit the values ​​of the predicted power generation, the predicted supplementary power and the power efficiency, so that the power efficiency calculated under this value is higher than the power efficiency calculated by predicting the power consumption, and accurate ranges can be set for a variety of pending values ​​to improve the operation efficiency of the optimization process and the power efficiency of the microgrid. The change of the power generation of the power generation equipment can be limited by the constraint conditions to avoid damage to the power generation equipment, and the calculated power, supplementary power and power consumption can be realized in reality through the restriction of the constraint conditions. When determining the objective function, the objective function can be used to optimize the value of the calculated power consumption, thereby optimizing the value of the power efficiency, the predicted power generation and the predicted supplementary power, so that the calculated power consumption can meet the power demand of the people using the microgrid. The objective function can also be used to optimize the value of the pending value of the power efficiency, so that the power efficiency of the microgrid can be maximized on the basis of meeting the power demand, thereby saving electricity and improving the efficiency of power use.

[0026] It should be understood that the above general description and the following detailed description are exemplary and explanatory only and do not limit the present invention. Other features and aspects of the present invention will become more apparent from the following detailed description of exemplary embodiments with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other embodiments can be obtained based on these drawings without creative work.

[0028] Figure 1 A schematic diagram of a process flow of a microgrid intelligent operation optimization method according to an embodiment of the present invention is exemplarily shown;

[0029] Figure 2 A block diagram of a microgrid intelligent operation optimization system according to an embodiment of the present invention is exemplarily shown. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.

[0031] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0032] Figure 1 A schematic flow chart of a microgrid intelligent operation optimization method according to an embodiment of the present invention is exemplarily shown, and the method includes:

[0033] Step S101, at multiple historical moments, obtaining and recording the historical power generation of multiple power generation devices, the historical power consumption of multiple power consumption devices, and the historical energy replenishment power of the external power grid;

[0034] Step S102, determining historical power efficiency at multiple historical moments according to the historical power generation power, the historical power consumption power and the historical energy replenishment power;

[0035] Step S103, training an electric power prediction model according to the historical generated power, the historical electric power, the historical energy replenishment power and the historical electric power efficiency to obtain a trained electric power prediction model;

[0036] Step S104, obtaining the current time and multiple external condition parameters at the current time, wherein the external condition parameters include temperature parameters, wind parameters, rainfall parameters and light parameters;

[0037] Step S105, determining a historical target time corresponding to the current time from a plurality of historical dates according to the current time and the plurality of external condition parameters;

[0038] Step S106, determining predicted power consumption of multiple electrical devices at multiple moments in a preset time period after the current moment according to the first historical power consumption of multiple electrical devices at multiple moments in a preset time period after the historical target moment;

[0039] Step S107, determining the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model;

[0040] Step S108, controlling a plurality of power generation devices according to the predicted power generation, and connecting the microgrid to an external power grid when the predicted energy replenishment power is not all zero.

[0041] According to the microgrid intelligent operation optimization method of the embodiment of the present invention, the power consumption prediction model can be trained based on the historical data of multiple historical moments, so as to determine the relationship between the power consumption of each power-consuming device and the power efficiency of the microgrid, and provide a data basis for optimizing the power generation of the power generation equipment and the energy replenishment power of the external power grid. The historical target moments can be screened based on multiple external condition parameters to obtain the predicted power consumption, improve the prediction accuracy, and then accurately plan the power generation of the power generation equipment and the energy replenishment power of the external power grid based on the predicted power consumption and the trained power consumption prediction model, so as to improve the overall power efficiency of the microgrid.

[0042] According to one embodiment of the present invention, the power consumption prediction model can be trained by using historical data obtained at multiple historical moments to determine the relationship between various power generation powers of different power generation equipment, various power consumption powers of different power consumption equipment, and the power consumption efficiency of the microgrid.

[0043] According to an embodiment of the present invention, in step S101, historical power generation of multiple power generation devices, historical power consumption of multiple power consumption devices, and historical energy replenishment power of the external power grid may be acquired and recorded at multiple historical moments.

[0044] According to one embodiment of the present invention, in step S102, the historical power generation power at the above-mentioned multiple historical moments, the historical power consumption power of multiple power-consuming devices, and the historical energy replenishment power of the external power grid can be read to calculate the historical power consumption efficiency at multiple historical moments.

[0045] According to an embodiment of the present invention, the historical power consumption efficiency at multiple historical moments is determined based on the historical power generation, the historical power consumption and the historical energy replenishment power, including: determining the historical power consumption efficiency at the i-th historical moment according to formula (1): ,

[0046] (1)

[0047] in, is the historical power consumption of the kth power-consuming equipment at the i-th historical moment, is the historical power generation of the j-th power generation equipment at the i-th historical moment, is the historical energy replenishment power at the i-th historical moment, n is the number of power generation equipment, m is the number of power consumption equipment, j≤n, k≤m, and i, j, n, k and m are all positive integers.

[0048] According to one embodiment of the present invention, the historical power generation is the power measured at the output port of the power generation equipment. The output port of the power generation equipment may not be directly connected to the transmission line of the microgrid. It may also need to connect to a transformer, rectifier and other equipment. After certain processing, it can be connected to the transmission line of the microgrid. Therefore, when the power generation equipment provides the generated electric energy to the microgrid, certain losses may occur. The historical power consumption is the power measured at the input port of the power consumption equipment. The input port of the power consumption equipment may not be directly connected to the transmission line of the microgrid. It may also need to connect to a transformer, rectifier and other equipment. After certain processing, it can receive the electric energy of the transmission line. Therefore, when the transmission line provides the electric energy to the power consumption equipment, certain losses may occur. Furthermore, the transmission line of the microgrid itself also has certain losses. In addition, when the external power grid provides energy replenishment power to the microgrid, there may also be certain losses. For example, there may be equipment such as a transformer between the external power grid and the microgrid, which will cause certain losses.

[0049] According to one embodiment of the present invention, based on the above-mentioned losses, the overall power efficiency of the microgrid is less than 100%, and the historical power efficiency at the i-th historical moment can be determined based on formula (1), that is, the ratio of the sum of the historical power consumption of all power-consuming devices to the sum of the provided power, wherein the sum of the provided power includes the sum of the historical power generation power of the power generation equipment and the sum of the historical energy replenishment power of the external power grid.

[0050] According to one embodiment of the present invention, in step S103, the historical power generation, historical power consumption, historical energy replenishment power, and the historical power consumption efficiency determined above can be used to train the power consumption prediction model, so as to determine the relationship between the various power generation of different power generation equipment and the various power consumption of different power consumption equipment and the power consumption efficiency of the microgrid. Some power generation equipment has high losses, while other power generation equipment has low losses. Therefore, when different power generation equipment is used, even if the same total power is generated, the power that can be used is different. Similarly, some power consumption equipment has high losses, while other power consumption equipment has low losses. Therefore, when different power consumption equipment is working, even if the total power consumption of the power consumption equipment is the same, its efficiency is different. Therefore, the above rules can be described by the power consumption prediction model.

[0051] According to one embodiment of the present invention, an electric power prediction model is trained based on the historical generated power, the historical electric power, the historical energy replenishment power and the historical electric power efficiency to obtain the trained electric power prediction model, including: inputting the historical generated power, the historical energy replenishment power and the historical electric power efficiency into the electric power prediction model to obtain the historical training electric power; obtaining the loss function of the electric power prediction model based on the historical training electric power, the historical electric power, the historical generated power, the historical energy replenishment power and the historical electric power efficiency; training the electric power prediction model based on the loss function of the electric power prediction model to obtain the trained electric power prediction model.

[0052] According to one embodiment of the present invention, in the process of training an electric power prediction model so that it can accurately determine the relationship between power generation power, energy replenishment power, electric power and power efficiency, the historical power generation power, historical energy replenishment power and historical power efficiency can be input into the electric power prediction model to obtain the historical training electric power. That is, under the action of a certain power generation power and energy replenishment power, different electric devices use different electric powers to work, and the power efficiency of the microgrid is different. Therefore, under the action of a certain power generation power and energy replenishment power, if the power efficiency of the microgrid is also a specific value, the power consumption of each electric device is also a specific value. Based on this, the electric power prediction model can be trained to determine the ability of the specific value of the electric power of each electric device based on the historical power generation power, historical energy replenishment power and historical power efficiency. The output value of the power consumption prediction model is the historical training power consumption. There is a gap between this specific value of the power consumption (i.e., the measured historical power consumption). The loss function can be determined based on the gap, and the power consumption prediction model can be trained based on the loss function to improve its accuracy. The accuracy of the historical training power consumption calculated by the power consumption prediction model can also be improved, that is, the error between the historical training power consumption and the measured historical power consumption can be reduced.

[0053] According to one embodiment of the present invention, according to the historical training power consumption, the historical power consumption, the historical power generation power, the historical energy replenishment power and the historical power efficiency, the loss function of the power consumption prediction model is obtained, including: according to formula (2), the loss function of the power consumption prediction model is obtained: ,

[0054] (2)

[0055] in, is the historical training power consumption of the kth power-consuming device at the i-th historical moment, N is the number of historical moments in a batch, and is the preset weight, i≤N, and i and N are both positive integers.

[0056] According to one embodiment of the present invention, in formula (2), is the difference between historical training power consumption and historical power consumption, The average value of the gap between the historical training power consumption and the historical power consumption of multiple power-consuming devices can be used as an item in the loss function. During the training process, the average value can be reduced, thereby reducing the gap between the historical training power consumption and the historical power consumption, and improving the accuracy of the specific value of the power consumption calculated by the power prediction model.

[0057] According to one embodiment of the present invention, in formula (2), It is the ratio of the sum of the historical training power consumption of each power-consuming device to the sum of the power provided (that is, the sum of the historical power generation power of each power generation device and the historical power replenishment power of the external power grid). This value is the training power efficiency. It is the gap between the training power efficiency and the measured historical power efficiency, which can be used as a loss function. During the training process, the gap can be narrowed, thereby narrowing the gap between the training power efficiency and the measured historical power efficiency, and further improving the accuracy of the historical training power power.

[0058] According to one embodiment of the present invention, the weighted sum of the above two items can be used as the loss function corresponding to the i-th historical moment, and the loss functions corresponding to multiple historical moments of a training batch can be averaged to obtain the loss function of the power consumption prediction model. The loss function of the power consumption prediction model can be back-propagated, and the parameters of the power consumption prediction model can be adjusted by the gradient descent method. The above training process can be performed multiple times to obtain the trained power consumption prediction model. In this way, the relationship between the generated power, the energy replenishment power, the power consumption and the power consumption efficiency can be accurately determined.

[0059] In this way, when training the power consumption prediction model, the loss function of the power consumption prediction model can be determined by the gap between the historical training power consumption and the historical power consumption, and the gap between the training power efficiency and the measured historical power efficiency. In this way, the calculation accuracy of the power consumption of each power-consuming device under a specific power efficiency is improved during the training process, thereby improving the training effect of the power consumption prediction model.

[0060] According to an embodiment of the present invention, in step S104, multiple external condition parameters at the current moment may be obtained, such as temperature parameters, wind parameters, rainfall parameters, and light parameters. The external condition parameters are related to future electricity demand. For example, when the temperature parameter is low, the heating demand of the people using the microgrid power supply is high, which will lead to an increase in electricity demand. When the light parameter is low, the lighting demand of the people using the microgrid power supply is high, which will lead to an increase in electricity demand.

[0061] According to one embodiment of the present invention, in step S105, the power demand at multiple moments in the preset time period after the current moment can be estimated by searching for a historical target moment similar to multiple external condition parameters at the current moment from multiple moments of the historical date, and then determining the power consumption conditions at multiple moments in the preset time period after the historical target moment. If multiple external condition parameters are similar, the power demand of people using the microgrid for power supply is similar, and therefore, the power consumption conditions in the predicted time period after the historical target moment with similar multiple external condition parameters can be used to estimate the power demand in the preset time period after the current moment.

[0062] According to one embodiment of the present invention, based on the current moment and the multiple external condition parameters, a historical target moment corresponding to the current moment is determined in multiple historical dates, including: determining the moment in multiple historical dates that is the same as the current moment as a candidate moment; obtaining multiple historical external condition parameters for each candidate moment, wherein the historical external condition parameters include historical temperature parameters, historical wind parameters, historical rainfall parameters and historical light parameters; obtaining an external condition parameter vector for the current moment based on the multiple external condition parameters; obtaining a historical external condition parameter vector for each candidate moment based on the multiple historical external condition parameters for each candidate moment; determining the similarity between the external condition parameter vector and the historical external condition parameter vector for each candidate moment; and determining the candidate moment corresponding to the historical external condition parameter vector whose similarity is greater than or equal to a similarity threshold as the historical target moment.

[0063] According to one embodiment of the present invention, first, among multiple historical dates, the same moment as the current moment is selected as a candidate moment, so that the candidate moment is comparable to the current moment. For example, if the current moment is 7:00 in the morning, the lighting parameters in the future preset time period (for example, within one hour) will be higher and higher, the temperature parameters will be higher and higher, and the electricity demand will be lower and lower. If other moments in the historical dates are selected, the electricity demand of the people using the microgrid for power supply may not have the above characteristics. Therefore, the same moment as the current moment in the historical date is determined as a candidate moment, and it is determined whether the multiple historical external condition parameters of the candidate moment are similar to the multiple external condition parameters of the current moment.

[0064] According to one embodiment of the present invention, multiple external condition parameters can be combined into an external condition parameter vector of the current moment, that is, each external condition parameter is used as a component of the external condition parameter vector. Similarly, multiple historical external condition parameters of each candidate moment can be used to form a historical external condition parameter vector of the candidate moment. Then, the similarity between the external condition parameter vector and the historical external condition parameter vectors of each candidate moment can be determined, for example, the cosine similarity can be calculated, and then the candidate moment corresponding to the historical external condition parameter vector whose similarity is greater than or equal to the similarity threshold can be determined as the historical target moment.

[0065] According to one embodiment of the present invention, in step S106, the historical external condition parameters at the historical target moment are relatively similar to the external condition parameters at the current moment. Therefore, the power demand of the people using the microgrid power supply in the preset time period after the historical target moment is similar to the power demand of the people using the microgrid power supply in the preset time period after the current moment. Therefore, the first historical power consumption of multiple power devices at multiple moments in the preset time period after the historical target moment can be used to determine the predicted power consumption of multiple power devices at multiple moments in the preset time period after the current moment. In the example, the average value of the first historical power consumption of the kth power device at the tth moment after multiple historical target moments can be used as the predicted power consumption of the kth power device at the tth moment after the current moment. Further, based on the above calculation method, the predicted power consumption of each power device at multiple moments in the preset time period after the current moment can be determined.

[0066] According to one embodiment of the present invention, in step S107, based on the predicted power consumption and the trained power consumption prediction model, the power generation power of the power generation equipment at multiple moments in the preset time period after the current moment and the energy replenishment power of the external power grid can be optimized to maximize the power efficiency of the microgrid, thereby improving the power utilization rate. The optimal power generation power and energy replenishment power obtained in the optimization process are the predicted power generation power and predicted energy replenishment power. The optimization process can be carried out using an optimization model, such as a nonlinear programming model or a genetic algorithm model, and the present invention does not limit the specific type of the optimization model. Constraints and objective functions can be set based on the predicted power consumption and the trained power consumption prediction model, so as to find the optimal solution that can achieve the goal described by the objective function to the greatest extent under the constraints, and the optimal solution is the predicted power generation power and the predicted energy replenishment power of the external power grid.

[0067] According to one embodiment of the present invention, based on the predicted power consumption and the trained power consumption prediction model, the predicted power generation power of the power generation equipment at multiple moments in a preset time period after the current moment and the predicted energy replenishment power of the external power grid are determined, including: determining the constraints of the power supply optimization model based on the predicted power consumption and the trained power consumption prediction model; determining the objective function of the power supply optimization model based on the trained power consumption prediction model; obtaining the predicted power generation power of the power generation equipment at multiple moments in the preset time period after the current moment and the predicted energy replenishment power of the external power grid based on the constraints and objective function of the power supply optimization model.

[0068] According to an embodiment of the present invention, determining the constraint conditions of the power supply optimization model according to the predicted power consumption and the trained power consumption prediction model includes: determining the constraint conditions of the power supply optimization model according to formula (3), formula (4), formula (5), formula (6), formula (7) and formula (8),

[0069] (3)

[0070] (4)

[0071] (5)

[0072] (6)

[0073] (7)

[0074] (8)

[0075] in, is the mapping function of the trained power consumption prediction model, is the undetermined value of the power efficiency at the tth moment, is the undetermined value of the predicted power generation of the first power generation equipment at the tth moment, is the undetermined value of the predicted power generation of the second power generation equipment at the tth moment, is the undetermined value of the predicted power generation of the j-th power generation equipment at the t-th moment, is the undetermined value of the predicted power generation of the nth power generation equipment at the tth moment, is the undetermined value of the predicted energy replenishment power at the tth moment, is the predicted power efficiency at the tth moment, is the calculated power consumption of the first power-consuming device at the tth moment, is the calculated power consumption of the second power-consuming device at the tth moment, is the calculated power consumption of the kth electrical equipment at the tth moment, is the calculated power consumption of the mth electrical equipment at the tth moment, is the predicted power consumption of the kth power-consuming device at the tth moment, is the undetermined value of the predicted power generation of the j-th power generation equipment at the t+1th time, is the preset scale factor, is the maximum power generation of the jth power generation equipment, is the maximum power supply of the external power grid, The maximum power consumption of the electrical equipment.

[0076] According to one embodiment of the present invention, in formula (3), it is indicated that the undetermined value of the predicted power generation and the undetermined value of the predicted energy supplement power of each power generation device at the tth moment, as well as the undetermined value of the power efficiency, are input into the trained power prediction model, and the calculated power consumption of each power consumption device at the tth moment is calculated by the power prediction model. Furthermore, in order to improve the power efficiency, formula (4) indicates that appropriate undetermined values ​​of the predicted power generation and the predicted energy supplement power are used to make the power efficiency corresponding to the calculated power consumption calculated by the power prediction model (This power efficiency is equal to ) is greater than or equal to the predicted power efficiency corresponding to the predicted power consumption , which can also indicate that the undetermined value of power efficiency should be Select within the range of .

[0077] According to one embodiment of the present invention, the constraint condition described in formula (5) can be used to constrain the relative difference in the power generation of the same power generation equipment between adjacent moments. The relative difference should not be too large, that is, less than or equal to the preset proportional coefficient, so as to protect the power generation equipment and prevent the power generation equipment from changing sharply and damaging the equipment. Formula (6), formula (7) and formula (8) indicate that the to-be-determined value of the power generation equipment is not higher than the maximum power generation, the to-be-determined value of the energy replenishment power is not higher than the maximum energy replenishment power, and the calculated power consumption of the power consumption equipment is not higher than the maximum power consumption, so that the calculation results of the power generation power, energy replenishment power and power consumption can be realized in reality.

[0078] In this way, the values ​​of the predicted power generation, predicted supplementary power and power efficiency can be limited by constraints, so that the power efficiency calculated under this value is higher than the power efficiency calculated by predicting power consumption, and accurate ranges can be set for multiple pending values ​​to improve the calculation efficiency of the optimization process and the power efficiency of the microgrid. The changes in the power generation of power generation equipment can be limited by constraints to avoid damage to the power generation equipment, and the calculated power, supplementary power and power consumption can be realized in reality by limiting the constraints.

[0079] According to an embodiment of the present invention, determining the objective function of the power supply optimization model according to the trained power consumption prediction model includes: determining the objective function of the power supply optimization model according to formulas (9) and (10),

[0080] (9)

[0081] (10)

[0082] Among them, min is the minimization function, max is the maximization function, and M is the number of moments in the preset time period.

[0083] According to one embodiment of the present invention, the objective function described in formula (9) represents minimizing the deviation between the calculated power and the predicted power calculated based on the above-mentioned pending value of power efficiency, the pending value of predicted power generation and the pending value of predicted energy supplement power, so as to meet the power demand of people using microgrid power supply, and provide a target value for the calculated power, so that the calculation of predicted power generation and predicted energy supplement power is more accurate. In addition, formula (9) minimizes the sum of the difference between the calculated power and the predicted power generation of each power device at each moment, that is, minimizes the two in a global sense, so as to achieve the purpose of global optimization by selecting the pending value of power efficiency, the pending value of predicted power generation and the pending value of predicted energy supplement power.

[0084] According to one embodiment of the present invention, the objective function described by formula (10) represents maximizing the average value of the electricity efficiency determined based on the calculated electricity power at each moment (i.e., the pending value of the electricity efficiency) so as to maximize the electricity efficiency in a global sense. The purpose of global optimization can also be achieved by selecting the pending value of the electricity efficiency, the pending value of the predicted generated power, and the pending value of the predicted energy replenishment power.

[0085] According to an embodiment of the present invention, through the above-set constraints and objective functions, an optimization solution process can be performed, so as to find the optimal solution that can achieve the goal described by the objective function to the greatest extent under the constraints of the constraints, that is, the optimal solution of the undetermined value of the predicted power generation and the optimal solution of the undetermined value of the predicted energy replenishment power, so as to obtain the predicted power generation and predicted energy replenishment power. In this way, the power efficiency of the microgrid is maximized on the basis of meeting the power demand of the people using the microgrid for power supply.

[0086] In this way, the value of the calculated power can be optimized through the objective function, thereby optimizing the value of the pending value of power efficiency, the pending value of the predicted power generation power, and the pending value of the predicted energy replenishment power, so that the calculated power can meet the power demand of the people using the microgrid. And the value of the pending value of power efficiency can be optimized through the objective function, so as to maximize the power efficiency of the microgrid on the basis of meeting the power demand, thereby saving electricity and improving the efficiency of power use.

[0087] According to an embodiment of the present invention, in step S108, based on the predicted power generation determined above, multiple power generation devices can be controlled to generate power at multiple times within a preset time period so that their power generation reaches the predicted power generation. And when the predicted energy replenishment power is not all 0, it indicates that an external power grid is required for energy replenishment, so the microgrid can be connected to the external power grid to receive energy replenishment from the external power grid.

[0088] According to the microgrid intelligent operation optimization method of the embodiment of the present invention, the power consumption prediction model can be trained based on historical data of multiple historical moments, so as to determine the relationship between the power consumption of each power-consuming device and the power consumption efficiency of the microgrid, and provide a data basis for optimizing the power generation power of the power generation equipment and the energy replenishment power of the external power grid. The historical target moments can be screened based on multiple external condition parameters to obtain the predicted power consumption, improve the prediction accuracy, and then accurately plan the power generation power of the power generation equipment and the energy replenishment power of the external power grid based on the predicted power consumption and the trained power consumption prediction model, so as to improve the overall power consumption efficiency of the microgrid. When training the power consumption prediction model, the loss function of the power consumption prediction model can be determined by the gap between the historical training power consumption and the historical power consumption, and the gap between the training power consumption efficiency and the measured historical power consumption efficiency, so as to improve the calculation accuracy of the power consumption of each power-consuming device under a specific power consumption efficiency during the training process, and improve the training effect of the power consumption prediction model. When determining the constraint conditions, the constraint conditions can be used to limit the values ​​of the predicted power generation, the predicted supplementary power and the power efficiency, so that the power efficiency calculated under this value is higher than the power efficiency calculated by predicting the power consumption, and accurate ranges can be set for a variety of pending values ​​to improve the operation efficiency of the optimization process and the power efficiency of the microgrid. The change of the power generation of the power generation equipment can be limited by the constraint conditions to avoid damage to the power generation equipment, and the calculated power, supplementary power and power consumption can be realized in reality through the restriction of the constraint conditions. When determining the objective function, the objective function can be used to optimize the value of the calculated power consumption, thereby optimizing the value of the power efficiency, the predicted power generation and the predicted supplementary power, so that the calculated power consumption can meet the power demand of the people using the microgrid. The objective function can also be used to optimize the value of the pending value of the power efficiency, so that the power efficiency of the microgrid can be maximized on the basis of meeting the power demand, thereby saving electricity and improving the efficiency of power use.

[0089] Figure 2 A block diagram of a microgrid intelligent operation optimization system according to an embodiment of the present invention is exemplarily shown, and the system includes:

[0090] A recording module is used to obtain and record the historical power generation of multiple power generation equipment, the historical power consumption of multiple power consumption equipment, and the historical energy replenishment power of the external power grid at multiple historical moments;

[0091] A historical electricity efficiency module, used to determine the historical electricity efficiency at multiple historical moments according to the historical generated power, the historical electricity power and the historical energy replenishment power;

[0092] A training module, used to train the power consumption prediction model according to the historical generated power, the historical power consumption, the historical energy replenishment power and the historical power consumption efficiency, so as to obtain the trained power consumption prediction model;

[0093] An external condition parameter module is used to obtain the current moment and multiple external condition parameters at the current moment, wherein the external condition parameters include temperature parameters, wind parameters, rainfall parameters and light parameters;

[0094] A historical target moment module, used to determine a historical target moment corresponding to the current moment from a plurality of historical dates according to the current moment and the plurality of external condition parameters;

[0095] A power consumption prediction module, configured to determine the predicted power consumption of multiple power consumption devices at multiple moments in a preset time period after a current moment according to the first historical power consumption of multiple power consumption devices at multiple moments in a preset time period after the historical target moment;

[0096] A prediction module, used to determine the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model;

[0097] The control module is used to control multiple power generation devices according to the predicted power generation power, and connect the microgrid to the external power grid when the predicted energy replenishment power is not all zero.

[0098] It should be understood by those skilled in the art that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments, and the embodiments of the present invention may be deformed or modified in any way without departing from the principles.

[0099] 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 replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A microgrid intelligent operation optimization method, characterized in that: include: At multiple historical moments, obtain and record the historical power generation of multiple power generation equipment, the historical power consumption of multiple power consumption equipment, and the historical energy replenishment power of the external power grid; Determining historical power efficiency at multiple historical moments according to the historical generated power, the historical power consumption and the historical energy replenishment power; According to the historical generated power, the historical power consumption, the historical energy replenishment power and the historical power consumption efficiency, an electric power consumption prediction model is trained to obtain a trained electric power consumption prediction model; Acquire the current time and multiple external condition parameters at the current time, wherein the external condition parameters include temperature parameters, wind parameters, rainfall parameters and light parameters; Determining a historical target moment corresponding to the current moment from among multiple historical dates according to the current moment and the multiple external condition parameters; Determine predicted power consumption of multiple electrical devices at multiple times within a preset time period after a current moment according to first historical power consumption of multiple electrical devices at multiple times within a preset time period after the historical target moment; Determine the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model; A plurality of power generation devices are controlled according to the predicted power generation power, and when the predicted energy supplement power is not all zero, the microgrid is connected to an external power grid.

2. The microgrid intelligent operation optimization method according to claim 1, characterized in that: Determining historical power efficiency at multiple historical moments according to the historical generated power, the historical power consumption and the historical energy replenishment power includes: According to the formula Determine the historical electricity efficiency at the i-th historical moment ,in, is the historical power consumption of the kth power-consuming equipment at the i-th historical moment, is the historical power generation of the j-th power generation equipment at the i-th historical moment, is the historical energy replenishment power at the i-th historical moment, n is the number of power generation equipment, m is the number of power consumption equipment, j≤n, k≤m, and i, j, n, k and m are all positive integers.

3. The microgrid intelligent operation optimization method according to claim 2, characterized in that: According to the historical generated power, the historical power consumption, the historical energy replenishment power and the historical power consumption efficiency, an electric power consumption prediction model is trained to obtain a trained electric power consumption prediction model, including: Inputting the historical generated power, the historical energy replenishment power and the historical power efficiency into a power prediction model to obtain historical training power; Obtaining a loss function of an electric power prediction model according to the historical training electric power, the historical electric power, the historical power generation power, the historical energy replenishment power and the historical electric power efficiency; The electric power prediction model is trained according to the loss function of the electric power prediction model to obtain a trained electric power prediction model.

4. The microgrid intelligent operation optimization method according to claim 3 is characterized in that: According to the historical training power consumption, the historical power consumption, the historical power generation power, the historical energy replenishment power and the historical power consumption efficiency, a loss function of the power consumption prediction model is obtained, including: According to the formula Obtain the loss function of the power consumption prediction model ,in, is the historical training power consumption of the kth power-consuming device at the i-th historical moment, N is the number of historical moments in a batch, and is the preset weight, i≤N, and i and N are both positive integers.

5. The microgrid intelligent operation optimization method according to claim 1, characterized in that: According to the current moment and the plurality of external condition parameters, determining a historical target moment corresponding to the current moment from a plurality of historical dates includes: Determine the time in multiple historical dates that is the same as the current time as the candidate time; Acquire multiple historical external condition parameters at each candidate moment, wherein the historical external condition parameters include historical temperature parameters, historical wind parameters, historical rainfall parameters, and historical light parameters; According to the plurality of external condition parameters, obtaining an external condition parameter vector at the current moment; According to multiple historical external condition parameters at each candidate moment, a historical external condition parameter vector of each candidate moment is obtained; Determining the similarity between the external condition parameter vector and the historical external condition parameter vectors at each candidate moment; The candidate moment corresponding to the historical external condition parameter vector whose similarity is greater than or equal to the similarity threshold is determined as the historical target moment.

6. The microgrid intelligent operation optimization method according to claim 1, characterized in that: Determining predicted power generation power of the power generation equipment and predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model, including: Determining the constraint conditions of the power supply optimization model according to the predicted power consumption and the trained power consumption prediction model; Determining the objective function of the power supply optimization model according to the trained power consumption prediction model; According to the constraints and objective function of the power supply optimization model, the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment are obtained.

7. The microgrid intelligent operation optimization method according to claim 6, characterized in that: Determining the constraint conditions of the power supply optimization model according to the predicted power consumption and the trained power consumption prediction model, including: According to the formula Determine the constraints of the power supply optimization model, where: is the mapping function of the trained power consumption prediction model, is the undetermined value of the power efficiency at the tth moment, is the undetermined value of the predicted power generation of the first power generation equipment at the tth moment, is the undetermined value of the predicted power generation of the second power generation equipment at the tth moment, is the undetermined value of the predicted power generation of the j-th power generation equipment at the t-th moment, is the undetermined value of the predicted power generation of the nth power generation equipment at the tth moment, is the undetermined value of the predicted energy replenishment power at the tth moment, is the predicted power efficiency at the tth moment, is the calculated power consumption of the first power-consuming device at the tth moment, is the calculated power consumption of the second power-consuming device at the tth moment, is the calculated power consumption of the kth electrical equipment at the tth moment, is the calculated power consumption of the mth electrical equipment at the tth moment, is the predicted power consumption of the kth power-consuming device at the tth moment, is the undetermined value of the predicted power generation of the j-th power generation equipment at the t+1th time, is the preset scale factor, is the maximum power generation of the jth power generation equipment, is the maximum power supply of the external power grid, The maximum power consumption of the electrical equipment.

8. The microgrid intelligent operation optimization method according to claim 7, characterized in that: According to the trained power consumption prediction model, determining the objective function of the power supply optimization model includes: According to the formula Determine the objective function of the power supply optimization model, where min is the minimization function, max is the maximization function, and M is the number of moments in a preset time period.

9. A microgrid intelligent operation optimization system, characterized in that: include: A recording module is used to obtain and record the historical power generation of multiple power generation equipment, the historical power consumption of multiple power consumption equipment, and the historical energy replenishment power of the external power grid at multiple historical moments; A historical electricity efficiency module, used to determine the historical electricity efficiency at multiple historical moments according to the historical generated power, the historical electricity power and the historical energy replenishment power; A training module, used to train the power consumption prediction model according to the historical generated power, the historical power consumption, the historical energy replenishment power and the historical power consumption efficiency, so as to obtain the trained power consumption prediction model; An external condition parameter module is used to obtain the current moment and multiple external condition parameters at the current moment, wherein the external condition parameters include temperature parameters, wind parameters, rainfall parameters and light parameters; A historical target moment module, used to determine a historical target moment corresponding to the current moment from a plurality of historical dates according to the current moment and the plurality of external condition parameters; A power consumption prediction module, configured to determine the predicted power consumption of multiple power consumption devices at multiple moments in a preset time period after a current moment according to the first historical power consumption of multiple power consumption devices at multiple moments in a preset time period after the historical target moment; A prediction module, used to determine the predicted power generation power of the power generation equipment and the predicted energy replenishment power of the external power grid at multiple moments within a preset time period after the current moment according to the predicted power consumption and the trained power consumption prediction model; The control module is used to control multiple power generation devices according to the predicted power generation power, and connect the microgrid to the external power grid when the predicted energy replenishment power is not all zero.

Images