A distribution network mobile energy storage scheduling method, device and storage medium based on rolling forward warning

Abstract

The present invention provides a distribution network mobile energy storage scheduling method, device and storage medium based on rolling forward warning, which belongs to the technical field of power system operation and scheduling. The method comprises: inputting historical operation data of a selected date into a linearized network flow model of the distribution network to obtain a distribution network node impedance matrix; performing a rolling forward forward prediction of each distribution transformer step-down flow, distributed power output prediction value, node load prediction value and balancing node voltage based on the historical operation data; and obtaining a conjugate complex matrix of the total injected power of each node; inputting the distribution network node impedance matrix, the balancing node voltage prediction value and the conjugate complex matrix of the total injected power of each node into the linearized network flow model of the distribution network to predict the node voltage phasor; and predicting the line load rate and distribution transformer load rate between nodes, thereby obtaining an optimized scheduling plan for mobile energy storage in the distribution network, realizing a forward-looking layout of mobile energy storage, and rationally using mobile energy storage system resources to improve the reliability of the distribution network.

Description

Technical Field

[0001] The present invention relates to a distribution network mobile energy storage dispatching method, equipment and storage medium based on rolling forward warning, belonging to the technical field of power system operation dispatching. Background Art

[0002] Against the backdrop of new power system construction, the network structure of distribution networks is becoming increasingly complex, the number of distributed power sources (wind power, photovoltaics, etc.) in distribution networks is increasing, and the requirements for distribution network power supply reliability in scenarios such as emergency supply, peak loads, and rural power grid upgrades are also constantly improving. Mobile energy storage technology is also being used more and more.

[0003] With the gradual enrichment of load, power supply and other elements in the distribution network, the load and distributed power output directly affect the flow and voltage level in the distribution network. At the same time, the load and distributed power output fluctuate greatly, and the difficulty of predicting the operating status of the distribution network and prejudging risks gradually increases.

[0004] Mobile energy storage technology can provide users with safe, reliable, and convenient power supply solutions with strong flexibility. However, the relatively high investment cost of mobile energy storage systems limits their large-scale application. For scenarios such as emergency power supply, how to rationally utilize mobile energy storage system resources to fully leverage their role in improving distribution network reliability, and how to optimize the layout of mobile energy storage, combined with forward-looking predictions of distribution network operating conditions, to reduce risks such as line flow over-limits, excessive transformer load factors, and system voltage over-limits in the distribution network through optimized mobile energy storage layout, is a question worthy of research. Summary of the Invention

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for dispatching mobile energy storage in a distribution network based on rolling forward warning. The method uses a rolling forward prediction model to make a forward-looking prediction of the operating status of the distribution network, and then combines the mobile energy storage warning optimization model to achieve a reasonable configuration of mobile energy storage.

[0006] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0007] In a first aspect, the present invention provides a method for dispatching mobile energy storage in a distribution network based on rolling forward warning, comprising:

[0008] Obtain the actual operation data of the distribution network within a preset historical period, collect the current operation data in real time, and obtain the historical actual operation data set based on all the operation data;

[0009] One day's actual operation data is selected from the historical actual operation data set, and input into a pre-built distribution network linearized network power flow model to calculate and obtain a distribution network node impedance matrix;

[0010] Input the historical actual operation data within a preset day a before the current date into the pre-trained rolling forward prediction model, and make rolling predictions on the voltage step-down flow of each distribution transformer, the output forecast value of the distributed generation, the node load forecast value, and the balanced node voltage within a preset day b after the current date;

[0011] According to the predicted values ​​of the step-down power flows of each distribution transformer, the predicted values ​​of the node loads, and the predicted values ​​of the outputs of the distributed power sources, a conjugate complex matrix of the total injected power of each node in the distribution network within a preset b days after the current date is obtained, wherein the total injected power of the load node in the distribution network is the same as the predicted value of the node load of the load node, and the total injected power of the power node in the distribution network is the same as the sum of the predicted values ​​of the outputs of the distributed power sources connected to the power node;

[0012] Input the distribution network node impedance matrix, the balanced node voltage prediction value and the total injected power conjugate complex matrix of each node into the distribution network linearized network power flow model to calculate the node voltage phasor within b days after the current date;

[0013] According to the node voltage phasor, obtain the line load rate and distribution transformer load rate between nodes in the distribution network within b days after the current date;

[0014] According to the line load rate, distribution transformer load rate and node voltage phasor between nodes in the distribution network, an optimized scheduling scheme of mobile energy storage in the distribution network is obtained.

[0015] Furthermore, obtaining the historical actual operation data set based on all the operation data includes: preprocessing all the operation data; the preprocessing includes: removing abnormal data, filling in missing data and normalizing.

[0016] Furthermore, the linearized network power flow model of the distribution network is:

[0017] ;

[0018] Where, is the node voltage phasor, , is the number of nodes, The balancing nodes connected to the upper power grid are not included; is the unit phasor, and all elements in the phasor are 1; is the distribution network node impedance matrix, which represents the impedance between connected nodes; is the conjugate complex matrix of the total injected power of each node; , the conjugate complex number is: ; 、 are the voltage amplitude and phase angle of the equilibrium node respectively.

[0019] Furthermore, the distribution network node impedance matrix is ​​calculated, including: collecting actual operation data of the selected date at a preset collection frequency to obtain historical data of c time nodes; inputting the historical data of the c time nodes into the distribution network linearized network flow model respectively to calculate c node impedance matrices; and averaging the node impedance matrices of the c time nodes to obtain the distribution network node impedance matrix Z.

[0020] Furthermore, the construction of the rolling forward prediction model includes:

[0021] Establish BP neural network model and select neural network structure;

[0022] Determine the number and connection mode of neurons in the input layer, hidden layer, and output layer, as well as the learning rate, number of iterations, loss function, and activation function; the input layer parameters include: the step-down power flow of each distribution transformer in the distribution network, the actual output of the distributed power source, and the amplitude and phase angle of all node voltages;

[0023] The output layer parameters include: the step-down flow of each distribution transformer in the distribution network, the output forecast value of the distributed power source, the node load forecast value and the balanced node voltage;

[0024] The training of the rolling forward prediction model includes:

[0025] The actual operation data of the distribution network within a preset historical period is divided into a training set and a validation set;

[0026] Calculate the output of the rolling forward forecast model using the data in the training set and calculate the forecast error based on the actual value;

[0027] According to the prediction error, adjust the weight and bias parameters of the rolling forward prediction model until the model converges;

[0028] Use the data in the validation set to evaluate the trained rolling forward forecast model and perform parameter optimization.

[0029] Furthermore, the method of obtaining an optimized dispatching scheme for mobile energy storage in the distribution network based on the line load rate, distribution transformer load rate, and node voltage phasor between nodes in the distribution network includes: issuing an early warning on the distribution network operation status based on the line load rate, distribution transformer load rate, or node voltage phasor between nodes in the distribution network and the corresponding preset limits; wherein the early warning levels include: red warning, yellow warning, and green warning;

[0030] Based on the principle of highest priority configuration in red warning areas and second priority configuration in yellow warning areas, an optimized dispatching plan for mobile energy storage in the distribution network is generated according to the warning results of the distribution network operation status.

[0031] Furthermore, an early warning of the distribution network operation status is provided based on the line load rate, distribution transformer load rate or node voltage phasor between nodes in the distribution network and the corresponding preset limits, including:

[0032] According to the preset upper limit of distribution transformer load rate and lower limit Early warning of distribution transformer load rate, including: responding to the distribution transformer load rate being less than , generates a green warning of the distribution transformer load rate, in response to the distribution transformer load rate being greater than or equal to and less than or equal to , generates a yellow warning of the distribution transformer load rate; in response to the distribution transformer load rate being greater than , generating a red warning for the distribution transformer load rate.

[0033] According to the preset line load rate upper limit and lower limit Early warning of line load rate, including: responding to line load rate less than , generates a green warning of line load rate, in response to the line load rate being greater than or equal to and less than or equal to , generates a yellow warning of line load rate; in response to the line load rate being greater than , generating a red alert for line load rate.

[0034] Calculate the node voltage deviation ratio based on the node voltage phasor and the preset node voltage deviation ratio first upper limit value , first lower limit , the second upper limit and the second lower limit Early warning of voltage amplitude, including:

[0035] Calculate the node voltage deviation ratio using the formula:

[0036] ;

[0037] Where, is the magnitude of the node voltage, is the rated voltage corresponding to the distribution network.

[0038] Response to node voltage deviation ratio or , generates a voltage amplitude red warning, in response to the node voltage deviation ratio or , generates a yellow warning of voltage amplitude, in response to the node voltage deviation ratio , generating a green warning of voltage amplitude.

[0039] In a second aspect, the present invention provides an electronic device comprising: a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the method for dispatching mobile energy storage in a distribution network based on rolling forward warning as described in the first aspect is implemented.

[0040] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for dispatching mobile energy storage in a distribution network based on rolling forward warning as described in the first aspect is implemented.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] (1) The distribution network mobile energy storage scheduling method provided by the present invention is based on the historical operation data of the distribution network. It extracts the operation characteristics of the distribution network through a rolling forward prediction model established based on the BP neural network, and uses this to make a real-time rolling prediction of the future operation status of the distribution network. According to the real-time prediction results of the rolling forward prediction model, an optimized scheduling plan for mobile energy storage in the distribution network is generated, thereby realizing the forward-looking layout of mobile energy storage and rationally using the resources of the mobile energy storage system to improve the reliability of the distribution network.

[0043] (2) The distribution network mobile energy storage scheduling method provided by the present invention monitors the distribution network according to the line load rate, distribution transformer load rate and node voltage phasor between nodes in the distribution network, and issues an early warning for the line load rate, distribution transformer load rate or node voltage phasor in the distribution network through a mobile energy storage early warning optimization model. According to the early warning results, the mobile energy storage is first configured to the red early warning area, then to the yellow early warning area, and finally the remaining mobile energy storage is configured to the green early warning area, so as to realize the monitoring of the operating status of the distribution network and reduce the risks of line current exceeding the limit, distribution transformer load rate being too high and system voltage exceeding the limit in the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Flowchart of the method for dispatching mobile energy storage in a distribution network based on rolling forward warning provided in Example 1 of the present invention;

[0045] Figure 2 A flow chart showing the classification of distribution transformer load rate warning levels provided in Example 2 of the present invention;

[0046] Figure 3 A flow chart showing the classification of line flow warning levels provided in Example 2 of the present invention;

[0047] Figure 4 This is a flow chart of voltage amplitude warning level classification provided by Example 2 of the present invention. DETAILED DESCRIPTION

[0048] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the embodiments described are only part of the embodiments of the present application, not all of the embodiments.

[0049] Example 1

[0050] This embodiment provides a method for dispatching mobile energy storage in a distribution network based on rolling forward warning. Figure 1 As shown, the method includes the following steps:

[0051] Step 1: Obtain the actual operation data of the distribution network within a preset historical period, collect the operation data at the current moment in real time, and obtain the historical actual operation data set based on all the operation data.

[0052] Specifically, the current time node is defined as t, and the historical voltage amplitude, phase angle, distribution transformer step-down flow and distributed power output data of the distribution network within (t-30) days are obtained from the dispatching system at an interval of 15 minutes. The abnormal data are first eliminated, and then the interpolation method is used to fill the missing data. Finally, the data is normalized to obtain the historical actual operation data set of the distribution network for study.

[0053] It should be noted that the historical actual operation data set is a data set that is updated in real time.

[0054] Step 2: Select one day's actual operation data from the historical actual operation data set, input it into the pre-built distribution network linearized network power flow model, and calculate the distribution network node impedance matrix.

[0055] Specifically, based on the historical actual operation data of the distribution network on the (t-1)th day, the actual operation data of the (t-1)th day is collected at a preset collection frequency to obtain historical data of c time nodes; the node voltage phasor of the distribution network at each time node, the node voltage amplitude and phase angle of the balance node, and the conjugate complex number of the apparent power flowing into each node are input into the distribution network linearized network flow model to calculate c node impedance matrices; the node impedance matrices of the c time nodes are averaged to obtain the node impedance matrix of the distribution network system.

[0056] The linearized network power flow model of the distribution network is:

[0057] ;

[0058] Where, is the node voltage phasor, , is the number of nodes, The balancing nodes connected to the upper power grid are not included; is the unit phasor, and all elements in the phasor are 1; is the distribution network node impedance matrix, which represents the impedance between connected nodes; is the conjugate complex matrix of the total injected power of each node; , the conjugate complex number is: ; 、 are the voltage amplitude and phase angle of the equilibrium node respectively.

[0059] In this embodiment, 96 sets of data can be collected for 24 hours a day at a time interval of 15 minutes. 96 node impedance matrices are calculated based on the historical data of these 96 time nodes, and the distribution network node impedance matrix Z can be obtained by averaging them.

[0060] Step 3: Input the historical actual operation data within 30 days before the current date into the pre-trained rolling forward prediction model, and make rolling predictions for each distribution transformer step-down flow, distributed generation output forecast value, node load forecast value, and balance node voltage within the preset 3 days after the current date.

[0061] Specifically, the rolling forward prediction model is used to predict future operating data based on the distribution transformer step-down flow, node voltage amplitude, phase angle (including nodes connected to the main grid) and actual output of distributed power sources in historical actual operating data, including: distribution transformer step-down flow, voltage amplitude and phase angle of balancing nodes, node load forecast value and distributed power source output forecast value.

[0062] Step 4: According to the predicted values ​​of the step-down power flows of the distribution transformers, the predicted values ​​of the node loads and the predicted values ​​of the distributed generation outputs, a conjugate complex matrix of the total injected power of each node in the distribution network within a preset b days after the current date is obtained.

[0063] Specifically, the nodes in the distribution network include: load nodes and power supply nodes. The total injected power of the load node is the same as the node load forecast value of the load node, and the total injected power of the power supply node is the same as the sum of the output forecast values ​​of the distributed power sources connected to the power supply node.

[0064] Step 5: Input the distribution network node impedance matrix, the balanced node voltage prediction value and the total injected power conjugate complex matrix of each node into the distribution network linearized network power flow model to calculate the node voltage phasor 3 days after the current date.

[0065] Step 6: Based on the node voltage phasors and a conventional power system power flow calculation method, obtain the line load rate and distribution transformer load rate between nodes in the distribution network three days after the current date;

[0066] Step 7: Obtain an optimized scheduling scheme for mobile energy storage in the distribution network based on the line load rate, distribution transformer load rate or node voltage phasor between nodes in the distribution network.

[0067] Example 2

[0068] Based on Example 1, this embodiment provides a process for obtaining an optimized scheduling solution for mobile energy storage in a distribution network based on line load rates, distribution transformer load rates, and node voltage phasors between nodes in the distribution network.

[0069] The mobile energy storage early warning optimization model is used to issue early warnings on the distribution network operation status based on the line load rate, distribution transformer load rate or node voltage phasor between nodes in the distribution network, as well as the corresponding preset limits.

[0070] In this embodiment, the warning levels include: red warning, yellow warning and green warning;

[0071] According to the preset line load rate upper limit and lower limit Early warning of line load rate, including: responding to line load rate less than , generates a green warning of line load rate, in response to the line load rate being greater than or equal to and less than or equal to , generates a yellow warning of line load rate; in response to the line load rate being greater than , generating a red alert for line load rate.

[0072] Specifically, such as Figure 2As shown in the figure, the upper limit of the line load rate is 100% and the lower limit is 80%. The warning rules for the line load rate include: in response to the line load rate being less than 80%, a green line load rate warning is generated; in response to the line load rate being greater than or equal to 80% and less than or equal to 100%, a yellow line load rate warning is generated; and in response to the line load rate being greater than 100%, a red line load rate warning is generated.

[0073] According to the preset upper limit of distribution transformer load rate and lower limit Early warning of distribution transformer load rate, including: responding to the distribution transformer load rate being less than , generates a green warning of the distribution transformer load rate, in response to the distribution transformer load rate being greater than or equal to and less than or equal to , generates a yellow warning of the distribution transformer load rate; in response to the distribution transformer load rate being greater than , generating a red warning for the distribution transformer load rate.

[0074] Specifically, such as Figure 3 As shown in the figure, the upper limit of the distribution transformer load factor is 100% and the lower limit is 80%. The warning rules for the distribution transformer load factor include: a green warning is generated in response to the distribution transformer load factor being less than 80%, a yellow warning is generated in response to the distribution transformer load factor being greater than or equal to 80% and less than or equal to 100%, and a red warning is generated in response to the distribution transformer load factor being greater than 100%.

[0075] Calculate the node voltage deviation ratio based on the node voltage phasor and the preset node voltage deviation ratio first upper limit value , first lower limit , the second upper limit and the second lower limit Early warning of voltage amplitude, including:

[0076] Calculate the node voltage deviation ratio using the formula:

[0077] ;

[0078] Where, is the magnitude of the node voltage, is the rated voltage corresponding to the distribution network.

[0079] like Figure 4 As shown, in response to the node voltage deviation ratio or , generates a voltage amplitude red warning, in response to the node voltage deviation ratio or , generates a yellow warning of voltage amplitude, in response to the node voltage deviation ratio , generating a green warning of voltage amplitude.

[0080] In this embodiment, and Depend on and Calculation results:

[0081] ;

[0082] ;

[0083] Based on the above warning results, mobile energy storage is preferentially allocated to distribution transformers in the red warning area, followed by those in the yellow warning area, and the remaining mobile energy storage is allocated to those in the green warning area, thereby generating an optimized scheduling plan for mobile energy storage in the distribution network.

[0084] Example 3

[0085] Based on Example 1, this example provides a process for constructing and training a rolling forward prediction model.

[0086] The rolling forward prediction model is constructed based on the BP neural network. The process includes: establishing the BP neural network model and selecting the neural network structure; determining the number and connection method of neurons in the input layer, hidden layer and output layer, as well as the learning rate, number of iterations, loss function and activation function.

[0087] Specifically, the input layer parameters of the rolling forward prediction model include: the voltage step-down flow of each distribution transformer in the distribution network, the actual output of distributed power sources, and the amplitude and phase angle of all node voltages; the output layer parameters include: the voltage step-down flow of each distribution transformer in the distribution network, the predicted output value of distributed power sources, the predicted value of node load, and the balanced node voltage.

[0088] Preferably, the rolling look-ahead prediction model uses a ReLU activation function.

[0089] The training of the rolling forward forecast model includes:

[0090] The actual operation data of the distribution network in the historical actual operation dataset is divided into a training set and a validation set.

[0091] Specifically, 70% of the historical actual operation data are selected as the training set and 30% of the historical actual operation data are selected as the validation set according to the time distance. That is, the historical actual operation data from (t-30) to (t-10) days are the training set, and the data from (t-9) to (t-1) days are the validation set.

[0092] The output of the rolling forward prediction model is calculated using the data in the training set, and the prediction error is calculated based on the actual value. Based on the prediction error, the weight and bias parameters of the rolling forward prediction model are adjusted until the model meets the convergence conditions.

[0093] Use the data in the validation set to evaluate the trained rolling forward forecast model and perform parameter optimization.

[0094] Example 4

[0095] This embodiment provides an electronic device, including: a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the distribution network mobile energy storage scheduling method based on rolling forward warning as described in Example 1 or Example 2 is implemented.

[0096] Example 5

[0097] This embodiment provides a computer storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for dispatching mobile energy storage in a distribution network based on rolling forward warning as described in Example 1 or Example 2 is implemented.

[0098] The storage medium may include, for example, a storage component of a tablet computer, a hard disk of a computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disk read-only memory (CD-ROM), a USB memory, or any combination thereof. The computer storage medium may be any combination of one or more computer storage media.

[0099] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application 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 application.

Claims

1. A method for dispatching mobile energy storage in a distribution network based on rolling forward warning, characterized in that: include: Obtain the actual operation data of the distribution network within a preset historical period, collect the current operation data in real time, and obtain the historical actual operation data set based on all the operation data; One day's actual operation data is selected from the historical actual operation data set, and input into a pre-built distribution network linearized network power flow model to calculate and obtain a distribution network node impedance matrix; Input the historical actual operation data within a preset day a before the current date into the pre-trained rolling forward prediction model, and make rolling predictions on the voltage step-down flow of each distribution transformer, the output forecast value of the distributed generation, the node load forecast value, and the balanced node voltage within a preset day b after the current date; According to the predicted values ​​of the step-down power flows of the distribution transformers, the predicted values ​​of the node loads, and the predicted values ​​of the outputs of the distributed power sources, a conjugate complex matrix of the total injected power of each node in the distribution network within a preset b days after the current date is obtained; wherein the total injected power of the load node in the distribution network is the same as the predicted value of the node load of the load node, and the total injected power of the power source node in the distribution network is the same as the sum of the predicted values ​​of the outputs of the distributed power sources connected to the power source node; Input the distribution network node impedance matrix, the balanced node voltage prediction value and the total injected power conjugate complex matrix of each node into the distribution network linearized network power flow model to calculate the node voltage phasor within b days after the current date; According to the node voltage phasor, obtain the line load rate and distribution transformer load rate between nodes in the distribution network within b days after the current date; Obtaining an optimized dispatching scheme for mobile energy storage in the distribution network based on line load rates, distribution transformer load rates, and node voltage phasors between nodes in the distribution network; The linearized network power flow model of the distribution network is: Where, is the node voltage phasor, , is the number of nodes, The balancing nodes connected to the upper power grid are not included; is the unit phasor, and all elements in the phasor are 1; is the distribution network node impedance matrix, which represents the impedance between connected nodes; is the conjugate complex matrix of the total injected power of each node; , the conjugate complex number is: ; 、 are the voltage amplitude and phase angle of the equilibrium node respectively.

2. The method for dispatching mobile energy storage in distribution network based on rolling forward warning according to claim 1 is characterized in that: The obtaining of the historical actual operation data set based on all the operation data includes: preprocessing all the operation data; the preprocessing includes: abnormal data removal, vacant data filling and normalization processing.

3. The method for dispatching mobile energy storage in distribution network based on rolling forward warning according to claim 1, characterized in that: Calculate the distribution network node impedance matrix, including: The actual operation data of the selected date is collected at the preset collection frequency to obtain historical data of c time nodes; Inputting historical data of c time nodes into the distribution network linearized network power flow model respectively, and calculating c node impedance matrices; The average of the c node impedance matrices is calculated to obtain the distribution network node impedance matrix Z.

4. The method for dispatching mobile energy storage in a distribution network based on rolling forward warning according to claim 3 is characterized in that: The construction of the rolling forward prediction model includes: Establish BP neural network model and select neural network structure; Determine the number and connection mode of neurons in the input layer, hidden layer, and output layer, as well as the learning rate, number of iterations, loss function, and activation function; the input layer parameters include: the step-down power flow of each distribution transformer in the distribution network, the actual output of the distributed power source, and the amplitude and phase angle of all node voltages; The output layer parameters include: the step-down flow of each distribution transformer in the distribution network, the output forecast value of the distributed power source, the node load forecast value and the balanced node voltage; The training of the rolling forward prediction model includes: The actual operation data of the distribution network within a preset historical period is divided into a training set and a validation set; Calculate the output of the rolling forward forecast model using the data in the training set and calculate the forecast error based on the actual value; According to the prediction error, adjust the weight and bias parameters of the rolling forward prediction model until the model converges; Use the data in the validation set to evaluate the trained rolling forward forecast model and perform parameter optimization.

5. The method for dispatching mobile energy storage in distribution network based on rolling forward warning according to claim 4 is characterized in that: Obtaining an optimized dispatching scheme for mobile energy storage in the distribution network based on line load rates, distribution transformer load rates, and node voltage phasors between nodes in the distribution network includes: Providing early warning of the distribution network operation status based on the line load rate, distribution transformer load rate or node voltage phasor between nodes in the distribution network and the corresponding preset limits; wherein the warning levels include: red warning, yellow warning and green warning; Based on the principle of highest priority configuration in red warning areas and second priority configuration in yellow warning areas, an optimized dispatching plan for mobile energy storage in the distribution network is generated according to the warning results of the distribution network operation status.

6. The method for dispatching mobile energy storage in a distribution network based on rolling forward warning according to claim 5, characterized in that: Early warning of the distribution network operation status is provided based on the line load rate, distribution transformer load rate or node voltage phasor between nodes in the distribution network and the corresponding preset limits, including: According to the preset upper limit of distribution transformer load rate and lower limit Early warning of distribution transformer load rate, including: In response to the load factor of the distribution transformer being less than , generating a green warning of the distribution transformer load rate, In response to the load factor of the distribution transformer being greater than or equal to and less than or equal to , generating a yellow warning of the distribution transformer load rate; In response to the load factor of the distribution transformer being greater than , generating a red warning for the distribution transformer load rate; According to the preset line load rate upper limit and lower limit Early warning of line load rate, including: In response to the line load rate being less than , generating a green warning of line load rate, In response to a line load factor greater than or equal to and less than or equal to , generating a yellow warning of line load rate; In response to a line load greater than , generating a red warning for line load rate; Calculate the node voltage deviation ratio based on the node voltage phasor and the preset node voltage deviation ratio first upper limit value , first lower limit , the second upper limit and the second lower limit Early warning of voltage amplitude, including: Calculate the node voltage deviation ratio using the formula: ； Where, is the magnitude of the node voltage, is the rated voltage corresponding to the distribution network; Response to node voltage deviation ratio or , generating a red warning of voltage amplitude, Response to node voltage deviation ratio or , generating a yellow warning of voltage amplitude, Response to node voltage deviation ratio , generating a green warning of voltage amplitude.

7. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the distribution network mobile energy storage scheduling method based on rolling forward warning according to any one of claims 1 to 6 is implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method for dispatching mobile energy storage in a distribution network based on rolling forward warning according to any one of claims 1 to 6 is implemented.

Images