Power control method, energy router, power control system and storage medium

By using the accuracy verification algorithm in electric vehicle charging equipment to screen high-accuracy prediction parameters, the problem of inaccurate power distribution when the electricity environment of electric vehicle charging equipment changes is solved, and the power control stability of energy routers is improved.

CN113809789BActive Publication Date: 2024-11-15CHINA SOUTHERN POWER GRID ELECTRIC VEHICLE SERVICE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110892995.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-08-04
Publication Date
2024-11-15
Estimated Expiration
2041-08-04

AI Technical Summary

Technical Problem

When the power control method of the existing electric vehicle charging equipment changes in the power environment, the prediction model prediction results are far from the actual power demand, resulting in inaccurate distribution of power of the energy router to the charging equipment, affecting stability.

Method used

By obtaining prediction parameters, using the accuracy verification algorithm to calculate the prediction accuracy, and determining whether it is a power calculation parameter based on the preset threshold, filtering out high-accuracy prediction parameters as the characteristic output power of the energy router, ensuring that the power distribution meets the charging device actual needs.

Benefits of technology

It improves the power control stability of the energy router, avoids interference and misjudgment caused by inaccurate prediction, and ensures that the power requirements of the charging equipment are accurately met.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113809789B_ABST
    Figure CN113809789B_ABST
Patent Text Reader

Abstract

The present application relates to a power control method, including: obtaining prediction parameters; performing accuracy calculation on the prediction parameters according to a preset accuracy verification algorithm to obtain prediction accuracy; comparing the prediction accuracy with a preset threshold value, and judging whether the prediction parameters are power calculation parameters according to the comparison result. The present application also relates to an energy router, power control system and storage medium using the above method. The prediction accuracy of the present invention is high and the power control is stable.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of power distribution technology, and in particular to a power control method, an energy router, a power control system and a storage medium. Background Art

[0002] With the continuous development of new energy technology, new energy charging technology has gradually entered people's lives, among which electric vehicles are the most widely used. As the application scope of electric vehicles becomes wider and wider, electric vehicles have become an important means of transportation in people's lives, and the charging problem of electric vehicles has become a major concern for people. The charging of electric vehicles mainly depends on the power control technology of the charging equipment.

[0003] In the related art, an energy router connected to the transformer is set up, and the energy router is connected to the charging device. Under the control of a preset power control method, the energy router distributes the power obtained from the transformer to the charging device. A prediction model can be constructed based on historical data of the control power allocated to the charging device, and then the prediction model can be used to perform predictions. A power control instruction is generated based on the prediction result (the prediction result is used to characterize the power demand prediction value of the charging device), and finally a charging power value corresponding to the prediction result is allocated to the charging device according to the power control instruction.

[0004] However, when the power consumption environment of the charging device changes, the prediction result obtained by the current prediction model is far from the actual value of the power demand of the charging device, that is, the prediction result is inaccurate. Continuing to control the power of the charging device using the inaccurate prediction result will cause great interference to the energy router when allocating charging power to the charging device, and even cause serious misjudgment, so that the power allocation effect of the energy router to the charging device cannot meet the actual charging power demand of the charging device, resulting in poor stability of the power control of the energy router. Summary of the invention

[0005] Based on this, it is necessary to provide a power control method, an energy router, a power control system and a storage medium to address the technical problem of poor stability of the above-mentioned power control.

[0006] The present invention provides a power control method, which is applied to an energy router and comprises:

[0007] Get prediction parameters;

[0008] Calculating the accuracy of the prediction parameters according to a preset accuracy verification algorithm to obtain the prediction accuracy;

[0009] The prediction accuracy is compared with a preset threshold, and whether the prediction parameter is a power calculation parameter is determined according to the comparison result; wherein the power calculation parameter is used to calculate the characteristic output power of the energy router, and the characteristic output power is used to instruct the energy router to distribute power to the charging device.

[0010] In one embodiment, the preset threshold includes a first threshold; and the step of judging whether the prediction parameter is a power calculation parameter according to the comparison result includes:

[0011] When the prediction accuracy is greater than or equal to the first threshold, determining that the prediction parameter is the power calculation parameter; and / or,

[0012] When the prediction accuracy is less than the first threshold, the process returns to the step of obtaining prediction parameters.

[0013] In one embodiment, the energy router is used to connect to a power supply device and / or a charging device; the step of obtaining the prediction parameter includes:

[0014] Acquire a modeling input parameter; wherein the modeling input parameter includes at least one of a first modeling parameter and a second modeling parameter, the first modeling parameter is a parameter used to characterize the operating state of the power supply device, and the second modeling parameter is a parameter used to characterize the operating state of the charging device connected to the energy router;

[0015] Establishing a prediction model according to the modeling input parameters;

[0016] A prediction calculation is performed according to the prediction model to obtain the prediction parameters.

[0017] In one embodiment, the preset threshold includes a first threshold and a second threshold, and the first threshold is greater than the second threshold; and the step of judging whether the prediction parameter is a power calculation parameter according to the comparison result includes:

[0018] When the prediction accuracy is less than the first threshold and greater than or equal to the second threshold, returning to the step of performing prediction calculation according to the prediction model to obtain the prediction parameter; and / or,

[0019] When the prediction accuracy is less than the second threshold, the method returns to the step of obtaining modeling input parameters to establish a new prediction model.

[0020] In one embodiment, the step of determining whether the prediction parameter is a power calculation parameter according to the comparison result further includes:

[0021] When the prediction accuracy is less than the second threshold, at least part of the data in the modeling input parameters is deleted according to a preset data cleaning rule.

[0022] In one embodiment, the preset threshold includes a first threshold; and the method further includes:

[0023] In the step of obtaining prediction parameters, the prediction parameters include at least two;

[0024] In the step of calculating the accuracy of the prediction parameters according to a preset accuracy verification algorithm to obtain the prediction accuracy, the prediction accuracy includes at least two, and the prediction accuracy is set in a one-to-one correspondence with the prediction parameters;

[0025] The step of comparing the prediction accuracy with a preset threshold and judging whether the prediction parameter is a power calculation parameter according to the comparison result includes:

[0026] Each of the prediction accuracies is compared with the first threshold value respectively, and the prediction parameter that meets the preset condition is determined as the power calculation parameter according to the comparison result; wherein the preset condition is used to characterize that the prediction accuracy corresponding to the prediction parameter is greater than or equal to the first threshold value.

[0027] In one embodiment, the step of determining the prediction parameter that meets the preset condition as the power calculation parameter according to the comparison result includes:

[0028] From the at least two prediction parameters that meet the preset conditions, the prediction parameter with the highest prediction accuracy is selected to be determined as the optimal power calculation parameter.

[0029] In one embodiment, the energy router is used to connect to a power supply device, and the power supply device is used to connect to a power-consuming device; the characteristic output power of the energy router includes the maximum output power of the energy router, and the method further includes:

[0030] According to a preset first calculation rule, the power calculation parameter and the first power parameter are calculated to obtain the maximum output power of the energy router; wherein the power calculation parameter includes the actual power prediction value of the power supply device, and the first power parameter includes the rated power of the power supply device, the maximum input power of the energy router and the actual power of the power-consuming device.

[0031] In one embodiment, the energy router is further used to connect to a charging device; the characteristic output power of the energy router also includes a charging control power, and the charging control power is used to instruct the charging device to perform charging work, and the method further includes:

[0032] According to a preset second calculation rule, the maximum output power of the energy router, the power calculation parameter and the second power parameter are calculated to obtain the charging control power; wherein the power calculation parameter also includes the actual power prediction value of the charging device and the required power prediction value of the charging device, and the second power parameter includes the actual power of the energy router and the required power of the energy router;

[0033] The charging control power is input to the charging device.

[0034] An energy router, comprising:

[0035] A prediction module, for obtaining prediction parameters; and

[0036] A control module is used to calculate the accuracy of the prediction parameter according to a preset accuracy verification algorithm to obtain the prediction accuracy; compare the prediction accuracy with a preset threshold, and determine whether the prediction parameter is a power calculation parameter according to the comparison result, wherein the power calculation parameter is used to calculate the characteristic output power of the energy router, and the characteristic output power is used to instruct the energy router to distribute power to the charging device.

[0037] A power control system comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned power control method when executing the computer program.

[0038] A computer-readable storage medium stores a computer program, which implements the steps of the power control method when executed by a processor.

[0039] In the above-mentioned power control method, energy router, power control system and storage medium, by calculating the prediction accuracy of the prediction parameter, and according to the comparison result of the prediction accuracy with the preset threshold value, it is judged whether the prediction parameter is suitable as the power calculation parameter for calculating the characteristic output power of the energy router, that is, through the above steps, a selection mechanism for power calculation parameters is constructed, and the prediction parameters with high prediction accuracy are screened as power calculation parameters, avoiding great interference or serious misjudgment when the energy router allocates charging power to the charging device due to inaccurate prediction parameters (that is, low prediction accuracy), and providing conditions for obtaining the characteristic output power of the energy router according to the power calculation parameters, so as to ensure that the effect of the energy router allocating power to the charging device according to the characteristic output power meets the actual charging power demand of the charging device, and effectively improves the stability of the power control of the energy router. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application 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 described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 is a schematic flow chart of a power control method in an embodiment;

[0042] Figure 2 for Figure 1 A flowchart of an embodiment of step 106;

[0043] Figure 3 A schematic diagram of a flow chart of steps for establishing a prediction model in an embodiment;

[0044] Figure 4 for Figure 1 A flowchart of another embodiment of step 106;

[0045] Figure 5 A schematic flow chart of the steps for calculating the characteristic output power of an energy router in one embodiment;

[0046] Figure 6 Schematic diagram of the module structure of an energy router in one embodiment. Specific embodiments

[0047] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0048] like Figure 1 As shown, in one embodiment, the present invention provides a power control method, the power control method is applied to Figure 6 The energy router 600 shown is used to be connected to the power supply device 20 and the charging device 30 respectively.

[0049] It should be noted that the power supply device 20 is used to connect to the external power supply device 10, and receive the electric energy input by the external power supply device 10, and use the electric energy input by the power supply device 10 as the total input power; the power supply device 20 is also used to connect to the energy router 600 and the power-consuming device 40 respectively, and provide electric energy to the energy router 600 and the power-consuming device 40 respectively.

[0050] The charging device 30 is used to be connected to the energy router 600 and the powered device respectively, receive the charging power input by the energy router 600, and use the charging power to charge the powered device connected thereto.

[0051] For example, in some embodiments, the power supply device 20 is a transformer, the power consumption device 40 includes but is not limited to at least one of household appliances used in residential areas and industrial machines in industrial parks, and the charging device 30 includes but is not limited to charging piles, and the power receiving device includes but is not limited to one of electric vehicles, electric bicycles and storage batteries; wherein the transformer is used to receive the total input power input by the external power supply device 10, and to distribute the input total power. Specifically, the transformer inputs a part of the total input power as power consumption to the power consumption device 40, and also uses a part of the total input power as charging control power, and distributes the charging power to the charging device 30 through the energy router 600 using the charging control power, and the charging device 30 can charge the electric vehicle according to the charging power.

[0052] The power control method includes:

[0053] Step 102, obtaining prediction parameters.

[0054] Among them, in step 102, the prediction parameters include but are not limited to at least one of the actual power prediction value of the power supply device 20, the actual power prediction value of the charging device 30 and the required power prediction value of the charging device 30. In addition, it is worth mentioning that the prediction parameter acquisition method can be specifically set according to the actual use needs. For example, in one embodiment, multiple model training samples can be collected before step 102, and the energy router 600 establishes a prediction model based on the multiple model training samples. The energy router 600 uses the prediction model to perform prediction operations to obtain prediction parameters; of course, in other embodiments, the process of using the prediction model to perform prediction operations to obtain prediction parameters can also be performed in an external prediction calculation device, and the energy router 600 receives the prediction parameters from the prediction calculation device.

[0055] Step 104: Calculate the accuracy of the prediction parameters according to a preset accuracy verification algorithm to obtain the prediction accuracy.

[0056] In step 104, the preset accuracy verification algorithms specifically include but are not limited to the following two:

[0057] The first type is that the accuracy verification algorithm includes at least one of the RMSE function, MAPE function and MAE function, and the accuracy of the prediction parameters is calculated based on at least one of the RMSE function, MAPE function and MAE function, and at least one of the RMSE value, MAPE value and MAE value is obtained as the prediction accuracy. Specifically:

[0058] (1) RMSE function:

[0059]

[0060] (2) MAPE function:

[0061]

[0062] (3) MAE function:

[0063]

[0064] In the above formulas (1)-(3), N is the number of model training samples of the prediction model, and the serial number of any model training sample is i, and the value range of i is [1, N]. i is the actual value corresponding to the i-th model training sample (i.e., the parameter value obtained by actual measurement corresponding to the i-th model training sample), p i It is represented as the predicted value corresponding to the i-th model training sample (that is, the parameter value obtained by performing a prediction operation on the i-th model training sample using the prediction model).

[0065] The second type of accuracy verification algorithm includes the ROC-AUC algorithm. First, draw the ROC curve, where the ROC curve refers to the curve that shows the degree of inaccurate predictions and serious impacts obtained by the subjects under different judgment criteria under specific stimulus conditions. Then, calculate the AUC (Area Under the Curve) value based on the ROC curve. The AUC value is the area of ​​the ROC curve, that is, the AUC value is the size of the area under the ROC curve; the AUC value is used as the prediction accuracy. The larger the AUC value, the better the prediction effect, and vice versa. The specific steps include:

[0066] Step 1: Build a binary classification model, where:

[0067] (1) The predicted value is greater than the actual value and does not exceed 10% of the actual value, which is recorded as TP (True Positive). TP indicates that the warning result is that the energy router 600 has an abnormality, and the abnormality actually occurs, which is a correct warning in the traditional sense.

[0068] (2) The predicted value is less than the actual value and the actual value exceeds the predicted value by 10% and is recorded as FP (False Positive). FP indicates that the warning result is that the energy router 600 is abnormal, but in fact no abnormality occurs, which is a false alarm in the traditional sense.

[0069] (3) If the predicted value is less than the actual value and the actual value does not exceed 10% of the predicted value, it is TN (True Negative). TN means that the warning result is that the energy router 600 has no abnormality, and in fact, no abnormality has occurred.

[0070] (4) A predicted value greater than the actual value and exceeding 10% of the actual value is recorded as FN (Fasle Negative). FN indicates that the warning result is that the energy router 600 has no abnormality, but an abnormality actually occurred, which is a missed alarm in the traditional sense.

[0071] Table 1

[0072]

[0073] Among them, TPR = TP / (TP+FN); FPR = FP / (FP+TN), TPR is the proportion of samples that actually have abnormalities and the early warning model correctly warns, and FPR is the proportion of samples that actually have no abnormalities and the early warning model does not report abnormalities.

[0074] Step 2: Draw the ROC curve with the above FPR as the horizontal axis and TPR as the vertical axis.

[0075] Step 3: Perform integral calculation based on the ROC curve to obtain the AUC (Area Under the Curve) value, that is, the AUC value is the size of the area under the ROC curve.

[0076] It should be noted that the above-mentioned RMSE value, MAPE value and MAE value are all used to characterize the prediction accuracy corresponding to any model training sample in the prediction model, and can better reflect the individual situation of the prediction accuracy of the prediction parameters obtained under any model training sample; and the above-mentioned AUC value is used to characterize the prediction accuracy corresponding to the prediction model as a whole, and can better reflect the overall situation of the prediction accuracy of all prediction parameters of the prediction model.

[0077] Step 106, comparing the prediction accuracy with a preset threshold, and determining whether the prediction parameter is a power calculation parameter based on the comparison result.

[0078] In step 106 , the power calculation parameter is used to calculate the characteristic output power of the energy router 600 , and the characteristic output power is used to instruct the energy router 600 to distribute power to the charging device 30 .

[0079] In some embodiments, the prediction accuracy includes but is not limited to at least one of the RMSE value, MAPE value, MAE value and AUC value. At least one of the RMSE value, MAPE value, MAE value and AUC value can be compared with a preset threshold, and whether the prediction parameter is a power calculation parameter can be determined based on the comparison result. It should be noted that the specific parameters for comparison can be set according to the actual setting requirements. For example, in some embodiments, only one of the RMSE value, MAPE value, MAE value and AUC value can be selected for comparison with the preset threshold; in some embodiments, at least one of the RMSE value, MAPE value and MAE value can be selected for comparison with the preset threshold, and the AUC value can also be selected for comparison with the preset threshold.

[0080] Specifically, the preset thresholds include a first threshold. When the prediction accuracy is greater than or equal to the preset threshold, the prediction parameter is determined to be a power calculation parameter. Of course, the value of the first threshold can be specifically set according to actual usage requirements. For example, in some embodiments, the first threshold is 80%. If the prediction accuracy is greater than 80%, the prediction parameter is determined to be a power calculation parameter. If not, the prediction parameter is immediately excluded from the selection range of the power calculation parameter.

[0081] In the above-mentioned power control method, energy router, power control system and storage medium, by calculating the prediction accuracy of the prediction parameter, and according to the comparison result of the prediction accuracy with the preset threshold value, it is judged whether the prediction parameter is suitable as the power calculation parameter for calculating the characteristic output power of the energy router 600, that is, through the above steps, a selection mechanism for power calculation parameters is constructed, and the prediction parameters with high prediction accuracy are screened as power calculation parameters, which avoids great interference or serious misjudgment when the energy router 600 allocates charging power to the charging device 30 due to inaccurate prediction parameters (that is, low prediction accuracy), and provides conditions for obtaining the characteristic output power of the energy router 600 according to the power calculation parameters, so as to ensure that the effect of the energy router 600 allocating power to the charging device 30 according to the characteristic output power meets the actual charging power demand of the charging device 30, and effectively improves the stability of the power control of the energy router 600.

[0082] like Figure 2 , as shown, in one embodiment, when the preset threshold includes a first threshold; the step of judging whether the prediction parameter is a power calculation parameter according to the comparison result includes at least one of the following steps:

[0083] Step 202: When the prediction accuracy is greater than or equal to a first threshold, it is determined that the prediction parameter is a power calculation parameter.

[0084] Wherein, in step 202, the prediction accuracy includes at least one of the RMSE value, the MAPE value, the MAE value and the AUC value, and the value of the first threshold value can be specifically set according to the actual use requirements. For example, in some embodiments, the first threshold value is 80%. When the AUC value is greater than or equal to 80%, the prediction parameter corresponding to the prediction accuracy greater than or equal to 80% is used as the power calculation parameter. It should be noted that after step 202, it is possible to return to step 102 and execute steps 102-104 again to compare the prediction accuracy corresponding to the prediction parameter obtained in the new round with the first threshold value. If the prediction accuracy corresponding to the prediction parameter in the new round is greater than or equal to the first threshold value, step 202 is executed to obtain the second power calculation parameter, and so on, to obtain several power calculation parameters, and the several power calculation parameters together constitute the feature parameter set.

[0085] Step 204: when the prediction accuracy is less than the first threshold, the process returns to the step of obtaining prediction parameters, that is, returns to the above step 102.

[0086] Among them, in step 204, for example, in some embodiments, if the first threshold is 80%, when the AUC value is less than 80%, the prediction parameters corresponding to the prediction accuracy less than 80% are excluded from the selection range of the power calculation parameters, and return to step 102.

[0087] After step 204, the next round of steps 102-104 is executed again, and the prediction accuracy corresponding to the prediction parameters obtained in the new round is compared with the first threshold. If the prediction accuracy corresponding to the prediction parameters of the new round is less than the first threshold, step 204 is executed and returns to step 102 again, and so on, until the prediction accuracy corresponding to the prediction parameters of the new round is greater than or equal to the first threshold, so as to execute step 202 to obtain the power calculation parameters.

[0088] Preferably, in some embodiments, after executing step 202, the first power calculation parameter can be obtained when executing step 202 for the first time. After obtaining the first power calculation parameter, step 102 is no longer returned to (i.e., execution of the next round of steps 102-106 is stopped), and the calculation steps of the characteristic output power of the energy router 600 after step 106 are directly executed, that is, calculation is performed according to the first power calculation parameter to obtain the characteristic output power of the energy router 600.

[0089] By setting the above steps, after obtaining the first power calculation parameter, the characteristic output power of the energy router 600 can be calculated, thereby realizing the rapid acquisition of the characteristic output power of the energy router 600, so that the energy router 600 can make a power control response according to the characteristic output power more quickly, which is beneficial to improving the response speed of the energy router 600.

[0090] Further, in some embodiments, before executing the above step 202 and / or step 204, step 102 may be executed multiple times to obtain at least two prediction parameters; then, step 104 is executed to obtain the prediction accuracy of multiple prediction parameters respectively, and the prediction accuracy is set in one-to-one correspondence with the prediction parameter; finally, each prediction accuracy is compared with the first threshold value respectively, and step 202 is executed for the prediction parameter that meets the preset condition according to the comparison result, wherein the preset condition is used to characterize that the prediction accuracy corresponding to the prediction parameter is greater than or equal to the first threshold value, that is, the prediction parameter that meets the preset condition is judged as a power calculation parameter, and at least two power calculation parameters are obtained, and then a characteristic parameter set composed of all power calculation parameters is obtained. Through the above step setting, conditions are provided for the subsequent energy router 600 to calculate and obtain the characteristic output power of multiple energy routers 600 according to the characteristic parameter set.

[0091] More preferably, in some embodiments, step 106 includes: selecting a prediction parameter with the highest prediction accuracy from at least two prediction parameters that meet preset conditions and determining it as the optimal power calculation parameter.

[0092] Specifically, the power calculation parameter with the highest prediction accuracy is selected from the characteristic parameter set as the optimal power calculation parameter, and then the calculation step of the characteristic output power of the energy router 600 is performed, that is, calculation is performed according to the optimal power calculation parameter to obtain the characteristic output power of the energy router 600.

[0093] Through the above-mentioned step settings, the power calculation parameter with the highest prediction accuracy based on the characteristic parameter set is used as the optimal power calculation parameter to calculate the characteristic output power of the energy router 600, which greatly improves the accuracy of the characteristic output power calculation of the energy router 600, so that the effect of the energy router 600 distributing power to the charging device 30 according to the characteristic output power is more in line with the actual charging power demand of the charging device 30, and further improves the stability of the power control of the energy router 600.

[0094] It should be noted that, in order to avoid the problem of data redundancy in the energy router 600, preferably, in some embodiments, the following steps are further included:

[0095] After the step of obtaining the first power calculation parameter, the prediction parameters whose prediction accuracy is less than the first threshold are deleted; or, after the step of selecting the power calculation parameter with the highest prediction accuracy from the feature parameter set, the prediction parameters whose prediction accuracy is less than the first threshold are deleted, and / or, the power calculation parameters other than the power calculation parameter with the highest prediction accuracy in the feature parameter set are partially deleted or completely deleted.

[0096] The above-mentioned steps can avoid data redundancy caused by storing too many power calculation parameters in the energy router 600, improve the utilization rate of the storage space of the energy router 600, and avoid increasing the amount of data calculation of the energy router 600 due to the large amount of data, thereby greatly improving the computing efficiency of the energy router 600.

[0097] like Figure 3 , 6 As shown, in one embodiment, the step of obtaining prediction parameters includes:

[0098] Step 302: Obtain modeling input parameters.

[0099] Wherein, in step 302, the modeling input parameter includes at least one of a first modeling parameter and a second modeling parameter.

[0100] Specifically, the first modeling parameter is a parameter used to characterize the operating status of the power supply device 20. For example, in some embodiments, the first modeling parameter includes but is not limited to at least one of the actual power value of the power supply device 20 at each moment (i.e., the total input power input to the power supply device 20), voltage, current, voltage rise and fall rates, current rise and fall rates, active power, reactive power, and the rise and fall rates of the actual power value.

[0101] In addition, the second modeling parameter is a parameter used to characterize the operating state of the charging device 30 connected to the energy router 600. For example, in some embodiments, the second modeling parameter includes but is not limited to at least one of the actual power value, the required power value, the voltage, the current, the rate of increase and decrease of the voltage, the rate of increase and decrease of the current, the active power, the reactive power, and the rate of increase and decrease of the actual power value of the charging device 30 at each moment.

[0102] Step 304: Establish a prediction model based on the modeling input parameters.

[0103] Among them, in step 304, first, obtain model training samples: collect and organize the modeling input parameters according to a preset time period to obtain multiple model training samples, and the multiple model training samples are set in one-to-one correspondence with multiple moments; then, establish a prediction model: use a preset neural network algorithm to perform deep learning on each moment and the model training samples corresponding to each moment to obtain a prediction model.

[0104] It is worth mentioning that the above-mentioned neural network algorithm can be pre-set according to actual use needs. For example, in some embodiments, the neural network algorithm includes an EMD-SLSTM combination algorithm. The EMD-SLSTM combination prediction algorithm is a prediction algorithm formed by combining two algorithms, the EMD algorithm (EMD, referring to Empirical Mode Decomposition) and the SLSTM algorithm (SLSTM, referring to Stack Long Short-term Memory).

[0105] It should be noted that in the above step 304 , the prediction model established according to the first modeling parameters is the actual power prediction model of the power supply device 20 ; and the prediction model established according to the second modeling parameters is the actual power and required power prediction model of the charging device 30 .

[0106] Step 306: perform prediction calculation according to the prediction model to obtain prediction parameters.

[0107] Among them, in step 306, after performing a prediction operation using the actual power prediction model of the above-mentioned power supply equipment 20, the obtained prediction parameters include the actual power prediction value of the power supply equipment 20; after performing a prediction operation using the actual power and demand power prediction model of the above-mentioned charging equipment 30, the obtained prediction parameters include the actual power prediction value of the charging equipment 30 and the demand power prediction value of the charging equipment 30.

[0108] Through the setting of the above steps 302-306, it is achieved that according to any moment when the power supply device 20 is currently running, the actual power prediction value of the power supply device 20 at any moment is predicted by the prediction model, which provides a condition for analyzing its load situation at any moment according to the actual power prediction value of the power supply device 20 at any moment, and / or, it is achieved that according to any moment when the charging device 30 is currently running, the actual power prediction value and the required power prediction value of the charging device 30 at any moment are predicted by the prediction model, which provides a condition for analyzing its charging demand situation at any moment according to the actual power prediction value and the required power prediction value of the charging device 30 at any moment, so that in the subsequent steps, at any moment, the charging power can be allocated to the charging device 30 at that moment according to the actual power prediction value of the power supply device 20, the actual power prediction value of the charging device 30 and the required power prediction value of the charging device 30 at the corresponding moment.

[0109] like Figure 4, as shown, in one embodiment, the preset threshold includes a first threshold and a second threshold, and the first threshold is greater than the second threshold. It is worth mentioning that the specific values ​​of the first threshold and the second threshold can be set according to actual needs. For example, in some embodiments, the first threshold is 80% and the second threshold is 60%.

[0110] The step of judging whether the prediction parameter is a power calculation parameter according to the comparison result includes at least one of the following steps:

[0111] Step 402, when the prediction accuracy is less than the first threshold and greater than or equal to the second threshold, the process returns to the step of performing prediction calculation according to the prediction model to obtain prediction parameters, that is, returns to the above step 306.

[0112] Among them, in step 402, in some embodiments, the AUC value is used as the prediction accuracy. If the AUC value is less than the first threshold and greater than or equal to the second threshold, that is, when the AUC value is between 60% and 80%, the prediction parameter corresponding to the AUC value in the range of 60% to 80% is excluded from the selection range of the power calculation parameter, and the process returns to step 306 to obtain a new round of prediction accuracy (i.e., AUC value) for re-judgment. Of course, in other embodiments, step 306 can be executed multiple times according to a preset number of times to obtain multiple prediction parameters, and then the AUC values ​​of the multiple prediction parameters are obtained through step 104.

[0113] If some AUC values ​​are greater than 80%, the prediction parameters corresponding to the AUC values ​​greater than or equal to 80% are used as power calculation parameters; if multiple AUC values ​​are less than 80%, the prediction parameters corresponding to the AUC values ​​less than 80% are excluded from the selection range of the power calculation parameters, and the process returns to step 302, and steps 302-304 are re-executed to obtain a new prediction model, thereby implementing the update processing of the prediction model, and performing prediction operations according to the updated prediction model in subsequent steps.

[0114] For example, in one embodiment, the preset number of times is 20 times, then step 306 is executed 20 times to obtain 20 prediction parameters, and then the AUC values ​​corresponding to the 20 prediction parameters are obtained according to step 104, and the prediction parameters corresponding to the AUC values ​​greater than or equal to 80% are used as power calculation parameters. If all 20 AUC values ​​are less than 80%, the 20 prediction parameters are excluded from the selection range of the power calculation parameters.

[0115] Step 404 , when the prediction accuracy is less than the second threshold, the process returns to the step of obtaining the modeling input parameters to establish a new prediction model, that is, returns to step 302 .

[0116] Among them, in step 404, in some embodiments, the AUC value is used as the prediction accuracy. When the AUC value is less than the second threshold, that is, when the AUC value is less than 60%, the prediction parameter corresponding to the AUC value less than 60% is excluded from the selection range of the power calculation parameter, and then the process returns to step 302, and steps 302-304 are re-executed to obtain a new prediction model, thereby realizing the update processing of the prediction model, and performing prediction operations according to the updated prediction model in subsequent steps.

[0117] Through the settings of the above steps 402-404, a prediction model verification and reconstruction mechanism is constructed. When the power consumption environment of the energy router 600 changes significantly, the prediction result obtained by the current prediction model is far different from the actual value of the power demand of the charging device 30. The current prediction model can no longer meet the prediction needs of the energy router 600 after the power consumption environment changes, that is, the current prediction model is no longer applicable. Through the above prediction model verification and reconstruction mechanism, it can be verified whether the prediction model is no longer applicable. If so, re-execute steps 302-304 to obtain a new prediction model that can meet the prediction needs of the energy router 600 after the power consumption environment changes, thereby further improving the stability of the power control of the energy router 600.

[0118] It should be noted that, in order to further avoid the problem of data redundancy in the energy router 600, preferably, in some embodiments, the step of determining whether the prediction parameter is a power calculation parameter according to the comparison result further includes the following data deletion step:

[0119] When the prediction accuracy is less than the second threshold, at least part of the data in the modeling input parameters is deleted according to a preset data cleaning rule.

[0120] In some embodiments, the second threshold is 60%, and the prediction accuracy may be an AUC value. When the AUC value is less than 60%, the above data deletion step is performed.

[0121] Specifically, in the above data deletion step, the data cleaning rule can be pre-set according to at least one parameter of the data ratio and the data acquisition time. The data cleaning rule includes but is not limited to the following methods:

[0122] The first method is to delete data according to a preset ratio. The preset ratio can be set according to actual needs. For example, in some embodiments, the preset ratio is 10%, and 10% of the data in the modeling input parameters is deleted.

[0123] The second method is to delete according to the preset acquisition time. The preset acquisition time can be set according to actual needs. For example, in some embodiments, the preset acquisition time is the previous day, and the modeling input parameters acquired the previous day are deleted. Alternatively, the preset time is all time before the current day, and the modeling input parameters acquired before the current day are deleted, that is, only the modeling input parameters acquired on the current day are retained.

[0124] The third method is to delete according to a preset ratio and a preset acquisition time. The preset ratio and the preset acquisition time can be set according to actual needs. For example, in some embodiments, the preset ratio is 10%, and the preset acquisition time is the earliest time to start acquiring the modeling input parameters, that is, deletion starts from the modeling input parameters acquired at the earliest time until the amount of data deleted from the modeling input parameters reaches 10% of the amount of data of all modeling input parameters before deletion.

[0125] Of course, it should be noted that in the above step 402, if

[0126] After executing step 306 multiple times according to a preset number of times to obtain multiple prediction parameters, and then obtaining the AUC values ​​of the multiple prediction parameters through step 104, if the multiple AUC values ​​are all less than 80%, at least part of the data in the modeling input parameters can be deleted according to the above-mentioned data cleaning rules. The specific steps are shown above and will not be repeated here.

[0127] like Figure 5 , 6 As shown, in one embodiment, the characteristic output power of the energy router 600 includes the maximum output power of the energy router 600; the power control method further includes:

[0128] Step 502: Calculate the power calculation parameter and the first power parameter according to the preset first calculation rule to obtain the maximum output power P of the energy router 600. max .

[0129] The power calculation parameter includes the actual power prediction value P1 of the power supply device 20, and the first power parameter includes the rated power P of the power supply device 20. b 、The maximum input power P of the energy router 600 a and the actual power P2 of the power-consuming device 40. The actual power P2 of the power-consuming device 40 is actually the actual output power of the energy router 600. b , P a The maximum output power P of the energy router 600 is calculated by P2 max .

[0130] like Figure 5 ,6 As shown, in one embodiment, the characteristic output power of the energy router 600 also includes the charging control power, and the characteristic output power of the energy router 600 also includes the charging control power, and the charging control power is used to instruct the charging device 30 to perform charging work. The above power control method also includes:

[0131] Step 504: Calculate the maximum output power, power calculation parameter and second power parameter of the energy router 600 according to the preset second calculation rule to obtain the charging control power P c .

[0132] In step 504, the power calculation parameters include the actual power prediction value P3 of the charging device 30 and the required power prediction value P4 of the charging device 30, and the second power parameter includes the maximum input power P of the energy router 600. a , the actual power P5 of the energy router 600 and the required power P6 of the energy router 600, that is, according to P a , P3, P4, P5 and P6 calculate the maximum output power P of the energy router 600 c .

[0133] Step 506 : Input charging power to the charging device 30 using the charging control power.

[0134] In step 506, the energy router 600 controls the charging power P according to c A power control instruction is sent to the charging device 30 to input charging power to the charging device 30 .

[0135] It should be understood that although Figure 1-5 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-5 At least part of the steps may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0136] See also Figure 6 As shown, in one embodiment, the present invention provides an energy router 600 to which the above power control method can be applied, and the energy router 600 is used to be connected to the power supply device 20 and the charging device 30 respectively.

[0137] The energy router 600 includes a prediction module 610, a control module 620, a power input module 630 and a power output module 640. The control module 620 is connected to the prediction module 610, the power input module 630 and the power output module 640 respectively, and the control module 620 is used to connect to the power supply device 20 and the charging device 30 respectively; the power input module 630 is connected to the power output module 640, the power input module 630 is used to connect to the power supply device 20, and the power output module 640 is used to connect to the charging device 30.

[0138] In some embodiments, the control module 620 can be communicatively connected with the charging device 30, and the control module 620 can receive information reported by the power supply device 20 and each charging device 30, wherein the above information includes but is not limited to parameters used to characterize the operating status of the power supply device 20 and parameters used to characterize the operating status of each charging device 30. At the same time, the control module 620 can also receive parameters used to characterize the operating status of the energy router 600.

[0139] It is worth mentioning that the communication connection method between the control module 620 and the charging device 30 is not limited. For example, in one embodiment, the energy router 600 implements power dispatch management (i.e., distribution and recovery of power supply) based on PLC (Power Line Communication) technology and IoT (Internet of Things) technology, wherein the control module 620 is provided with a PLC communication unit (Power Line Communication), and the control module 620 communicates with each charging device 30 through PLC communication.

[0140] in:

[0141] Prediction module 610, used to obtain prediction parameters;

[0142] The control module 620 is used to receive the prediction parameters of the prediction module 610, perform accuracy calculation on the prediction parameters according to a preset accuracy verification algorithm to obtain the prediction accuracy; compare the prediction accuracy with a preset threshold, and determine whether the prediction parameter is a power calculation parameter according to the comparison result, wherein the power calculation parameter is used to calculate the characteristic output power of the energy router 600, and the characteristic output power is used to instruct the energy router 600 to distribute power to the charging device 30.

[0143] In one embodiment, the preset threshold includes a first threshold; the control module 620 is also used to determine that the prediction parameter is a power calculation parameter when the prediction accuracy is greater than or equal to the first threshold; and / or, when the prediction accuracy is less than the first threshold, re-acquire the prediction parameter through the prediction module 610.

[0144] In one embodiment, the prediction module 610 is also used to obtain modeling input parameters, wherein the modeling input parameters include at least one of a first modeling parameter and a second modeling parameter, the first modeling parameter is a parameter used to characterize the operating state of the power supply device 20, and the second modeling parameter is a parameter used to characterize the operating state of the charging device 30 connected to the energy router 600; a prediction model is established based on the modeling input parameters; and a prediction calculation is performed based on the prediction model to obtain the prediction parameters.

[0145] In one embodiment, the preset threshold includes a first threshold and a second threshold, and the first threshold is greater than the second threshold; the control module 620 is also used to control the prediction module 610 to re-calculate the prediction based on the prediction model to obtain new prediction parameters when the prediction accuracy is less than the first threshold and greater than or equal to the second threshold; and / or, when the prediction accuracy is less than the second threshold, control the prediction module 610 to obtain new modeling input parameters to establish a new prediction model.

[0146] In one embodiment, the control module 620 is further configured to delete at least part of the data in the modeling input parameters according to a preset data cleaning rule when the prediction accuracy is less than a second threshold.

[0147] In one embodiment, the preset threshold includes a first threshold; the prediction parameters include at least two, and correspondingly, the prediction accuracy includes at least two, and the prediction accuracy is set in one-to-one correspondence with the prediction parameters;

[0148] The control module 620 is also used to compare each prediction accuracy with the first threshold value respectively, and judge the prediction parameter that meets the preset condition as the power calculation parameter according to the comparison result, wherein the preset condition is used to characterize that the prediction accuracy corresponding to the prediction parameter is greater than or equal to the first threshold value.

[0149] In one embodiment, the control module 620 is further configured to select a prediction parameter with the highest prediction accuracy from at least two prediction parameters that meet a preset condition and determine it as the optimal power calculation parameter.

[0150] In one of the embodiments, the characteristic output power of the energy router 600 includes the maximum output power of the energy router 600;

[0151] The control module 620 is also used to calculate the power calculation parameter and the first power parameter according to a preset first calculation rule to obtain the maximum output power of the energy router 600, wherein the power calculation parameter includes the actual power prediction value of the power supply device 20, and the first power parameter includes the rated power of the power supply device 20, the maximum input power of the energy router 600, and the actual power of the power-consuming device 40.

[0152] In one embodiment, the characteristic output power of the energy router 600 also includes a charging control power, and the charging control power is used to instruct the charging device 30 to perform charging work;

[0153] The control module 620 is further used to calculate the maximum output power, power calculation parameters and second power parameters of the energy router 600 according to a preset second calculation rule to obtain the charging control power, wherein the power calculation parameters also include the actual power prediction value of the charging device 30 and the required power prediction value of the charging device 30, and the second power parameters include the actual power of the energy router 600 and the required power of the energy router 600; and input the charging control power to the charging device 30.

[0154] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the energy router 600 to which the solution of the present application is applied. The specific energy router 600 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0155] A power control system comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned power control method when executing the computer program.

[0156] A computer-readable storage medium stores a computer program, which implements the steps of the power control method when executed by a processor.

[0157] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0158] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0159] The above-mentioned embodiments only express several embodiments of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A power control method, characterized in that: The method is applied to an energy router, and the energy router is used to connect to a power supply device and / or a charging device. The method includes: Acquire a modeling input parameter; wherein the modeling input parameter includes at least one of a first modeling parameter and a second modeling parameter, the first modeling parameter is a parameter used to characterize the operating state of the power supply device, and the second modeling parameter is a parameter used to characterize the operating state of the charging device connected to the energy router; Establishing a prediction model according to the modeling input parameters; Performing prediction calculation according to the prediction model to obtain prediction parameters; performing accuracy calculation on the prediction parameters according to a preset accuracy verification algorithm to obtain prediction accuracy; The prediction accuracy is compared with a preset threshold, the preset threshold includes a first threshold and a second threshold, the first threshold is greater than the second threshold, and when the prediction accuracy is greater than or equal to the first threshold, the prediction parameter is determined to be a power calculation parameter; wherein the power calculation parameter is used to calculate the characteristic output power of the energy router, and the characteristic output power is used to instruct the energy router to allocate power to the charging device; When the prediction accuracy is less than the second threshold, at least part of the data in the modeling input parameters is deleted according to a preset data cleaning rule, and the process returns to the step of obtaining the modeling input parameters to establish a new prediction model.

2. The power control method according to claim 1, characterized in that: Also includes: When the prediction accuracy is less than the first threshold, the process returns to the step of obtaining prediction parameters.

3. The power control method according to claim 1, characterized in that: The power control method further includes: When the prediction accuracy is less than the first threshold and greater than or equal to the second threshold, the step of returning to the step of performing prediction calculation according to the prediction model to obtain the prediction parameters.

4. The power control method according to claim 1, characterized in that: The method further comprises: In the step of obtaining prediction parameters, the prediction parameters include at least two; In the step of calculating the accuracy of the prediction parameters according to a preset accuracy verification algorithm to obtain the prediction accuracy, the prediction accuracy includes at least two, and the prediction accuracy is set in a one-to-one correspondence with the prediction parameters; The step of comparing the prediction accuracy with a preset threshold and judging whether the prediction parameter is a power calculation parameter according to the comparison result includes: Each of the prediction accuracies is compared with the first threshold value respectively, and the prediction parameter that meets the preset condition is determined as the power calculation parameter according to the comparison result; wherein the preset condition is used to characterize that the prediction accuracy corresponding to the prediction parameter is greater than or equal to the first threshold value.

5. The power control method according to claim 4, characterized in that: The step of determining the prediction parameter that meets the preset condition as the power calculation parameter according to the comparison result includes: From the at least two prediction parameters that meet the preset conditions, the prediction parameter with the highest prediction accuracy is selected to be determined as the optimal power calculation parameter.

6. The power control method according to claim 1, characterized in that: The energy router is used to connect to a power supply device, and the power supply device is used to connect to a power-consuming device; the characteristic output power of the energy router includes the maximum output power of the energy router, and the method further includes: According to a preset first calculation rule, the power calculation parameter and the first power parameter are calculated to obtain the maximum output power of the energy router; wherein the power calculation parameter includes the actual power prediction value of the power supply device, and the first power parameter includes the rated power of the power supply device, the maximum input power of the energy router and the actual power of the power-consuming device.

7. The power control method according to claim 6, characterized in that: The energy router is also used to connect to a charging device; the characteristic output power of the energy router also includes a charging control power, and the charging control power is used to instruct the charging device to perform charging work. The method also includes: According to a preset second calculation rule, the maximum output power of the energy router, the power calculation parameter and the second power parameter are calculated to obtain the charging control power; wherein the power calculation parameter also includes the actual power prediction value of the charging device and the required power prediction value of the charging device, and the second power parameter includes the actual power of the energy router and the required power of the energy router; The charging control power is input to the charging device.

8. An energy router, characterized in that: The energy router is used to connect to a power supply device and / or a charging device, and the energy router includes: A prediction module, used to obtain modeling input parameters; wherein the modeling input parameters include at least one of a first modeling parameter and a second modeling parameter, the first modeling parameter is a parameter used to characterize the operating state of the power supply device, and the second modeling parameter is a parameter used to characterize the operating state of the charging device connected to the energy router; according to the modeling input parameters, a prediction model is established; according to the prediction model, a prediction calculation is performed to obtain the prediction parameters; and, A control module is used to perform accuracy calculation on the prediction parameter according to a preset accuracy verification algorithm to obtain the prediction accuracy; compare the prediction accuracy with a preset threshold, the preset threshold includes a first threshold and a second threshold, the first threshold is greater than the second threshold, when the prediction accuracy is greater than or equal to the first threshold, the prediction parameter is judged to be a power calculation parameter, wherein the power calculation parameter is used to calculate the characteristic output power of the energy router, and the characteristic output power is used to instruct the energy router to distribute power to the charging device; when the prediction accuracy is less than the second threshold, at least part of the data in the modeling input parameter is deleted according to a preset data cleaning rule, and return to the step of obtaining the modeling input parameter to establish a new prediction model.

9. A power control system comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the power control method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the power control method according to any one of claims 1 to 7 are implemented.