Park power load prediction and risk monitoring method and system

Abstract

The invention discloses a park power load prediction and risk monitoring method and system.The method comprises the steps that firstly, historical weather data of a to-be-predicted area is packaged to form a historical weather data set, a load prediction model is trained through the historical load data set, and a corresponding prediction load data set is obtained; substituting the historical weather data set into the predicted load data set to obtain historical load data sets under different weather conditions, and obtaining reference load curves under different weather conditions through the historical load data sets and the predicted load data set under different weather conditions; and judging whether the electric power system of the park has risks or not according to the difference between the prediction result and the reference load curve. According to the method, the severe weather can be monitored, the coverage range of power grid operation risk identification can be effectively expanded, the power load of the park and the enterprise under different weather conditions can be predicted in combination with the weather forecast of the place where the park is located, and the prediction difficulty is effectively reduced.

Description

Technical Field

[0001] The present invention relates to a method for predicting electric load and monitoring risks, and particularly to a method and system for predicting electric load and monitoring risks in a park. Background Art

[0002] The power industry occupies a very important position in the development of society. In the global energy consumption, the proportion of electricity consumption in buildings is increasing. Among them, the electricity consumption in parks accounts for a relatively high proportion. In some parks, there are office areas and production areas. Due to the differences in electricity consumption and electricity usage time between these two areas, the peak electricity consumption period of the park is uncertain and difficult to accurately predict. If the electricity load is too large in a short period of time, it may affect the power supply equipment in the park and may also lead to some potential risks.

[0003] In addition to the above-mentioned situation where excessive load will affect the power supply equipment, bad weather will also cause operation risks. For example, strong winds, heavy rains, heavy snows, etc. will cause damage to the power supply lines and failure of important power supply equipment. In severe cases, it will cause fires and casualties. The existing monitoring and control often first collect the historical load data, historical temperature data and relevant date data of the power grid, and train a long short-term memory neural network through the collected data. Through the long short-term memory neural network, the predicted value of the electric load is predicted.

[0004] Although this method for predicting electric load can work, it still has the following defects:

[0005] 1. Ignoring the monitoring of bad weather, resulting in a reduced coverage of the identification of power grid operation risks.

[0006] 2. Only focusing on protecting the power supply system, without considering the non-power supply equipment in the park that cannot be suddenly powered off or powered off for a long time, which damages the interests of the owners in the park.

[0007] Disclosing the information of this background art section is only intended to increase the understanding of the overall background of the present application, and should not be regarded as an admission or any form of implication that this information constitutes the prior art known to those of ordinary skill in the art. Summary of the Invention

[0008] The object of the present invention is to overcome the defects in the prior art of ignoring the monitoring of bad weather and only protecting the power supply system, and provides a method and system for predicting electric load and monitoring risks in a park that can monitor bad weather and protect non-power supply equipment that cannot be suddenly powered off or powered off for a long time.

[0009] To achieve the above object, the technical solution of the present invention is:

[0010] A method for predicting power load and monitoring risks in a park, the prediction and risk monitoring method comprising the following steps:

[0011] S1. Acquisition and preprocessing of load data, acquiring historical load data of the area to be predicted, and packing the historical load data to form a historical load data set;

[0012] S2. Acquisition and preprocessing of weather data, acquiring historical weather data of the area to be predicted, and packing the historical weather data to form a historical weather data set Z, Z = {Z1, Z2, Z3…Zm}, Zm being the historical weather data of the m-th time unit, and classifying different types of weather in the historical weather data set Z;

[0013] S3. Constructing a load prediction model, training the load prediction model through the historical load data set to obtain a corresponding predicted load data set, substituting the historical weather data set Z into the predicted load data set to respectively obtain a historical load data set under good weather conditions and a historical load data set under bad weather conditions;

[0014] S4. Constructing a curve prediction model, training the curve prediction model through the predicted load data set, the historical load data set under good weather conditions and the historical load data set under bad weather conditions to respectively obtain a reference load curve under good weather conditions and reference load curves under different types of bad weather;

[0015] S5. Selecting an appropriate prediction curve according to the weather conditions of the date to be predicted to predict the power load;

[0016] S6. Judging whether there is a risk in the power system of the park according to the difference between the prediction result and the reference load curve, and giving an early warning when there is a risk.

[0017] In S1, the historical load data of the area to be predicted includes the historical load data of the park and the historical load data of enterprises. The historical load data of the park includes the historical power consumption load data of the whole park, and the historical load data of enterprises includes the historical power consumption load data of each enterprise in the park. The number of enterprises in the park is n;

[0018] Performing data cleaning and variational mode decomposition on the historical load data of the park and the historical load data of enterprises to respectively obtain a historical load data set X of the park and a historical load data set Y of enterprises, X = {Y1, Y2, Y3…Yn}, Yn = {Yn1, Yn2, Yn3…Ynm}, Ynm being the historical power usage data of the n-th enterprise at the m-th time unit;

[0019] The data cleaning includes filling null values in the data and removing data with a load less than a certain threshold within a set time period.

[0020] In S2, a weather classification model is constructed, and different types of weather in the historical weather data set Z are classified by the weather classification model to obtain a good weather data set Zg and a bad weather data set Zb. The bad weather is a weather state that has a significant negative impact on human life and production activities. The weather condition marks include typhoons, heavy rains, blizzards, cold waves, strong winds, sandstorms, high temperatures, droughts, lightning, hail, frost, heavy fog, haze and road icing.

[0021] The S3 includes training the load prediction model through the park historical load data set X and the enterprise historical load data set Y, and solving the load prediction model to obtain the park prediction load data set X' and the enterprise prediction load data set Yn', substituting the good weather data set Zg and the bad weather data set Zb into the park historical load data set X and the enterprise historical load data set Y for analysis, and obtaining the park historical load data set Xg under good weather conditions, the enterprise historical load data set Yg under good weather conditions, the park historical load data set Xb under bad weather conditions, and the enterprise historical load data set Yb under bad weather conditions.

[0022] The S4 includes training the curve prediction model through the park historical load data set Xg under good weather conditions, the enterprise historical load data set Yg under good weather conditions, the park historical load data set Xb under bad weather conditions, and the enterprise historical load data set Yb under bad weather conditions, and respectively obtaining the park benchmark load curve B under good weather conditions, the enterprise benchmark load curve C under good weather conditions, the park benchmark load curve D under different types of bad weather, and the enterprise benchmark load curve E under different types of bad weather;

[0023] The enterprise numbers of high power loads in the enterprise historical load data set Y are marked with enterprise numbers, and the dates of high power loads in the enterprise historical load data set Y are marked with dates;

[0024] A high power load variation curve A is constructed based on the enterprise number mark and date mark.

[0025] In S5, the weather forecast or weather data of the park location is obtained and analyzed, the weather conditions of the date to be predicted are determined by the weather forecast or weather data of the date to be predicted, and the weather conditions of the date to be predicted are marked.

[0026] When the forecast date is under good weather conditions:

[0027] Use the park benchmark load curve B under good weather conditions to predict the park power load;

[0028] When the enterprise has a non-high electricity load, the enterprise benchmark load curve C under good weather conditions is used to predict the power load of the park;

[0029] When the enterprise has a high electricity load, the high electricity load change curve A is substituted into the enterprise benchmark load curve C under good weather conditions to detect whether the prediction is accurate;

[0030] When the date to be predicted is under bad weather conditions, according to the weather condition mark of the date to be predicted, the corresponding park benchmark load curve D and enterprise benchmark load curve E are selected from the enterprise benchmark load curves E under different types of bad weather, and the following operations are carried out:

[0031] Use the corresponding park benchmark load curve D to predict the power load of the park;

[0032] When the enterprise has a non-high electricity load, the corresponding enterprise benchmark load curve E is used to predict the power load of the park;

[0033] When the enterprise has a high electricity load, the high electricity load change curve A is substituted into the corresponding enterprise benchmark load curve E to detect whether the prediction is accurate.

[0034] In step S6, when the difference between the prediction result and the benchmark load curve is greater than or equal to the threshold, there is a risk in the power system of the park;

[0035] If there is a risk, the risk warning will be notified to the management department of the park and the power management department of the location of the park through wired communication or wireless communication for risk early warning;

[0036] The management department of the park will notify the enterprises in the park of the risk early warning through wired communication or wireless communication.

[0037] A park power load prediction and risk monitoring system, which is used to execute the aforementioned park power load prediction and risk monitoring method, specifically includes:

[0038] A load data processing module, a weather data processing module, a load prediction model construction module, a curve prediction model construction module, a power load prediction module and a monitoring module;

[0039] The load data processing module is used to obtain the historical load data of the area to be predicted and pack the historical load data into a historical load data set;

[0040] The weather data processing module is used to obtain the historical weather data of the area to be predicted, and pack the historical weather data to form a historical weather data set Z, Z = {Z1, Z2, Z3…Zm}, where Zm is the historical weather data of the mth time unit, and classify different types of weather in the historical weather data set Z;

[0041] The load prediction model construction module is used to train the load prediction model through the historical load data set to obtain the corresponding predicted load data set, and substitute the historical weather data set Z into the predicted load data set to obtain the historical load data set under good weather conditions and the historical load data set under bad weather conditions respectively;

[0042] The curve prediction model construction module is used to train the curve prediction model through the predicted load data set, the historical load data set under good weather conditions and the historical load data set under bad weather conditions to obtain the reference load curve under good weather conditions and the reference load curves under different types of bad weather respectively;

[0043] The electric load prediction module is used to predict the electric load by selecting an appropriate prediction curve according to the weather conditions of the date to be predicted;

[0044] The monitoring module is used to judge whether there is a risk in the power system of the park according to the difference between the prediction result and the reference load curve, and give an early warning when there is a risk.

[0045] A park electric load prediction and risk monitoring device includes a memory and a processor. The memory is used to store computer program codes and transmit the computer program codes to the processor;

[0046] The processor is used to execute the aforementioned park electric load prediction and risk monitoring method according to the instructions in the computer program codes.

[0047] A computer program product includes a computer program, and the computer program is executed by the processor to perform the aforementioned park electric load prediction and risk monitoring method.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] 1. In a method for predicting power load and monitoring risks in a park in the present invention, a historical weather dataset is formed by packaging historical weather data of the area to be predicted. The load prediction model is trained with the historical load dataset to obtain the corresponding predicted load dataset. The historical weather dataset is substituted into the predicted load dataset to obtain the historical load dataset under good weather conditions and the historical load dataset under bad weather conditions respectively. Then, the baseline load curves under different weather conditions are obtained through the historical load datasets and the predicted load datasets under different weather conditions. Therefore, this design can monitor bad weather and effectively expand the coverage of power grid operation risk identification.

[0050] 2. In a method for predicting power load and monitoring risks in a park in the present invention, by constructing different prediction curves, the power load of the park and the power load of enterprises are predicted, and the enterprises with high electricity consumption are marked for key attention, thereby improving the accuracy of power load prediction for the park and enterprises and reducing the difficulty of prediction. Therefore, this design can mark the enterprises with high electricity consumption for key attention and effectively improve the accuracy of power load prediction for the park and enterprises.

[0051] 3. In a method for predicting power load and monitoring risks in a park in the present invention, the curves obtained by analyzing the weather conditions and power load can be combined with the weather forecast of the location of the park to predict the power load of the park and enterprises under different weather conditions, thereby improving the accuracy of power load prediction for the park and enterprises and reducing the difficulty of prediction. Therefore, this design can combine the weather forecast of the location of the park to predict the power load of the park and enterprises under different weather conditions and effectively reduce the difficulty of prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is the flowchart of the method of the present invention.

[0053] Figure 2 is the structural diagram of the system of the present invention.

[0054] Figure 3 is the structural diagram of the equipment in Embodiment 3. DETAILED DESCRIPTION OF THE INVENTION

[0055] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] Embodiment 1:

[0057] Refer to Figure 1 , a method for predicting power load and monitoring risks in a park. The prediction and risk monitoring method includes the following steps:

[0058] S1. Load data acquisition and preprocessing: Obtain the historical load data of the area to be predicted, and package the historical load data to form a historical load data set.

[0059] S2. Weather data acquisition and preprocessing: Obtain the historical weather data of the area to be predicted, and package the historical weather data to form a historical weather data set Z, Z = {Z1, Z2, Z3…Zm}, where Zm is the historical weather data of the mth time unit, and classify different types of weather in the historical weather data set Z.

[0060] S3. Build a load prediction model: Train the load prediction model with the historical load data set to obtain the corresponding predicted load data set. Substitute the historical weather data set Z into the predicted load data set to obtain the historical load data set under good weather conditions and the historical load data set under bad weather conditions respectively.

[0061] S4. Build a curve prediction model: Train the curve prediction model with the predicted load data set, the historical load data set under good weather conditions, and the historical load data set under bad weather conditions to obtain the reference load curve under good weather conditions and the reference load curves under different types of bad weather respectively.

[0062] S5. Select an appropriate prediction curve according to the weather conditions on the date to be predicted to predict the electric load.

[0063] S6. Judge whether there is a risk in the power system of the park according to the difference between the prediction result and the reference load curve, and give an early warning when there is a risk.

[0064] In S1, the historical load data of the area to be predicted includes the historical load data of the park and the historical load data of enterprises. The historical load data of the park includes the historical electricity consumption load data of the whole park, and the historical load data of enterprises includes the historical electricity consumption load data of each enterprise in the park. The number of enterprises in the park is n.

[0065] Perform data cleaning and variational mode decomposition on the historical load data of the park and the historical load data of enterprises to obtain the historical load data set X of the park and the historical load data set Y of enterprises respectively, X = {Y1, Y2, Y3…Yn}, Yn = {Yn1, Yn2, Yn3…Ynm}, where Ynm is the historical electricity consumption data of the nth enterprise at the mth time unit.

[0066] The data cleaning includes filling the null values in the data and removing the data with a load less than a certain threshold within a set time period.

[0067] In S2, a weather classification model is constructed. Different types of weather in the historical weather dataset Z are classified through the weather classification model to obtain a good weather dataset Zg and a severe weather dataset Zb. The severe weather refers to a weather condition that has a significant negative impact on human life and production activities. The weather condition marks include typhoon, heavy rain, heavy snow, cold wave, strong wind, sandstorm, high temperature, drought, lightning, hail, frost, fog, haze, and road icing.

[0068] When at least one of the following criteria is met, the weather condition is determined to be severe weather:

[0069] 1. The average wind force within 24 hours ≥ 8 levels, or the gust ≥ 10 levels;

[0070] 2. The rainfall within 2 hours ≥ 50 mm, or within 24 hours ≥ 100 mm;

[0071] 3. The snowfall within 12 hours ≥ 6 mm, or the snow depth ≥ 10 cm;

[0072] 4. The temperature drop within 48 hours ≥ 8 °C, and the minimum temperature ≤ 4 °C;

[0073] 5. The average wind force ≥ 6 levels, or the gust ≥ 8 levels;

[0074] 6. A sandstorm occurs, and the visibility < 1000 m, and the wind speed ≥ 3 levels;

[0075] 7. The daily maximum temperature ≥ 35 °C for three consecutive days;

[0076] 8. Moderate drought (CI ≤ -1.5);

[0077] 9. Lightning activities occur;

[0078] 10. Hail with a diameter ≥ 1 cm occurs;

[0079] 11. The ground minimum temperature ≤ 0 °C within 48 hours;

[0080] 12. Fog occurs, and the visibility < 500 m;

[0081] 13. The PM2.5 concentration ≥ 150 μg / m 3 , and the visibility < 3 km;

[0082] 14. The ice thickness of some sections of the road ≥ 2 mm;

[0083] The above-mentioned S3 includes training a load forecasting model with the park historical load dataset X and the enterprise historical load dataset Y, and solving the load forecasting model to obtain the park predicted load dataset X' and the enterprise predicted load dataset Yn'. Substitute the good weather dataset Zg and the bad weather dataset Zb into the park historical load dataset X and the enterprise historical load dataset Y for analysis, and obtain the park historical load dataset Xg under good weather conditions, the enterprise historical load dataset Yg under good weather conditions, the park historical load dataset Xb under bad weather conditions, and the enterprise historical load dataset Yb under bad weather conditions.

[0084] The above-mentioned S4 includes training a curve forecasting model with the park historical load dataset Xg under good weather conditions, the enterprise historical load dataset Yg under good weather conditions, the park historical load dataset Xb under bad weather conditions, and the enterprise historical load dataset Yb under bad weather conditions, and respectively obtaining the park benchmark load curve B under good weather conditions, the enterprise benchmark load curve C under good weather conditions, the park benchmark load curve D under different types of bad weather, and the enterprise benchmark load curve E under different types of bad weather;

[0085] Mark the enterprise numbers of the enterprises with high electricity consumption loads in the enterprise historical load dataset Y, and mark the dates of the high electricity consumption loads in the enterprise historical load dataset Y of the enterprises with high electricity consumption loads;

[0086] Construct a high electricity consumption load change curve A according to the enterprise number mark and the date mark.

[0087] In the above-mentioned S5, obtain and analyze the weather forecast or weather data of the location of the park, judge the weather conditions of the date to be predicted through the weather forecast or weather data on the day of the date to be predicted, and mark the weather conditions of the date to be predicted.

[0088] When the date to be predicted is under good weather conditions:

[0089] Use the park benchmark load curve B under good weather conditions to predict the park power load;

[0090] When the enterprise is under non-high electricity consumption load conditions, use the enterprise benchmark load curve C under good weather conditions to predict the park power load;

[0091] When the enterprise is under high electricity consumption load conditions, substitute the high electricity consumption load change curve A into the enterprise benchmark load curve C under good weather conditions to detect whether the prediction is accurate;

[0092] When the date to be predicted is under bad weather conditions, select the corresponding park baseline load curve D and enterprise baseline load curve E for the day to be predicted from the enterprise baseline load curves E under different types of bad weather according to the weather conditions of the date to be predicted, and perform the following operations:

[0093] Use the corresponding park baseline load curve D to predict the park's power load;

[0094] When the enterprise is under non-high power consumption load conditions, use the corresponding enterprise baseline load curve E to predict the park's power load;

[0095] When the enterprise is under high power consumption load conditions, substitute the high power consumption load change curve A into the corresponding enterprise baseline load curve E to detect whether the prediction is accurate.

[0096] In S6, when the difference between the prediction result and the baseline load curve is greater than or equal to the threshold, there is a risk in the park's power system;

[0097] If there is a risk, notify the park's management department and the local power management department of the park through wired communication or wireless communication for risk early warning;

[0098] The park's management department notifies the enterprises within the park of the risk early warning through wired communication or wireless communication.

[0099] Embodiment 2:

[0100] See Figure 2 , a park power load prediction and risk monitoring system, which is used to execute the park power load prediction and risk monitoring method as described in Embodiment 1, and specifically includes:

[0101] A load data processing module, a weather data processing module, a load prediction model construction module, a curve prediction model construction module, a power load prediction module, and a monitoring module;

[0102] The load data processing module is used to obtain the historical load data of the area to be predicted and package the historical load data into a historical load data set;

[0103] The weather data processing module is used to obtain the historical weather data of the area to be predicted and package the historical weather data into a historical weather data set Z, Z = {Z1, Z2, Z3... Zm}, where Zm is the historical weather data of the mth time unit, and classify different types of weather in the historical weather data set Z;

[0104] The load prediction model construction module is used to train a load prediction model through a historical load data set to obtain a corresponding predicted load data set, and substitute the historical weather data set Z into the predicted load data set to obtain the historical load data set under good weather conditions and the historical load data set under bad weather conditions respectively;

[0105] The curve prediction model construction module is used to train a curve prediction model through the predicted load data set, the historical load data set under good weather conditions and the historical load data set under bad weather conditions to obtain the reference load curve under good weather conditions and the reference load curves under different types of bad weather respectively;

[0106] The electric load prediction module is used to predict the electric load by selecting an appropriate prediction curve according to the weather conditions on the date to be predicted;

[0107] The monitoring module is used to judge whether there is a risk in the power system of the park according to the difference between the prediction result and the reference load curve, and give an early warning when there is a risk.

[0108] Embodiment 3:

[0109] See Figure 3 A power load prediction and risk monitoring device for a park, including a memory and a processor. The memory is used to store computer program codes and transmit the computer program codes to the processor;

[0110] The processor is used to execute the power load prediction and risk monitoring method for the park as described in Embodiment 1 according to the instructions in the computer program codes.

[0111] A computer program product includes a computer program, and the computer program is executed by a processor to perform the power load prediction and risk monitoring method for the park as described in Embodiment 1.

[0112] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. Any equivalent modification or change made by those of ordinary skill in the art according to the disclosure of the present invention shall be included in the protection scope recorded in the claims.

Claims

1. A method for predicting power load and monitoring risks in a park, characterized in that, The prediction and risk monitoring method includes the following steps: S1. Load data acquisition and preprocessing: Acquire the historical load data of the area to be predicted, and package the historical load data to form a historical load data set; S2. Weather data acquisition and preprocessing: Acquire the historical weather data of the area to be predicted, and package the historical weather data to form a historical weather data set Z, Z = {Z1, Z2, Z3…Zm}, where Zm is the historical weather data of the m-th time unit. Classify different types of weather in the historical weather data set Z; S3. Build a load prediction model: Train the load prediction model with the historical load data set to obtain the corresponding predicted load data set. Substitute the historical weather data set Z into the predicted load data set to obtain the historical load data set under good weather conditions and the historical load data set under bad weather conditions respectively; S4. Build a curve prediction model: Train the curve prediction model with the predicted load data set, the historical load data set under good weather conditions and the historical load data set under bad weather conditions to obtain the baseline load curve under good weather conditions and the baseline load curves under different types of bad weather respectively; S5. Select an appropriate prediction curve according to the weather conditions on the date to be predicted to predict the power load; S6. Judge whether there is a risk in the power system of the park according to the difference between the prediction result and the baseline load curve, and give an early warning when there is a risk.

2. The method for predicting power load and monitoring risks in a park according to claim 1, wherein, In S1, the historical load data of the area to be predicted includes the historical load data of the park and the historical load data of the enterprises. The historical load data of the park includes the historical power consumption load data of the whole park, and the historical load data of the enterprises includes the historical power consumption load data of each enterprise in the park. The number of enterprises in the park is n; Perform data cleaning and variational mode decomposition on the historical load data of the park and the historical load data of the enterprises to obtain the historical load data set X of the park and the historical load data set Y of the enterprises respectively. X = {Y1, Y2, Y3…Yn}, Yn = {Yn1, Yn2, Yn3…Ynm}, where Ynm is the historical power consumption data of the n-th enterprise at the m-th time unit; The data cleaning includes filling the null values in the data and removing the data with a load less than a certain threshold within a set time period.

3. A method for predicting power load and monitoring risks in a park according to claim 1, characterized in that, In S2, build a weather classification model. Classify different types of weather in the historical weather data set Z through the weather classification model to obtain a good weather data set Zg and a bad weather data set Zb. The bad weather is a weather state that has a significant negative impact on human life and production activities. The weather condition marks include typhoon, heavy rain, heavy snow, cold wave, strong wind, sandstorm, high temperature, drought, lightning, hail, frost, fog, haze and road icing.

4. The method for predicting power load and monitoring risks in a park according to claim 3, characterized in that, S3 includes training a load forecasting model with the park historical load dataset X and the enterprise historical load dataset Y, solving the load forecasting model to obtain the park predicted load dataset X' and the enterprise predicted load dataset Yn', substituting the good weather dataset Zg and the bad weather dataset Zb into the park historical load dataset X and the enterprise historical load dataset Y for analysis, and obtaining the park historical load dataset Xg under good weather conditions, the enterprise historical load dataset Yg under good weather conditions, the park historical load dataset Xb under bad weather conditions, and the enterprise historical load dataset Yb under bad weather conditions.

5. The method for predicting power load and monitoring risk in a park according to claim 4, characterized in that, S4 includes training a curve prediction model with the park historical load dataset Xg under good weather conditions, the enterprise historical load dataset Yg under good weather conditions, the park historical load dataset Xb under bad weather conditions, and the enterprise historical load dataset Yb under bad weather conditions, and respectively obtaining the park baseline load curve B under good weather conditions, the enterprise baseline load curve C under good weather conditions, the park baseline load curve D under different types of bad weather, and the enterprise baseline load curve E under different types of bad weather; Mark the enterprise numbers of the enterprises with high electricity consumption loads in the enterprise historical load dataset Y, and mark the dates of the high electricity consumption loads in the enterprise historical load dataset Y of the enterprises with high electricity consumption loads; Construct a high electricity consumption load change curve A based on the enterprise number mark and the date mark.

6. A method for predicting power load and monitoring risks in a park according to claim 5, characterized in that, In S5, obtain and analyze the weather forecast or weather data of the location of the park, judge the weather condition of the date to be predicted through the weather forecast or weather data on the day of the date to be predicted, and mark the weather condition of the date to be predicted.

7. A method for predicting power load and monitoring risks in a park according to claim 6, characterized in that, When the date to be predicted is under good weather conditions: Use the park baseline load curve B under good weather conditions to predict the park power load; When the enterprise is under non-high electricity consumption load conditions, use the enterprise baseline load curve C under good weather conditions to predict the park power load; When the enterprise is under high electricity consumption load conditions, substitute the high electricity consumption load change curve A into the enterprise baseline load curve C under good weather conditions to detect whether the prediction is accurate; When the date to be predicted is under bad weather conditions, select the corresponding park baseline load curve D and enterprise baseline load curve E of the date to be predicted from the enterprise baseline load curves E under different types of bad weather according to the weather condition mark of the date to be predicted, and perform the following operations: Use the corresponding park baseline load curve D to predict the park power load; When the enterprise is under non-high electricity consumption load conditions, use the corresponding enterprise baseline load curve E to predict the park power load; When the enterprise is under high electricity consumption load conditions, substitute the high electricity consumption load change curve A into the corresponding enterprise baseline load curve E to detect whether the prediction is accurate.

8. A method for predicting power load and monitoring risks in a park according to claim 7, characterized in that, In S6, when the difference between the prediction result and the baseline load curve is greater than or equal to the threshold, there is a risk in the power system of the park; If there is a risk, the risk warning will be notified to the management department of the park and the power management department of the park's location through wired communication or wireless communication for risk early warning. The management department of the park will notify the enterprises in the park of the risk early warning through wired communication or wireless communication.

9. A power load forecasting and risk monitoring system for a park, characterized in that, The system is used to execute the park power load prediction and risk monitoring method described in any one of claims 1 to 8, specifically including: The system includes a load data processing module, a weather data processing module, a load prediction model construction module, a curve prediction model construction module, a power load prediction module, and a monitoring module; The load data processing module is used to obtain the historical load data of the area to be predicted and package the historical load data to form a historical load data set. The weather data processing module is used to obtain the historical weather data of the area to be predicted and package the historical weather data to form a historical weather data set Z, Z = {Z1, Z2, Z3... Zm}, where Zm is the historical weather data of the mth time unit, and classify different types of weather in the historical weather data set Z. The load prediction model construction module is used to train the load prediction model through the historical load data set to obtain the corresponding predicted load data set, and substitute the historical weather data set Z into the predicted load data set to obtain the historical load data set under good weather conditions and the historical load data set under bad weather conditions respectively. The curve prediction model construction module is used to train the curve prediction model through the predicted load data set, the historical load data set under good weather conditions, and the historical load data set under bad weather conditions to obtain the reference load curve under good weather conditions and the reference load curves under different types of bad weather respectively. The power load prediction module is used to select an appropriate prediction curve according to the weather conditions of the date to be predicted to predict the power load. The monitoring module is used to judge whether there is a risk in the power system of the park according to the difference between the prediction result and the reference load curve, and give an early warning when there is a risk.

10. A power load forecasting and risk monitoring device for a park, characterized in that, It includes a memory and a processor. The memory is used to store computer program code and transmit the computer program code to the processor; The processor is used to execute the park power load prediction and risk monitoring method described in any one of claims 1 to 8 according to the instructions in the computer program code.

11. A computer program product, comprising a computer program, characterized in that, The computer program is executed by the processor to perform the park power load prediction and risk monitoring method described in any one of claims 1 to 8.

Images