Public building two-stage energy management method considering new energy uncertainty

Abstract

The invention provides a public building two-stage energy management method considering new energy uncertainty. The management method specifically comprises the following steps: constructing a day-ahead load prediction curve according to historical load data; setting real-time adjustment periods, and collecting real-time load data of the current real-time adjustment period when each real-time adjustment period is ended; comparing the current load prediction curve with a day-ahead load prediction curve to obtain a load deviation value; operation information of all loads in the public building energy supply system is obtained, and an adjustable resource list is determined; adjusting objects are selected from the adjustable resource list, the load adjusting quantity of each adjusting object is determined according to the load deviation value, and the adjusting objects participate in load adjustment of the next real-time adjusting period according to the corresponding load adjusting quantity. Load deviation caused by new energy power generation output fluctuation and the like can be compensated through real-time load adjustment, the problem of energy supply and demand imbalance is avoided, and operation safety of a public building energy supply system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of public building energy management, and in particular to a two-stage energy management method for public buildings taking into account the uncertainty of new energy sources. Background Art

[0002] Unlike traditional public building loads, the public building energy supply system is equipped with user-side new energy power generation equipment such as household rooftop photovoltaic panels and small wind turbines. Through the active participation of user-side new energy power generation equipment in energy management, two-way flow of energy can be achieved. It can not only obtain energy from the main grid to meet its own needs, but also sign an agreement with the electricity market to sell energy to the electricity market, thereby improving the energy utilization rate of the public building energy supply system and reducing the pressure of energy supply and demand.

[0003] However, when considering energy management of user-side new energy power generation equipment, the existing public building energy supply system mostly conducts energy distribution planning through load forecasting and other methods, thereby scheduling the demand-side load and the output of new energy power generation equipment. However, the output of new energy power generation equipment is greatly affected by environmental factors, and the output prediction accuracy of new energy power generation equipment is limited. When the actual scheduling of demand-side load and new energy power generation equipment output is carried out according to the established plan, it is easy to cause an imbalance in the supply and demand of the public building energy supply system due to the output fluctuation of new energy power generation, thereby affecting the operation safety of the public building energy supply system. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art that when using load forecasting to carry out energy planning of public building energy supply systems, the output prediction accuracy of new energy power generation equipment is limited. When actually implementing energy planning, it is easy to cause energy supply and demand imbalance problems in public building energy supply systems due to output fluctuations of new energy power generation. A two-stage energy management method for public buildings taking into account the uncertainty of new energy is provided, and the energy management of public buildings is divided into two stages: day-ahead prediction and real-time feedback adjustment. The energy management plan of the public building energy supply system is obtained through day-ahead load forecasting. In the process of executing energy management according to the day-ahead load forecasting result, the load deviation value caused by output fluctuations of new energy power generation during the actual implementation process is determined according to real-time load data, and specific adjustment objects are screened out for load adjustment to correct the load deviation value, avoid energy supply and demand imbalance problems in the public building energy supply system, and ensure the operation safety of the public building energy supply system.

[0005] The purpose of the present invention is achieved through the following technical solutions:

[0006] A two-stage energy management approach for public buildings considering the uncertainty of new energy sources, including:

[0007] Obtain historical load data of the energy supply system of public buildings, make a day-ahead load forecast based on the historical load data, and obtain a day-ahead load forecast curve;

[0008] Set a real-time adjustment cycle. At the end of each real-time adjustment cycle, collect the real-time load data of the public building system in the current real-time adjustment cycle and construct a corresponding real-time load curve;

[0009] Compare the real-time load curve with the day-ahead load forecast curve to obtain the load deviation value;

[0010] Obtain the operating information of all loads in the energy supply system of public buildings and determine the list of adjustable resources;

[0011] An adjustment object is selected from the adjustable resource list, and the load adjustment amount of each adjustment object is determined according to the load deviation value. The adjustment object participates in the load adjustment of the next real-time adjustment cycle according to the corresponding load adjustment amount.

[0012] Furthermore, the real-time load curve is compared with the day-ahead load forecast curve to obtain the load deviation value, including:

[0013] Determine the real-time regulation period to which the real-time load curve belongs;

[0014] Determine the time period according to the real-time adjustment cycle, and obtain the load sub-curve of the day-ahead load forecast curve within the corresponding time period;

[0015] Compare the real-time load curve with the load sub-curve to obtain the deviation value at each time point;

[0016] The deviation values ​​at each time point are summed to obtain the load deviation value.

[0017] Furthermore, the acquisition of the operating information of all loads in the public building energy supply system and the determination of the list of adjustable resources include:

[0018] Obtain the operating information of all loads in the public building energy supply system and filter out the loads that are currently in an idle state;

[0019] Obtain the load sub-curve of the day-ahead load forecast curve in the corresponding time period of the real-time regulation cycle, as well as the regulation objects and corresponding load regulation amounts determined in the previous real-time regulation cycle, and filter out loads with regulation margins in combination with the load operation information;

[0020] A list of adjustable resources is created based on the loads that are in an idle state and the loads that have adjustment margin.

[0021] Furthermore, the step of establishing an adjustable resource list according to the loads in an idle state and the loads with an adjustment margin includes:

[0022] Get the adjustable amount and load type of each load in idle state;

[0023] The loads in the idle state are divided into adjustment priorities according to the load type and the adjustable amount, and the loads in the idle state are sorted according to the corresponding adjustment priorities;

[0024] Obtaining the adjustable amount corresponding to each load having an adjustment margin;

[0025] After the loads in the idle state are sorted, the loads with adjustment margins are sorted according to the corresponding adjustable amounts to obtain an adjustable resource list.

[0026] Furthermore, the method selects an adjustable object from the adjustable resource list, and determines the load adjustment amount of each adjustable object according to the load deviation value. The adjustable object participates in the load adjustment of the next real-time adjustment cycle according to the corresponding load adjustment amount, including summing the adjustable amount of each load in the adjustable resource table in order of arrangement, and taking the load corresponding to the load when the sum reaches the load adjustment amount as the adjustment object, allocating the load deviation value according to the adjustable amount of each adjustable object, and determining the load adjustment amount of each adjustable object.

[0027] Furthermore, the step of comparing the real-time load curve with the day-ahead load forecast curve to obtain the load deviation value also includes:

[0028] Determine whether the next real-time adjustment cycle is the last real-time adjustment cycle;

[0029] When the next real-time regulation cycle is the last real-time regulation cycle, the real-time load curves in all completed real-time regulation cycles are retrieved;

[0030] According to all real-time load curves and the load regulation amount of each completed real-time regulation cycle, the actual load curve of the public building system on the day is constructed;

[0031] Predict the load deviation value of the last real-time adjustment cycle according to the actual load curve of the day to obtain the load deviation prediction value;

[0032] The load deviation value is corrected according to the load deviation prediction amount.

[0033] Furthermore, the acquisition of historical load data of the public building energy supply system, performing a day-ahead load forecast based on the historical load data, and obtaining a day-ahead load forecast curve includes:

[0034] Obtain electricity price information from the electricity market and build a transaction power decision model for the public building energy supply system based on historical load data and electricity price information;

[0035] Determine the energy trading strategy for the public building energy supply system to trade electricity with the superior energy management center based on the trading electricity decision model;

[0036] Construct the day-ahead load forecast curve based on the energy trading strategy.

[0037] Furthermore, the expression of the transaction power decision model is:

[0038]

[0039] Among them, F1 is the total economic cost, F e is the electricity cost, F t is the dissatisfaction cost, a and b are the importance coefficients of electricity cost and dissatisfaction cost respectively. is the electricity purchase price at time t, is the electricity price at time t, P buy (t) is the amount of electricity purchased at time t, P sell (t) is the electricity sales at time t, ΔT is the time period for making transaction electricity decisions, and f DL (t) is the dissatisfaction of the demand side load at time t.

[0040] The beneficial effects of the present invention are:

[0041] (1) The energy management of public buildings is divided into two stages: day-ahead prediction and real-time feedback regulation. The energy management plan of the public building energy supply system is obtained through day-ahead load prediction. In the process of executing energy management according to the day-ahead load prediction results, the load deviation value caused by the fluctuation of renewable energy power generation output in the actual implementation process is determined based on the real-time load data. Specific regulation objects are selected for load regulation to correct the load deviation value, avoid the imbalance of energy supply and demand in the public building energy supply system, and ensure the safe operation of the public building energy supply system.

[0042] (2) Before obtaining the load deviation value, a day-ahead load prediction curve corresponding to the day-ahead load prediction result and a real-time load curve corresponding to the real-time load data are constructed respectively. By comparing the curves, the deviation situation can be intuitively displayed, thereby accurately obtaining the load deviation value.

[0043] (3) A list of adjustable resources can be prepared according to the operating status of the load and the specific adjustable quantity. When adjusting the load, the adjustment object can be selected based on the list of adjustable resources to ensure the efficiency of load adjustment.

[0044] (4) Since the load deviation caused by the last real-time adjustment cycle cannot be adjusted by the next real-time adjustment cycle after the end of the last real-time adjustment cycle, the load deviation value of the last real-time adjustment cycle is predicted according to the current load operation status, and the load deviation value that may exist in the last real-time adjustment cycle is added to the load deviation value calculated in the previous real-time adjustment cycle to further avoid the occurrence of load deviation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of a process of the present invention. DETAILED DESCRIPTION

[0046] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0047] Example:

[0048] A two-stage energy management approach for public buildings considering the uncertainty of new energy sources, such as Figure 1 As shown, including,

[0049] Obtain historical load data of the energy supply system of public buildings, make a day-ahead load forecast based on the historical load data, and obtain a day-ahead load forecast curve;

[0050] Set a real-time adjustment cycle. At the end of each real-time adjustment cycle, collect the real-time load data of the public building system in the current real-time adjustment cycle and construct a corresponding real-time load curve;

[0051] Compare the real-time load curve with the day-ahead load forecast curve to obtain the load deviation value;

[0052] Obtain the operating information of all loads in the energy supply system of public buildings and determine the list of adjustable resources;

[0053] An adjustment object is selected from the adjustable resource list, and the load adjustment amount of each adjustment object is determined according to the load deviation value. The adjustment object participates in the load adjustment of the next real-time adjustment cycle according to the corresponding load adjustment amount.

[0054] In addition to new energy power generation equipment, the energy supply system of public buildings is also equipped with demand-side loads such as adjustable loads and uncontrollable loads, as well as backup power sources such as energy storage equipment and backup generators.

[0055] The energy supply system of public buildings can supply power to other loads through the output of its own renewable energy power generation equipment, and can sell the output of renewable energy power generation equipment to the main power grid.

[0056] After completing the transaction with the superior electricity market, the public building energy supply system needs to provide the superior electricity market with the load specified in the transaction. If the established energy trading target is not achieved, it may cause an imbalance in the supply and demand of the power system, thereby affecting the stable operation of the power system. The public building energy supply system will also receive corresponding punishment measures.

[0057] Therefore, a day-ahead load forecast is conducted based on the historical load data of the public building energy supply system to determine how to regulate and distribute the demand-side load of the public building energy supply system and the output of new energy power generation equipment, so as to meet the normal operating requirements of the demand-side load of the public building energy supply system while conducting electricity transactions with the superior energy management center with maximum benefits.

[0058] Specifically, the acquiring of historical load data of the public building energy supply system, performing a day-ahead load forecast based on the historical load data, and acquiring a day-ahead load forecast curve includes:

[0059] Obtain electricity price information from the electricity market and build a transaction power decision model for the public building energy supply system based on historical load data and electricity price information;

[0060] Determine the energy trading strategy for the public building energy supply system to trade electricity with the superior energy management center based on the trading electricity decision model;

[0061] Construct the day-ahead load forecast curve based on the energy trading strategy.

[0062] Among them, the day-ahead load forecast curve is based on the data of electricity trading. It is constructed according to the electricity allocated to electricity trading in the output of new energy power generation equipment. The day-ahead load forecast curve can reflect the amount of electricity trading between the public building energy supply system and the superior energy management center within the corresponding time period.

[0063] In order to avoid the impact of electricity trading on the normal operation of other loads in the public building energy supply system, when formulating the trading electricity, in addition to considering the output of new energy power generation equipment, it is also necessary to consider the operation of demand-side loads.

[0064] Therefore, electricity price information is obtained from the electricity market, and the output of new energy power generation equipment and the load demand of the demand side are predicted based on historical load data. On the basis of known output and load demand, a trading electricity decision model is constructed in combination with electricity price information to determine the energy trading strategy for electricity trading with the superior energy management center.

[0065] The energy trading strategy described in this embodiment includes trading electricity and trading time.

[0066] Among them, meteorological forecast data, such as wind speed forecast, solar irradiance forecast, etc., can be used as the basis for predicting the output of new energy power generation equipment. Then, by analyzing the historical power generation data of new energy power generation equipment in the historical load data, the output law of new energy power generation equipment can be determined, such as seasonal changes, daily changes, etc. Finally, combined with machine learning algorithms, the output of new energy power generation equipment can be predicted.

[0067] Similarly, the demand-side load demand can be predicted by analyzing historical load data to obtain the seasonality, periodicity, trend and other characteristics of the demand-side load, and using statistical methods or machine learning algorithms.

[0068] The expression of the transaction power decision model in this embodiment is:

[0069]

[0070] Among them, F1 is the total economic cost, F e is the electricity cost, F t is the dissatisfaction cost, a and b are the importance coefficients of electricity cost and dissatisfaction cost respectively. is the electricity purchase price at time t, is the electricity price at time t, P buy (t) is the amount of electricity purchased at time t, P sell (t) is the electricity sales at time t, ΔT is the time period for making transaction electricity decisions, and f DL (t) is the dissatisfaction of the demand side load at time t.

[0071] Since electricity trading of the output of new energy power generation equipment may affect the supply of demand-side loads, the dissatisfaction on the user side is bound to increase. Therefore, the dissatisfaction is used to measure the impact on the normal operation of the demand-side load, so as to determine how to trade electricity with the superior energy management center with higher profits while meeting the normal operation requirements of the demand-side load of the public building energy supply system.

[0072] The control strategy for demand-side load and the output allocation plan of new energy power generation equipment are both based on the output forecast results of new energy power generation equipment and the demand-side load demand forecast results. However, the output forecast results of new energy power generation equipment rely on data such as weather forecasts, whose accuracy cannot be estimated, which can easily lead to a large deviation between the actual output and the expected output.

[0073] When the output deviation of new energy power generation equipment is too large, in order to meet the power transaction, it is necessary to adjust the output distribution of new energy power generation equipment, which will cause the supply and demand imbalance of public building energy supply system. In order to avoid this problem, for the existing output deviation, under the premise of not changing the output distribution of new energy power generation equipment, the load deviation value caused by the output deviation is compensated by adjusting other loads.

[0074] In order to ensure the accuracy and timeliness of real-time regulation, a real-time regulation cycle is set, and the real-time regulation cycle is used as the feedback node for load regulation. The real-time load curve of each real-time regulation cycle is compared with the day-ahead load forecast curve to obtain the load deviation value, so as to perform specific load regulation according to the corresponding load deviation value in the next real-time regulation cycle.

[0075] Specifically, the real-time load curve is compared with the day-ahead load forecast curve to obtain the load deviation value, including:

[0076] Determine the real-time regulation period to which the real-time load curve belongs;

[0077] Determine the time period according to the real-time adjustment cycle, and obtain the load sub-curve of the day-ahead load forecast curve within the corresponding time period;

[0078] Compare the real-time load curve with the load sub-curve to obtain the deviation value at each time point;

[0079] The deviation values ​​at each time point are summed to obtain the load deviation value.

[0080] Since the day-ahead load forecast curve is established based on the output of new energy power generation equipment, in order to determine the accurate load deviation, a real-time load curve is constructed based on the actual output of new energy power generation equipment, thereby determining the load deviation value caused by the output fluctuation of new energy power generation equipment.

[0081] Moreover, the output of renewable energy power generation equipment is affected by environmental factors. There may be two situations: the output is greater than expected at a certain point in time, or the output is lower than expected at a certain point in time. In order to accurately obtain the load deviation value that needs to be compensated, a time period is used as the adjustment cycle, and the accumulated load deviation value in the time period is determined, and then the load deviation of this time period is compensated and adjusted in the next time period.

[0082] Because within a time period, the load deviation value may be positive or negative depending on whether the output is greater than expected or the output is less than expected at a certain point in time, after summing them up, it is possible to determine the load deviation value that is outside the adjustable range of the new energy power generation equipment itself and requires other equipment to assist in load regulation.

[0083] The obtaining of the operating information of all loads in the public building energy supply system and determining the list of adjustable resources includes:

[0084] Obtain the operating information of all loads in the public building energy supply system and filter out the loads that are currently in an idle state;

[0085] Obtain the load sub-curve of the day-ahead load forecast curve in the corresponding time period of the real-time regulation cycle, as well as the regulation objects and corresponding load regulation amounts determined in the previous real-time regulation cycle, and filter out loads with regulation margins in combination with the load operation information;

[0086] A list of adjustable resources is created based on the loads that are in an idle state and the loads that have adjustment margin.

[0087] In order to compensate for the load deviation value caused by the output fluctuation of new energy power generation equipment, other equipment needs to intervene to absorb the output or compensate for the electricity. Different types of loads have different load regulation efficiencies when participating in load regulation. In order to improve the load regulation efficiency, a corresponding adjustable resource list is established according to the loads that can participate in the load regulation, and the adjustment object for load regulation is directly selected according to the adjustable resource list.

[0088] Among them, the loads that can participate in load regulation include reducible loads, transferable loads, energy storage equipment, etc. For such loads in idle state, they can participate in load regulation by connecting to the public building energy supply system. Even if these loads have been connected to the public building energy supply system, they can participate in load regulation by changing their response strategies.

[0089] Therefore, the loads that are currently in an idle state and the loads that have participated in load regulation and have a regulation margin are selected as candidates for participating in load regulation in the next real-time regulation cycle.

[0090] The step of establishing an adjustable resource list according to the loads in an idle state and the loads with an adjustment margin includes:

[0091] Get the adjustable amount and load type of each load in idle state;

[0092] The loads in the idle state are divided into adjustment priorities according to the load type and the adjustable amount, and the loads in the idle state are sorted according to the corresponding adjustment priorities;

[0093] Obtaining the adjustable amount corresponding to each load having an adjustment margin;

[0094] After the loads in the idle state are sorted, the loads with adjustment margins are sorted according to the corresponding adjustable amounts to obtain an adjustable resource list.

[0095] Since different types of loads have different costs and regulation efficiencies when participating in load regulation, and the way of connecting idle loads to the public building energy supply system is more efficient than replanning the control strategy for the participating loads, when establishing a list of adjustable resources, idle loads are placed at the front.

[0096] The load in the idle state has different priorities in participating in load regulation depending on its load type and adjustable quantity.

[0097] For loads with higher importance and smaller adjustable amounts, their priority in participating in load regulation should be lowered, while for loads with lower importance and larger adjustable amounts, their priority in participating in load regulation needs to be increased.

[0098] Therefore, the importance of the load in the operation of the public building energy supply system is determined according to the load type, and the adjustment priority is determined in combination with its specific adjustable quantity. Specifically, the load adjustment priority can be divided by weighted summation and sorting.

[0099] As for loads with adjustment margin, since they have already participated in load regulation, the impact of importance is relatively low. Therefore, they are directly sorted according to the adjustable amount, and loads with more adjustable amounts are adjusted first to reduce the amount of loads participating in load regulation and reduce the adjustment cost.

[0100] The method selects an adjustment object from the adjustable resource list, and determines the load adjustment amount of each adjustment object according to the load deviation value. The adjustment object participates in the load adjustment of the next real-time adjustment cycle according to the corresponding load adjustment amount, including summing the adjustable amount of each load in the adjustable resource table in order of arrangement, and taking the load corresponding to the load when the summation result reaches the load adjustment amount as the adjustment object, allocating the load deviation value according to the adjustable amount of each adjustment object, and determining the load adjustment amount of each adjustment object.

[0101] Priority is given to the adjustment objects with higher rankings. After the load deviation value is met, the adjustable amount of each load is used as the corresponding adjustment amount. If the adjustment amount of the last adjustment object is greater than the remaining adjustment amount of the load deviation value, the remaining adjustment amount according to the load deviation value is used as its adjustment amount.

[0102] Moreover, the operation time of each load is not certain, and the adjustable resources available in each real-time adjustment cycle are not consistent. Therefore, a corresponding adjustable resource list is constructed in each real-time adjustment cycle to ensure load adjustment efficiency.

[0103] The step of comparing the real-time load curve with the day-ahead load forecast curve to obtain the load deviation value also includes:

[0104] Determine whether the next real-time adjustment cycle is the last real-time adjustment cycle;

[0105] When the next real-time regulation cycle is the last real-time regulation cycle, the real-time load curves in all completed real-time regulation cycles are retrieved;

[0106] According to all real-time load curves and the load regulation amount of each completed real-time regulation cycle, the actual load curve of the public building system on the day is constructed;

[0107] Predict the load deviation value of the last real-time adjustment cycle according to the actual load curve of the day to obtain the load deviation prediction value;

[0108] The load deviation value is corrected according to the load deviation prediction amount.

[0109] The real-time load curve reflects the actual output of new energy power generation equipment, while the load regulation reflects the adjustment of the load in order to maintain the balance between supply and demand. By matching and combining the two, the actual amount of electricity traded after adjustment can be determined.

[0110] The trend and periodic changes of the actual load curve of the day are judged. Through time series analysis, machine learning models and other methods, the actual load curve of the day in the last real-time adjustment cycle can be predicted, and then it is compared with the sub-curve corresponding to the last real-time adjustment cycle in the day-ahead load forecast curve to determine the load deviation prediction value.

[0111] After the last real-time adjustment cycle ends, the load deviation caused by it cannot be adjusted by the next real-time adjustment cycle. Therefore, the load deviation value of the last real-time adjustment cycle is predicted according to the current load operation status, and the predicted load deviation that may exist in the last real-time adjustment cycle is added to the load deviation value calculated in the previous real-time adjustment cycle to correct the load deviation value. When the load adjustment plan is carried out subsequently, the corrected load deviation value is used as the adjustment data basis to further avoid the occurrence of load deviation.

[0112] The above-described embodiment is only a preferred solution of the present invention and does not limit the present invention in any form. There are other variations and modifications without exceeding the technical solution described in the claims.

Claims

1. A two-stage energy management method for public buildings taking into account the uncertainty of new energy sources, characterized in that: Including, obtaining historical load data of the public building energy supply system, making a day-ahead load forecast based on the historical load data, and obtaining a day-ahead load forecast curve; Set a real-time adjustment cycle. At the end of each real-time adjustment cycle, collect the real-time load data of the public building system in the current real-time adjustment cycle and construct a corresponding real-time load curve; Compare the real-time load curve with the day-ahead load forecast curve to obtain the load deviation value; Obtain the operating information of all loads in the energy supply system of public buildings and determine the list of adjustable resources; An adjustment object is selected from the adjustable resource list, and the load adjustment amount of each adjustment object is determined according to the load deviation value. The adjustment object participates in the load adjustment of the next real-time adjustment cycle according to the corresponding load adjustment amount.

2. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 1 is characterized in that: The real-time load curve is compared with the day-ahead load forecast curve to obtain the load deviation value, including: Determine the real-time regulation period to which the real-time load curve belongs; Determine the time period according to the real-time adjustment cycle, and obtain the load sub-curve of the day-ahead load forecast curve within the corresponding time period; Compare the real-time load curve with the load sub-curve to obtain the deviation value at each time point; The deviation values ​​at each time point are summed to obtain the load deviation value.

3. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 1 is characterized in that: The operation information of all loads in the energy supply system of the public building is obtained to determine the list of adjustable resources. include, Obtain the operating information of all loads in the public building energy supply system and filter out the loads that are currently in an idle state; Obtain the load sub-curve of the day-ahead load forecast curve in the corresponding time period of the real-time regulation cycle, as well as the regulation objects and corresponding load regulation amounts determined in the previous real-time regulation cycle, and filter out loads with regulation margins in combination with the load operation information; A list of adjustable resources is created based on the loads that are in an idle state and the loads that have adjustment margin.

4. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 3 is characterized in that: The adjustable resource list is established according to the loads in the idle state and the loads with adjustment margin, include, Get the adjustable amount and load type of each load in idle state; The loads in the idle state are divided into adjustment priorities according to the load type and the adjustable amount, and the loads in the idle state are sorted according to the corresponding adjustment priorities; Obtaining the adjustable amount corresponding to each load having an adjustment margin; After the loads in the idle state are sorted, the loads with adjustment margins are sorted according to the corresponding adjustable amounts to obtain an adjustable resource list.

5. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 4 is characterized in that: The method selects an adjustment object from the adjustable resource list, and determines the load adjustment amount of each adjustment object according to the load deviation value. The adjustment object participates in the load adjustment of the next real-time adjustment cycle according to the corresponding load adjustment amount, including summing the adjustable amount of each load in the adjustable resource table in order of arrangement, and taking the load corresponding to the load when the summation result reaches the load adjustment amount as the adjustment object, allocating the load deviation value according to the adjustable amount of each adjustment object, and determining the load adjustment amount of each adjustment object.

6. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 1 is characterized in that: The step of comparing the real-time load curve with the day-ahead load forecast curve to obtain the load deviation value also includes: Determine whether the next real-time adjustment cycle is the last real-time adjustment cycle; When the next real-time regulation cycle is the last real-time regulation cycle, the real-time load curves in all completed real-time regulation cycles are retrieved; According to all real-time load curves and the load regulation amount of each completed real-time regulation cycle, the actual load curve of the public building system on the day is constructed; Predict the load deviation value of the last real-time adjustment cycle according to the actual load curve of the day to obtain the load deviation prediction value; The load deviation value is corrected according to the load deviation prediction amount.

7. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 1 is characterized in that: The method of obtaining the historical load data of the public building energy supply system, performing a day-ahead load forecast based on the historical load data, and obtaining a day-ahead load forecast curve includes: Obtain electricity price information from the electricity market and build a transaction power decision model for the public building energy supply system based on historical load data and electricity price information; Determine the energy trading strategy for the public building energy supply system to trade electricity with the superior energy management center based on the trading electricity decision model; Construct the day-ahead load forecast curve based on the energy trading strategy.

8. The two-stage energy management method for public buildings taking into account the uncertainty of new energy sources according to claim 7 is characterized in that: The expression of the transaction power decision model is: Among them, F1 is the total economic cost, F e is the electricity cost, F t is the dissatisfaction cost, a and b are the importance coefficients of electricity cost and dissatisfaction cost respectively. is the electricity purchase price at time t, is the electricity price at time t, P buy (t) is the amount of electricity purchased at time t, P sell (t) is the electricity sales at time t, ΔT is the time period for making transaction electricity decisions, and f DL (t) is the dissatisfaction of the demand side load at time t.

Images