Intelligent building point-to-point optimization scheduling method considering uncertainty transaction
By introducing predictive deviation modeling and flexible load adjustment of wind and light resources in smart buildings, a collaborative P2P trading market is designed, which solves the problem of low electricity trading efficiency caused by uncertainty of wind and light resources, and improves market operation efficiency and wind and light resources Utilization rate achieves active user participation and maximizes social benefits.
Patent Information
- Application Number
- CN202410979886.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-08
AI Technical Summary
Traditional energy scheduling solutions based on deterministic scenarios are difficult to adapt to the complex, multi-dimensional and uncertain power market operation after large-scale wind and light resources access, resulting in insufficient economic and flexibility of power transactions and lack of fully distributed P2P transactions. Centralized scheduling and regulatory agencies limit the scale and transparency of market transactions.
An intelligent building optimization scheduling method based on collaborative P2P transaction is proposed. By introducing the predicted deviation probability density function of the wind and light resources, the uncertainty of the wind and light resources is modeled, and the flexible load is divided into time flexibility and power flexibility Loads, by adjusting the capacity of these loads, absorbing the uncertain output of wind and light resources, designing a combined trading market for electricity-regulating capacity, improving market operation efficiency and utilization of wind and light resources.
It effectively improves the power generation rate of wind and light resources, eliminates the prediction deviation of wind and light resources, promotes users to participate in P2P transactions, maximizes social benefits, and reduces the occurrence of wind and light abandonment.
Smart Images

Figure CN118917603A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions. Background Art
[0002] In recent years, the in-depth reform of the electricity market has made market transactions more open and transparent, and has derived a variety of new interest entities. As the main body of urban energy consumption, smart buildings not only deploy a large number of demand-side loads, but also contain a large number of wind and photovoltaic (WPV) resources. Traditional energy scheduling solutions based on deterministic scenarios are no longer suitable for the complex, multi-dimensional, and uncertain new forms of electricity market operation generated after the access of large-scale WPV resources. Therefore, how to design an effective power scheduling solution to improve the economy and flexibility of power trading in buildings has become a problem that needs to be solved urgently.
[0003] After connecting to a large number of WPV resources, smart buildings have the ability to produce and consume electricity, allowing electricity trading to take place inside the building. Among them, peer-to-peer (P2P) electricity trading has become a current research hotspot because it can effectively reduce users' electricity costs. P2P electricity trading can be divided into the following two types: 1) Fully distributed P2P trading: The electricity trading between market users is completely determined by the buyer and seller, without the participation of a third-party organization. The literature establishes a fully distributed P2P optimization scheduling model that considers building characteristics and electricity trading, aiming to optimize the economic efficiency of the building while protecting the privacy information of users during the electricity trading process. 2) Collaborative P2P trading: A third-party organization is set up in the market to collect user information and is responsible for building a matching relationship between the two parties. However, due to the lack of centralized scheduling and regulatory agencies, fully distributed P2P electricity trading has limited market trading scale, reduced market transparency and trading efficiency, which in turn limits its application and development in large-scale electricity trading markets. Therefore, the market still needs a third-party management agency to integrate market trading data and optimize trading plans.
[0004] Considering that the high proportion of WPV in smart buildings has strong volatility, its actual power generation often deviates from the expected power generation. When the actual power generation of WPV is greater than the expected power generation, it will cause an increase in wind and solar power abandonment; when the actual power generation of WPV is less than the expected power generation, other energy supply equipment is required to make up for the power shortage in time. Therefore, modeling and analyzing the uncertainty of WPV output is crucial to improving the reliability of transactions. Existing studies mainly use probability distribution functions or fluctuation ranges to model the uncertainty of WPV output. Common methods include stochastic programming (SP), robust optimization (RO), and distributionally robust optimization (DRO). Based on the above modeling methods for the uncertainty of WPV output, relevant scholars have proposed to adjust the power consumption of flexible loads in smart buildings to achieve peak shaving and valley filling of WPV power generation. However, in the proposed flexible load optimization configuration scheme, the flexible loads in smart buildings need to be able to respond quickly and eliminate the forecast deviation when WPV has a forecast deviation, which requires the market to have high communication capabilities and fast response speed. Summary of the invention
[0005] In order to overcome the above-mentioned deficiencies in the prior art, the present invention proposes a point-to-point optimization scheduling method for intelligent buildings taking into account uncertain transactions.
[0006] This paper takes WPV resources and flexible loads in smart buildings as analysis objects, and proposes an intelligent building optimization scheduling strategy based on collaborative P2P transactions. First, the probability density function of WPV prediction deviation is introduced to model the uncertain output of WPV. Then, the flexible loads in the smart building are modeled as time flexibility (TF) loads and power flexibility (PF) loads, and the uncertain output of WPV is absorbed by scheduling the regulating capacity of these two types of loads. Secondly, this paper proposes a joint trading market for electricity and regulating capacity, in which electricity and regulating capacity are both market trading commodities and participate in P2P transactions, thereby matching the expected output and uncertain output of WPV with flexible loads at the same time, effectively improving the market's operating efficiency and the utilization rate of WPV. Finally, a numerical example is used to analyze the proposed method and model. The results show that the model proposed in this paper can not only effectively increase the local absorption of WPV, but also encourage users in the building to actively participate in P2P transactions.
[0007] In order to achieve the above object, the technical solution of the present invention is:
[0008] A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions includes the following steps:
[0009] S1: The operating framework of smart buildings mainly includes electricity users, wind and photovoltaic aggregators (WPVAs) and power grid companies, and electricity trading can be carried out within the building;
[0010] S2: Considering the diversity of loads in smart buildings, the loads are divided into time flexibility (TF) loads and power flexibility (PF) loads according to whether the loads can provide regulation capacity;
[0011] S3: Analyze how to use the load regulation capacity to eliminate the uncertainty of wind and solar resources output;
[0012] S4: Design the transaction process of smart buildings to ensure the feasibility of power trading;
[0013] S5: Design a market mechanism to complete the matching process between wind and solar resource aggregators and power users;
[0014] S6: In order to analyze the rationality of the model, it is necessary to conduct a rationality analysis of the model so that it can be applied to actual situations;
[0015] Among them, in step S1, the operation framework design of the intelligent building includes the following processes:
[0016] S1-1: What is the operating framework of a smart building? Figure 1 As shown in the figure, N represents the total number of smart buildings. The trading entities in the market include power users, wind and photovoltaic aggregators (WPVAs) and power grid companies. Among them, the power user side has flexible loads including lighting systems, EVs, air conditioners, etc.
[0017] S1-2: The tasks of market trading entities in the market are as follows: Electricity users: purchase electricity to meet their own needs, and at the same time obtain benefits by providing flexible load regulation capacity to WPVAs. WPVAs: aggregate WPV resources in different buildings, sell electricity, and purchase regulation capacity to avoid penalties due to WPV prediction errors. Grid companies: maintain the balance of electricity supply and demand in smart buildings. In addition, smart buildings also have market operators and distribution system operators (DSOs). Among them, the market operator is responsible for collecting the bidding information of participants and establishing a P2P matching relationship. The DSO is responsible for the safe and reliable operation of the distribution system and monitoring the operation of flexible loads.
[0018] Wherein, in step S2, the modeling of TF load and PF load includes the following contents:
[0019] S2-1: TF means that the working time of the load can be changed within a specific time interval. Here, the air humidifier (AH) in the smart building is used as a representative to analyze the characteristics of TF. The energy consumption mode of AH can be modeled as (1) and (2):
[0020]
[0021] In the formula, and P AH Respectively represent the working power and rated power of AH; is a binary variable, 1 and 0 represent the start and stop status of the load respectively; t AH represents the continuous working time of AH; [t0,t end ] is the working time interval of AH, and the load must complete the energy-consuming work within this time period.
[0022] Formula (1) shows that the working power requirement of AH is equal to the rated power; Formula (2) shows that AH cannot work intermittently.
[0023] S2-2: PF means that the working power of the load is adjustable. Here, the electric vehicle (EV) in the smart building is used as a representative to analyze the characteristics of PF. EV can be modeled as:
[0024]
[0025] In the formula, and They refer to the operating power, minimum and maximum charging power of the EV respectively; and Refers to the downward and upward regulation capacity provided by the EV at time t; Refers to the stored electric energy capacity of EV at time t; Refers to the electric energy capacity of EV at the beginning of the charging phase; η EV For charging efficiency.
[0026] Formula (3) indicates that the charging power of EV should be between the minimum and maximum charging powers; Formulas (4) and (5) indicate that the difference between the charging power of EV at time t and the rated minimum and maximum charging powers is the regulation capacity it can provide.
[0027] Among them, in step S3, the downward and upward adjustment capacity provided by EV can eliminate the predicted positive and negative errors of WPVA respectively, and the modeling steps are as follows:
[0028] S3-1: A positive prediction deviation of WPVA means that its actual output is greater than the expected output; a negative prediction deviation means that its actual output is less than the expected output. The more regulation capacity WPVA obtains, the more prediction errors it can eliminate. The concept of cumulative expected deviation (CED) is introduced here to measure the ability of regulation capacity to eliminate errors, and its definition is as follows:
[0029]
[0030] Among them, f t (x) is the probability density function of WPVA prediction error; c d,t and c u,t To adjust the capacity downward and upward; D represents the unit time period; D p,t and D n,t Represents the cumulative expected positive deviation (CEPD) and the cumulative expected negative deviation (CEND).
[0031] It is usually assumed that the prediction error of WPVA follows a normal distribution. This paper adopts a zero-mean Gaussian distribution function: t =N(0,σ 2 ), f t The expression of (x) is shown in formula (9).
[0032]
[0033] In order to more intuitively demonstrate the ability of adjusting capacity to eliminate WPVA prediction deviation, here we take the example of eliminating WPVA CEPD by downward adjusting capacity, and its relationship diagram is as follows Figure 2 shown.
[0034] pass Figure 2 It can be seen that the CEPD eliminated by downward regulating capacity shows a rapid growth trend in the early stage, and the growth gradually slows down in the later stage. This shows that as the regulating capacity obtained by WPVA increases, the amount of error eliminated also increases, but the prediction deviation of WPVA is limited. Obtaining too much regulating capacity will cause most of the regulating capacity to be used with a very small probability, increasing costs. This provides an important reference for WPVA when purchasing regulating capacity.
[0035] Among them, the transaction process of smart buildings is designed as follows:
[0036] S4-1: The operation process of smart buildings can be divided into the following steps: (1) In the day-ahead market, WPVAs predict WPV output based on historical data and submit bid electricity prices, predicted power curves and uncertainty parameter information to the market operator; electricity users submit load information to the market operator. (2) WPVAs are sorted from low to high according to bid electricity prices and matched with flexible loads in sequence. (3) WPVAs that have completed matching sign contracts with the loads that match them and exit the market. (4) When the WPVA with the highest bid completes the match, the P2P market is closed. The market operator broadcasts the matching relationship to market participants, and market participants who still have demand will trade with the power grid company. The flow chart is as follows: Figure 3 shown.
[0037] S4-2: Execution process of power trading in smart buildings: In the day-ahead stage, market participants complete P2P power trading. In the real-time stage, power users control the switching of flexible loads according to the matching results of the day-ahead market to consume the expected output of WPV. When the actual output power of WPV deviates from its expected power, the working power of the PF load matched with it will be adjusted appropriately to follow the actual output of WPV.
[0038] The market mechanisms include:
[0039] Pricing mechanism: In the market, the electricity price quoted by WPVAs in the day-ahead phase is the respective electricity transaction price. The price of the regulation capacity provided by PF loads can be set as a fixed value.
[0040] Trading mechanism: The P2P market designed in this paper aims to promote the local consumption of WPV in smart buildings. Here, the market optimization goal is to maximize the total income of WPVA, which can be expressed as:
[0041]
[0042] Where U i means the bid price of WPVAi; represents the estimated transaction power of WPVAi in the P2P market, Refers to the income from WPVAi selling positive deviation power to power users by downward adjustment of capacity. If the WPVAi does not match the downward adjustment capacity, the positive power deviation will be sold to the power grid company at a lower price; Refers to the compensation cost caused by negative power deviation avoided by WPVAi by adjusting capacity upward. If the WPVAi does not match the downward adjustment capacity, it needs to buy electricity from the grid company at a higher price as compensation; and The WPVAi is divided into price U u and U d Purchase regulation capacity and cost.
[0043] In addition, the above optimization problem is subject to the following constraints:
[0044]
[0045] In the formula, It is a binary variable, where 1 indicates that WPVAi and load l are matched, and 0 indicates that the match is not completed; Refers to the working power of load l; and are the upward and downward regulation capacities provided by load l, respectively; L is the total number of loads in the market; P t WPVi Refers to the estimated electrical energy output of WPVAi.
[0046] In addition, due to D p,t and D n,t About c d,t and c u,t In order to simplify the solution process, the 3σ rule can be used to set the three segmentation points σ, 2σ, and 3σ to perform piecewise linearization processing. The specific implementation is as follows: Figure 4 shown.
[0047] The rationality analysis of the model includes:
[0048] The "low price first matching" strategy adopted in the power trading market can effectively increase the vitality of the market and mobilize the enthusiasm of market participants. The rationality of the model can be analyzed from the following two aspects: (1) The on-grid electricity price and the price of purchasing electricity from the power grid company are naturally set as the lower and upper limits of the electricity price in the P2P market, and the bid price of WPVAs should be between these two prices. If the power bid of WPVAs is too low, its own profit will be very small; if the power bid is too high, power users will choose to purchase electricity from the power grid company. (2) Since PF loads can provide regulation capacity to WPVAs, it can avoid penalties caused by prediction errors, which prompts WPVAs in the market to prioritize matching with PF loads. Because the market adopts the "low price first matching" strategy, the probability of PF loads matching with WPVAs with lower power bids is higher, resulting in more profits for power users with PF loads, which effectively promotes the enthusiasm of power users to participate in P2P transactions and greatly reduces the occurrence of wind and solar power abandonment in the market.
[0049] The beneficial effects of the present invention are:
[0050] 1. It can effectively utilize wind and solar resources and improve the power generation rate of wind and solar resources.
[0051] 2. The regulation capacity provided by the flexible load can effectively eliminate the prediction deviation of WPV and maximize social benefits.
[0052] 3. In order to encourage electricity users to actively provide regulation capacity to WPVA, users who provide regulation capacity can not only purchase electricity at a lower price, but also sell the regulation capacity to obtain additional income. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] Figure 1 It is the operation framework diagram of the intelligent building of the present invention.
[0054] Figure 2 It is the cumulative expected deviation diagram of the wind and solar aggregator of the present invention.
[0055] Figure 3 It is a market transaction flow chart of the present invention.
[0056] Figure 4 It is a piecewise linear model diagram of the cumulative expected deviation and the adjustment capacity of the present invention.
[0057] Figure 5 Graph showing the estimated output and standard deviation of WPVAs in a small-scale market according to the present invention.
[0058] Figure 6 (a) is the matching result diagram of WPVA1 in the small-scale market of the present invention.
[0059] Figure 6 (b) is the matching result diagram of WPVA2 in the small-scale market of the present invention.
[0060] Figure 6 (c) is the matching result diagram of WPVA3 in the small-scale market of the present invention.
[0061] Figure 6 (d) is the matching result diagram of WPVA4 in the small-scale market of the present invention.
[0062] Figure 7 It is a diagram of the regulation capacity matching results of WPVAs in a small-scale market of the present invention.
[0063] FIG8( a ) is a CEDs diagram of WPVA1 in a small-scale market of the present invention.
[0064] FIG8( b ) is a CEDs diagram of WPVA2 in a small-scale market of the present invention.
[0065] FIG8( c ) is a CEDs diagram of WPVA3 in a small-scale market of the present invention.
[0066] FIG8( d ) is a diagram of CEDs of WPVA4 in a small-scale market of the present invention.
[0067] Fig. 9 Graph showing the estimated output and standard deviation of WPVAs for the mass market of the present invention.
[0068] FIG. 10( a ) is a diagram showing the matching results of WPVA1 in a large-scale market according to the present invention.
[0069] FIG. 10( b ) is a diagram showing the matching results of WPVA2 in a large-scale market according to the present invention.
[0070] FIG10( c ) is a diagram showing the matching results of WPVA3 in a large-scale market according to the present invention.
[0071] FIG10( d ) is a diagram showing the matching results of WPVA4 in a large-scale market according to the present invention.
[0072] FIG10( e ) is a diagram showing the matching results of WPVA5 in the large-scale market of the present invention.
[0073] FIG10( f ) is a diagram showing the matching results of WPVA6 in a large-scale market according to the present invention.
[0074] FIG. 11( a ) is a diagram showing the CEDs of WPVA1 in the mass market of the present invention.
[0075] FIG11( b ) is a diagram of CEDs of WPVA2 in the mass market of the present invention.
[0076] FIG11( c ) is a diagram showing the CEDs of WPVA3 in the mass market of the present invention.
[0077] FIG11( d ) is a diagram showing the CEDs of WPVA4 in the mass market of the present invention.
[0078] FIG11( e ) is a diagram showing the CEDs of WPVA5 in the mass market of the present invention.
[0079] FIG11( f ) is a diagram of CEDs of WPVA6 in the mass market of the present invention.
[0080] Fig.12 is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0081] The present invention will be further described below in conjunction with the accompanying drawings
[0082] Reference Figure 1 to Figure 1 1. A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions, the method comprising the following steps:
[0083] A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions, characterized in that the method comprises the following steps:
[0084] S1: The operating framework of smart buildings mainly includes electricity users, wind and photovoltaic aggregators (WPVAs) and power grid companies, and electricity trading can be carried out within the building;
[0085] S2: Considering the diversity of loads in smart buildings, the loads are divided into time flexibility (TF) loads and power flexibility (PF) loads according to whether the loads can provide regulation capacity;
[0086] S3: Analyze how to use the load regulation capacity to eliminate the uncertainty of wind and solar resources output;
[0087] S4: Design the transaction process of smart buildings to ensure the feasibility of power trading;
[0088] S5: Design a market mechanism to complete the matching process between wind and solar resource aggregators and power users;
[0089] S6: In order to analyze the rationality of the model, it is necessary to conduct a rationality analysis of the model so that it can be applied to actual situations;
[0090] In step S1, the operation framework design of the intelligent building includes the following processes:
[0091] S1-1: What is the operating framework of a smart building? Figure 1 As shown in the figure, N represents the total number of smart buildings. The trading entities in the market include power users, wind and photovoltaic aggregators (WPVAs) and power grid companies. Among them, the power user side has flexible loads including lighting systems, EVs, air conditioners, etc.
[0092] S1-2: The tasks of market trading entities in the market are as follows: Electricity users: purchase electricity to meet their own needs, and at the same time obtain benefits by providing flexible load regulation capacity to WPVAs. WPVAs: aggregate WPV resources in different buildings, sell electricity, and purchase regulation capacity to avoid penalties due to WPV prediction errors. Grid companies: maintain the balance of electricity supply and demand in smart buildings. In addition, smart buildings also have market operators and distribution system operators (DSOs). Among them, the market operator is responsible for collecting the bidding information of participants and establishing a P2P matching relationship. The DSO is responsible for the safe and reliable operation of the distribution system and monitoring the operation of flexible loads.
[0093] In step S2, the modeling of TF load and PF load includes the following contents:
[0094] S2-1: TF means that the working time of the load can be changed within a specific time interval. Here, the air humidifier (AH) in the smart building is used as a representative to analyze the characteristics of TF. The energy consumption mode of AH can be modeled as (15) and (16):
[0095]
[0096] In the formula, and P AH Respectively represent the working power and rated power of AH; is a binary variable, 1 and 0 represent the start and stop status of the load respectively; t AH represents the continuous working time of AH; [t0,t end ] is the working time interval of AH, and the load must complete the energy-consuming work within this time period.
[0097] Formula (15) shows that the working power requirement of AH is equal to the rated power; Formula (16) shows that AH cannot work intermittently.
[0098] S2-2: PF means that the working power of the load is adjustable. Here, the electric vehicle (EV) in the smart building is used as a representative to analyze the characteristics of PF. EV can be modeled as:
[0099]
[0100]
[0101] In the formula, and They refer to the operating power, minimum and maximum charging power of the EV respectively; and Refers to the downward and upward regulation capacity provided by the EV at time t; Refers to the stored electric energy capacity of EV at time t; Refers to the electric energy capacity of EV at the beginning of the charging phase; η EV For charging efficiency.
[0102] Formula (17) indicates that the charging power of EV should be between the minimum and maximum charging powers; Formulas (18) and (19) indicate that the difference between the charging power of EV at time t and the rated minimum and maximum charging powers is the regulation capacity it can provide.
[0103] In step S3, the downward and upward adjustment capacities provided by the EV can eliminate the positive and negative errors of the WPVA prediction respectively. The modeling steps are as follows:
[0104] S3-1: A positive prediction deviation of WPVA means that its actual output is greater than the expected output; a negative prediction deviation means that its actual output is less than the expected output. The more regulation capacity WPVA obtains, the more prediction errors it can eliminate. The concept of cumulative expected deviation (CED) is introduced here to measure the ability of regulation capacity to eliminate errors, and its definition is as follows:
[0105]
[0106] Among them, f t (x) is the probability density function of WPVA prediction error; c d,t and c u,t To adjust the capacity downward and upward; D represents the unit time period; D p,t and D n,t Represents the cumulative expected positive deviation (CEPD) and the cumulative expected negative deviation (CEND).
[0107] It is usually assumed that the prediction error of WPVA follows a normal distribution. This paper adopts a zero-mean Gaussian distribution function: t =N(0,σ 2 ), f t The expression of (x) is shown in formula (23).
[0108]
[0109] In order to more intuitively demonstrate the ability of adjusting capacity to eliminate WPVA prediction deviation, here we take the example of eliminating WPVA CEPD by downward adjusting capacity, and its relationship diagram is as follows: Figure 2 shown.
[0110] pass Figure 2 It can be seen that the CEPD eliminated by downward regulating capacity shows a rapid growth trend in the early stage, and the growth gradually slows down in the later stage. This shows that as the regulating capacity obtained by WPVA increases, the amount of error eliminated also increases, but the prediction deviation of WPVA is limited. Obtaining too much regulating capacity will cause most of the regulating capacity to be used with a very small probability, increasing costs. This provides an important reference for WPVA when purchasing regulating capacity.
[0111] The transaction process of smart buildings is designed as follows:
[0112] S4-1: The operation process of smart buildings can be divided into the following steps: (1) In the day-ahead market, WPVAs predict WPV output based on historical data and submit bid electricity prices, predicted power curves and uncertainty parameter information to the market operator; electricity users submit load information to the market operator. (2) WPVAs are sorted from low to high according to bid electricity prices and matched with flexible loads in sequence. (3) WPVAs that have completed matching sign contracts with the loads that match them and exit the market. (4) When the WPVA with the highest bid completes the match, the P2P market is closed. The market operator broadcasts the matching relationship to market participants, and market participants who still have demand will trade with the power grid company. The flow chart is as follows: Figure 3 shown.
[0113] S4-2: Execution process of power trading in smart buildings: In the day-ahead stage, market participants complete P2P power trading. In the real-time stage, power users control the switching of flexible loads according to the matching results of the day-ahead market to consume the expected output of WPV. When the actual output power of WPV deviates from its expected power, the working power of the PF load matched with it will be adjusted appropriately to follow the actual output of WPV.
[0114] The market mechanisms include:
[0115] Pricing mechanism: In the market, the electricity price quoted by WPVAs in the day-ahead phase is the respective electricity transaction price. The price of the regulation capacity provided by PF loads can be set as a fixed value.
[0116] Trading mechanism: The P2P market designed in this paper aims to promote the local consumption of WPV in smart buildings. Here, the market optimization goal is to maximize the total income of WPVA, which can be expressed as:
[0117]
[0118] Where U i means the bid price of WPVAi; represents the estimated transaction power of WPVAi in the P2P market, Refers to the income from WPVAi selling positive deviation power to power users by downward adjustment of capacity. If the WPVAi does not match the downward adjustment capacity, the positive power deviation will be sold to the power grid company at a lower price; Refers to the compensation cost caused by negative power deviation avoided by WPVAi by adjusting capacity upward. If the WPVAi does not match the downward adjustment capacity, it needs to buy electricity from the grid company at a higher price as compensation; and The WPVAi is divided into price U u and Ud Purchase regulation capacity and cost.
[0119] In addition, the above optimization problem is subject to the following constraints:
[0120]
[0121] In the formula, It is a binary variable, where 1 indicates that WPVAi and load l are matched, and 0 indicates that the match is not completed; Refers to the working power of load l; and are the upward and downward regulation capacities provided by load l, respectively; L is the total number of loads in the market; P t WPVi Refers to the estimated electrical energy output of WPVAi.
[0122] In addition, due to D p,t and D n,t About c d,t and c u,t In order to simplify the solution process, the 3σ rule can be used to set the three segmentation points σ, 2σ, and 3σ to perform piecewise linearization processing. The specific implementation is as follows: Figure 4 shown.
[0123] In step S6, the rationality analysis of the model includes:
[0124] The "low price first matching" strategy adopted in the power trading market can effectively increase the vitality of the market and mobilize the enthusiasm of market participants. The rationality of the model can be analyzed from the following two aspects: (1) The on-grid electricity price and the price of purchasing electricity from the power grid company are naturally set as the lower and upper limits of the electricity price in the P2P market, and the bid price of WPVAs should be between these two prices. If the power bid of WPVAs is too low, its own profit will be very small; if the power bid is too high, power users will choose to purchase electricity from the power grid company. (2) Since PF loads can provide regulation capacity to WPVAs, it can avoid penalties caused by prediction errors, which prompts WPVAs in the market to prioritize matching with PF loads. Because the market adopts the "low price first matching" strategy, the probability of PF loads matching with WPVAs with lower power bids is higher, resulting in more profits for power users with PF loads, which effectively promotes the enthusiasm of power users to participate in P2P transactions and greatly reduces the occurrence of wind and solar power abandonment in the market.
[0125] In order to enable those skilled in the art to better understand the present invention, the example analysis includes the following components:
[0126] 1. Case Description
[0127] The P2P market proposed here is modeled as a mixed integer linear programming. The optimization program is written in the MATLAB R2021a environment and the GUROBI 9.1.2 solver is called for solution.
[0128] In the example, the P2P trading market consists of 4 WPVAs, 5 TF loads and 5 PF loads. The electricity selling prices of WPVAs are shown in Table 1. It is worth pointing out that the electricity price of WPVAs needs to be set between the on-grid electricity price and the price of purchasing electricity from the power grid company.
[0129] 2. Simulation Example Settings
[0130] In this example, the on-grid electricity price is set at 0.5 yuan per kWh, while the price of electricity purchased from the power grid company is 0.7 yuan per kWh. If the electricity price offered by WPVAs exceeds this price range, the WPVAs will have too low a profit or will not be able to participate in P2P market transactions. The P2P transaction time interval set in the building is 6:00-17:00. The expected output curve of WPVAs is as follows: Figure 5 As shown in the figure, the solid line represents the expected output of WPVAs, and the dotted line represents the standard deviation of the probability density function of the WPVAs prediction error; the parameters of the flexible load are shown in Table 2-3; the price of the upward and downward adjustment capacity provided by the load is set to 0.087 yuan per kilowatt-hour.
[0131] Table 1 Electrical parameters of WPVAs
[0132]
[0133] Table 2 AHs parameters
[0134]
[0135] Table 3 EVs parameters
[0136]
[0137] 3. Simulation Analysis
[0138] In the above P2P market, the bid prices of WPVAs are ranked from low to high as WPVA1, WPVA2, WPVA3 and WPVA4. The matching results are shown in Table 4.
[0139] Table 4 Matching results of energy market
[0140]
[0141] From Table 4, we can see that the number of flexible loads matched by WPVA1 to WPVA4 is gradually decreasing. By the end of WPVA3, all loads in the market have been matched. WPVA4, which has not matched any flexible load, can only sell all its electricity to the grid company at the grid price. Therefore, in order to obtain more profits, WPVAs will appropriately reduce their own electricity bids in the day-ahead market to be more competitive.
[0142] The power supply of WPVAs to flexible loads at each time is as follows: Figure 6 As shown in the figure, the green bar graph represents the energy supply from WPVAs to AHs, and the blue bar graph represents the energy supply from WPVAs to EVs. As can be seen from the figure, most EVs that can provide regulation capacity for WPVAs are matched with WPVA1, which has the lowest electricity bid. These EVs have obtained the lowest electricity purchase price in the market, while AHs that cannot provide regulation capacity are matched with WPVAs in order from low to high prices, effectively encouraging electricity users to participate in local P2P electricity transactions.
[0143] The regulatory capacity obtained by WPVAs is as follows Figure 7 As shown in the figure, the upward and downward regulation capacities obtained by WPVAs are represented above and below the zero scale line, respectively. As shown in the figure, WPVA1, which has the lowest electricity bid, has obtained a large amount of regulation capacity, which can eliminate its own forecast errors in the real-time market through regulation capacity. On the contrary, WPVA4, which has the highest bid, has failed to obtain regulation capacity in the market. When WPVA4 sells electricity to the power grid company in the real-time stage, if there is a forecast error, it will be punished.
[0144] The CEDs obtained by WPVAs are shown in Figure 8. In this transaction, the maximum CEPD and CEND that WPVA1 can avoid are 8.38kW and 9.58kW respectively; the maximum CEPD and CEND that WPVA2 can avoid are 3.79kW and 1.89kW respectively; the maximum CEPD and CEND that WPVA3 can avoid are 1.58kW and 1.58kW respectively; and WPVA4 fails to match flexible loads in the market, and the maximum CEPD and CEND that can be avoided are both 0kW.
[0145] In summary, by obtaining regulation capacity, the uncertainty of WPV output can be effectively eliminated, thereby improving the local consumption rate of WPV.
[0146] In order to verify the performance of the model constructed in this paper when the forecast deviation of WPVA is large, the other data of the above example are kept unchanged, and only the standard deviation σ0 of the forecast error function of WPVA1 at each time is increased by 1 and 2 respectively and marked as σ1 and σ2 respectively. The market clearing results under the three standard deviations are compared. The CEDs obtained by WPVAs are shown in Table 5.
[0147] Table 5 CEDs obtained by WPVAs with different standard deviations
[0148]
[0149] As can be seen from the above table, the larger the σ is, the more CEDs WPVA can eliminate by adjusting capacity, that is, it can obtain more profits, which fully proves that the model constructed in this paper has a good performance effect, that is, it can effectively deal with the uncertainty of WPV output.
[0150] In order to verify the scalability of the model built in this paper, the scale of the smart building is expanded in this section. In this example, there are 6 WPVAs, 12 TF loads, and 12 PF loads. The expected output and standard deviation of WPVAs are shown in Fig. 9 The power user parameters are shown in Table 6-8.
[0151] Table 6 Power parameters of WPVAs
[0152]
[0153] Table 7 AHs parameters
[0154]
[0155] (hh)
[0156]
[0157] First, the WPVAs are sorted from low to high according to the electricity bids: WPVA1, WPVA2, …, WPVA6, and according to the “low price priority matching strategy”, they are matched with the flexible loads in the market in turn.
[0158] The clearing results of the P2P market are shown in Table 9. As can be seen from the table, as the electricity bids of WPVAs increase successively, their surplus electricity in the P2P market also increases successively, and the surplus electricity will be sold to the upper power grid at a lower price. WPVA1, which has the lowest electricity bid, has matched 7 AHs and 5 EVs in the market, leaving only 1.18% of electricity to be traded with the f power grid company. As of WPVA5, all flexible loads in the P2P market have been matched, so WPVA6, which has the highest electricity bid, has not been matched with flexible loads in the market, and all its electricity will be sold to the power grid company at the on-grid electricity price, greatly reducing the income from electricity sales.
[0159] Table 9 Matching results of energy market
[0160]
[0161] The power supply of WPVAs to flexible loads is shown in Figure 10, where the green and blue bars represent the energy supply of WPVAs to Ahs and EVs, respectively. It can be observed from the figure that WPVA1 and WPVA2, which have lower power bids, match more flexible loads, while the remaining WPVAs can only sell most of their power to the power grid company at a low price due to their higher bids.
[0162] The CED that WPVAs can avoid by adjusting capacity is shown in Figure 11. It can be observed that WPVA1 and WPVA2 with lower bids have stronger abilities to eliminate forecast errors, while the remaining WPVAs have weaker abilities to eliminate forecast errors due to the smaller adjustment capacity they obtain, making them more susceptible to market penalties.
[0163] In summary, after expanding the scale of the market, the market clearing results still meet the expected effects, which verifies the scalability of the model.
[0164] The contents described in the embodiments of this specification are merely an enumeration of the implementation forms of the inventive concept. The protection scope of the present invention should not be regarded as limited to the specific forms described in the embodiments. The protection scope of the present invention also extends to equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.
Claims
1. A point-to-point optimization scheduling method for smart buildings considering uncertain transactions. It is characterized in that The following steps are involved: S1: The operating framework of smart buildings includes electricity users, wind and photovoltaic aggregators (WPVAs) and power grid companies, and electricity trading can be carried out within the building; S2: Considering the diversity of loads in smart buildings, the loads are divided into time flexibility (TF) loads and power flexibility (PF) loads according to whether the loads can provide regulation capacity; S3: Analyze how to use the load regulation capacity to eliminate the uncertainty of wind and solar resources output; S4: Design the transaction process of smart buildings to ensure the feasibility of power trading; S5: Design a market mechanism to complete the matching process between wind and solar resource aggregators and power users; S6: In order to analyze the rationality of the model, a rationality analysis is conducted on the model so that it can be applied to actual situations.
2. A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions as claimed in claim 1, It is characterized in that In step S1, the design of the operation framework of the intelligent building includes the following processes: S1-1: Build the operating framework of smart buildings; the trading entities in the middle market include power users, wind and photovoltaic aggregators (WPVAs) and power grid companies. Among them, the power user side has flexible loads including lighting systems, EVs, air conditioners, etc. S1-2: The tasks of market trading entities in the market are as follows: Electricity users: purchase electricity to meet their own needs, and at the same time obtain benefits by providing flexible load regulation capacity to WPVAs; WPVAs: aggregate WPV resources in different buildings, sell electricity, and purchase regulation capacity to avoid penalties due to WPV prediction errors; Power grid companies: maintain the balance of electricity supply and demand in smart buildings; In addition, smart buildings also have market operators and distribution system operators (DSO); Among them, the market operator is responsible for collecting the bidding information of participants and establishing a P2P matching relationship; and the DSO is responsible for the safe and reliable operation of the distribution system and monitoring the operation of flexible loads.
3. A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions as described in claim 2, It is characterized in that In step S2, the modeling of TF load and PF load includes the following contents: S2-1: TF refers to the load working time that can be changed within a specific time interval. The air humidifier (AH) in the smart building is used as a representative to analyze the characteristics of TF. The energy consumption mode of AH can be modeled as equations (1) and (2): In the formula, and P AH Respectively represent the working power and rated power of AH; is a binary variable, 1 and 0 represent the start and stop status of the load respectively; t AH Indicates the continuous working time of AH; [t 0 ,t end ] is the working time interval of AH, and the load must complete the energy-consuming work within this time period; Formula (1) shows that the working power requirement of AH is equal to the rated power; Formula (2) shows that AH cannot work intermittently; S2-2: PF refers to the adjustable working power of the load; the electric vehicle (EV) in the smart building is used as a representative to analyze the characteristics of PF. The EV can be modeled as: In the formula, and They refer to the operating power, minimum and maximum charging power of the EV respectively; and Refers to the downward and upward regulation capacity provided by the EV at time t; Refers to the stored electric energy capacity of EV at time t; Refers to the electric energy capacity of EV at the beginning of the charging phase; η EV For charging efficiency; Formula (3) indicates that the charging power of EV should be between the minimum and maximum charging powers; Formulas (4) and (5) indicate that the difference between the charging power of EV at time t and the rated minimum and maximum charging powers is the regulation capacity it can provide.
4. A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions as claimed in claim 3, It is characterized in that In step S3, the downward and upward adjustment capacities provided by the EV can eliminate the positive and negative errors of the WPVA prediction respectively. The modeling steps are as follows: S3-1: A positive prediction deviation of WPVA means that its actual output is greater than the expected output; a negative prediction deviation means that its actual output is less than the expected output; the more regulation capacity WPVA obtains, the more prediction errors it can eliminate; the concept of cumulative expected deviation (CED) is introduced here to measure the ability of regulation capacity to eliminate errors, and its definition is as follows: Among them, f t (x) is the probability density function of WPVA prediction error; c d,t and c u,t To adjust the capacity downward and upward; D represents the unit time period; D p,t and D n,t Represents the cumulative expected positive deviation (CEPD) and the negative deviation (CEND); Assuming that the prediction error of WPVA follows a normal distribution, a zero-mean Gaussian distribution function is used: t =N(0,σ 2 ), f t The expression of (x) is shown in formula (9); 5. A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions as claimed in claim 4, It is characterized in that The transaction process of smart buildings is as follows: S4-1: The operation process of smart buildings can be divided into the following steps: (1) In the day-ahead market, WPVAs predict WPV output based on historical data and submit bid prices, predicted power curves, and uncertainty parameter information to the market operator; Electricity users submit load information to market operators; (2) Sort the WPVAs from low to high according to the bid electricity prices and match them with flexible loads in sequence; (3) The WPVAs that have completed the matching will sign a contract with the loads that match them and exit the market; (4) When the WPVA with the highest bid completes the matching, the P2P market is closed; the market operator broadcasts the matching relationship to market participants, and market participants that still have demand will trade with the power grid company; S4-2: Execution process of power trading in smart buildings: In the day-ahead stage, market participants complete P2P power trading; In the real-time stage, electricity users control the switching of flexible loads according to the matching results of the day-ahead market to consume the expected output of WPV; when the actual output power of WPV deviates from its expected power, the working power of the PF load matched with it will be adjusted appropriately to follow the actual output of WPV.
6. A point-to-point optimization scheduling method for intelligent buildings considering uncertain transactions as claimed in claim 5, It is characterized in that The market mechanisms described include: Pricing mechanism: In the market, the electricity price quoted by WPVAs in the day-ahead phase is the respective electricity transaction price; and the price of the regulation capacity provided by PF loads can be set as a fixed value; Trading mechanism: The P2P market aims to promote the local consumption of WPV in smart buildings, with the maximization of WPVA total revenue as the market optimization goal, which can be expressed as: Where U i means the bid price of WPVAi; represents the estimated transaction power of WPVAi in the P2P market, Refers to the income from WPVAi selling positive deviation power to power users by downward adjustment of capacity; if the WPVAi does not match the downward adjustment capacity, the positive power deviation will be sold to the power grid company at a lower price; Refers to the compensation cost caused by negative power deviation avoided by WPVAi by adjusting capacity upward. If the WPVAi does not match the downward adjustment capacity, it needs to buy electricity from the grid company at a higher price as compensation; and The WPVAi is divided into price U u and U d Purchase regulation capacity and Costs; In addition, the above optimization problem is subject to the following constraints: In the formula, It is a binary variable, where 1 indicates that WPVAi and load l are matched, and 0 indicates that the match is not completed; Refers to the working power of load l; and are the upward and downward regulation capacities provided by load l, respectively; L is the total number of loads in the market; Refers to the estimated power output of WPVAi; In addition, due to D p,t and D n,t About c d,t and c u,t A nonlinear function; in order to simplify the solution process, the 3σ rule can be used to set the three segmentation points σ, 2σ, and 3σ respectively to perform piecewise linearization processing.
7. A virtual inertia optimization scheduling method considering frequency-time-space distribution as claimed in claim 6, It is characterized in that In step S6, the rationality analysis of the model specifically includes: (1) The on-grid electricity price and the price of electricity purchased from the grid company are naturally set as the lower and upper limits of the electricity price in the P2P market. The bid price of WPVAs should be between these two prices. If the electricity price of WPVAs is too low, its own profit will be very small; if the electricity price is too high, electricity users will choose to purchase electricity from the grid company. (2) Since PF loads can provide regulation capacity to WPVAs, allowing them to avoid penalties due to prediction errors, this encourages WPVAs in the market to prioritize matching with PF loads. Also, because the market adopts a "low price first matching" strategy, the probability of PF loads matching with WPVAs with lower electricity quotations is higher, resulting in more profits for electricity users with PF loads, effectively promoting the enthusiasm of electricity users to participate in P2P transactions and greatly reducing the occurrence of wind and solar power abandonment in the market.
