Power distribution network electric energy distributed transaction method, system, equipment and medium

Abstract

The invention is suitable for the technical field of distributed transaction, and provides a power distribution network electric energy distributed transaction method, system, device and medium, and the method comprises the steps: a distributed energy demand end obtains power generation supply information data issued by each distributed energy supply end, and carries out the data arrangement and cleaning of the power generation supply information data; the distributed energy demand end sets power consumption demand variables, determines node variables of the Bayesian network based on the sorted and cleaned power generation supply information data and the power consumption demand variables, determines correlation between the node variables, constructs a Bayesian network structure and performs parameter learning; analyzing the transaction relevance with each distributed energy supply end according to the constructed Bayesian network structure and the learned parameters; and determining a target transaction supply end and a corresponding transaction strategy based on the transaction relevance. According to the method, the supply end matched with the demand end is screened by using the Bayesian algorithm, so that the efficiency of bilateral negotiation distributed transaction is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of distributed trading technology, and in particular to a method, system, device and medium for distributed trading of electric energy in a distribution network. Background Art

[0002] In traditional power grids, electricity is usually transmitted from large power stations to users, while distributed power transactions in distribution networks mainly rely on distribution networks, allowing owners of distributed power sources to directly trade electricity with other users in the distribution network. With the diversification of trading market entities, more owners of distributed energy sources are participating in distributed transactions, and trading models are also becoming more diversified and intelligent.

[0003] At present, in addition to centralized bidding, various distributed transactions can also choose bilateral negotiation as a trading method, but the current bilateral negotiation model has some shortcomings. On the one hand, there is a problem of information asymmetry, such as the degree of market understanding and value perception of products or services between the two parties. Bilateral negotiation requires repeated communication, negotiation and consultation between the two parties, which takes a lot of time and energy. On the other hand, it is also difficult to match a suitable counterparty in distributed transactions. Both parties need to spend time and energy to find a counterparty that matches their needs. This makes distributed transactions less efficient and more risky.

[0004] In view of this, a method, system, device and medium for distributed trading of electric energy in a distribution network are needed. Summary of the invention

[0005] The embodiments of the present application provide a method, system, device and medium for distributed trading of electric energy in a distribution network, which are used to solve the problems of low efficiency and high risk in distributed electric energy trading.

[0006] A first aspect of an embodiment of the present application provides a method for distributed trading of electric energy in a distribution network, comprising:

[0007] The distributed energy demand side obtains the power generation supply information data released by each distributed energy supply side, and sorts and cleans the power generation supply information data;

[0008] The distributed energy demand side sets the power demand variable, and determines the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variable, determines the correlation between the node variables, constructs the Bayesian network structure and performs parameter learning;

[0009] Analyze the transaction correlation with each distributed energy supply end according to the constructed Bayesian network structure and the learned parameters;

[0010] A target transaction supply end and a corresponding transaction strategy are determined based on the transaction relevance.

[0011] Furthermore, the data sorting and cleaning of the power generation supply information data includes:

[0012] Power generation supply information data collation includes data time series alignment, data format conversion, data classification and labeling, and power generation supply information data cleaning includes missing value processing, duplicate value processing, and outlier detection and processing.

[0013] Furthermore, the distributed energy demand side sets the power demand variable, and determines the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variable, determines the correlation between the node variables, constructs the Bayesian network structure and performs parameter learning, including:

[0014] Filter node variables based on the sorted and cleaned power supply information data;

[0015] Analyze the causal relationship and dependency relationship in the data based on the selected node variables and the set power demand variables, determine the directed edges between the node variables, and construct the Bayesian network structure;

[0016] Based on the constructed Bayesian network structure and the historical transaction data of the distributed energy supply side, the statistical learning method is used to estimate the conditional probability table of each node in the network.

[0017] Furthermore, the node variables in the power generation supply information data include power generation cost, power generation capacity, power generation power and price adjustment, and the power demand variables set on the distributed energy demand side include power capacity, expected price, power purchase decision and geographical location.

[0018] Furthermore, based on the constructed Bayesian network structure and the historical transaction data of the distributed energy supply side, the conditional probability table of each node in the network is estimated using a statistical learning method, including:

[0019] Calculate the prior probability of each node in different states based on the historical transaction data of distributed energy consumers and the historical transaction data corresponding to the set electricity demand variables;

[0020] The conditional probability of each node in the state of its parent node is calculated based on the historical transaction data of distributed energy consumers and the conditional information of the historical transaction data corresponding to the set electricity demand variable.

[0021] Furthermore, the transaction correlation between the constructed Bayesian network structure and the learned parameters and the distributed energy supply terminals includes:

[0022] Perform inference calculations based on the constructed Bayesian network structure and learned parameters;

[0023] Based on the results of inference calculation, the transaction impact and correlation between distributed energy consumers are analyzed.

[0024] Furthermore, the determining of the target transaction supply end and the corresponding transaction strategy based on the transaction relevance includes:

[0025] Determining transaction relevance indicators through transaction relevance with distributed energy consumers, the indicators including price relevance indicators, power relevance indicators and geographic location relevance indicators;

[0026] Calculating a comprehensive transaction relevance weight based on the price relevance index, the power relevance index, and the geographic location relevance index;

[0027] The transaction levels are divided according to the comprehensive weights, and the target transaction strategy is determined based on the transaction levels.

[0028] A second aspect of the embodiment of the present application provides a distributed power trading system for a distribution network, including:

[0029] A data acquisition and processing unit, used for the distributed energy demand side to acquire the power generation supply information data released by each distributed energy supply side, and to sort and clean the power generation supply information data;

[0030] A Bayesian network structure construction unit is used to set power demand variables at the distributed energy demand side, and determine the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variables, determine the correlation between the node variables, construct the Bayesian network structure and perform parameter learning;

[0031] A transaction relevance analysis unit, used for analyzing the transaction relevance with each distributed energy supply end according to the constructed Bayesian network structure and the learned parameters;

[0032] The target transaction supply end and the transaction strategy determination unit are used to determine the target transaction supply end and the corresponding transaction strategy based on the transaction relevance.

[0033] A third aspect of the embodiments of the present application provides a computer device, including:

[0034] memories, transceivers, processors, and bus systems;

[0035] Wherein, the memory is used to store programs;

[0036] The processor is used to execute the program in the memory, including executing the distribution network electric energy distributed trading method as described in any one of the above;

[0037] The bus system is used to connect the memory and the processor so that the memory and the processor can communicate with each other.

[0038] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, comprising instructions, which, when executed on a computer, enables the computer to execute a distribution network electric energy distributed trading method as described in any one of the above.

[0039] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0040] In the present invention, when the bilateral negotiation transaction mode is carried out, the distributed energy demand side first obtains the power generation supply information data released by each distributed energy supply side, and organizes and cleans the power generation supply information data; based on the set power demand variables and power generation supply information data, the system determines the node variables of the Bayesian network, determines the correlation between the node variables, constructs the Bayesian network structure and performs parameter learning; analyzes the transaction correlation with each distributed energy supply side according to the constructed Bayesian network structure and the learned parameters; finally, determines the appropriate transaction party and the corresponding transaction strategy based on the transaction correlation. Based on the setting of transaction demand variables and corresponding supply variables, the present invention uses the Bayesian algorithm to efficiently screen the supply side that matches the demand side, effectively improving the efficiency of bilateral negotiated distributed transactions. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The present invention is a flowchart of an embodiment of a method for distributed trading of electric energy in a distribution network. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0043] In this embodiment, the distributed power transaction method of the distribution network is used to screen the supply side that matches the demand side, and improve the efficiency of bilateral negotiated distributed transactions. The implementation method in this embodiment can be implemented in the system, can be implemented in the server, and can also be implemented in the terminal, and the specific details are not clearly limited.

[0044] Embodiment 1

[0045] See also Figure 1 An embodiment of a method for distributed trading of electric energy in a distribution network of the present invention comprises the following steps:

[0046] S11. The distributed energy demand side obtains the power generation supply information data released by each distributed energy supply side, and organizes and cleans the power generation supply information data;

[0047] The power supply information data includes the power generation index values ​​at different time points (such as every hour, every day, etc.), power generation cost data, technical parameters and operating status data of power generation equipment, and producer price adjustment decision records. The power demand information data includes the user's power consumption data, that is, information such as power consumption, power consumption time, power consumption, expected price data, and user's power consumption behavior characteristics.

[0048] In this embodiment, data sorting and cleaning of power supply information data includes:

[0049] Power generation supply information data collation includes data time series alignment, data format conversion, data classification and labeling, and power generation supply information data cleaning includes missing value processing, duplicate value processing, and outlier detection and processing.

[0050] Time series alignment is to ensure that the timestamps of the data correspond, data format conversion is to convert the data into a unified data format, and data classification and labeling is to classify and label the power generation data. Missing value processing, duplicate value processing, outlier detection and processing, and invalid data processing are to improve the quality of power generation supply information data and power demand information data, and provide a reliable data foundation for subsequent data analysis, transaction matching, and decision making.

[0051] S12. The distributed energy demand side sets the power demand variable, and determines the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variable, determines the correlation between the node variables, constructs the Bayesian network structure and performs parameter learning;

[0052] S13. Analyze the transaction correlation with each distributed energy supply end based on the constructed Bayesian network structure and the learned parameters;

[0053] Step S12 also includes the following:

[0054] 1. Filter node variables based on the sorted and cleaned power supply information data;

[0055] The node variables in the power generation supply information data include power generation cost, power generation capacity, power generation power and price adjustment; the power demand variables set on the distributed energy demand side include power capacity, expected price, power purchase decision and geographical location.

[0056] Specifically, the node variables in the power supply information data are:

[0057] The range of power generation cost C is assumed to be {C1,C2,…,Cn}, for example, {high cost, medium cost, low cost}, with C = C1, C = C2, ..., C = C n express;

[0058] The value range of power generation capacity V is set to {V1, V2, ..., V m}, for example, {small capacity, medium capacity, large capacity}, respectively use V=V1, V=V2,…, V=V m express;

[0059] The power generation P value range is set to {P1, P2, ..., P k}, for example, {low power, high power}, with P = P1, P = P k express;

[0060] The value range of price adjustment A is set to {A1, A2, A3}, for example, {increase, remain unchanged, decrease}, which is represented by A=A1, A=A2, and A=A3 respectively.

[0061] Electricity demand variables:

[0062] The value range of the power consumption capacity D is set to {D1,D2,…,D l}, for example {low demand, high demand}, that is, D=D1,D=D2,…,D=D l express;

[0063] The expected price E range is set to {E1, E2, …, E p}, for example {low price expectation, high price expectation}, that is, E=E1, E=E p express;

[0064] The value range of the power purchase decision B is set to {B1, B2, B3}, for example, {purchase increase, purchase unchanged, purchase decrease}, which is represented by B = B1, B = B2, and B = B3 respectively;

[0065] The value range of the geographical location G is set to {G1, G2, …, G q}, for example {region I, region II, ...}, that is, G = G1, G = G q ,…,G=G q express.

[0066] 2. Analyze the causal relationship and dependency relationship in the data based on the selected node variables and the set power demand variables, determine the directed edges between the node variables, and construct the Bayesian network structure;

[0067] The Bayesian network structure relationship here is determined based on the above received data and domain knowledge. It is assumed here that the Bayesian network structure relationship is determined as follows:

[0068] (1) The power generation cost C affects the power generation capacity V, power generation P and price adjustment A, that is, the directed edge relationships are C→V, C→P, C→A;

[0069] (2) The generation capacity V and the generation power P affect each other and jointly affect the price adjustment A. The directed edge relationships are V→A, P→A, and V→P. It should be noted that P may also affect V. For example, high-power operation may affect the maximum capacity that the equipment can reach. This is just a simplified mutual influence relationship. The specific setting needs to be determined according to the actual situation.

[0070] (3) Price adjustment A, power consumption capacity D, expected price E, and geographical location G jointly influence the power purchase decision B. The directed edge relationships are A→B, D→B, E→B, and G→B.

[0071] 3. Based on the constructed Bayesian network structure and the historical transaction data of the distributed energy supply side, the conditional probability table of each node in the network is estimated using statistical learning methods.

[0072] This includes the following:

[0073] The prior probability of each node in different states is calculated based on the historical transaction data of distributed energy consumers and the historical transaction data corresponding to the set electricity demand variables; the conditional probability of each node in the state of its parent node is calculated based on the conditional information of the historical transaction data of distributed energy consumers and the historical transaction data corresponding to the set electricity demand variables.

[0074] Step S13 includes the following: performing inference calculations based on the constructed Bayesian network structure and the learned parameters; and analyzing the transaction impact and correlation with the distributed energy consumers based on the results of the inference calculations.

[0075] The details are as follows:

[0076] (1) Calculate the prior probability:

[0077] For power generation costs, it is assumed that the probability of power generation costs being at different levels is obtained from the collected historical data, for example:

[0078] P(C=C1)=p1(the prior probability that the power generation cost is C1 is p1); P(C=C2)=p2(the prior probability that the power generation cost is C2 is p2); P(C=C3)=p3(the prior probability that the power generation cost is C3 is p3); then p1+p2+p3=1;

[0079] For power consumption capacity, it is assumed that the probability of power consumption capacity being at different demand levels is:

[0080] P(D=D1)=d1, P(D=D2)=d2, then d1+d2=1;

[0081] The prior probabilities of other nodes are similarly calculated based on historical data, such as the prior probabilities of different values ​​of nodes such as power generation capacity, power generation, price adjustment, expected price, power purchase decision and geographical location. For example:

[0082] P(V=V1), P(V=V2), P(V=V3), etc.; P(A=A1), P(A=A2), P(A=A3), etc.;

[0083] P(E=E1), P(E=E2), etc.; P(B=B1), P(B=B2), P(B=B3), etc.; P(G=G1), P(G=G2),…, etc.

[0084] (2) Calculate conditional probability:

[0085] The conditional probability P(V│C) of power generation capacity under different values ​​of power generation cost is:

[0086] Assume that according to data statistics:

[0087] When C=C1, P=(V=V1│C=C1)=v 11 (It means that when the power generation cost is C1, the probability that the power generation capacity is V1 is v 11 ), P = (V = V2 | C = C2) = v 12 , P = (V = V3 | C = C3) = v 13 , and v 11 +v 12 +v 13 =1;

[0088] When C=C2, similarly, P=9V=V1│C=C2)=v 21 , P = (V = V2 | C = C2) = v 22 , P = (V = V3 | C = C2) = v 23 ;

[0089] When C=C3, similarly, P=(V=V1│C=C3)=v 31 , P = (V = V2 | C = C3) = v 32 , P = (V = V3 | C = C3) = v 33 .

[0090] The conditional probability P(P│C,V) of power generation under different combinations of power generation cost and power generation capacity is:

[0091] For example:

[0092] When C=C1 and V=V1, P=(P=P1│C=C1,V=V1)=p111 (It means that when the power generation cost is C1 and the power generation capacity is V1, the probability that the power generation is P1 is p 111 ), P=(P=P2│C=C1,V=V1)=p 112 , and p 111 +p 112 =1;

[0093] For other different (C, V) combinations, such as (C=C1, V=V2), (C=C1, V=V3), (C=C2, V=V1), etc., similar statistical calculations are required to obtain the corresponding conditional probabilities.

[0094] The conditional probability P(A│C,V,P) of price adjustment under different combinations of power generation cost, power generation capacity and power generation power is:

[0095] Assume that the following strength data is obtained (actually a large amount of data statistics are required):

[0096] When C=C1、V=V1、P=P1,P=(A=A1│C=C1,V=V1,P=P1)=a 1111 (Indicates that the probability of price increase in the corresponding situation is a 1111 ), P=(A=A2│C=C1, V=V1, P=P1)=a 1112 ,P=(A=A3│C=C1,V=V1,P=P1)=a 1113 For other combinations of (C, V, P), the conditional probability of price adjustment A at different values ​​must be statistically calculated.

[0097] The conditional probability P(B│A,D,E,G) of the power purchase decision under different combinations of price adjustment, power consumption capacity, expected price and geographical location:

[0098] This is a more complex conditional probability that requires statistical analysis based on a large amount of transaction data. Assume the following example:

[0099] When A=A1, D=D1, E=E1, G=G1, P=(B=B1│A=A1, D=D1, E=E1, G=G1)=v 1111 (indicates the probability of purchase increase in the corresponding situation b 1111 ), P=(B=B2│A=A1, D=D1, E=E1, G=G1)=b 1112 , P=(B=B3│A=A1,D=D1,E=E1,G=G1)=b 1113 , and b 1111 +b 1112 +b 1113 =1;

[0100] For other combinations of values, such as (A=A1, D=D1, E=E1, G=G1), (A=A1, D=D1, E=E2, G=G1), etc., as well as various combinations when A has other values ​​(A1, A3), similar statistical calculations are required to obtain the corresponding conditional probabilities.

[0101] (3) Calculate the joint probability:

[0102] According to the chain rule of Bayesian networks, the joint probability can be expressed as:

[0103]

[0104] For example, for a specific instance including generation cost C, generation capacity V, generation power P, price adjustment A, power consumption capacity D, expected price E, power purchase decision B and geographical location G, the joint probability is:

[0105] P(C,V,P,A,D,E,B,G)=P(C)×P(V│C)×P(P│C,V)×P(A│C,V,P)×P(D)×P(E)×P(B│A,D,E,G)×P(G);

[0106] Suppose we want to calculate the joint probability when the power generation cost C=C1, power generation capacity V=V1, power generation power P=P1, price adjustment A=A1, power consumption capacity D=D1, expected price E=E1, power purchase decision B=B1, and geographical location G=G1. The calculation process is similar to the above and will not be repeated here.

[0107] Assume that we know the values ​​of the above probabilities, which are p1, v 11 、p 111 、a 1111 , d1, e1, b 1111 , g1 (here g1 represents the prior probability that the geographical location is G1), then the specific joint probability value can be calculated as:

[0108] F=p1×v 11 ×p 111 ×a 1111 ×d1×e1×b 1111 ×g1

[0109] (4) Calculate the posterior probability:

[0110] The posterior probability is calculated using Bayes' theorem, which is expressed as:

[0111]

[0112] To calculate P(C│A,B,D,E,G), first calculate the numerator P(C,A,B,D,E,G):

[0113] P(C,A,B,D,E,G)=P(C)×P(A│C)×P(B│A,D,E,G)×P(D)×P(E)×P(G)

[0114] Assume that C has n values, then the denominator ∑ J P(X J ,Y1,Y2,…,Y m ) is to sum all the values ​​of C:

[0115]

[0116] For example, to calculate the posterior probability that the power generation cost is C1 when the price adjustment A=A1, the power purchase decision B=B1, the power consumption capacity D=D1, the expected price E=E1, and the geographical location G=G1, the calculation method is similar to the above method and will not be repeated here.

[0117] After the above calculation process, we can obtain the posterior probability between different variables, and thus analyze the transaction correlation between producers and consumers.

[0118] The value of P(C=C1│A=A1,B=B1,D=D1,E=E1,G=G1) is large, which means that when the price is increased, the power purchase decision is to purchase more, the power consumption capacity is low demand, the expected price is the bottom price expectation and the geographical location is G1, the possibility of power generation cost being C1 is high, which reflects the correlation between these factors.

[0119] S14. Determine the target transaction supply side and the corresponding transaction strategy based on the transaction relevance.

[0120] In this embodiment, the demand side determines the target transaction strategy according to the transaction correlation of distributed energy consumers, including:

[0121] 1. Determine the transaction correlation index through the transaction correlation with distributed energy consumers, including price correlation index, electricity correlation index and geographical location correlation index;

[0122] Specifically, price correlation indicators include price elasticity coefficient and price matching degree. The price elasticity coefficient is calculated by the ratio of the percentage change in the amount of electricity purchased by consumers to the percentage change in price; the price matching degree is calculated by the ratio of the difference between the producer's quotation and the consumer's expected price to the expected price.

[0123] Electricity correlation indicators include the matching rate of power generation capacity and power demand and the synchronization rate of power generation and power consumption. The matching rate of power generation capacity and power demand is the proportion of the producer's power generation capacity that can meet the consumer's power demand; the synchronization rate of power generation and power consumption is the proportion of time when the producer's power generation peak period overlaps with the consumer's power consumption peak period within a certain period of time.

[0124] Geographical location correlation indicators include transmission distance and regional supply and demand balance. The transmission distance is calculated by the geographical distance between producers and consumers; the regional supply and demand balance is the ratio of the total power supply in the production area to the total electricity demand of consumers in the area.

[0125] 2. Calculate the comprehensive weight of transaction correlation based on the price correlation index, power correlation index and geographical location correlation index;

[0126] Comprehensive score calculation:

[0127] Assume that the weight vector of each correlation index after weight distribution is W = (w1, w2, ..., w m ), the indicator value vector is V = (v1, v2, ..., v m ), then the transaction correlation comprehensive score

[0128] 3. Divide the transaction level according to the comprehensive weight, and determine the target transaction supply side and corresponding transaction strategy based on the transaction level.

[0129] Finally, the transaction level is divided according to the comprehensive score, for example, it can be divided into three levels:

[0130] If the high level S≥0.8, it means that the supply side and the demand side that meet this classification level have a strong transaction correlation, the possibility of successful transaction is high, and the two sides have a good match in terms of price, electricity, geographical location, etc., which is suitable for giving priority to the transaction. The demand side can then choose the appropriate transaction strategy based on the determined supply side.

[0131] For the medium level 0.4≤S<0.8, the demand side and the supply side that meets this classification level have a good match in some indicators, but there may be certain issues that need to be coordinated, such as slightly higher prices or unstable electricity supply in certain periods of time.

[0132] If the low level S is less than 0.4, it may indicate that there is a large price gap between the demand side and the supply side that meets the classification level, which may be caused by reasons such as mismatch between electricity supply and demand or remote geographical location.

[0133] Embodiment 2

[0134] An embodiment of a power distribution network distributed trading system of the present invention comprises the following steps:

[0135] A data acquisition and processing unit is used for the distributed energy demand side to acquire the power generation supply information data released by each distributed energy supply side, and to organize and clean the power generation supply information data;

[0136] The Bayesian network structure construction unit is used to set the power demand variables at the distributed energy demand side, and determine the node variables of the Bayesian network based on the sorted and cleaned power supply information data and power demand variables, determine the correlation between the node variables, construct the Bayesian network structure and perform parameter learning;

[0137] A transaction relevance analysis unit, used to analyze the transaction relevance with each distributed energy supply end based on the constructed Bayesian network structure and learned parameters;

[0138] The target transaction supply end and the transaction strategy determination unit are used to determine the target transaction supply end and the corresponding transaction strategy based on the transaction relevance.

[0139] Embodiment 3

[0140] The present invention provides a computer device, comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the above method are implemented.

[0141] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0142] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored. In addition, each functional unit in each embodiment of the present invention can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

[0143] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0144] It can be understood that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for distributed trading of electric energy in a distribution network, characterized in that: include: The distributed energy demand side obtains the power generation supply information data released by each distributed energy supply side, and sorts and cleans the power generation supply information data; The distributed energy demand side sets the power demand variable, and determines the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variable, determines the correlation between the node variables, constructs the Bayesian network structure and performs parameter learning; Analyze the transaction correlation with each distributed energy supply end according to the constructed Bayesian network structure and the learned parameters; A target transaction supply end and a corresponding transaction strategy are determined based on the transaction relevance.

2. The method for distributed power trading in a distribution network according to claim 1, characterized in that: The step of arranging and cleaning the power generation supply information data includes: Power generation supply information data collation includes data time series alignment, data format conversion, data classification and labeling, and power generation supply information data cleaning includes missing value processing, duplicate value processing, and outlier detection and processing.

3. The method for distributed power trading in distribution network according to claim 1, characterized in that: The distributed energy demand side sets the power demand variable, and determines the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variable, determines the correlation between the node variables, constructs the Bayesian network structure and performs parameter learning, including: Filter node variables based on the sorted and cleaned power supply information data; Analyze the causal relationship and dependency relationship in the data based on the selected node variables and the set power demand variables, determine the directed edges between the node variables, and construct the Bayesian network structure; Based on the constructed Bayesian network structure and the historical transaction data of the distributed energy supply side, the statistical learning method is used to estimate the conditional probability table of each node in the network.

4. The method for distributed power trading in distribution network according to claim 3, characterized in that: The node variables in the power generation supply information data include power generation cost, power generation capacity, power generation power and price adjustment, and the power demand variables set at the distributed energy demand end include power capacity, expected price, power purchase decision and geographical location.

5. The method for distributed power trading in distribution network according to claim 3, characterized in that: Based on the constructed Bayesian network structure and the historical transaction data of the distributed energy supply side, the conditional probability table of each node in the network is estimated using a statistical learning method, including: Calculate the prior probability of each node in different states based on the historical transaction data of distributed energy consumers and the historical transaction data corresponding to the set electricity demand variables; The conditional probability of each node in the state of its parent node is calculated based on the historical transaction data of distributed energy consumers and the conditional information of the historical transaction data corresponding to the set electricity demand variable.

6. The method for distributed power trading in distribution network according to claim 1, characterized in that: The transaction correlation between the constructed Bayesian network structure and the learned parameters and the distributed energy supply terminals includes: Perform inference calculations based on the constructed Bayesian network structure and learned parameters; Based on the results of inference calculation, the transaction impact and correlation between distributed energy consumers are analyzed.

7. The method for distributed power trading in a distribution network according to claim 1, characterized in that: The determining of the target transaction supply end and the corresponding transaction strategy based on the transaction relevance includes: Determining transaction relevance indicators through transaction relevance with distributed energy consumers, the indicators including price relevance indicators, power relevance indicators and geographic location relevance indicators; Calculating a comprehensive transaction relevance weight based on the price relevance index, the power relevance index, and the geographic location relevance index; The transaction levels are divided according to the comprehensive weights, and the target transaction strategy is determined based on the transaction levels.

8. A distributed power trading system for distribution network, characterized in that: include: A data acquisition and processing unit, used for the distributed energy demand side to acquire the power generation supply information data released by each distributed energy supply side, and to sort and clean the power generation supply information data; A Bayesian network structure construction unit is used to set power demand variables at the distributed energy demand side, and determine the node variables of the Bayesian network based on the sorted and cleaned power supply information data and the power demand variables, determine the correlation between the node variables, construct the Bayesian network structure and perform parameter learning; A transaction relevance analysis unit, used for analyzing the transaction relevance with each distributed energy supply end according to the constructed Bayesian network structure and the learned parameters; The target transaction supply end and the transaction strategy determination unit are used to determine the target transaction supply end and the corresponding transaction strategy based on the transaction relevance.

9. A computer device, characterized in that: include: memories, transceivers, processors, and bus systems; Wherein, the memory is used to store programs; The processor is used to execute the program in the memory, including executing the distribution network electric energy distributed trading method according to any one of claims 1 to 7; The bus system is used to connect the memory and the processor so that the memory and the processor can communicate with each other.

10. A computer-readable storage medium, characterized in that: The method comprises instructions which, when executed on a computer, enable the computer to execute the method for distributed trading of electric energy in a power distribution network as claimed in any one of claims 1 to 7.

Images