Residential Load Power Consumption Period Prediction Method Based on Bayesian Network and Fuzzy Clustering Analysis

Abstract

The present invention discloses a method for predicting the electricity consumption period of residential load based on Bayesian network and fuzzy clustering analysis. The method includes: establishing a Bayesian network model with the corresponding relationship representing external environmental factors, user historical electricity consumption data, and the usage status of the load in each period; obtaining the external environmental factors affecting the residential load usage and the historical electricity consumption data of the load. The external environmental factors include environmental temperature and date information, and discretize the relevant data; obtaining the prior probability and conditional probability of the load opening status in different periods, calculating the posterior probability of the opening status in each period through the Bayesian formula to obtain the electricity consumption state probability matrix of the load, that is, the possibility of the load being used in each period; using a partial large semi-gradient membership function to conduct a fuzzy evaluation of the membership degree of the electricity consumption probability of the load in each period; using the fuzzy clustering analysis method to determine the electricity consumption time range of the load.

Description

Technical Field

[0001] The present invention relates to the field of intelligent home energy management, and particularly to a method for predicting the electricity consumption time period of residential loads based on Bayesian network and fuzzy clustering analysis. Background Art

[0002] Driven by economic development, the electricity consumption of residents in China has increased significantly, resulting in more tense power supply during peak hours. At the same time, residential loads have the characteristics of rich response resources, strong controllability, and large aggregation potential. The development of smart grid technology has made it possible for flexible resources on the residential side to participate in grid regulation. Flexible resources on the residential side, such as air conditioners, water heaters, and electric vehicles, have gradually been used in fields such as peak shaving and valley filling of the power system and new energy consumption.

[0003] In the environment of smart grid, various uncertain factors have a great impact on the implementation effect of the grid demand response scheduling strategy. Summary of the Invention

[0004] Objective: To overcome the deficiencies in the prior art, in view of the problem of uncertainty in residential load electricity consumption, the present invention provides a method and device for predicting the electricity consumption time period of residential loads based on Bayesian network and fuzzy clustering analysis; considering the time characteristics of residential loads and the correlation with external environmental factors, a probability model for the energy consumption of residential flexible resources is constructed based on Bayesian network, and the time period of flexible resource use is predicted by means of fuzzy clustering. Based on this, it is possible to avoid the influence on the comfort of residents caused by excessive participation of loads in scheduling during the scheduling process, and provide certain support for evaluating the potential of residential load demand response and participating in demand response scheduling.

[0005] Technical Solution: To solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0006] In the first aspect, a method for predicting the electricity consumption time period of residential loads based on Bayesian network and fuzzy clustering analysis is provided, including:

[0007] Obtain the external environmental factors affecting the use of residential loads and the historical electricity consumption data of the loads, and perform discretization processing on the external environmental factors and the historical electricity consumption data of the loads; wherein the external environmental factors include the temperature of the electricity consumption environment and the corresponding date information;

[0008] Based on the discretized external environmental factors and the historical electricity consumption data of the loads, use the pre-constructed Bayesian network model to obtain the prior probability and conditional probability of the load opening state at different time periods, calculate the posterior probability of the opening state at each time period, and obtain the electricity consumption state probability matrix of the load, that is, the possibility of the load being used at each time period; wherein the established Bayesian network model can represent the corresponding relationship between the external environmental factors, the historical electricity consumption data of the user load, and the use state of the load at each time period;

[0009] Based on the probability matrix of the load power consumption state, a fuzzy evaluation is carried out on the membership degree of the power consumption probability of each period of the load by using a semi-gradient membership function of the right-skewed type;

[0010] Based on the membership degree of the power consumption probability of each period of the load obtained by the evaluation, a fuzzy clustering analysis method is used to determine the load power consumption time range.

[0011] In some embodiments, the construction method of the Bayesian network model includes: taking the load historical usage state, the power consumption environment temperature information, the corresponding date information, and the load on / off state of each period as the node variables of the Bayesian network model, and building the Bayesian network model according to the correlation relationship between the variables.

[0012] In some embodiments, the discretization process includes:

[0013] Perform discretization processing on the load historical power consumption data, The sample set representing the usage situation of device a by the user on the i-th day, Indicating that the device is off, Indicating that the device is on, divide a day into T periods, t∈[1,2,3,4,…,T] represents any period in a day, d n Represents the usage situation matrix of device a by the user within i days;

[0014]

[0015] F = {F i,1 , F i,2 , F i,3 , F i,4 , …, F i,T} represents the power consumption environment temperature information of the user at different periods on the i-th day, F n Represents the power consumption environment temperature matrix of the user at different periods within i days;

[0016]

[0017] Divide the attribute range of the date variable:

[0018] The date variable X = {1,2,3,4,5,6,7,8}: 1 - 7 represent seven days from Monday to Sunday, and 8 represents holidays.

[0019] In some embodiments, obtaining the prior probability and conditional probability of the load on / off state at different periods includes:

[0020] According to the historical power consumption data of the user's power load in the previous i days and the load usage state at each period, that is, the external environment data, calculate the load state The frequency proportion of the total sample of the frequency is used to obtain the prior probability

[0021] According to the conditional independence judgment of the Bayesian network, given the state of the electrical equipment, the external factors are independent of each other; therefore, calculate the conditional probability function:

[0022]

[0023] Where is the conditional probability that the external factors F, D, and X appear simultaneously in ; F, D, and X represent the temperature of the electrical environment, historical electricity consumption data, and date, respectively.

[0024] In some embodiments, calculate the posterior probability of the on-state in each time period to obtain the load electricity consumption state probability matrix, including:

[0025] For the electricity consumption state of the load at time t on the (i + 1)-th day Calculate the posterior probability Indicates the likelihood of the load equipment being turned on during this time period:

[0026]

[0027] In the formula: is the probability that the external factors F, D, and X appear simultaneously in , Indicates The prior probability of, obtained from the frequency ratio of the load state frequency to the total sample frequency, and P(F, D, X) represents the probability that the external factors F, D, and X appear simultaneously;

[0028] Repeat the above steps to calculate the likelihood of the home appliance equipment being turned on in each time period of a day, so as to obtain the load electricity consumption state probability matrix P a =[P a,1 P a,2 … P a,T , where P a,t is the electricity consumption probability of the electrical equipment a at time t.

[0029] In some embodiments, a fuzzy evaluation of the membership degree of the electricity consumption probability of the load in each time period is performed using a partial large semi-gradient membership function, including:

[0030] Use a partial large semi-gradient membership function to calculate the membership degree u(P a,t ) of time period t with respect to the electricity consumption time period:

[0031]

[0032] In the formula: P a,tis the power consumption probability of the electrical device a in the time period t; n is the maximum value of the power consumption probabilities of the electrical device a in each time period; m is the minimum value of the power consumption probabilities of the electrical device a in each time period.

[0033] In some embodiments, a fuzzy clustering analysis method is used to determine the load power consumption time range, including:

[0034] 1) Taking each time period as the classification object and the membership degree of each time period as the statistical index, a data vector E = [e1 e2… e T T ;

[0035] 2) Standardize the data vector:

[0036]

[0037]

[0038]

[0039] In the formula: e′ t is the intermediate parameter, e t represents the membership degree of the power consumption time period of the time period t, represents the mean value of the membership degrees of the power consumption time periods of T time periods; v represents the variance of the membership degrees of the power consumption time periods of T time periods; t = 1, 2,…, T;

[0040] 3) Use the Euclidean distance method to establish a fuzzy similarity matrix;

[0041]

[0042] In the formula: x m , x n respectively represent any two data after the above standardization; r mn is the parameter of the constructed fuzzy similarity matrix; m, n = 1, 2,…, T; H is a determined constant such that all r mn ∈ [0, 1];

[0043] 4) Solve the quadratic equation for the calibrated fuzzy similarity matrix, that is, R → R 2 → … → R 2j → …, j = [1, 2, 3,…] is a constant, until R k oR k = R k , k = [1, 2,…, 2j,…], R k is the required transitive closure t(R);

[0044] Solve the λ-cut matrix of t(R), and there is:

[0045] R​λ = (λr mn ) T×T

[0046]

[0047] where: λ ∈ [0, 1] is a constant; m, n = 1, 2, …, T;

[0048] Let λ gradually decrease from 1, and classify each time period according to the corresponding R λ Take the number of clusters as 2 to obtain the clustering set T of the electricity consumption time period Y and the clustering set T of the non - electricity consumption time period N .

[0049] In a second aspect, the present invention provides a device for predicting the electricity consumption time period of residential load based on Bayesian network and fuzzy clustering analysis, including a processor and a storage medium;

[0050] The storage medium is used to store instructions;

[0051] The processor is used to operate according to the instructions to execute the steps of the method according to the first aspect.

[0052] In a third aspect, the present invention provides a storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method according to the first aspect are realized.

[0053] Beneficial effects: The method and device for predicting the electricity consumption time period of residential load based on Bayesian network and fuzzy clustering analysis provided by the present invention have the following advantages: The present invention takes into account the time characteristics of residential load and its correlation with external environmental factors, constructs a probability model of residential flexible resource energy use based on Bayesian network, and predicts the usage time period of flexible resources through fuzzy clustering. Based on this, it can avoid affecting the comfort of residents due to excessive participation of load in scheduling during the scheduling process, and provides certain support for evaluating the potential of residential load demand response and participating in demand response scheduling. Brief description of the drawings

[0054] Figure 1 is the flowchart of predicting the electricity consumption time period of the load in the embodiment of the present invention.

[0055] Figure 2 is the Bayesian network model diagram in the embodiment of the present invention.

[0056] Figure 3 is the probability distribution diagram of the electricity consumption of the washing machine in each time period in the embodiment of the present invention. Detailed implementation manners

[0057] To make the objectives, technical solutions, and advantages of the present invention more clear, the present invention will be 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 used to limit the present invention, that is, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0058] In the description of the present invention, the meaning of several is more than one, the meaning of multiple is more than two, greater than, less than, exceeding, etc. are understood as not including the number itself, above, below, within, etc. are understood as including the number itself. If there is a description of first and second, it is only for the purpose of distinguishing technical features and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features or implicitly indicating the sequence relationship of the indicated technical features.

[0059] In the description of the present invention, the description with reference to terms such as "one embodiment", "some embodiments", "schematic embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0060] Embodiment 1

[0061] A method for predicting the electricity consumption period of residential load based on Bayesian network and fuzzy clustering analysis, comprising:

[0062] Obtain the external environmental factors affecting the use of residential load and the historical electricity consumption data of the load, and perform discretization processing on the external environmental factors and the historical electricity consumption data of the load; wherein the external environmental factors include the electricity consumption environment temperature and the corresponding date information;

[0063] Based on the discretized external environmental factors and the historical electricity consumption data of the load, use the pre-constructed Bayesian network model to obtain the prior probability and conditional probability of the load opening state at different times, calculate the posterior probability of the opening state at each time, and obtain the load electricity state probability matrix, that is, the possibility of the load using at each time; wherein the establishment of the Bayesian network model can represent the corresponding relationship between the external environmental factors, the historical electricity consumption data of the user load, and the use state of the load at each time;

[0064] Based on the load electricity state probability matrix, use the partial large semi-gradient membership function to perform fuzzy evaluation on the membership degree of the electricity consumption probability of each time period of the load;

[0065] Based on the membership degrees of the electricity consumption probabilities of the load in each period obtained from the evaluation, the fuzzy clustering analysis method is used to determine the electricity consumption time range of the load.

[0066] In some specific embodiments, this embodiment uses the historical load electricity consumption data and external environment data of an electric water heater collected within 64 days before the prediction date of a certain user as the training sample set, and predicts the user's usage situation on the 65th day based on the load electricity consumption period prediction method. This method includes the following steps:

[0067] Step 1: Determine the network nodes in the Bayesian network model, and establish a Bayesian network model with the corresponding relationship between external environment factors, historical electricity consumption data of the user load, and the usage status of the load in each period;

[0068] Specifically in this example, Step 1 is refined as:

[0069] Take the historical usage status of the load, the electricity environment temperature information, the corresponding date information, and the load on - off status in each period as the node variables of the Bayesian network model. Build a Bayesian network model according to the correlation relationship between the variables. The directed edges connecting these nodes in the figure describe the causal relationship between the historical usage status of the load, the electricity environment temperature information, the corresponding date information, and the load on - off status in each period. The established Bayesian network topology structure diagram is as Figure 2 shown;

[0070] Step 2: Obtain the external environment factors affecting the use of the residential load and the historical electricity consumption data of the load. The external environment factors include the electricity environment temperature and date information, and discretize the relevant data;

[0071] Specifically in this example, Step 2 is refined as:

[0072] The user behavior of the electric water heater mainly focuses on the bathing time. Combining the bathing time in the actual situation, divide 24 periods to predict the behavior of the user using the electric water heater at different times of the day. Discretize the collected load historical electricity consumption data and external environment factor data at 1 - hour time intervals. Discretize the load historical electricity consumption data. Let \(t\in[1,2,3,4,\cdots,24]\) represent any period of a day, represents the sample set of the usage situation of the electric water heater equipment of the user on the \(i\) - th day, represents the equipment is closed, represents the equipment is turned on, \(d\) n represents the usage situation matrix of the electric water heater equipment of this user within 64 days.

[0073]

[0074] F = {F i,1 ,Fi,2 , F i,3 , F i,4 , …, F i,24} represents the power consumption ambient temperature information of the user at different time periods on the i-th day, F n represents the power consumption ambient temperature information of the user at different time periods within 64 days.

[0075]

[0076] Divide the attribute range of the date variable:

[0077] X = {1, 2, 3, 4, 5, 6, 7, 8}: 1 - 7 represent the seven days from Monday to Sunday, and 8 represents holidays.

[0078] Step 3: Calculate the posterior probability of the on-state at each time period through Bayes' formula:

[0079] Specifically in this example, Step 3 is refined as:

[0080] Calculate the frequency proportion of the electric water heater being in the on-state within 64 days in the total sample frequency to obtain the prior probability

[0081] According to the conditional independence determination of the Bayesian network, given the state of the electrical equipment, the external factors are independent of each other. Therefore, calculate the conditional probability function:

[0082]

[0083] Collect the historical power consumption data of the user's power load in the previous 64 days, the usage status of the load at each time period, that is, the external environment data, and for the power consumption status of the load at time period t on the 65th day calculate the posterior probability, which represents the likelihood of the load equipment being on in this time period:

[0084]

[0085] In the formula: is the probability that the external factors F, D, X appear simultaneously in , represents the prior probability of, which can be obtained according to the frequency proportion in the total sample frequency, and P(F, D, X) represents the probability that the external factors F, D, X appear simultaneously.

[0086] Repeat the above steps, calculate the likelihood of the home appliance equipment being on at each time period in a day to obtain the power consumption status probability matrix P of the equipment for one day a = [P WH,1 P WH,2 … PWH,24 , the calculated electricity consumption probabilities of the washing machine at different times of a day are as Figure 3 shown.

[0087] Step 4: Use a semi-gradient membership function of the larger type to perform fuzzy evaluation on the membership degrees of the electricity consumption probabilities at each time period of the load;

[0088] Specifically in this example, Step 4 is refined as follows:

[0089] Use a semi-gradient membership function of the larger type to calculate the membership degree u(P WH,t ) of time period t with respect to the electricity consumption time period:

[0090]

[0091] In the formula: P WH,t is the electricity consumption probability of the electric water heater device at time period t; n is the maximum value of the electricity consumption probabilities of the electric water heater devices at each time period; m is the minimum value of the electricity consumption probabilities of the electric water heater devices at each time period.

[0092] Step 5: Use the fuzzy clustering analysis method to determine the electricity consumption time range of the load:

[0093] Specifically in this example, Step 5 is refined as follows:

[0094] Taking each time period as the classification object and the membership degree of each time period as the statistical index, a data vector is obtained: E = [e1 e2… e 24 24 ;

[0095] Perform standardization processing on the data vector:

[0096]

[0097]

[0098] In the formula: e t ' is an intermediate parameter, e t represents the membership degree of the electricity consumption time period of time period t, represents the mean value of the membership degrees of the electricity consumption time periods of 24 time periods; v represents the variance of the membership degrees of the electricity consumption time periods of 24 time periods; t = 1, 2,…, 24;

[0099] Use the Euclidean distance method to establish a fuzzy similarity matrix;

[0100]

[0101] In the formula: x m , x n respectively represent the data after the above standardization processing; r mn ​is a parameter of the constructed fuzzy similarity matrix; m, n = 1, 2, …, 24; H is a determined constant such that all r mn ∈ [0, 1].

[0102] Solve the quadratic equation for the calibrated fuzzy similarity matrix, i.e., R → R 2 → … → R 2j → …, j = [1, 2, 3, …] is a constant until R k oR k = R k , k = [1, 2, …, 2j, …], R k which is the required transitive closure t(R).

[0103] Solve the λ-cut matrix of t(R), and we have:

[0104] R λ = (λr mn ) T×T

[0105]

[0106] where: λ ∈ [0, 1] is a constant; m, n = 1, 2, …, 24.

[0107] Let λ gradually decrease from 1, and classify each time period according to the corresponding R λ , taking the number of clusters as 2, to obtain the clustering sets T Y and T N . It is calculated that the electricity consumption time periods of the washing machine throughout the day are between 6:00 - 8:00 at noon and 20:00 - 22:00 at night.

[0108] So far, from step 1 to step 5, a method for predicting the electricity consumption time periods of residential loads based on Bayesian network and fuzzy clustering analysis is completed.

[0109] Example 2

[0110] In the second aspect, this embodiment provides a device for predicting the electricity consumption time periods of residential loads based on Bayesian network and fuzzy clustering analysis, including a processor and a storage medium;

[0111] The storage medium is used to store instructions;

[0112] The processor is used to operate according to the instructions to execute the steps of the method according to Example 1.

[0113] Example 3

[0114] Thirdly, this embodiment provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in Embodiment 1 are implemented.

[0115] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0116] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the specified functions in one Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0117] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the specified functions in one Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for implementing the specified functions in one Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0119] The above description is only a preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present invention.

Claims

1. A method for predicting the electricity consumption period of residential load based on Bayesian network and fuzzy clustering analysis, characterized in that, The method includes: Obtaining external environmental factors affecting residents' load usage and historical load electricity consumption data, and performing discretization processing on the external environmental factors and the historical load electricity consumption data; wherein the external environmental factors include the electricity usage environment temperature and the corresponding date information; Based on the discretized external environmental factors and the historical electricity consumption data of the load, use the pre-constructed Bayesian network model to obtain the prior probability and conditional probability of the load's on-state at different times, calculate the posterior probability of the on-state at each time, and obtain the load electricity state probability matrix, that is, the likelihood of the load being used at each time; wherein the established Bayesian network model can represent the corresponding relationship between the external environmental factors, the historical electricity consumption data of the user load, and the usage state of the load at each time; the construction method of the Bayesian network model includes: taking the historical usage state of the load, the electricity environment temperature information, the corresponding date information, and the load on-state at each time as the node variables of the Bayesian network model, and building the Bayesian network model according to the correlation between the variables; obtaining the prior probability and conditional probability of the load's on-state at different times, including: according to the user's electricity load in the previous i days of historical electricity consumption data, the usage state of the load at each time, that is, the external environmental data, calculate the load state The frequency proportion of the total sample is used to obtain the prior probability where represents that device a is turned on on the i-th day of the user; according to the conditional independence judgment of the Bayesian network, given the state of the electrical equipment, the external factors are independent of each other; therefore, calculate the conditional probability function: Among them is the conditional probability of the simultaneous occurrence of external factors F, D, and X in ; F, D, and X respectively represent the temperature of the electricity usage environment, historical electricity usage data, and date Based on the load electricity state probability matrix, using a semi-gradient membership function of the larger type to perform fuzzy evaluation on the membership degrees of the electricity consumption probabilities of each load time period; Based on the membership degrees of the electricity consumption probabilities of each load time period obtained from the evaluation, using a fuzzy clustering analysis method to determine the load electricity usage time range, wherein the fuzzy clustering analysis method takes each time period as the classification object and the membership degree of each time period as the statistical index.

2. The method for predicting the electricity consumption period of residential load based on Bayesian network and fuzzy clustering analysis according to claim 1, characterized in that, Performing discretization processing includes: Discretize the load historical power consumption data, The sample set representing the usage of device a by the user on the i-th day, Indicates that the device is off, Indicates that the device is on. Divide a day into T time periods, and t ∈ [1, 2, 3, 4, …, T] represents any time period in a day, d n Represents the usage matrix of device a by the user within i days; F = {F i,1 , F i,2 , F i,3 , F i,4 , …, F i,T} represents the ambient temperature information of the user's electricity consumption at different times on the i-th day, and F n represents the ambient temperature matrix of the user's electricity consumption at different times within the i-th day; Dividing the attribute range of the date variable: The date variable X = {1, 2, 3, 4, 5, 6, 7, 8}: 1 - 7 represent the seven days from Monday to Sunday, and 8 represents a holiday.

3. A method for predicting the electricity consumption period of residential loads based on Bayesian network and fuzzy clustering analysis according to claim 1, characterized in that, Calculating the posterior probability of the on-state of each time period to obtain the load electricity state probability matrix, including: The power consumption status of the i+1th day and the tth period of the load Calculate the posterior probability Indicates the likelihood of the load equipment being turned on during this period: In the formula: is the probability that external factors F, D, X appear simultaneously in , represents prior probability, which is obtained according to the frequency proportion of the load state frequency in the total sample frequency, and P(F, D, X) represents the probability that external factors F, D, X appear simultaneously; Repeat the above steps to calculate the likelihood of a household electrical appliance being turned on at each time period in a day, thereby obtaining the probability matrix P of the load power consumption status of the device for one day a = [P a,1 P a,2 … P a,T , where P a,t is the power consumption probability of electrical appliance a at time period t 4. A method for predicting the electricity consumption period of residential loads based on Bayesian network and fuzzy clustering analysis according to claim 1, characterized in that Using a semi-gradient membership function of the larger type to perform fuzzy evaluation on the membership degrees of the electricity consumption probabilities of each load time period, including: The membership degree u(P a,t ) of time period t with respect to the electricity consumption period is calculated using a semi-gradient membership function of the larger type: Where: P a,t is the power consumption probability of the electrical equipment a during the time period t; n is the maximum value among the power consumption probabilities of the electrical equipment a in each time period; m is the minimum value among the power consumption probabilities of the electrical equipment a in each time period.

5. A method for predicting the electricity consumption period of residential load based on Bayesian network and fuzzy clustering analysis according to claim 1, characterized in that Using a fuzzy clustering analysis method to determine the load electricity usage time range, including: 1) Taking each time period as the classification object and the membership degree of each time period as the statistical index, a data vector E = [e1 e2…e T T ;​ 2) Standardizing the data vector: where: e' t is an intermediate parameter, and e t represents the membership degree of the electricity consumption period at time period t, represents the mean value of the membership degrees of the electricity consumption periods in T time periods; v represents the variance of the membership degrees of the electricity consumption periods in T time periods; t = 1, 2, …, T; 3) Establishing a fuzzy similarity matrix using the Euclidean distance method; where: x m , x n respectively represent any two data after the above standardization processing; r mn is the parameter of the constructed fuzzy similarity matrix; m, n = 1, 2, …, T; H is a determined constant such that all r mn ∈[0, 1]; 4) Solve the quadratic equation for the already calibrated fuzzy similarity matrix, i.e., \(R\rightarrow R^{2}\) 2 \(\rightarrow\cdots\rightarrow R^{j}\), where \(j = [1, 2, 3,\cdots]\) is a constant, until 2j \(R^{j}=R^{j + 1}\) appears for the first time \(R^{j}\) k is the required transitive closure \(t(R)\); Solving the λ-cut matrix of t(R), there is: R λ = (λr mn ) T×T In the formula: λ ∈ [0, 1] is a constant; m, n = 1, 2,..., T; Let λ gradually decrease from 1, according to the corresponding R λ Classify each time period, take the number of clusters as 2, and obtain the clustering set T of the power consumption time period Y and the clustering set T of the non-power consumption time period N .

6. A device for predicting the electricity consumption period of residential loads based on Bayesian network and fuzzy clustering analysis, characterized in that, Including a processor and a storage medium; The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the steps of the method according to any one of claims 1 to 5.

7. A storage medium, on which a computer program is stored, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 5.

Images