Power facility power intelligent dispatching method and system

Abstract

The present invention provides a method for intelligent dispatching of electric power of electric power facilities, which comprises the following steps: S1, obtaining the electric power operation data under dispatching in a certain area within a certain time; S2, screening and preprocessing the electric power operation data and sending it to a background server; S3, the background server inputs the screened electric power operation data into a prediction model, and obtains the electric power operation dispatching instruction of the next time through the prediction model; S4, intelligently dispatching the electric power of the electric power facilities through the electric power operation dispatching instruction. The present invention also provides an electric power intelligent dispatching system for electric power facilities.

Description

Technical Field

[0001] The present invention relates to the technical field of power facility dispatching, and in particular to a method and system for intelligent power dispatching of power facilities. Background Art

[0002] Electric energy is an indispensable energy source in national production and people's daily life. The importance of power production and power transmission is beyond doubt. As the intermediate link between power production and power transmission, power dispatching determines the effective and healthy operation of the entire power grid. The automation and intelligence of power dispatching is a long-term work worthy of in-depth study. However, the current power automation dispatching system mainly serves the front desk of power production, and the dispatching and production management work in the background has always lacked complete technical support. How to coordinate the front desk control and back desk management in the power production process, build an organic and unified comprehensive intelligent system, form a complete and smooth information flow, and finally realize the integration of production planning, production process, and power sales, standardization, and improve the quality of power grid management has become an urgent problem to be solved. With the development of the scale of the power grid, the number of power plants and stations continues to increase, the amount of data in the power system increases sharply, and the difficulty of data analysis and management continues to increase; at the same time, in addition to traditional power dispatching technology and equipment, there are no other good technical solutions and devices that keep pace with the times. Summary of the invention

[0003] In view of this, the present invention proposes an electric power intelligent dispatching system and method for electric power facilities, which can realize fully automatic intelligent dispatching of electric power of electric power facilities.

[0004] A method for intelligent dispatching of electric power facilities, comprising the following steps:

[0005] S1. Obtain the power operation data under dispatch in a certain area within a certain period of time;

[0006] S2, screening and preprocessing the power operation data and sending it to the backend server;

[0007] S3. The backend server inputs the filtered power operation data into the prediction model, and obtains the power operation dispatch instruction for the next time through the prediction model;

[0008] S4. Intelligently dispatch the power of power facilities through power operation dispatch instructions.

[0009] In the electric power intelligent dispatching method for electric power facilities described in the present invention,

[0010] After step S4, the following steps are also included:

[0011] After intelligent dispatch, collect the power operation data under dispatch; obtain manual evaluation of the power operation data; and modify the prediction model based on the evaluation.

[0012] In the electric power intelligent dispatching method for electric power facilities described in the present invention,

[0013] The step S3 comprises:

[0014] The backend server inputs the screened power operation data as samples into the prediction model;

[0015] For each sample, by transferring the model P(X t+1 |X t ), the current state value of the sample X t Under the condition, for the next state value X t+1 Sampling is performed so that the samples propagate forward;

[0016] For each sample, by assigning the probability value P(e) of the new state to the sample t+1 |X t ) as weight;

[0017] Resample the population sample to generate a new N sample population, each new sample is selected from the current population; the probability of each specific sample being selected is proportional to the weight of the specific sample;

[0018] Set the sample population to start with the forward message f at time t 1:t =P(X t |e1: t ) is the correct representation; use N(X t |e 1:t ) means that after fully processing the observation, it has state X t The number of samples, so for N that is large enough to meet the requirements, the following relationship can be obtained:

[0019] N(X t |e 1:t )÷N=P(X t |e 1:t );

[0020] Given a sample condition at time t, propagate each sample forward by sampling the state variable corresponding to time t+1; from each X t Arrive at X t+1 The number of samples is equal to the transition probability multiplied by X t The total amount, so we reach state X t+1 The total number of samples is as follows:

[0021] N(X t+1 |e 1:t )=∑P(X t+1 |X t )N(X t |e 1:t );

[0022] Assign weights to samples based on the probability of each sample at time t+1; t+1 The sample gets the weight P(e t+1 X t+1 ); when evidence e is observed t+1 After that, it is in state X t+1 The total sample weights are as follows:

[0023] W(X t+1 |e 1:t+1 )=P(e t+1 |X t+1 )N(X t+1 |e 1:t );

[0024] The sampling step is repeated to obtain the cyclic sample population to represent the forward message at time t+1.

[0025] The present invention also provides an electric power intelligent dispatching system for electric power facilities, which includes the following units:

[0026] An operation data acquisition unit is used to acquire the power operation data under dispatch in a certain area within a certain period of time;

[0027] A data preprocessing and sending unit, used to filter and preprocess the power operation data and send it to the backend server;

[0028] The prediction analysis unit is used to input the screened power operation data into the prediction model through the background server, and obtain the power operation dispatch instruction at the next time through the prediction model;

[0029] The dispatching unit is used to intelligently dispatch the power of the power facilities through power operation dispatching instructions.

[0030] In the electric power intelligent dispatching system for electric power facilities described in the present invention,

[0031] The scheduling unit then further includes:

[0032] The correction unit is used to collect the power operation data under the dispatch after the intelligent dispatch; obtain the manual evaluation of the power operation data; and correct the prediction model according to the evaluation.

[0033] In the electric power intelligent dispatching system for electric power facilities described in the present invention,

[0034] The prediction analysis unit comprises:

[0035] The backend server inputs the screened power operation data as samples into the prediction model;

[0036] For each sample, by transferring the model P(X t+1 |X t ), the current state value of the sample X t Under the condition, for the next state value X t+1 Sampling is performed so that the samples propagate forward;

[0037] For each sample, by assigning the probability value P(e) of the new state to the sample t+1 |X t ) as weight;

[0038] Resample the population sample to generate a new N sample population, each new sample is selected from the current population; the probability of each specific sample being selected is proportional to the weight of the specific sample;

[0039] Set the sample population to start with the forward message f1 at time t: t =P(X t |e 1:t ) is the correct representation; use N(X t |e 1:t ) means that after fully processing the observation, it has state X t The number of samples, so for N that is large enough to meet the requirements, the following relationship can be obtained:

[0040] N(X t |e 1:t )÷N=P(X t |e 1:t );

[0041] Given a sample condition at time t, propagate each sample forward by sampling the state variable corresponding to time t+1; from each X t Arrive at X t+1 The number of samples is equal to the transition probability multiplied by X t The total amount, so we reach state X t+1 The total number of samples is as follows:

[0042] N(X t+1 |e 1:t )=∑P(X t+1 |X t )N(X t |e 1:t );

[0043] Assign weights to samples based on the probability of each sample at time t+1; t+1 The sample gets the weight P(e t+1 |X t+1 ); when evidence e is observed t+1 After that, it is in state X t+1The total sample weights are as follows:

[0044] W(X t+1 |e 1:t+1 )=P(e t+1 |X t+1 )N(X t+1 |e 1:t );

[0045] The sampling step is repeated to obtain the cyclic sample population to represent the forward message at time t+1.

[0046] The implementation of the electric power intelligent dispatching system and method for electric power facilities provided by the present invention has the following beneficial effects compared with the prior art: the present invention can realize the intelligent dispatching of electric power of electric power facilities, and after the intelligent dispatching, can collect the power operation data under the dispatching; obtain manual evaluation of the power operation data; and modify the prediction model according to the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flow chart of an improved electric power facility intelligent dispatching system and method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] like Figure 1 As shown, an embodiment of the present invention provides a method for intelligent dispatching of electric power facilities, which includes the following steps:

[0049] S1. Obtain the power operation data under dispatch in a certain area within a certain period of time;

[0050] S2, screening and preprocessing the power operation data and sending it to the backend server;

[0051] S3. The backend server inputs the filtered power operation data into the prediction model, and obtains the power operation dispatch instruction for the next time through the prediction model;

[0052] S4. Intelligently dispatch the power of power facilities through power operation dispatch instructions.

[0053] In the electric power intelligent dispatching method for electric power facilities described in the present invention,

[0054] After step S4, the following steps are also included:

[0055] After intelligent dispatch, collect the power operation data under dispatch; obtain manual evaluation of the power operation data; and modify the prediction model based on the evaluation.

[0056] In the electric power intelligent dispatching method for electric power facilities described in the present invention,

[0057] The step S3 comprises:

[0058] The backend server inputs the screened power operation data as samples into the prediction model;

[0059] For each sample, by transferring the model P(X t+1 |X t ), the current state value of the sample X t Under the condition, for the next state value X t+1 Sampling is performed so that the samples propagate forward;

[0060] For each sample, by assigning the probability value P(e) of the new state to the sample t+1 |X t ) as weight;

[0061] Resample the population sample to generate a new N sample population, each new sample is selected from the current population; the probability of each specific sample being selected is proportional to the weight of the specific sample;

[0062] Set the sample population to start with the forward message f at time t 1:t =P(X t |e 1:t ) is correctly expressed; use N(X t |e 1:t ) means that after fully processing the observation, it has state X t The number of samples, so for N that is large enough to meet the requirements, the following relationship can be obtained:

[0063] N(X t |e 1:t )÷N=P(X t |e 1:t );

[0064] Given a sample condition at time t, propagate each sample forward by sampling the state variable corresponding to time t+1; from each X t Arrive at X t+1 The number of samples is equal to the transition probability multiplied by X t The total amount, so we reach state X t+1 The total number of samples is as follows:

[0065] N(X t+1 |e 1:t )=∑P(X t+1 |X t )N(X t |e 1:t );

[0066] Assign weights to samples based on the probability of each sample at time t+1; t+1 The sample gets the weight P(et+1 |X t+1 ); when evidence e is observed t+1 After that, it is in state X t+1 The total sample weights are as follows:

[0067] W(X t+1 |e 1:t+1 )=P(e t+1 |X t+1 )N(X t+1 |e 1:t );

[0068] The sampling step is repeated to obtain the cyclic sample population to represent the forward message at time t+1.

[0069] The present invention also provides an electric power intelligent dispatching system for electric power facilities, which includes the following units:

[0070] An operation data acquisition unit is used to acquire the power operation data under dispatch in a certain area within a certain period of time;

[0071] A data preprocessing and sending unit, used to filter and preprocess the power operation data and send it to the backend server;

[0072] The prediction analysis unit is used to input the screened power operation data into the prediction model through the background server, and obtain the power operation dispatch instruction at the next time through the prediction model;

[0073] The dispatching unit is used to intelligently dispatch the power of the power facilities through power operation dispatching instructions.

[0074] In the electric power intelligent dispatching system for electric power facilities described in the present invention,

[0075] The scheduling unit then further includes:

[0076] The correction unit is used to collect the power operation data under the dispatch after the intelligent dispatch; obtain the manual evaluation of the power operation data; and correct the prediction model according to the evaluation.

[0077] In the electric power intelligent dispatching system for electric power facilities described in the present invention,

[0078] The prediction analysis unit comprises:

[0079] The backend server inputs the screened power operation data as samples into the prediction model;

[0080] For each sample, by transferring the model P(X t+1 |X t ), the current state value Xt of the sample, the next state value Xt t+1Sampling is performed so that the samples propagate forward;

[0081] For each sample, by assigning the probability value P(e) of the new state to the sample t+1 |X t ) as weight;

[0082] Resample the population sample to generate a new N sample population, each new sample is selected from the current population; the probability of each specific sample being selected is proportional to the weight of the specific sample;

[0083] Set the sample population to start with the forward message f at time t 1:t =P(X t |e 1:t ) is correctly expressed; use N(X t |e 1:t ) represents the number of samples with state Xt after the observation is completely processed, so for a sufficiently large N that tends to meet the requirements, the following relationship can be obtained:

[0084] N(X t |e 1:t )÷N=P(X t |e 1:t );

[0085] Given a sample condition at time t, propagate each sample forward by sampling the state variable corresponding to time t+1; from each X t Arrive at X t+1 The number of samples is equal to the transition probability multiplied by X t The total amount, so we reach state X t+1 The total number of samples is as follows:

[0086] N(X t+1 |e 1:t )=∑P(X t+1 |X t )N(X t |e 1:t );

[0087] Assign weights to samples based on the probability of each sample at time t+1; t+1 The sample gets the weight P(e t+1 |X t+1 ); when evidence e is observed t+1 After that, it is in state X t+1 The total sample weights are as follows:

[0088] W(X t+1 |e1: t+1 )=P(e t+1 |X t+1 )N(Xt+1 |e 1:t );

[0089] The sampling step is repeated to obtain the cyclic sample population to represent the forward message at time t+1.

[0090] The implementation of the electric power intelligent dispatching system and method for electric power facilities provided by the present invention has the following beneficial effects compared with the prior art: the present invention can realize the intelligent dispatching of electric power of electric power facilities, and after the intelligent dispatching, can collect the power operation data under the dispatching; obtain manual evaluation of the power operation data; and modify the prediction model according to the evaluation.

[0091] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory, a memory, a read-only memory, an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0092] It is understandable that a person skilled in the art can make other corresponding changes and modifications based on the technical concept of the present invention, and all these changes and modifications should fall within the protection scope of the claims of the present invention.

Claims

1. A method for intelligent power dispatching of electric power facilities, using an intelligent power dispatching system of electric power facilities for intelligent power dispatching; in, The electric power intelligent dispatching system for electric power facilities comprises the following units: an operation data acquisition unit, used to acquire the power operation data under dispatch in a certain area within a certain period of time; a data preprocessing and sending unit, used to screen and preprocess the power operation data and send it to the background server; The prediction analysis unit is used to input the screened power operation data into the prediction model through the background server, and obtain the power operation dispatch instruction at the next time through the prediction model; A dispatching unit, used to intelligently dispatch the power of the power facility through power operation dispatching instructions; The dispatching unit further includes: a correction unit, which is used to collect the power operation data under the dispatching after the intelligent dispatching; obtain the manual evaluation of the power operation data; and correct the prediction model according to the evaluation; the prediction analysis unit is specifically used to: input the screened power operation data as a sample into the prediction model through the background server; for each sample, through the transfer model P(X t+1 |X t ), the current state value of the sample X t Under the condition, for the next state value X t+1 Sampling is performed so that the sample propagates forward; for each sample, by assigning the probability value P(e t+1 |X t ) as the weight; resample the overall sample to generate a new N sample population, each new sample is selected from the current population; the probability of each specific sample being selected is proportional to the weight of the specific sample; set the sample population to start at time t forward message f 1：t =P(X t |e 1：t ) is the correct representation; use N(X t |e 1：t ) means that after fully processing the observation, it has state X t The number of samples, so for N that is large enough to meet the requirements, the following relationship can be obtained: N(X t |e 1：t )÷N=P(X t |e 1：t ); Given a sample condition at time t, propagate each sample forward by sampling the state variable corresponding to time t+1; from each X t Arrive at X t+1 The number of samples is equal to the transition probability multiplied by X t The total amount, so we reach state X t+1 The total number of samples is as follows: N(X t+1 |e 1：t )=∑P(X t+1 |X t )N(X t |e 1：t ); weight each sample according to its probability at time t+1; in state X t+1 The sample gets the weight P(e t+1 |X t+1 ); when evidence e is observed t+1 After that, it is in state X t+1 The total sample weight is as follows: W(X t+1 |e 1：t+1 )=P(e t+1 |X t+1 )N(X t+1 |e 1：t ); Repeat the sampling step to obtain the sample population after the cycle to represent the forward message at time t+1; The method for intelligent dispatching of electric power facilities comprises the following steps: S1, obtaining the dispatched electric power operation data in a certain area within a certain period of time; S2, screening and preprocessing the electric power operation data and sending it to a background server; S3, the background server inputs the screened electric power operation data into a prediction model, and obtains the electric power operation dispatching instruction for the next time through the prediction model; S4, intelligently dispatching the electric power of the electric power facility through the electric power operation dispatching instruction; The step S4 further includes: after the intelligent dispatch, collecting the power operation data under the dispatch; obtaining manual evaluation of the power operation data; and modifying the prediction model according to the evaluation; The step S3 includes: the backend server inputs the filtered power operation data as samples into the prediction model; for each sample, the transfer model P(X t+1 X t ), the current state value of the sample X t Under the condition, for the next state value X t+1 Sampling is performed so that the sample propagates forward; for each sample, by assigning the probability value P(e t+1 X t ) as the weight; resample the overall sample to generate a new N sample population, each new sample is selected from the current population; the probability of each specific sample being selected is proportional to the weight of the specific sample; set the sample population to start at time t forward message f 1：t =P(X t |e 1：t ) is the correct representation; use N(X t |e 1：t ) means that after fully processing the observation, it has state X t The number of samples, so for N that is large enough to meet the requirements, the following relationship exists: N(X t |e 1：t )÷N=P(X t |e 1：t ); Given a sample condition at time t, propagate each sample forward by sampling the state variable corresponding to time t+1; from each X t Arrive at X t+1 The number of samples is equal to the transition probability multiplied by X t The total amount, so we reach state X t+1 The total number of samples is as follows: N(X t+1 |e 1：t )=∑P(X t+1 |X t )N(X t |e 1：t ); weight each sample according to its probability at time t+1; in state X t+1 The sample gets the weight P(e t+1 |X t+1 ); when evidence e is observed t+1 After that, it is in state X t+1 The total sample weight is as follows: W(X t+1 |e 1：t+1 )=P(e t+1 |X t+1 )N(X t+1 |e 1：t ); Repeat the sampling step to obtain the cyclic sample population to represent the forward message at time t+1.

Images