Active power distribution network rolling optimization scheduling method

Abstract

The invention discloses a rolling optimization scheduling method for an active power distribution network, and relates to the field of power distribution networks. Making a day-ahead scheduling plan; performing intra-day rolling optimization scheduling; scheduling and adjusting in real time; and solving the optimization model by adopting an improved differential evolution algorithm MOEA / D-TSA. The method has remarkable advantages in solving the problems of intermittency and uncertainty of the distributed power supply, the economical efficiency and the reliability of the power distribution network are improved, and stable operation of a power system is ensured. The innovative achievement provides a new thought and solution for the management of the power distribution network, and is helpful for promoting the development of the power distribution network towards a more intelligent, efficient and sustainable direction.

Description

Technical Field

[0001] The present invention relates to the field of distribution networks, and in particular to an active distribution network rolling optimization scheduling method. Background Art

[0002] With the widespread access to distributed power sources (such as renewable energy sources such as solar and wind power) and demand response resources, the operating environment of active distribution networks has become more complex and dynamic. The access to these emerging energy sources and resources has brought unprecedented challenges to the management of distribution networks. Traditional distribution network scheduling methods, because they did not take the characteristics of these new energy sources into consideration during their design and implementation, are unable to cope with the current operating environment. They are difficult to adapt to new operating requirements, especially when dealing with the intermittent and uncertain issues of distributed power sources. Traditional methods are often helpless, which not only affects the economic efficiency of distribution networks, but also poses a threat to their reliability. Therefore, a new scheduling method is urgently needed to meet these challenges in order to improve the economic efficiency and reliability of distribution networks and ensure the stable operation of power systems. Summary of the Invention

[0003] The object of the present invention is to provide a method for rolling optimization scheduling of an active distribution network to solve the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A method for rolling optimization scheduling of an active distribution network comprises the following steps:

[0006] Data collection and forecasting: Real-time operating data of the distribution network is collected, including the load demand of each node and the real-time output of distributed power sources. After preprocessing the collected data, principal component analysis is used to filter out the most influential features from a large number of features to reduce model complexity. Finally, time series analysis and machine learning methods are used to forecast future load demand and distributed power output.

[0007] Develop a day-ahead dispatch plan: Based on the forecast results, establish a day-ahead dispatch model with the objective function of minimizing operating costs, active network losses, and node voltage deviations. Determine the tie-line switch status for distribution network reconstruction and the output plan for distributed generation.

[0008] Intraday rolling optimization scheduling: During the intraday phase, rolling optimization scheduling is performed based on real-time data and forecast results. The optimization problem is broken down into hourly and minute-by-minute levels to smooth out fluctuations in cooling and heating power and electrical power, respectively.

[0009] Real-time scheduling adjustment: Based on the normal distribution of the intraday forecast curve, real-time scheduling adjustment is carried out to ensure stable system operation;

[0010] The improved differential evolution algorithm MOEA / D-TSA is used to solve the optimization model.

[0011] As a further solution of the present invention, the data collected in the data collection step includes:

[0012] Load data: real-time load demand and historical load data of each node;

[0013] Distributed power data: real-time and historical output data of distributed power;

[0014] Meteorological data: weather data such as light intensity, wind speed, and temperature, which have a direct impact on the output of distributed power sources;

[0015] Demand response data: User-side adjustable load data, including the response capability and historical response data of demand response resources.

[0016] As a further solution of the present invention, the data preprocessing includes:

[0017] Data cleaning: remove noise data, fill missing values, and handle outliers;

[0018] Data standardization: convert the data into a unified format and normalize it to the range [0,1];

[0019] Feature extraction: Extracting periodic features of time series data that are helpful for prediction from the original data.

[0020] As a further embodiment of the present invention, step (3) specifically includes:

[0021] Upper-level optimization: the control time domain is 1 hour, and the scheduling time window is 2 hours, which are used to smooth out fluctuations in cooling and heating power. Lower-level optimization: the control time domain is 5 minutes, and the scheduling time window is 1 hour, which are used to smooth out fluctuations in electric power.

[0022] As a further solution of the present invention, the objective function and constraints of the improved differential evolution algorithm MOEA / D-TSA in step (5) are as follows:

[0023] Objective function: min f(x) = α·C 运行成本 (x)+β·C 网损 (x)+γ·C 电压偏差 (x); where C 运行成本 (x) represents the operating cost of the system, C 网损 (x) represents the active network loss, C 电压偏差 (x) represents the node voltage deviation, α, β, and γ are weight coefficients;

[0024] Constraints:

[0025] Power balance constraints:

[0026] Voltage Constraints:

[0027] Line flow constraints:

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The present invention proposes a rolling optimization scheduling method for active distribution networks, which fully considers the characteristics of distributed power sources and demand response resources, and realizes the effective smoothing of fluctuations in cold and heat energy power and electric power through the combination of upper and lower layer optimization. First, the upper layer optimization uses a longer control time domain and scheduling time window to smooth the fluctuations in cold and heat energy power, which helps to improve the efficiency of the distribution network in energy conversion and utilization. Secondly, the lower layer optimization uses a shorter control time domain and scheduling time window to quickly respond to fluctuations in electric power, ensuring the real-time balance and stable operation of the power system. In addition, the present invention also introduces an improved differential evolution algorithm MOEA / D-TSA, which further improves the performance and adaptability of the scheduling method by optimizing the objective function and constraints.

[0030] Compared with existing technologies, this invention offers significant advantages in addressing the intermittent and uncertain nature of distributed power sources. It not only improves the economic efficiency and reliability of distribution networks, but also ensures the stable operation of power systems. This innovative achievement provides new insights and solutions for distribution network management, helping to drive their development towards greater intelligence, efficiency, and sustainability. DETAILED DESCRIPTION

[0031] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0032] In an embodiment of the present invention, a method for rolling optimization scheduling of an active distribution network includes the following steps:

[0033] Data collection and forecasting: Real-time operating data of the distribution network is collected, including the load demand of each node and the real-time output of distributed power sources. After preprocessing the collected data, principal component analysis is used to filter out the most influential features from a large number of features to reduce model complexity. Finally, time series analysis and machine learning methods are used to forecast future load demand and distributed power output.

[0034] Develop a day-ahead dispatch plan: Based on the forecast results, establish a day-ahead dispatch model with the objective function of minimizing operating costs, active network losses, and node voltage deviations. Determine the tie-line switch status for distribution network reconstruction and the output plan for distributed generation.

[0035] Intraday rolling optimization scheduling: During the intraday phase, rolling optimization scheduling is performed based on real-time data and forecast results. The optimization problem is broken down into hourly and minute-by-minute levels to smooth out fluctuations in cooling and heating power and electrical power, respectively.

[0036] Real-time scheduling adjustment: Based on the normal distribution of the intraday forecast curve, real-time scheduling adjustment is carried out to ensure stable system operation;

[0037] The improved differential evolution algorithm MOEA / D-TSA is used to solve the optimization model.

[0038] As a further solution of the present invention, the data collected in the data collection step includes:

[0039] Load data: real-time load demand and historical load data of each node;

[0040] Distributed power data: real-time and historical output data of distributed power;

[0041] Meteorological data: weather data such as light intensity, wind speed, and temperature, which have a direct impact on the output of distributed power sources;

[0042] Demand response data: User-side adjustable load data, including the response capability and historical response data of demand response resources.

[0043] As a further solution of the present invention, the data preprocessing includes:

[0044] Data cleaning: remove noise data, fill missing values, and handle outliers;

[0045] Data standardization: convert the data into a unified format and normalize it to the range [0,1];

[0046] Feature extraction: Extracting periodic features of time series data that are helpful for prediction from the original data.

[0047] As a further embodiment of the present invention, step (3) specifically includes:

[0048] Upper-level optimization: the control time domain is 1 hour, and the scheduling time window is 2 hours, which are used to smooth out fluctuations in cooling and heating power. Lower-level optimization: the control time domain is 5 minutes, and the scheduling time window is 1 hour, which are used to smooth out fluctuations in electric power.

[0049] As a further solution of the present invention, the objective function and constraints of the improved differential evolution algorithm MOEA / D-TSA in step (5) are as follows:

[0050] Objective function: minf(x) = α·C 运行成本 (x)+β·C 网损 (x)+γ·C 电压偏差 (x); where C 运行成本 (x) represents the operating cost of the system, C 网损 (x) represents the active network loss, C 电压偏差 (x) represents the node voltage deviation, α, β, and γ are weight coefficients;

[0051] Constraints:

[0052] Power balance constraints:

[0053] Voltage Constraints:

[0054] Line flow constraints:

[0055] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations that come within the meaning and range of equivalents of the claims be embraced therein.

[0056] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for rolling optimization scheduling of an active distribution network, characterized in that: The steps include: (1) Data collection and data prediction: Collect real-time operation data of the distribution network, including the load demand of each node and the real-time output of distributed power sources; after preprocessing the collected data, use the principal component analysis method to filter out the most influential features from a large number of features to reduce the complexity of the model; finally, use time series analysis and machine learning methods to predict future load demand and distributed power output; (2) Formulate a day-ahead dispatch plan: Based on the forecast results, establish a day-ahead dispatch model with the objective function of minimizing operating costs, active network losses, and node voltage deviations; determine the switch status of the tie lines for distribution network reconstruction and the output plan of the distributed generation; (3) Intraday rolling optimization scheduling: In the intraday stage, rolling optimization scheduling is carried out based on real-time data and forecast results, and the optimization problem is decomposed into hourly and minute levels to smooth out fluctuations in cooling and heating power and electric power respectively; (4) Real-time scheduling adjustment: Based on the normal distribution of the intraday forecast curve, real-time scheduling adjustment is performed to ensure stable operation of the system; (5) The improved differential evolution algorithm MOEA / D-TSA is used to solve the optimization model.

2. The method for rolling optimization scheduling of an active distribution network according to claim 1, characterized in that: The data collected in the data collection step includes: Load data: real-time load demand and historical load data of each node; Distributed power data: real-time and historical output data of distributed power; Meteorological data: weather data such as light intensity, wind speed, and temperature, which have a direct impact on the output of distributed power sources; Demand response data: User-side adjustable load data, including the response capability and historical response data of demand response resources.

3. The method for rolling optimization scheduling of an active distribution network according to claim 1, characterized in that: The data preprocessing includes: Data cleaning: remove noise data, fill missing values, and handle outliers; Data standardization: convert the data into a unified format and normalize it to the range [0,1]; Feature extraction: Extracting periodic features of time series data that are helpful for prediction from the original data.

4. The method for rolling optimization scheduling of an active distribution network according to claim 1, characterized in that: The step (3) specifically includes: Upper-layer optimization: The control time domain is 1 hour, and the scheduling time window is 2 hours, which is used to smooth out fluctuations in cooling and heating power; Lower-level optimization: The control time domain is 5 minutes and the scheduling time window is 1 hour to smooth out power fluctuations.

5. The method for rolling optimization scheduling of an active distribution network according to claim 1, characterized in that: The objective function and constraints of the improved differential evolution algorithm MOEA / D-TSA in step (5) are as follows: Objective function: min f(x) = α·C 运行成本 (x)+β·C 网损 (x)+γ·C 电压偏差 (x); where C 运行成本 (x) represents the operating cost of the system, C 网损 (x) represents the active network loss, C 电压偏差 (x) represents the node voltage deviation, α, β, and γ are weight coefficients; Constraints: Power balance constraints: Voltage Constraints: Line flow constraints:

Images