Method for decomposing annual planned power generation amount curve of stepped hydropower plants

Abstract

To provide a method for decomposing an annual planned power amount curve of stepped hydroelectric power plant group that can obtain a reasonable power amount decomposition curve for each month of a year for a plurality of power networks of a stepped hydroelectric power plant, can effectively respond to a different peak adjustment demand in various receiving end power networks, and can improve response capability to a load.SOLUTION: In consideration of differentiated load adjustment demand of a plurality of receiving end power networks, a method constructs a peak adjustment reference model of power amount curve decomposition, determines a typical load curve of each receiving end power network in the peak adjustment reference model, and using a big-M method, introduces 0 to 1 variables bsc,d,k,m, a large number M, and two continuous variables xsc,d,k,m, osc,d,k,m. Then, a mathematical description of positive and negative deviations is abstracted into a general rule of power market contractual issues, and obtained using a mixed integer programming model.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to the field of scheduling of a hydropower system, and specifically to a method for decomposing the annual planned power generation curve of a cascade hydropower plant group.

Background Art

[0002] Many of the large hydropower plants in southwestern China are responsible for power transmission tasks across provinces and regions. Under the background of the power market reform, they are usually involved in the trading and fulfillment of power generation among multiple receiving markets and multiple time scales, and the scheduling has become more complex. Due to the large scale of the power generation, the main power generation of the current large hydropower plants is mainly the planned power generation, and it is usually necessary to participate in the normal power market trading for the increase in power generation. Under such circumstances, how to decompose and execute the planned power generation is of great significance to the operation of the entire power plant and the impact on the receiving power grid, and how to reasonably arrange the execution curve of the planned power generation is particularly important.

[0003] Regarding the problem of decomposing the annual planned power curve, around October every year, it is necessary to determine how to accurately distribute the planned power among power plants and each month. In addition, it is necessary to construct the decomposition criteria for the typical curves of each month for each power receiving province, and determine the typical power quantity curves of each month of each power plant for each province based on certain criteria. After the market is liberalized, it is necessary to reasonably distribute the marketized power among each month and each province based on the market conditions of different provinces. This problem requires considering various scheduling and operation constraints of hydropower plants, reservoirs, and power grids simultaneously, such as hydrological constraints like the water volume balance equation, reservoir water level limits, discharge flow limits, and power generation flow limits, output constraints like output limits, power quantity balance limits, and power balance limits, and system constraints like power transmission ratios across provinces and different load demands. This is a very complex optimization problem with high-dimensional, non-linear multi-time and space coupling constraints. Currently, many of the power quantity decomposition and contract fulfillment problems focus on the execution progress of contract power quantity or power generation benefits, with little consideration for the peak regulation needs of the power grid. How to consider the peak regulation needs of multiple receiving power grids and how to achieve an efficient solution for complex models are facing major challenges.

Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a method for decomposing the annual planned power curve of a cascade hydropower plant group, and the cascade hydropower plant group determines the decomposition results of the power curve among multiple power grids and multiple items. This method can obtain a reasonable monthly power decomposition curve of the cascade hydropower plant for multiple power grids in a year, and can effectively respond to different peak regulation needs in various receiving power grids, improving the response ability of the cascade hydropower system to different loads in the receiving power grid.

[0005] The technical solution adopted by the present invention is as follows. A method for decomposing the annual planned power curve of a cascade hydropower plant group, including the following steps.

[0006] JPEG0007714728000001.jpg62170JPEG0007714728000002.jpg47170

[0007] JPEG0007714728000003.jpg143170

[0008] Step 3: Solve the peak adjustment reference model to obtain the decomposition curve of the annual planned power quantity. Using the big-M method, 0-1 variable b sc,d,k,m , large number M, and two continuous variables x sc,d,k,m , o sc,d,k,m are introduced. At this time, the mathematical description of positive and negative deviations is abstracted into the general rules of the power market contract fulfillment problem and obtained using a mixed-integer programming model. Specifically, it is as follows. JPEG0007714728000004.jpg32170pcl sc,d,k,m represents pcl sc,d,k,m indicating that pcl sc,d,k,m , o sc,d,k,m represents two continuous variables respectively. pca sc,d,k,m represents the plan of power plant k in province m of scenario sc in month d. pcu sc,d,k,m represents the positive deviation of the planned power quantity of power plant k in province m of scenario sc in month d. pcd d,k,m represents the determined allocated planned power quantity of power plant k in province m of month d. M is the number of power plants.

[0009] A method for allocating load according to a ratio in the AGC control policy of the present invention, and the technical effects are as follows. 1) The present invention can obtain a reasonable monthly decomposition curve of the annual power quantity of a plurality of hydropower plants to multiple power grids and can respond to different peak adjustment demands in different receiving power grids. 2) Compared with the conventional methods that focused on the execution progress of the power quantity or the power generation profit of a single power grid, the method of the present invention can simultaneously adapt to the actual situation where a cascade hydropower station transmits power to multiple power grids, and by constructing peak regulation criteria that adapt to multiple power grids, it is possible to balance different load regulation requirements in each power grid. 3) The present invention provides a method for extracting typical load curves that adapt to different power grids, and based on historical load data, it can quickly generate the typical load curve process of each power grid. 4) The present invention combines the big-M method with the variable equivalent transformation policy, converts the original non-linear model into a mixed integer programming model, realizes the efficient solution of the model, thereby obtaining a reasonable monthly power quantity decomposition curve of a cascade hydropower station to multiple power grids throughout the year, and effectively responding to different peak regulation requirements in various receiving-end power grids.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Embodiments for Carrying Out the Invention

[0011] It is a method for decomposing the annual planned power generation curve of a stepped hydropower plant group. The present invention mainly constructs a peak regulation model for a plurality of receiving power grids for the problem of curve decomposition of the annual planned power generation amount, and obtains a typical daily power generation curve of each month of a reasonable stepped hydropower plant by considering hydraulic constraints, power constraints, and system operation constraints. The technical solution is as follows. Considering the differentiated load adjustment requirements of a plurality of receiving power grids, a plurality of power grid peak regulation criteria are constructed. A method for extracting a typical load curve adapted to different power grids is provided, and based on historical load data, a typical load curve process of each power grid is generated. Combining the big-M method with the variable equivalent conversion policy, the original non-linear model is converted into a mixed integer programming model to realize efficient solution of the model.

[0012] JPEG0007714728000005.jpg65170Where C d,k,m is the differentiated peak regulation coefficient of power plant d in province k and month m. lsx d,k,m is the maximum value of the residual load of power plant d in province k and month m. lsm d,k,m is the minimum value of the residual load of power plant d in province k and month m. d is the power plant number. k is the province number. m is the month number. D is the number of power plants. K d is the number of provinces to which power plant d transmits power. M is the number of months. JPEG0007714728000006.jpg68170

[0013] JPEG0007714728000007.jpg174170JPEG0007714728000008.jpg36170

[0014] JPEG0007714728000009.jpg89170

[0015] JPEG0007714728000010.jpg64170

[0016] (2): The second part is to determine the typical load curve of each receiving power grid in the peak regulation criterion model, including the following steps. Step (1): Select the daily load data of a certain power grid for a certain month, transmit the data to a computer, extract the load characteristic parameters daily by the computer using the Java language, perform statistical calculations, and obtain the daily load characteristic index vector C t =[C t1 ,C t2 ,C t3 ,C t4 ,C t5 . Here, C t1 ,C t2 ,C t3 ,C t4 ,C t5 represent the daily load rate, the ratio of peak to bottom per day, the peak period load rate, the flat period load rate, and the bottom period load rate on the t-th day, respectively. Step (2): Perform density function fitting on each load characteristic index vector using kernel density estimation, transmit the index results in (1) to a computer, and use the gaussian_kde function in the scipy.stats module of the Python language to estimate the probability density function to obtain the typical daily load characteristic index vector CTP = [CTP1, CTP2, CTP3, CTP4, CTP5]. Here, CTP1, CTP2, CTP3, CTP4, CTP5 represent the daily load rate, the ratio of peak to bottom per day, the peak period load rate, the flat period load rate, and the bottom period load rate of a typical day, respectively. JPEG0007714728000011.jpg108170

[0017] (III): The third part is to solve the peak regulation standard model to obtain the decomposition curve of the annual planned power consumption. Using the big-M method, introduce the 0-1 variable b sc,d,k,m , the large number M, and two continuous variables x sc,d,k,m , o sc,d,k,m . At this time, the mathematical description of the positive and negative deviations is abstracted into the general rules of the power market contract fulfillment problem, and is obtained using a mixed integer programming model. The model is modeled by the computer Python language and the pyomo modeling tool, and solved using Gurobi. Specifically, it is as follows. JPEG0007714728000012.jpg31170pcl sc,d,k,m is pcl sc,d,k,m indicates that it is the planned power generation negative deviation of the sc-scene d power plant in Province k in month m. x sc,d,k,m , o sc,d,k,m represents two continuous variables respectively. pca sc,d,k,m indicates the plan of the sc-scene d power plant in Province k in month m. pcu sc,d,k,m indicates the positive deviation of the planned power generation of the sc-scene d power plant in Province k in month m. pcd d,k,m indicates the determined allocated planned power generation of the d power plant in Province k in month m. M is the number of power plants.

Example

[0018] Taking the "Wudongde", "Baihetan", "Xiluodu", and "Xiangjiaba" hydropower plants in the lower reaches of the Jinsha River as the research objects, their installed capacities are 10200MW, 16000MW, 12600MW, and 6000MW respectively. They are the "main force" of China's "West-East Power Transmission", related to 7 power receiving provinces (municipalities), including Guangdong, Guangxi, Yunnan, Jiangsu, Zhejiang, Sichuan, Zhejiang, and Shanghai. Among them, the "Xiluodu" power plant has the characteristics of "two reservoirs, two factories, and two power grid scheduling systems", with different power transmission ratios during flood and dry seasons, and the scheduling relationship is very complex. Table 1 shows the allocation ratio of the planned power generation among multiple provincial-level power grids.

[0019]

Table 1

[0020] Adopt the method of the present invention to decompose the annual planned power generation curve of the cascade hydropower plants in the lower reaches of the Jinsha River. First, determine the typical load, and 84 typical loads of each receiving province and each month of the cascade power plants in the lower reaches of the Jinsha River can be extracted by using the second step of the present invention, thereby constructing a peak adjustment model with the peak standard of the first step as the target.

[0021] Table 2, Figures 2 and 3 provide some calculation results. The ratios of the peak to the bottom of the original loads in Guangdong and Sichuan in January are 25.6% and 23.7% respectively. After sharing the peak regulation information and calculating the target, the ratios of the peak to the bottom in Guangdong and Sichuan become 12.3% and 15.6%. After the independent peak regulation calculation for each power plant, the ratios of the peak to the bottom in Guangdong Province and Sichuan Province become 16.3% and 21.9%. For the situation where different stepped hydropower plants send electricity to the same province (Wudongde and the right bank of Xiluodu both send electricity to Guangdong, and Baihetan, the left bank of Xiluodu, and Xiangjiayan both send electricity to Sichuan), the peak regulation effect using shared peak regulation information is significantly better than the independent peak regulation effect of each power plant. Based on the calculation results, the method proposed in the present invention effectively takes into account the different peak regulation requirements for each province such as Guangdong Province and Sichuan Province.

[0022]

Table 2-1

[0023]

Table 2-2

[0024]

Table 2-3

Claims

1.

2.

3.

Images