A method for calculating incremental cost of distribution network nodes based on conditional value at risk

Abstract

The present invention discloses a method for calculating the incremental cost of a distribution network node based on conditional value at risk, comprising the following steps: (1) modeling the extreme value of line flow. The tail distribution of the line flow is fitted using the super-threshold method to obtain an analytical expression for the expected power outage loss of users based on the generalized Pareto distribution; (2) determining the initial investment time based on the expected future load loss, and triggering investment when the expected user power outage loss is equal to the future investment cost limit; (3) calculating the new line investment years after the node injects (or flows out) power; (4) calculating the net present value change caused by the node injecting (or flowing out) power, and finally obtaining the long-term incremental cost of the node from the annual incremental cost. The present invention can reflect the impact of the user power outage loss risk on the power grid investment by considering the conditional value at risk through the incremental cost calculation method of the distribution network node, can significantly delay the investment time of the power grid components, and better reflect the investment strategy of the power grid components when the power flow exceeds the limit risk.

Description

Technical Field

[0001] The present invention relates to a distribution network, and in particular to a method for calculating incremental costs of distribution network nodes based on conditional value at risk. Background Art

[0002] With the deepening of my country's power market reform and the gradual liberalization of distribution and retail sectors, the market is increasingly playing a decisive role in optimizing resource allocation. While integrating a high proportion of renewable energy into the distribution network reduces energy costs, the random and intermittent nature of its output also poses significant operational risks to the network and increases investment costs. In this context, transmission fees, as an important means of economic regulation, play a crucial role in promoting effective competition, quantifying the contributions of different market players to distribution network operations and planning, and promoting the optimal allocation of resources within the market environment.

[0003] However, my country's distribution market has long lacked a unified pricing system, and existing international access fee pricing methods are mostly based on traditional passive distribution networks. When a high proportion of renewable energy is integrated into the distribution network, the random intermittent nature of renewable energy output and reverse power flow will profoundly change existing distribution network analysis methods. Continuing to use the original access fee pricing system will significantly reduce market efficiency.

[0004] Therefore, it is necessary to propose a new transmission fee pricing model to help guide the healthy development of the distribution side market and provide correct price signals to market participants. Summary of the Invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a method for calculating the incremental cost of distribution network nodes based on conditional risk value, so as to reflect the impact of the load's ability to withstand power outage risks on the investment time and transmission fees of distribution network components.

[0006] Technical solution: The method for calculating the incremental cost of distribution network nodes based on conditional value at risk described in the present invention includes the following steps:

[0007] (1) Line power flow extreme value modeling: Using the super-threshold method to fit the tail distribution of the line power flow, an analytical expression for the user's expected power outage loss based on the generalized Pareto distribution is obtained;

[0008] (2) Determine the initial investment time based on risk, and trigger investment when the expected user power outage loss equals the future investment cost limit;

[0009] (3) Calculate the investment years of the new line after the power is injected or outflowed from the node;

[0010] (4) Calculate the net present value change caused by the power injected or outflowed from the node, and finally derive the long-term incremental cost of the node from the annual incremental cost.

[0011] The step (1) is specifically as follows:

[0012] The tail of the line power flow distribution can be expressed by the generalized Pareto distribution as follows:

[0013]

[0014] Where, l represents the feeder; X l represents the power flow through the feeder, which is a random variable; μ is the location parameter of the generalized Pareto distribution, σ is the scale parameter, and ξ is the shape parameter; and X l >μ,1+(X l -μ) / σ>0; μ is X l A larger threshold of the distribution, and less than the absorption capacity of feeder l This formula assumes that X l The distribution function of is maximum stable;

[0015] Random trend X after n years l The tail distribution of is expressed as:

[0016]

[0017] Where r represents the growth rate of random power flow; represents the power flow on the feeder n years later;

[0018] Greater than the feeder absorption capacity The probability is:

[0019]

[0020] Over-the-limit trend The probability distribution of is expressed as:

[0021]

[0022] The power outage cost for users caused by line current exceeding the limit is L·E l , which can be represented by the generalized Pareto distribution:

[0023]

[0024] Expected cost of power outage for users after n years:

[0025]

[0026] The above formula is called expected loss or conditional value at risk.

[0027] The step (2) is specifically as follows:

[0028] The present value of future investment costs is:

[0029]

[0030] Where, Asset l represents the current asset cost; d represents the discount rate; n represents the investment period;

[0031] When the expected power outage loss to users is equal to the present value of future investment costs, the investment is triggered. The investment time n can be obtained from the following formula:

[0032]

[0033] The step (3) is specifically as follows:

[0034] The new power flow distribution after node power injection (or outflow) is expressed as:

[0035]

[0036] The new line investment calculation formula is:

[0037]

[0038] The step (4) is specifically as follows:

[0039] The net present value of the line after the node increases power injection or outflow is:

[0040]

[0041] The net present value change caused by the power injected or taken out of the node is:

[0042]

[0043] The annual incremental cost of the line is:

[0044] IC l =ΔPV l ·AF(13)

[0045] Where AF represents the annuity factor, which is calculated as follows:

[0046]

[0047] The long-term incremental cost of a node is expressed as:

[0048]

[0049] Where i represents a specific node.

[0050] A computer storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned method for calculating the incremental cost of distribution network nodes based on conditional value at risk.

[0051] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for calculating the incremental cost of a distribution network node based on conditional value at risk is implemented.

[0052] Beneficial effects: Compared with the existing technology, the present invention has the following advantages: it considers the impact of random power injection such as distributed power sources and loads on pricing decisions in an analytical form, and uses the expected user power outage losses as the bottom line to trigger grid investment to guide grid investment. It can reflect the impact of user reliability requirements and node load levels on transmission fees, and effectively guide the planning and construction of regional power grids, conventional power sources, and new energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the present invention;

[0054] Figure 2 This is a diagram of a two-node test system;

[0055] Figure 3 is the line power flow distribution and generalized Pareto distribution;

[0056] Figure 4 is the relationship between load level and net present value;

[0057] Figure 5 is the relationship between load level and investment life;

[0058] Figure 6 is the transmission fee under different load loss factors;

[0059] Figure 7 It is the network access fee under different load levels. DETAILED DESCRIPTION

[0060] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0061] A method for calculating the incremental cost of distribution network nodes based on conditional value at risk includes the following steps:

[0062] (1) Line power flow extreme value modeling: Using the super-threshold method to fit the tail distribution of the line power flow, an analytical expression for the user's expected power outage loss based on the generalized Pareto distribution is obtained;

[0063] (2) Determine the initial investment time based on risk, and trigger investment when the expected user power outage loss equals the future investment cost limit;

[0064] (3) Calculate the investment years of the new line after the power is injected or outflowed from the node;

[0065] (4) Calculate the net present value change caused by the power injected or outflowed from the node, and finally derive the long-term incremental cost of the node from the annual incremental cost.

[0066] The step (1) is specifically as follows:

[0067] The tail of the line power flow distribution can be expressed by the generalized Pareto distribution as follows:

[0068]

[0069] Where, l represents the feeder; X l represents the power flow through the feeder, which is a random variable; μ is the location parameter of the generalized Pareto distribution, σ is the scale parameter, and ξ is the shape parameter; and X l >μ,1+(X l -μ) / σ>0; μ is X l A larger threshold of the distribution, and less than the absorption capacity of feeder l This formula assumes that X l The distribution function of is maximum stable;

[0070] Random trend X after n years l The tail distribution of is expressed as:

[0071]

[0072] Where r represents the growth rate of random power flow; represents the power flow on the feeder n years later;

[0073] Greater than the feeder absorption capacity The probability is:

[0074]

[0075] Over-the-limit trend The probability distribution of is expressed as:

[0076]

[0077] The power outage cost for users caused by line current exceeding the limit is L·E l , which can be represented by the generalized Pareto distribution:

[0078]

[0079] Expected cost of power outage for users after n years:

[0080]

[0081] The above formula is called expected loss or conditional value at risk.

[0082] The step (2) is specifically as follows:

[0083] The present value of future investment costs is:

[0084]

[0085] Where, Asset l represents the current asset cost; d represents the discount rate; n represents the investment period;

[0086] When the expected power outage loss to users is equal to the present value of future investment costs, the investment is triggered. The investment time n can be obtained from the following formula:

[0087]

[0088] The step (3) is specifically as follows:

[0089] The new power flow distribution after node power injection (or outflow) is expressed as:

[0090]

[0091] The new line investment calculation formula is:

[0092]

[0093] The step (4) is specifically as follows:

[0094] The net present value of the line after the node increases power injection or outflow is:

[0095]

[0096] The net present value change caused by the power injected or taken out of the node is:

[0097]

[0098] The annual incremental cost of the line is:

[0099] IC l =ΔPV l ·AF (13)

[0100] Where AF represents the annuity factor, which is calculated as follows:

[0101]

[0102] The long-term incremental cost of a node is expressed as:

[0103]

[0104] Where i represents a specific node.

[0105] A computer storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned method for calculating the incremental cost of distribution network nodes based on conditional value at risk.

[0106] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for calculating the incremental cost of a distribution network node based on conditional value at risk is implemented.

[0107] An embodiment is listed below:

[0108] In order to reveal its mechanism, the Figure 2 The R-LRIC model is tested on a two-node system shown in Figure 1. The feeder has a capacity of 40 MW (the line capacity is 45 MW) and the present value of the feeder cost is £3,193,400. The discount rate is 6.9% and the load growth rate is 1.6% per year. The probability distribution of the line load is shown in the figure. Figure 3 As shown, the initial load level is 20 MW. The 90% quantile of the line load distribution is used as the threshold to fit the generalized Pareto distribution parameters, [μ, σ, ξ] = [10.57, 4.49, -0.45]. The user outage loss factor is set to 32,000 £ / MW.

[0109] First, the proposed risk-oriented node long-term incremental cost (R-LRIC) method is compared with the existing classic long-term incremental cost (LRIC) pricing method. Then, the influence of user power outage loss factor and load level on R-LRIC is revealed by sensitivity analysis. When the initial load level is 20MW, the calculation results of the two LRICs are shown in Table 1. Figure 4 and attached Figure 5 The net present value and investment life curves of the two methods before and after power injection at different load levels are given respectively.

[0110] Table 1 Transmission charges calculated by two LRIC methods at 20MW load level

[0111]

[0112] As shown in Table 1, the proposed R-LRIC calculation method significantly reduces the incremental node cost compared to the classic LRIC calculation method. Furthermore, the investment lifespan after injecting 1 MW of power is minimal, decreasing by only approximately 1.5 years, compared to a 3-year reduction for the classic LRIC method. This is because the proposed algorithm is based on the expected value of the load tail. When the load tail is thin, its sensitivity to load growth is not as high as that of the maximum load. Furthermore, the proposed method explicitly provides the expected load loss, a key factor in determining distribution network investment.

[0113] Attachment Figure 4 As can be seen, when the load level is low, the NPVs calculated by the two methods are similar. However, when the load level increases to 15 MW, the NPVs of the two methods diverge, with the NPV calculated by the classical method exceeding that of the proposed conditional value-at-risk method. Furthermore, the NPV of the classical LRIC method increases exponentially with the load level, while the proposed method shows a nearly linear increase. This is closely related to the probability distribution of the load. When the load level is low, the variance of the load distribution is small, and the expected value of the tail of the load is not much different from the maximum value. This results in a similar difference in the initial NPV. However, when the load is scaled up, the variance of the load distribution also increases, and the tail of the load distribution is significantly elongated. The impact of power injection before and after injection on peak load is greater than on conditional value-at-risk, resulting in a smaller NPV obtained by R-LRIC than that of the classical algorithm.

[0114] Attachment Figure 5 The investment horizons for the two methods at different load levels are presented. Both methods start with the same investment horizon value and decrease as the load level increases. The rate of decline for R-LRIC begins to slow down compared to the classical method at 15 MW. The difference in investment horizon before and after power injection decreases for both methods, but the difference decreases more rapidly for the proposed method compared to the classical method. This is also because R-LRIC mitigates the impact of an increase in maximum power loss based on the expected power outage loss.

[0115] From formula (5), we can see that R-LRIC is related to the user's power outage loss function. Figure 6 As shown in Figure 2, the transmission fee increases monotonically with the user's power outage loss factor. However, the main driving force for the increase in transmission fee comes from the load level. Figure 7 The growth of the network fee obtained by the two algorithms with the increase of load level is given. Figure 7 It can be seen that the LRICs obtained by both methods increase monotonically relative to the load level, but the rate of increase varies with the load level. The rate of change of the classical algorithm increases rapidly with increasing load level. However, because the impact of load increment on user loss expectations decreases with load level, the change in the transmission fee for R-LRIC gradually decreases with load level, forming a gentle "S"-shaped curve.

Claims

1. A method for calculating the incremental cost of distribution network nodes based on conditional value at risk, characterized in that: The following steps are involved: (1) Line power flow extreme value modeling; The super-threshold method is used to fit the tail distribution of the line power flow, and an analytical expression for the user's expected power outage loss based on the generalized Pareto distribution is obtained. The tail of the line power flow distribution can be expressed by the generalized Pareto distribution as follows: Where, l represents the feeder; X l represents the power flow through the feeder, which is a random variable; μ is the location parameter of the generalized Pareto distribution, σ is the scale parameter, and ξ is the shape parameter; and X l >μ,1+(X l -μ) / σ>0; This formula assumes that X l The distribution function of is maximum-stable; Random trend X after n years l The tail distribution of is expressed as: Where r represents the growth rate of random power flow; represents the power flow on the feeder n years later; Greater than the feeder absorption capacity The probability is: Over-the-limit trend The probability distribution of is expressed as: The power outage cost for users caused by line current exceeding the limit is L·E l , which can be represented by the generalized Pareto distribution: Where L is the power outage loss that will be incurred by users due to load or power removal, and is a constant function. It is only related to the power of the load shedding; Expected cost of power outage for users after n years: The above formula is called expected loss or conditional value at risk; (2) Determine the initial investment time based on risk, and trigger investment when the expected user power outage loss is equal to the present value of future investment costs; (3) Calculate the investment years of the new line after the power is injected or outflowed from the node; (4) Calculate the net present value change caused by the power injected or outflowed from the node, and finally derive the long-term incremental cost of the node from the annual incremental cost, which is: The net present value of the line after the node increases power injection or outflow is: Where: n new The new investment period; The net present value change caused by the power injected or taken out of the node is: The annual incremental cost of the line is: IC l =ΔPV l ·OF Where AF represents the annuity factor, which is calculated as follows: The long-term incremental cost of a node is expressed as: Where i represents a specific node; ΔPI i is the change in the power injected or outflowed at node i.

2. A method for calculating the incremental cost of distribution network nodes based on conditional value at risk according to claim 1, characterized in that: The step (2) is specifically as follows: The present value of future investment costs is: Where, Asset l represents the current asset cost; d represents the discount rate; n represents the investment period; When the expected power outage loss to users is equal to the present value of future investment costs, the investment is triggered. The investment time n can be obtained from the following formula:

3. A method for calculating the incremental cost of distribution network nodes based on conditional value at risk according to claim 2, characterized in that: The step (3) is specifically as follows: The new power flow distribution after node power injection (or outflow) is expressed as: Where, is the new current flowing through the feeder; The new line investment calculation formula is:

4. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements a method for calculating the incremental cost of a distribution network node based on conditional value at risk as described in any one of claims 1 to 3.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements a method for calculating the incremental cost of a distribution network node based on conditional value at risk according to any one of claims 1 to 3.

Images