A method for optimizing configuration of channel capacity considering priority planning and market transaction

Abstract

The application discloses a channel capacity optimization configuration method considering priority planning and market transaction, takes maximum line annual investment income as an objective function, takes a newly-built line path as a decision variable, establishes a planning layer model with investment budget constraints, newly-built line quantity constraints and N-1 safety constraints as constraint conditions, and obtains a planning decision according to the planning layer model; based on a direct current optimal power flow model, a settlement layer model is obtained on the basis of priority planning power and direct transaction power, according to the planning decision, power market settlement is carried out through the settlement layer model, and network power flow after settlement is returned to the planning layer model; the upper planning layer model and the lower settlement layer model are coupled into a mixed integer programming model by using KKT conditions, and investment income is obtained according to the mixed integer programming model.

Description

Technical Field

[0001] The present invention relates to the field of power system planning, and specifically relates to a transmission channel capacity optimal configuration method considering priority plans and market transactions. Background Art

[0002] Under the background of the new round of power system reform, China has set the goals of "orderly liberalization of power generation and consumption plans" and establishing an efficient power market system. The scale of the power market continues to expand, and its structure is increasingly improved. In 2019, the trading volume organized by all power trading centers across the country reached 2.8344×10¹² kWh; in the same year, China's first 8 pilot provinces for the construction of power spot markets launched simulated trial operations. The power and electricity balance mechanism is transforming from the traditional "plan management" mode to the new "market regulation" mode.

[0003] In the transformation process, the dual-track quantity and price system with the coexistence of planned and market modes will exist for a long time. Planned electricity is based on factors such as important livelihood needs, ecological civilization construction, and safe operation of the power grid, and generally guarantees the power supply for non-market users at a government-approved price. On the premise of fully ensuring the safe and stable operation of the power grid, market-based electricity uses time-of-use price signals to promote trading competition and realize efficient optimal allocation of power resources. The difference in positioning and price systems between the two determines that spot market settlement will be affected by factors such as market entity behavior and electricity volume level; therefore, the connection mode between the plan and the market under the dual-track mode in the transition period will be one of the key designs of the power market. At the present stage, researches on the settlement balance of non-market and market-based power generation and consumption sides, as well as the influence of various factors of market operation on the benefit distribution of market entities and unbalanced funds under different modes are relatively scattered.

[0004] Against the background of the continuous advancement of power marketization, the benefit of power grid planning investment is related to the power flow transmitted by the channels and the channel utilization rate. Traditional power grid planning is difficult to reflect the planning willingness of market entities under the competitive market situation in the coming period. In terms of power grid investment returns, priority electricity is the basic part with low volatility, and how to leverage the electricity growth of the market transaction part is directly related to the investment recovery benefit.

[0005] In a high-proportion hydropower transmission system, during the wet season, due to the minimum output requirement of thermal power units, thermal power units are often required to operate at the minimum startup output, and the load increment part is preferentially supplemented by hydropower to reduce water abandonment; during the dry season, following the principle of no water abandonment, the generated hydropower quantity is preferentially consumed, and then the remaining electricity is met by thermal power. Due to the huge difference in marginal power generation costs between hydropower and thermal power, the nodal price difference at the same node between wet and dry seasons can reach hundreds of yuan per MWh. At present, there are few power grid planning studies that consider market transaction factors and the wet-dry characteristics of hydropower, and existing studies often only optimize the channel utilization rate by analyzing the difference in nodal electricity prices in wet and dry seasons, so as to achieve better annual channel utilization of the entire network.

[0006] Currently, power grid planning that considers market factors is mostly based on pure market transactions, without taking into account the limitations on the market decision-making space of the generation side when priority planned power volumes are included. Furthermore, the high proportion of hydropower transmission systems places high demands on the maintenance of power market order due to their fluctuating power availability. After considering the market rules for hydropower-thermal power complementarity, how to optimize the capacity of the system's external transmission channels to promote market transactions and fully mobilize market vitality is a pressing practical problem that needs to be solved. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for optimizing channel capacity allocation that considers priority planning and market transactions, comprising the following processes: A planning layer model is established with the objective function of maximizing the annual investment return of the line, the decision variable of the newly built line route, and the constraints of investment budget, the number of newly built lines, and N-1 safety constraints. Planning decisions are obtained based on the planning layer model. Based on the DC optimal power flow model, a clearing layer model is obtained on the basis of priority planned power and direct transaction power. According to the planning decision, the spot market is cleared through the clearing layer model, and the cleared network power flow is returned to the planning layer model. The upper planning layer model and the lower clearing layer model will be coupled into a mixed integer programming model using KKT conditions, and the investment returns will be obtained based on the mixed integer programming model.

[0008] Furthermore, the aforementioned planning layer model, which uses maximizing the annual investment return of the railway line as the objective function, the route of the newly constructed railway line as the decision variable, and investment budget constraints, constraints on the number of newly constructed railway lines, and N-1 safety constraints as constraints, includes the following process: The upper-level model considers the costs of line expansion planning and deep adjustment and renovation of generating units, with the goal of maximizing total investment returns. Constraints include the number of branch lines in the line expansion plan and line investment cost constraints.

[0009] in F 1 represents the investment returns in the power grid. C 1 represents the equivalent annual investment cost of the power grid planning scheme.

[0010] in To achieve a unified transmission and distribution price across the entire network, This represents the total power flow of the entire network.

[0011] The investment budget for the route expansion plan adopts the equivalent annual value method to calculate the annual equivalent investment cost based on the investment period. The expression is as follows: C 1 represents the equivalent annual investment cost of the power grid planning scheme; η 1 represents the discount rate; m Indicates the planning period; and These represent the investment cost per unit length of a newly built transmission corridor and the length of the transmission line, respectively. N w The number of lines to be selected; C 1max This represents the maximum investment budget.

[0012] Furthermore, the clearing layer model, derived from the DC optimal power flow model based on priority planned power and direct transaction power, includes the following process: The DC power flow model is based on: in N gen The total number of generators in the system. for s day i The price curve for the generator set at node number 1 for s day i The winning bid volume of node generators in the spot market; The constraints are power balance constraints, line power flow constraints, and generator output constraints: for s Kusaka i Node load size, T i For nodes i The power transmission distribution factor, For the line m - n The limit of trend delivery, For nodes i exist s Daily priority power output For nodes i exist s Daily medium- and long-term trading volume , Let N be the generator output range, and N be the total number of system nodes. Constraints on hydropower-thermal power exchange in direct electricity trading: in express s day i Medium and long-term hydropower transaction volume at key nodes express s day i Medium- and long-term trading volume of thermal power plants at nodes for s The ratio of water and fire mutual assistance on a daily basis; Priority electricity usage needs to be balanced on a rolling basis throughout the year, therefore the following equation constraint applies: in for i The total priority planned electricity volume for each node throughout the year is represented by this equation, which indicates that the planned electricity volume of each node can be balanced on a rolling basis throughout the year. The planned regional load is obtained based on the single-period load price elasticity model. Based on the network structure information and load data of the newly constructed lines included in the planning decisions transmitted from the upper layer, the network flow is recalculated, and the new network flow is returned to the upper layer for decision-making. The new network flow is as follows: in To account for node load after demand response, e The price elasticity of demand is the coefficient. , These are the marginal electricity prices before and after the planning stages, l The total number of branches in the system. The total power flow of the system lines after taking into account the elasticity of electricity prices.

[0013] Furthermore, the process of coupling the upper-level planning layer model and the lower-level clearing layer model into a mixed-integer programming model using KKT conditions includes the following steps: The KKT conditions are as follows: in P G This is the output matrix of the system's generators. P D The system load matrix; , These are the upper and lower limit matrices for generator output, respectively. Let be the power flow constraint matrix for the line. Generator price quote matrix; , , , These are the Lagrange multipliers corresponding to the constraints; the market clearing constraint and its KKT conditions are coupled with the upper-level planning layer model as constraints to form a single-layer mixed integer programming model.

[0014] The beneficial effects of this invention are as follows: In the construction of the planning model, this patent has discovered market elements that are compatible with market transactions and power grid planning incentives. That is, when the channel capacity is met, the market always tends to call low-cost electricity across regions. Under the new market model, the power grid investment returns are also related to the channel capacity. Therefore, by using the channel transmission flow, the market clearing model is well coupled with planning, so as to achieve the accuracy and efficiency of planning investment.

[0015] The market clearing model of this invention considers a spot clearing model where priority electricity and market transactions coexist. In actual operation, the proportion of spot transactions is currently small, and the market decision space for spot transactions is limited by priority electricity and medium- and long-term transactions. Discussing the impact of spot transactions on planning without considering these two factors results in a large deviation from reality. Therefore, this invention considers the optimal configuration of grid channel capacity under the influence of the entire cycle.

[0016] This model takes into account the actual situation of high-proportion hydropower transmission systems. There is an imbalance between the wet and dry seasons in the market bidding of thermal power and hydropower. Therefore, based on actual operating experience, the optimal power flow model considers the mutual support between hydropower and thermal power according to the wet and dry conditions at different times of the year, so that the power flow of the system is closer to the actual operation. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a channel capacity optimization configuration method that considers priority planning and market transactions; Figure 2 The following is a model logic diagram for an example. Detailed Implementation

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0019] like Figure 1 As shown, a channel capacity optimization allocation method considering priority planning and market transactions includes the following process: A planning layer model is established with the objective function of maximizing the annual investment return of the line, the decision variable of the newly built line route, and the constraints of investment budget, the number of newly built lines, and N-1 safety constraints. Planning decisions are obtained based on the planning layer model. Based on the DC optimal power flow model, a clearing layer model is obtained on the basis of priority planned power and direct transaction power. According to the planning decision, the spot market is cleared through the clearing layer model, and the cleared network power flow is returned to the planning layer model. The upper planning layer model and the lower clearing layer model will be coupled into a mixed integer programming model using KKT conditions, and the investment returns will be obtained based on the mixed integer programming model.

[0020] The aforementioned planning layer model, which uses maximizing the annual investment return of the railway line as the objective function, the route of the newly constructed railway line as the decision variable, and investment budget constraints, constraints on the number of newly constructed railway lines, and N-1 safety constraints as constraints, includes the following process: The upper-level model considers the costs of line expansion planning and deep adjustment and renovation of generating units, with the goal of maximizing total investment returns. Constraints include the number of branch lines in the line expansion plan and line investment cost constraints.

[0021] in F 1 represents the investment returns in the power grid. C 1 represents the equivalent annual investment cost of the power grid planning scheme.

[0022] in To achieve a unified transmission and distribution price across the entire network, This represents the total power flow of the entire network.

[0023] The investment budget for the route expansion plan adopts the equivalent annual value method to calculate the annual equivalent investment cost based on the investment period. The expression is as follows: C 1 represents the equivalent annual investment cost of the power grid planning scheme; η 1 represents the discount rate; m Indicates the planning period; and These represent the investment cost per unit length of a newly built transmission corridor and the length of the transmission line, respectively. N w The number of lines to be selected; C 1max This represents the maximum investment budget.

[0024] The aforementioned DC optimal power flow model, which derives a clearing layer model based on priority planned power and direct transaction power, includes the following process: The DC power flow model is based on: in Ngen The total number of generators in the system. for s day i The price curve for the generator set at node number 1 for s day i The winning bid volume of node generators in the spot market; The constraints are power balance constraints, line power flow constraints, and generator output constraints: for s Kusaka i Node load size, T i For nodes i The power transmission distribution factor, For the line m - n The limit of trend delivery, For nodes i exist s Daily priority power output For nodes i exist s Daily medium- and long-term trading volume , Let N be the generator output range, and N be the total number of system nodes. Constraints on hydropower-thermal power exchange in direct electricity trading: in express s day i Medium and long-term hydropower transaction volume at key nodes express s day i Medium- and long-term trading volume of thermal power plants at nodes for s The ratio of water and fire mutual assistance on a daily basis; Priority electricity usage needs to be balanced on a rolling basis throughout the year, therefore the following equation constraint applies: in for i The total priority planned electricity volume for each node throughout the year is represented by this equation, which indicates that the planned electricity volume of each node can be balanced on a rolling basis throughout the year. The planned regional load is obtained based on the single-period load price elasticity model. Based on the network structure information and load data of the newly constructed lines included in the planning decisions transmitted from the upper layer, the network flow is recalculated, and the new network flow is returned to the upper layer for decision-making. The new network flow is as follows: in To account for node load after demand response, e The price elasticity of demand is the coefficient. , These are the marginal electricity prices before and after the planning stages, l The total number of branches in the system. The total power flow of the system lines after taking into account the elasticity of electricity prices.

[0025] The process of coupling the upper-level planning layer model and the lower-level clearing layer model into a mixed-integer programming model using KKT conditions includes the following steps: The KKT conditions are as follows: in P G This is the output matrix of the system's generators. P D The system load matrix; , These are the upper and lower limit matrices for generator output, respectively. Let be the power flow constraint matrix for the line. Generator price quote matrix; , , , These are the Lagrange multipliers corresponding to the constraints; the market clearing constraint and its KKT conditions are coupled with the upper-level planning layer model as constraints to form a single-layer mixed integer programming model.

[0026] Specifically, the technical solution adopted by this invention to solve its technical problem is as follows: A channel capacity optimization configuration model considering both prioritized planned electricity and market-traded electricity is proposed. The upper layer is a planning layer model, aiming to maximize the investment returns of transmission lines. Constraints such as planned investment scale and budget are considered to form a planning decision, which is then transmitted to the lower layer. The lower layer, based on the DC optimal power flow model, performs spot market clearing based on prioritized planned electricity and direct-traded electricity, obtaining the clearing result. The cleared network power flow is then returned to the upper layer to calculate investment returns.

[0027] The upper-level model considers the costs of line expansion planning and deep adjustment and renovation of generating units, with the goal of maximizing total investment returns. Constraints include the number of branch lines in the line expansion plan and line investment cost constraints.

[0028] in F 1 represents the investment returns in the power grid. C 1 represents the equivalent annual investment cost of the power grid planning scheme.

[0029] in To achieve a unified transmission and distribution price across the entire network, This represents the total power flow of the entire network.

[0030] The investment budget for the route expansion plan adopts the equivalent annual value method to calculate the annual equivalent investment cost based on the investment period. The expression is as follows: C 1 represents the equivalent annual investment cost of the power grid planning scheme; η 1 represents the discount rate; m Indicates the planning period; and These represent the investment cost per unit length of a newly built transmission corridor and the length of the transmission line, respectively. N w This represents the number of lines to be selected. C 1max This represents the maximum investment budget.

[0031] The clearing layer model is based on priority power planning and direct power trading; therefore, the DC power flow model upon which this model is based is: The clearing layer is the DC optimal power flow model described above. Considering the clearing situation under different operating modes, the clearing model under a single typical daily operating mode is as follows, and its objective function expression is: in N gen The total number of generators in the system. for s day i The price curve for the generator set at node number 1 for s day iThe winning bid volume of node generators in the spot market. This paper assumes that the bidding behavior of power suppliers in the market is rational, that is, they bid based on the marginal generation cost of the units. The generator cost curve of thermal power units can be represented by a quadratic curve; for ease of calculation, a piecewise cost curve is used instead of a quadratic curve. The marginal cost of hydropower units is relatively small and can be approximately represented by a linear bidding curve.

[0032] The constraints consider system power balance, line power flow constraints, and generator output constraints. The generator output decision variable is the cleared electricity in the spot market, while the total output considers the sum of the advantageous electricity, the direct transaction electricity, and the spot transaction electricity. for s Kusaka i Node load size, T i For nodes i The power transmission distribution factor, For the line m - n The limit of trend delivery, For nodes i exist s Daily priority power output For nodes i exist s Daily medium- and long-term trading volume , Let N be the generator output range, and N be the total number of system nodes.

[0033] This model optimizes high-proportion hydropower transmission systems. In such systems, the direct trading volume should be balanced between hydropower and thermal power according to the different water seasons. During the wet season, trading should be conducted with hydropower exceeding thermal power, while during the dry season, trading should be conducted with thermal power exceeding hydropower. The following are the constraints for the balance between hydropower and thermal power in direct trading volume: in express s day i Medium and long-term hydropower transaction volume at key nodes express s day i Medium- and long-term trading volume of thermal power plants at nodes for s The ratio of water and fire in the day.

[0034] Priority electricity usage needs to be balanced on a rolling basis throughout the year, therefore the following equation constraint applies: in for i The total priority planned electricity for each node throughout the year is represented by this equation, which indicates that the planned electricity for each node can be balanced on a rolling basis throughout the year.

[0035] Based on the above optimal power flow model, the Lagrange multipliers of the optimized model are obtained, namely the marginal electricity price at system nodes. Based on the single-time load price elasticity model, the planned regional load is obtained: The network power flow is then recalculated based on the newly derived network parameters and load data, and the new network power flow is returned to the upper-level decision-making process.

[0036] in To account for node load after demand response, e The price elasticity of demand is the coefficient. , These are the marginal electricity prices before and after the planning stages, l This represents the total number of branches in the system. The total power flow of the system lines after taking into account the elasticity of electricity prices.

[0037] The upper layer of the model is an integer programming model, and the lower layer is a linear programming model. Therefore, the KKT conditions can be used to couple the lower layer model into a mixed integer programming model for stable solution.

[0038] First, based on the KKT conditions of the day-ahead optimization model, the upper-level planning model is coupled with the lower-level clearing model. The KKT conditions (matrix form) of the day-ahead optimization model are as follows: in P G This is the output matrix of the system's generators. P D The system load matrix; , These are the upper and lower limit matrices for generator output, respectively. Let be the power flow constraint matrix for the line. Generator price quote matrix; , , , These are the Lagrange multipliers corresponding to the constraints.

[0039] The market clearing constraint and its KKT conditions are coupled with the upper-level planning layer model to form a single-layer mixed integer programming model, which can be solved by calling Cplex.

[0040] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for optimizing channel capacity allocation considering priority planning and market transactions, characterized in that, The process includes the following: A planning layer model is established with the objective function of maximizing the annual investment return of the line, the decision variable of the newly built line route, and the constraints of investment budget, the number of newly built lines, and N-1 safety constraints. Planning decisions are obtained based on the planning layer model. Based on the DC optimal power flow model, a clearing layer model is obtained on the basis of priority planned power and direct transaction power. According to the planning decision, the spot market is cleared through the clearing layer model, and the cleared network power flow is returned to the planning layer model. The upper-level planning layer model and the lower-level clearing layer model will be coupled into a mixed-integer programming model using KKT conditions, and the investment return will be obtained based on the mixed-integer programming model. The aforementioned planning layer model, which uses maximizing the annual investment return of the railway line as the objective function, the route of the newly constructed railway line as the decision variable, and investment budget constraints, constraints on the number of newly constructed railway lines, and N-1 safety constraints as constraints, includes the following process: The upper-level model considers the cost of line expansion planning and deep adjustment and modification of generating units, with the goal of maximizing total investment returns. Constraints include the number of branch lines in the line expansion plan and line investment cost. F1 represents the investment income of the power grid. C 1 represents the equivalent annual investment cost of the power grid planning scheme; in To achieve a unified transmission and distribution price across the entire network, The sum of power flow across the entire network; The investment budget for the route expansion plan adopts the equivalent annual value method to calculate the annual equivalent investment cost based on the investment period. The expression is as follows: C 1 represents the equivalent annual investment cost of the power grid planning scheme; η 1 represents the discount rate; m Indicates the planning period; and These represent the investment cost per unit length of a newly built transmission corridor and the length of the transmission line, respectively. N w The number of lines to be selected; C 1max This represents the maximum investment budget. The aforementioned DC optimal power flow model, which derives a clearing layer model based on priority planned power and direct transaction power, includes the following process: The DC power flow model is based on: in N gen The total number of generators in the system. for s day i The price curve for the generator set at node number 1 for s day i The winning bid volume of node generators in the spot market; The constraints are power balance constraints, line power flow constraints, and generator output constraints: for s Kusaka i Node load size, T i For nodes i The power transmission distribution factor, For the line m - n The limit of trend delivery, For nodes i exist s Daily priority power output For nodes i exist s Daily medium- and long-term trading volume , Let N be the generator output range, and N be the total number of system nodes. Constraints on hydropower-thermal power exchange in direct electricity trading: in express s day i Medium and long-term hydropower transaction volume at key nodes express s day i Medium- and long-term trading volume of thermal power plants at nodes for s The ratio of water and fire mutual assistance on a daily basis; Priority electricity usage needs to be balanced on a rolling basis throughout the year, therefore the following equation constraint applies: in for i The total priority planned electricity volume for each node throughout the year is represented by this equation, which indicates that the planned electricity volume of each node can be balanced on a rolling basis throughout the year. The planned regional load is obtained based on the single-period load price elasticity model. Based on the network structure information and load data of the newly constructed lines included in the planning decisions transmitted from the upper layer, the network flow is recalculated, and the new network flow is returned to the upper layer for decision-making. The new network flow is as follows: in To account for node load after demand response, e The price elasticity of demand is the coefficient. , These are the marginal electricity prices before and after the planning stages, l The total number of branches in the system. The total power flow of the system lines after taking into account the elasticity of electricity prices.

2. The channel capacity optimization allocation method considering priority planning and market transactions according to claim 1, characterized in that, The process of coupling the upper-level planning layer model and the lower-level clearing layer model into a mixed-integer programming model using KKT conditions includes the following steps: The KKT conditions are as follows: in P G This is the output matrix of the system's generators. P D The system load matrix; , These are the upper and lower limit matrices for generator output. Let be the power flow constraint matrix for the line. Generator price quote matrix; , , , These are the Lagrange multipliers corresponding to the constraints; the market clearing constraint and its KKT conditions are coupled with the upper-level planning layer model as constraints to form a single-layer mixed integer programming model.

Images