Bidding method for cogeneration generators in the electricity-heat-gas combined market

By constructing the electric-thermal-gas market cleaning model and the optimal bidding model, combining binary expansion and original-dual complementary conditions, the bidding strategy optimization problem of gas-type cogeneration units in the electric-thermal-gas combined market is solved. Efficient resource allocation and competitiveness improvement have been achieved, and market status analysis has been simplified.

CN114511241BActive Publication Date: 2024-11-08SOUTHEAST UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210156356.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-02-21
Publication Date
2024-11-08
Estimated Expiration
2042-02-21

AI Technical Summary

Technical Problem

The bidding strategy of gas-type cogeneration units in the electric-thermal-gas combined market is difficult to optimize, resulting in insufficient allocation of market resources, insufficient competitiveness, and the existing technology is difficult to effectively solve the non-convex optimization problem, and low computing efficiency.

Method used

Build a clearing model for the power market, heat market and natural gas market, combine binary expansion and original-dual complementary conditions, and derive the optimal bidding equivalent conversion model, and solve the optimal bidding plan for gas-type CHP units through mixed integer linear planning. Control bid accuracy to optimize computing resource usage.

Benefits of technology

Transforming non-convex problems into solveable mixed integer linear programming problems improves bid accuracy and competitiveness, enables flexibly use of computing resources under different time conditions, and simplifies macro-analysis of market status.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114511241B_ABST
    Figure CN114511241B_ABST
Patent Text Reader

Abstract

The present invention discloses a bidding method for a cogeneration generator set in the electricity-heat-gas combined market. As the gas-fired CHP unit relies on its good economy, environmental protection and strong regulation ability, its installed capacity accounts for a rapid increase in the entire power system. The gas-fired CHP unit uses natural gas as fuel and produces electricity and heat at the same time. Therefore, it couples the electricity, heat and natural gas systems and is the most typical and most complex equipment in the electricity-heat-gas combined market. In order to adapt to the rapid marketization needs of the traditional monopoly industries of electricity, heat and natural gas in the future, it is necessary to study the optimal bidding strategy for the gas-fired CHP unit. The gas-fired CHP unit acts as a supplier in the electricity market and the heat market, and as a user in the natural gas market. Therefore, the present invention designs a two-stage trading mechanism for the gas-fired CHP unit with diverse market roles, uses binary expansion and primal-dual complementary conditions to obtain an optimal bidding model with controllable bidding accuracy, and converts the original non-convex problem into a mixed integer linear programming problem, thereby achieving a fast solution. In the optimal bidding strategy calculation process, if the diagonalization algorithm is embedded, the Nash equilibrium of the N-person Cournot game can be calculated, demonstrating the scalability of the present invention.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the market operation mechanism and optimal bidding field of a multi-energy market, and in particular to a bidding method for a cogeneration generator set in an electricity-heat-gas combined market. Background Art

[0002] The multi-energy complementary system relies on the continuous innovation of related technologies such as cogeneration equipment, energy conversion equipment and renewable energy, and couples multiple energy networks such as power grid, natural gas grid and heat grid. It is a new energy service system that realizes "multi-source complementarity and cascade utilization" of electricity, gas, heat, renewable energy, and highly coordinated "source-grid-load-storage" links. Among them, gas-fired cogeneration units, as the core equipment of the multi-energy complementary system, have increased their installed capacity in the energy system year by year due to their characteristics such as strong controllability and high environmental protection. With the deepening of energy market-oriented reforms, gas-fired cogeneration units need to shift from the traditional planned scheduling operation mode to the market bidding operation mode. Therefore, it is necessary to conduct a systematic study on the bidding of gas-fired cogeneration units in multiple energy markets, optimize their bidding curves and bidding strategies, and give full play to the market's optimal allocation of resources. Summary of the invention

[0003] Purpose of the invention: In view of the problems existing in the prior art, the present invention provides a bidding method for cogeneration generator sets in the electricity-heat-gas joint market. The method can provide the best bidding strategy for gas-fired CHP units and generate the best bidding curve, thereby improving the competitiveness of gas-fired CHP units in the market. In addition, the present invention also sorts out the collaborative clearing process of the future electricity-heat-gas market, ensuring the efficient and stable operation of the energy market.

[0004] Technical solution: The bidding method of the cogeneration generator set in the electricity-heat-gas joint market of the present invention includes:

[0005] (1) Construct a linear clearing model for power market trends;

[0006] (2) Construct a heat market clearing model;

[0007] (3) Construct a natural gas market clearing model;

[0008] (4) Construct the optimal bidding model for gas-fired CHP units and derive the equivalent conversion model of the optimal bidding based on binary expansion and primal-dual complementary conditions;

[0009] (5) Based on the proposed bidding process, find the optimal bidding solution for the gas-fired CHP unit.

[0010] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: it converts the optimization problem of non-convex problems that are difficult to solve into a mixed integer linear programming problem with relatively mature solving technology. Based on binary expansion, the control of bidding accuracy can be achieved. The higher the bidding accuracy, the longer the calculation time required. However, when the bidding accuracy reaches a certain level, it cannot further improve the economic benefits of the bidder. Users can use computing resources reasonably by controlling the bidding accuracy. When time is tight and the bidding cycle is short, a lower bidding accuracy can be used to speed up the solution and complete the bidding work before the bidding deadline. When time is ample and the bidding cycle is long, a higher bidding accuracy can be used. Although the calculation time is long, the optimal bidding strategy can be solved. In addition, the bidding process proposed by the present invention can also facilitate scholars to study the Nash equilibrium solution of the N-person oligopoly game, so as to analyze the market status from an overall macro perspective. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a schematic diagram of the segmented bidding curve of the present invention.

[0012] Figure 2 This is the optimal bidding flow chart of the gas-fired CHP unit of the present invention.

[0013] Figure 3 It is a diagram of the iterative convergence process of the present invention.

[0014] Figure 4 A schematic diagram of the electricity-heat-gas relationship of the present invention. DETAILED DESCRIPTION

[0015] This embodiment provides an optimization method for coordinating bilateral interests based on a two-level programming model of a multi-energy flow system, the method comprising:

[0016] (1) Under the premise of meeting all electricity demands, the goal is to purchase the least amount of electricity in the electricity market, and a linear power flow clearing model for the electricity market is constructed under the condition of linear power flow constraints.

[0017] The node electricity price is related to the geographical location of the node, line capacity, node voltage, etc. However, if the AC power flow is directly used for calculation, it will bring a huge calculation burden to the independent system operator (E-ISO) of the electricity market and cannot meet the needs of actual project operation. Therefore, it is necessary to further simplify the AC power flow model. Based on the original AC power flow model, the apparent power S ij It can be calculated by the following formula:

[0018]

[0019] Where V i is the node voltage of bus i, θ ijrepresents the voltage phase difference between busbar i and busbar j, r ij and x ij Represent the line resistance and reactance respectively. The above formula can be divided into two parts: real part and imaginary part:

[0020] S ij =p ij +jq ij (2)

[0021]

[0022]

[0023] In the formula, p ij / q ij Respectively represent the active and reactive power flowing through line ij. Active power loss It can be expressed as

[0024]

[0025] In general, the voltage phase difference between the two ends of the line is usually small, and the amplitude is controlled at about 1.0pu. i ≈1p.u.,θ ij ≈sinθ ij , so the active power loss can be calculated by the following formula:

[0026]

[0027] Taking (6) into account in the DC power flow, we can get

[0028]

[0029]

[0030] The reactive power flow can be further simplified as

[0031]

[0032] E-ISO is responsible for collecting and aggregating the segmented bidding curves of energy producers, as shown in the attached Figure 1 shown. represents the quantity and price of the bidding curve of unit g in segment b, p g,b represents the winning bid quantity of unit g in segment b. The bidding curve can be regarded as a set of price-quantity pairs. E-ISO can solve the following optimization problem to achieve clearing.

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041]

[0042] In the formula, the number after “:” indicates the Lagrange multiplier corresponding to the constraint condition. Bus indicates the bus set, Bus i represents the bus set connected to bus i, Line represents the line set, B g represents the bidding segment set of unit g, G represents the energy supplier set, G i represents the set of energy suppliers connected to bus i. Equation (10) indicates that E-ISO should minimize the power procurement cost under the premise of meeting demand and safety. Equations (11)-(12) are the active and reactive balance constraints of the node, where represents the active and reactive loads on bus i, are the Lagrange multipliers for active power balance constraint and reactive power balance constraint respectively. Constraints (13)-(15) represent the constraints on active and reactive power line flows, and their corresponding Lagrange multipliers are and are the maximum active capacity and reactive capacity of line ij respectively. Equation (16) represents the constraint on the size of the scalar quantity, and the corresponding multiplier is Formula (17) limits the reactive power output range of unit g. Respectively represent the minimum / maximum reactive output of the unit, and the corresponding multiplier is and Formula (18) limits the amplitude of the bus voltage. Represent the minimum and maximum values ​​of the voltage amplitude respectively, and the corresponding multiplier is and

[0043] in, That is, the node electricity price on bus i. Due to the nonlinear terms in equations (7)-(8) This will make the optimal power flow problem difficult to handle. Here, the incremental linearization method is used to linearize For processing, assume that the interval where the phase angle difference is Divide it into SN E segment, get against The constructed incremental linearization model is as follows:

[0044]

[0045]

[0046] 0≤d ij,k ≤1,k=1,2,...,SN E (twenty one)

[0047] d ij,k+1 ≤Bin ij,k ≤d ij,k ,k=1,2,...,SN E -1 (22)

[0048] Bin ij,k ∈{0,1},k=1,2,...,SN E -1 (23)

[0049] Where, d ij =[d ij,1 ,d ij,2 ,…,d ij,SN ],d ij,k is a continuous auxiliary variable with a value range of [0,1]. Equations (22)-(23) are filling constraints, which introduce a 0-1 auxiliary variable Bin ij,k .

[0050] right After linearization, the optimal power flow optimization model can be obtained:

[0051]

[0052] st type (11)-(18) (25)

[0053]

[0054]

[0055]

[0056]

[0057] In the formula, ε represents a small positive number, and the introduction of binary variables Bin is avoided by adding a penalty term to the objective function.ij,k . is the Lagrange multiplier of constraint (26). In the above optimal power flow model, the decision variable is p g,b ,d ij,k ,q g , and V i .

[0058] (2) Under the premise of meeting all heat demands, the goal is to purchase the least amount of heat in the heat market, and a heat market clearing model is constructed under the condition of safe operation constraints.

[0059] Assuming that the thermal system adopts mass regulation, the flow direction and mass flow rate of the working fluid in the pipeline are determined. Similar to the electricity market E-ISO, the independent system operator H-ISO of the thermal market is responsible for aggregating the bidding curves of all energy producers, which consists of multiple sets of price-quantity pairs. composition. It represents the price of the bth segment in the thermal bidding curve of unit g, in $ / MWh. It represents the number of the bth segment in the thermal bidding curve of unit group g, in MWh. H-ISO is responsible for solving the following thermal optimal power flow problem:

[0060]

[0061]

[0062]

[0063]

[0064]

[0065] In the formula, h g,b represents the winning heat of unit g in section b. The dual multiplier of formula (31) is Indicates the scalar h in the restriction g,b Must be less than the bid amount GG represents the collection of gas-fired CHP units. n represents the set of gas-fired CHP units connected to node n. HS represents the set of other heat sources except gas-fired CHP units. n Indicates the set of heat sources connected to node n except the gas-fired CHP unit. g represents the bidding section set of unit g. HN represents the node set of the district heating network. Equation (32) represents the node thermal power balance, and its dual multiplier It is the node heat price. They represent the set of pipes flowing out of node n and the set of pipes flowing into node n respectively.l It represents the product of the specific heat capacity and mass flow rate of hot water in pipe l. n represents the temperature of hot water flowing out from node n, Indicates the water temperature at the end of the pipe. represents the heat load on node n. The dual multiplier of equation (33) is φ n,l , Len l and l denote the length of pipe l and the temperature loss factor, respectively, and denote the temperature change of hot water in pipe l. Formula (34) limits the node temperature to a safe range. The dual multiplier of this constraint is and Respectively represent the minimum / maximum temperature of node n.

[0066] (3) Under the premise of meeting all natural gas demands, the total natural gas cost purchased by natural gas users is minimized, and a heat market clearing model is constructed under the condition of safe operation constraints.

[0067] Unlike the electricity and heat markets, the natural gas market operates asynchronously with the former two, and stops receiving market quotations later than the electricity and heat markets. The natural gas market independent system operator G-ISO is responsible for calculating the optimal natural gas flow and clearing it. Civilian loads have a higher priority than gas-fired power generation. Therefore, on the basis of ensuring civil loads, the gas-fired power generation load is gradually increased to calculate the segmented price of natural gas.

[0068] In order to facilitate the calculation and expression of power flow, we first construct the following correlation matrix A, where the internal element a mn The assignment rules are as follows:

[0069]

[0070] In the formula, inpipe(m) and outpipe(m) represent the natural gas pipelines flowing into and out of node m, respectively. Based on the association matrix A, the direction matrix can be further obtained:

[0071]

[0072]

[0073] Gas-fired CHP units are no longer energy producers in the natural gas system, but energy consumers. Their demand can affect natural gas prices to a certain extent. However, the price of natural gas sources has not yet been quoted in a segmented manner, so it is assumed that the unit price of natural gas sources is a known parameter. The purpose of G-ISO clearing is to clarify the winning bid gas volume and gas price for power generation gas loads. The optimal natural gas flow calculation model is as follows:

[0074]

[0075] stGS-GL-GC-GP=A MN F MN +A NM F NM :λ g (39)

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084] In the formula, the objective function is to minimize the system gas purchase cost, represents the unit price of natural gas purchased from gas source i during period t, SW i,t represents the amount of natural gas purchased from gas source i, G gas represents the set of natural gas sources, and Γ represents the set of time. Equation (39) is the node gas volume balance of natural gas, and its dual multiplier λ g It is the node gas price. The vectors representing gas source, gas load, compressor gas load and power generation gas load respectively. GN represents the number of nodes in the natural gas system. represents the vector of natural gas flow from node m to node n and from node n to node m, respectively, and LN represents the number of pipelines. Assuming that the natural gas system does not change the direction of the flow during the day, it is assumed that F MN All elements in F are positive numbers. NM All the elements in are negative. Formula (40) limits the pipeline flow to a safe range, and Line represents the set of pipelines. Formulas (41)-(42) are used to calculate the natural gas pipeline inventory L in the compressor-free pipeline mn. m,n,t ,pr m,t / pr n,t Respectively represent the air pressure at both ends of the pipeline mn in time period t. m,n It is related to the pipeline characteristics and is the pipe storage factor used to calculate the gas storage volume in the pipeline mn. Formula (43) limits the gas pressure to a safe range, pr m,min / prm,max Respectively represent the minimum / maximum safety gas pressure at node m. In order to ensure the smooth transportation of gas, there must be a certain amount of pipeline inventory in the pipeline. Formula (44) indicates that the pipeline inventory after a scheduling cycle must be greater than the safety value L min Formula (45) is used to describe that the pipeline pressure containing the compressor must satisfy the working characteristics of the compressor, CP m,n represents the set of compressors on the pipeline mn, σ min,k / σ max,k Respectively represent the minimum / maximum compression ratio of compressor k.

[0085] Among them, formula (47) is the Weymouth equation used to characterize the pressure and gas flow at both ends of the natural gas pipeline, W m,n is a constant coefficient, which is related to pipeline characteristics and gas characteristics and can be regarded as a constant under standard conditions. The Weymouth equation is non-convex, which brings considerable challenges to optimization. Second-order cone relaxation is often used to deal with the Weymouth equation, thereby converting the above model into second-order cone optimization.

[0086] There are two square terms in the Weymouth equation. One is about the mass flow rate f m,n,t The square of the node pressure pr m / n The square of . They are similar to the quadratic equation f(x) = x 2 Linearization. m,n,t and pr m / n The range of is divided into SN1 and SN2 segments, and an auxiliary variable δ1 with a value between 0 and 1 is introduced k,t and δ2 k,t The original natural gas optimal flow model is improved to the following model:

[0087]

[0088] st type (39)-(46) (49)

[0089]

[0090]

[0091]

[0092]

[0093] 0≤δ1 t,k ≤1,k=1,2,...,SN1 (54)

[0094] 0≤δ2 t,k ≤1,k=1,2,...,SN2 (55) In order to concisely express the natural gas optimal power flow model, the compact form of the above model is given here:

[0095] Min cost(x gs ,x flow ,x node ) (56)

[0096] stx gs -GD=MA·x flow :λ gas (57)

[0097] h(x gs ,x flow ,x node )=0 (58)

[0098] g(x gs ,x flow ,x node )≤0 (59)

[0099] In the formula, x gs , x flow and x node are vectors representing gas source output, natural gas flow in pipelines and node status respectively. GD represents natural gas load, and MA is the association matrix between nodes and pipelines. Equation (57) represents the node flow balance constraint, and equations (58) and (59) represent the equality and inequality constraints in the original model respectively.

[0100] 4) Build the optimal bidding model for the most complex gas-fired CHP units with the most coupled energy types, and derive the equivalent conversion model of the optimal bidding based on binary expansion and primal-dual complementary conditions.

[0101] Assume that the goal of the gas-fired CHP unit g connected to the grid bus i and the heating network node n is to maximize profit.

[0102]

[0103] In the formula, cost g represents the set of cost parameter segments of unit g, represents the natural gas cost of unit g in section f, e g,f Indicates the amount of natural gas consumed by unit g in section f.

[0104] Assuming that the unit has a high market share in both the electricity and heat markets and plays a leading role in the market, the constraints it is subject to are mainly divided into internal constraints and external constraints.

[0105] Internal constraints (technical and bidding rules constraints):

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113] Formula (61) is the energy conservation constraint, which is used to calculate the gas consumption of the CHP unit. g / TH g are the gas consumption for power generation and heating of unit g respectively. g,f It represents the amount of natural gas cost of unit g in segment f. They represent the maximum power generation and maximum heating power of unit g respectively. Equation (64) describes the thermal-electric coupling constraint of the CHP unit. They represent the rth segment boundary parameters of the feasible region of operation of unit g, R g is the boundary index set of the feasible domain of operation of unit g. Equations (65)-(67) are all bidding rule constraints. Equation (65) indicates that the first segment quantity of the bid cannot be greater than a certain percentage of the unit capacity δ∈(0,1), and different values ​​will be set according to different market rules. For example, the Greek power trading market stipulates δ=0.3. Equation (66) limits the maximum bid price, where π emax / π hmax They represent the maximum electricity bidding price and the maximum thermal bidding price respectively. Formula (67) shows that the bidding curve is monotonically rising.

[0114] External constraints (grid power flow and heating network power flow):

[0115] Assuming that the gas-fired CHP unit has a certain market power, has a certain understanding of the topological structure of the power grid and the heat network, and can estimate the information of other competitors based on history, the scenario generation method can be used to simulate the bids of competitors. Here, it is assumed that the scenario generation has been completed and the KKT constraints have been constructed for a certain scenario. Other scenarios only change the relevant parameters in the KKT conditions and have no effect on the overall model architecture and complexity.

[0116] Assume L eis the Lagrangian function of the power market flow clearing model. In addition to the equality constraints in the power clearing model, the inequality constraints can be rewritten as the following complementary constraints:

[0117]

[0118]

[0119]

[0120]

[0121] Assume L h is the Lagrangian function of the heat market flow clearing model. In addition to the equality constraints, the inequality constraints in the heat clearing model can be rewritten as the following complementary constraints:

[0122]

[0123]

[0124] In summary, the optimal bidding model of energy producers includes internal and external constraints, which can be transformed into a mathematical programming with equilibrium constraints (MPEC) based on the KKT condition:

[0125] Max(60) (74)

[0126] st internal constraints (61)-(67) (75)

[0127] External constraints (11)-(12), (32)-(33), (68)-(73) (76)

[0128] The complementary constraints in the external constraints can be transformed into linear constraints with binary variables by the Big-M method. However, the bilinear and The term still makes the model non-convex and difficult to solve. These bilinear terms can be handled by binary expansion, but it is necessary to consider which variable to use for binary expansion. In this model, there are 4 variables that can be directly selected, which are p g,b , and h g,b In the actual market, the winning bid power p g,b and the winning heat h g,b It is determined by the market ISO clearing, and energy producers cannot intervene. and node heat price They are the dual multipliers of the corresponding node power balance constraints, so it is even more impossible to expand them in binary, otherwise the problem may not be solved. Therefore, directly expanding the above four variables in binary is contrary to the actual situation and cannot be applied in practice. It is necessary to further derive the formula. From formula (68), we can get

[0129]

[0130]

[0131] Substitute equation (78) into equation (78) and eliminate the variable get

[0132]

[0133] From the above formula, we can see that the bilinear term Two new bilinear terms can be used and In these two bilinear terms, the bid price and bid electricity These are variables that energy producers can decide on their own, so it is more reasonable to perform a binary expansion of these two variables in the model.

[0134] Similarly, according to formula (72), we can get

[0135]

[0136]

[0137] Substitute equation (81) into equation (80) and eliminate the variable get

[0138]

[0139] From the above formula, we can see that the bilinear term Two new bilinear terms can be used and In these two bilinear terms, the bid heat price and bid heat Both are variables that energy producers can decide on their own, so the two variables will be binary expanded. Substituting equations (79) and (82) into the objective function equation (60), we can get

[0140]

[0141] right Expand as follows:

[0142]

[0143]

[0144] In the formula, It represents the upper and lower limits of the bidding price of unit g in segment b. The bidding accuracy is controlled by KE, and a binary auxiliary variable is introduced. It represents the minimum unit of change of the bid electricity price, which can also be regarded as the accuracy of the bid electricity price. g,b Multiplying by equation (84), we get

[0145]

[0146] In the formula, From the above formula, we can see that the bilinear term It can be expressed as a linear equation. It can be transformed into the following inequality group:

[0147]

[0148]

[0149] Similarly, other bilinear terms and It can also be replaced by the following system of equations containing integer variables.

[0150]

[0151]

[0152]

[0153] In the formula, The binary variables introduced include and They represent the minimum / maximum value of the bidding power of unit g in section b, They represent the minimum / maximum value of the bidding heat price of unit g in section b, They represent the minimum / maximum value of the bid heat of unit g in section b. KP, KH and KC control the accuracy of bid electricity, bid heat price and bid heat respectively. Indicates the bidding power accuracy of unit g in section b, Indicates the bid heat price accuracy of unit g in section b, Indicates the bid heat accuracy of unit g in section b.

[0154] By adding the constraint conditions (84)-(91) into the internal constraints, the bilinear terms in the objective function can be replaced, and the entire objective function is transformed into a linear equation:

[0155]

[0156] (5) According to the bidding process proposed, as attached Figure 2 The optimal bidder for the gas-fired CHP unit is as follows:

[0157] Step 1: The natural gas market G-ISO calculates the node gas price for power generation load on the premise of meeting the gas demand of all residents, and publishes this price information in the market bulletin board. Energy producers initialize cost parameters based on the natural gas price information and prepare for bidding;

[0158] Step 2: If only the optimal bidding curve of a single gas-fired CHP unit with market power is studied, after solving the optimization model (74)-(76), go to step 4;

[0159] Step 3: If the study is based on the N-person Cournot model to solve the Nash equilibrium, then the DA algorithm is embedded in this step. After energy producer 1 has completed the solution, its bidding curve is updated. The strategies of other suppliers remain unchanged, and energy producer 2 continues to solve models (74)-(76) to obtain a new bidding curve. This cycle continues until the bidding curves of all suppliers no longer change significantly, at which point the Nash equilibrium is reached;

[0160] Step 4: After all suppliers have updated their bidding curves in the electricity and heat markets, they submit new fuel demand to the natural gas market G-ISO, which clears and updates the node gas price. Energy producers update their cost parameters after receiving the new node gas price;

[0161] Step 5: Use the latest cost parameters to solve the optimization model (74)-(76). If the difference between the electricity and heat quotations of two consecutive bidding curves is greater than 0.1$ / MWh, go to step 3; if the difference between the electricity and heat quotations of two consecutive bidding curves is less than 0.1$ / MWh, stop the iteration; if the bidding curve is still not stable after multiple iterations, set the maximum number of iterations and select the bidding curve (quantity-price pair) with the highest profit from the many results as the final bidding information submitted to E-ISO and H-ISO.

[0162] Example description:

[0163] Since bidding and clearing are carried out in a certain cycle, this paper focuses on analyzing the bidding and market game results of energy producers in a single time period. The example uses the P5H8G7 system, which includes a 5-node power system, an 8-node thermal system and a 7-node natural gas system. Its topology is shown in the attached figure. Figure 3 As shown. There are three energy producers in the system, each with only one device. G1 and G2 both have gas-fired CHP units, and G3 is equipped with a coal-fired initial condensing unit. G1 occupies a leading position in the market and has greater market power. First, we focus on studying the optimal bidding strategy of G1 in the market. Table 1 gives the bidding parameters of G2 and G3, based on which the bidding curve can be drawn. Table 2 gives the parameters of all units.

[0164] Table 1 Bidding curves of energy producers G2 and G3 in the market

[0165]

[0166] Table 2 Unit parameters

[0167]

[0168] In this example, the maximum number of bidding segments is set to 5. The variables KE, KP, KH, and KC used by binary expansion to control the resolution of the bidding volume are all set to the same value K. Selecting different K values ​​can result in different bidding curves and different clearing results, as shown in Table 3-5. represents the electricity bid of G1 in segment b, represents the electricity bidding price of G1 in segment b, represents the heat bidding amount of G1 in section b, Represents the heat bidding price of G1 in section b.

[0169] Table 3 Bidding curves under different resolutions K

[0170]

[0171]

[0172] Table 4 Power bid quantity at different resolutions K

[0173]

[0174] Table 5 Thermal median values ​​at different resolutions K

[0175]

[0176] From Table 3, as the resolution K increases, the total profit continues to increase, but when K is greater than 7, the total profit no longer changes. This shows that the resolution K = 7 is sufficient to help G1 find the optimal bidding curve. In essence, the optimal bidding curve is not unique. When the bid price is lower than the opponent, this part of the bid volume can win the bid. When calculating the profit, it is not cleared according to the pay-as-bid method, but cleared according to the marginal cost of the system, so the quantity after the winning bid is decoupled from the bid electricity price of the corresponding segment, and the bid electricity price is not further utilized. Combining Tables 4 and 5, it can be seen that when the resolution K = 6, the thermal bid amount of energy producer G2 is 105MWh. Due to the thermal-electric coupling constraint of the CHP unit, the maximum power bid amount of G2 is limited to 113.7MWh. If there is no thermal-electric coupling constraint, all of G2's power generation capacity can win the bid.

Claims

1. A bidding method for a combined heat and power generating unit in a combined electricity-heat-gas market, characterized in that The method includes: Construct a linear clearing model for power market power flow; Construct a heat market clearing model; Construct a natural gas market clearing model; The optimal bidding model of gas-fired CHP units is constructed, and based on binary expansion and primal-dual complementary conditions, an equivalent conversion model of the optimal bidding is derived; Constructing a heat market clearing model based on safe operation constraints: Calculate the optimal natural gas flow and clear it, gradually increase the gas-fired power generation load on the basis of ensuring the civilian load, and thus calculate the segmented price of natural gas. Construct the following association matrix A, the internal element a mn The assignment is as follows: In the formula, inpipe(m) and outpipe(m) represent the natural gas pipelines flowing into node m and out of node m respectively. The direction matrix is ​​further obtained based on the association matrix A: Determine the unit price of natural gas source, clarify the winning gas volume and gas price of power generation type gas load, and the optimal natural gas flow calculation model is as follows: s.t.GS-GL-GC-GP=A MN F MN +A NM F NM :λ g In the formula, the objective function is to minimize the system gas purchase cost, represents the unit price of natural gas purchased from gas source i during period t, SW i,t represents the amount of natural gas purchased from gas source i, G gas represents the set of natural gas sources, Γ represents the set of time, and Equation (39) is the node gas volume balance of natural gas, and its dual multiplier λ g is the node gas price, They represent the vectors of gas source, gas load, compressor gas load and power generation gas load respectively. GN represents the number of nodes in the natural gas system. They represent the vectors of natural gas flow from node m to node n and from node n to node m, respectively. LN represents the number of pipelines. The natural gas system does not change the direction of the flow within a day. F MN All elements in F are positive numbers. NM All the elements in are negative numbers. Line represents a collection of pipelines. Used to calculate the natural gas pipeline inventory L in the pipeline without compressor mn m,n,t ,pr m,t / pr n,t They represent the air pressure at both ends of the pipeline mn in time period t, K m,n Pipeline storage factor used to calculate the gas storage capacity in pipeline mn, pr m,min / pr m,max Respectively represent the minimum / maximum safety air pressure at node m, Indicates that the pipe inventory after a scheduling cycle is greater than the safety value L min , Used to describe the pressure in the pipeline containing the compressor to meet the working characteristics of the compressor, CP m,n represents the set of compressors on the pipeline mn, σ min,k / σ max,k Respectively represent the minimum / maximum compression ratio of compressor k; f m,n,t and pr m / n The range of is divided into SN1 and SN2 segments, and an auxiliary variable δ1 with a value between 0 and 1 is introduced k,t and δ2 k,t , the original natural gas optimal flow model is improved to the following model: Min cost(x gs ,x flow ,x node ) Node traffic balance constraint: stx gs -GD=MA·x flow :λ gas The equality constraint h(x gs ,x flow ,x node )=0 The inequality constraint g(x gs ,x flow ,x node )≤0 In the formula, x gs , x flow and x node are the vectors representing the gas source output, the natural gas flow in the pipeline and the node status, respectively. GD represents the natural gas load. MA is the association matrix between the node and the pipeline. Formula (57) represents, and formulas (58) and (59) represent respectively; Based on binary expansion and primal-dual complementarity conditions, the equivalent transformation model of optimal bidding is derived as follows: The profit of the gas-fired CHP unit g connected to the power grid bus i and the heating network node n is maximized. In the above formula, cost g represents the set of cost parameter segments of unit g, represents the natural gas cost of unit g in section f, e g,f Indicates the amount of natural gas consumed by unit g in section f; The constraints imposed on the unit are divided into internal constraints and external constraints. Internal constraints: technical and bidding rules constraints; Energy conservation constraints: Where: TP g / TH g are the power generation gas consumption and heating gas consumption of unit g, E g,f represents the amount of natural gas cost of unit g in segment f, They represent the maximum power generation and heating power of unit g respectively; Thermal and electrical coupling constraints for CHP units: in: They represent the boundary parameters of the rth segment of the operating domain of unit g, R g is the boundary index set of the feasible region of operation of unit g, Bidding rules and regulations: External constraints: power grid flow and heat network flow; If L e is the Lagrangian function of the power market flow clearing model. In addition to the equality constraints in the power clearing model, the inequality constraints are rewritten as the following complementary constraints: If L h is the Lagrangian function of the heat market flow clearing model. In addition to the equality constraints in the heat clearing model, the inequality constraints are rewritten as the following complementary constraints: The optimal bidding model of energy producers includes internal constraints and external constraints, which are transformed into a mathematical programming with equilibrium constraints based on the KKT condition: right p g,b , and h g,b Before binary expansion of the four variables, the formula is further optimized: Further we get: Bilinear term With two new bilinear terms and The addition and substitution of the bidding price and bid electricity These are variables determined by energy producers themselves, so the model uses binary expansion for these two variables: get: From the above formula, we can see that the bilinear term With two new bilinear terms and In these two bilinear terms, the bid heat price and bid heat These are variables that are determined by energy producers themselves, so the two variables will be binary expanded: right Expand as follows: In the formula, It represents the upper and lower limits of the bidding price of unit g in segment b. The bidding accuracy is controlled by KE, and a binary auxiliary variable is introduced. Indicates the minimum unit of change in the bid electricity price, which is regarded as the accuracy of the bid electricity price. g,b Multiplying by equation (84), we get In the formula, From the above formula, we can see that the bilinear term Expressed as a linear equation, Eq. Transformed into the following inequality system: Similarly, other bilinear terms and We can also substitute the following system of equations with integer variables, In the formula, The binary variables introduced include and They represent the minimum / maximum value of the bidding power of unit g in section b, They represent the minimum / maximum value of the bidding heat price of unit g in section b, They represent the minimum / maximum value of the bid heat of unit g in section b. KP, KH and KC control the accuracy of bid electricity, bid heat price and bid heat respectively. Indicates the bidding power accuracy of unit g in section b, Indicates the bid heat price accuracy of unit g in section b, Indicates the bidding heat accuracy of unit g in section b, By adding the constraint condition into the internal constraint, the bilinear term in the objective function is replaced, and the entire objective function is transformed into a linear equation:

2. The bidding method for a cogeneration power generation unit in the electricity-heat-gas joint market according to claim 1, characterized in that: Collect and aggregate the segmented bidding curves of energy producers, represents the price of the bidding curve of unit g in segment b, p g,b Indicates the winning bid quantity of unit g in segment b. The bidding curve is composed of multiple price-quantity pairs. express.

3. The bidding method for a cogeneration power generation unit in the electricity-heat-gas joint market according to claim 2, characterized in that: The linear power flow clearing model of the power market is constructed based on the linear power flow constraint. Original AC power flow model: Where V i is the node voltage of bus i, θ ij represents the voltage phase difference between busbar i and busbar j, r ij and x ij Represent the line resistance and reactance respectively, splitting the above formula into real and imaginary parts: S ij =p ij +jq ij In the formula, p ij / q ij Respectively represent the active and reactive power flowing through line ij, and the active loss Expressed as The voltage phase difference at both ends of the line is controlled between 0.5-1.5pu, with V i ≈1p.u.,θ ij ≈sinθ ij , the power loss can be calculated by the following formula: Put the above formula into DC power flow, and we get Bring in The segmented bidding curve represented by: E-ISO aims to minimize electricity procurement costs while meeting demand and safety Node active power balance constraints: Node reactive power balance constraints: Among them, represents the active and reactive loads on bus i, They represent the Lagrange multipliers of active power balance constraint and reactive power balance constraint respectively; Constraints on active and reactive line flows The corresponding Lagrange multipliers are and are the maximum active capacity and reactive capacity of line ij respectively; The corresponding multiplier for the scalar size constraint is The reactive power output range of unit g is limited. Respectively represent the minimum / maximum reactive output of the unit, and the corresponding multiplier is and Limits the bus voltage amplitude in, Represent the minimum and maximum values ​​of the voltage amplitude respectively, and the corresponding multiplier is and In the above formula, Bus represents the bus set. i represents the bus set connected to bus i, Line represents the line set, B g represents the bidding segment set of unit g, G represents the energy supplier set, G i represents the set of energy suppliers connected to bus i; is the node electricity price on bus i. Linearization is performed, and the incremental linearization method is used to For processing, if the interval where the phase angle difference is Divide it into SN E segment, get against The constructed incremental linearization model is as follows: 0≤d ij,k ≤1,k=1,2,...,SN E Among them, d ij =[d ij,1 ,d ij,2 ,…,d ij,SN ],d ij,k is a continuous auxiliary variable with a value range of [0,1]; Filling constraint, introducing 0-1 auxiliary variable Bin ij,k : d ij,k+1 ≤Bin ij,k ≤d ij,k ,k=1,2,...,SN E -1 Am ij,k ∈{0,1},k=1,2,...,SN E -1 right After linearization, the optimal power flow optimization model is obtained: ε represents a positive number, which is achieved by adding a penalty term to the objective function.

4. The bidding method for a cogeneration power generation unit in a combined electricity-heat-gas market according to claim 1, characterized in that: The heat market clearing model is constructed based on the safety operation constraints, including: The thermal system adopts the method of quality regulation, which aggregates the bidding curves of energy producers. The bidding curves consist of multiple sets of price-quantity pairs. composition, It represents the price of the bth segment in the thermal bidding curve of unit g, in $ / MWh. represents the number of the bth segment in the thermal bidding curve of unit group g, in MWh, and solves the following thermal optimal power flow problem: The even multiplier is Node thermal power balance: The dual multiplier is φ n,l , In the formula, h g,b Indicates the winning heat of unit g in section b, limiting the winning quantity h g,b Must be less than the bid amount GG represents the collection of gas-fired CHP units. n represents the set of gas-fired CHP units connected to node n, HS represents the set of other heat sources except gas-fired CHP units, HS n represents the set of heat sources connected to node n except the gas-fired CHP unit, B g represents the bidding segment set of unit g, HN represents the node set of the district heating network, and its dual multiplier is the node heat price, They represent the set of pipes flowing out of node n and the set of pipes flowing into node n, respectively. l represents the product of the specific heat capacity and mass flow rate of hot water in pipe l, T n represents the temperature of hot water flowing out from node n, Indicates the water temperature at the end of the pipe. represents the heat load on node n, Len l represents the loss factor for the length of the pipe l, ζ l represents the loss factor of the pipe temperature, represents the temperature change of the hot water in the pipe l, and the dual multiplier is and Respectively represent the minimum / maximum temperature of node n.

5. The bidding method for a cogeneration power generation unit in a combined electricity-heat-gas market according to claim 1, characterized in that: It also includes solving the optimal bidding scheme for gas-fired CHP units according to the proposed bidding process: Step 1: Determine the node gas price of the power generation gas load, publish the price information, and energy producers initialize cost parameters based on the natural gas price information to participate in the bidding; Step 2: If only the optimal bidding curve of a single gas-fired CHP unit with market power is studied, then solve Internal constraint and external constraint model, go to step 4; Step 3: If the study is based on the N-person Cournot model to solve the Nash equilibrium, the DA algorithm is embedded in this step. After energy producer 1 has solved the problem, its bidding curve is updated. The strategies of other suppliers remain unchanged, and energy producer 2 continues to solve the model. The internal and external constraint models are used to obtain a new bidding curve, which is repeated until the bidding curves of all suppliers no longer change significantly, and the Nash equilibrium is reached. Step 4: After all suppliers have updated their bidding curves in the electricity and heat markets, they submit new fuel demand to the natural gas market G-ISO. G-ISO clears and updates the node gas price. Energy producers update their cost parameters after obtaining the new node gas price. Step 5: Solve the optimization model with the latest cost parameters For the internal and external constraint models, if the difference between the electricity and heat bids of two consecutive bidding curves is greater than 0.1$ / MWh, go to step 3; if the difference between the electricity and heat bids of two consecutive bidding curves is less than 0.1$ / MWh, stop the iteration; if the bidding curve is still not stable after multiple iterations, set the maximum number of iterations and select the bidding curve with the highest profit from the many results as the final bidding information submitted to E-ISO and H-ISO.