Dual stochastic electrothermal coupling optimization scheduling method considering thermal resistance uncertainty and wind power uncertainty

Abstract

The dual random electrothermal coupling optimization scheduling method taking into account the uncertainty of thermal resistance and the uncertainty of wind power includes the following steps: Step 1: Considering that the pipeline damage factor in the actual electrothermal coupling scheduling will cause the thermal resistance to change, thereby affecting the system, establish the thermal resistance uncertainty constraint of the thermal network; Step 2: Based on the thermal resistance uncertainty constraint of the thermal network established in Step 1, consider the uncertainty of wind power and the impact of the uncertainty of thermal resistance of the thermal network in the electrothermal coupling scheduling system structure, and establish a dual random electrothermal coupling optimization scheduling model taking into account the uncertainty of wind power and thermal resistance; Step 3: Based on the optimization scheduling model established in Step 2, propose an information gap robust optimization solution strategy to obtain the optimal scheduling result. The present invention takes into account the changes in the heat loss characteristics of the thermal network caused by the uncertainty of thermal resistance and its impact on the reliability of the electrothermal coupling scheduling decision, which is more in line with actual working conditions. In addition, the information gap robust optimization solution strategy is used to ensure the economy and stability of the system optimization scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of energy system control, and in particular to a dual random electrothermal coupling optimization scheduling method taking into account thermal resistance uncertainty and wind power uncertainty. Background Art

[0002] Cogeneration units typically operate in a "heat-to-power" mode, with their electrical power output largely limited and constrained by their thermal power output. This rigid constraint severely squeezes the grid accessibility of wind power, and wind power absorption is becoming a bottleneck in energy development, requiring further research and resolution. The emergence of coupled heat and power scheduling offers a solution to this problem, effectively alleviating the constraints of "heat-to-power" and increasing grid accessibility for wind power. However, in coupled heat and power scheduling, uncertainties in actual operating conditions pose challenges to the reliability of traditional deterministic decision-making plans. Analyzing and resolving these uncertainties plays a crucial role in ensuring the safe and reliable operation of coupled heat and power systems.

[0003] Existing designs for the study of uncertainty problems include: using scenario analysis to describe wind power uncertainty in the scheduling of electrothermal coupling systems; using Latin hypercube sampling to generate scenarios for wind power uncertainty and constructing its probability density function that conforms to the beta distribution; in the case of uncertainty problems on the load side of electrothermal coupling scheduling, interval models of heat pumps, power grids, heating networks and other equipment are established respectively, and an iterative algorithm based on interval expansion is used to solve the electrothermal coupling scheduling model; after sufficient training in uncertainty scenarios such as wind power, electricity prices, and electrothermal loads, approximate value functions are used to derive empirical knowledge to help the electrothermal coupling system make scheduling plans to deal with uncertainties.

[0004] However, most current designs focus on optimizing scheduling by considering the uncertainties of power system sources and loads. These designs assume that the insulation layer of heating network pipelines is in an ideal state, and when calculating heat losses in the heating network, the pipeline thermal resistance is assumed to be constant. This ignores the variability in the network's heat loss characteristics caused by thermal resistance uncertainty and its impact on the reliability of power-heat coupling scheduling decisions. However, in actual engineering applications, after a long period of operation, the supply and return water pipelines experience not only aging and damage to the insulation layer itself, but also the accumulation and corrosion of contaminants from the surrounding environment. The insulation layer inside the pipes also changes due to scaling caused by hot water. This leads to random errors between the actual thermal resistance and the designed value, which in turn causes variations in the calculated heat losses of the heating network. Summary of the Invention

[0005] In response to the above technical problems, the present invention provides a dual random electrothermal coupling optimization scheduling method that takes into account the uncertainty of thermal resistance and wind power uncertainty. Compared with the traditional thermal network pipeline model that has always used deterministic thermal resistance in electrothermal coupling scheduling, the present invention takes into account the changes in the heat loss characteristics of the thermal network caused by the uncertainty of thermal resistance and its impact on the reliability of electrothermal coupling scheduling decisions, which is more in line with actual working conditions; and uses an information gap robust optimization solution strategy to ensure the economy and stability of system optimization scheduling.

[0006] The technical solution adopted by the present invention is:

[0007] The dual stochastic electrothermal coupling optimization scheduling method considering thermal resistance uncertainty and wind power uncertainty includes the following steps:

[0008] Step 1: Considering the fact that pipeline damage in actual electrothermal coupling scheduling may cause changes in thermal resistance, thereby affecting the system, uncertainty constraints on the thermal resistance of the heating network are established.

[0009] Step 2: Based on the uncertainty constraints of the thermal resistance of the heating network established in step 1, the uncertainty of wind power and the uncertainty of the thermal resistance of the heating network are considered in the structure of the electric-thermal coupling scheduling system, and a dual stochastic electric-thermal coupling optimization scheduling model taking into account the uncertainties of wind power and thermal resistance is established;

[0010] Step 3: Based on the dual random electrothermal coupling optimization scheduling model established in step 2, an information gap robust optimization solution strategy is proposed to obtain the optimal scheduling result.

[0011] In the above step 1, in actual working conditions, the thermal resistance of the inner and outer layers of the water supply pipe and the return pipe will change due to damage factors, thereby affecting the system, as follows:

[0012] (1) Relationship between the change of thermal resistance and the change of thermal resistance heat loss:

[0013]

[0014] In formula (1), Δq is the fluctuation in heat loss in the heating network caused by changes in thermal resistance, ΔR is the change in thermal resistance, c is the specific heat capacity of hot water, ρ is the density of hot water, A is the internal cross-sectional area of ​​the pipe, m is the mass flow rate of hot water, q is the heat loss in the heating network, and Δx and Δt are the spatial and time steps in the difference format, respectively. It can be seen that the thermal resistance and the heat loss in the heating network show opposite trends.

[0015] (2) Relationship between heat loss changes in the heating network and pipeline temperature fluctuations:

[0016]

[0017] In formula (2), △T(x, t) is the temperature change caused by heat loss fluctuation at the distance x from the pipeline inlet at time t, and △q is the heat loss fluctuation of the heating network caused by the change of thermal resistance. It can be seen from formula (2) that the change of heat loss of the heating network will further affect the fluctuation of pipeline temperature.

[0018] The relationship between the supply and return water pipe temperature change and the heat loss of the heating network is quantified, and the relationship between the pipe temperature at any spatial position and the heat loss of the heating network is obtained, as shown in the following equations (3) and (4):

[0019]

[0020]

[0021] In formula (3) and formula (4), c is the specific heat capacity of hot water; m v and m r are the mass and flow rate of hot water in the supply and return pipes respectively; T v (x,t) and T r (x, t) is the temperature at the distance x from the inlet of the supply and return pipes at time t; T v (x-△x,t) and T r (x-△x,t) is the temperature at time t at a distance x-△x from the inlet of the supply and return pipes; is the historical temperature sequence of the pipe in space after differentiation, and are the symmetrical heat losses of the supply and return pipes with symmetrical thermal resistance, q a is the asymmetric heat loss of the asymmetric thermal resistance.

[0022] From Equations (3) and (4), it can be seen that the uncertainty of thermal resistance will lead to corresponding changes in the heat loss characteristics of the heating network, which in turn causes fluctuations in the temperature of the supply and return pipes, thereby affecting the optimal scheduling of the electrothermal coupling system.

[0023] In the step 1, uncertainty constraints on the thermal resistance of the thermal network are established.

[0024] Since factors such as surface corrosion and scaling in the pipe are deeply uncertain, the actual value of the thermal resistance obtained through parameter identification does not have the characteristics of probability distribution. Therefore, the envelope constraint method is used to express the uncertainty of the thermal resistance:

[0025] The uncertainty models of symmetric thermal resistance and asymmetric thermal resistance in the uncertainty problem are shown in the following equations (5) and (5):

[0026] (1-α1)R s ≤R′ s ≤(1+α1)R s (5);

[0027] (1-α2)Ra ≤R′ a ≤(1+α2)R a (6);

[0028] In formula (5) and formula (6), α1 and α2 are the uncertainty ranges of symmetric thermal resistance and asymmetric thermal resistance respectively; R s and R a The thermal resistances of the symmetrical and asymmetrical heat loss branches are respectively; R′ a , R′ s are the actual values ​​of asymmetric thermal resistance and symmetric thermal resistance respectively.

[0029] At the same time, the operation constraints of the original heat network pipeline need to be modified, and the deterministic constraints are replaced by uncertain constraints, as shown in the following equations (7) to (14):

[0030]

[0031]

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039] In formulas (7) to (14), C′1, C′2, C′3, C′4, and C′5 are weight coefficients; and They are the pipe temperatures during water supply and return respectively; and are the inlet temperatures of the water supply pipe e at time t-τ1 and t-τ2 respectively; and are the inlet temperatures of the return pipe e at time t-τ1 and t-τ2 respectively; T b is the ambient temperature; m e,K (τ2-τ) and m e,K+1 (τ-τ1) are the mass and flow rate of the mass at τ2-τ and τ-τ1 respectively; ξ e is the loss coefficient of pipeline temperature; l is the total length of the pipeline; m erepresents the mass flow rate of hot water; R' is the equivalent thermal resistance per unit length of the pipe; R' s , R′ a are the thermal resistances of the symmetrical heat loss and asymmetrical heat loss branches respectively.

[0040] In step 2, in the dual stochastic electrothermal coupling optimization scheduling model, the optimization goal is to minimize the total system scheduling cost while ensuring the maximum absorption of wind power, and wind power abandonment is added as a penalty term. The objective function is shown in the following formula:

[0041] min F total =F chp +F con +F wind (15);

[0042]

[0043]

[0044] F k,ST =u k,t (1-u k,t )f k,ST (18);

[0045]

[0046] In formulas (15) to (19), the objective function is defined as the total scheduling cost F total , F chp is the dispatching cost of the extraction steam cogeneration unit; F con is the dispatching cost of coal-fired power units; F wind In order to promote the consumption of wind power, a penalty cost for wind curtailment is set; t is a certain moment of model scheduling, T is the total scheduling time of the model, ψ chp , ψ con and ψ wind represent the collection of extraction steam cogeneration units, coal-fired thermal power units and wind turbine units, respectively; the subscripts g, k and i represent the unit numbers of extraction steam cogeneration units, coal-fired thermal power units and wind turbine units, respectively; f g,p1 and f g,p2 is the electric power cost coefficient of the cogeneration unit; f g,q1 and f g,q2 is the thermal power cost coefficient of the cogeneration unit; f g,pq and f g,pq0 is the electric heating cost coefficient.

[0047] f k0 、f k1 and f k2 are the power cost coefficients of fuel-fired, coal-fired and thermal power units k respectively; F k,STis the start-up and shutdown cost of coal-fired power unit k, u k,t is the operating status of coal-fired power unit k at time t, f k,ST is the single startup cost of coal-fired power unit k, δ i is the penalty factor of wind turbine i, P i,max is the maximum electric power of wind turbine i.

[0048] The operation constraints of the optimal dispatch model include power system operation constraints and thermal system operation constraints.

[0049] (1) The power system operation constraints mainly include: the operation constraints of extraction steam cogeneration units, the operation constraints of coal-fired power units, the operation constraints of wind turbines, the active power balance constraints of the power network, and the line flow operation constraints of the power network.

[0050] Operation constraints of extraction steam cogeneration unit g:

[0051] 0≤P g,t ≥r g Q g,t (20)

[0052] F g,min ≤ρ g,p P g,t +ρ g,q Q g,t ≤F g,max (twenty one)

[0053] 0≤Q g,t ≤Q g,max (twenty two)

[0054] In formulas (20) to (22), P g,t and Q g,t are the electric power and thermal power of the extraction steam cogeneration unit g at time t; r g is the electric-thermal coupling coefficient of the extraction steam cogeneration unit under back pressure conditions; ρ g,p and ρ g,q are the coal consumption rates of the extraction steam cogeneration unit for electric power and thermal power, respectively; Q g,max F is the upper limit of thermal power of the extraction steam cogeneration unit; g,max and F g,min The maximum and minimum values ​​of the unit's coal intake.

[0055] Operation constraints of coal-fired power unit k:

[0056] P k,min ≤P k,t ≤P k,max (twenty three)

[0057] -Rk,down ≤P k,t -P k,t-1 ≤R k,up (twenty four)

[0058]

[0059] In formulas (23) to (25), P k,t is the electric power of coal-fired power unit k at time t; P k,t-1 is the electric power of coal-fired power unit k at time t-1; P k,min and P k,max are the minimum and maximum power of coal-fired power unit k; R k,down and R k,up are the downward and upward climbing rates of coal-fired power unit k; U k,t-1 is the state quantity of coal-fired power unit k starting and stopping at time t-1; U k is the maximum allowed number of starts and stops of coal-fired power unit k.

[0060] Operation constraints of wind turbine i:

[0061] 0≤P i,t ≤P i,max (26)

[0062] In formula (26), P i,t is the electric power of wind turbine i at time t, P i,max is the maximum allowable power of wind turbine i at time t.

[0063] Active power balance constraints in power networks:

[0064]

[0065] In formula (27), P j,t is the power demand of system load j at time t; ψ chp , ψ con , ψ wind and ψ d They represent the collection of extraction steam cogeneration units, coal-fired power units, wind turbines and system electrical loads respectively; the subscript j represents the number of the system electrical load.

[0066] Line flow operation constraints of power networks:

[0067] L l,min ≤L l,t ≤L l,max (28);

[0068]

[0069] In formula (28) to formula (29), the subscript l represents the number of the power network line; G l- is the power allocation factor; L l,t is the transmission power of line l at time t; L l,min and L l,max are the minimum and maximum power transmission values ​​of line l respectively.

[0070] (2) The operation constraints of the thermal system include the node equations and pipeline equations of the thermal system:

[0071] Nodal equations of the thermal system:

[0072] Heat balance equation for heat source and heat load nodes:

[0073]

[0074]

[0075] In formula (30) to formula (31), m n is the mass flow rate of the heat source node and the heat load node n; c is the specific heat capacity; Q g,t and Q d,t are the heat power of heat source node n and the heat demand of heat load node n at time t respectively; and are the inlet temperature and outlet temperature of the heat source node and heat load node n at time t respectively; ψ HS represents the set of heat source nodes; ψ h Represents a collection of heat load nodes.

[0076] Heat balance equations for supply and return pipe network nodes:

[0077]

[0078]

[0079] In formula (32) to formula (33), and are the mass flow rates of the supply and return pipes e connected to node n, respectively; and are the outlet temperatures of the supply and return pipes e connected to node n at time t, respectively; e represents the pipe number; and They represent the front-side and back-side pipeline sets connected to the heating network node n respectively; ψ HS represents the set of heat source nodes; ψ HES Represents a collection of heat exchange stations.

[0080] Temperature conservation assumption for supply and return pipe network nodes:

[0081]

[0082]

[0083] In formula (34) to formula (35), and are the inlet temperatures of the supply and return pipes e connected to node n at time t; and are the temperatures at the heat source node and heat load node n at time t respectively; ψ HS represents the set of heat source nodes; ψ h represents the heat load node set; ψ HES Represents a collection of heat exchange stations.

[0084] Piping equations for the thermal system:

[0085]

[0086]

[0087]

[0088]

[0089] In formula (36) to formula (37), and are the pipe temperatures of water supply / return respectively; T b is the ambient temperature; e is the loss coefficient of pipeline temperature; R is the equivalent thermal resistance per unit length of pipeline; l is the total length of pipeline; R s and R a The thermal resistances of the symmetrical and asymmetrical heat loss branches are respectively; m e represents the mass flow rate of hot water; c is the specific heat capacity.

[0090] Taking into account the impact of wind power uncertainty, the uncertainty modeling of wind power is as follows:

[0091] (1-α3)P i,max ≤P′ i,max ≤(1+α3)P i,max (40);

[0092] α3 is the uncertainty range of the maximum wind power, P′ i,max is the actual value of the maximum wind power. The wind turbine operation constraints are rewritten as:

[0093] 0≤P i,t ≤P′ i,max (41);

[0094] Find the largest uncertainty range when the cost is allowed to a certain extent, so as to improve the stability of the system.

[0095] In step 3, the model is transformed using information gap robust optimization, and the optimization scheduling model is transformed into an upper-layer optimization scheduling model and a lower-layer optimization scheduling model:

[0096] (1) Upper-level optimization scheduling model:

[0097] The optimization goal of the upper-level planning is to maximize the system uncertainty range while meeting the basic operating constraints of the power system and thermal system.

[0098] maxα=k1α1+k2α2+k3α3 (42);

[0099] k1+k2+k3=1 (43);

[0100] In equations (42) and (43), k1, k2, and k3 are weight factors, which guide the model to obtain an uncertainty range that is consistent with the actual situation, so that the scheduling plan is more reasonable and reliable. α represents the fluctuation amplitude of the actual maximum power output, that is, the uncertainty radius.

[0101]

[0102] In formula (44), the objective function J up (x) corresponds to formula (42), x up are the decision variables of the upper-level planning, which are the power P of the cogeneration unit at time t g,t , thermal power Q of cogeneration unit g,t , thermal power unit output P k,t , wind turbine output P i,t ,y up are the state variables of the upper planning, which are the inlet temperature of the heat supply and return pipe e and Outlet temperature of heat network supply and return water pipe e and Inlet and outlet temperatures of the heat source node and heat load node n and m up (x up ,y up ) and n up (x up ,y up) are the equality constraints and inequality constraints of the upper-level planning, respectively, corresponding to the operation constraints of the extraction steam cogeneration unit shown in equations (20) to (22), the operation constraints of the coal-fired power unit shown in equations (23) to (25), the operation constraints of the wind turbine unit shown in equation (41), the active power balance constraint of the power network shown in equation (27), the power network line flow operation constraints shown in equations (28) to (29), the heat balance constraints of the heat source and heat load nodes shown in equations (30) to (31), the heat balance constraints of the supply and return water pipeline network nodes shown in equations (32) to (33), the temperature conservation assumption of the supply and return water pipeline network nodes shown in equations (34) to (35), the supply and return water pipeline operation constraints shown in equations (7) to (14), and the weight factor equation shown in equation (43).

[0103] (2) Lower-level optimization scheduling model:

[0104] The optimization goal of the lower layer is that when the system fluctuates within the uncertainty range, the maximum total scheduling cost of the system must be less than its expected target, as shown in the following formula:

[0105] max F total (45);

[0106] F total ≤F ro (46);

[0107] Where, F total is the total scheduling cost.

[0108] Based on the above optimization objectives, the lower-level planning of the constructed random electrothermal coupling information gap robust optimization scheduling model is shown in the following formula (47):

[0109]

[0110] In formula (47), the objective function J low (x) corresponds to formula (45), x low are the decision variables for the lower-level planning, which are the actual values ​​of the symmetrical thermal resistance R′ s , the actual value of asymmetric thermal resistance R' a And the actual maximum output of wind power P′ i,max , n low (x low ) is the inequality constraint of the lower-level planning, corresponding to the system uncertainty model constraints shown in Equations (5) to (6) and (40), and the economic constraints shown in Equation (46).

[0111] In step 3, before solving the model, the original planning problem needs to be reasonably relaxed, so that the lower-level planning of the scheduling model is equivalently replaced with its KKT conditions, and the lower-level planning model is processed:

[0112]

[0113] In formula (48), P g,t , Q g,t 、P k,t and P i,t are all decision variables of upper-level planning, R′ s , R′ a and P′ i,max are all decision variables of the lower-level planning. The equivalent KKT conditions of the lower-level planning are shown in the following equations (49) to (64):

[0114] -δ i +μ1δ i -μ2+μ3-μ4+μ5-μ6+μ7=0 (49)

[0115]

[0116] μ2((1-α1)R s -R′ s )=0 (51)

[0117] μ3(-(1+α1)R s +R′ s )=0 (52)

[0118] μ4((1-α2)R a -R′ a )=0 (53)

[0119] μ5(-(1+α2)R a +R′ a )=0 (54)

[0120] μ6((1-α3)P i,max -P′ i,max )=0 (55)

[0121] μ7(-(1+α3)P i,max +P′ i,max )=0 (56)

[0122] μ1, μ2, μ3, μ4, μ5, μ6, μ7≥0 (57)

[0123]

[0124] (1-α1)R s -R′ s ≤0 (59)

[0125] -(1+α1)R s+R′ s ≤0 (60)

[0126] (1-α2)R a -R′ a ≤0 (61)

[0127] -(1+α2)R a +R′ a ≤0 (62)

[0128] (1-α3)P i,max -P′ i,max ≤0 (63)

[0129] -(1+α3)P i,max +P′ i,max ≤0 (64)

[0130] In equations (49) to (64), μ1, μ2, μ3, μ4, μ5, μ6, and μ7 are all Lagrange multipliers. Equations (48) to (54) are the complementary relaxation conditions of the KKT condition. Equation (55) is the feasibility criterion of the dual problem of the KKT condition. Equations (56) and (62) are the feasibility criteria of the original problem of the KKT condition.

[0131] After the above relaxation strategy is applied, the solver can be used to solve it directly.

[0132] The present invention provides a dual stochastic electrothermal coupling optimization scheduling method that takes into account thermal resistance uncertainty and wind power uncertainty. The technical effects are as follows:

[0133] (1) The dual random electrothermal coupling optimization scheduling model designed by the method of the present invention, which takes into account the uncertainty of thermal resistance and wind power, is more in line with actual working conditions than the traditional electrothermal coupling scheduling model. Taking into account the changes in the heat loss characteristics of the thermal network caused by the uncertainty of thermal resistance and its impact on the reliability of the electrothermal coupling scheduling decision, after adopting the information gap robust optimization strategy, decision makers can choose different scheduling plans to coordinate the relationship between uncertainty and cost.

[0134] (2) The present invention sets up a double random electrothermal coupling scheduling optimization problem around the uncertainty changes of thermal resistance and the uncertainty problems of wind power, which further solves the impact of uncertainty problems on scheduling, contributes to the safe operation and promotion of electrothermal coupling scheduling, and has important theoretical significance and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0135] Figure 1 It is the structural diagram of the electric-thermal coupling scheduling model.

[0136] Figure 2 It is a graph of electric heat load demand and wind power forecast.

[0137] Figure 3 This is the heat loss curve of the return pipe in scenario 1.

[0138] Figure 4 This is the heat loss curve of the water supply pipeline in scenario 1.

[0139] Figure 5 This is a temperature change trend diagram of the supply and return pipes under different schemes in scenario 1.

[0140] Figure 6 This is the heat loss curve of the return pipe in scenario 2.

[0141] Figure 7 This is the heat loss curve of the water supply pipeline in scenario 2.

[0142] Figure 8 This is a temperature change trend diagram of the supply and return pipes under different schemes in scenario 2.

[0143] Figure 9 It is a trend chart of scheduling costs and system uncertainty range changes under different scenarios. DETAILED DESCRIPTION

[0144] The dual stochastic electrothermal coupling optimization scheduling method considering thermal resistance uncertainty and wind power uncertainty includes the following steps:

[0145] Step 1: Considering the fact that pipeline damage in actual electrothermal coupling scheduling may cause changes in thermal resistance, thereby affecting the system, uncertainty constraints on the thermal resistance of the heating network are established.

[0146] Step 2: Based on the uncertainty constraints of the thermal resistance of the heating network established in step 1, the uncertainty of wind power and the uncertainty of the thermal resistance of the heating network are considered in the structure of the electric-thermal coupling scheduling system, and a dual stochastic electric-thermal coupling optimization scheduling model taking into account the uncertainties of wind power and thermal resistance is established;

[0147] Step 3: Based on the optimization scheduling model established in step 2, an information gap robust optimization solution strategy is proposed to obtain the optimal scheduling result.

[0148] Through the above steps, the optimal scheduling of the dual random electrothermal coupling system taking into account the uncertainty of thermal resistance and wind power is achieved.

[0149] In step 1, in actual working conditions, the thermal resistance of the inner and outer layers of the water supply and return pipes may change due to damage and other factors, thus affecting the system:

[0150] (1) Relationship between the change of thermal resistance and heat loss and the change of thermal resistance:

[0151]

[0152] Where Δq is the fluctuation in heat loss from the heating network caused by changes in thermal resistance, ΔR is the change in thermal resistance, c is the specific heat capacity of the hot water, ρ is the density of the hot water, A is the internal cross-sectional area of ​​the pipe, m is the mass flow rate of the hot water, q is the heat loss from the heating network, and Δx and Δt are the spatial and time steps in the difference scheme, respectively. It can be seen that the thermal resistance and the heat loss from the heating network show opposite trends.

[0153] (2) Relationship between heat loss changes in the heating network and pipeline temperature fluctuations:

[0154]

[0155] Where ΔT(x,t) is the temperature change caused by heat loss fluctuations at time t at a distance x from the pipeline inlet, Δq is the heat loss fluctuation of the heating network caused by changes in thermal resistance, c is the specific heat capacity of the hot water, ρ is the density of the hot water, A is the internal cross-sectional area of ​​the pipeline, m is the mass flow rate of the hot water, Δx and Δt are the spatial and time steps in the difference scheme, respectively. This formula shows that changes in heat loss in the heating network will further affect pipeline temperature fluctuations.

[0156] The relationship between the supply and return water pipe temperature changes and the heat loss of the heating network is quantified, and the relationship between the pipe temperature at any spatial position and the heat loss of the heating network is obtained as shown in the following formula:

[0157]

[0158]

[0159] Where △x is the spatial step size in the difference format, T v (x,t) and T r (x, t) is the temperature at the distance x from the inlet of the supply and return pipes at time t, T v (x-△x,t) and T r (x-△x,t) is the temperature at time t at a distance x-△x from the inlet of the supply and return pipes, and is the historical temperature sequence of the pipe in space after differentiation. and are the symmetrical heat loss of the supply and return pipes with symmetrical thermal resistance, q a is the asymmetric heat loss due to the asymmetric thermal resistance. The formula shows that the uncertainty in thermal resistance will lead to corresponding changes in the heat loss characteristics of the heating network, which in turn causes fluctuations in the temperature of the supply and return pipes, thus affecting the optimal scheduling of the electrothermal coupling system.

[0160] In step 1, uncertainty constraints on the thermal network's thermal resistance are established. Because factors such as surface corrosion and scaling within the pipes are deeply uncertain, the actual thermal resistance value obtained through parameter identification lacks the characteristics of a probability distribution. Therefore, an envelope constraint method is used to describe the uncertainty of the thermal resistance.

[0161] The uncertainty model of symmetric thermal resistance and asymmetric thermal resistance in uncertainty problems is shown as follows:

[0162] (1-α1)R s ≤R′ s ≤(1+α1)R s (5)

[0163] (1-α2)R a ≤R′ a ≤(1+α2)R a (6)

[0164] Where α1 and α2 are the uncertainty ranges of symmetrical thermal resistance and asymmetrical thermal resistance respectively, R s and R a are the thermal resistances of the symmetrical and asymmetrical heat loss branches, respectively. At the same time, the operating constraints of the original heat network pipelines need to be modified, replacing the deterministic constraints with uncertain constraints, as shown in the following formula:

[0165]

[0166]

[0167]

[0168]

[0169]

[0170]

[0171]

[0172]

[0173]

[0174] Where C′1, C′2, C′3, C′4 and C′5 are weight coefficients; and They are the pipe temperatures during water supply / return respectively; and are the inlet temperatures of the water supply pipe e at time t-τ1 and t-τ2 respectively; and are the inlet temperatures of the return pipe e at time t-τ1 and t-τ2 respectively; T b is the ambient temperature; ξ′ e is the loss coefficient of pipeline temperature; R' is the equivalent thermal resistance per unit length of pipeline; l is the total length of pipeline; R' s and R′ aThe thermal resistances of the symmetrical and asymmetrical heat loss branches are respectively; m e represents the mass flow rate of hot water; c is the specific heat capacity.

[0175] The traditional electro-thermal coupling scheduling model adds the influence of wind power uncertainty on the basis of step one, and establishes a dual random electro-thermal coupling optimization scheduling model that takes into account the uncertainties of wind power and thermal resistance.

[0176] In the optimization scheduling model, the optimization goal is to minimize the total system scheduling cost while ensuring the maximum absorption of wind power, and wind power abandonment is added as a penalty item. The objective function is shown in the following formula:

[0177] min F total =F chp +F con +F wind (15)

[0178]

[0179]

[0180] F k,ST =u k,t (1-u k,t )f k,ST (18)

[0181]

[0182] In the formula, the objective function is defined as the total scheduling cost F total , F chp F is the dispatching cost of the extraction steam cogeneration unit; con is the dispatching cost of coal-fired power units; F wind In order to promote the consumption of wind power, a penalty cost for wind curtailment is set; t is a certain moment of model scheduling, T is the total scheduling time of the model, ψ chp , ψ con and ψ wind The subscripts g, k and i represent the unit numbers of the extraction steam cogeneration unit, coal-fired power unit and wind turbine unit, respectively. g,p1 、f g,p2 、f g,q1 、f g,q2 、f g,pq and f g,pq0 are the coal consumption cost coefficient of the extraction steam cogeneration unit g, f k0 、f k1 and f k2 are the power cost coefficient of coal-fired power unit k, F k,STis the start-up and shutdown cost of coal-fired power unit k, u k,t is the operating status of coal-fired power unit k at time t, f k,ST is the single startup cost of coal-fired power unit k, δ i is the penalty factor of wind turbine i, P i,max is the maximum electric power of wind turbine i.

[0183] The operation constraints of the optimal dispatch model include power system operation constraints and thermal system operation constraints.

[0184] (1) The power system operation constraints mainly include: the operation constraints of extraction steam cogeneration units, the operation constraints of coal-fired power units, the operation constraints of wind turbines, the active power balance constraints of the power network, and the line flow operation constraints of the power network.

[0185] Operation constraints of extraction steam cogeneration unit g:

[0186] 0≤P g,t ≥r g Q g,t (20)

[0187] F g,min ≤ρ g,p P g,t +ρ g,q Q g,t ≤F g,max (twenty one)

[0188] 0≤Q g,t ≤Q g,max (twenty two)

[0189] Where, P g,t and Q g,t are the electric power and thermal power of the extraction steam cogeneration unit g at time t; r g is the electric-thermal coupling coefficient of the extraction steam cogeneration unit under back pressure conditions; ρ g,p and ρ g,q are the coal consumption rates of the extraction steam cogeneration unit for electric power and thermal power, respectively; Q g,max F is the upper limit of thermal power of the extraction steam cogeneration unit; g,max and F g,min The maximum and minimum values ​​of the unit's coal intake.

[0190] Operation constraints of coal-fired power unit k:

[0191] P k,min ≤P k,t ≤P k,max (twenty three)

[0192] -R k,down ≤Pk,t -P k,t-1 ≤R k,up (twenty four)

[0193]

[0194] Where, P k,t is the electric power of coal-fired power unit k at time t; P k,min and P k,max are the minimum and maximum power of coal-fired power unit k; R k,down and R k,up are the downward and upward climbing rates of coal-fired power unit k; U k is the maximum allowed number of starts and stops of coal-fired power unit k.

[0195] Operation constraints of wind turbine i:

[0196] 0≤P i,t ≤P i,max (26)

[0197] Where, P i,t is the electric power of wind turbine i at time t.

[0198] Active power balance constraints in power networks:

[0199]

[0200] Where, P j,t is the power demand of system load j at time t; ψ chp , ψ con , ψ wind and ψ d They represent the collection of extraction steam cogeneration units, coal-fired power units, wind turbines and system electrical loads respectively, and the subscript j represents the number of the system electrical load.

[0201] Line flow operation constraints of power networks:

[0202] L l,min ≤L l,t ≤L l,max (28)

[0203]

[0204] Where, subscript l represents the number of the power network line; G l- is the power allocation factor; L l,t is the transmission power of line l at time t; L l,min and L l,max are the minimum and maximum power transmission values ​​of line l respectively.

[0205] (2) The operation constraints of the thermal system include the node equations and pipeline equations of the thermal system.

[0206] Nodal equations of the thermal system:

[0207] Heat balance equation for heat source and heat load nodes:

[0208]

[0209]

[0210] Where m n is the mass flow rate of the heat source node and the heat load node n; c is the specific heat capacity; Q g,t and Q d,t are the heat power of heat source node n and the heat demand of heat load node n at time t respectively; and are the inlet temperature and outlet temperature of the heat source node and heat load node n at time t respectively; ψ HS represents the set of heat source nodes; ψ h Represents a collection of heat load nodes.

[0211] Heat balance equations for supply and return pipe network nodes:

[0212]

[0213]

[0214] Where, and are the mass flow rates of the supply and return pipes e connected to node n, respectively; and are the outlet temperatures of the supply and return pipes e connected to node n at time t, respectively; e represents the pipe number; and They represent the front-side and back-side pipeline sets connected to the heating network node n respectively; ψ HS represents the set of heat source nodes; ψ HES Represents a collection of heat exchange stations.

[0215] Temperature conservation assumption for supply and return pipe network nodes:

[0216]

[0217]

[0218] Where, and are the inlet temperatures of the supply and return pipes e connected to node n at time t; and are the temperatures at the heat source node and heat load node n at time t respectively; ψ HS represents the set of heat source nodes; ψ h represents the heat load node set; ψ HES Represents a collection of heat exchange stations.

[0219] Piping equations for thermal systems

[0220]

[0221]

[0222]

[0223]

[0224] Where, and are the pipe temperatures of water supply / return respectively; T b is the ambient temperature; e is the loss coefficient of pipeline temperature; R is the equivalent thermal resistance per unit length of pipeline; l is the total length of pipeline; R s and R a The thermal resistances of the symmetrical and asymmetrical heat loss branches are respectively; m e represents the mass flow rate of hot water; c is the specific heat capacity.

[0225] Taking into account the impact of wind power uncertainty, the uncertainty modeling of wind power is as follows:

[0226] (1-α3)P i,max ≤P′ i,max ≤(1+α3)P i,max (40)

[0227] α3 is the uncertainty range of the maximum wind power, P′ i,max is the actual value of the maximum electric power of wind power. The operation constraints of wind turbines are rewritten as:

[0228] 0≤P i,t ≤P′ i,max (41)

[0229] Find the largest uncertainty range when the cost is allowed to a certain extent, so as to improve the stability of the system.

[0230] In step 3, the model is transformed using information gap robust optimization, and the original optimization model is transformed into an upper-layer optimization scheduling model and a lower-layer optimization scheduling model.

[0231] (1) Upper-level optimization scheduling model

[0232] The optimization goal of the upper-level planning is to maximize the system uncertainty range while meeting the basic operating constraints of the power system and thermal system.

[0233] maxα=k1α1+k2α2+k3α3 (42)

[0234] k1+k2+k3=1 (43)

[0235] Where k1, k2, and k3 are weight factors, which guide the model to derive an uncertainty range that is consistent with the actual situation, so that the scheduling plan is more reasonable and reliable. α represents the fluctuation amplitude of the actual maximum power output, that is, the uncertainty radius.

[0236]

[0237] Where, the objective function J up (x) corresponds to formula (42), x up are the decision variables of the upper-level planning, which are the power P of the cogeneration unit at time t g,t , thermal power Q of cogeneration unit g,t , thermal power unit output P k,t , wind turbine output P i,t ,y up are the state variables of the upper planning, which are the inlet temperature of the heat supply and return pipe e and Outlet temperature of heat network supply and return water pipe e and Inlet and outlet temperatures of the heat source node and heat load node n and m up (x up ,y up ) and n up (x up ,y up ) are the equality constraints and inequality constraints of the upper-level planning, respectively, corresponding to the operation constraints of the extraction steam cogeneration unit shown in equations (20) to (22), the operation constraints of the coal-fired power unit shown in equations (23) to (25), the operation constraints of the wind turbine unit shown in equation (41), the active power balance constraint of the power network shown in equation (27), the power network line flow operation constraints shown in equations (28) to (29), the heat balance constraints of the heat source and heat load nodes shown in equations (30) to (31), the heat balance constraints of the supply and return water pipeline network nodes shown in equations (32) to (33), the temperature conservation assumption of the supply and return water pipeline network nodes shown in equations (34) to (35), the supply and return water pipeline operation constraints shown in equations (7) to (14), and the weight factor equation shown in equation (43).

[0238] (2) Lower-level optimization scheduling model

[0239] The optimization goal of the lower layer is that when the system fluctuates within the uncertainty range, the maximum total scheduling cost of the system must be less than its expected target, as shown in the following formula:

[0240] max F total (45)

[0241] F total ≤F ro (46)

[0242] Where, F total is the total scheduling cost. Based on the above optimization objectives, the lower-level planning of the constructed random electrothermal coupling information gap robust optimization scheduling model is shown in the following formula:

[0243]

[0244] Where, the objective function J low (x) corresponds to formula (45), x low are the decision variables for the lower-level planning, which are the actual values ​​of the symmetrical thermal resistance R′ s , the actual value of asymmetric thermal resistance R' a And the actual maximum output of wind power P′ i,max , n low (x low ) is the inequality constraint of the lower-level planning, corresponding to the system uncertainty model constraints shown in Equations (5) to (6) and (40), and the economic constraints shown in Equation (46).

[0245] In step 3, before solving the model, the original planning problem needs to be reasonably relaxed so that the lower-level planning of the scheduling model is equivalently replaced by its KKT conditions. Process the lower-level planning model:

[0246]

[0247] Where, P g,t , Q g,t 、P k,t and P i,t are all decision variables of upper-level planning, R′ s , R′ a and P′ i,max are all decision variables for the lower-level planning. The equivalent KKT conditions for the lower-level planning are as follows:

[0248] -δ i +μ1δ i -μ2+μ3-μ4+μ5-μ6+μ7=0 (49)

[0249]

[0250] μ2((1-α1)R s -R′ s )=0 (51)

[0251] μ3(-(1+α1)R s +R′ s )=0 (52)

[0252] μ4((1-α2)R a -R′ a )=0 (53)

[0253] μ5(-(1+α2)R a +R′ a )=0 (54)

[0254] μ6((1-α3)P i,max -P′ i,max )=0 (55)

[0255] μ7(-(1+α3)P i,max +P′ i,max )=0 (56)

[0256] μ1、μ2、μ3、μ4、μ5、μ6、μ7≥0 (57)

[0257]

[0258] (1-α1)R s -R′ s ≤0 (59)

[0259] -(1+α1)R s +R′ s ≤0 (60)

[0260] (1-α2)R a -R′ a ≤0 (61)

[0261] -(1+α2)R a +R′ a ≤0 (62)

[0262] (1-α3)P i,max -P′ i,max ≤0 (63)

[0263] -(1+α3)P i,max +P′ i,max ≤0 (64)

[0264] Where μ1, μ2, μ3, μ4, μ5, μ6, and μ7 are all Lagrange multipliers. Equations (48) to (54) are the complementary relaxation conditions of the KKT condition. Equation (55) is the feasibility criterion of the dual problem of the KKT condition. Equations (56) and (62) are the feasibility criteria of the original problem of the KKT condition. After the above relaxation strategy, it can be directly solved using the solver.

[0265] Example:

[0266] The actual random electrothermal coupling system model of a certain area is used. The model structure diagram is shown in Figure 1 , the system electric and thermal load demand and wind power forecast maximum power output are detailed in Figure 2 , the other parameters are as follows:

[0267] 1. Power system parameters:

[0268] In this example, the power system network parameters are all taken from the IEEE standard 39-node model parameters. The wind power cost penalty coefficient is 100 kW / h, the minimum power output of the thermal power unit is 50 MW, the maximum power output of the thermal power unit is 210 MW, the electric-thermal coupling system of the cogeneration unit is 0.6, the maximum fuel intake of the cogeneration unit is 365 kg, the minimum fuel intake of the cogeneration unit is 200 kg, the fuel consumption rate of the cogeneration unit for electrical output is 1.2 kg / MW, the fuel consumption rate of the cogeneration unit for thermal output is 0.8 kg / MW, and the maximum thermal output of the cogeneration unit is 240 MW.

[0269] 2. Thermal system parameters:

[0270] In this embodiment, the pipe network parameters of the thermal system are all taken from the actual parameters of a certain region.

[0271] Finally, a corresponding mathematical simulation model was established in Matlab, and the rationality and advantages of considering the uncertainty of resistance and wind power in the electric-thermal coupling scheduling model were verified through simulation.

[0272] To analyze the specific impact of thermal resistance uncertainty on the heat loss characteristics of the heating network, the following scenario is set:

[0273] Scenario 1: A random electrothermal coupling optimization scheduling model that takes into account the uncertainty of asymmetric thermal resistance, without considering the impact of symmetric thermal resistance and wind power uncertainty.

[0274] Scenario 2: A random electrothermal coupling optimization scheduling model that takes into account the uncertainty of symmetrical thermal resistance, without considering the impact of asymmetrical thermal resistance and wind power uncertainty.

[0275] The relationship between the system uncertainty range α, cost deviation coefficient β and scheduling cost of different schemes under the two scenarios are shown in Table 1.

[0276] Table 1 Uncertainty ranges under different scenarios

[0277]

[0278] Depend on Figure 3 and Figure 4 It can be seen that the heat loss of the return pipe in Scheme 1 is lower than that in Scheme 2 and Scheme 3 at all times, and the total heat loss is reduced by 0.203% and 0.402% respectively compared with the other schemes. However, the heat loss of the water supply pipe shows the opposite change, and the total heat loss of the other schemes is reduced by 0.126% and 0.249% respectively compared with Scheme 1. The main reason for the change in heat loss is that in Scenario 1, under the action of asymmetric heat loss characteristics, as the uncertainty range of asymmetric thermal resistance increases, it makes it difficult for the heat of the water supply pipe to transfer to the return pipe through the insulation layer inside the pipe. While reducing the heat loss of the water supply pipe, it causes the heat loss of the return pipe to increase, thereby affecting the dynamic change process of the pipe temperature. Figure 5 It can be seen that the temperature change trends of the different schemes are similar. However, due to the change in heat loss from the heating network, the average supply pipe temperatures of Schemes 2 and 3 increased by 0.0001°C and 0.0002°C, respectively, compared to Scheme 1, while the average return pipe temperatures decreased by 0.0003°C and 0.0006°C, respectively. In this case, heating to the same outlet temperature requires the cogeneration unit to provide greater thermal power, which reduces the unit's adjustable flexibility and affects the overall dispatch output arrangement.

[0279] Depend on Figure 6 and Figure 7 It can be seen that the heat loss of the heating network pipeline in Scheme 1 is lower than that of other schemes at all times. The total heat loss of return water increases by 1.039% and 2.081% year-on-year, and the total heat loss of water supply increases by 0.722% and 1.447% year-on-year. With the change of heat loss of the heating network, the average temperature of the water supply pipeline in Schemes 2 and 3 decreases by 0.0004℃ and 0.0008℃ respectively compared with Scheme 1, and the average temperature of the return water pipeline decreases by 0.05℃ and 0.1℃ respectively. Figure 8 The main reason for this change is that as the uncertainty range of the symmetrical thermal resistance increases, the heat exchange process between the heating network pipeline and the surrounding environment is intensified, resulting in increased heat dissipation and a decrease in the inlet temperature at the heat source node. Similarly, when heated to the same outlet temperature, the thermal power provided by the cogeneration unit will be greater. In summary, considering the uncertainty of thermal resistance in the electric-thermal coupling scheduling will cause significant changes in the heat loss characteristics of the heating network, which in turn will cause fluctuations in the unit's output thermal power, thus affecting the output arrangement of the entire scheduling.

[0280] To analyze the impact of system uncertainty on the results of electric-thermal coupling scheduling, the following scenario is set:

[0281] Scenario 1: Stochastic electrothermal coupling scheduling with consideration of non-opposite heat loss in the heating network under a double-pipe common shell direct buried structure.

[0282] Scenario 2: Stochastic electric-thermal coupling scheduling without considering the asymmetric heat loss of the heating network under the double-pipe direct buried structure. Figure 9 As shown in the figure, under the same cost deviation coefficient, the uncertainty range of Scenario 1 is greater than that of Scenario 2, indicating that the risk tolerance after accounting for the asymmetric heat loss characteristics of the heating network is better than that without timely accounting. Similarly, when the two scenarios have the same risk tolerance, the scheduling cost under Scenario 1 is lower than that of Scenario 2, indicating that the economic efficiency of Scenario 1 is better than that of Scenario 2. Therefore, accounting for the asymmetric heat loss characteristics of the heating network will affect the uncertainty optimization scheduling strategy. The heat loss characteristics of the double-pipe common shell direct buried laying structure can achieve greater resistance to system uncertainty risks at a lower cost, which has a positive impact on uncertainty scheduling decisions.

Claims

1. A dual stochastic electrothermal coupling optimization scheduling method taking into account thermal resistance uncertainty and wind power uncertainty, characterized by The following steps are involved: Step 1: Considering the impact of pipeline damage in actual electric-thermal coupling scheduling, which may cause changes in thermal resistance and thus affect the system, establish uncertainty constraints on the thermal resistance of the heating network; Step 2: Based on the uncertainty constraints of the thermal resistance of the heating network established in step 1, the uncertainty of wind power and the uncertainty of the thermal resistance of the heating network are considered in the structure of the electric-thermal coupling scheduling system, and a dual stochastic electric-thermal coupling optimization scheduling model taking into account the uncertainties of wind power and thermal resistance is established; Step 3: Based on the dual stochastic electrothermal coupling optimization scheduling model established in step 2, an information gap robust optimization solution strategy is proposed to obtain the optimal scheduling result; In the above step 1, in actual working conditions, the thermal resistance of the inner and outer layers of the water supply pipe and the return pipe will change due to damage factors, thereby affecting the system, as follows: (1) Relationship between the change of thermal resistance and the change of thermal resistance heat loss: In formula (1), △q is the heat loss fluctuation of the heating network caused by the change of thermal resistance, △R is the change of thermal resistance, c is the specific heat capacity of hot water, ρ is the density of hot water, A is the internal cross-sectional area of ​​the pipeline, m is the mass flow rate of hot water, q is the heat loss of the heating network, △x and △t are the space and time steps in the difference format respectively; (2) Relationship between heat loss changes in the heating network and pipeline temperature fluctuations: In formula (2), △T(x,t) is the temperature change caused by heat loss fluctuation at the distance x from the pipeline inlet at time t, and △q is the heat loss fluctuation of the heating network caused by the change of thermal resistance; The relationship between the supply and return water pipe temperature change and the heat loss of the heating network is quantified, and the relationship between the pipe temperature at any spatial position and the heat loss of the heating network is obtained, as shown in the following equations (3) and (4): In formula (3) and formula (4), c is the specific heat capacity of hot water; m v and m r are the mass and flow rate of hot water in the supply and return pipes respectively; T v (x,t) and T r (x,t) is the temperature at time t at a distance x from the inlet of the supply and return pipes; T v (x-△x,t) and T r (x-△x,t) is the temperature at time t at a distance x-△x from the inlet of the supply and return pipes; is the historical temperature sequence of the pipe in space after differentiation, and are the symmetrical heat losses of the supply and return pipes with symmetrical thermal resistance, q a is the asymmetric heat loss of the asymmetric thermal resistance; In step 1, uncertainty constraints on the thermal resistance of the thermal network are established, and the envelope constraint method is used to express the uncertainty problem of the thermal resistance: The uncertainty models of symmetric thermal resistance and asymmetric thermal resistance in uncertainty problems are shown in the following equations (5) and (6): (1-α1)R s ≤R′ s ≤(1+α1)R s (5); (1-α2)R a ≤R′ a ≤(1+α2)R a (6); In formula (5) and formula (6), α1 and α2 are the uncertainty ranges of symmetric thermal resistance and asymmetric thermal resistance respectively; R s and R a The thermal resistances of the symmetrical and asymmetrical heat loss branches are respectively; R′ a , R′ s are the actual values ​​of asymmetric thermal resistance and symmetric thermal resistance respectively; At the same time, the operation constraints of the original heat network pipeline need to be modified, and the deterministic constraints are replaced by uncertain constraints, as shown in the following equations (7) to (14): In formulas (7) to (14), C′1, C′2, C′3, C′4, and C′5 are weight coefficients; and They are the pipe temperatures during water supply and return respectively; and are the inlet temperatures of the water supply pipe e at time t-τ1 and t-τ2 respectively; and are the inlet temperatures of the return pipe e at time t-τ1 and t-τ2 respectively; T b is the ambient temperature; m e,K (τ2-τ) and m e,K+1 (τ-τ1) are the mass and flow rate of the mass at τ2-τ and τ-τ1 respectively; ξ e ' is the loss coefficient of pipeline temperature; l is the total length of the pipeline; m e represents the mass flow rate of hot water; R' is the equivalent thermal resistance per unit length of pipe; R' s , R' a are the thermal resistances of the symmetrical heat loss and asymmetrical heat loss branches respectively; In step 2, in the dual stochastic electrothermal coupling optimization scheduling model, the optimization goal is to minimize the total system scheduling cost while ensuring the maximum absorption of wind power, and wind power abandonment is added as a penalty term; the objective function is shown in the following formula: min F total =F chp +F con +F wind (15); F k,ST =u k,t (1-u k,t )f k,ST (18); In formulas (15) to (19), the objective function is defined as the total scheduling cost F total , F chp F is the dispatching cost of the extraction steam cogeneration unit; con is the dispatching cost of coal-fired power units; F wind Setting wind curtailment penalty costs to promote wind power consumption; t is a certain moment of model scheduling, T is the total scheduling time of the model, ψ chp , ψ con and ψ wind represent the collection of extraction steam cogeneration units, coal-fired thermal power units and wind turbine units, respectively; the subscripts g, k and i represent the unit numbers of extraction steam cogeneration units, coal-fired thermal power units and wind turbine units, respectively; f g,p1 and f g,p2 is the electric power cost coefficient of the cogeneration unit; f g,q1 and f g,q2 is the thermal power cost coefficient of the cogeneration unit; f g,pq and f g,pq0 is the electric heating cost coefficient; f k0 、f k1 and f k2 are the electric power cost coefficients of fuel-fired, coal-fired and thermal power units k, respectively; F k,ST is the start-up and shutdown cost of coal-fired power unit k, u k,t is the operating status of coal-fired power unit k at time t, f k,ST is the single startup cost of coal-fired power unit k, δ i is the penalty factor of wind turbine i, P i,max is the maximum electric power of wind turbine i; Operational constraints of the dual stochastic electric-thermal coupling optimization dispatch model, including power system operation constraints and thermal system operation constraints; (1) Power system operation constraints mainly include: operation constraints of extraction steam cogeneration units, operation constraints of coal-fired power units, operation constraints of wind turbine units, active power balance constraints of the power network, and line flow operation constraints of the power network; Operation constraints of extraction steam cogeneration unit g: 0≤P g,t ≥r g Q g,t (20) F g,min ≤ρ g,p P g,t +ρ g,q Q g,t ≤F g,max (21) 0≤Q g,t ≤Q g,max (22) In formulas (20) to (22), P g,t and Q g,t are the electric power and thermal power of the extraction steam cogeneration unit g at time t; r g is the electric-thermal coupling coefficient of the extraction steam cogeneration unit under back pressure conditions; ρ g,p and ρ g,q are the coal consumption rates of the extraction steam cogeneration unit for electric power and thermal power, respectively; Q g,max The upper limit of thermal power of the extraction steam cogeneration unit; F g,max and F g,min Maximum and minimum values ​​of coal intake of the unit; Operation constraints of coal-fired power unit k: P k,min ≤P k,t ≤P k,max (23) -R k,down ≤P k,t -P k,t-1 ≤R k,up (24) In formulas (23) to (25), P k,t is the electric power of coal-fired power unit k at time t; P k,t-1 is the electric power of coal-fired power unit k at time t-1; P k,min and P k,max are the minimum and maximum power of coal-fired power unit k; R k,down and R k,up are the downward and upward climbing rates of coal-fired power unit k; U k,t-1 is the state quantity of coal-fired power unit k starting and stopping at time t-1; U k is the maximum allowed number of starts and stops of coal-fired power unit k; Operation constraints of wind turbine i: 0≤P i,t ≤P i,max (26) In formula (26), P i,t is the electric power of wind turbine i at time t, P i,max is the maximum allowable power of wind turbine i at time t; Active power balance constraints in power networks: In formula (27), P j,t is the power demand of system load j at time t; ψ chp , ψ con , ψ wind and ψ d They represent the extraction steam cogeneration unit, coal-fired power unit, wind turbine unit and the collection of system electrical loads respectively; subscript j represents the number of system electrical loads; Line flow operation constraints of power networks: L l,min ≤L l,t ≤L l,max (28); In formula (28) to formula (29), the subscript l represents the number of the power network line; G l- is the power allocation factor; L l,t is the transmission power of line l at time t; L l,min and L l,max are the minimum and maximum power transmission values ​​of line l, respectively; (2) The operation constraints of the thermal system include the node equations and pipeline equations of the thermal system: Nodal equations of the thermal system: Heat balance equation for heat source and heat load nodes: In formula (30) to formula (31), m n is the mass flow rate of the heat source node and the heat load node n; c is the specific heat capacity; Q g,t and Q d,t are the heat power of heat source node n and the heat demand of heat load node n at time t respectively; and are the inlet temperature and outlet temperature of the heat source node and heat load node n at time t respectively; ψ HS represents the set of heat source nodes; ψ h Represents a collection of heat load nodes; Heat balance equations for supply and return pipe network nodes: In formula (32) to formula (33), and are the mass flow rates of the supply and return pipes e connected to node n, respectively; and are the outlet temperatures of the supply and return pipes e connected to node n at time t, respectively; e represents the pipe number; and They represent the front-side and back-side pipeline sets connected to the heating network node n respectively; ψ HS represents the set of heat source nodes; ψ HES represents a collection of heat exchange stations; Temperature conservation assumption for supply and return pipe network nodes: In formula (34) to formula (35), and are the inlet temperatures of the supply and return pipes e connected to node n at time t; and are the temperatures at the heat source node and heat load node n at time t respectively; ψ HS represents the set of heat source nodes; ψ h represents the heat load node set; ψ HES represents a collection of heat exchange stations; Piping equations for the thermal system: In formula (36) to formula (37), and are the pipe temperatures of water supply / return respectively; T b is the ambient temperature; e is the loss coefficient of pipeline temperature; R is the equivalent thermal resistance per unit length of pipeline; l is the total length of pipeline; R s and R a The thermal resistances of the symmetrical and asymmetrical heat loss branches are respectively; m e represents the mass flow rate of hot water; c is the specific heat capacity; Taking into account the impact of wind power uncertainty, the uncertainty modeling of wind power is as follows: (1-α3)P i,max ≤P′ i,max ≤(1+α3)P i,max (40); α3 is the uncertainty range of the maximum wind power, P′ i,max is the actual value of the maximum electric power of wind power; the operation constraints of wind turbines are rewritten as: 0≤P i,t ≤P′ i,max (41)。 2. The dual stochastic electrothermal coupling optimization scheduling method taking into account thermal resistance uncertainty and wind power uncertainty according to claim 1 is characterized by: In step 3, the model is transformed using information gap robust optimization, and the optimization scheduling model is transformed into an upper-layer optimization scheduling model and a lower-layer optimization scheduling model: (1) Upper-level optimization scheduling model: The optimization goal of the upper-level planning is to maximize the system uncertainty range while satisfying the basic power system and thermal system operation constraints; maxα=k1α1+k2α2+k3α3 (42); k1+k2+k3=1 (43); In formulas (42) and (43), k1, k2, and k3 are weight factors, which are used to guide the model to obtain an uncertainty range that is consistent with the actual situation, so that the scheduling plan is more reasonable and reliable. α represents the fluctuation range of the actual maximum power output, that is, the uncertainty radius. In formula (44), the objective function J up (x) corresponds to formula (42), x up are the decision variables of the upper-level planning, which are the power P of the cogeneration unit at time t g,t , thermal power Q of cogeneration unit g,t , thermal power unit output P k,t , wind turbine output P i,t ,y up are the state variables of the upper planning, which are the inlet temperature of the heat supply and return pipe e and Outlet temperature of heat network supply and return water pipe e and Inlet and outlet temperatures of the heat source node and heat load node n and m up (x up ,y up ) and n up (x up ,y up ) are the equality constraints and inequality constraints of the upper-level planning respectively; (2) Lower-level optimization scheduling model: The optimization goal of the lower layer is that when the system fluctuates within the uncertainty range, the maximum total scheduling cost of the system must be less than its expected target, as shown in the following formula: max F total (45); F total ≤F ro (46); Where, F total is the total dispatch cost; Based on the above optimization objectives, the lower-level planning of the constructed random electrothermal coupling information gap robust optimization scheduling model is shown in the following formula (47): In formula (47), the objective function J low (x) corresponds to formula (45), x low are the decision variables for the lower-level planning, which are the actual values ​​of the symmetrical thermal resistance R′ s , the actual value of asymmetric thermal resistance R' a And the actual maximum output of wind power P′ i,max , n low (x low ) is the inequality constraint of the lower-level planning.

3. The dual stochastic electrothermal coupling optimization scheduling method taking into account thermal resistance uncertainty and wind power uncertainty according to claim 2 is characterized by: In step 3, before solving the model, the original planning problem needs to be reasonably relaxed so that the lower-level planning of the scheduling model is equivalently replaced by its KKT condition. Processing of the lower-level planning model: In formula (48), P g,t , Q g,t 、P k,t and P i,t are all decision variables of upper-level planning, R′ s , R′ a and P′ i,max are all decision variables of the lower-level planning; the equivalent KKT conditions of the lower-level planning are shown in the following equations (49) to (64): -d i +μ1d i -μ2+μ3-μ4+μ5-μ6+μ7=0 (49) μ2((1-α1)R s -R′ s )=0 (51) μ3(-(1+α1)R s +R′ s )=0 (52) μ4((1-α2)R a -R′ a )=0 (53) μ5(-(1+α2)R a +R′ a )=0 (54) μ6((1-α3)P i,max -P′ i,max )=0 (55) μ7(-(1+α3)P i,max +P′ i,max )=0 (56) μ1, μ2, μ3, μ4, μ5, μ6, μ7≥0 (57) (1-α1)R s -R′ s ≤0 (59) -(1+α1)R s +R′ s ≤0 (60) (1-α2)R a -R′ a ≤0 (61) -(1+α2)R a +R′ a ≤0 (62) (1-α3)P i,max -P′ i,max ≤0 (63) -(1+α3)P i,max +P′ i,max ≤0 (64) In equations (49) to (64), μ1, μ2, μ3, μ4, μ5, μ6, and μ7 are all Lagrange multipliers. Equations (48) to (54) are the complementary relaxation conditions of the KKT condition. Equation (55) is the feasibility criterion of the dual problem of the KKT condition. Equations (56) and (62) are the feasibility criteria of the original problem of the KKT condition. After being processed by the above relaxation strategy, the solver can be used to solve it directly.

4. A dual stochastic electrothermal coupling optimization scheduling system taking into account the uncertainties of wind power and thermal resistance, wherein the scheduling system operates according to the scheduling method of claim 1, and is characterized by: A dual stochastic electrothermal coupling optimization scheduling model that takes into account the uncertainty of wind power and thermal resistance is adopted. The dual stochastic electrothermal coupling optimization scheduling model takes minimizing the total system scheduling cost under the premise of ensuring the maximum absorption of wind power as the optimization goal, and adds the abandoned wind power as a penalty item. The objective function of the optimization scheduling model is shown as follows: min F total =F chp +F con +F wind (15); F k,ST =u k,t (1-u k,t )f k,ST (18); In formulas (15) to (19), the objective function is defined as the total scheduling cost F total , F chp is the dispatching cost of the extraction steam cogeneration unit; F con is the dispatching cost of coal-fired power units; F wind Setting wind curtailment penalty costs to promote wind power consumption; t is a certain moment of model scheduling, T is the total scheduling time of the model, ψ chp , ψ con and ψ wind represent the collection of extraction steam cogeneration units, coal-fired thermal power units and wind turbine units, respectively; the subscripts g, k and i represent the unit numbers of extraction steam cogeneration units, coal-fired thermal power units and wind turbine units, respectively; f g,p1 and f g,p2 is the electric power cost coefficient of the cogeneration unit; f g,q1 and f g,q2 is the thermal power cost coefficient of the cogeneration unit; f g,pq and f g,pq0 is the electric heating cost coefficient; f k0 、f k1 and f k2 are the electric power cost coefficients of fuel-fired, coal-fired and thermal power units k, respectively; F k,ST is the start-up and shutdown cost of coal-fired power unit k, u k,t is the operating status of coal-fired power unit k at time t, f k,ST is the single startup cost of coal-fired power unit k, δ i is the penalty factor of wind turbine i, P i,max is the maximum electric power of wind turbine i.

5. The dual stochastic electrothermal coupling optimization scheduling system taking into account the uncertainty of wind power and thermal resistance according to claim 4 is characterized by: The model is transformed using information gap robust optimization, and the optimization scheduling model is transformed into an upper-level optimization scheduling model and a lower-level optimization scheduling model: (1) Upper-level optimization scheduling model: The optimization goal of the upper-level planning is to maximize the system uncertainty range while satisfying the basic power system and thermal system operation constraints; maxα=k1α1+k2α2+k3α3 (42); k1+k2+k3=1 (43); In formulas (42) and (43), k1, k2, and k3 are weight factors, which are used to guide the model to obtain an uncertainty range that is consistent with the actual situation, so that the scheduling plan is more reasonable and reliable. α represents the fluctuation range of the actual maximum power output, that is, the uncertainty radius. In formula (44), the objective function J up (x) corresponds to formula (42), x up are the decision variables of the upper-level planning, which are the power P of the cogeneration unit at time t g,t , thermal power Q of cogeneration unit g,t , thermal power unit output P k,t , wind turbine output P i,t ,y up are the state variables of the upper planning, which are the inlet temperature of the heat supply and return pipe e and Outlet temperature of heat network supply and return water pipe e and Inlet and outlet temperatures of the heat source node and heat load node n and m up (x up ,y up ) and n up (x up ,y up ) are the equality constraints and inequality constraints of the upper-level planning respectively; (2) Lower-level optimization scheduling model: The optimization goal of the lower layer is that when the system fluctuates within the uncertainty range, the maximum total scheduling cost of the system must be less than its expected target, as shown in the following formula: max F total (45); F total ≤F ro (46); Where, F total is the total dispatch cost; Based on the above optimization objectives, the lower-level planning of the constructed random electrothermal coupling information gap robust optimization scheduling model is shown in the following formula (47): In formula (47), the objective function J low (x) corresponds to formula (45), x low are the decision variables for the lower-level planning, which are the actual values ​​of the symmetrical thermal resistance R′ s , the actual value of asymmetric thermal resistance R' a And the actual maximum output of wind power P′ i,max , n low (x low ) is the inequality constraint of the lower-level planning.

Images