Robust random hybrid optimization scheduling method and device based on period division

Abstract

The invention discloses a robust random hybrid optimization scheduling method and device based on period division, and the method comprises the steps: building a day-ahead pre-scheduling model based on a random optimization strategy; processing through the day-ahead pre-scheduling model to obtain a pre-scheduling scheme; constructing a scheduling period division model based on a robust optimization strategy; according to security constraints, performing period division processing on the pre-scheduling scheme through the scheduling period division model to obtain a period division result; constructing a robust random mixed day-ahead scheduling model according to the period division result; and solving the robust random mixed day-ahead scheduling model through a column and constraint generation algorithm to obtain an optimal scheduling scheme. The method not only can cope with extremely severe conditions, but also has good economical efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of operation optimization of energy systems, and particularly to a robust stochastic hybrid optimization scheduling method and device based on period division. Background Art

[0002] With the intensification of the energy crisis and environmental deterioration, the replacement of traditional fossil energy by renewable energy resources (RES) has become a trend. However, the large-scale application of renewable energy brings greater uncertainties, which poses challenges to the economic and reliable operation of integrated energy systems (IES).

[0003] At present, to address this issue, scholars at home and abroad have conducted extensive research. The main technical routes include stochastic programming (SP) and robust optimization (RO). The optimization scheduling method based on SP depicts the fluctuations of uncertain factors by generating typical scenarios, and this method can effectively improve the economy of the scheduling scheme under the influence of uncertain factors. However, the scheme obtained by the scheduling method based on SP is too optimistic, making it difficult for this method to cope with extremely harsh situations. The optimization scheduling method based on RO improves the robustness of the scheme by finding the worst-case scenario. This method can effectively avoid the impact of extremely harsh scenarios on system operation, but its economy is poor and it is prone to cause large resource waste.

[0004] Therefore, how to invent an energy system optimization scheduling method that can both cope with extremely harsh situations and has good economy has become an urgent problem to be solved. Summary of the Invention

[0005] For this purpose, the present invention provides a robust stochastic hybrid optimization scheduling method and device based on period division. Through a pre-scheduling model and a period division model, the periods prone to violating safety constraints and the periods not prone to violating safety constraints are divided. According to the period division results, through a hybrid robust stochastic optimization model, the economy and robustness of the scheduling are balanced, and the best economic scheme is given on the basis of meeting the robustness requirements. The present invention can both cope with extremely harsh situations and has good economy.

[0006] To achieve the above object, the present invention provides the following technical solution: A robust stochastic hybrid optimization scheduling method based on period division, comprising:

[0007] Construct a day-ahead pre-scheduling model based on a stochastic optimization strategy; through the processing of the day-ahead pre-scheduling model, obtain a pre-scheduling scheme;

[0008] Based on a robust optimization strategy, a scheduling period division model is constructed; according to the security constraints, the pre-scheduling plan is processed for period division through the scheduling period division model to obtain a period division result;

[0009] According to the period division result, a robust stochastic hybrid day-ahead scheduling model is constructed;

[0010] The robust stochastic hybrid day-ahead scheduling model is solved by a column-and-constraint generation algorithm to obtain an optimal scheduling plan.

[0011] As an optimal solution of a robust stochastic hybrid optimization scheduling method based on period division, the objective function of the day-ahead pre-scheduling model includes: the day-ahead scheduling cost and the expected value of the intra-day scheduling cost; the day-ahead scheduling cost includes: the day-ahead market power purchase cost, the CHP start-up cost, and the energy storage device maintenance cost; the intra-day scheduling cost includes: the intra-day market power purchase cost, the gas purchase cost, and the unit equipment maintenance cost;

[0012] The expression of the objective function of the day-ahead pre-scheduling model is:

[0013]

[0014] where c DA is the day-ahead scheduling cost; is the intra-day scheduling cost; S DA is the day-ahead scheduling strategy expression; is the intra-day scheduling strategy expression; S m is the uncertain variable expression; is the day-ahead electricity price; is the day-ahead power purchase quantity; is the start-up cost of the combined heat and power equipment; is the start-up cost of the gas boiler; is the depreciation cost of the energy storage equipment; p mai,ES is the equipment maintenance cost of the energy storage equipment; is the charging power of the energy storage equipment; is the discharging power of the energy storage equipment; p mai,HS is the equipment maintenance cost of the heat storage equipment; is the charging power of the heat storage equipment; is the discharging power of the heat storage equipment; is the intra-day electricity price; is the intra-day power purchase quantity; p gas is the gas price; is the gas purchase quantity; p mai,CHP is the equipment maintenance cost of the combined heat and power equipment; is the electric power of the combined heat and power equipment; p mai,HPis the equipment maintenance cost of the ground source heat pump; is the equipment electric power of the ground source heat pump; p mai,GB is the equipment maintenance cost of the gas boiler; is the equipment thermal power of the gas boiler; t is the time serial number; m is the scenario serial number.

[0015] As an optimal solution of a robust stochastic hybrid optimal scheduling method based on period division, the constraint conditions of the day-ahead pre-scheduling model include: power system constraints, thermal system constraints, natural gas system constraints, energy coupling equipment constraints, PV output constraints, on-load tap-changing transformer constraints, and energy storage equipment constraints.

[0016] As an optimal solution of a robust stochastic hybrid optimal scheduling method based on period division, the expression of the scheduling period division model is:

[0017]

[0018] In the formula, τ is the moment; n slack is the slack variable; N is the matrix corresponding to the slack variable; g(·) ≤ 0 is the inequality constraint in the above constraints; h(·) = 0 is the equality constraint in the above constraints; s DA is the day-ahead scheduling strategy; s ID is the intra-day scheduling strategy; ζ RO is the severe scenario in the uncertainty set.

[0019] As an optimal solution of a robust stochastic hybrid optimal scheduling method based on period division, the expression of the robust stochastic hybrid day-ahead scheduling model is:

[0020]

[0021] In the formula, is the stochastic intra-day strategy; is the scenario in the stochastic scenario set; e SP is the stochastic scenario set; e RO is the robust uncertainty set; ζ RO is the severe scenario in the uncertainty set; s ID,RO is the robust intra-day strategy; c ID,RO is the robust intra-day cost; is; E(·) is the mathematical expectation expression; is the robust stochastic hybrid intra-day strategy; is the robust stochastic hybrid scenario.

[0022] As an optimal solution of a robust stochastic hybrid optimization scheduling method based on period division, in the process of solving the robust stochastic hybrid day-ahead scheduling model by the column and constraint generation algorithm, the column and constraint generation algorithm decomposes the min-max two-layer optimization problem into a master problem and a sub-problem for solution; the expression of the master problem is:

[0023]

[0024] In the formula, x is the first-stage optimization variable; η is the constructed auxiliary variable; k is the iteration number of the algorithm; is the scenario set searched by the algorithm; is the scenario in the scenario set; y l is the second-stage optimization variable; a T is the linearized first-stage cost coefficient vector; b T is the linearized second-stage cost coefficient vector; A, B, C, D, E, F, G, H are the linearized constraint coefficient matrices; T is the matrix transpose;

[0025] The expression of the sub-problem is:

[0026]

[0027] In the formula, x * is the first-stage optimization variable obtained by the master problem; π is the dual variable corresponding to the inequality constraint; λ is the dual variable corresponding to the equality constraint; v is the constructed binary auxiliary variable; M is a sufficiently large number.

[0028] The present invention also provides a robust stochastic hybrid optimization scheduling device based on period division. Based on the above-mentioned robust stochastic hybrid optimization scheduling method based on period division, it includes:

[0029] A day-ahead pre-scheduling model construction and processing module, which is used to construct a day-ahead pre-scheduling model based on a stochastic optimization strategy; through the processing of the day-ahead pre-scheduling model, a pre-scheduling plan is obtained;

[0030] A scheduling period division model construction and processing module, which is used to construct a scheduling period division model based on a robust optimization strategy; according to safety constraints, perform period division processing on the pre-scheduling plan through the scheduling period division model to obtain a period division result;

[0031] A robust stochastic hybrid day-ahead scheduling model construction module, which is used to construct a robust stochastic hybrid day-ahead scheduling model according to the period division result;

[0032] A robust stochastic hybrid day-ahead scheduling model solving module, which is used to solve the robust stochastic hybrid day-ahead scheduling model by the column and constraint generation algorithm to obtain an optimal scheduling plan.

[0033] As an optimal solution of a robust stochastic hybrid optimal scheduling device based on period division, in the day-ahead pre-scheduling model construction and processing module, the objective function of the day-ahead pre-scheduling model includes: the day-ahead scheduling cost and the expected value of the intra-day scheduling cost; the day-ahead scheduling cost includes: the day-ahead market power purchase cost, the CHP start-up cost, and the energy storage device maintenance cost; the intra-day scheduling cost includes: the intra-day market power purchase cost, the gas purchase cost, and the unit equipment maintenance cost;

[0034] The expression of the objective function of the day-ahead pre-scheduling model is:

[0035]

[0036] In the formula, c DA is the day-ahead scheduling cost; is the intra-day scheduling cost; S DA is the day-ahead scheduling strategy expression; is the intra-day scheduling strategy expression; S m is the uncertain variable expression; is the day-ahead electricity price; is the day-ahead power purchase quantity; is the start-up cost of the combined heat and power equipment; is the start-up cost of the gas boiler; is the depreciation cost of the electricity storage equipment; p mai,ES is the equipment maintenance cost of the electricity storage equipment; is the charging power of the electricity storage equipment; is the discharging power of the electricity storage equipment; p mai,HS is the equipment maintenance cost of the heat storage equipment; is the charging power of the heat storage equipment; is the discharging power of the heat storage equipment; is the intra-day electricity price; is the intra-day power purchase quantity; p gas is the gas price; is the gas purchase quantity; p mai,CHP is the equipment maintenance cost of the combined heat and power equipment; is the equipment electric power of the combined heat and power equipment; p mai,HP is the equipment maintenance cost of the ground source heat pump; is the equipment electric power of the ground source heat pump; p mai,GB is the equipment maintenance cost of the gas boiler; is the equipment thermal power of the gas boiler; t is the time serial number; m is the scenario serial number.

[0037] As an optimal solution of a robust stochastic hybrid optimal scheduling device based on period division, in the day-ahead pre-scheduling model construction and processing module, the constraint conditions of the day-ahead pre-scheduling model include: power system constraints, thermal system constraints, natural gas system constraints, energy coupling equipment constraints, PV output constraints, on-load tap-changer constraints, and energy storage equipment constraints.

[0038] As an optimal solution of a robust stochastic hybrid optimal scheduling device based on period division, in the scheduling period division model construction and processing module, the expression of the scheduling period division model is:

[0039]

[0040] In the formula, τ is the time; n slack is the slack variable; N is the matrix corresponding to the slack variable; g(·) ≤ 0 is the inequality constraint in the above constraints; h(·) = 0 is the equality constraint in the above constraints; s DA is the day-ahead scheduling strategy; s ID is the intra-day scheduling strategy; ζ RO is the severe scenario in the uncertainty set.

[0041] As an optimal solution of a robust stochastic hybrid optimal scheduling device based on period division, in the robust stochastic hybrid day-ahead scheduling model construction module, the expression of the robust stochastic hybrid day-ahead scheduling model is:

[0042]

[0043] In the formula, is the stochastic intra-day strategy; is the scenario in the stochastic scenario set; e SP is the stochastic scenario set; e RO is the robust uncertainty set; ζ RO is the severe scenario in the uncertainty set; s ID,RO is the robust intra-day strategy; c ID,RO is the robust intra-day cost; is; E(·) is the mathematical expectation expression; is the robust stochastic hybrid intra-day strategy; is the robust stochastic hybrid scenario.

[0044] As an optimal solution of a robust stochastic hybrid optimal scheduling device based on period division, in the robust stochastic hybrid day-ahead scheduling model solving module, in the process of solving the robust stochastic hybrid day-ahead scheduling model by the column and constraint generation algorithm, the column and constraint generation algorithm decomposes the min-max two-layer optimization problem into a master problem and a sub-problem for solution; the expression of the master problem is:

[0045]

[0046] In the formula, x is the optimization variable in the first stage; η is the constructed auxiliary variable; k is the iteration number of the algorithm; is the scenario set searched by the algorithm; is the scenario in the scenario set; y l is the optimization variable in the second stage; a T is the cost coefficient vector of the first stage after linearization; b T is the cost coefficient vector of the second stage after linearization; A, B, C, D, E, F, G, H are the constraint coefficient matrices after linearization; T is the matrix transpose;

[0047] The expression of the sub-problem is:

[0048]

[0049] In the formula, x * is the optimization variable of the first stage obtained by the master problem; π is the dual variable corresponding to the inequality constraint; λ is the dual variable corresponding to the equality constraint; v is the constructed binary auxiliary variable; M is a sufficiently large number.

[0050] The present invention has the following advantages: Based on a stochastic optimization strategy, the present invention constructs a day-ahead pre-scheduling model; through the processing of the day-ahead pre-scheduling model, a pre-scheduling plan is obtained; based on a robust optimization strategy, a scheduling period division model is constructed; according to the security constraints, the pre-scheduling plan is subjected to period division processing through the scheduling period division model to obtain a period division result; according to the period division result, a robust stochastic hybrid day-ahead scheduling model is constructed; the robust stochastic hybrid day-ahead scheduling model is solved by a column and constraint generation algorithm to obtain an optimal scheduling plan. The present invention divides the periods that are prone to violating security constraints and the periods that are not prone to violating security constraints through the pre-scheduling model and the period division model. The scheduling day is divided into a stochastic scheduling period and a robust scheduling period. According to the period division result, through a hybrid robust stochastic optimization model, the economy and robustness of the scheduling are balanced, and by minimizing the expected cost in the stochastic scheduling period and the maximum cost in the robust scheduling period, a day-ahead scheduling plan is obtained, thereby realizing the balance between robustness and economy, and giving the best economic plan on the basis of meeting the robustness requirements. The present invention can not only cope with extremely harsh situations but also has good economy. Description of the Drawings

[0051] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings described below are only exemplary. For those of ordinary skill in the art, without creative efforts, other implementation drawings can also be obtained based on the provided drawings.

[0052] The structures, ratios, sizes, etc. illustrated in this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention. Therefore, they do not have technical essential significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed in the present invention.

[0053] Figure 1 Schematic flow diagram of a robust stochastic hybrid optimal scheduling method based on period division provided in Embodiment 1 of the present invention;

[0054] Figure 2 Schematic diagram of the structure of an electric-gas-thermal coupled integrated energy system in a possible embodiment provided in Embodiment 1 of the present invention;

[0055] Figure 3 Schematic diagram of the uncertainty factor data under each typical day in a possible embodiment provided in Embodiment 1 of the present invention; among them, (a) is the photovoltaic output curve; (b) is the electrical load curve; (c) is the heat load curve;

[0056] Figure 4 Schematic diagram of the scheduling period division result in a possible embodiment provided in Embodiment 1 of the present invention;

[0057] Figure 5 Schematic diagram of the unit start-stop status under each typical day in a possible embodiment provided in Embodiment 1 of the present invention;

[0058] Figure 6 Schematic diagram of the node voltage data under each typical day in a possible embodiment provided in Embodiment 1 of the present invention;

[0059] Figure 7 Schematic diagram of the water supply water temperature under each typical day in a possible embodiment provided in Embodiment 1 of the present invention;

[0060] Figure 8 Schematic diagram of the architecture of a robust stochastic hybrid optimal scheduling device based on period division provided in Embodiment 2 of the present invention. Detailed implementation manners

[0061] The following specific embodiments illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0062] Embodiment 1

[0063] Refer to Figure 1 , Embodiment 1 of the present invention provides a robust stochastic hybrid optimal scheduling method based on period division, including the following steps:

[0064] S1. Based on a stochastic optimization strategy, construct a day-ahead pre-scheduling model; through the processing of the day-ahead pre-scheduling model, obtain a pre-scheduling plan;

[0065] S2. Based on a robust optimization strategy, construct a scheduling period division model; according to security constraints, perform period division processing on the pre-scheduling plan through the scheduling period division model to obtain a period division result;

[0066] S3. According to the period division result, construct a robust stochastic hybrid day-ahead scheduling model;

[0067] S4. Solve the robust stochastic hybrid day-ahead scheduling model through a column and constraint generation algorithm to obtain an optimal scheduling plan.

[0068] In this embodiment, taking an electric-gas-thermal coupled integrated energy system as an example, the present invention is described. The equipment in this integrated energy system includes: photovoltaic power generation equipment (PV), combined heat and power equipment (CHP), ground source heat pump (HP), gas boiler (GB), electric energy storage equipment (ES), heat storage equipment (HS), and on-load tap changer (OLTC).

[0069] In this embodiment, in step S1, based on a stochastic optimization strategy, construct a day-ahead pre-scheduling model; through the processing of the day-ahead pre-scheduling model, obtain a pre-scheduling plan;

[0070] Specifically, the objective function of the day-ahead pre-scheduling model includes: the day-ahead scheduling cost and the expected value of the intra-day scheduling cost; the day-ahead scheduling cost includes: the day-ahead market power purchase cost, the CHP startup cost, and the energy storage device maintenance cost; the intra-day scheduling cost includes: the intra-day market power purchase cost, the gas purchase cost, and the unit equipment maintenance cost;

[0071] The expression of the objective function of the day-ahead pre-scheduling model is:

[0072]

[0073] Among them, the day-ahead scheduling strategy expression S DA is:

[0074] s DA =[P grid,DA ,U CHP ,U GB ,P ES,c ,P ES,d ,H HS,c ,H HS,d ,K OLTC (4)

[0075] The intra-day scheduling strategy expression is:

[0076]

[0077] The uncertain variable expression S m is:

[0078]

[0079] In the formula, c DA is the day-ahead scheduling cost; is the intra-day scheduling cost; S DA is the day-ahead scheduling strategy expression; is the intra-day scheduling strategy expression; S m is the uncertain variable expression; is the day-ahead electricity price; is the day-ahead power purchase quantity; is the startup cost of the combined heat and power equipment; is the startup cost of the gas boiler; is the depreciation cost of the electricity storage equipment; p mai,ES is the equipment maintenance cost of the electricity storage equipment; is the charging power of the electricity storage equipment; is the discharging power of the electricity storage equipment; p mai,HS is the equipment maintenance cost of the heat storage equipment; is the charging power of the heat storage equipment; is the discharging power of the heat storage equipment; is the intra-day electricity price; is the intra-day electricity purchase volume; p gas is the gas price; is the gas purchase volume; p mai,CHP is the equipment maintenance cost of the combined heat and power (CHP) equipment; is the equipment electric power of the combined heat and power (CHP) equipment; p mai,HP is the equipment maintenance cost of the ground source heat pump; is the equipment electric power of the ground source heat pump; p mai,GB is the equipment maintenance cost of the gas boiler; is the equipment heat power of the gas boiler; t is the time sequence number; m is the scenario number U CHP is the start-stop state of the CHP; U GB is the start-stop state of the GB; K OLTC is the tap position of the on-load tap-changer transformer; is the electric power of the CHP; is the heat power of the CHP; is the gas power of the CHP; is the electric power of the HP; is the heat power of the HP; is the heat power of the GB; is the gas power of the GB; is the maximum output of the photovoltaic; is the magnitude of the electrical load; is the magnitude of the heat load.

[0080] Among them, the start-up cost in Equation (2) can be calculated through the following constraints:

[0081]

[0082] In the formula, p j,sta is the single start-up cost; is the start-stop state of the equipment;

[0083] In this embodiment, the constraint conditions of the day-ahead pre-scheduling model include: power system constraint, thermal system constraint, natural gas system constraint, energy coupling equipment constraint, PV output constraint, on-load tap-changer transformer constraint and energy storage equipment constraint.

[0084] Among them, the power system constraint:

[0085] In this embodiment, the LinDistFlow model is used to model the power system, and its constraints can be expressed as:

[0086]

[0087] In the formula, is the active power injected into the node; Inject reactive power into the node; For the active power of the line; For the reactive power of the line; R n,n' For the line resistance; X n,n' For the line reactance; For the square of the node voltage; And V For the upper and lower limits of the node voltage; For the upper limit of the line capacity.

[0088] Among them, Equation (9) is the line voltage drop constraint; Equation (10) is the expression of the active power injected into the node; Equation (11) is the expression of the reactive power injected into the node; Equation (12) is the active power balance constraint of the node; Equation (13) is the active power balance constraint of the head node; Equation (14) is the reactive power balance constraint; Equation (15) is the reactive power balance constraint of the head node; Equation (16) is the transformer capacity constraint; Equation (17) is the upper and lower limit constraint of the node voltage; Equation (18) is the line capacity constraint.

[0089] Thermal system constraint:

[0090] In this embodiment, the thermal system operates in a mass regulation mode, and its constraint can be expressed as:

[0091]

[0092] In the formula, For the water temperature at the water supply node; For the water temperature at the return water node; For the water temperature at the outflow node; For the water temperature at the inflow node; For the water temperature of the pipeline; For the ambient temperature; q n',n For the mass flow rate of the line; q n For the mass flow rate from the water supply end to the return water end of the node; c p For the specific heat capacity of water; λ 0 For the heat conduction coefficient of the pipeline; L n,n' For the pipeline length.

[0093] Among them, Equation (19) is the pipeline temperature loss constraint; Equation (20) is the node thermal power balance constraint; Equation (21) is the water temperature mixing constraint; Equation (22) is the upper and lower limit constraint of the water supply node temperature; Equation (23) is the upper and lower limit constraint of the return water node temperature.

[0094] Natural gas system constraint:

[0095] In this embodiment, the natural gas system constraint can be expressed as:

[0096]

[0097] In the formula, is the natural gas injection power at the node; is the natural gas power of the line; is the upper limit of the natural gas power of the line.

[0098] Among them, formula (24) is the expression of the natural gas injection power at the node; formula (25) is the natural gas power balance constraint; formula (26) is the natural gas power balance constraint at the head node; formula (27) is the natural gas pipeline power constraint.

[0099] Energy coupling equipment constraint:

[0100] In this embodiment, the energy coupling link of the IES includes CHP, HP, EB and GB, and its operation constraints can be expressed as:

[0101]

[0102] Among them, formula (28) is the output constraint of CHP and GB; formula (29) is the output constraint of HP; formula (30) is the ramping constraint of CHP, HP and GB; formula (31) is the gas-electricity conversion constraint of CHP; formula (32) is the gas-heat conversion constraint; formula (33) is the electro-thermal conversion constraint of HP; formula (34) is the gas-heat conversion constraint of GB.

[0103] PV output constraint:

[0104] In this embodiment, it is assumed that the PV consumption power does not exceed its output in the MPPT mode, and its expression is:

[0105]

[0106] In the formula, is the equipment electric power of the photovoltaic power generation equipment; is the maximum output size of the photovoltaic power generation equipment.

[0107] On-load tap-changer constraint:

[0108] In this embodiment, the voltage at the connection point between the regulating system and the superior power grid is adjusted by the on-load tap-changer, and its expression is:

[0109]

[0110] In the formula, V 0 is the standard voltage; is the regulation ratio; Δk is the regulation step.

[0111] Energy storage equipment constraint:

[0112] In this embodiment, the operation constraints of ES and HS are relatively similar, so a general energy storage model is used for modeling. This model can be expressed as:

[0113]

[0114] Among them, equations (39)-(40) are the upper and lower limit constraints of charge and discharge power; equation (41) is the charge and discharge constraint; equation (42) is the state of charge (SOC) balance constraint; equation (43) is the upper and lower limit constraints of SOC; equation (44) is the initial and final SOC constraint.

[0115] In this embodiment, for simplicity of expression, the optimization problem is denoted in a compact format. The day-ahead pre-scheduling model can be expressed as:

[0116]

[0117] where g (·) ≤ 0 is the inequality constraint in the above constraints; h(·) = 0 is the equality constraint in the above constraints.

[0118] In this embodiment, in step S2, based on the robust optimization strategy, a scheduling period division model is constructed; according to the security constraints, the pre-scheduling plan is processed for period division through the scheduling period division model to obtain a period division result;

[0119] Specifically, based on the robust optimization strategy, a scheduling period division model is constructed; whether the pre-scheduling plan may violate the security constraints is verified through the scheduling period division model. By constructing a slack variable for the security constraint, the model searches for scenarios that may violate the security constraints in the robust uncertainty set. If it is necessary to verify the τ moment, the scheduling period division model can be expressed as:

[0120]

[0121] where τ is the moment; n slack is the slack variable; N is the matrix corresponding to the slack variable, and its construction rule is: if the g-th constraint in the inequality constraint g(·) ≤ 0 contains the variable in the time interval t, then N(g,t) = 1, otherwise N(g,t) = 0; s DA ; s DA is the day-ahead scheduling strategy; s ID is the intra-day scheduling strategy; ζ RO is the severe scenario in the uncertainty set.

[0122] After verifying each moment, the random scheduling period and the robust scheduling period can be divided according to the verification results. If it means that there will be no problem of violating the security constraints in the τ period, so it is divided into the random scheduling period; if It indicates that there may be problems violating the security constraints during the τ period, so it is classified into the robust scheduling period. Let the time set classified into the stochastic scheduling period be The time set classified into the robust scheduling period is The scheduling costs, scheduling strategies, and scheduling scenarios of the stochastic scheduling period and the robust scheduling period can be obtained as follows:

[0123]

[0124] Among them, Equation (47) is the intraday stochastic scheduling cost; Equation (48) is the intraday robust scheduling cost; Equation (49) is the intraday stochastic scheduling strategy; Equation (50) is the intraday robust scheduling strategy; Equation (51) is the stochastic scheduling scenario expression; Equation (52) is the robust scheduling scenario expression.

[0125] For the convenience of expression, denote the intraday scheduling strategy composed of the stochastic scheduling strategy and the robust scheduling strategy as the system integrated scheduling strategy, and the scenario composed of the stochastic scheduling scenario and the robust scheduling scenario as the system integrated scheduling scenario, that is:

[0126]

[0127] In the formula, is the system integrated scheduling strategy; is the system integrated scheduling scenario.

[0128] In this embodiment, in step S3, according to the period division result, a robust-stochastic hybrid day-ahead scheduling model is constructed;

[0129] Specifically, according to the period division result, a robust-stochastic hybrid day-ahead scheduling model is constructed. During the robust scheduling period, there may be situations where the stochastic optimization scheduling scheme violates the security constraints. Therefore, the scheduling cost under the worst scenario in this period is used as the objective function. During the stochastic scheduling period, the stochastic optimization scheduling scheme will not violate the security constraints. Therefore, the expected value of the scheduling costs of each scenario in this period is used as the objective function. The expression of the robust-stochastic hybrid day-ahead scheduling model is:

[0130]

[0131] In the formula, is the stochastic intraday strategy; is the scenario in the stochastic scenario set; e SP is the stochastic scenario set; e RO is the robust uncertainty set; ζ RO is the worst scenario in the uncertainty set; s ID,RO is the robust intraday strategy; c ID,RO is the robust intraday cost; is; E(·) is the mathematical expectation expression; It is a robust stochastic hybrid intraday strategy; It is a robust stochastic hybrid scenario.

[0132] In this embodiment, in step S4, the robust stochastic hybrid day-ahead scheduling model is solved by a column and constraint generation algorithm to obtain an optimal scheduling plan.

[0133] Specifically, the robust stochastic hybrid day-ahead scheduling model is solved by a column and constraint generation algorithm to obtain a day-ahead scheduling plan This plan is actually executed and passed to intraday for adjustment.

[0134] Among them, since there are non-linear constraints in the above optimization problem, in order to speed up the model solving speed, these non-linear constraints are linearized.

[0135] Specifically, the linearization of power capacity constraints:

[0136] The transformer capacity constraint in Equation (16) and the line capacity constraint in Equation (18) are non-linear constraints. Here, Equation (18) is taken as an example. The range of this constraint is circular and can be linearized by its inscribed polygon. Its expression is:

[0137]

[0138] In the formula, K is the segmentation parameter of the inscribed polygon. It can be seen that this constraint is transformed into 2K linear constraints through the inscribed polygon. In this embodiment, K = 4 is taken.

[0139] The linearization of energy storage device constraints:

[0140] In Equation (2), the calculation of the ES depreciation cost is a non-linear expression related to the depth of discharge:

[0141]

[0142] In the formula, is the depth of discharge; a, b, c are parameters obtained through experiments. This non-linear expression can be linearized by a piecewise linearization process.

[0143] The linearization of OLTC constraints:

[0144] The OLTC constraints in Equations (36)-(38) are non-linear constraints and can be linearized by the following formula:

[0145]

[0146] In the formula, N OLTC is the number of OLTC segments, satisfying is a binary auxiliary variable representing the state of the OLTC.

[0147] In this embodiment, since the robust random hybrid scheduling model is a min-max-min problem and cannot be directly solved. Therefore, the present invention solves this model through the Column-and-Constraint Generation (C&CG) algorithm.

[0148] The C&CG algorithm solves the min-max two-layer optimization problem by decomposing it into a master problem and a sub-problem. In the C&CG algorithm, the master problem solves the optimal strategy under a given scenario and passes it to the sub-problem for optimization; the sub-problem obtains the worst-case scenario under the given strategy and then sends it back to the master problem for optimization. Through iterative convergence, the C&CG algorithm can obtain the optimal solution under the worst-case scenario. Let the linearized model be:

[0149]

[0150] In the formula, x is the first-stage optimization variable; y is the second-stage optimization variable a T is the linearized first-stage cost coefficient vector; b T is the linearized second-stage cost coefficient vector; A, B, C, D, E, F, G, H are the linearized constraint coefficient matrices; T is the matrix transpose; u is the uncertain factor variable.

[0151] Then the master problem of C&CG can be expressed as:

[0152]

[0153] In the formula, η is the constructed auxiliary variable; k is the iteration number of the algorithm; is the scenario set searched by the algorithm; is the scenario in the scenario set; y l is the second-stage optimization variable;

[0154] The sub-problem transformed through the KKT conditions can be expressed as:

[0155]

[0156] In the formula, x * is the first-stage optimization variable obtained by the master problem; π is the dual variable corresponding to the inequality constraint; λ is the dual variable corresponding to the equality constraint; v is the constructed binary auxiliary variable; M is a sufficiently large number.

[0157] In this embodiment, the detailed steps of the C&CG algorithm are as follows:

[0158] T1. Initialization, let k = 1, UB = +∞, LB = -∞,

[0159] T2. Solve the main problem for the given set of scenarios, and update the solution result.

[0160] T3. Solve the sub-problem for the given policy, update the solution result k = k + 1, UB =

[0161]

[0162] T4. Determine whether |UB - LB| < ε is satisfied. If this condition is met, end the iteration and output the result; otherwise, return to step T2.

[0163] In a possible embodiment, a specific energy system scheduling example is provided as follows:

[0164] Taking a certain electric - gas - heat coupled IES in northern China as an example, a case study is carried out. The topology of this IES is as Figure 2 shown, where the power grid topology refers to the IEEE - 33 standard case, the heat network topology is a 14 - node heating network, and the gas network topology is a 3 - node gas supply network. The capacity configurations of various devices in the IES are as follows: CHP unit 1 is 2200 kW, CHP unit 2 is 2200 kW, ground - source heat pump is 600 kW, gas boiler is 1000 kW, electricity storage device is 3500 kWh, heat storage device is 4000 kWh, and photovoltaic device is 4500 kW.

[0165] The day - ahead electricity price in the park adopts time - of - use electricity price. The electricity price during peak hours is, the electricity price during normal hours is, the electricity price during valley hours is, and the intra - day electricity price is 1.2 times the day - ahead electricity price. The park adopts a fixed gas price of 0.28 yuan / kW. The day - ahead data of PV output, electricity load, and heat load adopt three groups of typical day data in summer, winter, and transition seasons, as Figure 3 shown, and 100 groups of data are generated by the Monte Carlo method on this basis for verification of intra - day data.

[0166] After operation, it can be obtained that the operating cost of a typical day in summer is 34,580 yuan, the operating cost of a typical day in winter is 64,010 yuan, and the operating cost of a typical day in the transition season is 64,050 yuan. Under the scheduling method proposed in the present invention, no violation of safety constraints occurs in each typical day. The scheduling period division result is as Figure 4 shown, the unit start - stop status is as Figure 5 shown, the node voltage data is as Figure 6 shown, and the water supply water temperature data is as Figure 7 shown.

[0167] During a typical summer day, the period from 9 to 11 is designated as the robust scheduling period; during a typical winter day, 8 o'clock is designated as the robust scheduling period; during a typical transitional season day, 5 o'clock is designated as the robust scheduling period. During the robust scheduling period, the IES usually faces problems such as relatively low node voltages, relatively high water supply temperatures, and units being at the critical start-stop state. At this time, the system is prone to violating safety constraints. Therefore, these periods are classified as robust periods to avoid violating safety constraints.

[0168] To verify the effectiveness of the present invention, three cases are set for comparative analysis:

[0169] Scenario 1: The hybrid robust stochastic optimization scheduling method proposed by the present invention is adopted. The most adverse scenario is considered during the robust scheduling period, and the typical scenario is considered during the stochastic scheduling period, and then the day-ahead scheduling plan is obtained.

[0170] Scenario 2: The robust optimization scheduling method is adopted to obtain the operation plan under the most adverse scenario in all periods as the day-ahead scheduling plan.

[0171] Scenario 3: The stochastic programming scheduling method is adopted to obtain the operation plan under the typical scenario in all periods as the day-ahead scheduling plan.

[0172] The comparison results of the scenarios are shown in Table 1.

[0173]

[0174] Table 1 Comparison of the operation costs and the probabilities of violating safety constraints of the scenarios

[0175] As can be seen from Table 1, Scenario 1 has a lower operation cost than Scenario 2, and there is no violation of safety constraints. This shows that the hybrid robust stochastic optimization scheduling method proposed by the present invention can improve the economy of scheduling as much as possible while ensuring the robustness of scheduling.

[0176] The operation costs of Scenario 2 are 0.80%, 0.95%, and 0.94% higher than those of Scenario 1 in summer, winter, and the transitional season respectively, and there are no cases of violating safety constraints in both scenarios. This scenario adopts the robust optimization method in all scheduling periods, which ensures the robustness of this scenario. Therefore, this scenario does not violate safety constraints. However, the overly conservative scheduling method also makes the scheduling cost of this scenario too high, which brings additional costs to the scheduling.

[0177] The operating cost of Scheme 3 is 0.19%, 0.05% and 0.03% lower than that of Scheme 1 in summer, winter and transitional seasons respectively. However, this scheme has violations of safety constraints of 3.96%, 1.50% and 1.33% in summer, winter and transitional seasons respectively. This scheme adopts the stochastic programming method during the entire scheduling period, which makes the economy of this scheme better. However, this scheme does not consider extreme situations, so the robustness of this scheme is poor and the possibility of violating safety constraints occurs.

[0178] In summary, the present invention constructs a day-ahead pre-scheduling model based on a stochastic optimization strategy; obtains a pre-scheduling plan through processing by the day-ahead pre-scheduling model; constructs a scheduling period division model based on a robust optimization strategy; performs period division processing on the pre-scheduling plan through the scheduling period division model according to safety constraints to obtain a period division result; constructs a robust stochastic hybrid day-ahead scheduling model according to the period division result; solves the robust stochastic hybrid day-ahead scheduling model through a column-and-constraint generation algorithm to obtain an optimal scheduling plan. The present invention divides the periods prone to violating safety constraints and the periods not prone to violating safety constraints through the pre-scheduling model and the period division model. The scheduling day is divided into a stochastic scheduling period and a robust scheduling period. According to the period division result, through a hybrid robust stochastic optimization model, the economy and robustness of the scheduling are balanced. By minimizing the expected cost in the stochastic scheduling period and the maximum cost in the robust scheduling period, a day-ahead scheduling plan is obtained, thereby achieving the balance between robustness and economy, and giving the best economic plan on the basis of meeting the robustness requirements. The present invention can not only cope with extreme situations but also has good economy.

[0179] It should be noted that the method of the embodiments of the present disclosure can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In this case of a distributed scenario, one of the multiple devices can only execute one or more steps of the method of the embodiments of the present disclosure, and these multiple devices will interact with each other to complete the described method.

[0180] It should be noted that some embodiments of the present disclosure have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the above embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0181] Embodiment 2

[0182] See Figure 8 In Embodiment 2 of the present invention, a robust stochastic hybrid optimal scheduling device based on period division is further provided, including:

[0183] A day-ahead pre-scheduling model construction and processing module 001, configured to construct a day-ahead pre-scheduling model based on a stochastic optimization strategy; and obtain a pre-scheduling plan through processing by the day-ahead pre-scheduling model;

[0184] A scheduling period division model construction and processing module 002, configured to construct a scheduling period division model based on a robust optimization strategy; and perform period division processing on the pre-scheduling plan through the scheduling period division model according to security constraints to obtain a period division result;

[0185] A robust stochastic hybrid day-ahead scheduling model construction module 003, configured to construct a robust stochastic hybrid day-ahead scheduling model according to the period division result;

[0186] A robust stochastic hybrid day-ahead scheduling model solving module 004, configured to solve the robust stochastic hybrid day-ahead scheduling model through a column and constraint generation algorithm to obtain an optimal scheduling plan.

[0187] In this embodiment, in the day-ahead pre-scheduling model construction and processing module 001, the objective function of the day-ahead pre-scheduling model includes: the day-ahead scheduling cost and the expected value of the intra-day scheduling cost; the day-ahead scheduling cost includes: the day-ahead market power purchase cost, the CHP startup cost, and the energy storage device maintenance cost; the intra-day scheduling cost includes: the intra-day market power purchase cost, the gas purchase cost, and the unit equipment maintenance cost;

[0188] The expression of the objective function of the day-ahead pre-scheduling model is:

[0189]

[0190] In the formula, c DA is the day-ahead scheduling cost; is the intra-day scheduling cost; S DA is the expression of the day-ahead scheduling strategy; is the expression of the intra-day scheduling strategy; S m is the expression of the uncertain variable; is the day-ahead electricity price; is the day-ahead power purchase quantity; is the startup cost of the combined heat and power equipment; is the startup cost of the gas boiler; is the depreciation cost of the energy storage equipment; p mai,ES is the equipment maintenance cost of the energy storage equipment; is the charging power of the energy storage equipment; is the discharging power of the energy storage equipment; p mai,HSis the equipment maintenance cost of the heat storage equipment; is the charging power of the heat storage equipment; is the discharging power of the heat storage equipment; is the intra-day electricity price; is the intra-day electricity purchase quantity; p gas is the gas price; is the gas purchase quantity; p mai,CHP is the equipment maintenance cost of the combined heat and power equipment; is the electrical power of the combined heat and power equipment; p mai,HP is the equipment maintenance cost of the ground source heat pump; is the electrical power of the ground source heat pump equipment; p mai,GB is the equipment maintenance cost of the gas boiler; is the heat power of the gas boiler equipment; t is the time serial number; m is the scenario serial number.

[0191] In this embodiment, in the day-ahead pre-scheduling model construction and processing module 001, the constraint conditions of the day-ahead pre-scheduling model include: power system constraint, thermal system constraint, natural gas system constraint, energy coupling equipment constraint, PV output constraint, on-load tap-changer constraint, and energy storage equipment constraint.

[0192] In this embodiment, in the scheduling period division model construction and processing module 002, the expression of the scheduling period division model is:

[0193]

[0194] In the formula, τ is the moment; n slack is the slack variable; N is the matrix corresponding to the slack variable; g(·) ≤ 0 is the inequality constraint in the above constraints; h(·) = 0 is the equality constraint in the above constraints; s DA is the day-ahead scheduling strategy; s ID is the intra-day scheduling strategy; ζ RO is the severe scenario in the uncertainty set.

[0195] In this embodiment, in the robust stochastic hybrid day-ahead scheduling model construction module 003, the expression of the robust stochastic hybrid day-ahead scheduling model is:

[0196]

[0197] In the formula, is the stochastic intra-day strategy; is the scenario in the stochastic scenario set; e SP is the stochastic scenario set; e RO is the robust uncertainty set; ζ RO is the severe scenario in the uncertainty set; s ID,RO is the robust intra-day strategy; cID,RO is the robust intraday cost; is; E(·) is the mathematical expectation expression; is the robust stochastic hybrid intraday strategy; is the robust stochastic hybrid scenario.

[0198] In this embodiment, in the robust stochastic hybrid day-ahead scheduling model solving module 004, during the process of solving the robust stochastic hybrid day-ahead scheduling model by the column and constraint generation algorithm, the column and constraint generation algorithm disassembles the min-max two-layer optimization problem into a master problem and a sub-problem for solution; the expression of the master problem is:

[0199]

[0200] In the formula, x is the first-stage optimization variable; η is the constructed auxiliary variable; k is the iteration number of the algorithm; is the scenario set searched by the algorithm; is the scenario in the scenario set; y l is the second-stage optimization variable; a T is the linearized first-stage cost coefficient vector; b T is the linearized second-stage cost coefficient vector; A, B, C, D, E, F, G, H are the linearized constraint coefficient matrices; T is the matrix transpose;

[0201] The expression of the sub-problem is:

[0202]

[0203] In the formula, x * is the first-stage optimization variable obtained from the master problem; π is the dual variable corresponding to the inequality constraint; λ is the dual variable corresponding to the equality constraint; v is the constructed binary auxiliary variable; M is a sufficiently large number.

[0204] It should be noted that for the information interaction, execution process, etc. between the above system modules, since they are based on the same concept as the method embodiment in Embodiment 1 of the present application, the technical effects brought by them are the same as those of the method embodiment of the present application. For the specific content, reference can be made to the description in the method embodiment shown in the foregoing of the present application, and details are not described herein again.

[0205] Embodiment 3

[0206] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which a program code of a robust stochastic hybrid optimization scheduling method based on period division is stored, and the program code includes instructions for executing a robust stochastic hybrid optimization scheduling method according to Embodiment 1 or any possible implementation manner thereof.

[0207] A computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that incorporates one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state disk (SSD)), etc.

[0208] Embodiment 4

[0209] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0210] The processor and the memory complete communication with each other through a bus; the memory stores program instructions executable by the processor, and the processor can execute a robust random hybrid optimization scheduling method based on period division according to Embodiment 1 or any possible implementation manner thereof by invoking the program instructions.

[0211] Specifically, the processor can be implemented by hardware or by software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor, which is implemented by reading software code stored in the memory. The memory can be integrated in the processor or can exist independently outside the processor.

[0212] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable systems. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (e.g., infrared, wireless, microwave, etc.).

[0213] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing system. They can be concentrated on a single computing system or distributed on a network composed of multiple computing systems. Optionally, they can be implemented by program codes executable by the computing system. Thus, they can be stored in a storage system and executed by the computing system. And in some cases, the steps shown or described can be executed in a sequence different from that here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0214] Although the present invention has been described in detail with general descriptions and specific embodiments above, based on the present invention, some modifications or improvements can be made, which are obvious to those skilled in the art. Therefore, these modifications or improvements made without departing from the spirit of the present invention all fall within the scope of the present invention claimed.

Claims

1. A robust random hybrid optimization scheduling method based on period division, characterized in that: include: Based on the stochastic optimization strategy, a day-ahead scheduling model is constructed; Obtaining a pre-scheduling plan by processing the day-ahead pre-scheduling model; Based on the robust optimization strategy, a scheduling period division model is constructed; according to the safety constraint, the pre-scheduling plan is period-divided by the scheduling period division model to obtain a period division result; According to the period division results, a robust random hybrid day-ahead scheduling model is constructed; The robust random hybrid day-ahead scheduling model is solved by a column and constraint generation algorithm to obtain the optimal scheduling solution.

2. The robust random hybrid optimization scheduling method based on period division according to claim 1 is characterized in that: The objective function of the day-ahead pre-dispatch model includes: the day-ahead dispatch cost and the expected value of the intraday dispatch cost; the day-ahead dispatch cost includes: the day-ahead market electricity purchase cost, CHP startup cost and energy storage equipment maintenance cost; the intraday dispatch cost includes: the intraday market electricity purchase cost, gas purchase cost and unit equipment maintenance cost; The objective function expression of the day-ahead scheduling model is: In the formula, c DA is the day-ahead scheduling cost; is the intraday scheduling cost; S DA is the day-ahead scheduling strategy expression; is the intraday scheduling strategy expression; S m is an expression of uncertain variables; is the day-ahead electricity price; The amount of electricity purchased on the day before; The start-up costs for the CHP plant; The start-up cost of the gas boiler; is the depreciation cost of the power storage equipment; p mai,ES Equipment maintenance costs for power storage equipment; The charging power of the power storage device; is the energy release power of the power storage device; p mai,HS Equipment maintenance costs for thermal storage equipment; The charging power of the thermal storage device; is the energy release power of the heat storage device; is the intraday electricity price; is the amount of electricity purchased during the day; p gas For gas price; is the gas purchase volume; p mai,CHP Equipment maintenance costs for cogeneration equipment; is the equipment power of the cogeneration equipment; p mai,HP Equipment maintenance costs for ground source heat pumps; is the equipment power of the ground source heat pump; p mai,GB Equipment maintenance costs for gas boilers; is the equipment thermal power of the gas boiler; t is the time sequence number; m is the scene sequence number.

3. The robust random hybrid optimization scheduling method based on period division according to claim 2 is characterized in that: The constraints of the day-ahead dispatch model include: power system constraints, thermal system constraints, natural gas system constraints, energy coupling equipment constraints, PV output constraints, on-load tap-changing transformer constraints and energy storage equipment constraints.

4. The robust random hybrid optimization scheduling method based on period division according to claim 3 is characterized in that: The expression of the scheduling period division model is: Where τ is the time; n slack is the slack variable; N is the matrix corresponding to the slack variable; g(·)≤0 is the inequality constraint in the above constraints; h(·)=0 is the equality constraint in the above constraints; s DA is the day-ahead scheduling strategy; ID It is an intraday scheduling strategy; For the bad scenarios with uncertain concentration.

5. The robust random hybrid optimization scheduling method based on period division according to claim 4 is characterized in that: The expression of the robust random hybrid day-ahead scheduling model is: In the formula, It is a random intraday strategy; is a scene in the random scene set; e SP is a random scene set; e RO is the robust uncertainty set; For the bad scenario with uncertain concentration; ID,RO is a robust intraday strategy; c ID,RO is the robust intraday cost; is; E(·) is the mathematical expectation expression; It is a robust random hybrid intraday strategy; for robust random mixing scenarios.

6. The robust random hybrid optimization scheduling method based on period division according to claim 5 is characterized in that: In the process of solving the robust random hybrid day-ahead scheduling model by the column and constraint generation algorithm, the column and constraint generation algorithm decomposes the min-max two-level optimization problem into a main problem and sub-questions for solution; the expression of the main problem is: In the formula, x is the first stage optimization variable; η is the constructed auxiliary variable; k is the number of iterations of the algorithm; The set of scenarios that the algorithm searches for; is the scene in the scene set; l Optimize variables for the second stage; a T is the first stage cost coefficient vector after linearization; b T is the linearized second-stage cost coefficient vector; A, B, C, D, E, F, G, H are the linearized constraint coefficient matrices; T is the matrix transpose; The expression of the sub-problem is: In the formula, x * is the first-stage optimization variable obtained for the main problem; π is the dual variable corresponding to the inequality constraint; λ is the dual variable corresponding to the equality constraint; v is the constructed binary auxiliary variable; M is a sufficiently large number.

7. A robust random hybrid optimization scheduling device based on period division, adopting a robust random hybrid optimization scheduling method based on period division according to any one of claims 1 to 6, characterized in that: include: The day-ahead pre-scheduling model construction and processing module is used to construct the day-ahead pre-scheduling model based on the stochastic optimization strategy; Obtaining a pre-scheduling plan by processing the day-ahead pre-scheduling model; A scheduling period division model construction and processing module is used to construct a scheduling period division model based on a robust optimization strategy; according to safety constraints, the pre-scheduling plan is processed by the scheduling period division model to obtain a period division result; A robust random hybrid day-ahead scheduling model construction module is used to construct a robust random hybrid day-ahead scheduling model according to the period division result; The robust random hybrid day-ahead scheduling model solving module is used to solve the robust random hybrid day-ahead scheduling model through a column and constraint generation algorithm to obtain an optimal scheduling solution.

8. The robust random hybrid optimization scheduling device based on period division according to claim 7 is characterized in that: In the day-ahead pre-dispatch model construction and processing module, the objective function of the day-ahead pre-dispatch model includes: the day-ahead dispatch cost and the expected value of the intraday dispatch cost; the day-ahead dispatch cost includes: the day-ahead market electricity purchase cost, CHP startup cost and energy storage equipment maintenance cost; the intraday dispatch cost includes: the intraday market electricity purchase cost, gas purchase cost and unit equipment maintenance cost; The objective function expression of the day-ahead scheduling model is: In the formula, c DA is the day-ahead scheduling cost; is the intraday scheduling cost; S DA is the day-ahead scheduling strategy expression; is the intraday scheduling strategy expression; S m is an expression of uncertain variables; is the day-ahead electricity price; The amount of electricity purchased on the day before; The start-up costs for the CHP plant; The start-up cost of the gas boiler; is the depreciation cost of the power storage equipment; p mai,ES Equipment maintenance costs for power storage equipment; The charging power of the power storage device; is the energy release power of the power storage device; p mai,HS Equipment maintenance costs for thermal storage equipment; The charging power of the thermal storage device; is the energy release power of the heat storage device; is the intraday electricity price; is the amount of electricity purchased during the day; p gas For gas price; is the gas purchase volume; p mai,CHP Equipment maintenance costs for cogeneration equipment; is the equipment power of the cogeneration equipment; p mai,HP Equipment maintenance costs for ground source heat pumps; is the equipment power of the ground source heat pump; p mai,GB Equipment maintenance costs for gas boilers; is the equipment thermal power of the gas boiler; t is the time sequence number; m is the scene sequence number.

9. The robust random hybrid optimization scheduling device based on period division according to claim 8 is characterized in that: In the day-ahead pre-dispatch model construction and processing module, the constraints of the day-ahead pre-dispatch model include: power system constraints, thermal system constraints, natural gas system constraints, energy coupling equipment constraints, PV output constraints, on-load tap-changing transformer constraints and energy storage equipment constraints.

10. The robust random hybrid optimization scheduling device based on period division according to claim 9, characterized in that: In the scheduling period division model construction and processing module, the expression of the scheduling period division model is: Where τ is the time; n slack is the slack variable; N is the matrix corresponding to the slack variable; g(·)≤0 is the inequality constraint in the above constraints; h(·)=0 is the equality constraint in the above constraints; s DA is the day-ahead scheduling strategy; ID It is an intraday scheduling strategy; For the bad scenarios with uncertain concentration.

Images