Multi-time-scale Scheduling Method, Device, Equipment and Medium for Virtual Power Plant

Abstract

The present invention relates to the technical field of energy scheduling, and particularly relates to a multi-time scale scheduling method, device, equipment and medium for a virtual power plant. The method includes: solving a day-ahead optimization scheduling model of a target virtual power plant to obtain a day-ahead scheduling plan of the virtual power plant, and performing parameter identification on the plan to obtain an optimal parameter set for comprehensive time periods; based on a rolling correction optimization model, using the optimal parameter set for comprehensive time periods to iteratively fine-tune multiple time period parameters to obtain an optimal parameter set for the t+m-1 time period, and using the set to correct the actual parameters of the target t+m time period to obtain the corrected parameters of the target t+m time period, and formulating a multi-time scale scheduling plan for the virtual power plant by using the corrected parameters of the target t+m time period. Thus, the problems that there are a large number of energy quantities and variable operation modes in the existing virtual power plant, it is difficult to accurately obtain the load-side model parameters, and thus it is difficult to construct an efficient virtual power plant optimization scheduling plan are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy scheduling, and particularly relates to a multi-time scale scheduling method, device, equipment and medium of a virtual power plant based on load side state feedback. Background Art

[0002] The attention to energy and environmental issues has become a key bottleneck restricting global economic and social development. In order to cope with the dual crises of energy and environment, many countries have promulgated various renewable energy policies and energy development plans. Vigorously increase the proportion of renewable energy power generation to reduce carbon emissions. However, the uncertainty and intermittency of renewable energy have brought huge challenges to the power system. Sufficient flexibility is needed to cope with the fluctuations of renewable energy and maintain the balance between load and energy supply. In this situation, it is urgent to further optimize resource planning and upgrade the development of the energy system. The virtual power plant is an effective way to promote the optimized development of the energy system by uniformly managing distributed energy, energy storage, and flexible electrical and thermal loads, etc. However, there are many difficulties in the implementation process, and the main difficulties are as follows:

[0003] (1) There are many types of virtual power plant equipment, the models are complex, the operation modes are changeable, and it is difficult to accurately obtain the load side model parameters, and thus it is difficult to construct an efficient virtual power plant optimal scheduling scheme;

[0004] (2) Uncertain factors such as wind and light output and load fluctuations exist in the virtual power plant. How to ensure the robust operation of the virtual power plant scheduling plan in a multi-source uncertain environment is another difficulty;

[0005] (3) Existing research does not consider using the virtual power plant as a means to simultaneously optimize multiple types of energy under the background of the energy revolution, so as to break the barriers of independent operation between energies and promote energy complementarity and consumption. Only consider the overall operation safety of the system, volatility processing or multi-energy green trading, and ignore the operation characteristics and setting value of the units in the multi-energy system during the balance interaction process. Summary of the Invention

[0006] The present invention provides a multi-time scale scheduling method, device, equipment and medium of a virtual power plant to solve the problem that there are many types of energy in the existing virtual power plant, the operation modes are changeable, it is difficult to accurately obtain the load side model parameters, and thus it is difficult to construct an efficient virtual power plant optimal scheduling scheme.

[0007] An embodiment of the first aspect of the present invention provides a multi-time scale scheduling method for a virtual power plant, including the following steps: solving a day-ahead optimization scheduling model of a pre-constructed target virtual power plant to obtain a day-ahead scheduling plan for the virtual power plant; performing parameter identification on the day-ahead scheduling plan of the virtual power plant to obtain an optimal parameter set for the comprehensive time period; based on a pre-constructed rolling correction optimization model, using the optimal parameter set for the comprehensive time period to iteratively fine-tune the parameters of multiple time periods to obtain an optimal parameter set for the (t + m - 1)th time period, where 0 ≤ t ≤ 23 and m ≥ 1; correcting the actual parameters of the target (t + m)th time period of the target virtual power plant according to the optimal parameter set for the (t + m - 1)th time period to obtain the corrected parameters of the target (t + m)th time period, and formulating a multi-time scale scheduling plan for the virtual power plant using the corrected parameters of the target (t + m)th time period.

[0008] Optionally, the step of solving a day-ahead optimization scheduling model of a pre-constructed target virtual power plant to obtain a day-ahead scheduling plan for the virtual power plant includes:

[0009] Decomposing the day-ahead optimization scheduling model into a mixed-integer linear programming model and a non-convex max-min model;

[0010] Using a column constraint generation algorithm to solve the mixed-integer linear programming model and the non-convex max-min model to obtain the day-ahead scheduling plan for the virtual power plant.

[0011] Optionally, the step of, based on a pre-constructed rolling correction optimization model, using the optimal parameter set for the comprehensive time period to iteratively fine-tune the parameters of multiple time periods to obtain an optimal parameter set for the (t + m - 1)th time period includes:

[0012] Performing interval sampling on the optimal parameter set for the comprehensive time period to obtain an initial population;

[0013] Based on the rolling correction optimization model, using the initial population to iteratively fine-tune the parameters of the (t - 1)th time period of the target virtual power plant until the number of iterations reaches a preset value, and outputting an optimal parameter set for the (t - 1)th time period;

[0014] Iteratively performing the interval sampling and fine-tuning process on the optimal parameter set for the (t - 1)th time period until the parameter fine-tuning of the (t + m - 1)th time period is completed, and outputting the optimal parameter set for the (t + m - 1)th time period.

[0015] Optionally, the step of, based on the rolling correction optimization model, using the initial population to iteratively fine-tune the parameters of the (t - 1)th time period of the target virtual power plant until the number of iterations reaches a preset value, and outputting an optimal parameter set for the (t - 1)th time period includes:

[0016] Construct the virtual power plant optimal scheduling plan for the next 24 hours at time t-1 of the target virtual power plant according to the initial population, and determine the planned output at time t-1 according to the virtual power plant optimal scheduling plan for the next 24 hours at time t-1;

[0017] Obtain the actual output at time t-1 of the target virtual power plant, and construct a balance cost function according to the planned output at time t-1 and the actual output at time t-1;

[0018] Solve the balance cost function based on the quantum genetic algorithm, obtain multiple fitness values, and sort the multiple fitness values to obtain the sorted fitness values;

[0019] Perform quantum rotation and evolution on the sorted fitness values to obtain the next generation population;

[0020] Based on the rolling correction optimization model, iteratively execute the foregoing process on the next generation population until the number of iterations reaches a preset value, complete the parameter fine-tuning at time t-1, and output the optimal parameter set at time t-1.

[0021] An embodiment of the second aspect of the present invention provides a multi-time scale scheduling device for a virtual power plant, including:

[0022] A solving module, configured to solve a day-ahead optimal scheduling model of a pre-constructed target virtual power plant to obtain a virtual power plant day-ahead scheduling plan;

[0023] A parameter identification module, configured to perform parameter identification on the virtual power plant day-ahead scheduling plan to obtain an optimal parameter set for the comprehensive time period;

[0024] A rolling correction module, configured to iteratively fine-tune the parameters of multiple time periods based on a pre-constructed rolling correction optimization model by using the optimal parameter set for the comprehensive time period to obtain an optimal parameter set at time t+m-1, where 0≤t≤23 and m≥1;

[0025] A correction module, configured to correct the actual parameters of the target virtual power plant at the target time t+m according to the optimal parameter set at time t+m-1 to obtain the corrected parameters at the target time t+m, and formulate a multi-time scale scheduling plan for the virtual power plant by using the corrected parameters at the target time t+m.

[0026] Optionally, the solving module includes:

[0027] A decomposition unit, configured to decompose the day-ahead optimal scheduling model into a mixed integer linear programming model and a non-convex max-min model;

[0028] A solution unit, configured to solve the mixed-integer linear programming model and the non-convex max-min model by using a column constraint generation algorithm, so as to obtain the day-ahead scheduling plan of the virtual power plant.

[0029] Optionally, the rolling correction module includes:

[0030] A sampling unit, configured to perform interval sampling on the comprehensive period optimal parameter set to obtain an initial population;

[0031] A fine-tuning unit, configured to iteratively fine-tune the parameters of the (t - 1)th period of the target virtual power plant by using the initial population based on the rolling correction optimization model until the number of iterations reaches a preset value, and output the optimal parameter set of the (t - 1)th period;

[0032] A multi-period iteration unit, which iteratively executes the interval sampling and fine-tuning processes on the optimal parameter set of the (t - 1)th period until the parameter fine-tuning of the (t + m - 1)th period is completed, and outputs the optimal parameter set of the (t + m - 1)th period.

[0033] Optionally, the fine-tuning unit includes:

[0034] A first construction subunit, configured to construct an optimized scheduling plan for the virtual power plant in the next 24 hours of the (t - 1)th period of the target virtual power plant according to the initial population, and determine the planned output of the (t - 1)th period according to the optimized scheduling plan for the virtual power plant in the next 24 hours of the (t - 1)th period;

[0035] A second construction subunit, configured to obtain the actual output of the (t - 1)th period of the target virtual power plant, and construct a balance cost function according to the planned output of the (t - 1)th period and the actual output of the (t - 1)th period;

[0036] A sorting subunit, configured to solve the balance cost function based on a quantum genetic algorithm to obtain a plurality of fitness values, and sort the plurality of fitness values to obtain the sorted fitness values;

[0037] An evolution subunit, configured to perform quantum rotation and evolution on the sorted fitness values to obtain a next-generation population;

[0038] A single-period iteration unit, configured to iteratively execute the foregoing process on the next-generation population based on the rolling correction optimization model until the number of iterations reaches a preset value, complete the parameter fine-tuning of the (t - 1)th period, and output the optimal parameter set of the (t - 1)th period.

[0039] An embodiment of the third aspect of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the multi-time scale scheduling method of the virtual power plant as described in the foregoing embodiment.

[0040] In the fourth aspect of the present invention, an embodiment provides a computer-readable storage medium that stores a computer program, and when the program is executed by a processor, it implements the multi-time scale scheduling method of the virtual power plant as described above.

[0041] The multi-time scale scheduling method, device, equipment, and medium of the virtual power plant proposed in the embodiments of the present invention fully consider the characteristics of the virtual power plant, such as a large number of device types, complex models, and variable operation modes, as well as the prediction uncertainty problems on the power supply side and load side of the virtual power plant. A two-stage robust optimization model of the virtual power plant with a min-max-min structure is constructed, and state parameters are obtained by using the two-stage robust optimization model of the virtual power plant, so as to solve problems such as difficult to accurately obtain load-side model parameters, and update the load-side model parameters by using the state parameters, thereby realizing the simultaneous optimization of multiple energy sources, breaking the barrier of independent operation between energy sources, promoting energy complementarity and full consumption, and fully considering the operation characteristics and setting value of the units in the multi-energy system during the balance interaction process.

[0042] Some of the additional aspects and advantages of the present invention will be given in the following description, some will become obvious from the following description, or will be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The above and / or additional aspects and advantages of the present invention will become obvious and easy to understand from the following description of the embodiments in conjunction with the drawings, where:

[0044] Figure 1 is a schematic flowchart of a multi-time scale scheduling method of a virtual power plant provided by an embodiment of the present invention;

[0045] Figure 2 is a schematic flowchart for solving the day-ahead optimal scheduling model of the target virtual power plant provided by an embodiment of the present invention;

[0046] Figure 3 is a schematic flowchart for iteratively fine-tuning multiple time period parameters with the comprehensive time period optimal parameter set provided by an embodiment of the present invention;

[0047] Figure 4 is a schematic block diagram of a multi-time scale scheduling device of a virtual power plant provided by an embodiment of the present invention;

[0048] Figure 5 is a schematic structural diagram of the electronic equipment provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as limiting the present invention.

[0050] Figure 1 It is a schematic flow chart of a multi-time scale scheduling method for a virtual power plant provided by an embodiment of the present invention.

[0051] As Figure 1 shown, the multi-time scale scheduling method for the virtual power plant includes the following steps:

[0052] In step S101, the day-ahead optimal scheduling model of the pre-constructed target virtual power plant is solved to obtain the day-ahead scheduling plan of the virtual power plant.

[0053] In some embodiments, solving the day-ahead optimal scheduling model of the pre-constructed target virtual power plant to obtain the day-ahead scheduling plan of the virtual power plant includes:

[0054] Decompose the day-ahead optimal scheduling model into a mixed integer linear programming model and a non-convex max-min model;

[0055] Use the column constraint generation algorithm to solve the mixed integer linear programming model and the non-convex max-min model to obtain the day-ahead scheduling plan of the virtual power plant.

[0056] It should be noted that the main bodies managed by the virtual power plant include gas turbines, combined heat and power units, wind turbines, photovoltaics, storage batteries, and electrical and thermal loads. Among them, wind turbines and photovoltaics are the main power output units for electricity. The storage battery and the distribution network play a role in balancing the power supply and demand of the system. The gas turbine uses natural gas as fuel to mainly produce electricity, but the waste heat generated by the machine can also be recycled.

[0057] In the actual execution process, the day-ahead optimal scheduling model of the pre-constructed target virtual power plant includes an objective function, inequality constraints, and equality constraints. Among them,

[0058] The objective function of the day-ahead optimal scheduling model is divided into six parts, and F(t) represents the total operating cost of the virtual power plant in the t-th period:

[0059]

[0060] Among them, is the total cost of power and heat co-generation units supplying power and heat in the virtual power plant in the t-th period:

[0061]

[0062] Where α, β, γ, θ, δ, are the operating cost coefficients of the combined heat and power unit, is the electrical output of the combined heat and power unit at time t, with the unit of MW, is the heat output of the combined heat and power unit at time t, with the unit of MW.

[0063] The total cost of power supply of the distributed thermal power unit in the t-th period is:

[0064]

[0065] Where a, b, c are the operating cost coefficients of the distributed thermal power unit, is the penalty or compensation cost of the distributed thermal power unit under the peak shaving ancillary service market mechanism, P D t G is the electrical output of the distributed thermal power unit at time t, with the unit of MW.

[0066] is the charging and discharging cost of the energy storage in the t-th period, is the compensation obtained by the energy storage participating in the ancillary peak shaving market:

[0067]

[0068] Where C ESS is the charging and discharging cost coefficient of the energy storage, are the charging power and discharging power of the energy storage at time t respectively.

[0069] The compensation cost of the adjustable load in the t-th period is:

[0070]

[0071] Where is the adjustable load compensation cost coefficient, is the change in adjustable load power.

[0072] The revenue from external power purchase and sale of the virtual power plant in the t-th period is:

[0073]

[0074] Where is the external power purchase and sale price at time t, is the external power purchase and sale electric power at time t.

[0075] The revenue from supplying power to the load by the virtual power plant in the t-th period is:

[0076]

[0077] Wherein, are the revenue coefficients of the virtual power plant supplying electricity and heat loads at time t, respectively, are the powers of the virtual power plant supplying electricity and heat loads at time t, respectively.

[0078] The inequality constraints of the day-ahead optimal dispatch model include:

[0079] The power output constraints of wind power, photovoltaic, gas turbine, and heat power generation units, and their expressions are:

[0080]

[0081] Wherein, is the power output of each unit at time t, with the unit of MW, are the upper and lower limits of the power output of each unit, with the unit of MW, respectively.

[0082] The heat output constraint of the combined heat and power generation unit, and its expression is:

[0083]

[0084] Wherein, is the heat output of the combined heat and power generation unit at time t, with the unit of MW, and are the upper and lower limits of the heat output of the combined heat and power generation unit, with the unit of MW, respectively.

[0085] The thermoelectric coupling constraint of the combined heat and power generation unit, and its expression is:

[0086]

[0087] Wherein, C m is the thermoelectric ratio of the combined heat and power generation unit under back pressure conditions, and K is a constant, is the power output of the combined heat and power generation unit at time t, with the unit of MW, P emax is the upper limit of the electric power of the combined heat and power generation unit, c v is the absolute value of the slope of the curve of the combined heat and power generation unit under pure condensation working conditions, is the heat output of the combined heat and power generation unit at time t, with the unit of MW.

[0088] The heat power constraint, and its expression is:

[0089]

[0090] Wherein, is the heat output of the combined heat and power supply unit in the t period, with the unit of MW, is the heat output of the electric heating equipment in period t, with the unit of MW. is the heat load in period t, with the unit of MW.

[0091] The energy storage charge and discharge constraint, and its expression is:

[0092]

[0093] In the formula, and are the discharge or charge power of the energy storage at time t, respectively, with the unit of MW, is the energy state of the energy storage at time t, with the unit of MWh; SOC max are the upper limits of the charge and discharge power and the energy state of the energy storage, respectively, with the unit of MW.

[0094] The adjustable load constraint, and its expression is:

[0095]

[0096] In the formula, is the upper limit of the control capacity of the controlled load, is the change amount of the adjustable load power.

[0097]

[0098] In the formula, S is the period when dispatching control measures can be taken.

[0099] The indoor temperature constraint in the heating area, and its expression is:

[0100] T in min ≤ T in (t) ≤ T in max

[0101] In the formula, T in (t) is the indoor temperature in the heating area at time t, T in min and T in max are the upper and lower limits of the indoor temperature for heating in the north in winter.

[0102] The equality constraints of the day-ahead optimal dispatching model include:

[0103] The electric power balance constraint, and its expression is:

[0104]

[0105] In the formula, on the left side of the equation, are the thermal power units, gas turbines, photovoltaics, wind power outputs and the energy storage discharge power at time t, respectively, with the unit of MW. On the right side of the equation, They are the energy storage discharge power, electrical load, electric heating load, the power of the virtual power plant participating in the power purchase and sale of the power grid, and the change in the adjustable load power in the source-network-load-storage integrated system at time t, with the unit of MW.

[0106] The electric-heat conversion power constraint, and its expression is:

[0107]

[0108] In the formula, is the electric power consumed by the electric heating equipment at time t, with the unit of MW, is the heat output of the electric heating equipment at time t, with the unit of MW, ζ eb is the electric-heat conversion efficiency.

[0109] The energy state constraint of the energy storage equipment, and its expression is:

[0110]

[0111] In the formula, λ soc is the energy storage efficiency, η ESS is the energy storage charging efficiency, and are the discharge or charging power of the energy storage at time t, with the unit of MW, is the energy state of the energy storage at time t-1, with the unit of MWh.

[0112] The thermal model constraint of the flexible load heating area, and its expression is:

[0113] k1Q DLC (t)+k4(T out (t-1)-T in (t-1))=k3T in (t)-k2T in (t-1)

[0114] In the formula, Q DLC (t) is the heat dissipation of the flexible thermal load at time t, T in (t) is the indoor temperature of the heating area at time t, T out (t) is the outdoor temperature of the heating area at time t, and k1, k2, k3, and k4 are the corresponding constant coefficients.

[0115] Furthermore, due to the strong uncertainty of the fan, photovoltaic output, and base load, the influence of uncertain factors needs to be added to the model. Among them, the constructed uncertain set is:

[0116]

[0117] In the formula, μ wt (t), μpv (t), μ p (t), μ q (t) are the uncertain variables of the fan, photovoltaic output power, and electrical and thermal base loads respectively, are the predicted value column vectors of the fan, photovoltaic output power, and electrical and thermal base loads at time t respectively, are the maximum allowable deviations of the fan, photovoltaic output power, and electrical and thermal base loads respectively, which are positive numbers, and N is the number of time periods.

[0118] If only the deterministic model is considered, the matrix representation form of the deterministic optimal scheduling model of the virtual power plant is:

[0119]

[0120] In the formula, c is the coefficient column vector of the objective function, and D, K, F, G, I μ are the coefficient matrices corresponding to the constraint conditions, and d, ξ, k are the constant column vectors, are the constant column vectors of the predicted values in each time period, and x, y are the optimization variables.

[0121] The idea of robust optimization is to consider the scheduling plan that optimizes the objective function when the uncertain variable μ changes towards the worst case within the uncertainty set U according to the constructed uncertainty set. Therefore, the robust optimization model (i.e., the day-ahead optimal scheduling model) established in the embodiments of the present invention is:

[0122]

[0123] The min outside the curly brackets is the outer first-stage optimization problem, and the optimization variable is x. The max min inside the curly brackets is the inner second-stage optimization problem, and the optimization variables are μ and x. Ω is the value range of y under the selected conditions of x and μ, and can be expressed as:

[0124]

[0125] In the formula, y| represents the value range of y under the four limiting conditions of Dy≥d, Ky = ξ, Fx + Gy≥k, I μ y = μ, and γ, λ, ν, π are the corresponding dual variables in the inner second-stage optimization problem.

[0126] Furthermore, as Figure 2 shown, the embodiments of the present invention decompose the day-ahead optimal scheduling model into two problems for solution. Among them, the main problem is a mixed-integer linear programming problem, and the sub-problem is a non-convex max-min problem. The two problems are iterated with each other and the upper and lower bounds are continuously updated to make the problem converge faster. When the difference between the upper and lower bounds reaches the allowable range, the algorithm terminates, and finally the optimal solution of the problem is found.

[0127] Among them, the main problem structure is:

[0128]

[0129] In the formula, α is an auxiliary variable, n is the current iteration number, y l is the solution after the l-th iteration, is the value of the uncertain variable μ under the worst condition obtained after the l-th iteration.

[0130] The sub-problem structure is:

[0131] max μ∈U min y∈Ω(x,μ) c T y

[0132] Since the formula of the sub-problem is a non-convex maximum problem and is difficult to solve directly, the embodiment of the present invention also uses the duality theorem to convert the max-min problem into a single-level bilinear programming problem, which can be expressed as:

[0133]

[0134] According to the above derivation, the column constraint generation (C&CG) algorithm is used to solve the main problem, the single-level bilinear programming problem and the sub-problem of the mixed integer linear optimization model, and the day-ahead scheduling scheme of the virtual power plant is obtained.

[0135] In step S102, parameter identification is performed on the day-ahead scheduling scheme of the virtual power plant to obtain the optimal parameter set for the comprehensive time period.

[0136] In step S103, based on the pre-constructed rolling correction optimization model, the optimal parameter set for the comprehensive time period is used to iteratively fine-tune the parameters for multiple time periods to obtain the optimal parameter set for the t+m-1 time period, where 0≤t≤23 and m≥1.

[0137] In some embodiments, based on the pre-constructed rolling correction optimization model, using the optimal parameter set for the comprehensive time period to iteratively fine-tune the parameters for multiple time periods to obtain the optimal parameter set for the t+m-1 time period includes:

[0138] Performing interval sampling on the optimal parameter set for the comprehensive time period to obtain an initial population;

[0139] Based on the rolling correction optimization model, using the initial population to iteratively fine-tune the parameters for the t-1 time period of the target virtual power plant until the number of iterations reaches a preset value, and output the optimal parameter set for the t-1 time period;

[0140] Iteratively execute the interval sampling and fine-tuning process on the optimal parameter set for the t-1 time period until the parameter fine-tuning for the t+m-1 time period is completed, and output the optimal parameter set for the t+m-1 time period.

[0141] It should be noted that the pre-constructed rolling correction optimization model is as follows:

[0142]

[0143] Among them, y t+1 is the aggregation parameter for time period t + 1, and x t+1 is the scheduling plan vector of all devices in the next scheduling time period. A t+1 , b t+1 is the (t + 1)-th row of the corresponding matrix. x t is the actual operation of the device in scheduling cycle t.

[0144] According to information such as the current operating state and the energy storage SOC, the matrix G in the constraint conditions is updated to G'(x t ), the vector P is updated to P', and the virtual power plant obtains the adjustable capacity range during the next scheduling.

[0145] The intra-day real-time correction optimization problem of the virtual power plant becomes a deterministic optimization problem. Therefore, real-time online correction can reduce or even eliminate the conservatism caused by uncertainty, and the rolling correction result can provide a basis for parameter selection in the day-ahead aggregation optimization problem. For example, if the lower limit of the system power is often modified, the weight coefficient of the lower limit of the modified scheduling time period can be appropriately increased. The heuristic parameter correction method can be expressed as:

[0146]

[0147] Among them, π pm,t is the weight coefficient corresponding to the lower limit of scheduling cycle t, k is a constant column vector, is the correction weight coefficient of scheduling cycle t, P DA and p RO are the lower bound parameters in the day-ahead optimal scheduling model and the intra-day rolling correction model, respectively.

[0148] In the actual implementation process, first, the comprehensive period optimal parameter set is obtained as the initial parameter set for the t-1 period. Interval sampling is performed on the initial parameter set for the t-1 period to obtain the initial population. Based on the rolling correction optimization model, according to the initial population, an optimal dispatching plan for the virtual power plant for the next 24 hours in the t-1 period of the target virtual power plant is constructed, and the planned output for the t-1 period is determined according to the optimal dispatching plan for the virtual power plant for the next 24 hours in the t-1 period. The actual output of the target virtual power plant in the t-1 period is obtained, and a balance cost function is constructed based on the planned output and the actual output in the t-1 period. The balance cost function is solved based on the quantum genetic algorithm to obtain multiple fitness values, and the multiple fitness values are sorted to obtain the sorted fitness values. Quantum rotation and evolution are performed on the sorted fitness values to obtain the first-generation population. Then, a new optimal dispatching plan for the virtual power plant for the next 24 hours in the t-1 period of the target virtual power plant is constructed using the first-generation population, and the new planned output for the t-1 period is determined according to the new optimal dispatching plan for the virtual power plant for the next 24 hours in the t-1 period. A new balance cost function is constructed based on the new planned output for the t-1 period and the initially obtained actual output in the t-1 period. The new balance cost function is solved based on the quantum genetic algorithm to obtain new multiple fitness values, and the new multiple fitness values are sorted to obtain the new sorted fitness values. Quantum rotation and evolution are performed on the new sorted fitness values to obtain the second-generation population. Then, the same operation process as the first-generation population is performed on the second-generation population, and this process is iterated to continuously update the population until the number of iterations reaches the preset value G, the parameter fine-tuning for the t-1 period is completed, and the optimal parameter set for the t-1 period is output.

[0149] Furthermore, as Figure 3As shown, the optimal parameter set in the (t - 1) period is used as the initial parameter set in the t period, and interval sampling is performed on the initial parameter set in the t period to obtain the initial population in the (t - 1) period. Based on the rolling correction optimization model, a 24-hour virtual power plant optimal scheduling plan for the t period of the target virtual power plant is constructed according to the initial population in the (t - 1) period, and the planned output in the t period is determined according to the 24-hour virtual power plant optimal scheduling plan for the t period. The actual output in the t period of the target virtual power plant is obtained, and a balance cost function is constructed according to the planned output and the actual output in the t period. Based on the quantum genetic algorithm, the balance cost function is solved to obtain multiple fitness values, and the multiple fitness values are sorted to obtain the sorted fitness values. Quantum rotation and evolution are performed on the sorted fitness values to obtain the first-generation population. Then, a new 24-hour virtual power plant optimal scheduling plan for the t period of the target virtual power plant is constructed using the first-generation population, and the new planned output in the t period is determined according to the new 24-hour virtual power plant optimal scheduling plan for the t period. A new balance cost function is constructed according to the new planned output and the initially obtained actual output in the t period. Based on the quantum genetic algorithm, the new balance cost function is solved to obtain new multiple fitness values, and the new multiple fitness values are sorted to obtain the new sorted fitness values. Quantum rotation and evolution are performed on the new sorted fitness values to obtain the second-generation population. Then, the same operation process as the first-generation population is performed on the second-generation population, and this process is iterated to continuously update the population until the number of iterations reaches the preset value, completing the parameter fine-tuning in the t period, and outputting the optimal parameter set in the (t + 1) period.

[0150] Iterate the aforementioned interval sampling and fine-tuning process according to the actual demand (the period t + m that needs to be corrected) until the parameter fine-tuning of the previous period of the period that needs to be corrected is completed, and output the optimal parameter set of the previous period of the period that needs to be corrected (i.e., the optimal parameter set in the (t + m - 1) period). For example, if m = 2 in the period that needs to be corrected, then it is necessary to fine-tune the (t - 1) period and the t period, and finally obtain the optimal parameter set in the (t + 1) period.

[0151] In step S104, the actual parameters of the target t + m period of the target virtual power plant are corrected according to the optimal parameter set in the (t + m - 1) period to obtain the corrected parameters of the target t + m period, and a multi-time scale scheduling plan for the virtual power plant is formulated using the corrected parameters of the target t + m period.

[0152] In the actual execution process, the optimal parameter set in the (t + m - 1) period is used as the initial parameter set in the (t + m) period, the actual parameters of the target t + m period of the target virtual power plant are corrected using the initial parameter set in the (t + m) period, and finally, a multi-time scale scheduling plan for the virtual power plant is formulated according to the corrected parameters of the target t + m period.

[0153] The multi-time scale scheduling method of a virtual power plant proposed according to an embodiment of the present invention has the following beneficial effects:

[0154] (1) It fully considers the characteristics of a virtual power plant, such as a large number of equipment types, complex models, and variable operation modes, and constructs an operation constraint and operation cost model for each flexibility resource model.

[0155] (2) It fully considers the prediction uncertainty problems on the power supply side and load side of the virtual power plant, and constructs a two-stage robust optimization model of the virtual power plant with a min-max-min structure.

[0156] (3) Based on state feedback, the load side model parameters are updated in real time, which can be used to solve problems such as difficult accurate acquisition of load side model parameters.

[0157] (4) Considering the uncertainty of renewable energy and load prediction in the virtual power plant and the inaccuracy of load side model parameters, a scheduling method combining day-ahead scheduling and intra-day correction of the virtual power plant is formed, and a multi-time scale scheduling scheme of the virtual power plant is obtained.

[0158] (5) Taking the virtual power plant as a means, multiple energy sources are optimized simultaneously, thereby breaking the barriers of independent operation between energy sources, promoting energy complementarity and full consumption, and fully considering the operation characteristics and setting values of units in the multi-energy system during the balance interaction process.

[0159] Next, a multi-time scale scheduling device of a virtual power plant proposed according to an embodiment of the present invention is described with reference to the accompanying drawings.

[0160] Figure 4 It is a block diagram of a multi-time scale scheduling device of a virtual power plant according to an embodiment of the present invention.

[0161] As Figure 4 shown, the multi-time scale scheduling device 40 of the virtual power plant includes: a solution module 401, a parameter identification module 402, a rolling correction module 403, and a correction module 404.

[0162] Among them, the solution module 401 is used to solve the day-ahead optimal scheduling model of the pre-constructed target virtual power plant to obtain the day-ahead scheduling plan of the virtual power plant. The parameter identification module 402 is used to identify the parameters of the day-ahead scheduling plan of the virtual power plant to obtain the optimal parameter set for the comprehensive time period. The rolling correction module 403 is used to iteratively fine-tune the parameters of multiple time periods based on the pre-constructed rolling correction optimization model using the optimal parameter set for the comprehensive time period to obtain the optimal parameter set for the t+m-1 time period, where 0≤t≤24 and m≥1. The correction module 404 is used to correct the actual parameters of the target virtual power plant for the target t+m time period according to the optimal parameter set for the t+m-1 time period to obtain the corrected parameters for the target t+m time period, and formulate the multi-time scale scheduling plan of the virtual power plant using the corrected parameters for the target t+m time period.

[0163] In some embodiments, the solution module 401 includes:

[0164] A decomposition unit, configured to decompose the day-ahead optimal scheduling model into a mixed integer linear programming model and a non-convex max-min model;

[0165] A solution unit, configured to solve the mixed integer linear programming model and the non-convex max-min model using a column constraint generation algorithm to obtain the day-ahead scheduling plan of the virtual power plant.

[0166] In some embodiments, the rolling correction module 403 includes:

[0167] A sampling unit, configured to perform interval sampling on the optimal parameter set for the comprehensive time period to obtain an initial population;

[0168] A fine-tuning unit, configured to iteratively fine-tune the parameters of the t-1 time period of the target virtual power plant based on the rolling correction optimization model using the initial population until the number of iterations reaches a preset value, and output the optimal parameter set for the t-1 time period;

[0169] A multi-time period iteration unit, which iteratively performs the interval sampling and fine-tuning processes on the optimal parameter set for the t-1 time period until the parameter fine-tuning for the t+m-1 time period is completed, and outputs the optimal parameter set for the t+m-1 time period.

[0170] In some embodiments, the fine-tuning unit includes:

[0171] A first construction subunit, configured to construct a 24-hour virtual power plant optimal scheduling plan for the future of the t-1 time period of the target virtual power plant according to the initial population, and determine the planned output for the t-1 time period according to the 24-hour virtual power plant optimal scheduling plan for the future of the t-1 time period;

[0172] A second construction subunit, configured to obtain the actual output of the target virtual power plant for the t-1 time period, and construct a balance cost function according to the planned output and the actual output for the t-1 time period;

[0173] A sorting subunit, configured to solve a balance cost function based on a quantum genetic algorithm, obtain multiple fitness values, and sort the multiple fitness values to obtain sorted fitness values;

[0174] An evolution subunit, configured to perform quantum rotation and evolution on the sorted fitness values to obtain a next-generation population;

[0175] A single-period iteration unit, configured to iteratively execute the foregoing process on the next-generation population based on a rolling correction optimization model until the number of iterations reaches a preset value, complete fine-tuning of parameters in the (t - 1)th period, and output an optimal parameter set for the (t - 1)th period.

[0176] It should be noted that the foregoing explanation of the embodiments of the multi-time scale scheduling method for a virtual power plant also applies to the multi-time scale scheduling device of the virtual power plant in this embodiment, and details are not described herein again.

[0177] The multi-time scale scheduling device of a virtual power plant according to an embodiment of the present invention has the following beneficial effects:

[0178] (1) Fully consider the characteristics of a large number of equipment types, complex models, and variable operation modes in a virtual power plant, and construct operation constraints and operation cost models for each type of flexibility resource model;

[0179] (2) Fully consider the prediction uncertainty problems on the power supply side and load side of the virtual power plant, and construct a two-stage robust optimization model of the virtual power plant with a min-max-min structure;

[0180] (3) Based on state feedback, the load-side model parameters are updated in real time, which can be used to solve problems such as difficult accurate acquisition of load-side model parameters;

[0181] (4) Consider the uncertainty of renewable energy and load prediction in the virtual power plant and the inaccuracy of load-side model parameters, form a scheduling method combining day-ahead scheduling and intra-day correction of the virtual power plant, and obtain a multi-time scale scheduling scheme for the virtual power plant;

[0182] (5) Taking the virtual power plant as a means, simultaneously optimize multiple types of energy, thereby breaking the barriers of independent operation between energies, promoting energy complementarity and full consumption, and fully considering the operation characteristics and setting values of units in the multi-energy system during the balance interaction process.

[0183] Figure 5 The structure diagram of an electronic device provided by an embodiment of the present invention. The electronic device may include:

[0184] A memory 501, a processor 502, and a computer program stored on the memory 501 and executable on the processor 502.

[0185] When the processor 502 executes the program, it implements the multi-time scale scheduling method of the virtual power plant provided in the above embodiments.

[0186] Furthermore, the electronic device further includes:

[0187] A communication interface 503 for communication between the memory 501 and the processor 502.

[0188] A memory 501 for storing a computer program that can run on the processor 502.

[0189] The memory 501 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.

[0190] If the memory 501, the processor 502, and the communication interface 503 are implemented independently, the communication interface 503, the memory 501, and the processor 502 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 5 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0191] Optionally, in a specific implementation, if the memory 501, the processor 502, and the communication interface 503 are integrated on a chip, the memory 501, the processor 502, and the communication interface 503 can communicate with each other through an internal interface.

[0192] The processor 502 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0193] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the multi-time scale scheduling method of the virtual power plant as described above.

[0194] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples", etc., mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0195] In addition, the terms "first" and "second" are used only for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0196] Any process or method description shown in a flowchart or described in other ways herein can be understood to represent a module, segment, or portion of code including one or N executable instructions for implementing a customized logic function or process, and the scope of the preferred embodiments of the present invention includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0197] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or N wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically by optically scanning the paper or other media, followed by editing, interpretation, or other appropriate processing as necessary, and then stored in a computer memory.

[0198] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0199] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the method of the above-described embodiments can be completed by a program instructing relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0200] In addition, in each embodiment of the present invention, each functional unit can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0201] The above-mentioned storage medium can be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A multi-time scale scheduling method for a virtual power plant, characterized in that: The following steps are involved: Solve the pre-built day-ahead optimization dispatch model of the target virtual power plant and obtain the day-ahead dispatch plan of the virtual power plant; Identify parameters of the day-ahead dispatching plan of the virtual power plant to obtain an optimal parameter set for a comprehensive period; Based on the pre-built rolling correction optimization model, the comprehensive period optimal parameter set is used to iteratively fine-tune the parameters of multiple periods to obtain the optimal parameter set of the t+m-1 period, wherein the pre-built rolling correction optimization model is: In the formula, y t+1 is the aggregation parameter for time period t+1, x t+1 is the scheduling plan vector of all devices in the next scheduling period t+1, A t+1 , b t+1 are the t+1th row of matrices A and b, respectively, and x t is the actual operation status of the equipment in the scheduling period t, 0≤t≤23, m≥1; According to the optimal parameter set for the t+m-1 period, the actual parameters of the target virtual power plant in the target t+m period are corrected to obtain the corrected parameters of the target t+m period, and the multi-time scale scheduling plan of the virtual power plant is formulated using the corrected parameters of the target t+m period.

2. The multi-time scale scheduling method of a virtual power plant according to claim 1, characterized in that: The method of solving the pre-built day-ahead optimization scheduling model of the target virtual power plant to obtain the day-ahead scheduling plan of the virtual power plant includes: Decomposing the day-ahead optimization scheduling model into a mixed integer linear programming model and a non-convex maximum-minimum model; The mixed integer linear programming model and the non-convex maximum-minimum model are solved by using a column constraint generation algorithm to obtain the day-ahead dispatching plan for the virtual power plant.

3. The multi-time scale scheduling method of a virtual power plant according to claim 1, characterized in that: The pre-built rolling correction optimization model uses the comprehensive period optimal parameter set to iteratively fine-tune multiple period parameters to obtain the t+m-1 period optimal parameter set, including: Performing interval sampling on the optimal parameter set for the comprehensive period to obtain an initial population; Based on the rolling correction optimization model, the initial population is used to iteratively fine-tune the parameters of the target virtual power plant during the t-1 period until the number of iterations reaches a preset value, and the optimal parameter set for the t-1 period is output; The interval sampling and fine-tuning process is iteratively performed on the optimal parameter set for the t-1 period until the parameter fine-tuning for the t+m-1 period is completed, and the optimal parameter set for the t+m-1 period is output.

4. The multi-time scale scheduling method of a virtual power plant according to claim 3, characterized in that: Based on the rolling correction optimization model, the t-1 period parameters of the target virtual power plant are iteratively fine-tuned using the initial population until the number of iterations reaches a preset value, and the optimal parameter set for the t-1 period is output, including: Constructing an optimal scheduling plan for the future 24 hours of the virtual power plant in the t-1 period of the target virtual power plant according to the initial population, and determining the planned output in the t-1 period according to the optimal scheduling plan for the future 24 hours of the virtual power plant in the t-1 period; Obtaining the actual output of the target virtual power plant during period t-1, and constructing a balancing cost function according to the planned output during period t-1 and the actual output during period t-1; Solving the equilibrium cost function based on a quantum genetic algorithm to obtain a plurality of fitness values, and sorting the plurality of fitness values ​​to obtain sorted fitness values; Performing quantum rotation and evolution on the fitness values ​​after the sorting to obtain the next generation population; The above process is iteratively performed on the next generation population based on the rolling correction optimization model until the number of iterations reaches a preset value, the parameter fine-tuning of the t-1 period is completed, and the optimal parameter set of the t-1 period is output.

5. A multi-time scale scheduling device for a virtual power plant, characterized in that: include: A solution module is used to solve the pre-built day-ahead optimization scheduling model of the target virtual power plant to obtain the day-ahead scheduling plan of the virtual power plant; A parameter identification module, used to identify parameters of the day-ahead dispatching plan of the virtual power plant to obtain an optimal parameter set for a comprehensive period; The rolling correction module is used to iteratively fine-tune the parameters of multiple time periods based on the pre-built rolling correction optimization model using the comprehensive time period optimal parameter set to obtain the optimal parameter set for the t+m-1 time period, wherein the pre-built rolling correction optimization model is: In the formula, y t+1 is the aggregation parameter for time period t+1, x t+1 is the scheduling plan vector of all devices in the next scheduling period t+1, A t+1 , b t+1 are the t+1th row of matrices A and b, respectively, and x t is the actual operation status of the equipment in the scheduling period t, 0≤t≤23, m≥1; A correction module is used to correct the actual parameters of the target t+m period of the target virtual power plant according to the optimal parameter set of the t+m-1 period, obtain the corrected parameters of the target t+m period, and use the corrected parameters of the target t+m period to formulate a multi-time scale scheduling plan for the virtual power plant.

6. The multi-time scale scheduling device for a virtual power plant according to claim 5, characterized in that: The solution module comprises: A decomposition unit, used for decomposing the day-ahead optimization scheduling model into a mixed integer linear programming model and a non-convex maximum-minimum model; A solving unit is used to solve the mixed integer linear programming model and the non-convex maximum and minimum model by using a column constraint generation algorithm to obtain the day-ahead scheduling plan of the virtual power plant.

7. The multi-time scale scheduling device for a virtual power plant according to claim 5, characterized in that: The rolling correction module comprises: A sampling unit, used for performing interval sampling on the optimal parameter set of the comprehensive time period to obtain an initial population; A fine-tuning unit, configured to iteratively fine-tune the parameters of the target virtual power plant in the t-1 period by using the initial population based on the rolling correction optimization model until the number of iterations reaches a preset value, and output an optimal parameter set for the t-1 period; The multi-period iteration unit iteratively performs interval sampling and fine-tuning processes on the optimal parameter set of the t-1 period until the parameter fine-tuning of the t+m-1 period is completed, and outputs the optimal parameter set of the t+m-1 period.

8. The multi-time scale scheduling device for a virtual power plant according to claim 7, characterized in that: The fine-tuning unit comprises: The first construction subunit is used to construct the optimal scheduling scheme of the virtual power plant in the future 24 hours of the target virtual power plant in the t-1 period according to the initial group, and determine the planned output in the t-1 period according to the optimal scheduling scheme of the virtual power plant in the future 24 hours of the t-1 period; The second construction subunit is used to obtain the actual output of the target virtual power plant in the t-1 period, and to construct a balancing cost function according to the planned output in the t-1 period and the actual output in the t-1 period; A sorting subunit, used for solving the equilibrium cost function based on a quantum genetic algorithm to obtain a plurality of fitness values, and sorting the plurality of fitness values ​​to obtain sorted fitness values; An evolution subunit, used for performing quantum rotation and evolution on the fitness values ​​after the sorting to obtain the next generation population; The single-period iteration unit is used to iteratively execute the above process on the next generation population based on the rolling correction optimization model until the number of iterations reaches a preset value, completes the t-1 period parameter fine-tuning, and outputs the t-1 period optimal parameter set.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the multi-time scale scheduling method for a virtual power plant as described in any one of claims 1 to 4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the multi-time scale scheduling method of a virtual power plant as described in any one of claims 1-4.

Images