Comprehensive energy system optimized operation method considering pumped storage

Abstract

Aiming at the problem that the real-time requirement is difficult to meet in the prior art, the invention discloses an integrated energy system optimization operation method considering pumped storage, and belongs to the field of integrated energy optimization operation, and the method comprises the following steps: step 1, establishing renewable energy source, energy conversion equipment and energy storage equipment models; step 2, establishing a flexible load model; 3, the models in the step 1 and the step 2 are combined, a comprehensive energy system multi-scale optimization scheduling model considering the comprehensive demand response is established, and the comprehensive energy system multi-scale optimization scheduling model comprises a day-ahead scheduling layer and a real-time scheduling layer; and step 4, substituting the electricity, heat and gas load prediction data into the objective functions and constraint conditions of the day-ahead scheduling layer and the real-time scheduling layer, and solving the objective functions by adopting a distributed optimization algorithm to realize system operation cost minimization and obtain an equipment output plan. According to the invention, the efficiency, flexibility and robustness of system scheduling are effectively improved.

Description

Technical Field

[0001] The present invention belongs to the field of integrated energy optimized operation, and particularly relates to an optimized operation method for an integrated energy system considering pumped - storage energy storage. Background Art

[0002] With the wide application of renewable energy, the power system is facing increasing dispatching challenges. Traditional dispatching methods often struggle to cope with the volatility and uncertainty of renewable energy such as wind energy and solar energy. Integrated demand response realizes precise matching of supply and demand at multiple time scales by aggregating various types of flexible resources on the user side to interact synergistically with the power grid, greatly improving energy utilization efficiency and supporting the consumption of a high proportion of renewable energy.

[0003] As a large - scale and long - cycle energy storage technology, pumped - storage energy storage can provide strong support for integrated demand response. Its fast charge - discharge characteristics can make up for the delay of demand response resources, while its large - capacity energy storage capacity can smooth the supply - demand fluctuations at long time scales. By jointly dispatching pumped - storage energy storage and demand response resources, the fine - tuning advantages of demand - side resources can be exerted, and the stable output characteristics of pumped - storage energy storage can be utilized. However, existing dispatching models mostly adopt fixed - efficiency assumptions and single - time - scale optimization, resulting in the difficulty of fully adapting the actual operation efficiency of pumped - storage energy storage to the regulation effect of demand response.

[0004] Current traditional dispatching algorithms often adopt a centralized solution method, resulting in huge computational amounts, slow solution speeds, and easy to fall into local optima when facing complex non - linear constraints and large - scale systems, making it difficult to meet real - time requirements. Summary of the Invention

[0005] Aiming at the problems of insufficient coordination between pumped - storage energy storage and integrated demand response, low accuracy of the dispatching model, and fragmentation of multiple time scales in the prior art, the present invention provides an optimized operation method for an integrated energy system considering pumped - storage energy storage. Aiming at the problems of enhanced system volatility and increased difficulty in supply - demand matching under the access of a high proportion of renewable energy, the present invention constructs a day - ahead - real - time two - layer optimization framework by establishing a dynamic efficiency model of pumped - storage energy storage and a flexible load response mechanism, realizing the collaborative and complementary utilization of the long - cycle energy storage characteristics of pumped - storage energy storage and the short - time flexible regulation ability of demand response resources, and distributedly coordinating multiple energy sub - problems by a distributed algorithm; and introducing a dynamic penalty factor to correct the deviation of the energy storage plan, achieving efficient coordination across time scales. This method significantly improves the joint regulation ability of pumped - storage energy storage and demand response resources, while reducing the system operation cost and supporting the efficient consumption of a high proportion of renewable energy.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] An optimized operation method for an integrated energy system considering pumped - storage energy storage, comprising the following steps:

[0008] Step 1: Establish models for renewable energy, energy conversion equipment, and energy storage equipment;

[0009] Step 2: Establish a flexible load model;

[0010] Step 3: Combine the models in Step 1 and Step 2 to establish a multi-scale optimal scheduling model for an integrated energy system considering integrated demand response. The multi-scale optimal scheduling model for the integrated energy system includes a day-ahead scheduling layer and a real-time scheduling layer. The day-ahead scheduling layer is used to output the target energy storage state of pumped-storage, the demand response plan, and the equipment output plan based on load prediction data, renewable energy output prediction curves, economic parameters, and pumped-storage dynamic efficiency parameters. The real-time scheduling layer is used to output the pumped-storage output, the demand response plan, and the implementation of the power purchase plan based on real-time load and renewable energy output data;

[0011] Step 4: Substitute the predicted data of electricity, heat, and gas loads into the objective function and constraint conditions of the day-ahead scheduling layer and the real-time scheduling layer, and use a distributed optimization algorithm to solve the objective function to minimize the system operation cost and obtain the equipment output plan.

[0012] To optimize the above technical solution, the specific measures taken also include:

[0013] Further, in Step 1, the renewable energy includes wind power and photovoltaic power, the energy conversion equipment includes absorption chillers and combined heat and power equipment, and the energy storage equipment includes batteries and pumped-storage systems.

[0014] Further, in Step 2, the flexible load model includes a shiftable load model, a curtailable load model, and a shiftable load model.

[0015] Further, in Step 3, the objective function of the day-ahead scheduling layer is:

[0016] J ahead =min(C chp +C ph +C s +C res )

[0017] Where J ahead is the day-ahead scheduling cost, C chp is the combined heat and power cost, C ph is the pumped-storage operation cost, C s is the shiftable load cost, C res is the renewable energy cost;

[0018] The formula for the combined heat and power cost is as follows:

[0019]

[0020] In the formula, t is the time period, and λ gas is the natural gas price, and P CHP,g (t) is the gas input power of the combined heat and power equipment;

[0021] The formula for the operating cost of the pumped-storage energy is as follows:

[0022] C ph = C water + C wear

[0023]

[0024] In the formula, C water is the total water consumption cost, C wear is the total equipment maintenance cost, c water represents the unit water consumption cost, V pump,t represents the pumped water volume in the t time period, V gen,t represents the water consumption during power generation in the t time period, c wear represents the unit equipment maintenance cost, P pump,t is the pumping power, P gen,t is the power generation power;

[0025] The formula for the cost of the shiftable load is as follows:

[0026]

[0027] In the formula, φ s represents the unit compensation coefficient of the shiftable load, and ΔP s (t) represents the shiftable load power in the t time period, and T is the scheduling period;

[0028] The formula for the cost of the renewable energy is as follows:

[0029]

[0030] In the formula, λ curt represents the cost of wind and solar curtailment, represents the maximum value of wind energy, P wind (t) represents the wind energy magnitude in the t time period, represents the maximum value of solar energy, P solar (t) represents the solar energy magnitude in the t time period.

[0031] Furthermore, in step 3, the objective function of the real-time scheduling layer is:

[0032] J inday = min(C grid + C c + C m + Cstore +ρ(t)·|E t -E target |)

[0033] In the formula, J inday represents the real-time scheduling cost, C grid represents the grid power purchase cost, C c represents the cost of load curtailment, C m represents the compensation cost of shiftable load, C store represents the energy storage cost, ρ(t) represents the dynamic penalty factor, E t represents the actual energy storage state of pumped-storage at time t, E target represents the target energy storage state of the day-ahead scheduling layer;

[0034] The formula for the grid power purchase cost is as follows:

[0035]

[0036] In the formula, λ grid (t) represents the grid power purchase unit price, P grid (t) represents the grid power purchase quantity;

[0037] The formula for the cost of load curtailment is as follows:

[0038]

[0039] In the formula, φ c represents the unit compensation coefficient of load curtailment, T represents the scheduling period, ΔP c (t) is the magnitude of the load curtailment value in the t time period;

[0040] The formula for the compensation cost of shiftable load is as follows:

[0041]

[0042] In the formula, φ m represents the unit compensation coefficient of shiftable load, ΔP m (t) is the magnitude of the shiftable load value in the t time period;

[0043] The formula for the energy storage cost is as follows:

[0044]

[0045] In the formula, λ store represents the charge-discharge loss cost coefficient, P charge (t) represents the battery charging power, P discharge (t) represents the battery discharging power.

[0046] Furthermore, the dynamic penalty factor is used to correct the deviation of the day-ahead plan and coordinate the energy storage state and real-time economy. The expression of the dynamic penalty factor is as follows:

[0047]

[0048] In the formula, ρ0 represents the initial penalty factor, E t represents the actual energy storage capacity state of the pumped-storage at time t, and E target represents the target energy storage state of the day-ahead scheduling layer.

[0049] Furthermore, the constraint conditions include the operation constraints of wind turbines, the operation constraints of photovoltaic power generation, the operation constraints of energy conversion equipment, the operation constraints of batteries, the state update constraints of pumped-storage devices, the state constraints of flexible loads, and the user experience constraints.

[0050] Furthermore, the dynamic efficiency model of the pumped-storage system is:

[0051]

[0052] In the formula, η pump,t represents the pumping efficiency at time t, η turbine,t represents the power generation efficiency at time t, η pump,0 represents the reference efficiency under the pumping condition, η turbine,0 represents the reference efficiency under the power generation condition, and α, β, γ, δ all represent efficiency decay coefficients; h t represents the current reservoir water level, h max and h min represent the upper and lower limits of the water level respectively. Ppump,t is the pumping power, and Pgen,t is the power generation power. represents the maximum pumping power, represents the maximum power generation power;

[0053] The actual energy storage state of the pumped-storage system is updated as follows:

[0054]

[0055] In the formula, E t represents the actual energy storage capacity state of the pumped-storage at time t, and E t-1 represents the actual energy storage capacity state of the pumped-storage at time t-1, and Δt represents the time step.

[0056] Furthermore, in step 4, the distributed optimization algorithm is specifically the alternating direction multiplier method. The specific process of solving the objective function using the alternating direction multiplier method is:

[0057] Step 4.1: Set initial local scheduling decision variables and corresponding Lagrange multipliers for each subsystem; in the day-ahead scheduling layer, initialize the output plans of each subsystem, where the output plans include the target states of pumped-storage energy and the demand response plans; in the real-time scheduling layer, initialize the real-time coordination variables, where the real-time coordination variables include the grid power purchase and the actual states of energy storage.

[0058] Step 4.2: Under the conditions of fixing the global coordination variables and Lagrange multipliers, the power, heat, and flexible load subsystems solve their respective optimization problems in parallel and update the local scheduling decision variables.

[0059] Step 4.3: Update the global coordination variables by fusing the solution results of all subsystems.

[0060] Step 4.4: Adjust the Lagrange multipliers according to the real-time operation data and dynamically correct the penalty factors.

[0061] Step 4.5: Judge the convergence condition. When the deviation between the output plan of each subsystem and the real-time adjustment is less than the preset threshold, stop the iteration and output the final global optimal scheduling plan; otherwise, increment the iteration count by one and return to Step 4.2.

[0062] The beneficial effects of the present invention are as follows:

[0063] (1) The present invention incorporates pumped-storage energy into the optimal scheduling of an integrated energy system with integrated demand response. Through the complementary coordination of the long-cycle energy storage characteristics of pumped-storage energy and the short-time flexible regulation ability of demand response resources, it can effectively suppress the volatility problems brought by the high proportion of renewable energy access - pumped-storage energy stores excess electric energy during low electricity prices or peak periods of wind and light output, while demand response quickly matches the supply-demand deviation through real-time regulation of shiftable and reducible loads, significantly reducing the curtailment rate of wind and light; at the same time, the combination of peak-valley arbitrage of pumped-storage energy and the load transfer strategy of demand response optimizes the economic operation of the system, reducing both the power purchase cost during high-price periods and the demand response compensation cost.

[0064] (2) Multi-scale optimal scheduling significantly improves system flexibility, economy, and reliability by hierarchically coordinating day-ahead plans and real-time adjustments. At the day-ahead level, the target energy storage state of pumped-storage and demand response plans are optimized on a 24-hour cycle to balance long-term energy storage maintenance and short-term economic requirements, reducing the curtailment rate of wind and solar power. At the real-time level, the deviation of energy storage output is corrected in real-time through a dynamic penalty factor to reduce the penalty cost caused by plan deviation, and a distributed algorithm is used to solve multi-energy sub-problems in a distributed parallel manner, improving the computational efficiency. This framework not only achieves efficient coordination across time scales but also enhances the system's adaptability to fluctuations in wind and solar power and load mutations by finely adjusting the complementary characteristics of pumped-storage and flexible loads. Ultimately, while reducing operating costs, it improves the consumption capacity of renewable energy, providing a solution with economy, real-time performance, and robustness for high-proportion renewable energy systems. Description of the Drawings

[0065] Figure 1 Flow chart of the optimal operation method for the integrated energy system considering pumped-storage proposed by the present invention;

[0066] Figure 2 Structural diagram of the integrated energy system of the present invention;

[0067] Figure 3 Optimization process of the multi-scale optimal scheduling model for the integrated energy system;

[0068] Figure 4 Flow chart of the principle of the ADMM algorithm. Detailed Implementation Modes

[0069] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0070] Embodiment 1

[0071] The present invention proposes an optimal operation method for an integrated energy system considering pumped-storage. The flow of this method is as Figure 1 shown, and the structure of the integrated energy system is as Figure 2 shown. The method includes the following steps:

[0072] Step 1: Establish models for renewable energy, energy conversion equipment, and energy storage equipment; renewable energy includes wind power and photovoltaic power, energy conversion equipment includes absorption chillers and combined heat and power equipment, and energy storage equipment includes batteries and pumped-storage systems.

[0073] The wind power generation model is established as follows:

[0074] The output power P of the wind turbine W The relationship with the wind speed U and the operating constraints are as follows:

[0075]

[0076] Where: P rw is the rated output power of the wind turbine, U cin , U cout , U r are the cut-in, cut-out and rated wind speeds respectively.

[0077] Establish a photovoltaic power generation model:

[0078] The output power of photovoltaic power generation is mainly determined by the solar radiation intensity. The operating constraints of photovoltaic power generation can be expressed by the following formula:

[0079] P t PV = G t ·A PV ·η PV

[0080] Where: G t is the solar radiation intensity (W / m 2 ) at time period t, A PV is the total area of the photovoltaic array (m 2 ), η PV is the conversion efficiency of the photovoltaic module, and P t PV is the photovoltaic power generation power at time period t.

[0081] Establish a model of the absorption chiller as follows:

[0082] The output power situation of the absorption chiller is expressed by the following formula:

[0083] P EC,t = E EC,t ·η EC

[0084] Where: E EC,t , P EC,t represent the input and output powers of the electric chiller respectively; the refrigeration efficiency is expressed as η EC .

[0085] The combined heat and power equipment is an energy conversion device that generates electricity by burning natural gas and supplies the waste heat generated during the power generation process to the heat load. Its operating constraints are:

[0086]

[0087] Where: Pg,CHP (t) represents the gas input power, represents the power generation efficiency of the combined heat and power generation, P CHP,e (t) represents the power generation power of the combined heat and power generation system, represents the heat generation efficiency of the combined heat and power generation, P CHP,h (t) represents the heat generation power of the combined heat and power generation.

[0088] The SOC model of the battery is as follows:

[0089] At any time, the battery SOC is:

[0090]

[0091] In the formula, SOC(t) is the state of charge of the battery energy storage at time t; Δt is the time step; E bat is the rated capacity of the battery energy storage; P t is the net power flow at time t (a positive value indicates charging, and a negative value indicates discharging); η ch is the charging efficiency of the battery energy storage; η dis is the discharging efficiency of the battery energy storage.

[0092] The operating constraint conditions of the battery are:

[0093] -P max ≤ P t ≤ P max

[0094] SOC min ≤ SOC t ≤ SOC max

[0095] SOC0 = SOC initial

[0096] SOC T ≥ SOC target

[0097] In the formula, P max represents the maximum allowable value of the battery charge and discharge power, SOC min , SOC max represents the minimum and maximum capacities allowed for the battery SOC to operate, SOC0, SOC initial represents the value of SOC at the initial moment, SOC T represents the value of SOC at the end of the scheduling period, SOC target represents the SOC target value.

[0098] Establish a pumped - storage energy system model:

[0099] The dynamic efficiency model of the pumped - storage energy system is:

[0100]

[0101] In the formula, η pump,t represents the pumping efficiency at time t, η turbine,t represents the power generation efficiency at time t, η pump,0 represents the reference efficiency of the pumping condition, η turbine,0 represents the reference efficiency of the power generation condition, and α, β, γ, and δ all represent efficiency decay coefficients; h t represents the current reservoir water level, h max and h min respectively represent the upper and lower limits of the water level, Ppump,t is the pumping power, Pgen,t is the power generation power, represents the maximum pumping power, represents the maximum power generation power;

[0102] The actual energy storage state of the pumped - storage system is updated as follows:

[0103]

[0104] In the formula, E t represents the actual energy storage capacity state of the pumped - storage during the t - th period, E t-1 represents the actual energy storage capacity state of the pumped - storage during the (t - 1) - th period, and Δt represents the time step.

[0105] Energy storage capacity constraint of the pumped - storage system:

[0106] E min ≤E t ≤E max

[0107] In the formula: E min and E max are the minimum and maximum energies allowed for the pumped - storage system

[0108] Pumping and power generation power limit constraint:

[0109]

[0110] Step 2: Establish a flexible load model; the flexible load model includes a shiftable load model, a curtailable load model, and a shiftable load model.

[0111] The shiftable load is defined as the user can flexibly adjust the energy consumption in each period under the constraint of meeting the total energy consumption unchanged within a scheduling period. The shiftable load has a high operation flexibility, and there is no continuous constraint on the transfer duration and time period. Typical shiftable loads include electric energy storage, thermal energy storage, and battery swapping stations for new energy vehicles.

[0112] Electric energy storage is a typical shiftable electric load:

[0113] P es (t + 1)= ΔtE es (t)+P es (t)(1 - γ es )

[0114] Its state constraint conditions are as follows:

[0115]

[0116] P es (0)= P es (1)

[0117] Where: P es (t), P es (t + 1) are the electric energy before and after the charge and discharge of the electric energy storage respectively; γ es is the self-discharge rate of the electric energy storage; E es (t) is the actual charge and discharge power of the electric energy storage device in the t time period. The electric energy storage charging is positive and discharging is negative; Δt is the unit scheduling duration; are the charge and discharge efficiencies of the electric energy storage respectively; are the maximum values of the charge and discharge powers of the electric energy storage respectively; are the minimum and maximum power values of the battery energy storage; P es (0), P es (1) are the stored electric powers of the battery at the beginning and end of a scheduling cycle respectively.

[0118] The compensation cost of the shiftable load within a scheduling cycle is:

[0119]

[0120] Where φ s represents the unit compensation coefficient of the shiftable load, and ΔP s (t) represents the shiftable load power in the t time period.

[0121] The load that can be curtailed mainly refers to the load that can appropriately curtail the energy consumption according to the changes in people's daily life needs or the corresponding electricity prices.

[0122] ΔP c (t)= f b (p b (t), Δp c (t), λ c (t), μ c (t))

[0123] Where, ΔP c (t) is the magnitude of the load that can be curtailed in the t time period; p b(t) is the basic load in the t period; Δp c (t) is the change in the amount of load that can be curtailed in the t time period; λ c (t) is the self-elasticity coefficient of the load in the t time period; μ c (t) is the curtailment rate based on the user's independent response. f b (t) represents the load curtailment response function.

[0124] Curtailing the load makes it difficult for users to obtain a high energy consumption comfort level. It is necessary to set a user experience constraint, the upper and lower limits of the constraint duration, and the number of curtailments:

[0125]

[0126] In the formula: a c (t) represents the load curtailment status in the t time period; 1 means being curtailed, and 0 means no load curtailment; T c min 、T c max respectively represent the minimum and maximum values of the continuous curtailment duration; Tc is the continuous curtailment duration of the load that can be curtailed; M represents the maximum value of the periods that can be curtailed within a scheduling period. τ represents the initial period of t.

[0127] The compensation cost of the load that can be curtailed within a scheduling period is:

[0128]

[0129] In the formula φ c represents the unit compensation coefficient of the load that can be curtailed.

[0130] Shiftable load:

[0131] ΔP m (t) = f c (t + Δt(Δp)) - f c (t)

[0132] In the formula ΔP m (t) is the magnitude of the shiftable load value in the t time period; f c (t) is the initial energy consumption of the user; Δt(Δp) represents the time shift of the load after the user's independent response; f c (t + Δt(Δp)) is the user's energy consumption curve after adding the shiftable load.

[0133] Constraint conditions:

[0134] T m = [t0, t1 - t s + 1]

[0135] In the formula T mis the set of shiftable start time periods; t0 and t1 are the start time and end time respectively within the shiftable time period in the scheduling cycle; t s represents the duration of the shiftable load.

[0136] The compensation cost of the shiftable load within one scheduling cycle is:

[0137]

[0138] In the formula, φ m represents the unit compensation coefficient of the shiftable load.

[0139] Step 3: Combine the models in Step 1 and Step 2 to establish a multi-scale optimal scheduling model for the integrated energy system considering integrated demand response. The optimization process of the multi-scale optimal scheduling model for the integrated energy system is as Figure 3 shown. The multi-scale optimal scheduling model for the integrated energy system includes a day-ahead scheduling layer and a real-time scheduling layer. The day-ahead scheduling layer is used to output the target energy storage state of pumped storage, the demand response plan, and the equipment output plan based on the load prediction data, the renewable energy output prediction curve, the economic parameters, and the pumped storage dynamic efficiency parameters. The real-time scheduling layer is used to output the pumped storage output, the demand response plan, and the implementation of the electricity purchase plan based on the real-time load and renewable energy output data.

[0140] The objective function of the day-ahead scheduling layer is:

[0141] J ahead = min(C chp + C ph + C s + C res )

[0142] In the formula, J ahead is the day-ahead scheduling cost, C chp is the combined heat and power cost, C ph is the pumped storage operation cost, C s is the cost of shiftable load, C res is the renewable energy cost;

[0143] The formula for the combined heat and power cost is as follows:

[0144]

[0145] In the formula, t is the time period, λ gas is the natural gas price, P CHP,g (t) is the gas input power of the combined heat and power equipment;

[0146] The formula for the pumped storage operation cost is as follows:

[0147] Cph = C water + C wear

[0148]

[0149] In the formula, C water is the total water consumption cost, C wear is the total equipment maintenance cost, c water represents the unit water consumption cost, V pump,t represents the water extraction volume at time t, V gen,t represents the water consumption during power generation at time t, c wear represents the unit equipment maintenance cost, P pump,t is the pumping power, P gen,t is the power generation power;

[0150] The formula for the cost of the shiftable load is as follows:

[0151]

[0152] In the formula, φ s represents the unit compensation coefficient of the shiftable load, ΔP s (t) represents the shiftable load power at time t, and T is the scheduling period;

[0153] The formula for the cost of renewable energy is as follows:

[0154]

[0155] In the formula, λ curt represents the cost of wind and light curtailment, represents the maximum value of wind energy, P wind (t) represents the wind energy magnitude at time t, represents the maximum value of light energy, P solar (t) represents the light energy magnitude at time t.

[0156] The objective function of the real-time scheduling layer is:

[0157] J inday = min(C grid + C c + C m + C store + ρ(t)·|E t - E target |)

[0158] In the formula, J inday represents the real-time scheduling cost, C grid represents the cost of purchasing electricity from the power grid, C c represents the cost of the load that can be curtailed, C m represents the compensation cost of the load that can be shiftedstore Denote the energy storage cost, ρ(t) denote the dynamic penalty factor, and E t denote the actual energy storage state of pumped-storage at time t, and E target denote the target energy storage state at the day-ahead scheduling layer; the dynamic penalty factor ρ(t) is used to correct the day-ahead plan deviation and coordinate the energy storage state and real-time economy. The expression of the dynamic penalty factor is as follows:

[0159]

[0160] In the formula, ρ0 denotes the initial penalty factor, and E t denote the actual energy storage capacity state of pumped-storage at time t, and E target denote the target energy storage state at the day-ahead scheduling layer.

[0161] The formula for the grid power purchase cost is as follows:

[0162]

[0163] In the formula, λ grid (t) denotes the grid power purchase unit price, and P grid (t) denotes the grid power purchase quantity;

[0164] The formula for the cost of the load that can be curtailed is as follows:

[0165]

[0166] In the formula, φ c denotes the unit compensation coefficient of the load that can be curtailed, T denotes the scheduling period, and ΔP c (t) is the magnitude of the load that can be curtailed in the t time period;

[0167] The formula for the compensation cost of the load that can be shifted is as follows:

[0168]

[0169] In the formula, φ m denotes the unit compensation coefficient of the load that can be shifted, and ΔP m (t) is the magnitude of the load that can be shifted in the t time period;

[0170] The formula for the energy storage cost is as follows:

[0171]

[0172] In the formula, λ store denotes the charge-discharge loss cost coefficient, and P charge (t) denotes the battery charging power, and P discharge (t) denotes the battery discharging power.

[0173] Step 4: Substitute the predicted data of electricity, heat, and gas loads into the objective functions and constraint conditions of the day-ahead scheduling layer and the real-time scheduling layer. The constraint conditions include the operation constraints of wind turbines, the operation constraints of photovoltaic power generation, the operation constraints of energy conversion equipment, the operation constraints of energy storage batteries, the state update constraints of pumped-storage devices, the state constraints of flexible loads, and the user experience constraints. Use a distributed optimization algorithm to solve the objective function to minimize the system operation cost and obtain the equipment output plan.

[0174] The operation constraints of wind turbines are as follows:

[0175]

[0176] In the formula, P W is the output power of the wind turbine, P rw is the rated output power of the wind turbine, U is the wind speed, U cin , U cout , U r are the cut-in wind speed, cut-out wind speed, and rated wind speed respectively.

[0177] The operation constraints of photovoltaic power generation can be expressed by the following formula:

[0178] P t PV = G t · A PV · η PV

[0179] In the formula, P t PV is the photovoltaic power generation power at time period t, G t is the solar radiation intensity at time period t (W / m 2 ), A PV is the total area of the photovoltaic array (m 2 ), η PV is the conversion efficiency of the photovoltaic module.

[0180] The operation constraints of energy conversion equipment include the operation constraints of absorption chillers and the operation constraints of combined heat and power equipment;

[0181] The operation constraints of absorption chillers are:

[0182] P EC,t = E EC,t · η EC

[0183] In the formula: E EC,t , P EC,t respectively represent the input and output powers of the electric chiller; the refrigeration efficiency is expressed as η EC .

[0184] The operation constraints of combined heat and power equipment are:

[0185]

[0186] Wherein: P g,CHP (t) represents the gas input power, represents the power generation efficiency of the combined heat and power generation, P CHP,e (t) represents the power generation power of the combined heat and power generation system, represents the heat generation efficiency of the combined heat and power generation, P CHP,h (t) represents the heat generation power of the combined heat and power generation.

[0187] The operating constraints of the battery are:

[0188] -P max ≤P t ≤P max

[0189] SOC min ≤SOC t ≤SOC max

[0190] SOC0 = SOC initial

[0191] SOC T ≥SOC target

[0192] Wherein, P max represents the maximum allowable value of the battery charge and discharge power, SOC min 、SOC max represent the minimum and maximum capacities allowed for the battery SOC to operate, SOC0, SOC initial represent the value of SOC at the initial moment, SOC T represents the value of SOC at the end of the scheduling period, SOC target represents the SOC target value.

[0193] The state update constraints of the pumped-storage device are as follows:

[0194]

[0195] Wherein, E t represents the actual energy storage capacity state of the pumped-storage at time t, E t-1 represents the actual energy storage capacity state of the pumped-storage at time t-1, Δt represents the time step. η pump,t represents the pumping efficiency at time t, Ppump,t is the pumping power, Pgen,t is the power generation power, η turbine,t represents the power generation efficiency at time t.

[0196] Flexible load status constraints, including transferable load status constraints and shiftable load status constraints.

[0197] The transferable load status constraint is:

[0198]

[0199] P es (0) = P es (1)

[0200] In the formula: P es (t), P es (t + 1) are the electric energy before and after the charge and discharge of the electrical energy storage respectively; γ es is the self-discharge rate of the electrical energy storage; E es (t) is the actual charge and discharge power of the electrical energy storage device in the t time period, positive for charging and negative for discharging; Δt is the unit scheduling duration; are the charge and discharge efficiencies of the electrical energy storage respectively; are the maximum values of the charge and discharge powers of the electrical energy storage respectively; are the minimum and maximum power values of the battery energy storage; P es (0), P es (1) are the storage electric powers of the battery at the beginning and end of a scheduling cycle respectively.

[0201] The shiftable load status constraint is:

[0202] T m = [t0, t1 - t s + 1]

[0203] In the formula, T m is the set of shiftable start time periods; t0 and t1 are the start time and end time of the shiftable time period within the scheduling cycle respectively; t s represents the duration of the shiftable load.

[0204] The user experience degree constraint is used to constrain the upper and lower limits of the duration and the number of curtailments. The expression is:

[0205]

[0206] In the formula: a c (t) represents the load curtailment status in the t time period; 1 represents being curtailed, and 0 represents no load curtailment; T c min 、T c max represent the minimum and maximum values of the continuous curtailment duration respectively; Tc is the continuous curtailment duration of the curtailable load; M represents the maximum value of the curtailable time periods within a scheduling cycle. τ represents the initial time period of t.

[0207] The distributed optimization algorithm in this embodiment is specifically the Alternating Direction Method of Multipliers (ADMM). Figure 4 As shown in the ADMM algorithm flowchart, the specific process of using the Alternating Direction Method of Multipliers to solve the objective function is as follows:

[0208] Step 4.1: Set the initial local scheduling decision variables and corresponding Lagrange multipliers for each subsystem; in the day-ahead scheduling layer, initialize the output plans of each subsystem, and the output plans include the target state of pumped-storage energy and the demand response plan; in the real-time scheduling layer, initialize the real-time coordination variables, and the real-time coordination variables include the grid power purchase and the actual state of energy storage.

[0209] Step 4.2: Under the condition of fixing the global coordination variables and Lagrange multipliers, the power, heat, and flexible load subsystems solve their respective optimization problems in parallel and update the local scheduling decision variables; the formula is as follows:

[0210]

[0211] In the formula, x k+1 represents the local scheduling decision variable at the (k + 1)-th iteration, f(x) represents the objective function of the electrical and thermal subsystem problem, ρ is a constant penalty factor, A and B are linear transformation matrices representing the coupling relationship between each subsystem, x represents the local scheduling decision variable, y k represents the global coordination variable at the k-th iteration, c is a constant vector, and z k is the Lagrange multiplier at the k-th iteration;

[0212] Step 4.3: Update the global coordination variables by fusing the solution results of all subsystems; the formula is as follows:

[0213]

[0214] In the formula, y k+1 represents the global coordination variable at the (k + 1)-th iteration, g(y) represents the objective function of the flexible load subsystem problem, and y represents the global coordination variable. ρ is a constant penalty factor, A and B are linear transformation matrices representing the coupling relationship between each subsystem, x k+1 represents the local scheduling decision variable at the (k + 1)-th iteration, c is a constant vector, and z k is the Lagrange multiplier at the k-th iteration;

[0215] Step 4.4: Adjust the Lagrange multiplier according to the real-time operation data (such as fluctuations in wind and light output and load mutations), and dynamically correct the penalty factor;

[0216] z k+1 = z k + ρ(Ax k+1 ]>+ Byk+1 -c)

[0217] where z k+1 is the Lagrange multiplier of the (k + 1)-th iteration, and z k is the Lagrange multiplier of the k-th iteration.

[0218] Step 4.5: Judge the convergence condition. When the deviation between the output plan of each subsystem and the real-time adjustment is less than the preset threshold, stop the iteration and output the final global optimal scheduling scheme; otherwise, increment the iteration count by one and return to Step 4.2.

[0219] The convergence condition is expressed by the following formula:

[0220]

[0221] where r k+1 , s k+1 are the original residual and the dual residual calculated after the (k + 1)-th iteration respectively, ε pri is the set original residual tolerance threshold, and ε dual is the set dual residual tolerance threshold, and the superscript T represents transpose.

[0222] Utilizing the distributed solution characteristics of ADMM not only accelerates the solution speed of large-scale systems, but also reduces the computational burden of the central coordinator, ensuring the economy, flexibility, and robustness of system operation.

[0223] Analysis of algorithm convergence and real-time performance:

[0224] To ensure the effectiveness of the algorithm in practical applications, the present invention conducts a theoretical proof of the convergence of ADMM and improves the system response speed through parameter adjustment (such as penalty factor, step size control, etc.). This algorithm can achieve global optimal scheduling in a short time and meet the requirements of real-time scheduling for complex multi-energy systems.

[0225] On the time scale, the optimized target state and equipment output plan of pumped storage are adopted at the current 24-hour scheduling layer, combined with the 15-minute real-time scheduling layer dynamic adjustment strategy. The input data includes the baseline prediction of electricity, gas, and heat loads, the output curves of wind power and photovoltaic power, time-of-use electricity prices, and equipment dynamic efficiency parameters. A distributed optimization algorithm is used for solution to verify the effectiveness of this optimized scheduling method in reducing system costs, improving the utilization rate of renewable energy, and optimizing load regulation.

[0226] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed in this application can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0227] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the concept of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should be regarded as within the protection scope of the present invention.

Claims

1. An optimal operation method for an integrated energy system considering pumped storage energy storage, characterized in that It includes the following steps: Step 1: Establish models for renewable energy, energy conversion equipment, and energy storage equipment; Step 2: Establish a flexible load model; Step 3: Combine the models in Step 1 and Step 2 to establish a multi-scale optimal scheduling model for an integrated energy system considering integrated demand response. The multi-scale optimal scheduling model for the integrated energy system includes a day-ahead scheduling layer and a real-time scheduling layer. The day-ahead scheduling layer is used to output the target energy storage state of pumped-storage, the demand response plan, and the equipment output plan based on load forecast data, renewable energy output forecast curves, economic parameters, and pumped-storage dynamic efficiency parameters. The real-time scheduling layer is used to output the pumped-storage output, the demand response plan, and the implementation of the electricity purchase plan based on real-time load and renewable energy output data; Step 4: Substitute the electricity, heat, and gas load forecast data into the objective function and constraint conditions of the day-ahead scheduling layer and the real-time scheduling layer, and use a distributed optimization algorithm to solve the objective function to minimize the system operation cost and obtain the equipment output plan.

2. The optimized operation method of the integrated energy system considering pumped storage as claimed in claim 1, wherein In Step 1, the renewable energy includes wind power and photovoltaic power, the energy conversion equipment includes absorption chillers and combined heat and power equipment, and the energy storage equipment includes batteries and pumped-storage systems.

3. The optimal operation method of the integrated energy system considering pumped-storage energy as claimed in claim 1, wherein In Step 2, the flexible load model includes a shiftable load model, a curtailable load model, and a shiftable load model.

4. The optimal operation method of the integrated energy system considering pumped storage energy as claimed in claim 1, wherein, In Step 3, the objective function of the day-ahead scheduling layer is: J ahead = min(C chp + C ph + C s + C res ) Where, J ahead is the day-ahead scheduling cost, C chp is the combined heat and power cost, C ph is the operating cost of pumped storage, C s is the transferable load cost, C res for renewable energy costs; The formula for the combined heat and power cost is as follows: where t is the time period, and λ gas is the natural gas price, and P CHP,g (t) is the gas input power of the combined heat and power equipment; The formula for the pumped-storage operation cost is as follows: C ph = C water + C wear Wherein, C water is the total water consumption cost, C wear is the total equipment maintenance cost, c water represents the unit water consumption cost, V pump,t represents the water pumping volume at time t, V gen,t represents the water consumption during power generation at time t, c wear represents the unit equipment maintenance cost, P pump,t is the pumping power, P gen,t is the power generation power; The formula for the shiftable load cost is as follows: where φ s represents the unit compensation coefficient of the transferable load, and ΔP s (t) represents the transferable load power at time t, and T is the scheduling period; The formula for the renewable energy cost is as follows: where λ curt represents the cost of curtailment of wind and solar power, represents the maximum value of wind energy, and P wind (t) represents the magnitude of wind energy at time t, represents the maximum value of solar energy, and P solar (t) represents the magnitude of solar energy at time t.

5. The optimal operation method of the integrated energy system considering pumped-storage energy storage according to claim 1, wherein In Step 3, the objective function of the real-time scheduling layer is: J inday = min(C grid + C c + C m + C store + ρ(t)·|E t - E target |) Where, J inday represents the real-time scheduling cost, C grid represents the power purchase cost from the power grid, C c represents the cost of load curtailment, C m represents the compensation cost of shiftable load, C store represents the energy storage cost, ρ(t) represents the dynamic penalty factor, E t represents the actual energy storage state of pumped storage at time t, E target represents the target energy storage state of the day-ahead scheduling layer; The formula for the grid electricity purchase cost is as follows: where λ grid (t) represents the unit price of electricity purchased from the power grid, and P grid (t) represents the electricity quantity purchased from the power grid; The formula for the curtailable load cost is as follows: where φ c represents the unit compensation coefficient of the load that can be curtailed, T represents the scheduling period, and ΔP c (t) is the magnitude of the load that can be curtailed in the time period t; The formula for the compensation cost of the shiftable load is as follows: where φ m represents the unit compensation coefficient of the shiftable load, and ΔP m (t) is the magnitude of the shiftable load value in the time period t; The formula for the energy storage cost is as follows: Where, λ store represents the charge and discharge loss cost coefficient, P charge (t) represents the battery charging power, P discharge (t) represents the battery discharging power.

6. The optimal operation method of the integrated energy system considering pumped storage as claimed in claim 5, characterized in that The dynamic penalty factor is used to correct the deviation of the day-ahead plan and coordinate the energy storage state and real-time economy. The expression of the dynamic penalty factor is as follows: where ρ0 represents the initial penalty factor, and E t represents the actual energy storage capacity state of pumped storage at time t, and E target represents the target energy storage state of the day-ahead dispatch layer.

7. The optimized operation method of the integrated energy system considering pumped-storage energy as claimed in claim 1, wherein The constraint conditions include wind turbine operation constraints, photovoltaic power generation operation constraints, energy conversion equipment operation constraints, battery operation constraints, pumped-storage device state update constraints, flexible load state constraints, and user experience constraints.

8. The optimized operation method of the integrated energy system considering pumped storage energy as claimed in claim 2, wherein The dynamic efficiency model of the pumped-storage system is: where η pump,t represents the pumping efficiency at time t, η turbine,t represents the power generation efficiency at time t, η pump,0 represents the reference efficiency under the pumping condition, η turbine,0 represents the reference efficiency under the power generation condition, and α, β, γ, and δ all represent efficiency decay coefficients; h t represents the current reservoir water level, h max and h min represent the upper and lower limits of the water level respectively, Ppump,t is the pumping power, Pgen,t is the power generation power, represents the maximum pumping power, represents the maximum power generation power; The actual energy storage state of the pumped-storage system is updated as follows: where E t represents the actual energy storage capacity state of pumped-storage at time t, and E t-1 represents the actual energy storage capacity state of pumped-storage at time t - 1, and Δt represents the time step.

9. The optimal operation method of the integrated energy system considering pumped-storage energy storage according to claim 1, wherein In Step 4, the distributed optimization algorithm is specifically the alternating direction method of multipliers. The specific process of using the alternating direction method of multipliers to solve the objective function is: Step 4.1: Set initial local scheduling decision variables and corresponding Lagrange multipliers for each subsystem; in the day-ahead scheduling layer, initialize the output plan of each subsystem, and the output plan includes the target state of pumped-storage and the demand response plan; In the real-time scheduling layer, initialize the real-time coordination variables, and the real-time coordination variables include the grid electricity purchase power and the actual energy storage state; Step 4.2: Under the condition of fixing the global coordination variables and Lagrange multipliers, the power, heat, and flexible load subsystems solve their respective optimization problems in parallel and update the local scheduling decision variables; Step 4.3: Update the global coordination variables by fusing the solution results of all subsystems. Step 4.4: Adjust the Lagrange multipliers according to the real-time operation data and dynamically correct the penalty factors. Step 4.5: Judge the convergence condition. When the deviation between the output plan of each subsystem and the real-time adjustment is less than the preset threshold, stop the iteration and output the final global optimal scheduling scheme; otherwise, increment the iteration count by one and return to Step 4.2.

Images