Comprehensive energy supply system multi-time scale dynamic scheduling method considering hysteresis effect of energy supply equipment

Abstract

The invention discloses an integrated energy supply system multi-time scale dynamic scheduling method considering a hysteresis effect of energy supply equipment. Aiming at the ultra-short-term scheduling optimization problem of the comprehensive energy supply system, environment, economy and feasibility are considered, and a multi-time-scale dynamic optimization model is established. According to the method, flexible scheduling of the energy supply system under different time granularities is realized by considering multiple time scales and equipment response differences. The multi-time-scale dynamic optimization model can dynamically adjust a scheduling strategy according to changes of equipment characteristics, demand response and energy output, and ensures high efficiency and sustainability of energy supply. Compared with a traditional single time scale scheduling method, the method can deal with supply and demand changes and equipment capacity limitation more finely, and achieves high efficiency and feasibility of system scheduling.

Description

Technical Field

[0001] The present invention relates to the field of design, planning and optimal operation of integrated energy systems. Specifically, it relates to a multi-time scale dynamic scheduling method for an integrated energy supply system considering the hysteresis effect of energy supply equipment, aiming to solve the ultra-short-term scheduling problem. Background Art

[0002] With the continuous development of the global economy and the increasing awareness of environmental protection, the demand for clean energy shows a significant growth trend. However, the large-scale access of renewable energy sources such as wind energy and solar energy poses unprecedented challenges to the traditional single energy supply system due to their inherent intermittency and uncertainty. These challenges specifically include: the intensification of dynamic imbalance between energy supply and demand, the urgent need to improve energy conversion efficiency, and the increasingly prominent problems of environmental pollution and climate change caused by fossil fuel consumption.

[0003] To address the above challenges, the integrated energy supply system has emerged as an integrated solution and developed rapidly. By integrating various energy forms such as electricity, heat, and natural gas and relying on the coordinated operation of cross-energy networks, this system not only significantly improves energy utilization efficiency but also effectively promotes the transformation of the energy structure towards cleaner and more sustainable directions. In the integrated energy supply system, multi-time scale scheduling, as a core link, is crucial for the rational allocation and optimal control of resources within the system. Multi-time scale scheduling refers to predicting, planning, and controlling the operating state of the system according to different time spans.

[0004] Currently, the multi-time scale scheduling of integrated energy supply systems is experiencing a trend of evolving from static to dynamic, from single-objective to multi-objective, and from centralized to distributed. Most existing scheduling strategies are based on a hierarchical architecture, following the basic framework of "day-ahead - intra-day - real-time", and using a step-by-step refinement method to address prediction uncertainty and operation volatility. However, the existing methods still have deficiencies in the following aspects: Existing methods usually fail to accurately model and handle the inherent time lag effects of different types of energy, such as the dynamic response characteristics of heat energy and natural gas. This simplified treatment may lead to a reduction in prediction accuracy, thereby affecting the effectiveness of scheduling decisions. Although there has been an increased focus on demand-side management, there is still room for improvement in effectively tapping the adjustment potential at the user end, especially in using means such as prices and incentives to guide users to participate in system operation optimization and improving the system's adaptive ability to demand changes. Most existing methods tend to adopt fixed scheduling time intervals and fail to dynamically adjust the time resolution according to the dynamic differences of different energy characteristics. This choice of fixed time granularity ignores the differences in the response speeds of different energies and may limit the effectiveness and implementation effect of scheduling strategies.

[0005] In view of the above analysis, in order to further improve the operating performance of the integrated energy supply system, it is urgent to explore new dispatching theories and technical means, especially to make breakthroughs in aspects such as multi-time scale coordination, accurate modeling of energy dynamic response, and design of advanced demand response mechanisms. Summary of the Invention

[0006] The object of the present invention is to overcome the deficiencies of the prior art and provide a multi-time scale dynamic dispatching method for an integrated energy supply system considering the hysteresis effect of energy supply equipment. Aiming at the ultra-short-term dispatching optimization problem of the integrated energy supply system, considering environmental benefits, economic benefits, and operational feasibility, a multi-time scale dynamic optimization model is constructed. The dispatching time window of this model can be set to one hour, and this hour is divided into finer time granularities for optimization. The model input includes the predicted multi-energy demand sequence for the next hour, and this demand sequence can be dynamically adjusted according to price, incentive mechanisms, and the inertia of the energy flow network. The model output includes the load distribution plan of each device within the system for the next hour. The optimization goal of the present invention is to maximize the overall economic benefits of the system on the premise of fully considering environmental benefits.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A multi-time scale dynamic dispatching method for an integrated energy supply system considering the hysteresis effect of energy supply equipment, comprising the following steps:

[0009] Step S1, construction of the energy supply equipment model; determine the energy supply equipment and its input and output variables, identify the hysteresis parameters of the energy supply equipment output variables relative to the input variables, and construct the energy supply equipment model based on the identified hysteresis parameters;

[0010] Step S2: Construction of the user-side demand response model and the energy flow network inertia model; for rigid demand loads and flexible demand loads, perform elastic modeling on the user-side demand to obtain the user-side demand response model; model the pipeline storage capacity and pressure regulation flexibility existing in the longitudinal space of the energy flow network, as well as the flow inertia of the energy medium and the flow velocity differences of different media existing in the transverse space to obtain the energy flow network inertia model;

[0011] Step S3: Construction of a multi-time scale dynamic optimization model; aiming at the overall economic benefits of the integrated energy supply system, model the multi-time scale dynamic optimization model, and use the user-side demand response model and the energy flow network inertia model as the constraints of the multi-time scale dynamic optimization model;

[0012] Step S4: Dynamic scheduling optimization of the integrated energy supply system; based on the energy supply equipment model, considering the multi-time scale response differences of different equipment, select multiple minimum time granularities; at different minimum time granularities, solve the multi-time scale dynamic optimization model to obtain an optimized scheduling plan, and schedule the integrated energy supply system according to the obtained optimized scheduling plan. The multi-time scale dynamic optimization model calculates the optimized scheduling plan for the next time period based on the actual operation data fed back by the integrated energy supply system; based on the calculation efficiency and model prediction accuracy, determine the best minimum time granularity, and perform scheduling optimization based on the best minimum time granularity.

[0013] Further, in the above-mentioned step S1:

[0014] Adopt a data-driven method to identify the lag parameters of various energy flows in different energy supply equipment based on correlation analysis. Specifically, the input is the historical data E of energy a in the same time period a , the historical data E of energy b b and the maximum lag duration L max , and the output is the lag duration τ of energy a relative to energy b. Based on the identified lag parameters, construct a mathematical model (i.e., the energy supply equipment model) that can reflect the dynamic response characteristics of the equipment.

[0015] The specific steps are as follows:

[0016] Step S11, determine the energy supply equipment and its input and output variables and collect data, including the following steps:

[0017] Step S111, determine the energy supply equipment that needs to consider lag parameters. The specific types of energy supply equipment include combined heat and power units, carbon capture equipment, power-to-gas equipment, electric boilers, and electric chillers;

[0018] Step S112, determine all input and output variables of the energy supply equipment. These variables represent various input variables entering the equipment and various output variables leaving the equipment. Taking a combined heat and power unit as an example, the input variables include coal feeding flow rate, steam turbine inlet valve opening, and medium-pressure extraction valve opening, and the output variables include low-pressure steam flow rate, medium-pressure steam flow rate, and electric power; collect the operation data of each input and output variable of each energy supply equipment in the same time period to ensure that all collected operation data are strictly aligned in time.

[0019] Step S12, use the fast Fourier transform to identify the lag parameters of the output variables of the energy supply equipment relative to the input variables. The specific steps are as follows:

[0020] Step S121, select the historical data of a certain output variable a and a certain input variable b of the same energy supply equipment in the same time period from the collected operation data, and normalize the two to obtain Ea and E b ;

[0021] Step S122, respectively apply the fast Fourier transform (FFT) to E a and E b to transform it from the time domain to the frequency domain, obtaining the input variable spectrum and the output variable spectrum, specifically expressed as:

[0022]

[0023]

[0024] where f represents frequency;

[0025] Step S123, cross-spectrum calculation, calculate the product of the output variable spectrum and the complex conjugate of the input variable spectrum to obtain the cross-spectrum S ab (f):

[0026]

[0027] where conj(.) represents the complex conjugate operation;

[0028] Step S124, apply the inverse fast Fourier transform (IFFT) to the calculated cross-spectrum S ab (f) to transform it back from the frequency domain to the time domain, obtaining the cross-correlation function R ab (τ) of the output variable and the input variable:

[0029] R ab (τ) = IFFT(S ab (f))

[0030] where τ represents the lag time;

[0031] Step S125, lag parameter identification, analyze the cross-correlation function R ab (τ), and identify the τ corresponding to the position where its maximum value is located max . This position τ max is the lag parameter of the historical data of the output variable a relative to the historical data of the input variable b.

[0032] Step S13, the energy supply equipment model includes a combined heat and power unit dynamic model, a carbon capture equipment dynamic model, a power-to-gas equipment dynamic model, an electric boiler equipment dynamic model, and an electric chiller equipment dynamic model. The construction of the energy supply equipment model specifically includes the following steps:

[0033] Step S131, construct a dynamic model of a combined heat and power unit. For a combined heat and power unit, the input variables include the coal feeding flow rate, the opening degree of the steam inlet valve of the steam turbine, and the opening degree of the medium-pressure extraction steam valve; the output variables include the low-pressure steam flow rate, the medium-pressure steam flow rate, and the electric power. Considering the time inertia in the steam generation and transportation process, construct the following dynamic model of the combined heat and power unit:

[0034]

[0035] In the formula, v TP，out1，t , v TP，out0，t , P CHP，t are respectively the low-pressure steam flow rate, the medium-pressure steam flow rate, and the electric power of the combined heat and power equipment at time t; v TP，out1，t+1 , v TP，out0，t+1 , P CHP，t+1 are respectively the low-pressure steam flow rate, the medium-pressure steam flow rate, and the electric power of the combined heat and power equipment at time t + 1; are respectively the coal consumption flow rate, the opening degree of the steam inlet valve of the steam turbine, and the opening degree of the medium-pressure extraction steam valve of the combined heat and power equipment at time t - τ 1 ; τ 1 is the time inertia parameter between steam and coal consumption; A CHP0，0 is the influence coefficient of the low-pressure steam flow rate at time t on the low-pressure steam flow rate at time t + 1; A CHP0，1 is the influence coefficient of the medium-pressure steam flow rate at time t on the low-pressure steam flow rate at time t + 1; A CHP0，2 is the influence coefficient of the electric power at time t on the low-pressure steam flow rate at time t + 1; A CHP1，0 is the influence coefficient of the low-pressure steam flow rate at time t on the medium-pressure steam flow rate at time t + 1; A CHP1，1 is the influence coefficient of the medium-pressure steam flow rate at time t on the medium-pressure steam flow rate at time t + 1; A CHP1，2 is the influence coefficient of the electric power at time t on the medium-pressure steam flow rate at time t + 1; A CHP2，0 is the influence coefficient of the low-pressure steam flow rate at time t on the electric power at time t + 1; A CHP2，1 is the influence coefficient of the medium-pressure steam flow rate at time t on the electric power at time t + 1; A CHP2，2 is the influence coefficient of the electric power at time t on the electric power at time t + 1; B CHP0，0 is the influence coefficient characterizing the coal feeding flow rate at time t - τ 1 on the low-pressure steam flow rate at time t + 1; B CHP0，1 is the influence coefficient of the opening degree of the steam inlet valve of the steam turbine at time t - τ 1 on the low-pressure steam flow rate at time t + 1; B CHP0，2 is the influence coefficient of the opening degree of the medium-pressure extraction steam valve at time t - τ 1 on the low-pressure steam flow rate at time t + 1; B CHP1，0 is at time t - τ1 The influence coefficient of the coal feeding flow rate at time t on the medium-pressure steam flow rate at time t + 1; B CHP1，1 is t - τ 1 The influence coefficient of the steam turbine inlet valve opening at time t on the medium-pressure steam flow rate at time t + 1; B CHP1，2 is t - τ 1 The influence coefficient of the medium-pressure extraction valve opening at time t on the medium-pressure steam flow rate at time t + 1; B CHP2，0 is t - τ 1 The influence coefficient of the coal feeding flow rate at time t on the electric power at time t + 1; B CHP2，1 is t - τ 1 The influence coefficient of the steam turbine inlet valve opening at time t on the electric power at time t + 1; B CHP2，2 is t - τ 1 The influence coefficient of the medium-pressure extraction valve opening at time t on the electric power at time t + 1; C CHPy，0 are the model constants for the cogeneration unit to produce different output variables. Here, y is 0, 1, 2, corresponding to the constant terms of the low-pressure steam flow rate, medium-pressure steam flow rate, and electric power models respectively. The above time inertia parameters, production and energy consumption model coefficients, and model constants of the cogeneration unit are all obtained by the data identification method.

[0036] Step S132, construct the dynamic model of the carbon capture equipment: For the carbon capture equipment, its input variable is the electric power, and the output variable is the captured carbon dioxide flow rate. The dynamic model is as follows:

[0037]

[0038] In the formula, v ccus，out0，t is the carbon dioxide capture flow rate of the CCUS (carbon capture) equipment at time t; v ccus，out0，t+1 is the carbon dioxide capture flow rate of the CCUS equipment at time t + 1; are the electric powers of the CCUS equipment at t - τ 2 time respectively; τ 2 is the time inertia parameter between the carbon dioxide capture flow rate and the power consumption; A ccus0，0 is the model coefficient for the CCUS equipment to capture carbon dioxide, C ccus0，0 is the model constant for the CCUS equipment to capture carbon dioxide; B ccus0，2 is the energy consumption model coefficient for the CCUS equipment to capture carbon dioxide.

[0039] Step S133, construct the dynamic model of the power-to-gas equipment: For the power-to-gas (P2G) equipment, its input variables are the carbon dioxide flow rate and the electric power, and the output variable is the natural gas flow rate. The model is as follows:

[0040]

[0041] Wherein, P P2G，t is the natural gas flow rate of the P2G device at time t; P P2G，t+1 is the natural gas flow rate of the P2G device at time t + 1; are respectively the carbon dioxide consumption flow rate and the power consumption of the P2G device at time t - τ 3 ; τ 3 is the time inertia parameter between the natural gas flow rate and the power consumption; A P2G0，0 is the model coefficient of the P2G device for producing natural gas, and C P2G0，0 is the model constant of the P2G device for producing natural gas; B P2G0，0 , B P2G0，1 are the energy consumption model coefficients of the P2G device for producing natural gas.

[0042] Step S134, construct a dynamic model of the electric boiler device: For the electric boiler device, its input is the electric power and the output is the hot water flow rate. The model is as follows:

[0043]

[0044] Wherein, P EB，t is the hot water flow rate of the electric boiler device at time t; is the power consumption of the electric boiler device at time t - τ 4 ; τ 4 is the time inertia parameter between the hot water flow rate and the power consumption; A EB0，0 is the model coefficient of the electric boiler device for producing hot water; C EB0，0 is the model constant of the electric boiler device for producing hot water; B EB0，0 is the energy consumption model coefficient of the electric boiler device for producing hot water.

[0045] Step S135, construct a dynamic model of the electric chiller device: For the electric chiller device, its input is the electric power and the output is the refrigerant flow rate. The model is as follows:

[0046]

[0047] Wherein, P EC，t is the refrigerant flow rate of the electric chiller device at time t; is the power consumption of the electric chiller device at time t - τ 5 ; τ 5 is the time inertia parameter between the refrigerant flow rate and the power consumption; A EC0，0 is the model coefficient of the electric chiller device for producing refrigerant; C EC0，0 is the model constant of the electric chiller device for producing refrigerant; B EC0，0 is the energy consumption model coefficient of the electric chiller device for producing refrigerant.

[0048] Furthermore, the said Step S2 includes the following steps:

[0049] Step S21, construct a user-side demand response model: Demand response originally referred to guiding users to change their energy consumption behaviors through price signals. The user side can interact with the energy supply side in various ways. The electricity consumption types of users are divided into three categories, specifically including discrete loads, shiftable loads, and continuously adjustable loads. Model each of the three electricity consumption types respectively, and use rigid demand loads and flexible demand loads as constraints to obtain the user-side demand response model. The specific steps are as follows:

[0050] Step S211, model the discrete load. The adjustment of discrete load means that high-energy-consuming users reduce or increase a certain amount of load within a specified time according to the grid dispatching instructions, and need to maintain a certain stable operation time and cannot be continuously adjusted. It is specifically expressed as follows:

[0051] P ld，t =P ld，t-1 -(1 - x ld，t )×d ld，t

[0052] In the formula, P ld，t is the discrete load value at time t; P ld，t-1 is the discrete load value at time t - 1; x ld，t is whether the discrete load at time t participates in the adjustment, 0 for participation and 1 for non-participation; d ld，t is the downward adjustment amount of the discrete load at time t.

[0053] Step S212, model the shiftable load. The shiftable load means that users can transfer the time point of energy consumption demand to other time points to keep the total load within a certain time range unchanged, thereby reducing the load value at peak times. It is specifically expressed as follows:

[0054] P lt，t =P lt，t - 1 +x lt，t ×d lt，t

[0055] In the formula, P lt，t is the shiftable load value at time t; P lt，t-1 is the shiftable load value at time t - 1; x lt，t is whether there is an adjustment value of other shiftable loads at time t; d lt，t is the load value transferred to time t.

[0056] Step S213, model the continuously adjustable load. The continuously adjustable load means that the user side can continuously and quickly respond to the dispatching instructions issued by the grid in each time period, support adjustment in each time period, and does not need to maintain a stable operation time. It is specifically expressed as follows:

[0057]

[0058] In the formula, P lc，t is the continuously adjustable load value at time t; d lc，n，t is the output value of the nth continuously adjustable unit at time t; N lc is the number of continuously adjustable units.

[0059] Step S214, construct the rigid demand load constraint: For user demands that need to be strictly met, such as medium - pressure and low - pressure steam, the constraint conditions are constructed as shown in the following formula:

[0060]

[0061] In the formula, is the demand value of demand i (specifically referring to the energy type, such as electricity, steam, etc.) at time t p ; t p is the time point of the known predicted demand; L demand，d，lb，i is the maximum downward adjustment amount of demand i; I demand，s，i is a 0 - 1 indicator matrix indicating whether the demand is strictly met, where 1 indicates that the corresponding demand needs to be strictly met; L demand，d，ub，i is the maximum upward adjustment amount of demand i; After selecting the time - interval granularity, the entire scheduling interval T is discretized with each t p as the boundary point, t a is the start time of the discrete interval where t p is located, 0 ≤ t a ≤ t p ≤ T; is the energy supply amount of demand i at time t.

[0062] Step S215, construct the flexible demand load constraint. For non - strictly met demand quantities, such as natural gas, the following constraints are constructed:

[0063]

[0064] Step S22, construct the inertia model of the energy - flow network. Due to the inertia characteristics of fluid energy, it provides a certain degree of flexibility for the operation of the integrated energy supply system, thus helping to smooth out load fluctuations. The energy - flow network has flexibility in pipeline storage capacity and pressure regulation in the longitudinal space, and in the transverse direction, it shows the flow inertia of energy media and the flow - speed differences of different media. The characteristics of the energy - flow network in the longitudinal and transverse directions jointly provide flexibility for operation scheduling.

[0065] Step S221: The present invention regards the fluid energy flow network as a large-scale energy storage unit. During the scheduling process, by utilizing the pressure regulation range of the energy flow network and analyzing the difference between the current steam supply flow rate and the user demand, the energy supply of the equipment is adjusted under the condition of ensuring the safety of the pipe network pressure, so as to achieve flexible response. Due to the existence of the energy flow network, there is an adjustment space for the difference between the inflow and outflow of the fluid energy. Therefore, the supply capacity of the energy flow network can be used to analyze the difference between the current fluid energy supply flow rate and the predicted user demand, and the smoothness of the equipment load can be maintained as much as possible on the premise of ensuring the safety of the pipe network pressure. Model the pipe storage capacity and pressure regulation flexibility existing in the longitudinal space of the energy flow network, which is specifically expressed as:

[0066]

[0067] L vs，lb ≤v smo ≤L vs，ub

[0068] v smo，0 =v r

[0069] In the formula, p smo is the smoothed pressure of the fluid energy pipe network, which needs to be kept within the safe range; p r is the actual pressure of the fluid energy pipe network; v pre is the energy demand at the user end at a future time interval; v smo is the smoothed fluid energy steam supply; θ r is the actual temperature of the fluid energy pipe network; v rated is the volume of the fluid energy pipe network; v smo，0 is v smo 's initial value; v r is the actual steam supply of the fluid energy; L vs，lb is the lower limit of the fluid energy steam supply; L vs，ub is the upper limit of the fluid energy steam supply; M is the molar mass; R is the gas constant; d t is the coefficient measuring the sensitivity of the energy flow network pressure to the supply-demand imbalance.

[0070] Step S222: Considering the flow inertia of the energy medium and the difference in flow velocity, there is an optimal operating range for the pipe network pressure. The present invention further realizes the dynamic calculation of the optimal pressure value of the pipe network by fitting the relationship curve between the energy load and the pipe network pressure. This method can convert the static pipe network pressure range into a dynamic range, so as to more accurately adapt to the energy supply-demand change and make the most of the flexibility of the energy flow network. Model the flow inertia of the energy medium existing in the transverse space of the energy flow network and the difference in flow velocity of different media, which is specifically:

[0071]

[0072] L p，ub = p * + p d

[0073] L p，lb = p * - p d

[0074] L p，lb ≤ p r -(v pre - v smo )×d t ×θ r ×R / M / v rated ≤ L p，ub

[0075] Wherein, p * is the optimal value of the pipe network pressure; g is the energy load and the optimal pressure function of the pipe network; is the average value of the energy demand at the user end under a future time interval; L p，ub is the upper limit of the pipe network pressure; p d is the adjustable range of the optimal pressure; L p，lb is the lower limit of the pipe network pressure for demand i.

[0076] All in all, the flexibility brought by the inertia of the energy flow network is achieved by adjusting v smo and its flexible adjustment range depends on the upper and lower limits of the pipe network pressure, and the dynamic calculation of the optimal value of the pipe network pressure further expands the flexibility interval.

[0077] Furthermore, the construction of the multi-time scale dynamic optimization model described in step S3 includes the following steps:

[0078] Step S31, for the multi-time scale dynamic optimization model of the integrated energy supply system studied in the present invention, its overall economic income includes external sales of steam, external sales of hot water, electricity sales, external sales of natural gas, external sales of carbon emission rights, etc., and the expenditure items include penalties for abandoning wind and light, coal consumption costs, and penalty costs for unmet demands. The goal of the multi-time scale dynamic optimization of the integrated energy supply system is to maximize the overall benefit of the system. Therefore, the objective function of the multi-time scale dynamic optimization model is as follows:

[0079]

[0080] Wherein, E total is the total benefit of the supply system; T is the entire scheduling period; v supply，i，tThe quantity of the externally sold energy \(i\) at time \(t\), where \(i\) takes values of 1, 2, 3, 4, 5, 6, 7, corresponding to low-pressure steam, medium-pressure steam, hot water, refrigerant, carbon dioxide, electricity, and natural gas respectively; The price of the externally sold energy \(i\) at time \(t\); \(v\) punish，j，t The quantity of the penalty item \(j\) at time \(t\), where \(j\) is the abandoned wind and solar energy, coal consumption, and demand non-satisfaction in sequence; The penalty cost coefficient of the penalty item \(j\) at time \(t\).

[0081] Step S32: Establish the constraint conditions of the multi-time-scale dynamic optimization model. The model constraint conditions include the upper and lower limits of equipment load, the minimum load constraint for equipment startup, the equipment operation performance constraint, the demand constraint, the energy flow inertia constraint, the energy storage device capacity constraint, the equipment ramp constraint, the node energy balance constraint, and the initial value constraint:

[0082] Upper and lower limits of equipment load:

[0083] \(L\) in，m，lb \(\leq u\) m，t \(\leq L\) in，m，ub

[0084] \(L\) out，n，lb \(\leq y\) n，t \(\leq L\) out，n，ub

[0085] In the formula, \(L\) in，m，lb is the lower limit of the energy consumption \(m\) of the equipment; \(u\) m，t is the quantity of the energy consumption \(m\) of the equipment at time \(t\); \(L\) in，m，ub is the upper limit of the energy consumption \(m\) of the equipment; \(L\) out，n，lb is the lower limit of the production capacity \(n\) of the equipment; \(y\) n，t is the quantity of the production capacity \(n\) of the equipment at time \(t\); \(L\) out，n，ub is the upper limit of the production capacity \(n\) of the equipment.

[0086] Minimum load constraint when the equipment is started:

[0087] \((u\) m，t \(- 0)(u\) m，t \(- L\) in，m，ub \(\times r\) m \()\geq0\)

[0088] In the formula, \(r\) m is the proportion of the minimum operating load of the energy consumption \(m\) of the equipment when it is started to its maximum load.

[0089] Equipment performance constraint:

[0090] is the energy supply equipment model equation constructed in Step S1.

[0091] Demand constraint:

[0092] It is the user-side demand response model equation constructed in step S2.

[0093] Energy flow inertia constraint:

[0094] It is the energy flow network inertia model equation constructed in step S2;

[0095] Energy storage device capacity constraint:

[0096]

[0097] L sx，k，lb ≤P sx，k，t ≤L sx，k，ub

[0098] In the formula, P sx，k，t+1 is the stored energy of energy storage device k at time t + 1; P sx，in，k，t is the charging energy of energy storage device k at time t; η sx，in，k is the charging efficiency of energy storage device k; P sx，out，k，t is the discharging energy of energy storage device k at time t; η sx，out is the discharging power of energy storage device; L sx，k，lb is the lower limit of the stored energy of energy storage device k; L sx，k，ub is the upper limit of the stored energy of energy storage device k.

[0099] Device ramp-up constraint:

[0100] |u m，t+1 -u m，t |≤L demand，d，ub，m

[0101] In the formula, L demand，d，ub，m is the upper limit of the ramp-up ability of device energy consumption m.

[0102] Node in-out energy conversion constraint:

[0103] The energy flow branch node needs to satisfy the constraint that the input energy is equal to the output energy.

[0104] Initial value constraint:

[0105] P sx，k，t =P sx，k，init

[0106] u m，t =u m，init

[0107] y n，0 =y n，init

[0108] In the formula, P sx，k，init is the initial value of the stored energy of energy storage device k; u m，init is the initial load value of device energy consumption m; yn，init is the initial production capacity for equipment production capacity n.

[0109] The above multi-time scale dynamic optimization model is an optimization model with quadratic constraints and can be solved using the interior point method.

[0110] Furthermore, in step S4:

[0111] The optimization scheduling time is divided into four sub-periods, and each sub-period is discretized into different minimum time granularities; during online dynamic optimization, the input parameter is the demand forecast value within the next T time; taking T / 4 as the time interval unit, the input parameter is the demand forecast matrix v composed of the demand forecast values at the end times of the four sub-periods. demand ; According to the demand forecast matrix and the multi-time scale dynamic optimization model, calculate the optimization scheduling scheme at different minimum time granularities, and select the best minimum time granularity according to the calculation efficiency and model prediction accuracy;

[0112] Use the multi-time scale dynamic optimization model to calculate the optimization scheduling scheme of the integrated energy supply system within the next T time, as follows:

[0113] P sx , u, y = f(P sx，k，init , u m，init , y n，init , v demand )

[0114] In the formula, f is the constructed multi-time scale dynamic optimization model; P sx , u, y are the optimization scheduling schemes; P sx is the scheduling matrix of the stored energy of all energy storage devices within the next T time with the minimum time granularity as the scheduling period, used to guide the charging and discharging behaviors of all energy storage devices in each scheduling period; u is the scheduling matrix of the energy consumption of all devices within the next T time with the minimum time granularity as the scheduling period, used to specify the energy consumption plans of all energy supply devices in each scheduling period; y is the scheduling matrix of the production capacity of all devices within the next T time with the minimum scheduling time as the scheduling period, used to specify the production capacity plans of all energy supply devices in each scheduling period.

[0115] Schedule the integrated energy supply system according to the obtained optimization scheduling scheme. The specific method is as follows:

[0116] 1) Send the scheduling scheme of the first minimum time granularity of the first sub-period to the integrated energy supply system for execution; subsequently, collect the actual execution data of the integrated energy supply system and feedback it to the multi-time scale dynamic optimization model to update the scheduling scheme of the second time granularity of the first sub-period; and so on until the first sub-period is executed;

[0117] 2) When the first sub - cycle is completed, the multi - time - scale dynamic optimization model will receive the latest demand prediction matrix and recalculate the optimized scheduling schemes for the new four sub - cycles based on the operations in step 1).

[0118] 3) The operations in step 2) are carried out in a loop. Each time, only the scheme of the first sub - cycle is adjusted and executed until all sub - cycles within the future T time are completed.

[0119] Different from the traditional scheduling optimization method, that is, the scheduling decision at time t corresponds to the demand at time t, the method of the present invention takes into account the difference in equipment response time. That is, the input of the equipment at time t may correspond to the output at time t + k, and the output at time t + k does not strictly meet the demand at time t + k, leaving room for flexibility. The difference in time scale leads to the "dynamic" performance of decision optimization.

[0120] The present invention adopts a dynamic optimization method to ensure meeting the constraints of cumulative demand. The operation plan for the first sub - cycle in the future is updated according to the actual execution effect. The scheduling scheme has a smaller time granularity and can handle the changes in supply and demand and equipment capacity limitations more precisely.

[0121] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0122] The multi - time - scale dynamic scheduling method for an integrated energy supply system by considering the hysteresis effect of energy - supply equipment has significant technical advantages and benefits compared with the traditional energy scheduling method.

[0123] (1) The present invention innovatively adopts a data - driven method based on the fast Fourier transform (FFT) to identify the hysteresis parameters of energy - supply equipment and incorporates these parameters into the energy - supply equipment model, significantly improving the accuracy of the energy - supply equipment model. This improvement enables the model to more realistically reflect the dynamic behavior of the equipment, overcoming the defect that the traditional model cannot accurately describe the time delay. Ultimately, a more accurate energy - supply equipment model can provide a more reliable basis for the flexible multi - time - scale scheduling of the integrated energy supply system, thereby formulating a better scheduling scheme and improving the overall operation efficiency and economy of the system.

[0124] (2) The present invention classifies the user's electricity consumption types into discrete load, shiftable load, and continuously adjustable load, and models them separately, more precisely depicting the demand response potential on the user side. The present invention also considers the mandatory satisfaction characteristics of rigid demand loads and the flexible adjustment space of flexible demand loads, constructs a more practical user - side demand response model, fully explores the flexibility on the demand side while ensuring the basic needs of users, and provides a larger space for system optimization scheduling.

[0125] (3) The present invention innovatively considers the pipeline storage capacity and pressure regulation flexibility of the energy flow network in the longitudinal space, as well as the energy medium flow inertia and the flow velocity differences of different media existing in the transverse space, and constructs an energy flow network inertia model that more conforms to the actual operation conditions. By modeling the dynamic characteristics of the energy flow network, it can more accurately predict the operation state of the pipe network, avoid safety problems caused by instantaneous supply-demand imbalance, and improve the reliability and feasibility of the scheduling scheme.

[0126] (4) The dynamic scheduling optimization method proposed by the present invention can update the optimized scheduling scheme for the next time period in real time according to the actual operation data fed back by the integrated energy supply system, effectively cope with prediction errors and system disturbances, and improve the robustness and adaptability of the scheduling. The present invention realizes the balance between scheduling efficiency and accuracy by solving the optimization model at different minimum time granularities and determining the optimal minimum time granularity according to the calculation efficiency and model prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0127] Figure 1 is a flowchart of the multi-time scale dynamic scheduling method for the integrated energy supply system considering the hysteresis effect of the energy supply equipment of the present invention;

[0128] Figure 2 is a schematic diagram of the multi-time scale dynamic optimization method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0129] The following further describes the present invention in detail with reference to the drawings, so that those skilled in the art can implement it according to the description in the specification. The drawings are all simplified schematic diagrams, only showing the basic structure of the present invention in a schematic way, so they only show the components related to the present invention.

[0130] It should be understood that the terms such as "having", "comprising" and "including" used herein do not exclude the presence or addition of one or more other elements or their combinations.

[0131] The present invention takes an integrated energy supply system as an example to describe the method of the present invention in detail. As Figure 1 is a flowchart of the multi-time scale dynamic scheduling method for the integrated energy supply system considering the hysteresis effect of the energy supply equipment of the present invention, which specifically includes the following steps:

[0132] Step S1: Construction of the energy supply equipment model; determine the energy supply equipment and its input and output variables, identify the hysteresis parameters of the output variables of the energy supply equipment relative to the input variables, and construct an energy supply equipment model based on the identified hysteresis parameters;

[0133] Step S2: Construction of the user-side demand response model and the energy flow network inertia model; for rigid demand loads and flexible demand loads, elastic modeling of the user-side demand is performed to obtain the user-side demand response model; for the pipeline storage capacity and pressure regulation flexibility existing in the longitudinal space of the energy flow network, and its flow inertia of energy media and the flow velocity differences of different media existing in the transverse space, modeling is carried out to obtain the energy flow network inertia model;

[0134] Step S3: Construction of a multi-time scale dynamic optimization model; with the overall economic benefit of the integrated energy supply system as the goal, a multi-time scale dynamic optimization model is modeled, and the user-side demand response model and the energy flow network inertia model are used as the constraints of the multi-time scale dynamic optimization model;

[0135] Step S4: Dynamic scheduling optimization of the integrated energy supply system; based on the energy supply equipment model, considering the multi-time scale response differences of different equipment, multiple minimum time granularities are selected; at different minimum time granularities, the multi-time scale dynamic optimization model is solved to obtain an optimized scheduling plan, and the integrated energy supply system is scheduled according to the obtained optimized scheduling plan. The multi-time scale dynamic optimization model calculates the optimized scheduling plan for the next time period based on the actual operation data fed back by the integrated energy supply system; based on the calculation efficiency and model prediction accuracy, the best minimum time granularity is determined, and scheduling optimization is carried out based on the best minimum time granularity.

[0136] The said Step S1 includes the following steps:

[0137] Step S11, determining the energy supply equipment and its input and output variables and collecting data, including the following steps:

[0138] Step S111, determining the energy supply equipment that needs to consider lag parameters, and the specific types of energy supply equipment include combined heat and power units, carbon capture equipment, power-to-gas equipment, electric boilers, and electric chillers;

[0139] Step S112, determining all input and output variables of the energy supply equipment, these variables represent various input variables entering the equipment, and various output variables leaving the equipment. Taking the combined heat and power unit as an example, the input variables include coal feeding flow rate, steam turbine inlet valve opening degree, and medium-pressure extraction valve opening degree, and the output variables include low-pressure steam flow rate, medium-pressure steam flow rate, and electric power; collect the operation data of each input and output variable of each energy supply equipment in the same time period to ensure that all the collected operation data are strictly aligned in time.

[0140] Step S12, using the fast Fourier transform to identify the lag parameters of the output variables of the energy supply equipment relative to the input variables, and the specific steps are as follows:

[0141] Step S121: Select the historical data of a certain output variable a and a certain input variable b of the same energy supply device within the same time period from the collected operation data, and perform normalization processing on the two to obtain E a and E b ;

[0142] Step S122: Apply the fast Fourier transform (FFT) to E a and E b respectively, to transform them from the time domain to the frequency domain, obtaining the input variable spectrum and the output variable spectrum, specifically expressed as:

[0143]

[0144] where f represents frequency;

[0145] Step S123: Cross-spectrum calculation, calculate the product of the output variable spectrum and the complex conjugate of the input variable spectrum to obtain the cross-spectrum S ab (f):

[0146]

[0147] where conj(.) represents the complex conjugate operation;

[0148] Step S124: Apply the inverse fast Fourier transform (IFFT) to the calculated cross-spectrum S ab (F) to transform it back from the frequency domain to the time domain, obtaining the cross-correlation function R ab (τ) of the output variable and the input variable:

[0149] R ab (τ) = IFFT(S ab (f))

[0150] where τ represents the lag time;

[0151] Step S125: Lag parameter identification, analyze the cross-correlation function R ab (τ), and identify the τ corresponding to the position where its maximum value is located max . This position τ max is the lag parameter of the historical data of the output variable a relative to the historical data of the input variable b.

[0152] Step S13: The energy supply device model includes a combined heat and power unit dynamic model, a carbon capture device dynamic model, an electric-to-gas device dynamic model, an electric boiler device dynamic model, and an electric chiller device dynamic model. The construction of the energy supply device model specifically includes the following steps:

[0153] Step S131: Build a dynamic model of a combined heat and power unit. For a combined heat and power unit, the input variables include the coal feeding flow rate, the opening degree of the steam inlet valve of the steam turbine, and the opening degree of the medium-pressure extraction steam valve; the output variables include the low-pressure steam flow rate, the medium-pressure steam flow rate, and the electric power. Considering the time inertia in the steam generation and transportation process, build the following dynamic model of the combined heat and power unit:

[0154]

[0155] In the formula, v TP，out1 , t , v TP，out0，t , P CHP，t are respectively the low-pressure steam flow rate, the medium-pressure steam flow rate, and the electric power of the combined heat and power equipment at time t; v TP，out1，t+1 , v TP，out0，t+1 , P CHP，t+1 are respectively the low-pressure steam flow rate, the medium-pressure steam flow rate, and the electric power of the combined heat and power equipment at time t + 1; are respectively the coal consumption flow rate, the opening degree of the steam inlet valve of the steam turbine, and the opening degree of the medium-pressure extraction steam valve of the combined heat and power equipment at time t - τ 1 ; τ 1 is the time inertia parameter between steam and coal consumption; A CHP0，0 is the influence coefficient of the low-pressure steam flow rate at time t on the low-pressure steam flow rate at time t + 1; A CHP0，1 is the influence coefficient of the medium-pressure steam flow rate at time t on the low-pressure steam flow rate at time t + 1; A CHP0，2 is the influence coefficient of the electric power at time t on the low-pressure steam flow rate at time t + 1; A CHP1，0 is the influence coefficient of the low-pressure steam flow rate at time t on the medium-pressure steam flow rate at time t + 1; A CHP1，1 is the influence coefficient of the medium-pressure steam flow rate at time t on the medium-pressure steam flow rate at time t + 1; A CHP1，2 is the influence coefficient of the electric power at time t on the medium-pressure steam flow rate at time t + 1; A CHP2，0 is the influence coefficient of the low-pressure steam flow rate at time t on the electric power at time t + 1; A CHP2，1 is the influence coefficient of the medium-pressure steam flow rate at time t on the electric power at time t + 1; A CHP2，2 is the influence coefficient of the electric power at time t on the electric power at time t + 1; B CHP0，0 is the influence coefficient characterizing the coal feeding flow rate at time t - τ 1 on the low-pressure steam flow rate at time t + 1; B CHP0，1 is the influence coefficient of the opening degree of the steam inlet valve of the steam turbine at time t - τ 1 on the low-pressure steam flow rate at time t + 1; B CHP0，2 is the influence coefficient of the opening degree of the medium-pressure extraction steam valve at time t - τ 1 on the low-pressure steam flow rate at time t + 1; BCHP1，0 is the coal feeding flow rate at time \(t - \tau\) 1 The influence coefficient of the coal feeding flow rate at time \(t - \tau\) on the medium-pressure steam flow rate at time \(t + 1\); \(B\) CHP1，1 is the coal feeding flow rate at time \(t - \tau\) 1 The influence coefficient of the opening of the steam turbine inlet valve at time \(t - \tau\) on the medium-pressure steam flow rate at time \(t + 1\); \(B\) CHP1，2 is the coal feeding flow rate at time \(t - \tau\) 1 The influence coefficient of the opening of the medium-pressure extraction steam valve at time \(t - \tau\) on the medium-pressure steam flow rate at time \(t + 1\); \(B\) CHP2，0 is the coal feeding flow rate at time \(t - \tau\) 1 The influence coefficient of the coal feeding flow rate at time \(t - \tau\) on the electric power at time \(t + 1\); \(B\) CHP2，1 is the coal feeding flow rate at time \(t - \tau\) 1 The influence coefficient of the opening of the steam turbine inlet valve at time \(t - \tau\) on the electric power at time \(t + 1\); \(B\) CHP2，2 is the coal feeding flow rate at time \(t - \tau\) 1 The influence coefficient of the opening of the medium-pressure extraction steam valve at time \(t - \tau\) on the electric power at time \(t + 1\); \(C\) CHPy，0 are the model constants for the cogeneration unit to produce different output variables. Here, \(y\) is 0, 1, 2, corresponding to the constant terms of the low-pressure steam flow rate, medium-pressure steam flow rate, and electric power models respectively. The above time inertia parameters, production and energy consumption model coefficients, and model constants of the cogeneration unit are all obtained by the data identification method.

[0156] Step S132, construct the dynamic model of the carbon capture equipment: For the carbon capture equipment, its input variable is the electric power, and the output variable is the captured carbon dioxide flow rate. The dynamic model is as follows:

[0157]

[0158] In the formula, \(v\) ccus，out0，t is the carbon dioxide capture flow rate of the CCUS (carbon capture) equipment at time \(t\); \(v\) ccus，out0，t+1 is the carbon dioxide capture flow rate of the CCUS equipment at time \(t + 1\); are the electric powers of the CCUS equipment at time \(t - \tau\) 2 respectively; \(\tau\) 2 is the time inertia parameter between the carbon dioxide capture flow rate and the power consumption; \(A\) ccus0，0 is the model coefficient for the CCUS equipment to capture carbon dioxide; \(C\) ccus0，0 is the model constant for the CCUS equipment to capture carbon dioxide; \(B\) ccus0，2 is the energy consumption model coefficient for the CCUS equipment to capture carbon dioxide.

[0159] Step S133, construct the dynamic model of the power-to-gas equipment: For the power-to-gas (P2G) equipment, its input variables are the carbon dioxide flow rate and the electric power, and the output variable is the natural gas flow rate. The model is as follows:

[0160]

[0161] Wherein, P P2G，t is the natural gas flow rate of the P2G device at time t; P P2G，t+1 is the natural gas flow rate of the P2G device at time t+1; are respectively the carbon dioxide consumption flow rate and the power consumption of the P2G device at time t-τ 3 ; τ 3 is the time inertia parameter between the natural gas flow rate and the power consumption; A P2G0，0 is the model coefficient of the P2G device for producing natural gas, C P2G0，0 is the model constant of the P2G device for producing natural gas; B P2G0，0 , B P2G0，1 are the energy consumption model coefficients of the P2G device for producing natural gas.

[0162] Step S134, construct a dynamic model of the electric boiler device: For the electric boiler device, its input is the electric power and the output is the hot water flow rate. The model is as follows:

[0163]

[0164] Wherein, P EB，t is the hot water flow rate of the electric boiler device at time t; is the power consumption of the electric boiler device at time t-τ 4 ; τ 4 is the time inertia parameter between the hot water flow rate and the power consumption; A EB0，0 is the model coefficient of the electric boiler device for producing hot water; C EB0，0 is the model constant of the electric boiler device for producing hot water; B EB0，0 is the energy consumption model coefficient of the electric boiler device for producing hot water.

[0165] Step S135, construct a dynamic model of the electric chiller device: For the electric chiller device, its input is the electric power and the output is the refrigerant flow rate. The model is as follows:

[0166]

[0167] Wherein, P EC，t is the refrigerant flow rate of the electric chiller device at time t; is the power consumption of the electric chiller device at time t-τ 5 ; τ 5 is the time inertia parameter between the refrigerant flow rate and the power consumption; A EC0，0 is the model coefficient of the electric chiller device for producing refrigerant; C EC0，0 is the model constant of the electric chiller device for producing refrigerant; B EC0，0 is the energy consumption model coefficient of the electric chiller device for producing refrigerant.

[0168] In the described step S2:

[0169] The described step S2 includes the following steps:

[0170] Step S21, constructing a user-side demand response model: Demand response originally referred to guiding users to change their energy consumption behaviors through price signals. Specifically, the user side can interact with the energy supply side in various ways. The electricity consumption types of users are divided into three categories, specifically including discrete load, shiftable load, and continuously adjustable load. Models are respectively established for the three electricity consumption types, and the rigid demand load and flexible demand load are used as constraints to obtain the user-side demand response model; the specific steps are as follows:

[0171] Step S211, modeling the discrete load. The regulation of discrete load means that high-energy-consuming users reduce or increase a certain amount of load within a specified time according to the power grid dispatching instructions, and need to maintain a certain stable operation time and cannot be continuously adjusted. It is specifically expressed as follows:

[0172] P ld，t = P ld，t-1 -(1 - x ld，t ) × d ld，t

[0173] In the formula, P ld，t is the discrete load value at time t; P ld，t-1 is the discrete load value at time t - 1; x ld，t is whether the discrete load at time t participates in regulation, 0 for participation and 1 for non-participation; d ld，t is the downward adjustment amount of the discrete load at time t.

[0174] Step S212, modeling the shiftable load. The shiftable load means that users can transfer the time point of energy consumption demand to other time points to keep the total load within a certain time range unchanged, thereby reducing the load value at peak times. It is specifically expressed as follows:

[0175] P lt，t = P lt，t-1 + x lt，t × d lt，t

[0176] In the formula, P lt，t is the shiftable load value at time t; P lt，t-1 is the shiftable load value at time t - 1; x lt，t is whether there is an adjustment value of other shiftable loads at time t; d lt，t is the load value transferred to time t.

[0177] Step S213: Model the continuously adjustable load. The continuously adjustable load refers to the load on the user side that can continuously and quickly respond to the dispatching instructions issued by the power grid at each time period, support adjustment in each time period, and does not need to maintain a stable operating time, which is specifically expressed as follows:

[0178]

[0179] In the formula, P lc，t is the continuously adjustable load value at time t; d lc，n，t is the output value of the nth continuously adjustable unit at time t; N lc is the number of continuously adjustable units.

[0180] Step S214: Construct the rigid demand load constraint. For user demands that need to be strictly met, such as medium - pressure and low - pressure steam, the constraint conditions are constructed as shown in the following formula:

[0181]

[0182] In the formula, is the demand value of demand i (specifically referring to the energy type, such as electricity, steam, etc.) at time t p ; t p is the time point of the known predicted demand; L demand，d，lb，i is the maximum downward adjustment amount of demand i; I demand，s，i is a 0 - 1 indicator matrix indicating whether the demand is strictly met, where 1 indicates that the corresponding demand needs to be strictly met; L demand，d，ub，i is the maximum upward adjustment amount of demand i; After selecting the time - interval granularity, the entire scheduling interval T is discretized with each t p as the boundary point, and t a is the start time of the discrete interval where t p is located, 0 ≤ t a ≤ t p ≤ T; is the energy supply amount of demand i at time t.

[0183] Step S215: Construct the flexible demand load constraint. For non - strictly met demand quantities, such as natural gas, the constraint is constructed as follows:

[0184]

[0185] Step S22: Construct an inertia model of the energy flow network. Due to the inertia characteristics of fluid energy, it provides a certain degree of flexibility for the operation of the integrated energy supply system, thus helping to suppress load fluctuations. The energy flow network has flexibility in pipeline storage capacity and pressure regulation in the longitudinal space, and in the transverse direction, it is manifested as the flow inertia of energy media and the difference in flow velocities of different media. The characteristics of the energy flow network in the longitudinal and transverse directions jointly provide flexibility for operation scheduling.

[0186] Step S221: The present invention regards the fluid energy flow network as a large-scale energy storage unit. During the scheduling process, by utilizing the pressure regulation range of the energy flow network and analyzing the difference between the current steam supply flow and the user demand, the energy supply of the equipment is adjusted under the condition of ensuring the safety of the pipe network pressure, so as to achieve flexible response. Due to the existence of the energy flow network, there is room for adjustment of the difference between the inflow and outflow of fluid energy. Therefore, the supply capacity of the energy flow network can be used to analyze the difference between the current fluid energy supply flow and the predicted user demand, and the smoothness of the equipment load can be maintained as much as possible on the premise of ensuring the safety of the pipe network pressure. Model the flexibility of the pipeline storage capacity and pressure regulation existing in the energy flow network in the longitudinal space, which is specifically expressed as:

[0187]

[0188] L vs，lb ≤v smo ≤L vs，ub

[0189] v smo，0 =v r

[0190] In the formula, p smo is the smoothed pressure of the fluid energy pipe network, which needs to be kept within a safe range; p r is the actual pressure of the fluid energy pipe network; v pre is the energy demand at the user end at a future time interval; v smo is the smoothed fluid energy steam supply; θ r is the actual temperature of the fluid energy pipe network; v rated is the volume of the fluid energy pipe network; v smo，0 is the initial value of v smo ; v r is the actual steam supply of the fluid energy; L vs，lb is the lower limit of the fluid energy steam supply; L vs，ub is the upper limit of the fluid energy steam supply; M is the molar mass; R is the gas constant; d t is the coefficient measuring the sensitivity of the energy flow network pressure to the imbalance between supply and demand.

[0191] Step S222: Considering the flow inertia of the energy medium and the flow velocity difference, there is an optimal operating range for the pipe network pressure. The present invention further realizes the dynamic calculation of the optimal pressure value of the pipe network by fitting the relationship curve between the energy load and the pipe network pressure. This method can convert the static pipe network pressure interval into a dynamic interval, so as to more accurately adapt to the changes in energy supply and demand and maximize the utilization of the flexibility of the energy flow network. Modeling the flow inertia of the energy medium existing in the horizontal space of the energy flow network and the flow velocity difference of different media, specifically:

[0192]

[0193] L p，ub = p * + p d

[0194] L p，lb = p * - p d

[0195] L p，lb ≤ p r -(v pre - v smo )×d t ×θ r ×R / M / v rated ≤ L p，ub

[0196] In the formula, p * is the optimal value of the pipe network pressure; g is the function of the energy load and the optimal pipe network pressure; is the average value of the energy demand at the user end at a future time interval; L p，ub is the upper limit of the pipe network pressure; p d is the adjustable interval of the optimal pressure; L p，lb is the lower limit of the pipe network pressure for demand i.

[0197] Generally speaking, the flexibility brought by the inertia of the energy flow network is realized by adjusting v smo , and its flexible adjustment space depends on the upper and lower limits of the pipe network pressure, and the calculation of the optimal value of the pipe network pressure further expands the flexibility interval.

[0198] The construction of the multi-time scale dynamic optimization model described in step S3 includes the following steps:

[0199] Step S31. Multi-time scale dynamic optimization model. For the integrated energy supply system studied in the present invention, its overall economic income includes external sales of steam, external sales of hot water, electricity sales, external sales of natural gas, external sales of carbon emission rights, etc., and the expenditure items include penalties for abandoned wind and light, coal consumption costs, and penalty costs for unmet demand. The goal of multi-time scale dynamic optimization of the integrated energy supply system is to maximize the overall system benefit. Therefore, the objective function of the multi-time scale dynamic optimization model is as follows:

[0200]

[0201] In the formula, E total is the total benefit of the supply system; T is the entire scheduling period; v supply，i，t is the quantity of externally sold energy i at time t, where i is 1, 2, 3, 4, 5, 6, 7 respectively, corresponding to low-pressure steam, medium-pressure steam, hot water, refrigerant, carbon dioxide, electricity, and natural gas; is the price of externally sold energy i at time t; v punish，j，t is the quantity of penalty item j at time t, and j is the abandoned wind and light quantity, coal consumption quantity, and unmet demand quantity in sequence; is the penalty cost coefficient of penalty item j at time t.

[0202] Step S32. Establish the constraint conditions of the multi-time scale dynamic optimization model. The model constraint conditions include upper and lower limits of equipment load constraints, minimum load constraints for equipment startup, equipment operation performance constraints, demand constraints, energy flow inertia constraints, energy storage device capacity constraints, equipment ramp-up constraints, node energy balance constraints, and initial value constraints:

[0203] Upper and lower limits of equipment load constraints:

[0204] L in，m，lb ≤u m，t ≤L in，m，ub

[0205] L out，n，lb ≤y n，t ≤L out，n，ub

[0206] In the formula, L in，m，lb is the lower limit of equipment energy consumption m; u m，t is the quantity of equipment energy consumption m at time t; L in，m，ub is the upper limit of equipment energy consumption m; L out，n，lb is the lower limit of equipment production capacity n; y n，t is the quantity of equipment production capacity n at time t; L out，n，ub is the upper limit of equipment production capacity n.

[0207] Minimum load constraints when the equipment is started:

[0208] (u m，t-0)(u m，t -L in，t，ub ×r m )≥0

[0209] In the formula, r m is the ratio of the minimum operating load of the energy consumption m of the device at startup to its maximum load.

[0210] Device performance constraint:

[0211] is the energy supply device model equation constructed in step S1.

[0212] Demand constraint:

[0213] is the user-side demand response model equation constructed in step S2.

[0214] Energy flow inertia constraint:

[0215] is the energy flow network inertia model equation constructed in step S2;

[0216] Energy storage device capacity constraint:

[0217]

[0218] L sx，k，lb ≤P sx，k，t ≤L sx，k，ub

[0219] In the formula, P sx，k，t+1 is the stored energy of the energy storage device k at time t + 1; P sx，in，k，t is the charging energy of the energy storage device k at time t; η sx，in，k is the charging efficiency of the energy storage device k; P sx，out，k，t is the discharging energy of the energy storage device k at time t; η sx，out is the discharging power of the energy storage device; L sx，k，lb is the lower limit of the stored energy of the energy storage device k; L sx，k，ub is the upper limit of the stored energy of the energy storage device k.

[0220] Device ramp constraint:

[0221] |u m，t+1 -u m，t |≤L demand，d，ub，m

[0222] In the formula, L demand，d，ub，m is the upper limit of the ramp-up capacity of the device energy consumption m.

[0223] Node in-out energy conversion constraint:

[0224] The energy flow branch node needs to satisfy the constraint that the input energy is equal to the output energy.

[0225] Initial value constraint:

[0226] P sx，k，t = P sx，k，init

[0227] u m，t = u m，init

[0228] y n，0 = y n，init

[0229] Wherein, P sx，k，init is the initial stored energy value of the energy storage device k; u m，init is the initial load value of the device energy consumption m; y n，init is the initial production capacity value of the device production capacity n.

[0230] The above multi-time scale dynamic optimization model is an optimization model with quadratic constraints and is solved using the interior point method.

[0231] In the said step S4:

[0232] Such as Figure 2 , the optimized scheduling time at the current time Tr (i.e., the future T time) is taken, and it is divided into four sub-periods with T / 4 as the time interval unit. Each sub-period is further divided into the minimum scheduling time τ 0 (i.e., the minimum time granularity); during the online dynamic optimization, the input parameters are the different demand prediction values v demand，i，t (v demand，i，t is the demand prediction value of demand i at time t. i are 1, 2, 3, 4, 5, 6, 7 respectively, corresponding to low-pressure steam, medium-pressure steam, hot water, refrigerant, carbon dioxide, electricity, natural gas, and t are T1, T2, T3, T4 respectively) that make up the demand prediction matrix v demand ;

[0233] According to the input demand prediction matrix and the multi-time scale dynamic optimization model, with the minimum scheduling time as the scheduling period, calculate the optimized scheduling scheme at different minimum scheduling times, and select the best minimum scheduling time according to the calculation efficiency and the model prediction accuracy;

[0234] Use the multi-time scale dynamic optimization model to calculate the optimized scheduling scheme of the integrated energy supply system within the future T time, as shown in the following formula:

[0235] P sx , u, y = f(P sx，k，init , u m，init , y n，init , v demand )

[0236] In the formula, f is the constructed multi-time scale dynamic optimization model, which takes the parameters in the brackets as inputs and outputs the optimized scheduling plan; P sx，k，init is the initial energy storage capacity of energy storage device k; u m，init is the initial load of equipment energy consumption m; y n，init is the initial production capacity of equipment production n.

[0237] P sx , u, and y are the optimized scheduling plans, where: P sx is the scheduling matrix of the energy storage capacities of all energy storage devices in the future T time with the minimum scheduling time as the scheduling period, guiding the charging and discharging behaviors of all energy storage devices in each scheduling period; u is the scheduling matrix of the energy consumptions of all equipment in the future T time with the minimum scheduling time as the scheduling period, specifying the energy consumption plans of all energy supply equipment in each scheduling period; y is the scheduling matrix of the production capacities of all equipment in the future T time with the minimum scheduling time as the scheduling period, specifying the production capacity plans of all energy supply equipment in each scheduling period.

[0238] The integrated energy supply system is scheduled according to the obtained optimized scheduling plan. The execution of the optimized scheduling plan is a dynamic and rolling process, with the minimum scheduling time as the basic execution unit, and the plan is updated and adjusted at the sub-cycle level. The specific method is as follows:

[0239] 1) At the current time T r , according to the optimized scheduling plan of the integrated energy supply system in the future T time calculated by the multi-time scale dynamic optimization model, the scheduling plan of the first minimum scheduling time in the first sub-cycle is sent to the integrated energy supply system for execution; subsequently, the actual execution data of the integrated energy supply system (the actual energy storage capacity of the energy storage device, the actual load and production capacity of each energy supply equipment, and the actual demand) are collected and fed back to the multi-time scale dynamic optimization model; after receiving the feedback data, the optimization model will recalculate and update the scheduling plan for the remaining time periods in the future T time based on the current actual state, so as to update the scheduling plan for the second scheduling time in the first sub-cycle; and so on until the first sub-cycle is completed; this update is still carried out in units of the minimum scheduling time, aiming to correct the deviation that may be caused by prediction errors or system disturbances and provide more accurate guidance for subsequent minimum scheduling times. This process of "execution - collection - feedback - update" will continue in the first sub-cycle until the first sub-cycle ends (reaching the T1 moment).

[0240] 2) When the first sub-cycle ends, the multi-time-scale dynamic optimization model will receive the latest demand prediction matrix for the next T time periods and the current actual state of the system (i.e., at time T1), and recalculate a new optimized scheduling plan for the next four sub-cycles based on the operations in step 1).

[0241] 3) Repeat the operation in step 2), adjusting and executing only the plan for the first sub-cycle each time, until all sub-cycles within the next T time periods have been executed.

[0242] Taking a combined heat and power (CHP) unit as an example, considering the time lag in the dynamic response of the CHP unit, with a lag parameter of τ 1 , assuming at a certain moment t a + τ 1 within the next T time periods, according to the production capacity plan of the optimized scheduling plan, the integrated energy supply system requires the CHP unit to reach a production capacity target matrix y 0 (t a + τ 1 ) consisting of specific low-pressure steam flow rate, medium-pressure steam flow rate, and electric power; at the current time t a , the system will pre-set an energy consumption scheduling matrix u 0 (t a ) consisting of the coal feed flow rate of the CHP unit, the opening degree of the steam turbine inlet valve, and the opening degree of the medium-pressure extraction steam valve, to ensure that at time t a + τ 1 it can more accurately reach the expected production capacity target matrix y 0 (t a + τ 1 ).

[0243] Based on the understanding and prediction of the dynamic characteristics of the CHP unit, the dynamic optimization model determines the optimal energy consumption scheduling matrix u a (t 0 ) at time t a such that after a time delay of τ 1 , i.e., at time t a + τ 1 , the actual output of the CHP unit can be as close as possible to the expected production capacity target y 0 (t a + τ 1 ). It should be emphasized that due to the errors in model prediction and various uncertainties in system operation, the actual output may not exactly equal the target value, but the goal of the scheduling strategy is to minimize this deviation as much as possible.

[0244] Figure 2 The t shown in b , t c , td represent similar future time points. When making scheduling decisions at these time points, corresponding input variables are also set in advance based on the corresponding target outputs and the lag characteristics of the combined heat and power units, with the aim of achieving the desired production capacity level at the future target time. For example, at time t b , the model will consider the production capacity target at t b +τ 1 time, and set the energy consumption scheduling matrix at t b time.

[0245] Traditional scheduling optimization methods usually formulate a static scheduling plan at a certain time point t, and this plan is mainly to meet the demand at time t, with less consideration of the future and the dynamic response of the equipment. While the present invention takes into account the lag time between the equipment input and output, which means that the impact of the equipment input decision made at time t will be reflected at the future time t + k. The present invention adopts a dynamic optimization method to ensure that the cumulative demand can be met within a sub-cycle, rather than just focusing on the instantaneous demand. Update the operation plan for the first future sub-cycle according to the actual execution effect. The scheduling scheme has a smaller scheduling time and can handle the supply and demand changes and equipment capacity limitations more precisely. By continuously receiving actual data and updating the future scheduling scheme, the scheduling decision can adapt to the dynamic changes of the system.

Claims

1. A multi-time-scale dynamic scheduling method for an integrated energy supply system considering the hysteresis effect of energy supply equipment. It is characterized in that The steps include: Step S1: constructing an energy supply device model; determining the energy supply device and its input and output variables, identifying the hysteresis parameters of the output variables of the energy supply device relative to the input variables, and constructing the energy supply device model based on the identified hysteresis parameters; Step S2: Construction of user-side demand response model and energy flow network inertia model; elastic modeling of user-side demand for rigid demand load and flexible demand load to obtain the user-side demand response model; modeling of pipeline storage capacity and pressure regulation flexibility in the longitudinal space of the energy flow network, and flow inertia of energy media in the transverse space and flow velocity differences of different media to obtain the energy flow network inertia model; Step S3: construct a multi-time scale dynamic optimization model; taking the overall economic benefits of the integrated energy supply system as the goal, modeling the multi-time scale dynamic optimization model, and taking the user-side demand response model and the energy flow network inertia model as constraints of the multi-time scale dynamic optimization model; Step S4: Dynamic scheduling optimization of the integrated energy supply system; based on the energy supply equipment model, multiple minimum time granularities are selected taking into account the multi-time scale response differences of different equipment; under different minimum time granularities, the multi-time scale dynamic optimization model is solved to obtain an optimized scheduling plan, and the integrated energy supply system is scheduled according to the obtained optimized scheduling plan, and the multi-time scale dynamic optimization model calculates the optimized scheduling plan for the next time period according to the actual operation data fed back by the integrated energy supply system; based on the computational efficiency and model prediction accuracy, the optimal minimum time granularity is determined, and scheduling optimization is performed based on the optimal minimum time granularity.

2. The multi-time scale dynamic scheduling method of the integrated energy supply system considering the hysteresis effect of energy supply equipment according to claim 1 is characterized in that: The step S1 comprises the following steps: Step S11, determining the energy supply equipment and its input and output variables and collecting data, includes the following steps: Step S111, determining energy supply equipment for which hysteresis parameters need to be considered, specifically including cogeneration units, carbon capture equipment, power-to-gas equipment, electric boilers, and electric refrigerators; Step S112, determining all input and output variables of the energy supply equipment, and collecting the operation data of each input and output variable of each energy supply equipment in the same time period, ensuring that all collected operation data are strictly aligned in time; Step S12, using fast Fourier transform to identify the lag parameter of the output variable of the energy supply device relative to the input variable, the specific steps are as follows: Step S121: select historical data of an output variable a and historical data of an input variable b of the same energy supply device in the same time period from the collected operation data, and normalize the two to obtain E a and E b ; Step S122, respectively a and E b Apply Fast Fourier Transform FFT to convert it from time domain to frequency domain to obtain the input variable spectrum And the output variable spectrum Specifically expressed as: Where, f represents frequency; Step S123, cross spectrum calculation, calculate the output variable spectrum With input variable spectrum The cross spectrum S is obtained by multiplying the complex conjugate of ab (f): Among them, conj(.) represents the complex conjugate operation; Step S124, the calculated cross spectrum S ab (f) Apply inverse fast Fourier transform IFFT to convert it from frequency domain back to time domain and obtain the cross-correlation function R between the output variable and the input variable ab (τ): R ab (τ)=IFFT(S ab (f)) Where, τ represents the lag time; Step S125, hysteresis parameter identification, analysis of cross-correlation function R ab (τ), identify the τ corresponding to the position of its maximum value max ; τ max That is, the lag parameter of the historical data of the output variable a relative to the historical data of the input variable b; Step S13, constructing an energy supply equipment model, wherein the energy supply equipment model includes a dynamic model of a cogeneration unit, a dynamic model of a carbon capture device, a dynamic model of a power-to-gas device, a dynamic model of an electric boiler device, and a dynamic model of an electric refrigerator device. The specific steps are as follows: Step S131, constructing a dynamic model of a cogeneration unit. For a cogeneration unit, input variables include coal feed flow, turbine steam inlet valve opening, and medium-pressure extraction valve opening; output variables include low-pressure steam flow, medium-pressure steam flow, and electric power; considering the time inertia of steam generation and transportation process, the following dynamic model of the cogeneration unit is constructed: In the formula, v TP,out1,t 、v TP,out0,t , P CHP,t are the low-pressure steam flow, medium-pressure steam flow, and electric power of the cogeneration equipment at time t; v TP,out1,t+1 、v TP,out0,t+1 , P CHP,t+1 are the low-pressure steam flow, medium-pressure steam flow, and electric power of the cogeneration equipment at time t+1, respectively; are the coal consumption flow rate, turbine steam inlet valve opening, and medium-pressure extraction valve opening of the cogeneration equipment at time t-τ1; τ1 is the time inertia parameter between steam and coal consumption; A CHP0,0 A is the influence coefficient of low-pressure steam flow at time t on low-pressure steam flow at time t+1; CHP0,1 A is the influence coefficient of the medium-pressure steam flow at time t on the low-pressure steam flow at time t+1; CHP0,2 A is the influence coefficient of electric power at time t on the low-pressure steam flow at time t+1; CHP1,0 A is the influence coefficient of low-pressure steam flow at time t on medium-pressure steam flow at time t+1; CHP1,1 A is the influence coefficient of the medium-pressure steam flow at time t on the medium-pressure steam flow at time t+1; CHP1,2 A is the influence coefficient of electric power at time t on the medium pressure steam flow at time t+1; CHP2,0 A is the influence coefficient of low-pressure steam flow at time t on the electric power at time t+1; CHP2,1 A is the influence coefficient of medium-pressure steam flow at time t on the electric power at time t+1; CHP2,2 B is the influence coefficient of the electric power at time t on the electric power at time t+1; CHP0,0 B is the coefficient that represents the influence of the coal flow rate at time t-τ1 on the low-pressure steam flow rate at time t+1; CHP0,1 B is the influence coefficient of the steam inlet valve opening of the steam turbine at time t-τ1 on the low-pressure steam flow at time t+1; CHP0,2 B is the influence coefficient of the opening of the medium-pressure extraction valve at time t-τ1 on the low-pressure steam flow at time t+1; CHP1,0 B is the influence coefficient of coal flow rate at time t-τ1 on medium pressure steam flow rate at time t+1; CHP1,1 B is the influence coefficient of the opening of the steam inlet valve of the steam turbine at time t-τ1 on the medium pressure steam flow at time t+1; CHP1,2 B is the influence coefficient of the opening of the medium-pressure extraction valve at time t-τ1 on the medium-pressure steam flow at time t+1; CHP2,0 B is the influence coefficient of coal flow rate at time t-τ1 on the electric power at time t+1; CHP2,1 B is the influence coefficient of the steam inlet valve opening of the steam turbine at time t-τ1 on the electric power at time t+1; CHP2,2 is the influence coefficient of the opening of the medium pressure extraction valve at time t-τ1 on the electric power at time t+1; C CHPy,0 is the model constant for producing different output variables of the cogeneration unit, where y is 0, 1, and 2, corresponding to the constant terms of the low-pressure steam flow, medium-pressure steam flow, and electric power model, respectively; Step S132, constructing a dynamic model of the carbon capture equipment: For the carbon capture equipment, its input variable is electric power, and its output variable is carbon dioxide capture flow rate. The dynamic model of the carbon capture equipment is as follows: In the formula, v ccus,out0,t is the CO2 capture flow of CCUS equipment at time t; v ccus,out0,t+1 is the CO2 capture flow of the CCUS equipment at time t+1; are the power consumption of CCUS equipment at time t-τ2; τ2 is the time inertia parameter between carbon dioxide capture flow and power consumption; A ccus0,0 B is the influence coefficient of the carbon dioxide capture flow at time t on the carbon dioxide capture flow at time t+1; ccus0,2 C is the influence coefficient of the electric power at time t-τ2 on the carbon dioxide capture flow at time t+1; ccus0,0 Model constants for CO2 capture in CCUS plants; Step S133, constructing a dynamic model of a power-to-gas device: For a power-to-gas (P2G) device, the input variables are the carbon dioxide flow rate and the electric power, and the output variable is the natural gas flow rate. The dynamic model of the power-to-gas device is as follows: Where P P2G,t is the natural gas flow of the P2G device at time t; P P2G,t+1 is the natural gas flow of the P2G equipment at time t+1; are the carbon dioxide flow rate and power consumption of the P2G equipment at time t-τ3; τ3 is the time inertia parameter between the natural gas flow rate and power consumption; A P2G0,0 is the influence coefficient of the natural gas flow at time t on the natural gas flow at time t+1; B P2G0,0 B is the influence coefficient of carbon dioxide flow at time t-τ3 on natural gas flow at time t+1; P2G0,1 is the influence coefficient of the electric power at time t-τ3 on the natural gas flow at time t+1; C P2G0,0 is the model constant for natural gas production in P2G equipment; Step S134, constructing a dynamic model of an electric boiler device: For an electric boiler device, its input is electric power and its output is hot water flow. The dynamic model of the electric boiler device is as follows: Where P EB,t is the hot water flow of the electric boiler at time t; P EB,t+1 is the hot water flow of the electric boiler equipment at time t+1; is the power consumption of the electric boiler equipment at time t-τ4; τ4 is the time inertia parameter between the hot water flow rate and the power consumption; A EB0,0 B is the influence coefficient of the hot water flow rate at time t on the hot water flow rate at time t+1; EB0,0 is the influence coefficient of the electric power at time t-τ4 on the hot water flow at time t+1; C EB0,0 is the model constant for hot water production by electric boiler equipment; Step S135, constructing a dynamic model of an electric refrigerator: For an electric refrigerator, its input is electric power and its output is refrigerant flow. The dynamic model of the electric refrigerator is as follows: Where P EC,t is the refrigerant flow rate of the electric refrigerator at time t; P EC,t+1 is the refrigerant flow rate of the electric refrigerator at time t+1; is the power consumption of the electric refrigerator at time t-τ5; τ5 is the time inertia parameter between the refrigerant flow rate and the power consumption; A EC0,0 B is the influence coefficient of the refrigerant flow rate at time t on the refrigerant flow rate at time t+1; EC0,0 is the influence coefficient of the electric power at time t-τ5 on the refrigerant flow at time t+1; C EC0,0 is the model constant for the refrigerant produced by the electric refrigeration equipment.

3. According to claim 2, a multi-time scale dynamic scheduling method for an integrated energy supply system considering the hysteresis effect of energy supply equipment is characterized in that: The step S2 comprises the following steps: Step S21, constructing a user-side demand response model: dividing the user's electricity consumption types into three categories, specifically including discrete loads, transferable loads and continuously adjustable loads, modeling the three types of electricity consumption respectively, and taking rigid demand loads and flexible demand loads as constraints to obtain the user-side demand response model; the specific steps are: Step S211, modeling the discretizable load, which is specifically expressed as follows: P ld,t =P ld,t-1 -x ld,t ×d ld,t Where P ld,t is the discrete load value at time t; P ld,t-1 is the discrete load value at time t-1; x ld,t Whether the discrete load at time t participates in regulation, if it participates, it is 1, if it does not participate, it is 0; d ld,t is the discrete load reduction at time t; Step S212, modeling the transferable load, which is expressed as follows: P lt,t =P lt,t-1 +x lt,t ×d lt,t Where P lt,t is the transferable load value at time t; P lt,t-1 is the transferable load value at time t-1; x lt,t Is there any other transferable load adjustment value at time t, if yes, then it is 1, if not, then it is 0; d lt,t is the load value transferred to time t; Step S213, modeling the continuously adjustable load is as follows: Where P lc,t is the continuously adjustable load value at time t; d lc,n,t is the output value of the nth continuously adjustable unit at time t; N lc The number of units can be continuously adjusted; Step S214, constructing a rigid demand load constraint, specifically: In the formula, For demand i at t p Demand value at time t p is the time point at which the forecast demand is known; L demand,d,lb,i is the maximum amount of downward adjustment that can be made for demand i; I demand,s,i L is a 0-1 indicator matrix indicating whether the requirement is strictly met, 1 means the corresponding requirement needs to be strictly met; demand,d,ub,i is the maximum adjustable amount of demand i; after selecting the time interval granularity, the entire scheduling interval T is divided into p Discretize the boundary points, t a t p The start time of the discrete interval, 0≤t a ≤t p ≤T; is the energy supply for demand i at time t; Step S215, construct the flexible demand load constraint as follows: Step S22, constructing an energy flow network inertia model, specifically includes the following steps: Step S221, modeling the pipeline storage capacity and pressure regulation flexibility of the energy flow network in the longitudinal space, specifically: v smo,0 =v r In the formula, p smo The smoothed fluid energy network pressure needs to be kept within a safe range; r is the actual pressure of the fluid energy network; v pre is the energy demand of the user at a certain time interval in the future; smo is the smoothed fluid energy steam supply; θ r is the actual temperature of the fluid energy network; v rated is the volume of the fluid energy network; v smo,0 v smo The initial value of v r is the actual steam supply of fluid energy; L vs,lb L is the lower limit of the steam supply of fluid energy; vs,ub is the upper limit of the steam supply of fluid energy; M is the molar mass; R is the gas constant; d t A coefficient that measures the sensitivity of energy flow network pressure to supply-demand imbalance; Step S222, modeling the flow inertia of energy media existing in the energy flow network in the lateral space and the flow speed difference of different media, specifically: L p,ub =p * +p d L p,lb =p* - p d L p,lb ≤p r -(v pre -v smo )×d t ×θ r ×R / M / v rated ≤L p,ub In the formula, p * is the optimal value of the pipe network pressure; g is the function of energy load and the optimal pressure of the pipe network; is the average energy demand of the user end in a certain time interval in the future; L p,ub is the upper limit of the pipe network pressure; p d is the adjustable range of optimal pressure; L p,lb is the lower limit of the network pressure for demand i.

4. According to claim 3, a multi-time scale dynamic scheduling method for an integrated energy supply system considering the hysteresis effect of energy supply equipment is characterized in that: The construction of the multi-time scale dynamic optimization model in step S3 includes the following steps: Step S31, taking the overall economic benefits of the integrated energy supply system as the goal, the objective function of the multi-time scale dynamic optimization model is as follows: In the formula, E total is the total benefit of the supply system; T is the entire scheduling period; v supply,i,t is the amount of energy i sold at time t, where i is 1, 2, 3, 4, 5, 6, and 7, which are low-pressure steam, medium-pressure steam, hot water, refrigerant, carbon dioxide, electricity, and natural gas, respectively; is the price of energy i sold at time t; v punish,j,t is the number of penalty projects j at time t, where j represents the amount of abandoned wind and solar power, coal consumption, and unmet demand; is the penalty cost coefficient of penalty item j at time t; Step S32, establishing constraints of a multi-time scale dynamic optimization model, wherein the model constraints include upper and lower limits of equipment load, minimum load constraints when equipment is turned on, equipment performance constraints, demand constraints, energy flow inertia constraints, energy storage equipment storage capacity constraints, equipment climbing constraints, node in and out energy conversion constraints, and initial value constraints; Equipment load upper and lower limit constraints: L in,m,lb ≤u m,t ≤L in,m,ub L out,n,lb ≤y n,t ≤L out,n,ub Where, L in,m,lb is the lower limit of the equipment energy consumption m; u m,t is the amount of energy consumption m of the equipment at time t; L in,m,ub is the upper limit of the equipment energy consumption m; L out,n,lb is the lower limit of equipment capacity n; y n,t is the number of equipment capacities n at time t; L out,n,ub is the upper limit of the equipment capacity n; Minimum load constraint when the device is turned on: (and m,t -0)(and m,t -L in,m,ub ×r m )≥0 In the formula, r m It is the ratio of the minimum operating load of the equipment energy consumption m to the maximum load when the equipment is turned on; Equipment performance constraints: Specifically, the energy supply equipment model equation; Requirements constraints: Specifically, it is the user-side demand response model equation; Energy flow inertia constraint: Specifically, it is the energy flow network inertia model equation; Energy storage device storage capacity constraints: Where P sx,k,t+1 and P sx,k,t are the storage energy of energy storage device k at time t+1 and t respectively; P sx,in,k,t is the charging energy of energy storage device k at time t; η sx,in,k is the charging efficiency of energy storage device k; P sx,out,k,t is the energy released by the energy storage device k at time t; η sx,out L is the energy storage device discharge power; sx,k,lb is the lower limit of the energy storage capacity of energy storage device k; L sx,k,ub is the upper limit of the energy storage capacity of the energy storage device k; Equipment climbing constraints: |in m,t+1 -in m,t |≤L demand,d,ub,m Where, L demand,d,ub,m is the upper limit of the climbing capacity of the equipment energy consumption m; Node in and out energy conversion constraints: The energy flow branch nodes satisfy the constraint that the input energy is equal to the output energy; Initial value constraints: P sx,k,t =P sx,k,init in m,t =in m,init and n,0 =and n,init Where P sx,k,init is the initial value of the energy storage capacity of the energy storage device k; u m,init is the initial load value of the equipment energy consumption m; y n,init is the initial capacity value of equipment capacity n.

5. According to claim 4, a multi-time scale dynamic scheduling method for an integrated energy supply system considering the hysteresis effect of energy supply equipment is characterized in that: In step S4: The optimization scheduling time is divided into four sub-periods, and each sub-period is discretized into different minimum time granularities; during the online dynamic optimization, the input parameter is the demand forecast value in the future T time; taking T / 4 as the time interval unit, the input parameter is the demand forecast matrix v composed of the demand forecast values ​​at the end of the four sub-periods demand ; According to the demand forecast matrix and multi-time scale dynamic optimization model, calculate the optimal scheduling scheme under different minimum time granularities, and select the best minimum time granularity based on the calculation efficiency and model prediction accuracy; The multi-time scale dynamic optimization model is used to calculate the optimal scheduling scheme of the integrated energy supply system in the future T time, as shown in the following formula: P sx ,u,y=f(P sx,k,init ,u m,init ,y n,init ,v demand ) Where, f is the constructed multi-time scale dynamic optimization model; P sx ,u,y are the optimized scheduling schemes, where P sx is the scheduling matrix of the energy storage of all energy storage devices in the future T time with the minimum time granularity as the scheduling period, u is the scheduling matrix of the energy consumption of all equipment in the future T time with the minimum time granularity as the scheduling period, and y is the scheduling matrix of the production capacity of all equipment in the future T time with the minimum scheduling time as the scheduling period; The integrated energy supply system is dispatched according to the obtained optimal dispatching scheme, and the specific method is as follows: 1) Send the scheduling plan of the first minimum time granularity of the first sub-cycle to the integrated energy supply system for execution; then, collect the actual execution data of the integrated energy supply system and feed it back to the multi-time scale dynamic optimization model to update the scheduling plan of the second time granularity of the first sub-cycle; and so on, until the first sub-cycle is completed; 2) When the first sub-cycle is completed, the multi-time scale dynamic optimization model will receive the latest demand forecast matrix and recalculate the optimized scheduling scheme for the new four sub-cycles based on the operation in step 1); 3) Repeat the operation of step 2) each time, adjusting and executing only the plan of the first sub-cycle until all sub-cycles within the next T time are completed.

Images