CN110707745A - Multi-time scale economic dispatching method of electric heating integrated system based on improved VMD - Google Patents

Abstract

The invention discloses an improved VMD-based multi-time scale economic dispatching method of an electric heating integrated system, which comprises the steps of firstly, improving the VMD into a self-adaptive decomposition method capable of being based on specified central frequency; then establishing a multi-time scale scheduling model of the electric heating comprehensive system containing mixed energy storage, determining the central frequency according to the long-term wind power trend, the energy storage and the response speed of the thermal part, and performing 3-layer decomposition on the wind power signal by adopting an improved VMD (variable mode decomposition); the layer 1 is matched with the long-term trend of wind power and used for formulating unit start and stop, unit and electric boiler initial output, the layer 2 frequency is adaptive to the response speed of the energy storage device and used for determining the charging and discharging plan of the energy storage device and the regulating quantity of the unit and the electric boiler, the layer 3 frequency is adaptive to the power storage device and used for formulating the charging and discharging plan of the energy storage device, and then the charging and discharging plan and the real-time scheduling model are corrected. By adopting the improved VMD, the air abandoning quantity and the load shedding quantity can be effectively reduced.

Description

Multi-time scale economic dispatching method of electric heating integrated system based on improved VMD

Technical Field

The invention relates to an improved VMD-based multi-time scale economic dispatching method for an electric heating integrated system, and belongs to the field of wind power uncertainty analysis and new energy grid-connected dispatching.

Background

In recent years, with the more serious problems of environmental pollution and energy crisis, renewable energy represented by wind power is rapidly developed to attract people to pay attention, but due to the characteristics of wind power volatility and randomness, the phenomenon of wind abandon is serious, and huge loss is brought to economy.

In the multi-time scale optimization scheduling of the electric heating integrated system, the energy storage devices and the electric boiler with different response speeds can play a role of wind curtailment, and the electric boiler can work when the wind curtailment is serious, so that the output of a cogeneration unit is reduced, and the wind power consumption space is improved; the energy storage device can transfer energy in a space-time range, can be charged when wind is abandoned seriously, and can discharge when wind power is at a valley, so that fluctuation can be well stabilized, and wide operation is realized. Energy storage devices are mainly classified into two categories according to different response speeds: energy type and power type. The energy type energy storage device has high energy density, but has slow response speed, cannot be charged and discharged frequently, and mainly represents storage battery energy storage and pumped storage; the power type energy storage device has high response speed, can be charged and discharged frequently, but has low energy density, and is mainly represented by a super capacitor and superconducting energy storage. The wind power output power can be divided into a low frequency band, a medium frequency band and a high frequency band according to the frequency, and the energy storage device with low response speed or the electric boiler can not timely absorb the wind power high frequency band power due to different response speeds of the energy storage device, the power storage device and the electric boiler, so that the research on the multi-time scale optimization scheduling of the electric heating comprehensive system for determining the power of the energy storage device and the electric boiler which are adaptive to the frequency of the wind power according to the power of the wind power in different frequency bands has practical significance. Because the energy storage device and the power storage device have complementary advantages in performance, most of the existing researches combine the energy storage device and the power storage device to stabilize wind power fluctuation, wherein the hybrid energy storage mode of a storage battery-super capacitor is the most typical. The research mainly focuses on the establishment of the hybrid energy storage device and the optimized scheduling in the system, and different response speeds of the hybrid energy storage device require different frequency decomposition of the wind power signal, so that the selection of a reasonable power signal decomposition method is particularly important.

Therefore, researchers put forward various signal decomposition methods, such as a hybrid energy storage control strategy based on wavelet packet decomposition, power signals are divided into low, medium and high frequencies through wavelet packet decomposition and reconstruction of the power signals and further distributed according to different response speeds of an energy storage device, although wavelet analysis has good time-frequency localization characteristics, the decomposition effect depends on selection of basis functions and threshold values, the self-adaptability is poor, the wind power signal characteristics are complex, strong uncertainty exists, and the wind power signal decomposition by utilizing the wavelet brings difficulty to the selection of the basis functions and the threshold values, so the wavelet analysis is not suitable for the wind power signal decomposition; if an learner combines wind power integration fluctuation standards and performance characteristics of a hybrid energy storage system, a self-adaptive wavelet packet decomposition method is provided, but only the wind power integration fluctuation standards are used as constraint conditions of a first layer of wavelet decomposition, and for the remaining medium-frequency and high-frequency parts, the wind power signals in two frequency ranges are simply divided by the response speed dividing points of energy type and power type energy storage devices, and the wind power frequency band cannot be accurately determined around the response frequency of the energy storage devices. Due to the poor adaptation characteristics of wavelet decomposition, scholars have proposed Empirical Mode Decomposition (EMD) methods that can be used to process nonlinear and non-stationary signals with the advantage of full adaptation. The learner proposes a smooth wind power fluctuation method adopting Empirical Mode Decomposition (EMD), but the requirement of intermittent energy power fluctuation change is not fully considered, and the EMD has a certain modal aliasing phenomenon. Some learners apply an energy storage system control method of fuzzy clustering empirical mode decomposition (EEMD), but the EEMD has the problems of overlarge data calculation amount and the like. The wavelet decomposition, EMD and EEMD methods have the problems of poor adaptability, modal aliasing and large calculation amount respectively. The Variable Mode Decomposition (VMD) is a new signal self-adaptive decomposition estimation method, and the VMD converts a signal into a non-recursive and variable mode decomposition mode which is a plurality of self-adaptive wiener filtering groups and shows better noise robustness. Compared with the three decomposition methods, the VMD has good robustness, high operation efficiency, solid theoretical basis and excellent performance in the aspects of modal separation and reconstruction of signals with similar frequencies. In the wind power signal, the VMD may adaptively decompose the wind power signal into K frequency band signals. Therefore, the scholars apply the hybrid energy storage device to the power of decomposing wind power and photovoltaic power so as to utilize the hybrid energy storage device to stabilize the fluctuation of the wind power and the photovoltaic power. The method is characterized in that a scholars carries out variational modal decomposition on photovoltaic original power in a self-adaptive mode by combining a photovoltaic power fluctuation standard and the characteristics of an energy storage element, so that primary power distribution is realized, but only the photovoltaic power fluctuation standard is used as a constraint condition of a first layer of VMD decomposition to carry out K value self-adaptive selection, but the remaining medium-frequency signal and high-frequency signal of photovoltaic are also divided by dividing points of different response speeds of an energy storage device and cannot be accurately adapted to the response speed of the energy storage device, and the phenomenon that the energy storage device with low response speed cannot timely stabilize photovoltaic fluctuation may exist. When the proposed decomposition method is suitable for the hybrid energy storage device, the decomposed signal is divided into ranges by the dividing points of different response speeds of the energy storage device, and the decomposed signal is not determined around the center frequency suitable for the energy storage device. In the electric heating comprehensive system, due to the characteristics of wind power fluctuation and randomness, wind power needs to be decomposed in a self-adaptive manner according to the response frequency of the energy storage device and the electric boiler, so that the advantage of stabilizing the wind power fluctuation of the energy storage device and the electric boiler can be better played.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention provides the multi-time scale economic dispatching method of the electric heating integrated system based on the improved VMD, and wind power is decomposed into frequency bands which are adaptive to the hybrid energy storage devices and the electric boiler with different response speeds, so that the power distribution of the hybrid energy storage devices and the electric boiler is formulated, and the air abandoning amount and the load shedding amount can be effectively reduced.

The invention is realized by the following technical scheme: the multi-time scale economic dispatching method of the electric heating integrated system based on the improved VMD is characterized by comprising the following steps of:

step 1: the VMD is improved into an adaptive decomposition method capable of being based on a specified center frequency;

step 2: determining the central frequency according to the long-term wind power trend, the stored energy and the response speed of the thermal part, and performing 3-layer decomposition on the wind power signal by adopting the improved VMD in the step 1;

and step 3: and (3) establishing a multi-time scale scheduling model of the electric heating comprehensive system containing the mixed energy storage based on the improved VMD method in the step (1) and the wind power signal 3-layer decomposition result in the step (2), and formulating the output of the unit, the energy storage device and the electric boiler which are adaptive to different wind power frequency bands.

The step 1 comprises the following specific steps:

step 1.1: with centre frequency ω being driven by the speed of response of the energy storage device or other coupling device of the electrothermal system_kSet to a known quantity with the goal of knowing the center frequency ω of each mode_kIn this case, the sum of the estimated bandwidths of each mode is minimized, i.e., the sparsity is minimized, and the objective function is:

in the formula: { u_k}＝{u₁,…,u_kThe method comprises the steps of (1) setting all sub-modes as a set; { omega [ [ omega ] ]_k}＝{ω₁,…,ω_kIs the set of corresponding center frequencies, center frequency ω_kIs a constant;

step 1.2: using a quadratic penalty factor alpha and a Lagrange multiplier lambda^v(t) changing the constrained variance problem into an unconstrained variance problem, where α guarantees the reconstruction accuracy of the signal, λ^v(t) keeping the strictness of the constraint condition, and expanding Lagrangian expressions as follows:

step 1.3: the above variational problem is solved by adopting an alternative direction multiplier method, and u is updated alternately_k ⁿ⁺¹And λ^v,n+1Seeking an optimal point to extend the Lagrangian expression, where u_k ⁿ⁺¹The transform to the frequency domain can be done using a fourier equidistant transform:

step 1.4: transforming the formula in the step 1.3 to a frequency domain by utilizing Fourier equidistant transformation to obtain a solution of a secondary optimization problem:

in the formula:corresponding to the current residual amount

Wiener filtering of (1); to pair

Performing inverse Fourier transform to obtain real part of u_k(t) }, center frequency ω_kIs a constant.

The multi-time scale scheduling model of the electric heating integrated system containing the hybrid energy storage in the step 3 comprises the following steps:

step 3.1 scheduling model day ahead: day-ahead scheduling is a scheduling plan 24h before the prediction time, and mainly passes through the data P of the layer 1 data in the day-ahead of wind power^D,W0,U1Setting the starting and stopping of a conventional unit, and the initial output of the conventional unit, the cogeneration unit and the electric boiler,

the method is characterized in that the cogeneration unit is set to be in a normally open state, only the operation cost of the cogeneration unit is considered, and the operation cost of the cogeneration unit, the starting cost of a conventional unit, the operation cost of the conventional unit, the operation cost of an electric boiler and the wind abandoning cost are taken as objective functions, and the formula is as follows:

min F^D＝min(C_NC ^D+C_NG，s ^D+C_NG，p ^D+C_EB ^D+C_loss ^D)

wherein

In the formula: f^DExpressed as the total day-ahead cost of the system, C_NC ^DExpressed as a function of the running cost of the cogeneration unit, C, day ahead_NG,s ^DExpressed as a conventional unit start-up cost function, C_NG,p ^DExpressed as a function of the running cost of the day-ahead conventional unit, C_EB ^DExpressed as a function of the running cost of the electric boiler before day, C_loss ^DExpressed as a curtailment cost function; a is_i ^chp，b_i ^chp，c_i ^chpExpressed as the cost coefficient of the ith cogeneration unit; a is_i，b_i，c_iExpressed as the ith conventional unit cost coefficient; CS_i,tThe starting cost coefficient of the ith thermal power generating unit at the time t is represented; u shape_i,t ^DThe starting and stopping states of the ith thermal power generating unit at the moment t before the day are shown, wherein 1 is starting and 0 is stopping; p_i,t ^D,chp，P_i,t ^D，P_i,t ^D,EB，P_i,t ^D,W0,U1And P_i,t ^D,W,U1Respectively expressing the output of the ith cogeneration unit, the output of a conventional unit, the output of an electric boiler, the predicted first-layer active output of the wind power and the actually-scheduled active output of the first-layer wind power at the moment t before the day; NC, NG, EB and NW are the number of the cogeneration unit, the conventional unit, the electric boiler and the wind farm, T is the scheduling period,

the constraint conditions include:

1) electric power balance constraint

In the formula: p_t ^D,LExpressed as total electrical load power at time t before day;

2) heat power balance condition

In the formula: q_i,t ^D,chp，Q_i,t ^D,EBExpressed as the thermal power, Q, generated by the ith cogeneration unit and the electric boiler_t ^D,LExpressed as total thermal load power at time t day ahead,

3) upper and lower limits of unit output and climbing restraint

P_i,min ^chp≤P_i,t ^chp≤P_i,max ^chp

U_i,tP_i,min≤P_i,t≤U_i,tP_i,max

-R_D,i ^chpΔt≤P_i,t ^chp-P_i,t-1 ^chp≤R_U,i ^chpΔt

-R_D,iΔt≤P_i,t-P_i,t-1≤R_U,iΔt

In the formula: p_i,min ^chpAnd P_i,max ^chpRespectively the lower limit and the upper limit of the electric power of the ith cogeneration unit, P_i,minAnd P_i,maxRespectively the lower limit and the upper limit of the electric power of the ith conventional unit, R_U,i ^chpAnd R_D,i ^chpRespectively represent the upper limit of the electric power of the ith cogeneration unit, R_U,iAnd R_D,iRespectively representing the upper limit of the electric power up-down climbing of the ith conventional unit,

4) wind power output constraint

0≤P_i,t ^D,W,U1≤P_i,t ^D,W0,U1

5) Cogeneration thermoelectric coupling constraints

Q_i,t ^chp＝η_chpP_i,t ^chp

Q_i,min ^chp≤Q_i,t ^chp≤Q_i,max ^chp

In the formula: eta_chpExpressing the thermoelectric ratio of the cogeneration unit, the invention takes eta_i,chp＝0.75，Q_i,min ^chpAnd Q_i,max ^chpRespectively is the lower limit and the upper limit of the thermal power of the ith cogeneration unit,

6) electric boiler restraint

Q_i,t ^EB＝η_EBP_i,t ^EB

P_i,min ^EB≤p_i,t ^EB≤P_i,max ^EB

-R_D,i ^EBΔt≤P_i,t ^EB-P_i,t-1 ^EB≤R_U,i ^EBΔt

In the formula: eta_EBExpressing the thermal efficiency of the electric boiler, the invention takes eta_EB＝0.98，P_i,min ^EBAnd P_i,max ^EBRespectively the lower limit and the upper limit of the electric power of the ith electric boiler, R_U,i ^EBAnd R_D,i ^EBRespectively representing the upper limit of the electric power up-down climbing of the ith electric boiler;

step 3.2, rolling the correction model within a day: the in-day correction is a plan 4 h-15 min before a predicted point, and is mainly based on a decision result in the day ahead and first-layer data P in a wind power day^S,W0,U1Setting the output regulating quantity of the set and the electric boiler and the second layer wind power day data P^S,W0,U2Making charging and discharging plan of energy type electricity storage device and output regulating quantity of machine set and electric boiler,

taking the total system cost as an objective function, the formula is as follows:

min F^S＝min(C_NC ^S+C_NG，s ^S+C_NG，p ^S+C_EB ^S+C_loss ^S)

wherein

In the formula: delta P_i,t ^chp、ΔP_i,tAnd Δ P_i,t ^EBExpressed as the electric power regulating quantities of the cogeneration unit, the conventional unit and the electric boiler,

the constraint conditions include:

1) electric power balance constraint

In the formula: p_t ^S,LExpressed as total electrical load power at time t of day, P_t ^S,WExpressed as the actual wind power dispatch power in the day, P_i,t ^ESS,nExpressed as the i-th energy storage device charging and discharging power, P_i,t ^ESS,n>0 represents charging, P_i,t ^ESS,n<0 represents the discharge of the electric current and,

2) heat power balance condition

In the formula: delta Q_i,t ^chpAnd Δ Q_i,t ^EBExpressed as the thermal power regulating quantity, Q, of the electric boiler of the cogeneration unit_t ^S,LExpressed as total thermal load power at time t of the day,

3) unit and electric boiler output correction amount constraint

Comprehensively considering the maximum slope climbing of the monotonous time interval delta t and the upper and lower limits of the output of the unit and the electric boiler to make constraints, the constraint is expressed as follows:

4) restraint of electricity storage device

In the formula: u shape_i,t ^chAnd U_i,t ^dcRespectively indicate the ith power storage device at tCharging state and discharging state, wherein 1 is working state, and 0 represents non-working state; p_i,t ^chAnd P_i,t ^dcRespectively representing the charging power and the discharging power of the ith power storage device at the time t; p_i,min ^chAnd P_i,max ^chRespectively representing a lower and an upper charging power limit, P_i,min ^dcAnd P_i,max ^dcRespectively representing a lower limit and an upper limit of the discharge power, E_i,t ^ESSExpressed as capacity of the i-th power storage device at time t, E_i,min ^ESSAnd E_i,max ^ESSRespectively represents the lower limit and the upper limit of the capacity of the i-th power storage device, beta_chAnd beta_dcRespectively representing a charge coefficient and a discharge coefficient,

wind power output constraint, combined heat and power generation electric heating coupling constraint and electric boiler constraint are consistent with the step 3.1;

step 3.3, real-time model correction: the real-time correction plan is made 15min before the prediction point, and the second layer data P is real-time by the wind power^F,W0,U2And setting the charging and discharging regulating quantity of the energy type electricity storage device and the output regulating quantity of the unit and the electric boiler. Then real-time third-layer data P of wind power^F,W0,U3And making a charging and discharging plan of the power type electricity storage device.

Real-time second-layer data P by wind power^F,W0,U2The objective function for formulating the charging and discharging adjustment quantity of the energy type electricity storage device and the output adjustment quantity of the unit and the electric boiler is similar to the scheduling model in the step 3.2 day, the total cost of the system is minimum, and the real-time third-layer data P of the wind power is generated^F,W0,U3When a charging and discharging plan of the power type energy storage device is formulated, the air abandoning quantity and the load shedding quantity are minimum, namely the charging and discharging quantity of the power type energy storage device is maximum as an objective function, and the formula is as follows: min F^F＝min[P^F,W0,U3(P^dc-P^ch)]

The constraint conditions include:

real-time third layer power P of wind power^F,W0,U3Balance constraint

In the formula: p_i,t ^F,W,U3Expressed as the real-time third-layer actual scheduling power of wind power, P_i,t ^ESS,gExpressed as the i-th energy storage device charging and discharging power, P_i,t ^ESS,g>0 represents charging, P_i,t ^ESS,g<0 represents the discharge of the electric current and,

and (3) keeping the wind power output constraint, the combined heat and power generation electric-thermal coupling constraint and the electric boiler constraint in the same step 3.1.

The invention has the beneficial effects that: the invention relates to an improved VMD-based multi-time scale economic dispatching method of an electric heating integrated system, which comprises the following steps of firstly, improving VMD into a self-adaptive decomposition method capable of being based on specified central frequency; then establishing a multi-time scale scheduling model of the electric heating comprehensive system containing mixed energy storage, determining the central frequency according to the long-term wind power trend, the energy storage and the response speed of the thermal part, and performing 3-layer decomposition on the wind power signal by adopting an improved VMD (variable mode decomposition); the decomposition result of the 1 st layer is matched with the long-term trend of wind power and is used for formulating the unit start and stop, the unit and the electric boiler primarily output, the frequency of the decomposition result of the 2 nd layer is adaptive to the response speed of the energy type energy storage device and is used for determining the charging and discharging plan of the energy storage device and the output regulating quantity of the unit and the electric boiler, the frequency of the decomposition result of the 3 rd layer is adaptive to the power type energy storage device and is used for formulating the charging and discharging plan of the energy storage device, and then the daily and real-time scheduling models are. The method can decompose the wind power into the frequency bands which are adaptive to the hybrid energy storage devices and the electric boiler with different response speeds, so that the power distribution of the hybrid energy storage devices and the electric boiler is formulated, and the air abandoning amount and the load shedding amount can be effectively reduced.

Description of the drawings:

FIG. 1 is a schematic diagram of a raw wind power signal;

FIG. 2 is a schematic diagram of a VMD decomposition signal before wind power layer 3 improvement;

FIG. 3 is a schematic diagram of a VMD decomposition signal after wind power layer 3 improvement;

FIG. 4 is a diagram of the relationship between the wind power frequency band and the devices with different response speeds in the electric heating integrated system;

FIG. 5 is a flow diagram of a multi-time scale rolling scheduling of an electrothermal synthesis system with improved VMD decomposition;

FIG. 6 is a view showing an example structure;

FIG. 7 is a comparison of the air loss and load shedding of Case1 and Case 2;

FIG. 8 shows the wind power U2 and the charging and discharging conditions of a storage battery after the real-time layer 2 data of wind power is adjusted by a set before the Case1 improves the VMD and after the Case2 improves the VMD;

FIG. 9 shows the charging and discharging conditions of the wind power real-time layer 3 data U3 and the super capacitor before the Case1 improves the VMD and after the Case2 improves the VMD;

FIG. 10 is a comparison of the air loss and load shedding of Case2 and Case 3;

FIG. 11 shows the charging and discharging conditions of the wind power U2 and the storage battery of Case2, the charging and discharging conditions of the wind power U3 and the super capacitor of Case2, and the charging and discharging conditions of the wind power U2 and U3 and the storage battery of the Case3 after the wind power is adjusted by a generator set;

FIG. 12 is a comparison of the air loss and load shedding of Case2 and Case 4.

The specific implementation mode is as follows:

the invention is further explained below with reference to the drawings.

The multi-time scale economic dispatching method of the electric heating integrated system based on the improved VMD comprises the following steps:

step 1: the VMD is improved into an adaptive decomposition method capable of being based on a specified center frequency, and the specific steps are as follows:

in the electric heating comprehensive system, the energy type energy storage device, the power type energy storage device and the electric boiler have different response speeds, so that the wind power frequency ranges adapted to the energy type energy storage device, the power type energy storage device and the electric boiler are different. And the VMD decomposition is completely self-adaptive decomposition according to signal characteristics, and the decomposed signal center frequency is possibly not suitable for the response speed of equipment, so that an energy storage device with low response speed cannot absorb high-frequency-band abandoned wind. Therefore, the invention provides an improvement on VMD decomposition, and takes the response frequency of an energy storage device or other coupling devices in an electrothermal system as a central frequency, so that the VMD is decomposed adaptively according to the signal characteristics on the basis of the specified central frequency.

Step 1.1: with centre frequency ω being driven by the speed of response of the energy storage device or other coupling device of the electrothermal system_kSet to a known quantity with the goal of knowing the center frequency ω of each mode_kUnder the circumstancesThe sum of the estimated bandwidths of each mode is minimized, that is, the sparsity is minimized, and the objective function is as follows:

in the formula: { u_k}＝{u₁,…,u_kThe method comprises the steps of (1) setting all sub-modes as a set; { omega [ [ omega ] ]_k}＝{ω₁,…,ω_kIs the set of corresponding center frequencies, center frequency ω_kIs a constant;

step 1.2: using a quadratic penalty factor alpha and a Lagrange multiplier lambda^v(t) changing the constrained variance problem into an unconstrained variance problem, where α guarantees the reconstruction accuracy of the signal, λ^v(t) keeping the strictness of the constraint condition, and expanding Lagrangian expressions as follows:

step 1.3: the above variational problem is solved by adopting an alternative direction multiplier method, and u is updated alternately_k ⁿ⁺¹And λ^v,n+1Seeking an optimal point to extend the Lagrangian expression, where u_k ⁿ⁺¹The transform to the frequency domain can be done using a fourier equidistant transform:

step 1.4: transforming the formula in the step 1.3 to a frequency domain by utilizing Fourier equidistant transformation to obtain a solution of a secondary optimization problem:

in the formula:

corresponding to the current residual amountWiener filtering of (1); to pair

Performing inverse Fourier transform to obtain real part of u_k(t) }, center frequency ω_kIs a constant.

Step 2: determining the central frequency according to the long-term wind power trend, the stored energy and the response speed of the thermal part, and performing 3-layer decomposition on the wind power signal by adopting the improved VMD in the step 1, wherein the specific steps are as follows:

in the electric heating comprehensive system, the response frequency is determined according to the difference of the long-term trend of wind power, the energy storage and the response speed of an electric boiler and is used as the central frequency, the improved VMD is adopted to carry out 3-layer decomposition on the wind power signal, and meanwhile, the improved VMD corresponds to the low frequency band, the medium frequency band and the high frequency band of the wind power signal. The decomposed wind power signal of the 1 st layer is mainly used for starting and stopping and outputting power of a unit, the wind power signal of the 2 nd layer is used for making a plan by using an energy storage device with slow response time and an electric boiler, and the wind power signal of the third layer is used for making a plan by using an energy storage device (super capacitor) with fast response time.

And respectively decomposing the wind power signals by utilizing a VMD method before improvement and a VMD method after improvement, wherein the data source is operation data disclosed by an Elia Belgium power operator, and the sampling resolution is 15 min. The wind power data curve for the first 500 samples beginning at 5.7.2019 is shown in fig. 1. The 3-layer decomposition signal obtained by the VMD decomposition before the improvement is shown in fig. 2, and the 3-layer decomposition signal obtained by the VMD decomposition after the improvement is shown in fig. 3.

The energy storage device response characteristics are compared as shown in table 1. The results of the VMD decomposition before the modification of FIG. 2 gave three layers with center frequencies of 4.1221e-4, 0.0114 and 0.0416, respectively. The time constant reflected by the central frequency 4.1221e-4 of the first layer of the wind power signal output by the VMD before improvement is less than 60min, and the time constant reflected by the central frequency 0.0114 of the second layer is less than 10min, so that the decomposed wind power signal shows the minimum bandwidth of all modal estimation, but the decomposed wind power signal cannot be well adapted to the time constant of the energy storage device, and the energy storage device may not timely stabilize wind power fluctuation.

TABLE 1 energy storage device characterization comparison

In fig. 3, the wind power center frequencies of the first layer, the second layer and the third layer of the wind power are set to be 2.78e-6,5e-4 and 0.05 respectively through the improved VMD decomposition. The decomposed wind power signal is adaptive to the long-term trend of wind power, energy type energy storage and power type energy storage response time constant. The improved VMD method is not only suitable for the response speed of the energy storage device and the electric boiler in the electric heating comprehensive system, but also suitable for equipment with other response speeds in the energy Internet, the response frequency of the energy storage device can be determined according to the response speed of other types of energy storage devices, the determined response frequency is used as the central frequency, and the wind power is decomposed into the frequency band corresponding to the central frequency so as to be suitable for the response speed of the equipment, so that the improved VMD method has universality and adaptability. Fig. 4 shows the relationship between the wind power frequency band and the devices with different response speeds in the electrothermal synthesis system.

Different devices in the electric heating system have different response speeds, the super capacitor, the superconducting energy storage and the flywheel energy storage have high response speeds, the storage battery, the pumped storage and the electric boiler are replaced by the conventional unit and the thermoelectric unit, and the devices with different response speeds form a system, so that the wind power change frequency bands which can be followed are also different, and the improved VMD is required to decompose the wind power at the specified central frequency according to the response speed. As shown in fig. 4, when the wind power frequency is between (0, 6.94e-4), the frequency band corresponds to the long-term wind power trend component and is used for determining the center frequency of the layer 1 decomposed by the improved VMD, and the frequency band can be used for making a start-stop plan and a preliminary output of a conventional unit and a thermoelectric unit; when the wind power frequency is between (6.94e-4, 4.16e-4), the frequency band corresponds to the wind power fluctuation component and is used for determining that the improved VMD decomposes the central frequency of the layer 2, and the storage battery, the pumped storage and the electric boiler with low response speed stabilize the wind power fluctuation in the frequency band; when the wind power frequency is between (4.16e-4, infinity), the frequency band corresponds to the wind power random component and is used for determining the center frequency of the layer 3 decomposed by the improved VMD, the frequency band is a high frequency band, and the influence of the wind power randomness is required to be reduced by the super capacitor, the superconducting energy storage and the flywheel energy storage with high response speed.

And step 3: and (3) establishing a multi-time scale scheduling model of the electric heating comprehensive system containing the mixed energy storage based on the improved VMD method in the step (1) and the wind power signal 3-layer decomposition result in the step (2), and formulating the output of the unit, the energy storage device and the electric boiler which are adaptive to different wind power frequency bands. The method comprises the following specific steps:

the scheduling model of the invention is divided into 3 time scales: day-ahead scheduling model ([ T, T + T)]) And intra-day rolling correction model ([ t.t +16 ]]) And real-time correction model ([ t, t + 1)]) 15min is used as 1 scheduling period. The first part firstly decomposes the predicted power data P of the wind power day ahead through the improved VMD proposed by the invention^D,W0Divided into 3 layers, and composed of the first layer data P before wind power day^D,W0,U1Setting the start and stop of the unit and the initial output of the unit and the electric boiler; the second part is to decompose the wind power prediction power data P in the day by improving VMD^{S ,W0}Dividing into 3 layers, and using the data P of the 1 st layer in the wind power day^S,W0,U1Setting the output regulating quantity of the generator set and the electric boiler and the data P of the 2 nd layer in the wind power day^S,W0,U2Making a charging and discharging plan of the energy type electricity storage device and output adjustment quantities of a unit and an electric boiler; the third part is to decompose the wind power real-time prediction data P by improving VMD^F,W0Divided into 3 layers and processed by wind power real-time layer 2 data P^F,W0,U2Setting up the charging and discharging regulating quantity of the energy type electricity storage device and the output regulating quantity of the unit and the electric boiler, and using the real-time data P of the 3 rd layer of the wind power^F,W0,U3And making a charging and discharging plan of the power type electricity storage device. Fig. 5 shows the overall flow of multi-time scale rolling scheduling.

Step 3.1 scheduling model day ahead: day-ahead scheduling is a scheduling plan 24h before the prediction time, and mainly passes through the data P of the layer 1 data in the day-ahead of wind power^D,W0,U1Setting the starting and stopping of a conventional unit, and the initial output of the conventional unit, the cogeneration unit and the electric boiler,

the cogeneration unit set by the invention is set to be in a normally open state, only the running cost of the cogeneration unit is considered, and the running cost of the cogeneration unit, the starting cost of a conventional unit, the running cost of the conventional unit, the running cost of an electric boiler and the cost of abandoned wind are taken as objective functions, and the formula is as follows:

min F^D＝min(C_NC ^D+C_NG，s ^D+C_NG，p ^D+C_EB ^D+C_loss ^D)

wherein

In the formula: f^DExpressed as the total day-ahead cost of the system, C_NC ^DExpressed as a function of the running cost of the cogeneration unit, C, day ahead_NG,s ^DExpressed as a conventional unit start-up cost function, C_NG,p ^DExpressed as a function of the running cost of the day-ahead conventional unit, C_EB ^DExpressed as a function of the running cost of the electric boiler before day, C_loss ^DExpressed as a curtailment cost function; a is_i ^chp，b_i ^chp，c_i ^chpExpressed as the cost coefficient of the ith cogeneration unit; a is_i，b_i，c_iExpressed as the ith conventional unit cost coefficient; CS_i,tThe starting cost coefficient of the ith thermal power generating unit at the time t is represented; u shape_i,t ^DThe starting and stopping states of the ith thermal power generating unit at the moment t before the day are shown, wherein 1 is starting and 0 is stopping; p_i,t ^D,chp，P_i,t ^D，P_i,t ^D,EB，P_i,t ^D,W0,U1And P_i,t ^D,W,U1Respectively expressing the output of the ith cogeneration unit, the output of a conventional unit, the output of an electric boiler, the predicted first-layer active output of the wind power and the actually-scheduled active output of the first-layer wind power at the moment t before the day; NC, NG, EB and NW are the number of the cogeneration unit, the conventional unit, the electric boiler and the wind power plant, and T is the scheduling period.

The constraint conditions include:

1) electric power balance constraint

In the formula: p_t ^D,LExpressed as total electrical load power at time t of day.

2) Heat power balance condition

In the formula: q_i,t ^D,chp，Q_i,t ^D,EBExpressed as the thermal power, Q, generated by the ith cogeneration unit and the electric boiler_t ^D,LExpressed as total heat load power at time t day ahead.

3) Upper and lower limits of unit output and climbing restraint

P_i,min ^chp≤P_i,t ^chp≤P_i,max ^chp

U_i,tP_i,min≤P_i,t≤U_i,tP_i,max

-R_D,i ^chpΔt≤P_i,t ^chp-P_i,t-1 ^chp≤R_U,i ^chpΔt

-R_D,iΔt≤P_i,t-P_i,t-1≤R_U,iΔt

In the formula: p_i,min ^chpAnd P_i,max ^chpRespectively the lower limit and the upper limit of the electric power of the ith cogeneration unit, P_i,minAnd P_i,maxRespectively the lower limit and the upper limit of the electric power of the ith conventional unit, R_U,i ^chpAnd R_D,i ^chpRespectively represent the upper limit of the electric power of the ith cogeneration unit, R_U,iAnd R_D,iRespectively representing the upper limit of the electric power up-down climbing of the ith conventional unit.

4) Wind power output constraint

0≤P_i,t ^D,W,U1≤P_i,t ^D,W0,U1

5) Cogeneration thermoelectric coupling constraints

Q_i,t ^chp＝η_chpP_i,t ^chp

Q_i,min ^chp≤Q_i,t ^chp≤Q_i,max ^chp

In the formula: eta_chpExpressing the thermoelectric ratio of the cogeneration unit, the invention takes eta_i,chp＝0.75，Q_i,min ^chpAnd Q_i,max ^chpThe lower limit and the upper limit of the thermal power of the ith cogeneration unit are respectively set.

6) Electric boiler restraint

Q_i,t ^EB＝η_EBP_i,t ^EB

P_i,min ^EB≤p_i,t ^EB≤P_i,max ^EB

-R_D,i ^EBΔt≤P_i,t ^EB-P_i,t-1 ^EB≤R_U,i ^EBΔt

In the formula: eta_EBExpressing the thermal efficiency of the electric boiler, the invention takes eta_EB＝0.98，P_i,min ^EBAnd P_i,max ^EBRespectively the lower limit and the upper limit of the electric power of the ith electric boiler, R_U,i ^EBAnd R_D,i ^EBRespectively representing the upper limit of the electric power up-down climbing of the ith electric boiler.

Step 3.2, rolling the correction model within a day: the in-day correction is a plan 4 h-15 min before a predicted point, and is mainly based on a decision result in the day ahead and first-layer data P in a wind power day^S,W0,U1Setting the output regulating quantity of the set and the electric boiler and the second layer wind power day data P^S,W0,U2Making charging and discharging plan of energy type electricity storage device and output regulating quantity of machine set and electric boiler,

taking the total system cost as an objective function, the formula is as follows:

min F^S＝min(C_NC ^S+C_NG，s ^S+C_NG，p ^S+C_EB ^S+C_loss ^S)

wherein

In the formula: delta P_i,t ^chp、ΔP_i,tAnd Δ P_i,t ^EBExpressed as the electric power adjustment quantities of the cogeneration unit, the conventional unit and the electric boiler, the constraint conditions include:

1) electric power balance constraint

In the formula: p_t ^S,LExpressed as total electrical load power at time t of day, P_t ^S,WExpressed as the actual wind power dispatch power in the day, P_i,t ^ESS,nExpressed as the i-th energy storage device charging and discharging power, P_i,t ^ESS,n>0 represents charging, P_i,t ^ESS,n<0 represents discharge.

2) Heat power balance condition

In the formula: delta Q_i,t ^chpAnd Δ Q_i,t ^EBExpressed as the thermal power regulating quantity, Q, of the electric boiler of the cogeneration unit_t ^S,LExpressed as total heat load power at time t of day

3) Unit and electric boiler output correction amount constraint

Comprehensively considering the maximum slope climbing of the monotonous time interval delta t and the upper and lower limits of the output of the unit and the electric boiler to make constraints, the constraint is expressed as follows:

4) restraint of electricity storage device

In the formula: u shape_i,t ^chAnd U_i,t ^dcRespectively showing the ith station power storage deviceSetting the charging state and the discharging state at the moment t, wherein 1 is the working state, and 0 represents the non-working state; p_i,t ^chAnd P_i,t ^dcRespectively representing the charging power and the discharging power of the ith power storage device at the time t; p_i,min ^chAnd P_i,max ^chRespectively representing a lower and an upper charging power limit, P_i,min ^dcAnd P_i,max ^dcRespectively representing a lower limit and an upper limit of the discharge power, E_i,t ^ESSExpressed as capacity of the i-th power storage device at time t, E_i,min ^ESSAnd E_i,max ^ESSRespectively represents the lower limit and the upper limit of the capacity of the i-th power storage device, beta_chAnd beta_dcRepresenting the charge and discharge coefficients, respectively.

Wind power output constraint, combined heat and power generation electric heating coupling constraint and electric boiler constraint are consistent with the step 3.1;

step 3.3, real-time model correction: the real-time correction plan is made 15min before the prediction point, and the second layer data P is real-time by the wind power^F,W0,U2And setting the charging and discharging regulating quantity of the energy type electricity storage device and the output regulating quantity of the unit and the electric boiler. Then real-time third-layer data P of wind power^F,W0,U3And making a charging and discharging plan of the power type electricity storage device.

Real-time second-layer data P by wind power^F,W0,U2And (3) establishing an objective function of the charging and discharging adjustment quantity of the energy type electricity storage device and the output adjustment quantity of the unit and the electric boiler, wherein the objective function is similar to the scheduling model in the step 3.2, and the total cost of the system is the minimum. Wind power real-time third-layer data P^F,W0,U3When a charging and discharging plan of the power type energy storage device is formulated, the air abandoning quantity and the load shedding quantity are minimum, namely the charging and discharging quantity of the power type energy storage device is maximum as an objective function, and the formula is as follows: min F^F＝min[P^F,W0,U3(P^dc-P^ch)]

The constraint conditions include:

real-time third layer power P of wind power^F,W0,U3Balance constraint

In the formula: p_i,t ^F,W,U3Expressed as the real-time third-layer actual scheduling power of wind power, P_i,t ^ESS,gExpressed as the i-th energy storage device charging and discharging power, P_i,t ^ESS,g>0 represents charging, P_i,t ^ESS,g<0 represents discharge. .

And (3) keeping the wind power output constraint, the combined heat and power generation electric-thermal coupling constraint and the electric boiler constraint in the same step 3.1.

The modified IEEE-39 node system is selected for verification and comprises 2 cogeneration units, 8 thermal power units, 1 electric boiler, 1 wind power plant, 1 energy type electricity storage device and 1 power type electricity storage device. The energy type electricity storage device is a storage battery, and the power type electricity storage device is a super capacitor. The wind power plant data come from 2019, 8 months and 7 days of operation data disclosed by Elia wind power plants in Belgium, and are normalized according to the installed capacity of 2000 MW. The power storage device parameters are shown in table 2. The energy storage device is connected to node 38 and the power storage device is connected to node 39. The structure of the example is shown in figure 6.

TABLE 2 electric storage device parameters

In order to illustrate the influence of the improved VMD decomposition method and the response speed of the hybrid energy storage device on the air abandoning amount and the load shedding amount, the influence of the single energy storage device and the hybrid energy storage device on the air abandoning amount and the load shedding amount and the influence of the comparative electric boiler on the energy storage capacity in the electric heating integrated system, the following 4 cases are set:

1) case 1: a VMD decomposition method before improvement is adopted, and a scheduling model consists of day-ahead, day-in and real-time scheduling models. The electric storage device is a storage battery and a super capacitor, and the system is not provided with an electric boiler. The method comprises the steps that the wind power first layer data is used for formulating unit starting and stopping, unit and electric boiler preliminary output, the wind power second layer data is used for formulating storage battery charging and discharging plans and unit and electric boiler regulating quantity, the wind power third layer data is used for formulating super capacitor charging and discharging plans, and then the super capacitor charging and discharging plans are modified through a daily and real-time scheduling model.

2) Case 2: and adopting an improved VMD decomposition method, wherein the scheduling model consists of a day-ahead scheduling model, a day-in scheduling model and a real-time scheduling model. The electric storage device consists of a storage battery and a super capacitor, and the system is not provided with an electric boiler. The scheduling plan is the same as Case 1.

3) Case 3: and adopting an improved VMD decomposition method, wherein the scheduling model consists of a day-ahead scheduling model and a day-within scheduling model. The electric storage device is a single energy type electric storage device (storage battery), and the system is not provided with an electric boiler. The wind power first layer data is used for formulating the unit starting and stopping, the unit and the electric boiler preliminary output, the wind power second layer data and the third layer data are used for formulating the storage battery charging and discharging plan and the unit and electric boiler regulating quantity, and then the daily model is used for correcting.

4) Case 4: and adopting an improved VMD decomposition method, wherein the scheduling model consists of a day-ahead scheduling model, a day-in scheduling model and a real-time scheduling model. The electric storage device consists of a storage battery and a super capacitor, and the system is provided with an electric boiler. The scheduling plan is the same as Case 1.

Case1 and Case2 analyses:

comparing Case1 and Case2, comparing the influences of wind power decomposition power and energy storage devices with different response speeds on the wind abandoning amount and the load shedding amount before and after the VMD decomposition method is improved by the electric heating system. The air curtailment amount and the load shedding amount of Case1 and Case2 are shown in fig. 7. Wherein positive values represent the air removal amount and negative values represent the load shedding amount. Wind power U2 and storage battery charging and discharging conditions of the real-time layer 2 wind power data after the unit adjustment before the Case1 improves the VMD and after the Case2 improves the VMD are shown in FIG. 8. The charging and discharging conditions of the wind power real-time layer 3 data U3 and the super capacitor before the Case1 improves the VMD and after the Case2 improves the VMD are shown in FIG. 9.

As shown in FIG. 7, after the improved VMD method is used, the sum of the air abandoning amount and the load cutting amount of the system is reduced from 5525.5MW to 448.3MW, which is reduced by 91.8%. The air abandoning amount is reduced from 5512.7MW to 214.1MW, and the load cutting amount is increased from 1.81MW to 234.1 MW. When the VMD before improvement is adopted to decompose the wind power signal, the center frequencies of three layers are decomposed to be 3.1685e-4, 0.0118 and 0.0314 respectively. Before improvement, the time constant reflected by the central frequency 3.1685e-4 of the first layer of the wind power signal decomposed by the VMD is less than 60min, and the time constant reflected by the central frequency 0.0118 of the second layer is less than 10 min. And the data of the layer 1 containing the medium-frequency-range wind power and the data of the layer two containing the high-frequency-range wind power are analyzed. As can be seen from FIG. 8, before improvement, the wind curtailment phenomenon of wind power U2 occurs in the time periods of 1-6h and 12-24h after the real-time layer 2 data of wind power is decomposed by the VMD and is adjusted by the set, and the wind curtailment power exceeds 500MW in most time periods, and the load shedding situation occurs only once. The method is characterized in that before improvement, a VMD decomposes a wind power high-frequency-band signal in a second layer of wind power data, a 1 layer of wind power data contains a medium-frequency-band power, before improvement, the VMD decomposes the 2 layer of wind power data, the wind power fluctuation adaptive to the response speed of a storage battery cannot be accurately reflected, the fluctuation cannot be adaptive to the response speed of the storage battery, the storage battery can only absorb a small part of abandoned wind for charging, and the storage battery discharges in a load cutting period, so that the storage battery cannot absorb the abandoned wind in the frequency band. After improvement, the VMD decomposes that the wind power real-time layer 2 data has abandoned wind in 1-5h and has load shedding in 7-14h, the decomposed wind power can be adapted to the response speed of the storage battery, and the storage battery is charged when the wind is abandoned and discharged when the load shedding is carried out, so that the abandoned wind quantity and the load shedding quantity can be reduced to the maximum extent. As can be seen from FIG. 9, the wind power signals of the 3 rd layer of wind power decomposed by the two methods are adapted to the response speed of the super capacitor, so that the super capacitor can well reduce the wind curtailment amount and the load shedding amount of the frequency band. Therefore, in the electric heating system, the wind power frequency band decomposed by the improved VMD can adapt to the response speed of the storage battery and the super capacitor, and the abandoned wind quantity and the load shedding quantity are reduced to the maximum extent.

Case2 and Case3 analyses:

comparing Case2 and Case3, wind power is decomposed through the improved VMD, and the influences of an electric heating system on the air abandoning amount and the load shedding amount under the conditions of a single power storage device and a mixed power storage device are compared. The air curtailment amounts and the load curtailment amounts of Case2 and Case3 are shown in fig. 10. Wherein positive values represent the air removal amount and negative values represent the load shedding amount. The charging and discharging conditions of the wind power U2 and the storage battery of the real-time 2-layer data of the Case2 wind power after being adjusted by the generator set, the charging and discharging conditions of the wind power U3 and the super capacitor of the real-time 3-layer data of the Case2 after being adjusted by the generator set, and the wind power U2 and U3 and the storage battery of the real-time data of the Case3 after being adjusted by the generator set are shown in fig. 11.

As shown in fig. 10, after the hybrid electric storage device is used, the sum of the system air abandonment amount and the system load shedding amount is reduced from 1229.5MW to 448.3MW, which is reduced by 63.5%. The air reject volume is reduced from 635.2MW to 214.1MW, which is reduced by 66.2%. The load rejection is reduced from 594.2MW to 234.1MW, which is reduced by 60.6%. The air abandoning quantity and the load cutting quantity are greatly reduced. In fig. 11, when t is 3, after charging through the hybrid power storage device, the abandoned wind volume is reduced to 0MW by 140.4MW, and it is thus seen that the single power storage device is not enough to consume abandoned wind, and the hybrid power storage device is charged by different power storage devices with different response speeds after the wind power is divided into an intermediate frequency and a high frequency, and wind power with different frequencies can be timely consumed, and the super capacitor in the graph can consume the high-frequency band wind power at this time when t is 3. When t is 10, the load cut is reduced from 165.8MW to 71MW after the discharge by the hybrid power storage device. Therefore, the single energy type electricity storage device is not enough in discharge to compensate wind power fluctuation, the storage battery is used for completely compensating the wind power fluctuation amount in the wind power intermediate frequency range of the hybrid electricity storage device, and the super capacitor can be timely charged and discharged to compensate wind power insufficiency in the wind power high frequency range. Therefore, in the electric heating system, when wind power is decomposed into frequency bands suitable for the power storage devices with different response speeds, the mixed power storage device can reduce the air abandoning amount and the load shedding amount of the system more than a single power storage device.

Case2 and Case4 analyses:

comparing Case2 with Case4, the effect on the capacity of the hybrid power storage device after the electric boiler is added to the electric heating system is compared. The air curtailment amount and the load shedding amount of Case2 and Case4 are shown in fig. 12. The capacity of a storage battery in Case2 is 2000MW, and the capacity of a super capacitor is 500 MW; the battery capacity in Case4 is 1800MW, and the super capacitor capacity is 500 MW.

As shown in FIG. 12, after the electric boiler is added, the sum of the air abandoning amount and the load cutting amount of the system is reduced from 448.3MW to 145.2MW, which is reduced by 67.6%. As can be seen from FIG. 12, after the electric boiler is added, the original 1-3h abandoned wind can be consumed, so that the heat load is provided for the system, the output of the cogeneration unit is reduced, and the wind power consumption space is increased.

And due to the addition of the electric boiler, the capacity of the energy type power storage device is reduced from 2000MW to 1800 MW. Therefore, after the electric boiler is added into the electric heating system, the air abandoning amount, the load cutting amount and the total system operation cost of the system can be reduced, and the capacity of the electricity storage device can be reduced.

In order to analyze the differences between the total system cost and the air reject rate and the load shedding amount in cases 1 to 4, the total system cost and the sum of the air reject rate and the load shedding amount in cases 1 to 4 are shown in table 4.

TABLE 4 Total cost of Case 1-Case 4 systems

It can be seen from Case1 and Case2 that wind power signals decomposed by wind power after VMD improvement can adapt to the response speed of the energy storage device, so that the wind energy absorption and abandoning capacity of the energy storage device is exerted to a greater extent, and the system cost is reduced. From Case2 and Case3, compared with a single energy storage device, the hybrid energy storage device can not only reduce the air abandoning amount and the load shedding amount, but also reduce the system operation cost. It can be seen from Case3 and Case4 that after the electric boiler is introduced into the electric heating system, the output of the cogeneration unit can be reduced, the wind power consumption space can be increased, the air abandonment quantity and the load shedding quantity can be reduced, and the system cost and the capacity of the energy storage device can be reduced.

The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims (3)

1. The multi-time scale economic dispatching method of the electric heating integrated system based on the improved VMD is characterized by comprising the following steps of:

step 1: the VMD is improved into an adaptive decomposition method capable of being based on a specified center frequency;

step 2: determining the central frequency according to the long-term wind power trend, the stored energy and the response speed of the thermal part, and performing 3-layer decomposition on the wind power signal by adopting the improved VMD in the step 1;

and step 3: and (3) establishing a multi-time scale scheduling model of the electric heating comprehensive system containing the mixed energy storage based on the improved VMD method in the step (1) and the wind power signal 3-layer decomposition result in the step (2), and formulating the output of the unit, the energy storage device and the electric boiler which are adaptive to different wind power frequency bands.

2. The improved VMD-based electric heat integrated system multi-time scale economic scheduling method of claim 1, characterized in that the following specific steps are included in the step 1:

step 1.1: with centre frequency ω being driven by the speed of response of the energy storage device or other coupling device of the electrothermal system_kSet to a known quantity with the goal of knowing the center frequency ω of each mode_kIn this case, the sum of the estimated bandwidths of each mode is minimized, i.e., the sparsity is minimized, and the objective function is:

in the formula: { u_k}＝{u₁,…,u_kThe method comprises the steps of (1) setting all sub-modes as a set; { omega [ [ omega ] ]_k}＝{ω₁,…,ω_kIs the set of corresponding center frequencies, center frequency ω_kIs a constant;

step 1.2: using a quadratic penalty factor alpha and a Lagrange multiplier lambda^v(t) changing the constrained variance problem into an unconstrained variance problem, where α guarantees the reconstruction accuracy of the signal, λ^v(t) keeping the strictness of the constraint condition, and expanding Lagrangian expressions as follows:

step 1.3: the above variational problem is solved by adopting an alternative direction multiplier method, and u is updated alternately_k ⁿ⁺¹And λ^v,n+1Seeking an optimal point to extend the Lagrangian expression, where u_k ⁿ⁺¹The transform to the frequency domain can be done using a fourier equidistant transform:

step 1.4: transforming the formula in the step 1.3 to a frequency domain by utilizing Fourier equidistant transformation to obtain a solution of a secondary optimization problem:

in the formula:

corresponding to the current residual amountWiener filtering of (1); to pair

Performing inverse Fourier transform to obtain real part of u_k(t) }, center frequency ω_kIs a constant.

3. The improved VMD-based electric heating integrated system multi-time scale economic scheduling method of claim 1, wherein the step 3 electric heating integrated system multi-time scale scheduling model with hybrid energy storage comprises the steps of:

step 3.1 scheduling model day ahead: day-ahead scheduling is a scheduling plan 24h before the prediction time, and mainly passes through the data P of the layer 1 data in the day-ahead of wind power^D,W0,U1Setting the starting and stopping of a conventional unit, and the initial output of the conventional unit, the cogeneration unit and the electric boiler,

the method is characterized in that the cogeneration unit is set to be in a normally open state, only the operation cost of the cogeneration unit is considered, and the operation cost of the cogeneration unit, the starting cost of a conventional unit, the operation cost of the conventional unit, the operation cost of an electric boiler and the wind abandoning cost are taken as objective functions, and the formula is as follows:

minF^D＝min(C_NC ^D+C_NG，s ^D+C_NG，p ^D+C_EB ^D+C_loss ^D)

wherein

In the formula: f^DExpressed as the total day-ahead cost of the system, C_NC ^DExpressed as a function of the running cost of the cogeneration unit, C, day ahead_NG,s ^DExpressed as a conventional unit start-up cost function, C_NG,p ^DExpressed as a function of the running cost of the day-ahead conventional unit, C_EB ^DExpressed as a function of the running cost of the electric boiler before day, C_loss ^DExpressed as a curtailment cost function; a is_i ^chp，b_i ^chp，c_i ^chpExpressed as the cost coefficient of the ith cogeneration unit; a is_i，b_i，c_iExpressed as the ith conventional unit cost coefficient; CS_i,tThe starting cost coefficient of the ith thermal power generating unit at the time t is represented; u shape_i,t ^DThe starting and stopping states of the ith thermal power generating unit at the moment t before the day are shown, wherein 1 is starting and 0 is stopping; p_i,t ^D,chp，P_i,t ^D，P_i,t ^D,EB，P_i,t ^D,W0,U1And P_i,t ^D,W,U1Respectively expressing the output of the ith cogeneration unit, the output of a conventional unit, the output of an electric boiler, the predicted first-layer active output of the wind power and the actually-scheduled active output of the first-layer wind power at the moment t before the day; NC, NG, EB and NW are the number of the cogeneration unit, the conventional unit, the electric boiler and the wind farm, T is the scheduling period,

the constraint conditions include:

1) electric power balance constraint

In the formula: p_t ^D,LExpressed as total electrical load power at time t before day;

2) heat power balance condition

In the formula: q_i,t ^D,chp，Q_i,t ^D,EBExpressed as the thermal power, Q, generated by the ith cogeneration unit and the electric boiler_t ^D,LExpressed as total thermal load power at time t day ahead,

3) upper and lower limits of unit output and climbing restraint

P_i,min ^chp≤P_i,t ^chp≤P_i,max ^chp

U_i,tP_i,min≤P_i,t≤U_i,tP_i,max

-R_D,i ^chpΔt≤P_i,t ^chp-P_i,t-1 ^chp≤R_U,i ^chpΔt

-R_D,iΔt≤P_i,t-P_i,t-1≤R_U,iΔt

In the formula: p_i,min ^chpAnd P_i,max ^chpRespectively the lower limit and the upper limit of the electric power of the ith cogeneration unit, P_i,minAnd P_i,maxRespectively the lower limit and the upper limit of the electric power of the ith conventional unit, R_U,i ^chpAnd R_D,i ^chpRespectively represent the upper limit of the electric power of the ith cogeneration unit, R_U,iAnd R_D,iRespectively representing the upper limit of the electric power up-down climbing of the ith conventional unit,

4) wind power output constraint

0≤P_i,t ^D,W,U1≤P_i,t ^D,W0,U1

5) Cogeneration thermoelectric coupling constraints

Q_i,t ^chp＝η_chpP_i,t ^chp

Q_i,min ^chp≤Q_i,t ^chp≤Q_i,max ^chp

In the formula: eta_chpExpressing the thermoelectric ratio of the cogeneration unit, the invention takes eta_i,chp＝0.75，Q_i,min ^chpAnd Q_i,max ^chpRespectively is the lower limit and the upper limit of the thermal power of the ith cogeneration unit,

6) electric boiler restraint

Q_i,t ^EB＝η_EBP_i,t ^EB

P_i,min ^EB≤p_i,t ^EB≤P_i,max ^EB

-R_D,i ^EBΔt≤P_i,t ^EB-P_i,t-1 ^EB≤R_U,i ^EBΔt

In the formula: eta_EBExpressing the thermal efficiency of the electric boiler, the invention takes eta_EB＝0.98，P_i,min ^EBAnd P_i,max ^EBRespectively the lower limit and the upper limit of the electric power of the ith electric boiler, R_U,i ^EBAnd R_D,i ^EBRespectively representing the upper limit of the electric power up-down climbing of the ith electric boiler;

step 3.2, rolling the correction model within a day: the in-day correction is a plan 4 h-15 min before a predicted point, and is mainly based on a decision result in the day ahead and first-layer data P in a wind power day^S,W0,U1Setting the output regulating quantity of the set and the electric boiler and the second layer wind power day data P^S,W0,U2Making charging and discharging plan of energy type electricity storage device and output regulating quantity of machine set and electric boiler,

taking the total system cost as an objective function, the formula is as follows:

minF^S＝min(C_NC ^S+C_NG，s ^S+C_NG，p ^S+C_EB ^S+C_loss ^S)

wherein

In the formula: delta P_i,t ^chp、ΔP_i,tAnd Δ P_i,t ^EBExpressed as combined heat and power unitsThe electric power regulating quantity of the conventional unit and the electric boiler,

the constraint conditions include:

1) electric power balance constraint

In the formula: p_t ^S,LExpressed as total electrical load power at time t of day, P_t ^S,WExpressed as the actual wind power dispatch power in the day, P_i,t ^ESS,nExpressed as the i-th energy storage device charging and discharging power, P_i,t ^ESS,n>0 represents charging, P_i,t ^ESS,n<0 represents the discharge of the electric current and,

2) heat power balance condition

In the formula: delta Q_i,t ^chpAnd Δ Q_i,t ^EBExpressed as the thermal power regulating quantity, Q, of the electric boiler of the cogeneration unit_t ^S,LExpressed as total thermal load power at time t of the day,

3) unit and electric boiler output correction amount constraint

Comprehensively considering the maximum slope climbing of the monotonous time interval delta t and the upper and lower limits of the output of the unit and the electric boiler to make constraints, the constraint is expressed as follows:

4) restraint of electricity storage device

In the formula: u shape_i,t ^chAnd U_i,t ^dcRespectively showing the charging state and the discharging state of the ith power storage device at the time t, wherein 1 is an operating state, and 0 is an inoperative state; p_i,t ^chAnd P_i,t ^dcRespectively representing the charging power and the discharging power of the ith power storage device at the time t; p_i,min ^chAnd P_i,max ^chRespectively representing a lower and an upper charging power limit, P_i,min ^dcAnd P_i,max ^dcRespectively representing a lower limit and an upper limit of the discharge power, E_i,t ^ESSExpressed as capacity of the i-th power storage device at time t, E_i,min ^ESSAnd E_i,max ^ESSRespectively represents the lower limit and the upper limit of the capacity of the i-th power storage device, beta_chAnd beta_dcRespectively representing a charge coefficient and a discharge coefficient,

wind power output constraint, combined heat and power generation electric heating coupling constraint and electric boiler constraint are consistent with the step 3.1;

step 3.3, real-time model correction: the real-time correction plan is made 15min before the prediction point, and the second layer data P is real-time by the wind power^F,W0,U2And setting the charging and discharging regulating quantity of the energy type electricity storage device and the output regulating quantity of the unit and the electric boiler. Then real-time third-layer data P of wind power^F,W0,U3And making a charging and discharging plan of the power type electricity storage device.

Real-time second-layer data P by wind power^F,W0,U2The objective function for formulating the charging and discharging adjustment quantity of the energy type electricity storage device and the output adjustment quantity of the unit and the electric boiler is similar to the scheduling model in the step 3.2 day, the total cost of the system is minimum, and the real-time third-layer data P of the wind power is generated^F,W0,U3When a charging and discharging plan of the power type energy storage device is formulated, the air abandoning quantity and the load shedding quantity are minimum, namely the charging and discharging quantity of the power type energy storage device is maximum as an objective function, and the formula is as follows:

minF^F＝min[P^F,W0,U3(P^dc-P^ch)]

the constraint conditions include:

real-time third layer power P of wind power^F,W0,U3Balance constraint

In the formula: p_i,t ^F,W,U3Expressed as wind powerReal-time third layer actual scheduling power, P_i,t ^ESS,gExpressed as the i-th energy storage device charging and discharging power, P_i,t ^ESS,g>0 represents charging, P_i,t ^ESS,g<0 represents the discharge of the electric current and,

and (3) keeping the wind power output constraint, the combined heat and power generation electric-thermal coupling constraint and the electric boiler constraint in the same step 3.1.

Images