CN115544856A - Day-ahead optimized scheduling method for electric heating integrated energy system - Google Patents

Day-ahead optimized scheduling method for electric heating integrated energy system Download PDF

Info

Publication number
CN115544856A
CN115544856A CN202110727038.9A CN202110727038A CN115544856A CN 115544856 A CN115544856 A CN 115544856A CN 202110727038 A CN202110727038 A CN 202110727038A CN 115544856 A CN115544856 A CN 115544856A
Authority
CN
China
Prior art keywords
power
scheduling
electric heating
unit
day
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110727038.9A
Other languages
Chinese (zh)
Inventor
刘鼎
李婉婷
胡博
刘玉奇
臧传治
曾鹏
顾洪群
罗桓桓
吕旭明
周桂平
王磊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenyang Institute of Automation of CAS
Original Assignee
Shenyang Institute of Automation of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenyang Institute of Automation of CAS filed Critical Shenyang Institute of Automation of CAS
Priority to CN202110727038.9A priority Critical patent/CN115544856A/en
Publication of CN115544856A publication Critical patent/CN115544856A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/08—Probabilistic or stochastic CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/70—Wind energy
    • Y02E10/76—Power conversion electric or electronic aspects

Abstract

The invention discloses a day-ahead optimized scheduling method for an electric heating comprehensive energy system. The method comprises the following steps: analyzing the characteristics of the wind power output data, and establishing a non-Gaussian distribution model of the uncertainty of the wind power output; training and predicting historical data through a GRU network to obtain a load predicted value of each scheduling period in the day ahead; establishing a unit model of the electric heating comprehensive energy system; based on uncertainty of wind power output in a scene tree structure description system, according to a load predicted value in the day ahead and a system unit model, a random model prediction control method is used for obtaining day-ahead scheduling parameter control quantity of the electric heating comprehensive energy system; and outputting the control quantity of the day-ahead scheduling parameters to a power supply and heat supply unit for regulation and control. The multi-time scale optimization scheduling method provided by the invention performs characteristic analysis on wind power output data, predicts the load, performs optimization scheduling on the electric heating comprehensive energy system in the future, saves the operation cost of the system, and ensures the stability and robustness of the system.

Description

Day-ahead optimized scheduling method for electric heating integrated energy system
Technical Field
The invention relates to the field of comprehensive regulation and control of an electric power system and a thermodynamic system, in particular to a day-ahead optimal scheduling method of an electric heating comprehensive energy system.
Background
The electric heating integrated energy system combines an electric power system and a thermodynamic system, meets the requirements of heat load and electric load in the system by coordinately adjusting the output of the thermodynamic system and the electric power system, and is widely applied to the electric heating integrated energy system based on a cogeneration unit especially in the northern area with heating requirement. However, the operation mode of the thermoelectric generator set in the electric heating comprehensive energy system for fixing the power by heat limits the internet access space of renewable energy wind power generation, and the coordination and optimization scheduling of the electric heating comprehensive energy system faces new challenges along with the improvement of the grid-connected proportion of the wind power output.
The wind power output has uncertainty such as randomness, intermittence and volatility, which brings challenges to the safety, reliability and operation of the electric heating comprehensive energy system. The coordinated optimization scheduling of the electric heating comprehensive energy system can be divided into day-ahead scheduling according to the scheduling time scale, a reasonable scheduling optimization scheme can cope with system uncertainty, the operation reliability and safety of the electric heating comprehensive energy system are guaranteed, and the gas consumption and the coal consumption of the system are reduced.
In recent years, advanced optimal scheduling techniques have improved the responsiveness and performance of the system, but these methods do not fully consider the uncertainty factors in the system, and cannot achieve optimal scheduling of the system. The day-ahead scheduling mode can make reasonable scheduling arrangement in advance, but cannot deal with real-time uncertainty of the system.
Disclosure of Invention
In order to solve the influence of uncertainty of wind power output on an electric heating comprehensive energy system, the invention provides an optimal scheduling method for day-ahead scheduling.
A day-ahead optimization scheduling method for an electric heating integrated energy system comprises the following steps:
step 1: performing characteristic analysis on historical wind power output data, and establishing a probability distribution model for describing the uncertainty of the wind power output;
step 2: predicting the load by utilizing a GRU network according to the environmental factors of the wind power output and historical data to obtain a load predicted value of each scheduling time interval in the day ahead;
and step 3: establishing a unit model of each component in the electric heating comprehensive energy system;
and 4, step 4: according to the wind power output uncertainty model in the step 1, describing uncertainty of wind power output in the system based on a scene tree structure; according to the load predicted value of each scheduling time interval in the day before obtained in the step 2 and the system model established in the step 3, solving by using a random model prediction control method to obtain the day-ahead scheduling parameter control quantity of the electric heating integrated energy system;
and 5: and outputting the control quantity of the day-ahead scheduling parameters to a power supply unit, a heat supply unit and an electric heating boiler, and regulating and controlling the power output of the unit at each scheduling time interval and the power input of the electric heating boiler to realize the optimal scheduling of the system.
The step 1) is specifically as follows:
the method comprises the steps of cleaning and analyzing characteristics of historical output data of wind power generation to obtain probability distribution characteristics of wind power output data, establishing a non-Gaussian distribution model of wind power output, and obtaining a probability density function of the wind power distribution model in an original non-Gaussian form.
The step 2) is specifically as follows:
and training and predicting the load by utilizing the established prediction network structure according to the environmental factors influencing the wind power output and the historical load data, and outputting the load demand predicted value at the day before.
The established network structure comprises a CNN layer, a GRU network and an Attention mechanism layer which are connected in sequence; the input of the network structure is wind speed, illumination, temperature and load at a sampling moment, and the output of the Attention layer is a load predicted value of each scheduling period in the day ahead: electric load for k-time system
Figure BDA0003139075240000031
Thermal load power
Figure BDA0003139075240000032
The electric heating comprehensive energy system comprises an electric power system and a thermodynamic system, the electric power system and the thermodynamic system are coupled through a thermoelectric unit, the thermodynamic system provides heat energy for a heat load through the thermoelectric unit and an electric heating boiler, the electric power system provides electric energy for the heat load through the thermoelectric unit, a conventional generator set and a fan, meanwhile, the system further comprises an energy storage system used for storing and releasing the electric energy of the electric heating comprehensive energy system, and unit models of all parts are respectively established.
The unit model of each part comprises:
(1) Thermoelectric unit operation model
Figure BDA0003139075240000033
Wherein the content of the first and second substances,
Figure BDA0003139075240000034
the generated power at the k moment of the ith set of CHP unit, eta GT For the comprehensive conversion efficiency of the gas turbine in the CHP unit, H ng Is a high heating value natural gas, and the natural gas has high heating value,
Figure BDA0003139075240000035
for the natural gas flow consumed by the ith set of CHP unit k at the moment, in the constant heat/power ratio mode, the heat power of the CHP unit is related to the output electric power of the unit, and can be expressed as follows:
Figure BDA0003139075240000036
wherein the content of the first and second substances,
Figure BDA0003139075240000037
the heat power of the CHP unit at the moment k; c. C CHP The heat-electricity ratio of the CHP unit;
(2) Thermal operation model of electric heating boiler
Figure BDA0003139075240000038
Wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003139075240000039
heating power of ith electric boiler at time k, eta EB In order to improve the electric heat conversion efficiency of the electric boiler,
Figure BDA0003139075240000041
the electricity power is used for the ith electric boiler at the moment k;
(3) Conventional generator set operation model
Figure BDA0003139075240000042
Wherein the content of the first and second substances,
Figure BDA0003139075240000043
representing the coal consumption of the ith conventional generating set at the moment k and representing the generated power of the conventional generating set, a i ，b i ，c i Representing coefficients related to the operation of the unit;
(4) Energy storage system model
In the dispatching of the electric heating comprehensive energy system, the energy storage system has the characteristic of high response speed, can compensate the power balance deviation of the system under the interference, and is an important component in the system; the energy storage system charging and discharging power model can be described as follows:
Figure BDA0003139075240000044
wherein the content of the first and second substances,
Figure BDA0003139075240000045
the capacity of the ith energy storage system at the moment k is shown, and delta T represents a unit scheduling time interval;
Figure BDA0003139075240000046
charging and discharging power for the energy storage system at k moment; eta c And η d Respectively charging and discharging efficiencies of the energy storage system;
Figure BDA0003139075240000047
the charge-discharge state of the energy storage system at k moment, when the energy storage system is charged
Figure BDA0003139075240000048
During discharge
Figure BDA0003139075240000049
(5) Plant operating constraints
The constraint of the upper and lower power limits of each unit device in the electric heating comprehensive energy system is required to be met at each operating moment, and the constraint is expressed as follows:
Figure BDA00031390752400000410
Figure BDA00031390752400000411
Figure BDA00031390752400000412
Figure BDA00031390752400000413
Figure BDA00031390752400000414
Figure BDA0003139075240000051
wherein the content of the first and second substances,
Figure BDA0003139075240000052
Figure BDA0003139075240000053
respectively representing the upper and lower limit values of the operating power of the ith conventional generator set, the CHP generator set, the wind turbine generator set, the energy storage equipment and the electric boiler;
Figure BDA0003139075240000054
and
Figure BDA0003139075240000055
respectively is the minimum value and the maximum value of the charge-discharge capacity of the energy storage system; n is a radical of G ,N CHP ,N WT ,N E ,N EB The number of the corresponding unit equipment in the system is set;
(6) Power change rate constraints
The control operation of each unit equipment in the system is required to be satisfied within a safety range, and is defined
Figure BDA0003139075240000056
Figure BDA0003139075240000057
The control power variation of the conventional generator set, the CHP generator set, the wind turbine generator set and the electric boiler in a unit scheduling period delta T is used for regulating and controlling the power output of the generator set in each scheduling period, and the following constraints are required to be met:
Figure BDA0003139075240000058
Figure BDA0003139075240000059
Figure BDA00031390752400000510
Figure BDA00031390752400000511
wherein the content of the first and second substances,
Figure BDA00031390752400000512
and
Figure BDA00031390752400000513
and
Figure BDA00031390752400000514
and
Figure BDA00031390752400000515
and
Figure BDA00031390752400000516
are respectively an ith conventional generatorThe upper and lower limit values of the power change of the group, the CHP unit, the wind turbine generator and the electric boiler in a unit scheduling period delta T;
(7) Electric and thermal power balance constraint
The electric heating comprehensive energy system needs to meet the electric load and heat load requirements of users by adjusting the parameter control quantity of each unit device in the operation process. The power balance constraint of the system is:
Figure BDA0003139075240000061
Figure BDA0003139075240000062
wherein the content of the first and second substances,
Figure BDA0003139075240000063
and
Figure BDA0003139075240000064
the power generation power of the conventional generator set and the power generation power of the wind generating set at the moment k are respectively;
Figure BDA0003139075240000065
the system power load is used at the moment k,
Figure BDA0003139075240000066
the thermal load power at time k of the system.
The wind power output uncertainty model according to the step 1) is used for describing uncertainty of wind power output in the system based on a scene tree structure; the method comprises the following steps:
establishing a scene tree structure-based wind power change explicit expression structure according to the non-Gaussian distribution model of the wind power output in the step 1) to obtain the probability omega of different wind power scenes s s 。
The method for solving and obtaining the parameter control quantity of the day-ahead scheduling of the electric heating integrated energy system by using a random model predictive control method according to the load predicted value of each scheduling time interval in the day-ahead obtained in the step 2) and the system model established in the step 3) comprises the following steps:
a. establishing an optimization problem model based on a model predictive control MPC method according to models of all units and equipment of the electric heating comprehensive energy system;
selecting the real-time power of each unit device in the k-time system and the capacity of the energy storage system as a system state variable x k Selecting the variable quantity of each equipment power in unit scheduling period delta T as the system optimization control parameter variable u k Respectively defined as:
Figure BDA0003139075240000067
Figure BDA0003139075240000068
Figure BDA0003139075240000071
in the operation process of the electric heating comprehensive energy scheduling system, under the condition of meeting the real-time electricity and heat load requirements of users and the self operation constraint conditions of each unit and equipment in the system, the aim of improving wind power consumption and reducing comprehensive operation cost is fulfilled, and the generation and heating power of each equipment participating in the adjustment of supply and demand balance is optimized;
gas consumption of power generation and heating equipment of electric heating comprehensive energy scheduling system
Figure BDA0003139075240000072
And the electricity purchase cost of the electric equipment
Figure BDA0003139075240000073
Respectively expressed as:
Figure BDA0003139075240000074
Figure BDA0003139075240000075
where k denotes the current time, R ng The gas purchase price of the system is represented,
Figure BDA0003139075240000078
purchasing electricity price for the system; running performance index l of system at moment k k Can be expressed as:
Figure BDA0003139075240000076
the objective function is as follows:
Figure BDA0003139075240000077
wherein J represents an objective function, N p The time domain is predicted for the system. Wherein:
Q＝diag{c 1 ,c 2 ,...,c NG ,0,...,0}
Figure BDA0003139075240000081
Figure BDA0003139075240000082
the four constraints are as follows:
s.t.x k+j+1|k ＝A·x k+j|k +B·u k+j|k
Figure BDA0003139075240000083
p x ·x k+j|k ≤q x
p u ·u k+j|k ≤q u
wherein the first constraint condition is a state space equation of the electric heating comprehensive energy system, wherein
Figure BDA0003139075240000084
N A ＝N G +N CHP +N EB +N E
N B ＝N G +N CHP +N EB
Figure BDA0003139075240000085
Wherein the content of the first and second substances,
Figure BDA0003139075240000086
charging or discharging efficiency of the ith set of energy storage system at the moment k;
the second constraint is the adaptation of the electrical-thermal equilibrium constraint equations in the system model, where:
C＝[d P ,d Q ] T ,D＝[1,0] T
Figure BDA0003139075240000087
Figure BDA0003139075240000088
Figure BDA0003139075240000091
Figure BDA0003139075240000092
wherein the content of the first and second substances,
Figure BDA0003139075240000093
is a predicted value of the electric heating load at each moment in the day, w k+j|k Predicting the wind power generation power value in the time domain;
the third constraint condition and the fourth constraint condition describe the constraint of the upper and lower power limits of the unit equipment and the constraint of the variable quantity of the control power in the scheduling period delta T, wherein:
Figure BDA0003139075240000094
Figure BDA0003139075240000095
Figure BDA0003139075240000096
b. establishing a random model predictive control model (SMPC) based on the scene tree model, and solving to obtain a day-ahead scheduling scheme of the electric heating comprehensive energy system; the method comprises the following steps:
at the moment k, the control variable of the stochastic economic model predictive control on the prediction time domain can be expressed as:
Figure BDA0003139075240000097
wherein N is s The total number of the branches of the scene tree, namely the total number of the optimized paths of the system; s represents an s-th scene tree structure branch; j denotes the time j-th in the prediction time domain. The corresponding predicted state variables are:
Figure BDA0003139075240000101
further improving the objective function J based on the scene tree to obtain N s Probability omega combining different wind power scenarios s And calculating the weighted sum of the system operation performance indexes under all probability situations
Figure BDA0003139075240000102
Figure BDA0003139075240000103
Wherein the content of the first and second substances,
Figure BDA0003139075240000104
weighted sum, omega, of day-ahead scheduling operation indices for an electric heating integrated energy system s Representing the probability of occurrence under different wind power scenarios s obtained by the scenario tree,
Figure BDA0003139075240000105
indicating the performance index of the correction after introducing the relaxation variable,
Figure BDA0003139075240000106
represents an introduced relaxation variable;
the corresponding constraint conditions are:
Figure BDA0003139075240000107
Figure BDA0003139075240000108
Figure BDA0003139075240000109
Figure BDA00031390752400001010
Figure BDA0003139075240000111
Figure BDA0003139075240000112
Figure BDA0003139075240000113
solving the random model predictive control problem to obtain the unit parameter control quantity of the electric heating comprehensive energy system at the moment k under the day-ahead optimized scheduling strategy of the system operation: the control power variation of the conventional generator set, the CHP set and the electric boiler in a unit scheduling period delta T is as follows:
Figure BDA0003139075240000114
and the charging and discharging power of the energy storage system
Figure BDA0003139075240000115
A day-ahead optimized dispatching system of an electric heating comprehensive energy system comprises an electric power system, a thermodynamic system, an energy storage system and an upper computer controller; the thermodynamic system comprises a thermoelectric unit and an electric heating boiler and is used for generating electric heating energy under the instruction of the controller; the power system comprises a thermoelectric generating set, a conventional generating set and a fan, and the energy storage system is used for storing and releasing electric energy of the power system; the upper computer controller comprises a processor and a memory, wherein the memory stores programs, the processor loads the programs to execute the steps of the method, and the control power variation of the conventional generator set, the CHP set and the electric boiler in a unit scheduling period delta T is obtained:
Figure BDA0003139075240000116
and charging and discharging power of energy storage system
Figure BDA0003139075240000117
And outputting the data to each unit on site to realize optimized scheduling.
The invention has the following effects and advantages:
1. the invention provides the optimized scheduling of the electric heating comprehensive energy system from a plurality of time scales, and the air abandoning rate of the system and the gas consumption and coal consumption of the system are reduced.
2. The established random model prediction control method utilizes the scene tree structure to express the uncertainty of the wind power output of the system, fully considers the random variation condition of the wind power, reduces the interference of the wind power randomness in the system optimization scheduling, and improves the system robustness.
3. The model and the method established by the invention can be used for scheduling optimization of the electric heating comprehensive energy system, and take the error of the prediction of the wind power output at the present into consideration, thereby laying a foundation for wind power consumption and reliable operation of the system.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a block diagram of a GRU predictive network of the method of the present invention;
FIG. 3 is an electric heat integrated energy system configuration of the method of the present invention;
FIG. 4 is a scene tree structure of the method of the present invention;
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below to enable one skilled in the art to better understand the present invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein, but rather should be construed as modified in the spirit and scope of the present invention as set forth in the appended claims.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
As shown in fig. 1, the method for optimizing and scheduling the electric heating integrated energy system in the day ahead comprises the following steps:
1) Carrying out characteristic analysis on the wind power output data to obtain the distribution characteristics of the data, and establishing a probability distribution model of the uncertainty of the wind power output;
2) According to the environmental factors of the wind power output and the historical load data, predicting the load by using a GRU network to obtain a load predicted value of each scheduling time period in the day before the wind power output;
3) Establishing a unit model of each component in the electric heating comprehensive energy system;
4) According to the wind power output uncertainty model in the step 1), describing uncertainty of wind power output in the system based on a scene tree structure, obtaining a day-ahead scheduling scheme of the electric heating comprehensive energy system by using a random model prediction control method according to the load prediction value of each scheduling time period in the day-ahead obtained in the step 2) and the system unit model established in the step 3);
5) And 5: and outputting the control quantity of the day-ahead scheduling parameters to a power supply unit, a heat supply unit and an electric heating boiler, and regulating and controlling the power output of the unit in each scheduling period to realize optimized scheduling. The unit parameter control quantity under the day-ahead optimized scheduling strategy is as follows: the control power variation of the conventional generator set, the CHP set and the electric boiler in a unit scheduling period delta T is as follows:
Figure BDA0003139075240000131
Figure BDA0003139075240000132
and charging and discharging power of energy storage system
Figure BDA0003139075240000133
In the step 1), the data set adopted by the invention is derived from CAISO, the data set comprises 2019 wind power output data, load, real-time electricity price data and the like, and the sampling time interval is 1 hour. The method comprises the steps of firstly cleaning and preprocessing historical wind power output data, completing missing data in an original data file by adopting an interpolation method, and sequencing the data according to sampling time. And then, carrying out feature analysis on the wind power output data by using a scimit-spare toolkit based on a python programming language.
As can be seen from a scatter diagram of the wind power output data, the wind power output data has extremely strong uncertainty, and the distribution of the wind power output data does not accord with standard Gaussian distribution, so an improved wind power non-Gaussian probability transformation model is provided, wind power sequences which do not meet the standard Gaussian distribution are transformed into new sequences which meet the standard Gaussian distribution, and the transformation model is expressed as follows:
Figure BDA0003139075240000141
wherein the non-linear transformation function f i (. The ith wind generating set history non-Gaussian distribution sequence P WT,i Transforming into a sequence of Gaussian distributions
Figure BDA0003139075240000142
ψ 1 ,ψ 2 ,ψ 3 For transforming the parameters, and psi 1 ,ψ 2 Is more than or equal to 0. Thus, the post-sequence is transformed
Figure BDA0003139075240000143
Can be expressed as the following gaussian distribution model:
Figure BDA0003139075240000144
the probability density function of the wind power distribution model in the original non-gaussian form calculated according to the formula above can be expressed as:
Figure BDA0003139075240000145
because the actual wind power values are all larger than or equal to zero, the nonlinear transformation function f i And monotonous on a definition domain, namely, the inverse function has a numerical solution in a feasible domain, and for each Gaussian distribution random value which accords with the formula, a value which accords with the original distribution of the wind power sequence can be found to correspond to the value.
In the step 2), the GRU is a special RNN, which has a structure similar to the LSTM, and referring to fig. 2, the GRU replaces an input gate and a forgetting gate of the LSTM with an update gate, and accesses a reset gate to determine how much the neuron hidden layer output at the previous time is reset. The reduction of gate units allows GRUs to have fewer training sessions, increasing the speed of training, and also allowing them to be more predictive than LSTM.
In order to improve the prediction accuracy, a CNN convolutional network is added before GRU for processing historical data, and an Attention mechanism is added after GRU, so that the learning capability of the GRU network on information correlation is further improved, and the prediction accuracy is improved.
The data set for GRU training is 8760 lines of environmental factors and load data related to wind power output in 2019 all the year around, and each line represents a wind speed, illumination, temperature and load demand sample at one sampling moment. The data from 20 days before each month were selected as the training set for 5760 lines and the data from the remaining days as the test set for 3000 lines. Each sample contained historical wind speed, light and temperature data for the past 24 hours, for 75 columns.
Firstly, an input layer is connected to a CNN layer, and data characteristics are calculated:
y k ＝f(x k-1 u k-1 +x k u k +x k+1 u k+1 +b k )
in the formula y k Output of a characteristic value representing time k, x k The data sample value representing the k moment is a four-dimensional vector which represents the wind speed, the illumination, the temperature and the load at a sampling moment, u k Representing the convolution kernel at time k, b k Is a constant.
Then, the output of the CNN layer is used as the input of the GRU network, and the operation process of each hidden layer in the network is as follows:
r k ＝σ(W r *h k-1 +U r *y k +b r )
z k ＝σ(W z *h k-1 +U z *y k +b z )
where σ denotes the activation function, b r And b z Respectively represent reset gates r k And an update gate z k Bias constant of (W) r ，U r ，W z ，U z Respectively representing parameters of the GRU network to be trained. Inputting the characteristic data y of the current moment k By resetting the gate r k And an update gate z k Respectively determining the hidden layer state h of the previous time k-1 How much information is forgotten and how much information is transferred to the hidden layer h at the current moment k In (1).
After obtaining the gating information, the gating information r is reset k And the previous time hidden layer state h k-1 Hadamard product is performed to determine the hidden layer h at the previous time k-1 How much information will be forgotten, the obtained reset information and the input y at the current moment k The binding is put into the activation function tanh.
h′ k ＝tanh(W c *(r k *h k-1 )+U c *y k +b c )
Wherein tan h is the activation function, b c Representing the bias constant of the forgetting gate. W c ，U c A parameter matrix trained for the network.
Finally, the door z is updated according to the previous obtained t Determining the hidden layer h at the current time t 。
h k ＝(1-z t )*h k-1 +z t *h′ k
An Attention mechanism is accessed behind a GRU network layer and is used for learning output data of the GRU network. And the output of the Attention layer is the final load predicted value of each scheduling period in the day ahead.
Compared with the prediction results of the GRU network and the LSTM network, the prediction precision is improved by 14.25%.
In the step 3), the electric heating comprehensive energy system comprises a thermal power system and an electric power system, the thermal power system provides heat energy for a thermal load through a thermoelectric unit and an electric heating boiler, the electric power system provides electric energy for the electric load and the electric heating boiler through the thermoelectric unit, the thermal power system is coupled with the electric power system through the thermoelectric unit, meanwhile, the system also comprises an energy storage system used for guaranteeing the supply and demand balance of the electric power system, and unit models of all parts are as follows. The structure of the electric heating comprehensive energy system is shown in the attached figure 3.
(1) Thermoelectric unit operation model
Figure BDA0003139075240000161
Wherein the content of the first and second substances,
Figure BDA0003139075240000162
the generated power at k moment of the ith set of CHP unit is eta GT For the comprehensive conversion efficiency of the gas turbine in the CHP unit, H ng Is a high heating value natural gas, and has high heat value,
Figure BDA0003139075240000163
for the natural gas flow consumed by the ith set of CHP unit k at the moment, in the constant heat/power ratio mode, the heat power of the CHP unit is related to the output electric power of the unit, and can be expressed as follows:
Figure BDA0003139075240000171
wherein the content of the first and second substances,
Figure BDA0003139075240000172
the heat power of the CHP unit at the moment k; c. C CHP The heat-electricity ratio of the CHP unit.
(2) Thermal power output of electric heating boiler
Figure BDA0003139075240000173
Figure BDA0003139075240000174
Wherein the content of the first and second substances,
Figure BDA0003139075240000175
for electric boilers at kThermal power of etching, eta EB In order to improve the electric heat conversion efficiency of the electric boiler,
Figure BDA0003139075240000176
and the electric power is used for the ith electric boiler at the moment k.
(3) Conventional generator set operation model
Figure BDA0003139075240000177
Figure BDA0003139075240000178
Wherein the content of the first and second substances,
Figure BDA0003139075240000179
representing the coal consumption of the ith conventional generator set at the moment k, and the coefficient a i ,b i ,c i Representing parameters related to the operation of the unit.
(4) Energy storage system model
In the dispatching of the electric heating comprehensive energy system, the power balance deviation of the system under the interference can be compensated by utilizing the characteristic of high response speed of the energy storage system, and the electric heating comprehensive energy system is an important component in the system. The energy storage system charge-discharge power model can be described as:
Figure BDA00031390752400001710
wherein the content of the first and second substances,
Figure BDA00031390752400001711
the capacity of the ith energy storage system at the moment k;
Figure BDA00031390752400001712
charging and discharging power for the energy storage system at k moment; eta c And η d Respectively charging and discharging efficiencies of the energy storage system;
Figure BDA00031390752400001713
for storing energyThe charge-discharge state of the system at the k moment and the charge of the energy storage system
Figure BDA00031390752400001714
During discharge
Figure BDA00031390752400001715
(5) Plant operating constraints
The constraint of the upper and lower power limits of each unit device in the electric heating comprehensive energy system is required to be met at each operating moment, and the constraint is expressed as follows:
Figure BDA0003139075240000181
Figure BDA0003139075240000182
Figure BDA0003139075240000183
Figure BDA0003139075240000184
Figure BDA0003139075240000185
Figure BDA0003139075240000186
wherein the content of the first and second substances,
Figure BDA0003139075240000187
Figure BDA0003139075240000188
respectively represents the ith conventional generator set, the CHP generator set, the wind turbine generator set,The upper and lower limit values of the operating power of the energy storage equipment and the electric boiler;
Figure BDA0003139075240000189
and
Figure BDA00031390752400001810
the minimum value and the maximum value of the charge-discharge capacity of the energy storage system are respectively.
(6) Power rate of change constraint
The control operation of each unit equipment in the system is required to be satisfied within a safety range, and is defined
Figure BDA00031390752400001811
Figure BDA00031390752400001812
The power changes of a conventional generator set, a CHP generator set, a wind turbine generator set and an electric boiler in a unit scheduling period T respectively need to satisfy the following constraints:
Figure BDA00031390752400001813
Figure BDA00031390752400001814
Figure BDA00031390752400001815
Figure BDA00031390752400001816
wherein the content of the first and second substances,
Figure BDA00031390752400001817
and
Figure BDA00031390752400001818
and
Figure BDA00031390752400001819
and
Figure BDA00031390752400001820
and
Figure BDA00031390752400001821
the power change upper and lower limit values of the ith conventional generator set, the CHP generator set, the wind turbine generator set and the electric boiler in the unit scheduling period delta T are respectively.
(7) Constraint of balance of electric and thermal power
The electric heating comprehensive energy system needs to meet the electric load and heat load requirements of users by adjusting the parameter control quantity of each unit device in the operation process. The power balance constraint of the system is:
Figure BDA0003139075240000191
Figure BDA0003139075240000192
wherein the content of the first and second substances,
Figure BDA0003139075240000193
and
Figure BDA0003139075240000194
the power generation power of the conventional generator set and the power generation power of the wind generating set at the moment k are respectively;
Figure BDA0003139075240000195
the system power load is used at the moment k; n is a radical of G ,N CHP ,N WT ,N E ,N EB The number of the corresponding unit equipment in the system.
Figure BDA0003139075240000196
The thermal load power at time k of the system.
In the step 4), the scene tree structure is used for describing the random variation condition of the wind power, the scene tree branch is selected based on the wind power output transformation model established in the step 1) to more accurately describe the wind power variation characteristic, and then the model prediction control method is used for solving, so that the uncertainty degree that the control quantity output of the system at the future moment depends on the wind power output at the previous moment is represented, and the method is a closed-loop robust control strategy.
Referring to FIG. 3, state x at time k of the system k Random quantity of different wind power
Figure BDA00031390752400001912
Under the action of (1), n scene tree branches are generated, corresponding to n different operation states of the system at the moment of (k + 1)
Figure BDA0003139075240000197
Similarly, the state of the scene tree at time k +1
Figure BDA0003139075240000198
Figure BDA0003139075240000199
There are also n branches corresponding to the state of the system at time k +1
Figure BDA00031390752400001910
Different operation states of
Figure BDA00031390752400001911
The system has n from k to k +2 2 Branches, and so on. In order to prevent the calculation amount from exponentially increasing due to excessive situations, a suitable robust time domain N is selected r ≤N P And the branch number n of a single node of the scene tree, and the uncertainty variable is kept unchanged after the robust time domain, so that the closed-loop performance is ensured, and meanwhile, the calculation load is reduced.
Although the scenario tree structure in fig. 3 cannot represent all possibilities of wind power change, a method for effectively selecting branch scenarios of each node is designed according to historical wind power information, and uncertain dynamic change conditions of a system at future time can be represented better. Because the actual wind power output sequence does not meet the standard Gaussian distribution, and the modeling can not be carried out by using a known distribution. Therefore, an improved wind power non-gaussian probability transformation model is proposed:
Figure BDA0003139075240000201
selecting the real-time power of each unit device in the k-time system and the capacity of the energy storage system as a system state variable x k Selecting the variable quantity of each device power in unit scheduling period delta T as the optimized control parameter variable u of the system k Respectively defined as:
Figure BDA0003139075240000202
Figure BDA0003139075240000203
in the operation process of the electric heating comprehensive energy scheduling system, under the condition of meeting the real-time electricity and heat load requirements of users and the self operation constraint conditions of each unit and equipment in the system, the electric heating comprehensive energy scheduling system aims to improve wind power consumption and reduce comprehensive operation cost, and optimizes the participation of each equipment in the adjustment of power generation and heating power of supply and demand balance.
Gas consumption of power generation and heating equipment of electric heating comprehensive energy scheduling system
Figure BDA0003139075240000204
And the electricity purchase cost of the electric equipment
Figure BDA0003139075240000211
Respectively expressed as:
Figure BDA0003139075240000212
Figure BDA0003139075240000213
where k denotes the current time, R ng The gas purchase price of the system is represented,
Figure BDA0003139075240000219
purchasing electricity price for the system; running performance index l of system at moment k k Can be expressed as:
Figure BDA0003139075240000214
the objective function is as follows:
Figure BDA0003139075240000215
wherein J represents an objective function, N p The time domain is predicted for the system. Wherein:
Figure BDA00031390752400002110
Figure BDA0003139075240000216
Figure BDA0003139075240000217
the four constraints are as follows:
s.t.x k+j+1|k ＝A·x k+j|k +B·u k+j|k
Figure BDA0003139075240000218
p x ·x k+j|k ≤q x
p u ·u k+j|k ≤q u
wherein, the first constraint condition is a state space equation of the electric heating comprehensive energy system, wherein:
Figure BDA0003139075240000221
N A ＝N G +N CHP +N EB +N E
N B ＝N G +N CHP +N EB
Figure BDA0003139075240000222
wherein the content of the first and second substances,
Figure BDA0003139075240000223
and charging or discharging efficiency of the ith set of energy storage system k at the moment.
The second constraint is the system's electrical-thermal equilibrium equation, where:
C＝[d P ,d Q ] T ,D＝[1,0] T
Figure BDA0003139075240000224
Figure BDA0003139075240000225
Figure BDA0003139075240000226
Figure BDA0003139075240000227
wherein the content of the first and second substances,
Figure BDA0003139075240000228
is a predicted value of the electric heating load at each moment in the day k+j|k To predict the wind power generation power value in the time domain.
The third and fourth constraint conditions are the rewriting of the power upper and lower limit constraint of the unit equipment and the power variation constraint in the scheduling period delta T, wherein:
Figure BDA0003139075240000229
Figure BDA0003139075240000231
Figure BDA0003139075240000232
b. obtaining scenes under different wind power groups based on a scene tree model, establishing a random model predictive control model (SMPC), and solving to obtain a day-ahead scheduling scheme of the electric heating integrated energy system; the method comprises the following steps:
at the moment k, the control variable of the stochastic economic model predictive control on the prediction time domain can be expressed as:
Figure BDA0003139075240000233
wherein N is s The total number of the branches of the scene tree, namely the total number of the optimized paths of the system; s represents an s-th scene tree structure branch; j denotes the time j-th in the prediction time domain. The corresponding predicted state variables are:
Figure BDA0003139075240000234
target function J based on scene treeFurther improvement, to obtain N s Probability omega combining different wind power scenarios s And calculating the weighted sum of the system operation performance indexes under all probability situations
Figure BDA0003139075240000241
Figure BDA0003139075240000242
Wherein the content of the first and second substances,
Figure BDA0003139075240000243
the final output electric heating comprehensive energy system is in N s Weighted summation of performance indicators, omega, under different wind power scenarios s Representing the probability of occurrence under different wind power scenarios s obtained by the scenario tree,
Figure BDA0003139075240000244
the performance index of the correction after the relaxation variable is introduced is shown,
Figure BDA0003139075240000245
represents the introduced relaxation variable, and λ is a penalty factor.
The corresponding constraint conditions are:
Figure BDA0003139075240000246
Figure BDA0003139075240000247
the system needs to satisfy constraints under any uncertain situation:
Figure BDA0003139075240000248
Figure BDA0003139075240000249
because the system can not accurately predict the size of uncertain variables at future time, corresponding control quantity can not be selected according to the size, and the same control variable u is adopted at each scene tree node k To satisfy the "future invisibility" of the wind speed, i.e. the "immeasurable constraints" of the system:
Figure BDA00031390752400002410
Figure BDA00031390752400002411
Figure BDA0003139075240000251
uncertain variables in FIG. 4
Figure BDA0003139075240000252
For the system in state x k And (4) possible values of n wind power uncertainty variables. The invention sets the same scene branch number at each node in the robust time domain. Selecting a prediction time domain N from 96 sampling points every day P At 96, i.e. scheduled 24 hours (days ago) in advance, each time interval Δ T is 15min. And solving the random model predictive control problem by adopting IPOPT based on CasADi technology to obtain the unit parameter control quantity under the day-ahead optimized scheduling strategy of the electric heating comprehensive energy system when the system operates at the moment k. The hardware environment is Intel Core i5-7400,6G RAM and Matlab2018a.
In conclusion, the invention provides a day-ahead optimization scheduling method for an electric heating integrated energy system, which considers the uncertainty of wind power output and analyzes the historical data of the wind power output to obtain a probability distribution model of the uncertainty of the wind power output; then predicting the load by utilizing the GRU; and then according to the predicted day-ahead load data, a day-ahead scheduling scheme is obtained by using a random prediction control method, and the day-ahead scheduling method can improve the consumption of the wind power output of the renewable energy source and reduce the coal consumption of the system on the premise of ensuring the reliability and the safety of the system.
Although the present invention has been described in detail in the above examples, the scope of application of the present invention is not limited thereto. It will be apparent to those skilled in the art that the system may be scaled and complicated without departing from the scope of the claims.

Claims (9)

1. A day-ahead optimization scheduling method for an electric heating comprehensive energy system is characterized by comprising the following steps:
step 1: performing characteristic analysis on historical wind power output data, and establishing a probability distribution model for describing the uncertainty of the wind power output;
step 2: predicting the load by utilizing a GRU network according to the environmental factors of the wind power output and the historical data to obtain the load predicted value of each scheduling time period in the day ahead;
and step 3: establishing a unit model of each component in the electric heating comprehensive energy system;
and 4, step 4: according to the wind power output uncertainty model in the step 1, describing uncertainty of wind power output in the system based on a scene tree structure; according to the load predicted value of each scheduling time interval in the day before obtained in the step 2 and the system model established in the step 3, solving by using a random model prediction control method to obtain the day-ahead scheduling parameter control quantity of the electric heating integrated energy system;
and 5: and outputting the control quantity of the day-ahead scheduling parameters to a power supply unit, a heat supply unit and an electric heating boiler, and regulating and controlling the power output of the unit at each scheduling time interval and the power input of the electric heating boiler to realize the optimal scheduling of the system.
2. The day-ahead optimization scheduling method for the electric-thermal integrated energy system according to claim 1, wherein the step 1) is specifically as follows:
the method comprises the steps of cleaning and analyzing characteristics of historical output data of wind power generation to obtain probability distribution characteristics of wind power output data, establishing a non-Gaussian distribution model of wind power output, and obtaining a probability density function of the wind power distribution model in an original non-Gaussian form.
3. The day-ahead optimization scheduling method of the electric heating integrated energy system according to claim 1, wherein the step 2) is specifically as follows:
and training and predicting the load by utilizing the established prediction network structure according to the environmental factors influencing the wind power output and the historical load data, and outputting the load demand predicted value at the day before.
4. The electric heating comprehensive energy system day-ahead optimization scheduling method of claim 3, wherein the established network structure comprises a CNN layer, a GRU network and an Attention mechanism layer which are connected in sequence; the input of the network structure is wind speed, illumination, temperature and load at a sampling moment, and the output of the Attention layer is a load predicted value of each scheduling time period in the day ahead: electric load for k-time system
Figure FDA0003139075230000021
Power of thermal load
Figure FDA0003139075230000022
5. The day-ahead optimized scheduling method of an electrothermal integrated energy system according to claim 1, wherein the electrothermal integrated energy system comprises an electric power system and a thermal system, the electric power system and the thermal system are coupled through a thermoelectric unit, the thermal system provides heat energy for a thermal load through the thermoelectric unit and an electric heating boiler, the electric power system provides electric energy for the electrical load through the thermoelectric unit, a conventional generator set and a fan, and the system further comprises an energy storage system for storing and releasing the electric energy of the electrothermal integrated energy system to respectively establish unit models of each part.
6. The electric heating integrated energy system day-ahead optimization scheduling method of claim 1, wherein the unit model of each part comprises:
(1) Thermoelectric unit operation model
Figure FDA0003139075230000023
Wherein the content of the first and second substances,
Figure FDA0003139075230000024
the generated power at the k moment of the ith set of CHP unit, eta GT For the comprehensive conversion efficiency of the gas turbine in the CHP unit, H ng Is a high heating value natural gas, and the natural gas has high heating value,
Figure FDA0003139075230000025
for the natural gas flow consumed by the ith set of CHP unit k at the moment, in the constant heat/power ratio mode, the heat power of the CHP unit is related to the output electric power of the unit, and can be expressed as follows:
Figure FDA0003139075230000026
wherein, the first and the second end of the pipe are connected with each other,
Figure FDA0003139075230000027
the heat power of the CHP unit at the moment k; c. C CHP The heat-electricity ratio of the CHP unit;
(2) Thermal operation model of electric heating boiler
Figure FDA0003139075230000031
Wherein the content of the first and second substances,
Figure FDA0003139075230000032
heating power of ith electric boiler at time k, eta EB In order to improve the electric heat conversion efficiency of the electric boiler,
Figure FDA0003139075230000033
the electricity power is used for the ith electric boiler at the moment k;
(3) Conventional generator set operation model
Figure FDA0003139075230000034
Wherein the content of the first and second substances,
Figure FDA0003139075230000035
represents the coal consumption of the ith conventional generating set at the moment k, represents the generating power of the conventional generating set, and a i ，b i ，c i Representing coefficients related to the operation of the unit;
(4) Energy storage system model
In the dispatching of the electric heating comprehensive energy system, the energy storage system has the characteristic of high response speed, can compensate the power balance deviation of the system under the interference, and is an important component in the system; the energy storage system charging and discharging power model can be described as follows:
Figure FDA0003139075230000036
wherein the content of the first and second substances,
Figure FDA0003139075230000037
the capacity of the ith energy storage system at the moment k is shown, and delta T represents a unit scheduling time interval;
Figure FDA0003139075230000038
charging and discharging power for the energy storage system at k moment; eta c And η d Respectively charging and discharging efficiencies of the energy storage system;
Figure FDA0003139075230000039
the energy storage system is in a k-time charging and discharging state when being charged
Figure FDA00031390752300000310
During discharge
Figure FDA00031390752300000311
(5) Plant operating constraints
The constraint of the upper and lower power limits of each unit device in the electric heating comprehensive energy system is required to be met at each operating moment, and the constraint is expressed as follows:
Figure FDA00031390752300000312
Figure FDA0003139075230000041
Figure FDA0003139075230000042
Figure FDA0003139075230000043
Figure FDA0003139075230000044
Figure FDA0003139075230000045
wherein the content of the first and second substances,
Figure FDA0003139075230000046
Figure FDA0003139075230000047
respectively representing the upper and lower limit values of the operating power of the ith conventional generator set, the CHP generator set, the wind turbine generator set, the energy storage equipment and the electric boiler;
Figure FDA0003139075230000048
and
Figure FDA0003139075230000049
respectively is the minimum value and the maximum value of the charge-discharge capacity of the energy storage system; n is a radical of G ，N CHP ，N WT ，N E ，N EB The number of the corresponding unit equipment in the system is set;
(6) Power change rate constraints
The control operation of each unit equipment in the system needs to be satisfied within a safety range, and the information is determined
Figure FDA00031390752300000410
Figure FDA00031390752300000411
The control power variation of the conventional generator set, the CHP generator set, the wind turbine generator set and the electric boiler in a unit scheduling period delta T is used for regulating and controlling the power output of the generator set in each scheduling period, and the following constraints are required to be met:
Figure FDA00031390752300000412
Figure FDA00031390752300000413
Figure FDA00031390752300000414
Figure FDA00031390752300000415
wherein the content of the first and second substances,
Figure FDA00031390752300000416
and
Figure FDA00031390752300000417
and
Figure FDA00031390752300000418
and
Figure FDA00031390752300000419
and
Figure FDA00031390752300000420
the power change upper and lower limit values of the ith conventional generator set, the CHP generator set, the wind turbine generator set and the electric boiler in a unit scheduling period delta T are respectively set;
(7) Electric and thermal power balance constraint
The electric heating comprehensive energy system needs to meet the electric load and heat load requirements of users by adjusting the parameter control quantity of each unit device in the operation process. The power balance constraint of the system is:
Figure FDA0003139075230000051
Figure FDA0003139075230000052
wherein the content of the first and second substances,
Figure FDA0003139075230000053
and
Figure FDA0003139075230000054
respectively a conventional generator set and a wind generating set at kThe generated power at that moment;
Figure FDA0003139075230000055
the system power load is used at the moment k,
Figure FDA0003139075230000056
the thermal load power at time k of the system.
7. The day-ahead optimized dispatching method for the electric-thermal integrated energy system according to claim 1, characterized in that according to the wind-electricity output uncertainty model of step 1), the uncertainty of wind-electricity output in the system is described based on a scenario tree structure; the method comprises the following steps:
establishing a scene tree structure-based wind power change explicit expression structure according to the non-Gaussian distribution model of the wind power output in the step 1) to obtain the probability omega of different wind power scenes s s 。
8. The method for optimizing and scheduling an electric heating integrated energy system in the day-ahead according to claim 1, wherein the method for solving and obtaining the parameter control quantity of the electric heating integrated energy system in the day-ahead scheduling by using a random model prediction control method according to the load prediction value of each scheduling time interval in the day-ahead obtained in the step 2) and the system model established in the step 3) comprises the following steps:
a. establishing an optimization problem model based on a model predictive control MPC method according to models of all units and equipment of the electric heating comprehensive energy system;
selecting the real-time power of each unit device in the k-time system and the capacity of the energy storage system as a system state variable x k Selecting the variable quantity of each device power in unit scheduling period delta T as the optimized control parameter variable u of the system k Respectively defined as:
Figure FDA0003139075230000061
Figure FDA0003139075230000062
in the operation process of the electric heating comprehensive energy scheduling system, under the condition of meeting the real-time electricity and heat load requirements of users and self operation constraint conditions of units and equipment in the system, the electric heating comprehensive energy scheduling system aims at improving wind power consumption and reducing comprehensive operation cost, and optimizes the participation of the equipment in the adjustment of power generation and heating power of supply and demand balance;
gas consumption of power generation and heating equipment of electric heating comprehensive energy scheduling system
Figure FDA0003139075230000063
And the electricity purchase cost of the electric equipment
Figure FDA0003139075230000064
Respectively expressed as:
Figure FDA0003139075230000065
Figure FDA0003139075230000066
wherein k represents the current time, R ng The gas purchase price of the system is represented,
Figure FDA0003139075230000067
purchasing electricity price for the system; running performance index l of system at moment k k Can be expressed as:
Figure FDA0003139075230000068
the objective function is as follows:
Figure FDA0003139075230000071
wherein J represents an objective function, N p The time domain is predicted for the system. Wherein:
Figure FDA0003139075230000072
Figure FDA0003139075230000073
Figure FDA0003139075230000074
the four constraints are as follows:
s.t.x k+j+1|k ＝A·x k+j|k +B·u k+j|k
Figure FDA0003139075230000075
p x ·x k+j|k ≤q x
p u ·u k+j|k ≤q u
wherein, the first constraint condition is the state space equation of the electric heating comprehensive energy system, wherein:
Figure FDA0003139075230000076
N A ＝N G +N CHP +N EB +N E
N B ＝N G +N CHP +N EB
Figure FDA0003139075230000077
wherein the content of the first and second substances,
Figure FDA0003139075230000078
charging or discharging efficiency of the ith set of energy storage system at the moment k;
the second constraint is the adaptation of the electrical-thermal equilibrium constraint equations in the system model, where:
C＝[d P ，d Q ] T ，D＝[1，0] T
Figure FDA0003139075230000081
Figure FDA0003139075230000082
Figure FDA0003139075230000083
Figure FDA0003139075230000084
wherein the content of the first and second substances,
Figure FDA0003139075230000085
is a predicted value of the electric heating load at each moment in the day, w k+j|k Predicting the wind power generation power value in the time domain;
the third constraint condition and the fourth constraint condition describe the constraint of the upper and lower power limits of the unit equipment and the constraint of the variable quantity of the control power in the scheduling period delta T, wherein:
Figure FDA0003139075230000086
Figure FDA0003139075230000087
Figure FDA0003139075230000088
b. establishing a random model predictive control model (SMPC) based on the scene tree model, and solving to obtain a day-ahead scheduling scheme of the electric heating comprehensive energy system; the method comprises the following steps:
at the moment k, the control variable of the stochastic economic model predictive control on the prediction time domain can be expressed as:
Figure FDA0003139075230000091
wherein, N s The total number of the branches of the scene tree, namely the total number of the optimized paths of the system; s represents an s-th scene tree structure branch; j denotes the time j-th in the prediction time domain. The corresponding predicted state variables are:
Figure FDA0003139075230000092
further improving the objective function J based on the scene tree to obtain N s Probability omega combining different wind power scenarios s And calculating the weighted sum of the system operation performance indexes under all probability situations
Figure FDA0003139075230000093
Figure FDA0003139075230000094
Wherein the content of the first and second substances,
Figure FDA0003139075230000095
weighted sum, omega, of day-ahead scheduling operation indexes for an electric heating integrated energy system s Representing the probability of occurrence under different wind power scenarios s obtained by the scenario tree,
Figure FDA0003139075230000096
the performance index of the correction after the relaxation variable is introduced is shown,
Figure FDA0003139075230000101
represents the introduced relaxation variables;
the corresponding constraint conditions are:
Figure FDA0003139075230000102
Figure FDA0003139075230000103
Figure FDA0003139075230000104
Figure FDA0003139075230000105
Figure FDA0003139075230000106
Figure FDA0003139075230000107
Figure FDA0003139075230000108
solving the random model predictive control problem to obtain the unit parameter control quantity under the day-ahead optimized scheduling strategy of the electric heating integrated energy system at the moment k by the system operation: the control power variation of the conventional generator set, the CHP set and the electric boiler in a unit scheduling period delta T is as follows:
Figure FDA0003139075230000109
and the charging and discharging power of the energy storage system
Figure FDA00031390752300001010
9. A day-ahead optimized dispatching system of an electric heating comprehensive energy system is characterized by comprising an electric power system, a thermodynamic system, an energy storage system and an upper computer controller; the thermodynamic system comprises a thermoelectric unit and an electric heating boiler and is used for generating electric heating energy under the instruction of the controller; the power system comprises a thermoelectric generating set, a conventional generating set and a fan, and the energy storage system is used for storing and releasing electric energy of the power system; the upper computer controller comprises a processor and a memory, wherein programs are stored in the memory, the processor loads the programs to execute the method steps according to any one of claims 1 to 8, and the control power variation of a conventional generator set, a CHP (combined cycle Power) set and an electric boiler in a unit scheduling period delta T is acquired:
Figure FDA00031390752300001011
and charging and discharging power of energy storage system
Figure FDA00031390752300001012
And outputting the data to each unit on site to realize optimized scheduling.
CN202110727038.9A 2021-06-29 2021-06-29 Day-ahead optimized scheduling method for electric heating integrated energy system Pending CN115544856A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110727038.9A CN115544856A (en) 2021-06-29 2021-06-29 Day-ahead optimized scheduling method for electric heating integrated energy system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110727038.9A CN115544856A (en) 2021-06-29 2021-06-29 Day-ahead optimized scheduling method for electric heating integrated energy system

Publications (1)

Publication Number Publication Date
CN115544856A true CN115544856A (en) 2022-12-30

Family

ID=84717111

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110727038.9A Pending CN115544856A (en) 2021-06-29 2021-06-29 Day-ahead optimized scheduling method for electric heating integrated energy system

Country Status (1)

Country Link
CN (1) CN115544856A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116646933A (en) * 2023-07-24 2023-08-25 北京中能亿信软件有限公司 Big data-based power load scheduling method and system

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116646933A (en) * 2023-07-24 2023-08-25 北京中能亿信软件有限公司 Big data-based power load scheduling method and system
CN116646933B (en) * 2023-07-24 2023-10-10 北京中能亿信软件有限公司 Big data-based power load scheduling method and system

Similar Documents

Publication Publication Date Title
Karadede et al. Breeder hybrid algorithm approach for natural gas demand forecasting model
Khotanzad et al. A neuro-fuzzy approach to short-term load forecasting in a price-sensitive environment
Gareta et al. Forecasting of electricity prices with neural networks
Siahkali et al. Electricity generation scheduling with large-scale wind farms using particle swarm optimization
CN114676941B (en) Electric-thermal load combined self-adaptive prediction method and device for park comprehensive energy system
Dolatabadi et al. A novel model-free deep reinforcement learning framework for energy management of a PV integrated energy hub
CN113822481A (en) Comprehensive energy load prediction method based on multi-task learning strategy and deep learning
Yi et al. Deep reinforcement learning based optimization for a tightly coupled nuclear renewable integrated energy system
CN104036328A (en) Self-adaptive wind power prediction system and prediction method
Jang et al. Optimal HVAC system operation using online learning of interconnected neural networks
CN106779175B (en) Time-interval-considered nonlinear photovoltaic output prediction method
Zhang et al. Research on AGC performance during wind power ramping based on deep reinforcement learning
Kusiak et al. Minimization of wind farm operational cost based on data-driven models
Mehmood et al. Optimal power generation in energy-deficient scenarios using bagging ensembles
Sretenović et al. Support vector machine for the prediction of heating energy use
CN115795992A (en) Park energy Internet online scheduling method based on virtual deduction of operation situation
CN109976155B (en) Method and system for randomly optimizing and controlling interior of virtual power plant participating in gas-electricity market
CN114723230A (en) Micro-grid double-layer scheduling method and system for new energy power generation and energy storage
CN115544856A (en) Day-ahead optimized scheduling method for electric heating integrated energy system
Alharbi et al. Short-term wind speed and temperature forecasting model based on gated recurrent unit neural networks
CN114021783A (en) Two-stage monthly unit combination and maintenance plan optimization method considering social carbon-like emission factors and short-term benefits
CN114119273A (en) Park comprehensive energy system non-invasive load decomposition method and system
CN115001002B (en) Optimal scheduling method and system for solving problem of energy storage participation peak clipping and valley filling
Zhang et al. Data-driven model predictive and reinforcement learning based control for building energy management: A survey
Li et al. Quantifying the impact of building load forecasts on optimizing energy storage systems

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination