CN116300755A - Double-layer optimal scheduling method and device for heat storage-containing heating system based on MPC - Google Patents
Double-layer optimal scheduling method and device for heat storage-containing heating system based on MPC Download PDFInfo
- Publication number
- CN116300755A CN116300755A CN202310301701.8A CN202310301701A CN116300755A CN 116300755 A CN116300755 A CN 116300755A CN 202310301701 A CN202310301701 A CN 202310301701A CN 116300755 A CN116300755 A CN 116300755A
- Authority
- CN
- China
- Prior art keywords
- heat supply
- supply unit
- model
- storage device
- heat storage
- 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
Links
- 238000005338 heat storage Methods 0.000 title claims abstract description 188
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000010438 heat treatment Methods 0.000 title claims description 90
- 238000005457 optimization Methods 0.000 claims abstract description 108
- 230000006870 function Effects 0.000 claims abstract description 59
- 230000009471 action Effects 0.000 claims abstract description 56
- 238000011217 control strategy Methods 0.000 claims abstract description 35
- 238000005096 rolling process Methods 0.000 claims abstract description 26
- 238000012937 correction Methods 0.000 claims abstract description 19
- 238000004422 calculation algorithm Methods 0.000 claims description 71
- 241000272194 Ciconiiformes Species 0.000 claims description 44
- 238000012549 training Methods 0.000 claims description 22
- 239000002245 particle Substances 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 14
- 238000005065 mining Methods 0.000 claims description 12
- 238000010801 machine learning Methods 0.000 claims description 10
- 230000007246 mechanism Effects 0.000 claims description 10
- 238000013507 mapping Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 6
- 238000009825 accumulation Methods 0.000 claims description 6
- 229910052799 carbon Inorganic materials 0.000 claims description 6
- 238000013499 data model Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 238000004088 simulation Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 5
- 230000000739 chaotic effect Effects 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008878 coupling Effects 0.000 claims description 3
- 238000010168 coupling process Methods 0.000 claims description 3
- 238000005859 coupling reaction Methods 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 230000035772 mutation Effects 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 238000011161 development Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000002411 adverse Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000000750 progressive effect Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 230000009191 jumping Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 241000283973 Oryctolagus cuniculus Species 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000003763 carbonization Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000003749 cleanliness Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41865—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32252—Scheduling production, machining, job shop
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses a double-layer optimal scheduling method of a heat supply system containing heat storage based on MPC, which comprises the following steps: configuring a heat storage device in a heat supply system, setting a scheduling instruction by using a predicted output value of a renewable energy heat supply unit, using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, optimizing the heat storage device by using a model prediction control strategy, and optimizing and solving model prediction control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device; the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, the output of the traditional heat supply unit is optimized, and the action instruction of the traditional heat supply unit is obtained.
Description
Technical Field
The invention belongs to the technical field of intelligent heat supply, and particularly relates to a double-layer optimal scheduling method of a heat supply system containing heat storage based on MPC.
Background
The autonomous optimization operation of the intelligent heating system is that the system constructs a digital twin model of the heating system through mechanism modeling and data identification, combines with artificial intelligent technologies such as model predictive control, real-time optimization and the like, and carries out overall optimization operation strategy automatic issuing control system on the heating system from multiple levels so as to realize full-flow automatic safe closed-loop control, thereby enabling the heating system to have self-sensing, self-learning, self-adaption and self-adjustment capabilities.
With the development trend of low carbonization, cleanliness and intelligence of heat supply, a high proportion of renewable energy sources need to be connected into a heat supply system. However, renewable energy sources have the characteristics of randomness, volatility and uncertainty, when renewable energy sources are utilized for heating, fluctuation of a heating system is likely to be caused, safe and stable operation and heating quality of the heating system are affected, and the scheduling difficulty of the heating system is greatly increased, so that adverse effects of the renewable energy sources on the heating scheduling are reduced, the fluctuation of output of the renewable energy sources is stabilized, and the economical efficiency and the reliability of the operation of the heating system are improved.
Based on the technical problems, a new double-layer optimal scheduling method of the heat storage-containing heating system based on the MPC needs to be designed.
Disclosure of Invention
The invention aims to solve the technical problems of overcoming the defects of the prior art and providing a double-layer optimizing scheduling method of a heat supply system containing heat storage based on MPC, wherein the output of a heat storage device can be effectively compensated by establishing a double-layer optimizing scheduling model, the adverse effect, uncertainty and fluctuation of the renewable energy source caused by the uncertainty of the renewable energy source to the scheduling of the heat supply system are reduced, the residual load deficiency is filled by the output of a traditional heat supply unit, the output of the traditional heat supply unit is optimized by utilizing an intelligent optimizing algorithm and is used as the second layer optimization of scheduling, and the economical efficiency and the reliability of the heat supply system can be improved.
In order to solve the technical problems, the technical scheme of the invention is as follows:
the invention provides a double-layer optimal scheduling method of a heat supply system containing heat storage based on MPC, which comprises the following steps:
establishing a digital twin model of the heating system by adopting a mechanism modeling and data identification method; the heat supply unit in the heat supply system at least comprises a renewable energy heat supply unit and a traditional heat supply unit;
Acquiring heat supply historical operation data and outdoor meteorological data based on a heat supply system digital twin model, and establishing a heat supply system load prediction model and a renewable energy heat supply unit output prediction model by adopting a machine learning algorithm to acquire a heat supply system total load requirement and a renewable energy heat supply unit output prediction value;
establishing a first-layer optimized scheduling model: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, optimizing the heat storage device by using the model predictive control strategy by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, and optimizing and solving the model predictive control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device;
establishing a second-layer optimal scheduling model: the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, and the output of the traditional heat supply unit is optimized through an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
Further, the method for establishing the digital twin model of the heating system by adopting mechanism modeling and data identification comprises the following steps:
establishing a physical model, a logic model, a simulation model and a data model of a renewable energy heat supply unit, a traditional heat supply unit, a primary network, a secondary network, a heating station and a tail end building in a heat supply system, performing mutual coupling, multi-layer and multi-scale integration among the physical model, the logic model, the simulation model and the data model, and establishing a digital twin model of the heat supply system after mapping and reconstructing a physical entity in a physical space in a virtual space;
adopting an improved self-adaptive inertia weight particle swarm optimization algorithm to optimize parameters of a digital twin model of the heating system: the combination of the inertia weight nonlinear decremental updating strategy and the mutation operation is introduced into a PSO algorithm to form an improved self-adaptive inertia weight particle swarm optimization algorithm; determining parameters to be identified by a digital twin model of the heating system, setting the value range of each parameter, and converting each parameter optimization problem into a particle position optimization problem; introducing a position variable, solving the deviation of a digital twin model of the heating system by means of training sample data acquired in the heating system, and selecting the optimal position of particles by adopting an improved adaptive inertia weight particle swarm optimization algorithm according to the magnitude of the deviation value to acquire the parameters of the digital twin model of the optimal heating system; and (3) selecting the root mean square error and the average absolute percentage error as measurement indexes, and verifying the performance of the digital twin model of the heating system.
Further, the method for obtaining heat supply historical operation data and outdoor meteorological data based on the heat supply system digital twin model, establishing a heat supply system load prediction model by adopting a machine learning algorithm to obtain a heat supply system total load requirement and a renewable energy heat supply unit output prediction value comprises the following steps:
acquiring historical renewable energy unit output, traditional heat supply unit output, heat supply system load, heat station operating parameters, outdoor temperature, humidity and tail end building indoor temperature based on a heat supply system digital twin model, and taking the historical renewable energy unit output, traditional heat supply unit output, heat supply system load, heat supply station operating parameters, outdoor temperature, humidity and tail end building indoor temperature as a heat supply system load prediction model training sample; acquiring historical renewable energy unit output, heat station operating parameters, outdoor temperature, humidity and indoor temperature of a tail end building, and taking the historical renewable energy unit output, heat station operating parameters, outdoor temperature, humidity and indoor temperature of the tail end building as a renewable energy heat supply unit output prediction model training sample;
after optimizing CNN-REGST model parameters by adopting a POA (point of application) pelican optimization algorithm, establishing a heating system load prediction model and a renewable energy heating unit output prediction model by adopting an optimized CNN-REGST model: after feature extraction is carried out on a model training sample through a CNN convolutional neural network model, adopting REGST stacked regression model theory, respectively inputting data after feature extraction into a plurality of basic learners to obtain a plurality of heat supply system load predicted values and a plurality of renewable energy heat supply unit output predicted values, and then adopting an SVM model as a metaregression to calculate the plurality of predicted values to obtain the heat supply system load predicted values and the renewable energy heat supply unit output predicted values.
Further, the optimizing the parameters of the CNN-REGST model by adopting the POA-pelicant optimization algorithm comprises the following steps:
initializing CNN-REGST model parameters including the number of neurons, the learning rate, the number of nodes and filters of a full connection layer;
optimizing the CNN-REGST model parameters by using a POA pelican optimization algorithm: initializing the number of pelican population members and the maximum iteration number; generating an initial population and calculating an objective function; updating the objective function in the exploration stage and the exploitation stage until the optimal candidate parameters are output;
wherein, the population matrix of pelicans is expressed as:
x is a pelican population matrix, the pelican population members are identified through the population matrix, each row of the matrix represents a candidate solution, and each column represents a suggested value of a problem variable; x is X i Is the i-th pelican; i=1, 2,. -%, N; j=1, 2, m; n is the number of population members; m is the number of problem variables;
the value of the objective function is expressed as:
b is an objective function vector; b (B) i Objective function value for the i-th candidate solution;
in the exploration phase, the process of the pelicans moving to the hunting position is expressed as:
If the value of the objective function is improved at this location, then a new location of the pelican is accepted, expressed as:
during the mining phase, the hunting process of the pelican is expressed as:
in the mining phase, the new location of the pelican is denoted as:
Further, the regt stacked regression model theory includes:
let k basic learners v be provided 1 (x),v 1 (x),…,v k (x) Model training was performed using the same model training samples, expressed as: l= { (y) n ,x n ),n＝1,2,...,N}；x n An input vector for an nth base learner; y is n Is the output vector of the nth base learner;
the combination of n base learners is expressed as:
w k the weight of each basic learner in the combination is calculated; v k (x) Is the kth learner.
Further, the first layer optimized scheduling model is established: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, and optimizing the heat storage device by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and a minimum action quantity of the heat storage device as control optimization targets through the model predictive control strategy, and optimizing and solving model predictive control through rolling optimization and feedback correction links to obtain the action instruction of the heat storage device, wherein the method comprises the following steps:
Configuring a heat storage device in a heat supply system, and introducing a model predictive control strategy into the heat supply system configured with the heat storage device;
taking the average value of the predicted values of different time periods as a scheduling instruction of the scheduling period through the predicted value of the renewable energy heat supply unit output;
the common output tracking and scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device are used as control and optimization targets;
optimizing the heat storage device through a model predictive control strategy, and optimizing and solving the model predictive control through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device;
before the heat storage device is optimized through the model predictive control strategy, the energy is converted into a state space model based on the power and energy balance of a heat supply system integrated by the renewable energy heat supply unit and the heat storage device, and the state space model is expressed as:
the state variable x (k) comprises output power of a renewable energy heat supply unit containing heat storage at the moment k and heat surplus of the heat storage device; the control variable u (k) comprises the heat storage and release power increment of the heat storage device at the moment k; the system output variable y (k) comprises the output power of the renewable energy heat supply unit containing heat storage at the moment k; the uncontrollable variable d (k) of the power calculation comprises the output power increment of the renewable energy heat supply unit at the moment k; A. b (B) 1 、B 2 And C is a known coefficient matrix; the output power of the renewable energy heat supply unit containing heat storage comprises the heat storage and release power of the heat storage device and the output power of the renewable energy heat supply unit;
the method takes a common output tracking and scheduling instruction of a renewable energy heat supply unit and a heat storage device and the minimum action quantity of the heat storage device as a control optimization target, and is expressed as follows:
y (k+j) is the prediction output; r (k+j) is a scheduling instruction of the system at the future time k+j, j=1, 2. t is t w 、t u The error output weight coefficient and the increment weight coefficient of the control variable are respectively; m is a control time domain, and m is less than or equal to p;
the constraint conditions are the operation constraint of the heat storage device, including the capacity constraint of the heat storage device and the heat storage and release power constraint of the heat storage device;
optimizing the heat storage device through a model predictive control strategy, optimizing and solving the model predictive control through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device, wherein the method comprises the following steps: according to the idea of model predictive control, measuring a system output variable value y (k) at the k moment, reversely deducing a state variable x (k) at the k moment, predicting p time periods from the k moment to the back to obtain y (k+ 1|k), y (k+ 2|k) in sequence, wherein y (k+p|k) is the system output response at the k+p moment predicted at the k moment, solving y (k+p|k) to obtain a control variable u (k) at the k moment, acting u (k+ 1|k) on the next moment, and circularly rolling and optimizing and feedback correcting the solved u (k+ 1|k) and the measured y (k+1) at the k+1 moment to obtain an optimal action instruction of the heat storage device;
j=0, 1,..m-1; when m.ltoreq.j.ltoreq.p, u (k+j|k) =u (k+m-1|k); when 1.ltoreq.j.ltoreq.p, d (k+j|k) =d (k).
Further, the second layer optimized scheduling model is established: the utility model provides a renewable energy heat supply unit and heat accumulation device's joint output reduces traditional heat supply unit load, utilizes residual load to optimize traditional heat supply unit to traditional heat supply unit running cost, renewable energy running cost and heat accumulation device running cost minimum are objective function in the dispatch period, optimize traditional heat supply unit output through intelligent optimization algorithm, obtain traditional heat supply unit's action instruction, include:
taking the common output of the renewable energy heat supply unit and the heat storage device as a known condition, subtracting the common output of the renewable energy heat supply unit and the heat storage device from the total load demand of a heat supply system to obtain a residual load, and optimizing the traditional heat supply unit by using the residual load;
the minimum operation cost of the traditional heat supply unit, the minimum operation cost of renewable energy sources and the minimum operation cost of the heat storage device in the scheduling period are used as an objective function, and the method is expressed as follows:
F i (t) is the operating cost of i different conventional heating units; f (F) w (t) the running cost of w different renewable energy heating units; f (F) b (t) is the running cost of the heat storage device b; t is an optimized scheduling period;
setting a power balance constraint condition of a heating system, an output constraint condition of a traditional heating unit, an operation constraint condition of a heat storage device and an operation constraint condition of a renewable energy heating unit;
and adopting an improved Harris eagle optimization algorithm to perform optimization solution on the output of the traditional heat supply unit, and obtaining the action instruction of the traditional heat supply unit.
Further, the improved harris eagle optimization algorithm comprises:
the Tent chaotic map is utilized to improve the initial population, expressed as:
X i ＝lb+Y i (ub-lb)；
i=1, 2,3, …, N-1, the first individual random generation of the population maps onto (0, 1) space, denoted Y 1 The method comprises the steps of carrying out a first treatment on the surface of the The individuals in the remaining N-1 mapping spaces are denoted as Y i+1 ；Y i An ith individual that is a mapping space; x is X i An ith individual that is the initial population; ub and lb are the upper and lower limits of the population position;
the escape energy using non-linearization is expressed as:
E 0 a random number (-1, 1); t is the iteration number; t is the maximum iteration number;
the positions of the population are updated by adopting a golden sine algorithm, which is expressed as follows:
R 1 is [0,2 pi ]]A random number within; r is R 2 Is [0, pi ]]A random number within; τ is the golden section number; r is R 3 And R is 4 The global optimizing capability of the algorithm is determined; x (t) is the position of the t generation population.
Furthermore, when the second-layer optimal scheduling model is established, besides taking the minimum running cost of the traditional heat supply unit, the minimum running cost of renewable energy sources and the minimum running cost of the heat storage device in the scheduling period as an objective function, the minimum carbon emission of the heat supply system can be set as an objective function, the minimum running cost of the heat supply system and the minimum carbon emission are taken as multiple objective functions, the multiple objective functions are converted into a single objective function, and the intelligent optimization algorithm is adopted to optimize the output of the traditional heat supply unit, so that the action instruction of the traditional heat supply unit is obtained.
The invention also provides a double-layer optimizing and dispatching device of the heat supply system containing heat storage of the MPC, which comprises:
the digital twin model building unit is used for building a digital twin model of the heating system by adopting a mechanism modeling and data identification method; the heat supply unit in the heat supply system at least comprises a renewable energy heat supply unit and a traditional heat supply unit;
the prediction model training unit is used for acquiring heat supply historical operation data and outdoor meteorological data based on a heat supply system digital twin model, and establishing a heat supply system load prediction model and a renewable energy heat supply unit output prediction model by adopting a machine learning algorithm to acquire a heat supply system total load requirement and a renewable energy heat supply unit output prediction value;
The double-layer optimal scheduling model building unit is used for building a first-layer optimal scheduling model: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, optimizing the heat storage device by using the model predictive control strategy by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, and optimizing and solving the model predictive control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device; and is also used for establishing a second-layer optimized scheduling model: the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, and the output of the traditional heat supply unit is optimized through an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
The beneficial effects of the invention are as follows:
according to the invention, a model predictive control strategy is introduced into a renewable energy heat supply unit configured with a certain capacity of heat storage device in a first layer of optimal scheduling model, the heat storage device in the renewable energy heat supply unit is optimized through the model predictive control strategy, the common output of the renewable energy and the heat storage device can be close to the expected output which is set in advance, and the action quantity of the heat storage device is minimum as a control optimization target, the model predictive control performs optimal solution through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device, the output of the heat storage device can effectively compensate the output power of the renewable energy, the adverse effect and the uncertainty and the fluctuation of the heat supply system caused by the uncertainty of the renewable energy are reduced, the schedulability is commonly provided, and the optimized common output is equivalent to a benign schedulable heat source for the heat supply system; in the second-layer optimal scheduling model, since the first-layer optimal scheduling model obtains the output of the renewable energy heat supply unit and the heat storage device, when the output of the traditional heat supply unit is optimized, the part is known and is equivalent to subtracting the equivalent load, the rest load deficiency is filled by the output of the traditional heat supply unit, and the output of the traditional heat supply unit is optimized by using an intelligent optimization algorithm to serve as the second-layer optimal scheduling, so that the economy and the reliability of the heat supply system can be improved; in addition, a digital twin model of the heating system is built, a load prediction model of the heating system and an output prediction model of the renewable energy heating unit are built by adopting a machine learning algorithm, so that accurate total load requirements of the heating system and output prediction values of the renewable energy heating unit are obtained, a data foundation is built for double-layer optimal scheduling of the heating system, and accuracy and effectiveness of double-layer optimal scheduling of the whole heating system are improved.
Additional features and advantages will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and drawings.
In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a double-layer optimized dispatching method of a heat supply system containing heat storage based on MPC;
FIG. 2 is a schematic diagram of a double-layer optimized dispatching device of a heat supply system containing heat storage based on MPC.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
FIG. 1 is a flow chart of a double-layer optimal scheduling method of a heat storage-containing heat supply system based on MPC.
As shown in fig. 1, this embodiment 1 provides a double-layer optimization scheduling method for a heat storage-containing heating system based on MPC, which includes:
establishing a digital twin model of the heating system by adopting a mechanism modeling and data identification method; the heat supply unit in the heat supply system at least comprises a renewable energy heat supply unit and a traditional heat supply unit;
acquiring heat supply historical operation data and outdoor meteorological data based on a heat supply system digital twin model, and establishing a heat supply system load prediction model and a renewable energy heat supply unit output prediction model by adopting a machine learning algorithm to acquire a heat supply system total load requirement and a renewable energy heat supply unit output prediction value;
Establishing a first-layer optimized scheduling model: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, optimizing the heat storage device by using the model predictive control strategy by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, and optimizing and solving the model predictive control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device;
establishing a second-layer optimal scheduling model: the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, and the output of the traditional heat supply unit is optimized through an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
In the first layer of optimal scheduling model, a model prediction control strategy is introduced into a renewable energy heating unit configured with a certain capacity of heat storage device, the heat storage device in the renewable energy heating unit is optimized through the model prediction control strategy, the common output of the renewable energy and the heat storage device can be close to the expected output which is set in advance, and the action quantity of the heat storage device is minimum as a control optimization target, the model prediction control is optimized and solved through a rolling optimization and feedback correction link to obtain the action command of the heat storage device, the heat storage device can effectively compensate the output power of the renewable energy, the adverse effect and the uncertainty and the fluctuation caused by the uncertainty of the renewable energy to the heat supply system scheduling are reduced, the common output has schedulability, and the optimized common output is equivalent to a benign schedulable heat source for the heat supply system; in the second-layer optimal scheduling model, since the first-layer optimal scheduling model obtains the output of the renewable energy heat supply unit and the heat storage device, when the output of the traditional heat supply unit is optimized, the part is known, which is equivalent to subtracting the equivalent load, the rest load deficiency is filled by the output of the traditional heat supply unit, and the output of the traditional heat supply unit is optimized by utilizing the intelligent optimization algorithm, and as the second-layer optimization of scheduling, the economy and the reliability of the heat supply system can be improved.
In this embodiment, the method for establishing a digital twin model of a heating system by adopting mechanism modeling and data identification includes:
establishing a physical model, a logic model, a simulation model and a data model of a renewable energy heat supply unit, a traditional heat supply unit, a primary network, a secondary network, a heating station and a tail end building in a heat supply system, performing mutual coupling, multi-layer and multi-scale integration among the physical model, the logic model, the simulation model and the data model, and establishing a digital twin model of the heat supply system after mapping and reconstructing a physical entity in a physical space in a virtual space;
adopting an improved self-adaptive inertia weight particle swarm optimization algorithm to optimize parameters of a digital twin model of the heating system: the combination of the inertia weight nonlinear decremental updating strategy and the mutation operation is introduced into a PSO algorithm to form an improved self-adaptive inertia weight particle swarm optimization algorithm; determining parameters to be identified by a digital twin model of the heating system, setting the value range of each parameter, and converting each parameter optimization problem into a particle position optimization problem; introducing a position variable, solving the deviation of a digital twin model of the heating system by means of training sample data acquired in the heating system, and selecting the optimal position of particles by adopting an improved adaptive inertia weight particle swarm optimization algorithm according to the magnitude of the deviation value to acquire the parameters of the digital twin model of the optimal heating system; and (3) selecting the root mean square error and the average absolute percentage error as measurement indexes, and verifying the performance of the digital twin model of the heating system.
In this embodiment, the obtaining heat supply historical operation data and outdoor weather data based on the heat supply system digital twin model, and establishing a heat supply system load prediction model by using a machine learning algorithm, to obtain a heat supply system total load requirement and a renewable energy heat supply unit output prediction value, includes:
acquiring historical renewable energy unit output, traditional heat supply unit output, heat supply system load, heat station operating parameters, outdoor temperature, humidity and tail end building indoor temperature based on a heat supply system digital twin model, and taking the historical renewable energy unit output, traditional heat supply unit output, heat supply system load, heat supply station operating parameters, outdoor temperature, humidity and tail end building indoor temperature as a heat supply system load prediction model training sample; acquiring historical renewable energy unit output, heat station operating parameters, outdoor temperature, humidity and indoor temperature of a tail end building, and taking the historical renewable energy unit output, heat station operating parameters, outdoor temperature, humidity and indoor temperature of the tail end building as a renewable energy heat supply unit output prediction model training sample;
after optimizing CNN-REGST model parameters by adopting a POA (point of application) pelican optimization algorithm, establishing a heating system load prediction model and a renewable energy heating unit output prediction model by adopting an optimized CNN-REGST model: after feature extraction is carried out on a model training sample through a CNN convolutional neural network model, adopting REGST stacked regression model theory, respectively inputting data after feature extraction into a plurality of basic learners to obtain a plurality of heat supply system load predicted values and a plurality of renewable energy heat supply unit output predicted values, and then adopting an SVM model as a metaregression to calculate the plurality of predicted values to obtain the heat supply system load predicted values and the renewable energy heat supply unit output predicted values.
In this embodiment, the optimizing the CNN-regt model parameters using the POA pelican optimization algorithm includes:
initializing CNN-REGST model parameters including the number of neurons, the learning rate, the number of nodes and filters of a full connection layer;
optimizing the CNN-REGST model parameters by using a POA pelican optimization algorithm: initializing the number of pelican population members and the maximum iteration number; generating an initial population and calculating an objective function; updating the objective function in the exploration stage and the exploitation stage until the optimal candidate parameters are output;
wherein, the population matrix of pelicans is expressed as:
x is a pelican population matrix, the pelican population members are identified through the population matrix, each row of the matrix represents a candidate solution, and each column represents a suggested value of a problem variable; x is X i Is the i-th pelican; i=1, 2,. -%, N; j=1, 2, m; n is the number of population members; m is the number of problem variables;
the value of the objective function is expressed as:
b is an objective function vector; b (B) i Objective function value for the i-th candidate solution;
in the exploration phase, the process of the pelicans moving to the hunting position is expressed as:
If the value of the objective function is improved at this location, then a new location of the pelican is accepted, expressed as:
during the mining phase, the hunting process of the pelican is expressed as:
in the mining phase, the new location of the pelican is denoted as:
It should be noted that, training CNN to extract the prediction feature, regt as regression operator algorithm, the hybrid model has more accurate prediction precision than the single model; and the parameters of the CNN-REGST model are optimized by using the pelican optimization algorithm POA, so that the generalization capability and the actual operability of the model are improved.
In this embodiment, the regt stacked regression model theory includes:
let k basic learners v be provided 1 (x),v 1 (x),...,v k (x) Model training was performed using the same model training samples, expressed as: l= { (y) n ,x n ),n＝1,2,...,N}；x n An input vector for an nth base learner; y is n Is the output vector of the nth base learner;
the combination of n base learners is expressed as:
w k the weight of each basic learner in the combination is calculated; v k (x) Is the kth learner.
In this embodiment, the first layer optimization scheduling model is established: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, and optimizing the heat storage device by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and a minimum action quantity of the heat storage device as control optimization targets through the model predictive control strategy, and optimizing and solving model predictive control through rolling optimization and feedback correction links to obtain the action instruction of the heat storage device, wherein the method comprises the following steps:
configuring a heat storage device in a heat supply system, and introducing a model predictive control strategy into the heat supply system configured with the heat storage device;
taking the average value of the predicted values of different time periods as a scheduling instruction of the scheduling period through the predicted value of the renewable energy heat supply unit output;
the common output tracking and scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device are used as control and optimization targets;
Optimizing the heat storage device through a model predictive control strategy, and optimizing and solving the model predictive control through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device;
before the heat storage device is optimized through the model predictive control strategy, the energy is converted into a state space model based on the power and energy balance of a heat supply system integrated by the renewable energy heat supply unit and the heat storage device, and the state space model is expressed as:
the state variable x (k) comprises output power of a renewable energy heat supply unit containing heat storage at the moment k and heat surplus of the heat storage device; the control variable u (k) comprises the heat storage and release power increment of the heat storage device at the moment k; the system output variable y (k) comprises the output power of the renewable energy heat supply unit containing heat storage at the moment k; the uncontrollable variable d (k) of the power calculation comprises the output power increment of the renewable energy heat supply unit at the moment k; A. b (B) 1 、B 2 And C is a known coefficient matrix; the output power of the renewable energy heat supply unit containing heat storage comprises the heat storage and release power of the heat storage device and the output power of the renewable energy heat supply unit;
the method takes a common output tracking and scheduling instruction of a renewable energy heat supply unit and a heat storage device and the minimum action quantity of the heat storage device as a control optimization target, and is expressed as follows:
y (k+j) is the prediction output; r (k+j) is a scheduling instruction of the system at the future time k+j, j=1, 2. t is t w 、t u The error output weight coefficient and the increment weight coefficient of the control variable are respectively; m is a control time domain, and m is less than or equal to p;
the constraint conditions are the operation constraint of the heat storage device, including the capacity constraint of the heat storage device and the heat storage and release power constraint of the heat storage device;
optimizing the heat storage device through a model predictive control strategy, optimizing and solving the model predictive control through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device, wherein the method comprises the following steps: according to the idea of model predictive control, measuring a system output variable value y (k) at the k moment, reversely deducing a state variable x (k) at the k moment, predicting p time periods from the k moment to the back to obtain y (k+ 1|k), y (k+ 2|k) in sequence, wherein y (k+p|k) is the system output response at the k+p moment predicted at the k moment, solving y (k+p|k) to obtain a control variable u (k) at the k moment, acting u (k+ 1|k) on the next moment, and circularly rolling and optimizing and feedback correcting the solved u (k+ 1|k) and the measured y (k+1) at the k+1 moment to obtain an optimal action instruction of the heat storage device;
j=0, 1,..m-1; when m.ltoreq.j.ltoreq.p, u (k+j|k) =u (k+m-1|k); when 1.ltoreq.j.ltoreq.p, d (k+j|k) =d (k).
It should be noted that, the MPC has more rolling optimization links than the conventional control, and is different from the conventional control in that the MPC solves the optimal solution at each moment instead of fixing the unique global optimal solution, and the optimization process is not completed once, but is repeatedly performed online, and the optimization of the performance index only considers from the current moment to a certain limited moment in the future, but the optimization process is repeated until the next moment, that is, the optimization period advances forward, and errors caused by factors such as model mismatch and environmental interference can be compensated by solving the optimal solution at each moment, so that the final optimization control result can be closer to reality. Feedback correction: the MPC has the characteristic of rolling optimization, each moment obtains an optimal control sequence from the current moment to some finite time in the future, in order to avoid larger errors in system control caused by some influencing factors, each moment repeatedly carries out forward rolling optimization, each moment only adopts an optimal control solution of the current moment, namely, each time when the control is implemented, only uses a first control variable in the control sequence to predict an output value of the future moment, and the real-time measurement information of an actual system is introduced as a feedback value, and the rolling optimization process of model prediction is not only model-based optimization, but also utilizes the actual measurement information of the system to form a closed-loop rolling optimization feedback system.
In this embodiment, the establishing a second layer optimized scheduling model: the utility model provides a renewable energy heat supply unit and heat accumulation device's joint output reduces traditional heat supply unit load, utilizes residual load to optimize traditional heat supply unit to traditional heat supply unit running cost, renewable energy running cost and heat accumulation device running cost minimum are objective function in the dispatch period, optimize traditional heat supply unit output through intelligent optimization algorithm, obtain traditional heat supply unit's action instruction, include:
taking the common output of the renewable energy heat supply unit and the heat storage device as a known condition, subtracting the common output of the renewable energy heat supply unit and the heat storage device from the total load demand of a heat supply system to obtain a residual load, and optimizing the traditional heat supply unit by using the residual load;
the minimum operation cost of the traditional heat supply unit, the minimum operation cost of renewable energy sources and the minimum operation cost of the heat storage device in the scheduling period are used as an objective function, and the method is expressed as follows:
F i (t) operation for i different conventional heating unitsCost; f (F) w (t) the running cost of w different renewable energy heating units; f (F) b (t) is the running cost of the heat storage device b; t is an optimized scheduling period;
Setting a power balance constraint condition of a heating system, an output constraint condition of a traditional heating unit, an operation constraint condition of a heat storage device and an operation constraint condition of a renewable energy heating unit;
and adopting an improved Harris eagle optimization algorithm to perform optimization solution on the output of the traditional heat supply unit, and obtaining the action instruction of the traditional heat supply unit.
In this embodiment, the improved harris eagle optimization algorithm includes:
the Tent chaotic map is utilized to improve the initial population, expressed as:
X i ＝lb+Y i (ub-lb)；
i=1, 2, 3..n-1, the first individual random generation of the population maps onto (0, 1) space, denoted Y 1 The method comprises the steps of carrying out a first treatment on the surface of the The individuals in the remaining N-1 mapping spaces are denoted as Y i+1 ；Y i An ith individual that is a mapping space; x is X i An ith individual that is the initial population; ub and lb are the upper and lower limits of the population position;
the escape energy using non-linearization is expressed as:
E 0 a random number (-1, 1); t is the iteration number; t is the maximum iteration number;
the positions of the population are updated by adopting a golden sine algorithm, which is expressed as follows:
R 1 is [0,2 pi ]]A random number within; r is R 2 Is [0, pi ]]A random number within; τ is the golden section number; r is R 3 And R is 4 The global optimizing capability of the algorithm is determined; x (t) is the position of the t generation population.
It should be noted that the harris eagle optimization algorithm includes 3 stages: a global exploration phase, a conversion phase from the exploration phase to the development phase and a local development phase; global exploration phase: the halisk is a rag and is a group of domestic animals, and is generally predated in a collective way, when the escape energy absolute value of the prey is larger than 1, the prey is full of physical force, the halisk is far away from the prey, and the halisk is in a stage of exploring the prey; from the exploration phase to the development transition phase: the Harris eagle converts different hunting stages according to different escape energies of the hunting, and enters a global exploration stage when the absolute value of the escape energy is greater than or equal to 1; otherwise, entering a local development stage; local development stage: there are 4 possible attack strategies according to the escape behavior of the prey and the chase strategy of the harris eagle, when the escape probability is less than 0.5, it is shown that the rabbit successfully escapes before the attack; when the escape probability is more than or equal to 0.5, indicating that the rabbit escapes; when the absolute value of escape energy is more than or equal to 0.5 and less than 1, and the escape probability is more than or equal to 0.5, the Harris hawk adopts a soft surrounding strategy; when the absolute value of escape energy is smaller than 0.5 and the escape probability is larger than or equal to 0.5, adopting a hard surrounding strategy; when the absolute value of escape energy is more than or equal to 0.5 and less than 1, and the escape probability is less than 0.5, the Harris eagle adopts a progressive dive rapid soft surrounding strategy; when the escape energy absolute value is less than 0.5 and the escape probability is less than 0.5, the harris eagle adopts a progressive dive rapid hard surrounding strategy.
Soft surround strategy means that the prey cannot escape the surround but still has enough energy to escape, at which stage the prey will confuse the harris eagle with jumping action; the hard enclosure strategy means that the prey fails to escape the enclosure and does not have enough energy to escape, at which time the harris eagle adopts the hard enclosure strategy and carries out the attack; the progressive dive rapid soft wrapping strategy means that the prey successfully escapes from wrapping of the harris eagle and enough energy successfully escapes, the harris eagle can perform dive attack on the prey, and if the dive attack is unsuccessful, the harris eagle can perform random dive attack on the prey based on the Lewy flight; the progressive dive rapid hard wrapping strategy means that the prey successfully escapes from the wrapping ring, but insufficient energy escapes, the harris eagle shortens the average distance with the prey first, then the prey is attacked, and if the attack is unsuccessful, the random attack is carried out.
The original Harris hawk optimizing algorithm initial population is randomly generated, which is not beneficial to the rapid convergence of the algorithm, and the initial population which is distributed uniformly can be obtained by introducing the Tent chaotic map, so that the convergence speed of the algorithm is increased; in the original Harris hawk optimization algorithm, escape energy is linearly decreased along with iteration times, however, the size of the escape energy directly influences the global exploration and local development capacity of the Harris hawk optimization algorithm, and the adoption of nonlinear escape energy is more beneficial to the convergence of the algorithm; and updating the position of the population by adopting a golden sine algorithm with strong global optimizing capability, and jumping out of local optimization. The Harris eagle optimization algorithm is improved, so that the capability of jumping out of local optimum is improved, the convergence rate of the algorithm is accelerated, the maximum iteration times are greatly reduced, and the stability of the algorithm is enhanced.
In this embodiment, when the second-layer optimized scheduling model is established, in addition to taking the minimum of the operation cost, the renewable energy source operation cost and the operation cost of the heat storage device of the traditional heat supply unit in the scheduling period as an objective function, the minimum carbon emission of the heat supply system can be set as an objective function, the minimum of the operation cost and the carbon emission of the heat supply system is a multi-objective function, the multi-objective function is converted into a single-objective function, and the output of the traditional heat supply unit is optimized by adopting an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
Example 2
FIG. 2 is a schematic diagram of a double-layer optimized dispatching device of a heat supply system containing heat storage based on MPC.
As shown in fig. 2, this embodiment 2 provides a double-layer optimizing and dispatching device for a heat storage-containing heating system of MPC, which includes:
the digital twin model building unit is used for building a digital twin model of the heating system by adopting a mechanism modeling and data identification method; the heat supply unit in the heat supply system at least comprises a renewable energy heat supply unit and a traditional heat supply unit;
the prediction model training unit is used for acquiring heat supply historical operation data and outdoor meteorological data based on a heat supply system digital twin model, and establishing a heat supply system load prediction model and a renewable energy heat supply unit output prediction model by adopting a machine learning algorithm to acquire a heat supply system total load requirement and a renewable energy heat supply unit output prediction value;
The double-layer optimal scheduling model building unit is used for building a first-layer optimal scheduling model: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, optimizing the heat storage device by using the model predictive control strategy by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, and optimizing and solving the model predictive control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device; and is also used for establishing a second-layer optimized scheduling model: the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, and the output of the traditional heat supply unit is optimized through an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
In the several embodiments provided in this application, it should be understood that the disclosed systems and methods may be implemented in other ways as well. The system embodiments described above are merely illustrative, for example, of the flowcharts and block diagrams in the figures that illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present invention may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part. The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored on a computer readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
With the above-described preferred embodiments according to the present invention as an illustration, the above-described descriptions can be used by persons skilled in the relevant art to make various changes and modifications without departing from the scope of the technical idea of the present invention. The technical scope of the present invention is not limited to the description, but must be determined according to the scope of claims.
Claims (10)
1. The double-layer optimal scheduling method for the heat storage-containing heating system based on the MPC is characterized by comprising the following steps of:
establishing a digital twin model of the heating system by adopting a mechanism modeling and data identification method; the heat supply unit in the heat supply system at least comprises a renewable energy heat supply unit and a traditional heat supply unit;
acquiring heat supply historical operation data and outdoor meteorological data based on a heat supply system digital twin model, and establishing a heat supply system load prediction model and a renewable energy heat supply unit output prediction model by adopting a machine learning algorithm to acquire a heat supply system total load requirement and a renewable energy heat supply unit output prediction value;
establishing a first-layer optimized scheduling model: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, optimizing the heat storage device by using the model predictive control strategy by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, and optimizing and solving the model predictive control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device;
Establishing a second-layer optimal scheduling model: the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, and the output of the traditional heat supply unit is optimized through an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
2. The heating system double-layer optimized dispatching method according to claim 1, wherein the method for establishing a heating system digital twin model by adopting mechanism modeling and data identification comprises the following steps:
establishing a physical model, a logic model, a simulation model and a data model of a renewable energy heat supply unit, a traditional heat supply unit, a primary network, a secondary network, a heating station and a tail end building in a heat supply system, performing mutual coupling, multi-layer and multi-scale integration among the physical model, the logic model, the simulation model and the data model, and establishing a digital twin model of the heat supply system after mapping and reconstructing a physical entity in a physical space in a virtual space;
Adopting an improved self-adaptive inertia weight particle swarm optimization algorithm to optimize parameters of a digital twin model of the heating system: the combination of the inertia weight nonlinear decremental updating strategy and the mutation operation is introduced into a PSO algorithm to form an improved self-adaptive inertia weight particle swarm optimization algorithm; determining parameters to be identified by a digital twin model of the heating system, setting the value range of each parameter, and converting each parameter optimization problem into a particle position optimization problem; introducing a position variable, solving the deviation of a digital twin model of the heating system by means of training sample data acquired in the heating system, and selecting the optimal position of particles by adopting an improved adaptive inertia weight particle swarm optimization algorithm according to the magnitude of the deviation value to acquire the parameters of the digital twin model of the optimal heating system; and (3) selecting the root mean square error and the average absolute percentage error as measurement indexes, and verifying the performance of the digital twin model of the heating system.
3. The method for double-layer optimized dispatching of heat supply system according to claim 1, wherein the obtaining heat supply history operation data and outdoor meteorological data based on the heat supply system digital twin model, establishing a heat supply system load prediction model by adopting a machine learning algorithm, obtaining a heat supply system total load demand and a renewable energy heat supply unit output prediction value, comprises:
Acquiring historical renewable energy unit output, traditional heat supply unit output, heat supply system load, heat station operating parameters, outdoor temperature, humidity and tail end building indoor temperature based on a heat supply system digital twin model, and taking the historical renewable energy unit output, traditional heat supply unit output, heat supply system load, heat supply station operating parameters, outdoor temperature, humidity and tail end building indoor temperature as a heat supply system load prediction model training sample; acquiring historical renewable energy unit output, heat station operating parameters, outdoor temperature, humidity and indoor temperature of a tail end building, and taking the historical renewable energy unit output, heat station operating parameters, outdoor temperature, humidity and indoor temperature of the tail end building as a renewable energy heat supply unit output prediction model training sample;
after optimizing CNN-REGST model parameters by adopting a POA (point of application) pelican optimization algorithm, establishing a heating system load prediction model and a renewable energy heating unit output prediction model by adopting an optimized CNN-REGST model: after feature extraction is carried out on a model training sample through a CNN convolutional neural network model, adopting REGST stacked regression model theory, respectively inputting data after feature extraction into a plurality of basic learners to obtain a plurality of heat supply system load predicted values and a plurality of renewable energy heat supply unit output predicted values, and then adopting an SVM model as a metaregression to calculate the plurality of predicted values to obtain the heat supply system load predicted values and the renewable energy heat supply unit output predicted values.
4. A heating system double-layer optimized scheduling method according to claim 3, wherein said optimizing CNN-regt model parameters using POA-pelargon optimization algorithm comprises:
initializing CNN-REGST model parameters including the number of neurons, the learning rate, the number of nodes and filters of a full connection layer;
optimizing the CNN-REGST model parameters by using a POA pelican optimization algorithm: initializing the number of pelican population members and the maximum iteration number; generating an initial population and calculating an objective function; updating the objective function in the exploration stage and the exploitation stage until the optimal candidate parameters are output;
wherein, the population matrix of pelicans is expressed as:
x is the population matrix of pelicans, and pelicans is identified through the population matrixEach row of the matrix representing a candidate solution and each column representing a proposed value of the problem variable; x is X i Is the i-th pelican; i=1, 2,. -%, N; j=1, 2, m; n is the number of population members; m is the number of problem variables;
the value of the objective function is expressed as:
b is an objective function vector; b (B) i Objective function value for the i-th candidate solution;
in the exploration phase, the process of the pelicans moving to the hunting position is expressed as:
If the value of the objective function is improved at this location, then a new location of the pelican is accepted, expressed as:
during the mining phase, the hunting process of the pelican is expressed as:
in the mining phase, the new location of the pelican is denoted as:
5. A heating system double-layer optimization scheduling method according to claim 3, wherein the regt stacked regression model theory comprises:
let k basic learners v be provided 1 (x),v 1 (x),…,v k (x) Model training was performed using the same model training samples, expressed as: l= { (y) n ,x n ),n＝1,2,...,N}；x n An input vector for an nth base learner; y is n Is the nth radicalAn output vector of the base learner;
the combination of n base learners is expressed as: 
w k The weight of each basic learner in the combination is calculated; v k (x) Is the kth learner.
6. The heating system double-layer optimal scheduling method according to claim 1, wherein the first layer optimal scheduling model is established: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, and optimizing the heat storage device by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and a minimum action quantity of the heat storage device as control optimization targets through the model predictive control strategy, and optimizing and solving model predictive control through rolling optimization and feedback correction links to obtain the action instruction of the heat storage device, wherein the method comprises the following steps:
configuring a heat storage device in a heat supply system, and introducing a model predictive control strategy into the heat supply system configured with the heat storage device;
taking the average value of the predicted values of different time periods as a scheduling instruction of the scheduling period through the predicted value of the renewable energy heat supply unit output;
the common output tracking and scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device are used as control and optimization targets;
Optimizing the heat storage device through a model predictive control strategy, and optimizing and solving the model predictive control through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device;
before the heat storage device is optimized through the model predictive control strategy, the energy is converted into a state space model based on the power and energy balance of a heat supply system integrated by the renewable energy heat supply unit and the heat storage device, and the state space model is expressed as:
the state variable x (k) comprises output power of a renewable energy heat supply unit containing heat storage at the moment k and heat surplus of the heat storage device; the control variable u (k) comprises the heat storage and release power increment of the heat storage device at the moment k; the system output variable y (k) comprises the output power of the renewable energy heat supply unit containing heat storage at the moment k; the uncontrollable variable d (k) of the power calculation comprises the output power increment of the renewable energy heat supply unit at the moment k; A. b (B) 1 、B 2 And C is a known coefficient matrix; the output power of the renewable energy heat supply unit containing heat storage comprises the heat storage and release power of the heat storage device and the output power of the renewable energy heat supply unit;
the method takes a common output tracking and scheduling instruction of a renewable energy heat supply unit and a heat storage device and the minimum action quantity of the heat storage device as a control optimization target, and is expressed as follows:
y (k+j) is the prediction output; r (k+j) is a scheduling instruction of the system at the future time k+j, j=1, 2. t is t w 、t u The error output weight coefficient and the increment weight coefficient of the control variable are respectively; m is a control time domain, and m is less than or equal to p;
the constraint conditions are the operation constraint of the heat storage device, including the capacity constraint of the heat storage device and the heat storage and release power constraint of the heat storage device;
optimizing the heat storage device through a model predictive control strategy, optimizing and solving the model predictive control through a rolling optimization and feedback correction link to obtain an action instruction of the heat storage device, wherein the method comprises the following steps: according to the idea of model predictive control, measuring a system output variable value y (k) at the k moment, reversely deducing a state variable x (k) at the k moment, predicting p time periods from the k moment to the back to obtain y (k+ 1|k), y (k+ 2|k) in sequence, wherein y (k+p|k) is the system output response at the k+p moment predicted at the k moment, solving y (k+p|k) to obtain a control variable u (k) at the k moment, acting u (k+ 1|k) on the next moment, and circularly rolling and optimizing and feedback correcting the solved u (k+ 1|k) and the measured y (k+1) at the k+1 moment to obtain an optimal action instruction of the heat storage device;
j=0, 1,..m-1; when m.ltoreq.j.ltoreq.p, u (k+j|k) =u (k+m-1|k); when 1.ltoreq.j.ltoreq.p, d (k+j|k) =d (k).
7. The heating system double-layer optimal scheduling method according to claim 1, wherein the second-layer optimal scheduling model is established: the utility model provides a renewable energy heat supply unit and heat accumulation device's joint output reduces traditional heat supply unit load, utilizes residual load to optimize traditional heat supply unit to traditional heat supply unit running cost, renewable energy running cost and heat accumulation device running cost minimum are objective function in the dispatch period, optimize traditional heat supply unit output through intelligent optimization algorithm, obtain traditional heat supply unit's action instruction, include:
taking the common output of the renewable energy heat supply unit and the heat storage device as a known condition, subtracting the common output of the renewable energy heat supply unit and the heat storage device from the total load demand of a heat supply system to obtain a residual load, and optimizing the traditional heat supply unit by using the residual load;
the minimum operation cost of the traditional heat supply unit, the minimum operation cost of renewable energy sources and the minimum operation cost of the heat storage device in the scheduling period are used as an objective function, and the method is expressed as follows:
F i (t) is the operating cost of i different conventional heating units; f (F) w (t) the running cost of w different renewable energy heating units; f (F) b (t) isThe operating cost of the heat storage device b; t is an optimized scheduling period;
setting a power balance constraint condition of a heating system, an output constraint condition of a traditional heating unit, an operation constraint condition of a heat storage device and an operation constraint condition of a renewable energy heating unit;
and adopting an improved Harris eagle optimization algorithm to perform optimization solution on the output of the traditional heat supply unit, and obtaining the action instruction of the traditional heat supply unit.
8. The heating system double-layer optimized scheduling method of claim 7, wherein the improved harris eagle optimization algorithm comprises:
the Tent chaotic map is utilized to improve the initial population, expressed as:
X i ＝lb+Y i (ub-lb)；
i=1, 2,3, …, N-1, the first individual random generation of the population maps onto (0, 1) space, denoted Y 1 The method comprises the steps of carrying out a first treatment on the surface of the The individuals in the remaining N-1 mapping spaces are denoted as Y i+1 ；Y i An ith individual that is a mapping space; x is X i An ith individual that is the initial population; ub and lb are the upper and lower limits of the population position;
the escape energy using non-linearization is expressed as:
E 0 a random number (-1, 1); t is the iteration number; t is the maximum iteration number;
the positions of the population are updated by adopting a golden sine algorithm, which is expressed as follows:
R 1 is [0,2 pi ]]A random number within; r is R 2 Is [0, pi ]]A random number within; τ is the golden section number; r is R 3 And R is 4 The global optimizing capability of the algorithm is determined; x (t) is the position of the t generation population.
9. The method for double-layer optimal scheduling of a heating system according to claim 1, wherein when the second-layer optimal scheduling model is established, besides the minimum running cost of the traditional heating unit, the minimum running cost of renewable energy sources and the minimum running cost of the heat storage device in a scheduling period are taken as target functions, the minimum carbon emission of the heating system is taken as a target function, the minimum running cost of the heating system and the minimum carbon emission are taken as multiple target functions, the multiple target functions are converted into a single target function, and the output of the traditional heating unit is optimized by adopting an intelligent optimization algorithm, so that the action instruction of the traditional heating unit is obtained.
10. A double-layer optimized dispatching device for a heat storage-containing heating system of an MPC, comprising:
the digital twin model building unit is used for building a digital twin model of the heating system by adopting a mechanism modeling and data identification method; the heat supply unit in the heat supply system at least comprises a renewable energy heat supply unit and a traditional heat supply unit;
the prediction model training unit is used for acquiring heat supply historical operation data and outdoor meteorological data based on a heat supply system digital twin model, and establishing a heat supply system load prediction model and a renewable energy heat supply unit output prediction model by adopting a machine learning algorithm to acquire a heat supply system total load requirement and a renewable energy heat supply unit output prediction value;
The double-layer optimal scheduling model building unit is used for building a first-layer optimal scheduling model: introducing a model predictive control strategy into a heat supply system for configuring a heat storage device, setting a scheduling instruction by using a renewable energy heat supply unit output predictive value, optimizing the heat storage device by using the model predictive control strategy by using a common output tracking scheduling instruction of the renewable energy heat supply unit and the heat storage device and the minimum action quantity of the heat storage device as control optimization targets, and optimizing and solving the model predictive control by using rolling optimization and feedback correction links to obtain the action instruction of the heat storage device; and is also used for establishing a second-layer optimized scheduling model: the load of the traditional heat supply unit is reduced through the common output of the renewable energy heat supply unit and the heat storage device, the residual load is utilized to optimize the traditional heat supply unit, the running cost of the renewable energy and the running cost of the heat storage device in a scheduling period are minimum as objective functions, and the output of the traditional heat supply unit is optimized through an intelligent optimization algorithm, so that the action instruction of the traditional heat supply unit is obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310301701.8A CN116300755A (en) | 2023-03-24 | 2023-03-24 | Double-layer optimal scheduling method and device for heat storage-containing heating system based on MPC |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310301701.8A CN116300755A (en) | 2023-03-24 | 2023-03-24 | Double-layer optimal scheduling method and device for heat storage-containing heating system based on MPC |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116300755A true CN116300755A (en) | 2023-06-23 |
Family
ID=86783228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310301701.8A Pending CN116300755A (en) | 2023-03-24 | 2023-03-24 | Double-layer optimal scheduling method and device for heat storage-containing heating system based on MPC |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116300755A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118195279A (en) * | 2024-05-16 | 2024-06-14 | 北京观天执行科技股份有限公司 | Heating management scheduling method and system based on renewable energy intelligent coupling |
-
2023
- 2023-03-24 CN CN202310301701.8A patent/CN116300755A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118195279A (en) * | 2024-05-16 | 2024-06-14 | 北京观天执行科技股份有限公司 | Heating management scheduling method and system based on renewable energy intelligent coupling |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112653142B (en) | Wind power prediction method and system for optimizing depth transform network | |
| CN112186743B (en) | Dynamic power system economic dispatching method based on deep reinforcement learning | |
| CN112733462B (en) | Ultra-short-term wind power plant power prediction method combining meteorological factors | |
| CN112116144B (en) | Regional power distribution network short-term load prediction method | |
| CN103729695A (en) | Short-term power load forecasting method based on particle swarm and BP neural network | |
| CN113705922B (en) | Improved ultra-short-term wind power prediction algorithm and model building method | |
| CN114676941B (en) | Electric-thermal load combined self-adaptive prediction method and device for park comprehensive energy system | |
| CN111897240B (en) | Simulation method and system based on nuclear power system operation | |
| Kalogirou | Artificial neural networks and genetic algorithms for the modeling, simulation, and performance prediction of solar energy systems | |
| CN116300755A (en) | Double-layer optimal scheduling method and device for heat storage-containing heating system based on MPC | |
| CN114909706B (en) | Two-level network balance regulation and control method based on reinforcement learning algorithm and differential pressure control | |
| CN115640901A (en) | Small sample load prediction method based on hybrid neural network and generation countermeasure | |
| CN115765050A (en) | Power system safety correction control method, system, equipment and storage medium | |
| CN115907122A (en) | Regional electric vehicle charging load prediction method | |
| CN111401659A (en) | Ultra-short-term or short-term photovoltaic power generation power prediction method based on case reasoning | |
| CN114429248A (en) | Transformer apparent power prediction method | |
| CN113609762A (en) | Electric cooling and heating load combined prediction method and system based on GRU-MTL | |
| CN116681154A (en) | Photovoltaic power calculation method based on EMD-AO-DELM | |
| CN111027760A (en) | Power load prediction method based on least square vector machine | |
| CN110543942A (en) | Multi-space-time long and short memory depth network accurate prediction method | |
| CN114204546B (en) | Unit combination optimization method considering new energy consumption | |
| CN115828726A (en) | Machine learning-based solar cross-season heat storage and supply system optimization control method | |
| CN115544856A (en) | Day-ahead optimized scheduling method for electric heating integrated energy system | |
| CN114759579A (en) | Power grid active power optimization control system, method and medium based on data driving | |
| CN113283638A (en) | Load extreme curve prediction method and system based on fusion model |
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 |