CN115965156A - Scheduling method and scheduling device of energy system - Google Patents
Scheduling method and scheduling device of energy system Download PDFInfo
- Publication number
- CN115965156A CN115965156A CN202310044061.7A CN202310044061A CN115965156A CN 115965156 A CN115965156 A CN 115965156A CN 202310044061 A CN202310044061 A CN 202310044061A CN 115965156 A CN115965156 A CN 115965156A
- Authority
- CN
- China
- Prior art keywords
- energy system
- optimization model
- energy
- historical data
- steam
- 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
Images
Classifications
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention provides a scheduling method of an energy system, a scheduling device of the energy system and a computer readable storage medium. The scheduling method of the energy system comprises the following steps: developing a deterministic energy system optimization model according to the process mechanism and the operation characteristics of the energy system; acquiring historical data of an energy system; performing cluster analysis on the historical data to map the historical data into clusters; determining an uncertain set and corresponding probabilities of a plurality of scenes in a deterministic energy system optimization model; determining a random robust optimization model according to the deterministic energy system optimization model, the historical data, the cluster and the uncertain set; and determining a scheduling scheme of the energy system via the stochastic robust optimization model.
Description
Technical Field
The present invention relates to the field of industrial energy systems, and in particular, to a method and an apparatus for scheduling an energy system, and a computer readable storage medium thereof.
Background
With the rapid development of economy in China, the consumption of various forms of energy in industry is higher and higher, and petroleum occupies a significant proportion in total energy consumption. Gradual exhaustion of traditional fossil energy and increasingly prominent environmental problems promote energy structure transformation and carbon emission reduction to become an important part in future industrial development strategies. The industrial energy system is designed to reduce the cost and the environmental pollution, promote the transition to the low-carbon industry, improve the overall efficiency of energy and tightly develop around the system economy and the environment. Renewable energy supply is an effective technology for reducing greenhouse gas emission, and energy sources such as solar energy, wind energy and biological energy have been widely researched for the sustainable development of the system.
In the field of energy system scheduling considering uncertainty factors, the following three methods can be generally adopted: opportunity constrained planning, stochastic optimization, and robust optimization. Opportunistic constraint planning must make a decision before the realization of a random variable is predicted, and the decision is considered to be made so that the probability that a constraint condition is established is not less than a certain confidence level, which can flexibly realize the balance between profitability and reliability, but the method is often a non-convex problem with multivariate integration, which leads to the increase of the difficulty of solving and calculating. In random optimization, uncertain parameters are generally described by probability distribution, the expected value of an objective function gradually tends to be optimal during optimization, but in practical application, the scene number gradually increases along with higher data dimension, and the calculation difficulty of random optimization is higher and higher. The robust optimization has the characteristics of high calculation efficiency and the like, and is widely applied to energy system optimization under an uncertain environment. The establishment of the uncertain set has a key influence on the effect of robust optimization, and the traditional robust optimization method adopts a box-shaped uncertain set, and adopts a thinking based on machine learning to obtain a more compact uncertain set, such as kernel-based support vector clustering and a Dirichlet process mixed model, so as to improve the conservative degree of optimization. The robust optimization can balance the performance and robustness of the solution, but cannot be directly applied to the renewable energy comprehensive system optimization under the uncertain condition. Since few renewable energy sources are available in the worst case, conventional robust optimization methods tend to be overly conservative without considering the penetration of renewable energy sources.
In order to overcome the above defects in the prior art, there is an urgent need in the art for a scheduling method for an energy system, which performs simple, fast and practical scheduling optimization on an industrial system by using a coupling relationship between data, thereby effectively improving the economic benefits of operation of the energy system. In addition, the result of the multi-objective optimization can provide a multi-optimal scheme for a decision maker to guide the sustainable development of the industry.
Disclosure of Invention
The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements of all aspects nor delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.
In order to overcome the defects in the prior art, the invention provides an energy system scheduling method, an energy system scheduling device and a corresponding computer readable storage medium thereof, which can perform simple, quick and practical scheduling optimization on an industrial system by using the coupling relation between data, thereby effectively improving the economic benefits of the operation of the energy system. In addition, the result of the multi-objective optimization can provide a multi-optimal scheme for a decision maker to guide the sustainable development of the industry.
Specifically, the scheduling method of the energy system provided by the first aspect of the invention includes the following steps: developing a deterministic energy system optimization model according to the process mechanism and the operation characteristics of the energy system; acquiring historical data of the energy system; performing cluster analysis on the historical data to map the historical data into clusters; determining an uncertainty set and corresponding probabilities for a plurality of scenes in the deterministic energy system optimization model; determining a stochastic robust optimization model according to the deterministic energy system optimization model, the historical data, the cluster, and the uncertainty set; and determining a scheduling scheme for the energy system via the stochastic robust optimization model.
Further, in some embodiments of the invention, the energy system is selected from one or more of a steam generation system, a steam turbine network, an electrical power system, a cooling water system and a renewable energy infiltration system, a waste heat recovery system and boiler to produce extra-high pressure steam, a steam turbine to meet power requirements of the process machinery, a steam turbine to drive a compressor, a de-superheater to balance the steam network, a cooling tower for water circulation, a heat exchanger, a wind turbine, and a solar collector, wherein the energy to be converted and transported includes one or more of fuel, steam, electricity, water, and renewable energy.
Further, in some embodiments of the invention, the assumed condition of the deterministic energy system optimization model is selected from at least one of: the temperature and the pressure in the steam pipe networks of multiple grades are constant values, the temperature and the pressure of steam inlet and steam extraction of the steam turbine are equal to the temperature and the pressure of the corresponding pipe networks, the selected equipment operates according to the rated power of the selected equipment, steam produced by the waste heat recovery system and consumed water and fuel are constant values, the steam requirement in the production process is constant value, candidate boilers, candidate cooling towers or candidate circulating water pumps of the same type have the same structural parameters, the energy loss in the steam pipe network balance is ignored, and/or given information of the deterministic energy system optimization model is selected from at least one of the following information: parameters in the plant efficiency fitting function, fuel and water consumed in the waste heat recovery system, ultra high pressure steam flow produced in the waste heat recovery system, steam demand at multiple levels of the production process, rated power of water pumps and/or cooling tower motors, local ambient temperature, humidity, wind speed and/or solar radiation, weighting coefficients for multiple forms of energy, and historical data of mass flow, temperature and/or pressure at key sites of the cracked gas compression system and cold box unit energy systems, and/or constraints of the deterministic energy system optimization model are selected from at least one of: a mass constraint for a plurality of cells, an energy balance constraint for each of the cells, a system balance constraint for each of the cells, and a range constraint for a decision variable.
Further, in some embodiments of the present invention, the step of performing cluster analysis on the historical data to map the historical data into clusters includes: preprocessing historical data regarding local wind speed and/or solar radiation; and mapping the historical data regarding local wind speed and/or solar radiation into a set of the clusters through a two-tier unsupervised machine learning framework.
Further, in some embodiments of the present invention, the step of performing cluster analysis on the historical data to map the historical data into clusters further comprises: constructing a cluster set by adopting fuzzy C-means clustering; using a generalized least squares error function:
where N is the number of data, u j Is the center ε of the jth data set, the s-th cluster s Expressed as:
the membership of the data samples is expressed as:
Further, in some embodiments of the invention, the step of determining an uncertainty set and corresponding probabilities for a plurality of scenarios in the deterministic energy system optimization model comprises: according to the description distribution of the historical data, an uncertainty set is constructed by adopting a data driving method based on kernel learning, and vector clustering is supported to describe the data into a closed sphere with the minimum volume:
minR 2
s.t.||ψ(u (i) )-q|| 2 ≤R 2 i =1, \8230, wherein, N is the circle center of the closed sphere, and R is the spherical radius of the closed sphere; through introducing a relaxation variable, re-expression is carried out by utilizing a KKT condition and duality, and a quadratic programming problem is obtained:
where i denotes the index of the support vector, v i As auxiliary variables, Q = Γ 2, Γ representing the covariance matrix of uncertainties; adopting random optimization, regarding the expected value of the calculated uncertain parameters as a multi-scenario optimization problem, representing the randomly optimized scenario by the result of the cluster analysis, and calculating the occurrence probability corresponding to each scenario, wherein the probability of each category s is defined as:
wherein, I (u) j ) = s denotes data set u j Belonging to cluster s.
Further, in some embodiments of the invention, the step of determining a stochastic robust optimization model based on the deterministic energy system optimization model, the historical data, the cluster, and the uncertainty set comprises: integrating the historical data according to the deterministic energy system optimization model, taking two-stage stochastic programming as an external problem for realizing an expected optimal value, and taking robust optimization as an internal problem for hedging the worst case; and determining the random robust optimization model according to the external problem, the internal problem and the uncertain set.
Further, in some embodiments of the invention, the decision variables of the stochastic robust optimization model are selected from one or more of boiler, steam turbine, bleed valve, actual load of regenerator, process power consumer drive source, whether boiler is employed, solar collector, thermal energy storage, solar collector area, number of wind turbines.
Further, in some embodiments of the present invention, the scheduling objective of the stochastic robust optimization model is a total annual operational cost, wherein the total annual operational cost at least comprises a new unit annual cost, an operational cost of a plurality of energy types, and/or a carbon tax.
Further, in some embodiments of the invention, the stochastic robust optimization model is expressed as a multi-layer problem:
wherein C is the number of data classes, s is the index of the data classes, and the step of determining the scheduling scheme of the energy system via the stochastic robust optimization model comprises: segmenting the random robust optimization model into a main problem and a plurality of sub-problems; iteratively solving the main problem and the sub-problems under each scene; and carrying out sensitivity analysis on the cluster number of the cluster analysis according to the main problem and the optimal solution obtained by each sub-problem under the condition of meeting the stop criterion so as to determine the scheduling scheme of the energy system.
Further, in some embodiments of the present invention, the step of iteratively solving the main problem and the sub-problems under each scenario includes: relaxing the stochastic robust optimization model to represent the main problem as:
wherein the master question provides a lower bound for a target value; representing the sub-problem solved iteratively as: 22A524 1CNCN
And by applying strong duality, the sub-problem is rewritten as:
wherein, eta, gamma, lambda i And μ i is the Lagrangian multiplier; obtaining an upper bound and a feasible solution of the target value by solving the sub-problem to generate a group of new extreme points; and adding an optimal cut in the constraint condition of the main problem to generate a new lower bound.
Further, according to a second aspect of the present invention, there is provided a scheduling apparatus of an energy system, comprising: a memory; and the processor is connected with the memory and is configured with the dispatching method of the energy system.
Furthermore, according to a third aspect of the present invention, there is provided a computer readable storage medium having stored thereon computer instructions, wherein the computer instructions, when executed by a processor, implement the scheduling method of the energy system described above.
Drawings
The above features and advantages of the present disclosure will be better understood upon reading the detailed description of embodiments thereof in conjunction with the following drawings. In the drawings, components are not necessarily drawn to scale, and components having similar relative characteristics or features may have the same or similar reference numerals.
Fig. 1 illustrates an architecture diagram of a scheduling apparatus of an energy system provided in accordance with some embodiments of the present invention;
fig. 2 illustrates a flow diagram of a method of scheduling an energy system provided according to some embodiments of the invention;
figure 3 illustrates an architectural diagram of a scheduling scheme for an energy system provided according to some embodiments of the invention;
FIG. 4 illustrates the total annual cost of an industrial energy system for different scenarios provided according to some embodiments of the invention;
FIG. 5 illustrates scheduling optimization results at different cluster numbers provided in accordance with some embodiments of the invention;
fig. 6 shows an algorithm data flow diagram of an industrial multi-type energy system data-driven stochastic robust optimization method under an uncertain environment according to some embodiments of the present invention.
Fig. 7 is a schematic diagram illustrating a scheduling scheme of an industrial multi-type energy system data-driven stochastic robust optimization method in an uncertain environment according to some embodiments of the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided for illustrative purposes, and other advantages and effects of the present invention will become apparent to those skilled in the art from the present disclosure. While the invention will be described in connection with the preferred embodiments, there is no intent to limit its features to those embodiments. On the contrary, the invention is described in connection with the embodiments for the purpose of covering alternatives or modifications that may be extended based on the claims of the present invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. The invention may be practiced without these particulars. Moreover, some of the specific details have been omitted from the description in order not to obscure or obscure the focus of the present invention.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Additionally, the terms "upper," "lower," "left," "right," "top," "bottom," "horizontal," "vertical" and the like as used in the following description are to be understood as referring to the segment and the associated drawings in the illustrated orientation. The relative terms are used for convenience of description only and do not imply that the described apparatus should be constructed or operated in a particular orientation and therefore should not be construed as limiting the invention.
It will be understood that, although the terms "first", "second", "third", etc. may be used herein to describe various elements, regions, layers and/or sections, these elements, regions, layers and/or sections should not be limited by these terms, but rather should be used to distinguish one element, region, layer and/or section from another. Thus, a first component, region, layer or section discussed below could be termed a second component, region, layer or section without departing from some embodiments of the present invention.
As described above, in the field of energy system scheduling considering uncertainty factors, the following three methods can be generally adopted: opportunity constrained planning, stochastic optimization, and robust optimization. The opportunistic constraint planning needs to make a decision before the implementation of a random variable is predicted, and needs to consider that the probability that the decision is made so that the constraint condition is satisfied is not less than a certain confidence level, so that the balance between profitability and reliability can be flexibly realized, but the method is a non-convex problem with multivariate integration, which causes the difficulty in solving and calculating. In random optimization, uncertain parameters are generally described by probability distribution, the optimization gradually tends to the optimal expected value of an objective function, but in practical application, the calculation difficulty of random optimization is increased along with the fact that the data dimension is higher, the scene number is gradually increased, and the calculation difficulty of random optimization is higher and higher. The robust optimization has the characteristics of high calculation efficiency and the like, and is widely applied to energy system optimization under an uncertain environment. The establishment of the uncertain set has a key influence on the effect of robust optimization, and the traditional robust optimization method adopts a box-shaped uncertain set, and adopts a thinking based on machine learning to obtain a more compact uncertain set, such as kernel-based support vector clustering and a Dirichlet process mixed model, so as to improve the conservative degree of optimization. Although the robust optimization can balance the performance and robustness of the solution, the robust optimization cannot be directly applied to the renewable energy comprehensive system optimization under the uncertain condition. Since few renewable energy sources are available in the worst case, conventional robust optimization methods tend to be overly conservative without considering the penetration of renewable energy sources.
In order to overcome the defects in the prior art, the invention provides an energy system scheduling method, an energy system scheduling device and a corresponding computer readable storage medium thereof, which can perform simple, quick and practical scheduling optimization on an industrial system by using the coupling relation between data, thereby effectively improving the economic benefits of the operation of the energy system. In addition, the result of the multi-objective optimization can provide a multi-optimal scheme for a decision maker to guide the sustainable development of the industry.
In some non-limiting embodiments, the scheduling method of the energy system provided by the first aspect of the present invention may be implemented via the scheduling apparatus of the energy system provided by the second aspect of the present invention. Specifically, the dispatching device of the energy system is provided with a memory and a processor. The memory includes, but is not limited to, the above-described computer-readable storage medium provided by the third aspect of the invention having computer instructions stored thereon. The processor is connected to the memory and configured to execute the computer instructions stored in the memory to implement the scheduling method of the energy system provided by the first aspect of the present invention.
Referring first to fig. 1, fig. 1 is a diagram illustrating an architecture of a dispatching device of an energy system according to some embodiments of the present invention.
Fig. 1 illustrates a dispatching device of an energy system provided according to some embodiments of the invention. The scheduling device of the energy system includes an internal communication bus 301, a processor (processor) 302, a Read Only Memory (ROM) 303, a Random Access Memory (RAM) 304, a communication port 305, and a hard disk 307. The internal communication bus 301 may enable data communication between the dispatcher components of the energy system. Processor 302 may make the determination and issue a prompt. In some embodiments, processor 302 may be comprised of one or more processors. The communication port 305 may enable data transmission and communication between the scheduling apparatus of the energy system and an external input/output device. In some embodiments, the dispatch device of the energy system may send and receive information and data from the network through the communication port 305. In some embodiments, the dispatching device of the energy system may transmit and communicate data between the external input/output device and the input/output terminal 306 in a wired manner. The dispatching means of the energy system also comprise various forms of program storage units and data storage units, such as a hard disk 307, a Read Only Memory (ROM) 303 and a Random Access Memory (RAM) 304, capable of storing various data files for computer processing and/or communication use, as well as possible program instructions for execution by the processor 302. The processor 302 executes these instructions to implement the main parts of the method. The results of the processing by the processor 302 are communicated to an external output device via the communication port 305 for display on a user interface of the output device.
The working principle of the scheduling device of the energy system will be described below with reference to some embodiments of the scheduling method of the energy system. It will be appreciated by those skilled in the art that these examples of the dispatching method are only some non-limiting embodiments provided by the present invention, and are intended to clearly illustrate the main concepts of the present invention and provide some specific solutions convenient for the public to implement, rather than to limit the overall function or overall operation of the dispatching device of the energy system. Similarly, the energy system dispatching device is only a non-limiting embodiment provided by the present invention, and the execution subject of each step in the energy system dispatching method is not limited.
Referring to fig. 2, fig. 2 is a flowchart illustrating a scheduling method of an energy system according to some embodiments of the invention.
As shown in step S1 of fig. 2, in the energy system dispatching process, the energy system dispatching method may first develop a deterministic energy system optimization model according to the process mechanism and the operation characteristics of the energy system. Thereafter, as shown in step S2 of fig. 2, the method may acquire historical data of the energy system. Thereafter, as shown in step S3 of fig. 2, the method may perform cluster analysis on the historical data to map the historical data into a cluster. Thereafter, as shown in step S4 of fig. 2, the method may determine an uncertainty set and corresponding probabilities of a plurality of scenarios in the deterministic energy system optimization model. Thereafter, as shown in step S5 of fig. 2, the method may determine a stochastic robust optimization model according to the deterministic energy system optimization model, the historical data, the cluster, and the uncertainty set. After determining the stochastic robust optimization model, the method may determine a dispatch plan for the energy system via the stochastic robust optimization model.
Further, in some embodiments of the invention, the energy system may be selected from one or more of a steam generation system, a steam turbine network, an electrical power system, a cooling water system and a renewable energy infiltration system, a waste heat recovery system and a boiler for producing extra-high pressure steam, a steam turbine to meet power requirements of a process machine, a steam turbine to drive a compressor, a de-superheater to balance a steam network, a cooling tower for water circulation, a heat exchanger, a wind turbine, and a solar collector, wherein the energy to be converted and transported includes one or more of fuel, steam, electricity, water, and renewable energy.
Further, in some embodiments of the invention, the assumed condition of the deterministic energy system optimization model may be selected from at least one of the following: the temperature and the pressure in the steam pipe networks of multiple grades are constant values, the temperature and the pressure of steam inlet and steam extraction of the steam turbine are equal to the temperature and the pressure of the corresponding pipe networks, selected equipment operates according to the rated power of the selected equipment, steam produced by the waste heat recovery system, consumed water and fuel are constant values, the steam requirement in the production process is constant value, candidate boilers, candidate cooling towers or candidate circulating water pumps of the same type have the same structural parameters, and energy loss in the steam pipe network balance is ignored. The given information of the deterministic energy system optimization model is selected from at least one of: parameters in the equipment efficiency fitting function, fuel and water consumed in the waste heat recovery system, ultrahigh pressure steam flow produced in the waste heat recovery system, steam requirements of multiple levels in the production process, rated power of a water pump and/or a cooling tower motor, local ambient temperature, humidity, wind speed and/or solar radiation, weighting coefficients of various forms of energy, and historical data of mass flow, temperature and/or pressure of key sites of a cracked gas compression system and a cold box unit energy system. The constraint condition of the deterministic energy system optimization model is selected from at least one of the following: a mass constraint for a plurality of cells, an energy balance constraint for each of said cells, a system balance constraint for each of said cells, and a range constraint for a decision variable.
Further, in some embodiments of the present invention, the method may first pre-process the historical data regarding local wind speed and solar radiation during the cluster analysis of the historical data to map the historical data into clusters. The method may then map the historical data regarding local wind speed and/or solar radiation into a set of clusters through a two-tier unsupervised machine learning framework.
In particular, the method may require data of not less than 20 years when collecting solar radiation and wind speed data, preferably including as many operating conditions as possible, such as dry weather, wet weather, rainy weather, and the like. As will be appreciated by those skilled in the art, a set of data is generally referred to herein as data collected at a certain point in time for a selected monitored variable.
Further, in some embodiments of the present invention, the step of performing cluster analysis on the historical data to map the historical data into a cluster further includes:
constructing a cluster set by adopting fuzzy C-means clustering;
using a generalized least squares error function:
where N is the number of data, u j Is the center ε of the jth data set, the s-th cluster s Expressed as:
the membership of the data samples is expressed as:
where p is a blurring parameter indicating the overlap between clusters.
Further, in some embodiments of the present invention, in the determining the uncertainty set and the corresponding probability of the multiple scenes in the deterministic energy system optimization model, the method may adopt a data-driven method based on kernel learning to construct the uncertainty set according to the description distribution of the historical data, and support vector clustering to describe the data as a closed sphere with a minimum volume:
minR 2
s.t.||ψ(u (i) )-q|| 2 ≤R 2 ,i＝1,…,N
wherein q is the center of the closed sphere, and R is the spherical radius of the closed sphere;
through introducing a relaxation variable, a quadratic programming problem is obtained by restating the KKT condition and duality:
wherein, K (u) (i) ,u (j) ) Is a kernel function, kappa is a regularization parameter for adjusting robustness, and kappa is more than or equal to 0;
the above-mentioned indeterminate set is represented as:
where i denotes the index of the support vector, v i As auxiliary variables, Q = Γ 2, Γ representing the covariance matrix of uncertainties;
adopting random optimization, regarding the expected value of the calculated uncertain parameters as a multi-scene optimization problem, representing the randomly optimized scene by the result of the cluster analysis, and calculating the occurrence probability corresponding to each scene, wherein the probability of each class s is defined as:
where I (u) = s indicates that the data set uj belongs to the cluster s.
Further, in some embodiments of the present invention, in the process of determining a stochastic robust optimization model according to the deterministic energy system optimization model, the historical data, the cluster and the uncertainty set, the method may integrate the historical data according to the deterministic energy system optimization model, take two-stage stochastic programming as an external problem for achieving a desired optimal value, and take robust optimization as an internal problem for conflicting worst cases. Then, the method may determine the stochastic robust optimization model according to the external problem, the internal problem, and the uncertainty set.
Further, in some embodiments of the invention, the decision variables of the stochastic robust optimization model may be selected from one or more of boiler, steam turbine, bleed valve, actual load of regenerator, process power user drive source, whether boiler, solar collector, thermal energy storage, solar collector area, number of wind turbines is employed.
Further, in some embodiments of the invention, the scheduling objective of the above-described random robust optimization model may be the total annual operational cost. Here, the annual total operational cost may include at least an annual cost of the additional units, an operational cost of the plurality of energy types, and a carbon tax.
Further, in some embodiments of the present invention, during the scheduling process of the method, the stochastic robust optimization model may be expressed as a multi-layer problem:
where C is the number of data classes, and s is the index of the data classes.
In the process of determining the scheduling scheme of the energy system through the stochastic robust optimization model, the method can firstly segment the stochastic robust optimization model into a main problem and a plurality of sub-problems. After the main problem and the sub-problems are divided, the method can iteratively solve the main problem and the sub-problems under each scene. The method can obtain an optimal solution according to the main problem and each sub-problem under the condition of meeting the stopping criterion. Then, the method can perform sensitivity analysis on the cluster number of the cluster analysis to determine the scheduling scheme of the energy system.
Specifically, in the process of iteratively solving the main problem and determining the sub-problem, the method may relax the stochastic robust optimization model to express the main problem as:
wherein the above-mentioned main question provides a lower bound for the target value;
the above sub-problem to be solved iteratively can be expressed as:
and by applying strong duality, the above sub-problem is rewritten as:
wherein, eta, gamma and lambda i And μ i is the Lagrangian multiplier.
By solving the sub-problem to obtain the upper bound and the feasible solution of the target value, the method can generate a group of new extreme points, and adds an optimal cut in the constraint condition of the main problem to generate a new lower bound.
Further, in some embodiments of the invention, the method may perform sensitivity analysis by clustering the number of clusters. Specifically, when the number of clusters is 1, the random robust optimization is equivalent to the robust optimization, and when the number of clusters increases, the robustness also increases.
In other embodiments of the invention, the method may economically use fuel to meet the heat demand and wind power to meet the power demand.
In some embodiments of the invention, the method may begin by designing a multi-type energy system comprising two boilers, four extraction condensing turbines, four HS-LS back pressure turbines, four MS-LS back pressure turbines, a solar collector panel, and 89 wind turbines. The method may then map the data into clusters using fuzzy C-means clustering (FCM), the number of clusters C =9. Then, the method can establish an uncertain set and the corresponding probability of each type as follows:
then, the method can establish an objective function of a stochastic robust optimization model
After adding the objective function, the method may also add mass and energy balance constraints, system balance constraints, variable range constraints, and other constraints. After adding constraints, the method can also turn the mixed integer nonlinear model into a single-layer optimization problem based on a decomposition algorithm, wherein uncertain variables are wind speed and solar radiation. After the mixed integer nonlinear model is changed into a single-layer optimization problem through a decomposition algorithm, the method can also set the expectation that the objective of the stochastic robust optimization is to minimize the objective under different situations. The goal is dependent on a worst case implementation, and the above method can continuously change the cluster number to observe the optimized target value.
Referring to fig. 3, fig. 3 illustrates a scheduling scheme of an optimal structure and operating conditions of an energy system under an industrial multi-type energy system data-driven stochastic robust optimization method in an uncertain environment.
In some embodiments of the present invention, as shown in fig. 3, the method can provide a scheduling scheme of the optimal structure and operation condition of the energy system under the random robust optimization method driven by the industrial multi-type energy system data in the uncertain environment.
Please refer to fig. 4 and fig. 5 in combination. Fig. 4 illustrates the total annual cost of an industrial energy system under different scenarios provided according to some embodiments of the invention. Fig. 5 illustrates scheduling optimization results at different cluster numbers provided according to some embodiments of the invention.
As shown in fig. 4 and 5, the experimental result of the method for scheduling and optimizing the energy system shows that the method can perform scheduling and optimizing on the industrial system by using the coupling relationship between data. Compared with the traditional robust optimization, the method has higher robustness and lower robust cost, effectively reduces the emission of greenhouse gases, obviously reduces the annual cost, and has more economic and environmental benefits in the optimized scheduling scheme.
Furthermore, as shown in fig. 5, the above method can analyze sensitivity according to the number of clusters, and although robustness increases as the number of clusters increases, more clusters does not mean worse results. The goal of SRO is to minimize the expectation of the target under different scenarios, which depends on the worst case implementation. In the embodiment shown in fig. 5, 9 clustering cases can achieve a relatively low annual total cost while ensuring robustness.
Further, in other embodiments of the present invention, the above method may use the same industrial energy system data and meteorological data as the above embodiments and simultaneously introduce the solar and wind energy independent industrial energy system Deterministic Optimization (DO), the solar and wind energy dependent industrial energy system deterministic optimization (DO-new), the Robust Optimization (RO) and the random robust optimization (SRO) proposed by the present invention to perform the scheduling optimization experiment, and the experimental results are shown in table 1.
Table 1: problem scaling and optimization results for different methods
22A524 1CNCN
As shown in table 1, the above method can introduce renewable energy to effectively reduce the total annual cost while reducing greenhouse gas emissions. The method can reduce the annual total cost by 1507730$/a and reduce the greenhouse gas emission by 26352ton/a. The annual total cost of RO is higher than DO-new due to the consideration of the worst implementation of uncertainty, and the SRO method has increased computational complexity compared to other methods due to the consideration of a variety of scenarios. Although the above method requires more computation time by using the SRO algorithm, both the PoR (2.91%) and the robustness (1.54% higher than RO) are much smaller. As fuel consumption increases, more wind turbines introduce SROs. Under renewable energy uncertainty conditions, process power consumers prefer steam turbine drives.
In addition, in other embodiments provided by the present invention, the method may introduce the results of scheduling optimization experiments performed by Deterministic Optimization (DO) of the industrial energy system without consideration of solar energy and wind energy, deterministic optimization (DO-new) of the industrial energy system with consideration of solar energy and wind energy, robust Optimization (RO) and Stochastic Robust Optimization (SRO) proposed by the present invention, and define the optimal level of robustness cost evaluation robustness sacrifice, where the index is calculated by:
PoR＝(obj UO -obj DO )/obj DO ，
wherein obj DO Target value of DO, obj UO Are optimized target values under uncertainty (RO and SRO). The lower the PoR value, the less optimality sacrifice to ensure robustness, and the easier optimization under uncertain conditions can be selected.
Please refer to fig. 6. Fig. 6 shows an algorithm data flow diagram of an industrial multi-type energy system data-driven stochastic robust optimization method under an uncertain environment according to some embodiments of the present invention.
Furthermore, in some embodiments of the present invention, as shown in fig. 6, the method may first perform data mining on information such as light radiation and wind speed through two-layer data mining. Then, the method may determine a probability corresponding to the data information according to the data information, and convert the probability into an indeterminate set. And then, the method can carry out robust and random optimization according to the uncertain set and the data information. And then, the method can schedule the industrial system according to the random robust optimization model after random optimization. Therefore, the invention can simply, quickly and practically schedule and optimize the industrial system by utilizing the coupling relation among the data, thereby effectively improving the operation economic benefit of the energy system. In addition, the result of the multi-objective optimization can provide a multi-optimal scheme for a decision maker to guide the sustainable development of the industry.
Please refer to fig. 7. Fig. 7 is a schematic diagram illustrating a scheduling scheme of an industrial multi-type energy system data-driven stochastic robust optimization method in an uncertain environment according to some embodiments of the present invention.
In some embodiments of the present invention, as shown in fig. 7, the energy system may include a solar collector, a thermal energy storage system, a wind turbine, and a waste heat recoverer. The energy scheduling method can perform energy scheduling through the embodiment shown in fig. 7, so as to perform simple, quick and practical scheduling optimization on the industrial system by using the coupling relationship between data, thereby effectively improving the operation economic benefit of the energy system.
While, for purposes of simplicity of explanation, the methodologies are shown and described as a series of acts, it is to be understood and appreciated that the methodologies are not limited by the order of acts, as some acts may, in accordance with one or more embodiments, occur in different orders and/or concurrently with other acts from that shown and described herein or not shown and described herein, as would be understood by one skilled in the art.
The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other variations without departing from the spirit or scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (13)
1. A scheduling method of an energy system is characterized by comprising the following steps:
developing a deterministic energy system optimization model according to the process mechanism and the operation characteristics of the energy system;
acquiring historical data of the energy system;
performing cluster analysis on the historical data to map the historical data into clusters;
determining an uncertainty set and corresponding probabilities for a plurality of scenarios in the deterministic energy system optimization model;
determining a stochastic robust optimization model according to the deterministic energy system optimization model, the historical data, the cluster, and the uncertainty set; and
determining a scheduling scheme for the energy system via the stochastic robust optimization model.
2. The scheduling method of claim 1 wherein the energy system is selected from one or more of a steam generation system, a steam turbine network, an electrical power system, a cooling water system and a renewable energy infiltration system, a waste heat recovery system and a boiler to produce extra high pressure steam, a steam turbine to meet power requirements of a process machine, a steam turbine to drive a compressor, a de-superheater to balance a steam network, a cooling tower for water circulation, a heat exchanger, a wind turbine, and a solar collector, wherein the energy to be converted and transported comprises one or more of fuel, steam, electricity, water, and renewable energy.
3. The scheduling method of claim 2 wherein the assumed conditions of the deterministic energy system optimization model are selected from at least one of: the temperature and the pressure in the steam pipe networks of multiple grades are constant values, the temperature and the pressure of steam inlet and steam extraction of the steam turbine are equal to the temperature and the pressure of the corresponding pipe networks, the selected equipment operates according to the rated power of the selected equipment, the steam produced by the waste heat recovery system, the consumed water and the fuel are constant values, the steam demand in the production process is constant value, the candidate boilers, the candidate cooling towers or the candidate circulating water pumps of the same type have the same structural parameters, the energy loss in the balance of the steam pipe networks is ignored, and/or
The given information of the deterministic energy system optimization model is selected from at least one of: parameters in the equipment efficiency fitting function, fuel and water consumed in the waste heat recovery system, ultrahigh pressure steam flow produced in the waste heat recovery system, steam requirements of multiple levels in the production process, rated power of a water pump and/or a cooling tower motor, local ambient temperature, humidity, wind speed and/or solar radiation, and weighted weighting of multiple forms of energy
Coefficients, and historical data of mass flow, temperature and/or pressure at key sites of cracked gas compression system and cold box unit energy system, and/or
Constraints of the deterministic energy system optimization model are selected from at least one of: a mass constraint for a plurality of cells, an energy balance constraint for each of the cells, a system balance constraint for each of the cells, and a range constraint for a decision variable.
4. The scheduling method of claim 3 wherein the step of cluster analyzing the historical data to map the historical data into clusters comprises:
preprocessing historical data regarding local wind speed and/or solar radiation; and
mapping the historical data on local wind speed and/or solar radiation into a set of the clusters through a two-tier unsupervised machine learning framework.
5. The scheduling method of claim 4 wherein the step of cluster analyzing the historical data to map the historical data into clusters further comprises:
constructing a cluster set by adopting fuzzy C-means clustering;
using a generalized least squares error function:
where N is the number of data, u j Is the center ε of the jth data set, the s-th cluster s Expressed as:
the membership of the data samples is expressed as:
where p is a blurring parameter indicating the overlap between clusters.
6. The scheduling method of claim 1 wherein the step of determining an uncertainty set and corresponding probabilities for a plurality of scenarios in the deterministic energy system optimization model comprises:
according to the description distribution of the historical data, an uncertainty set is constructed by adopting a data driving method based on kernel learning, and the data is described as a closed sphere with the minimum volume by adopting support vector clustering:
min R 2
s.t.||ψ(u (i) )-q|| 2 ≤R 2 ,i＝1,…,N
wherein q is the circle center of the closed sphere, and R is the spherical radius of the closed sphere;
through introducing a relaxation variable, a quadratic programming problem is obtained by restating the KKT condition and duality:
s.t.0≤ω i ≤1/Nκ,i＝1,…,N
wherein, K (u) (i) ,u (j) ) Is a kernel function, kappa is a regularization parameter for adjusting robustness, and kappa is more than or equal to 0;
representing the indeterminate set as:
where i denotes the index of the support vector, v i As auxiliary variable, Q = Γ 2 Γ represents the covariance matrix of uncertainty;
adopting random optimization, regarding expected values of multiple scenes as an optimization problem, representing the scenes of random optimization by using the result of the cluster analysis, and calculating the occurrence probability corresponding to each scene, wherein the probability of each class s is defined as:
wherein, I (u) j ) = s denotes data set u j Belonging to cluster s.
7. The scheduling method of claim 1 wherein the step of determining a stochastic robust optimization model based on the deterministic energy system optimization model, the historical data, the cluster, and the uncertainty set comprises:
integrating the historical data according to the deterministic energy system optimization model, taking two-stage stochastic programming as an external problem for realizing an expected optimal value, and taking robust optimization as an internal problem for hedging the worst case; and
and determining the random robust optimization model according to the external problem, the internal problem and the uncertain set.
8. The scheduling method of claim 7 wherein the decision variables of the stochastic robust optimization model are selected from one or more of boiler, steam turbine, bleed valve, actual load of regenerator, process work user drive source, whether boiler, solar collector, thermal energy storage, solar collector area, number of wind turbines are employed.
9. The scheduling method of claim 7, wherein the scheduling objective of the stochastic robust optimization model is a total annual operational cost, wherein the total annual operational cost comprises at least a new unit annual cost, an operational cost of multiple energy types, and/or a carbon tax.
10. The scheduling method of claim 7 wherein the stochastic robust optimization model is expressed as a multi-layer problem:
s.t.Ax≥g
Px+Ry s +Tu≥h
wherein C is the number of data classes, s is the index of the data classes, and the step of determining the scheduling scheme of the energy system via the stochastic robust optimization model comprises:
segmenting the random robust optimization model into a main problem and a plurality of sub-problems;
iteratively solving the main problem and the sub-problems under each scene; and
and carrying out sensitivity analysis on the cluster number of the cluster analysis according to the main problem and the optimal solution obtained by each sub-problem under the condition of meeting the stop criterion so as to determine the scheduling scheme of the energy system.
11. The scheduling method of claim 10 wherein said step of iteratively solving said main problem and sub-problems for each of said scenarios comprises:
relaxing the stochastic robust optimization model to represent the main problem as:
s.t.Ax≥g
Px+Ry s +Tu≥h
wherein the main question provides a lower bound for a target value;
representing the sub-problem solved iteratively as:
Px+Ry s +Tu≥h
and by applying strong duality, rewriting the sub-problem as:
γ＞0,η＞0
wherein, eta, gamma and lambda i And mu i Is a lagrange multiplier;
obtaining an upper bound and a feasible solution of the target value by solving the sub-problem to generate a group of new extreme points; and
an optimal cut is added to the constraints of the main problem to generate a new lower bound.
12. A scheduling apparatus of an energy system, comprising:
a memory; and
a processor connected to the memory and configured to implement the energy system scheduling method of any one of claims 1 to 11.
13. A computer-readable storage medium having stored thereon computer instructions, which, when executed by a processor, implement the method of scheduling of an energy system according to any one of claims 1 to 11.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310044061.7A CN115965156A (en) | 2023-01-29 | 2023-01-29 | Scheduling method and scheduling device of energy system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310044061.7A CN115965156A (en) | 2023-01-29 | 2023-01-29 | Scheduling method and scheduling device of energy system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115965156A true CN115965156A (en) | 2023-04-14 |
Family
ID=87361758
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310044061.7A Pending CN115965156A (en) | 2023-01-29 | 2023-01-29 | Scheduling method and scheduling device of energy system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115965156A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116629029A (en) * | 2023-07-19 | 2023-08-22 | 天津大学 | Data-driven-based flow industry user flexibility assessment method and related equipment |
-
2023
- 2023-01-29 CN CN202310044061.7A patent/CN115965156A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116629029A (en) * | 2023-07-19 | 2023-08-22 | 天津大学 | Data-driven-based flow industry user flexibility assessment method and related equipment |
| CN116629029B (en) * | 2023-07-19 | 2023-09-29 | 天津大学 | Data-driven-based flow industry user flexibility assessment method and related equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Combined multi-objective optimization and robustness analysis framework for building integrated energy system under uncertainty | |
| Li et al. | An improved two-stage robust optimization model for CCHP-P2G microgrid system considering multi-energy operation under wind power outputs uncertainties | |
| Sleiti et al. | Digital twin in energy industry: Proposed robust digital twin for power plant and other complex capital-intensive large engineering systems | |
| Qu et al. | Prediction of electricity generation from a combined cycle power plant based on a stacking ensemble and its hyperparameter optimization with a grid-search method | |
| Zhi et al. | Improved butterfly optimization algorithm for CCHP driven by PEMFC | |
| Ghenai et al. | Recent trends of digital twin technologies in the energy sector: A comprehensive review | |
| CN103258142B (en) | Based on the short-term wind speed forecasting method of wind farm of synoptic model | |
| Musbah et al. | Energy management using multi-criteria decision making and machine learning classification algorithms for intelligent system | |
| Zavala et al. | Proactive energy management for next-generation building systems | |
| CN115965156A (en) | Scheduling method and scheduling device of energy system | |
| Liu et al. | Two-layer multiple scenario optimization framework for integrated energy system based on optimal energy contribution ratio strategy | |
| Abokersh et al. | Sustainability insights on emerging solar district heating technologies to boost the nearly zero energy building concept | |
| Sharif et al. | Generation of whole building renovation scenarios using variational autoencoders | |
| Nasiri et al. | Data analytics and information technologies for smart energy storage systems: A state-of-the-art review | |
| Ledmaoui et al. | Forecasting solar energy production: A comparative study of machine learning algorithms | |
| Fragiacomo et al. | Semi‐empirical development of a novel and versatile multiobjective optimization tool for co/trigeneration energy system design | |
| Qin et al. | Two-stage robust optimal scheduling framework applied for microgrids: Combined energy recovery and forecast | |
| CN117078047B (en) | LSTM-based heat load prediction and distribution optimization method and system | |
| CN113708418A (en) | Micro-grid optimization scheduling method | |
| Garlík | The application of artificial intelligence in the process of optimizing energy consumption in intelligent areas | |
| CN113313410B (en) | Multi-energy coupling modeling evaluation method and device and terminal equipment | |
| Vannoni et al. | Combined Cycle, Heat Pump, and Thermal Storage Integration: Techno-Economic Sensitivity to Market and Climatic Conditions Based on a European and United States Assessment | |
| CN114358601A (en) | Method and device for constructing multi-dimensional evaluation index system of multi-energy system | |
| Zhao et al. | A multi-criteria optimization for a CCHP with the fuel cell as primary mover using modified Harris Hawks optimization | |
| Tang | Analysis of the driving factors of building energy consumption growth in smart cities based on the STIRPAT 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 |