CN113516269A - Management method of multi-energy complementary energy hub equipment - Google Patents

Management method of multi-energy complementary energy hub equipment Download PDF

Info

Publication number
CN113516269A
CN113516269A CN202010278164.6A CN202010278164A CN113516269A CN 113516269 A CN113516269 A CN 113516269A CN 202010278164 A CN202010278164 A CN 202010278164A CN 113516269 A CN113516269 A CN 113516269A
Authority
CN
China
Prior art keywords
energy
neural network
output
micro
data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010278164.6A
Other languages
Chinese (zh)
Inventor
李俊辉
周海明
韩笑
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
China Electric Power Research Institute Co Ltd CEPRI
Original Assignee
State Grid Corp of China SGCC
China Electric Power Research Institute Co Ltd CEPRI
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, China Electric Power Research Institute Co Ltd CEPRI filed Critical State Grid Corp of China SGCC
Priority to CN202010278164.6A priority Critical patent/CN113516269A/en
Publication of CN113516269A publication Critical patent/CN113516269A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Marketing (AREA)
  • Biophysics (AREA)
  • General Business, Economics & Management (AREA)
  • Biomedical Technology (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Primary Health Care (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a management method of a multi-energy complementary energy hub device, which comprises the following steps: constructing a micro energy network system of the multi-energy complementary energy hub equipment; acquiring energy input data of the micro energy network system of the multi-energy complementary energy hub device: acquiring load output data of the micro energy network system of the multi-energy complementary energy hub equipment; forming a neural network model training data sample of the micro-energy network system with multiple inputs and multiple outputs; constructing a neural network model of the micro energy network system according with the relation between energy input and load output of the micro energy network system environment; and inputting a load demand, predicting the output of the neural network model, and making a related system input regulation strategy to realize more effective management on the multi-energy complementary energy hub equipment.

Description

Management method of multi-energy complementary energy hub equipment
Technical Field
The invention relates to the technical field of energy Internet, in particular to a management method of a multi-energy complementary energy hub device.
Background
In the energy internet, the main goal of the energy internet comprehensive energy hub development is to realize the fusion, coordination and linkage of various energy sources by utilizing the complementarity among various energy sources, so that the comprehensive energy utilization rate and the renewable energy consumption capacity are promoted. The energy hub in the energy internet has various forms and complex scenes, has different forms of energy hub systems such as a cold, heat and electricity combined supply system, a solar energy and methane power supply hybrid system and the like, has different application scenes such as a garden, a smart ecological city, a living community and the like, and is one of key difficult problems of the energy internet multi-energy complementary system in terms of effective carbon emission and energy efficiency comprehensive monitoring and evaluation.
Aiming at the characteristics of various forms, complex scenes, multi-target constraint and the like of an energy internet multi-energy complementary hub system, the existing methods for solving the difficult problems mainly comprise two categories: one class is based on fundamental physical modeling methods and the other is based on data-driven methods. The accuracy of the physical model method mainly depends on the domain knowledge and modeling accuracy of an actual energy hub system, but the comprehensive utilization efficiency and renewable energy consumption of the energy hub are also influenced by factors such as the climate environment of an application area, the use scale, the user demand response and the like. The data-driven method is mainly based on the analysis of the real-time acquired data by the system in the actual operation process, and the parameters of the multi-energy complementary energy hub system are dynamically planned and dynamically adjusted, so that the method has high flexibility and universality, and has high applicability in project practice. With the development of the evolution technology, modern heuristic algorithms such as genetic algorithm, ant colony algorithm, particle swarm optimization algorithm, artificial neural network algorithm and the like arouse the research interest of people and show great application potential.
However, in the process of constructing the multi-energy comprehensive system by the energy internet, a scientific assessment and monitoring method for the systematicness of the multi-energy comprehensive system is lacked, and the method can be used for effectively managing the real-time operation process of the comprehensive energy system, facilitating the optimization of system parameter configuration and promoting the real-time online multi-target optimization of the comprehensive energy system aiming at the characteristics of various forms, complex scenes, multi-target constraints and the like of the energy internet multi-energy complementary system.
Disclosure of Invention
The invention provides a monitoring method of a multi-energy complementary energy hub device from the perspective of a deep neural network by utilizing the characteristics of strong universality and wide scale expansion of the neural network, the deep two-way long-short memory neural network based on the long-short memory neural network designed by the invention is an important branch of the development of a recurrent neural network, the phenomena of gradient explosion and gradient dispersion in an RNN neural network are relieved by skillful design of a memory unit, an input gate, a forgetting gate and an output gate, and the monitoring method has stronger generalization capability and long-time memory capability and can accurately realize multi-energy complementary dynamic comprehensive management evaluation and online dynamic system parameter optimization configuration of an energy internet.
A management method of a multi-energy complementary energy hub device comprises the following steps:
step 1, constructing a micro-energy network system of a multi-energy complementary energy hub device;
step 2, collecting energy input data of the micro energy network system of the multi-energy complementary energy hub device:
step 3, collecting load output data of the micro energy network system of the multi-energy complementary energy hub device;
step 4, preprocessing the energy input data and the load output data of the micro-energy network system;
step 5, forming a multi-input multi-output neural network model training data sample of the micro energy network system according to the energy input data and the load output data;
step 6, determining various energy input distribution coefficients of the micro energy network system, and constructing a micro energy network system neural network model according with the energy input and load output relation of the micro energy network system environment by combining the neural network model training data sample;
step 7, predicting the output of the neural network model, namely the distribution coefficient of various energy inputs by utilizing the built neural network model of the micro energy network system through inputting load requirements; the energy input data is used as the input of the neural network model, and the output of the neural network model is the distribution coefficient of the load demand corresponding to various energy inputs;
and 8, making a related system input regulation and control strategy according to the distribution coefficients of the various energy inputs, and realizing the management of the multi-energy complementary energy hub equipment.
The invention aims to solve the problem that a scientific evaluation and monitoring method for the systematicness of a multi-energy comprehensive system is lacked in the process of constructing the multi-energy comprehensive system by an energy internet, and aims at the characteristics of various forms, complex scenes, multi-target constraint and the like of the multi-energy complementary system of the energy internet, a data driving method is utilized to expand the memory characteristic of an LSTM neural network (long-short memory neural network), a bidirectional long-short memory neural network is realized, memory cells in the LSTM are utilized, the information of the past and the future is kept, the prediction capability of the neural network in the system monitoring and evaluation process is enhanced, a deep framework of the bidirectional long-short memory neural network is realized on the basis of the bidirectional long-short memory neural network, and the standardized distribution of the parameters of the comprehensive energy system is realized through a multilayer framework, so that the parameters are better utilized, the number of memories is reduced, and the nonlinear processing of the input data and the parameters of the energy comprehensive system is increased, therefore, the comprehensive energy system can be managed more effectively in the real-time operation process, the system parameter configuration can be optimized conveniently, and the real-time online multi-objective optimization of the comprehensive energy system is promoted.
Drawings
Embodiments of the invention are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a deep bidirectional long-short memory neural network model according to the present invention.
Detailed Description
In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced without these specific details. The following detailed description includes references to the accompanying drawings, which form a part of the detailed description. The drawings show illustrations in accordance with exemplary embodiments. These exemplary embodiments, which are also referred to herein as "examples," are described in sufficient detail to enable those skilled in the art to practice the present subject matter. The embodiments may be combined, other embodiments may be utilized, or structural and logical changes may be made without departing from the scope of the claims. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope is defined by the appended claims and their equivalents.
As shown in fig. 1, the present invention provides a management method for a multi-energy complementary energy hub device, including:
step 1, constructing a micro-energy network system of a multi-energy complementary energy hub device;
step 2, collecting energy input data of the micro energy network system of the multi-energy complementary energy hub device:
step 3, collecting load output data of the micro energy network system of the multi-energy complementary energy hub device;
step 4, preprocessing the energy input data and the load output data of the micro-energy network system;
step 5, forming a multi-input multi-output neural network model training data sample of the micro energy network system according to the energy input data and the load output data;
step 6, determining various energy input distribution coefficients of the micro energy network system, and constructing a micro energy network system neural network model according with the energy input and load output relation of the micro energy network system environment by combining the neural network model training data sample;
step 7, predicting the output of the neural network model, namely the distribution coefficient of various energy inputs by utilizing the built neural network model of the micro energy network system through inputting load requirements; the energy input data is used as the input of the neural network model, and the output of the neural network model is the distribution coefficient of the load demand corresponding to various energy inputs;
and 8, making a related system input regulation and control strategy according to the distribution coefficients of the various energy inputs, and realizing the management of the multi-energy complementary energy hub equipment.
Preferably, the micro-energy network system of the multi-energy complementary energy hub device is a solar-biogas micro-energy network formed by combining solar heat collection, photovoltaic power generation, a cogeneration unit and a biogas digester.
Preferably, in the step 2, energy input data of the micro energy network system of the multi-energy complementary energy hub device is collected, where the energy input data includes load demand variables, environmental parameters, and device node data of the micro energy network system; the environmental parameters comprise illumination conditions, environment average temperature, photovoltaic panel power temperature and unit volume methane concentration; the equipment node data is acquired by a sensor and comprises system energy input quantity, system configuration parameter quantity, energy carbon emission quantity and energy transfer efficiency.
Preferably, in step 4, the preprocessing of the energy input data and the load output data of the micro-energy network system specifically includes: and prejudging the rationality of the data according to a threshold value, if the data is abnormal, rejecting the data, and otherwise, keeping the data.
Preferably, in step 6, determining various energy input distribution coefficients of the micro energy network system, and constructing a micro energy network system deep bidirectional long and short memory neural network model according with the energy input and load output relationship of the micro energy network system environment by combining the neural network training data samples, specifically including:
step 6-1, constructing a deep bidirectional long and short memory neural network model:
stacking a plurality of layers of bidirectional long and short memory neural networks to form a deep bidirectional long and short memory neural network model, and obtaining a basic formula of the deep bidirectional long and short memory neural network:
Figure BDA0002445523930000041
Figure BDA0002445523930000042
Figure BDA0002445523930000043
Figure BDA0002445523930000044
Figure BDA0002445523930000045
Figure BDA0002445523930000046
wherein the content of the first and second substances,
Figure BDA0002445523930000047
g，
Figure BDA0002445523930000048
respectively representing the output values of an input gate, a forgetting gate, a current input unit state transition and an output gate in the first layer of the neural network at the moment t;
Figure BDA0002445523930000049
respectively representing the weight matrixes of an input gate, a forgetting gate, a current input unit state transition gate and an output gate in the first layer of the neural network;
Figure BDA00024455239300000410
respectively representing bias items of an input gate, a forgetting gate, a current input unit state and an output gate in the first layer of the neural network;
Figure BDA00024455239300000411
respectively representing the states of a hidden layer of a current layer and a hidden layer of a previous layer in the first layer of the neural network;
Figure BDA00024455239300000412
respectively representing the states of a cell pre-layer and a current layer in the l layer of the neural network;
in the deep bidirectional long and short memory neural network, the first layer takes characteristic data as input, and the input of each other layer is the output of the previous layer;
the characteristic data comprises power grid output power, solar energy output power and methane output power;
where, σ is the activation function,
Figure BDA0002445523930000051
characteristic data input at the time t for the first layer of the neural network;
step 6-2, constructing a forward propagation multilayer long and short memory neural network:
Figure BDA0002445523930000052
Figure BDA0002445523930000053
Figure BDA0002445523930000054
Figure BDA0002445523930000055
Figure BDA0002445523930000056
Figure BDA0002445523930000057
6-3, constructing a backward propagation multilayer long and short memory neural network:
Figure BDA0002445523930000058
Figure BDA0002445523930000059
Figure BDA00024455239300000510
Figure BDA00024455239300000511
Figure BDA00024455239300000512
Figure BDA00024455239300000513
wherein, the arrow → represents the forward propagation of the multi-layer long and short memory neural network to obtain the output value
Figure BDA00024455239300000514
Arrow ← representing backward propagation of multilayer long-and-short memory neural networks to obtain output value
Figure BDA00024455239300000515
6-4, transmitting the output values of the multi-layer long and short memory neural network forwards and backwards
Figure BDA0002445523930000061
And
Figure BDA0002445523930000062
and (3) combining, constructing a final deep bidirectional long and short memory neural network, namely a micro-energy network system neural network, and obtaining an output result:
Figure BDA0002445523930000063
wherein the deep bidirectional long and short memory neural network outputs stIs the distribution coefficient of various energy inputs corresponding to the load demand. The deep bidirectional long and short memory neural network model output stThe load demand can be obtained by training the neural network through characteristic data input and load demand historical data
Figure BDA0002445523930000064
The weight value is the distribution coefficient of various energy power inputs corresponding to the W weight value.
The power supply in the experimental design distribution network of the embodiment of the invention mainly comprises a Diesel Engine (DE), a wind power generator (WT), a photovoltaic power generator (PV) and a power Battery (BA), and relevant parameters are shown in the following table:
TABLE 1
Figure BDA0002445523930000065
The method comprises the steps of calculating the generated energy and the load demand of the distributed power supply in each time period by utilizing a photovoltaic physical model and a wind power physical model, and simulating the hourly power output and user load data of a diesel engine, a wind driven generator, a photovoltaic generator and a fuel cell in one year by software.
The target function for the example design:
cp is the minimum cost for treating the pollutant environment; cy is the minimum cost for system operation:
Figure BDA0002445523930000071
Figure BDA0002445523930000072
wherein:
Figure BDA0002445523930000073
the processing cost (dollar) for SO 2;
Figure BDA0002445523930000074
is the cost (dollar) of CO2 disposal;
Figure BDA0002445523930000075
cost (dollar) for NOx disposal; cDGiA fixed investment cost for the ith distributed power supply; wDGiThe maintenance cost (dollar) is overhauled for the ith distributed power supply.
MinZ＝a1Cp+a2Cy
Wherein: a1 is more than or equal to 0 and less than or equal to 1, a2 is more than or equal to 0 and less than or equal to 1, and a1+ a2 is equal to 1. The objective function comprehensively considers the operation cost and the environmental benefit and realizes multi-objective optimization.
The constraint of equation:
Figure BDA0002445523930000076
Ploadttotal load (kW) in the system at time t; pDGitThe distributed power supply outputs power (kW) for the time t.
By using historical data of software simulation, load requirements at the time t are input through a neural network designed in the text, the output power of a diesel engine, a wind driven generator, photovoltaic power generation and a fuel cell at the time t is predicted, and the operating cost and the pollutant treatment cost of the whole energy network system at the time t are calculated according to the parameters in the table 1.
Embodiment policy design was performed as follows reference example:
control mode 1: and the power battery is supported by a power grid, and wind power generation, photovoltaic power generation and diesel engine power generation are circularly charged.
The DC/AC inverter constitutes the grid and is the main support for the energy system grid. The genset is DC-coupled (via an AC/DC inverter) that serves as a backup unit when the power battery soc (state of charge) is low. The generator set operates in a full-load cyclic charging mode after being put into operation, and simultaneously supplies power to the load and charges the battery. When the battery reaches the cycle charge SOC set point, the generator set is shut down. In this control mode, the daily generation of wind energy WT, solar energy PV is typically complete charging the battery.
Control mode 2: hybrid master control, generator set cyclic charging (ac coupling):
both the battery converter and the diesel generator set may form a grid, although interchangeable, i.e. a single changing master control structure, where the designated master controller changes depending on the operating conditions. The control mode is a battery-dominated control mode, in which the genset is used as a backup unit and is operated only when the battery state of charge is low or the load exceeds the battery converter rating. As with mode 1, the genset is operated in a cyclic charge mode.
Control mode 3: hybrid Master control, Generator set load following (AC coupling)
Although interchangeable, both the power cell converter and the generator set may form an electrical grid. This control mode is typically battery dominated, where the genset acts as a backup unit and operates only when the battery state of charge is low or the load exceeds the battery converter rating. The genset is configured in a load following mode, never charging the battery.
Control mode 4: generator set power grid support and battery slope control
The diesel-electric generator set forms the electric grid and is the main support (together with the solar PV). The cells are used only for ramp control to account for changes in solar PV output so that during a fixed daily operating schedule (e.g., 10 am to 3 pm), the solar PV system can provide a firm preset output.
Different control modes are selected according to the output power prediction requirements of various units and the constraint requirements of the running cost and the environmental benefit.
In addition, the method is aimed at a deep bidirectional long and short memory neural network model, Google Tensorflow is used for building, a dropout (random deactivation) value is set to be 0.5, batch _ size of a training set is set to be 32, batch _ size of a testing set is set to be 64, dimensionality of a hidden state in the front-back direction in Bi-LSTM is set to be 100, training learning rate is set to be 0.001, an Adaptation moment estimator (adaptive moment estimator) am optimizer is adopted, and all training data are disturbed before training of each round.
CNN, LSTM and Deep Bi-LSTM neural networks are simulated and compared through the embodiment, model evaluation is carried out by adopting recall rate, accuracy and F1 value, and the experimental results are compared as follows:
TABLE 2
Figure BDA0002445523930000081
The experimental comparison result shows that the Deep bidirectional long and short memory neural network model based on Deep Bi-LSTM is better than the CNN and LSTM methods as a whole, because the Deep bidirectional long and short memory neural network model can acquire the characteristics from the sequence to the front and the back, more comprehensive sequence knowledge information can be acquired, and meanwhile, the Deep bidirectional long and short memory neural network model is constructed by combining with the Deep neural network, so that the model can acquire the sequence characteristics better.
The invention solves the problem that a scientific evaluation and monitoring method for the systematicness of a multi-energy comprehensive system is lacked in the process of constructing the multi-energy comprehensive system by an energy internet, and aims at the characteristics of various forms, complex scenes, multi-target constraint and the like of the multi-energy complementary system of the energy internet, a data driving method is utilized to expand the memory characteristic of an LSTM neural network (long-short memory neural network), a bidirectional long-short memory neural network is realized, memory cells in the LSTM are utilized, the information of 'past' and 'future' is kept, the prediction capability of the neural network in the process of system monitoring and evaluation is enhanced, a deep framework of the bidirectional long-short memory neural network is realized on the basis of the bidirectional long-short memory neural network, and the standardized distribution of the parameters of the comprehensive energy system is realized through a multilayer framework, so that the parameters are better utilized, the number of memories is reduced, and the nonlinear processing of the input data and the parameters of the energy comprehensive system is increased, therefore, the comprehensive energy system can be monitored more effectively in the real-time operation process, the system parameter configuration can be optimized conveniently, and the real-time online multi-target optimization of the comprehensive energy system is promoted.
While embodiments have been described with reference to specific exemplary embodiments thereof, it will be evident that various modifications and changes may be made to these embodiments without departing from the broader spirit and scope of the inventive subject matter. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.

Claims (6)

1. A management method for a multi-energy complementary energy hub device is characterized by comprising the following steps:
step 1, constructing a micro-energy network system of a multi-energy complementary energy hub device;
step 2, collecting energy input data of the micro energy network system of the multi-energy complementary energy hub device:
step 3, collecting load output data of the micro energy network system of the multi-energy complementary energy hub device;
step 4, preprocessing the energy input data and the load output data of the micro-energy network system;
step 5, forming a multi-input multi-output neural network model training data sample of the micro energy network system according to the energy input data and the load output data;
step 6, determining various energy input distribution coefficients of the micro energy network system, and constructing a micro energy network system neural network model according with the energy input and load output relation of the micro energy network system environment by combining the neural network model training data sample;
step 7, predicting the output of the neural network model, namely the distribution coefficient of various energy inputs by utilizing the built neural network model of the micro energy network system through inputting load requirements; the energy input data is used as the input of the neural network model, and the output of the neural network model is the distribution coefficient of the load demand corresponding to various energy inputs;
and 8, making a related system input regulation and control strategy according to the distribution coefficients of the various energy inputs, and realizing the management of the multi-energy complementary energy hub equipment.
2. The method of claim 1, wherein the multi-energy complementary energy hub equipment micro energy network system is a solar-biogas micro energy network combined by solar heat collection, photovoltaic power generation, cogeneration units and biogas digesters.
3. The method according to claim 1, wherein, in the step 2, energy input data of the micro energy network system of the multi-energy complementary energy hub device is collected, and the energy input data comprises load demand variables, environmental parameters and device node data of the micro energy network system; the environmental parameters comprise illumination conditions, environment average temperature, photovoltaic panel power temperature and unit volume methane concentration; the equipment node data is acquired by a sensor and comprises system energy input quantity, system configuration parameter quantity, energy carbon emission quantity and energy transfer efficiency.
4. The method of claim 1, wherein the load output data comprises a power load output, a cold load output, a heat load output, a gas load output of the micro energy source network.
5. The method according to claim 1, wherein the step 4 of preprocessing the energy input data and the load output data of the micro energy network system specifically comprises: and prejudging the rationality of the data according to a threshold value, if the data is abnormal, rejecting the data, and otherwise, keeping the data.
6. The method according to any one of claims 1 to 5, wherein the step 6 of determining the distribution coefficients of the types of energy inputs of the micro energy network system and combining the neural network model training data samples to construct the neural network model of the micro energy network system that conforms to the energy input and load output relationship of the micro energy network system environment specifically comprises:
step 6-1, constructing a basic depth bidirectional long and short memory neural network model:
stacking a plurality of layers of bidirectional long and short memory neural network models to form a deep bidirectional long and short memory neural network model, and obtaining a basic formula of the deep bidirectional long and short memory neural network model:
Figure FDA0002445523920000021
Figure FDA0002445523920000022
Figure FDA0002445523920000023
Figure FDA0002445523920000024
Figure FDA0002445523920000025
Figure FDA0002445523920000026
wherein the content of the first and second substances,
Figure FDA0002445523920000027
ft l，g，
Figure FDA0002445523920000028
respectively representing the output values of an input gate, a forgetting gate, a current input unit state transition and an output gate in the first layer of the neural network at the moment t;
Wi l，
Figure FDA0002445523920000029
respectively representing the weight matrixes of an input gate, a forgetting gate, a current input unit state transition gate and an output gate in the first layer of the neural network;
Figure FDA00024455239200000210
respectively representing bias items of an input gate, a forgetting gate, a current input unit state and an output gate in the first layer of the neural network;
Figure FDA00024455239200000211
respectively representing the states of a hidden layer of a current layer and a hidden layer of a previous layer in the first layer of the neural network;
Figure FDA00024455239200000212
respectively representing the states of a cell pre-layer and a current layer in the l layer of the neural network;
in the deep bidirectional long and short memory neural network, the first layer takes characteristic data as input, and the input of each other layer is the output of the previous layer;
the characteristic data comprises power grid output power, solar energy output power and methane output power;
where, σ is the activation function,
Figure FDA0002445523920000031
characteristic data input at the time t for the first layer of the neural network;
6-2, constructing a forward propagation multilayer long and short memory neural network model:
Figure FDA0002445523920000032
Figure FDA0002445523920000033
Figure FDA0002445523920000034
Figure FDA0002445523920000035
Figure FDA0002445523920000036
Figure FDA0002445523920000037
6-3, constructing a back propagation multi-layer long and short memory neural network model:
Figure FDA0002445523920000038
Figure FDA0002445523920000039
Figure FDA00024455239200000310
Figure FDA00024455239200000311
Figure FDA00024455239200000312
Figure FDA00024455239200000313
wherein, the arrow → represents the forward propagation of the multi-layer long and short memory neural network to obtain the output value
Figure FDA00024455239200000314
Arrow ← representing backward propagation of multilayer long-and-short memory neural networks to obtain output value
Figure FDA00024455239200000315
6-4, transmitting the output values of the forward propagation multilayer long and short memory neural network model and the backward propagation multilayer long and short memory neural network model
Figure FDA0002445523920000041
And
Figure FDA0002445523920000042
and (3) combining, constructing a final deep bidirectional long and short memory neural network model, namely a neural network model of the micro energy network system, and obtaining an output result:
Figure FDA0002445523920000043
wherein the deep bidirectional long-short memory neural network model outputs stIs the distribution coefficient of various energy inputs corresponding to the load demand.
CN202010278164.6A 2020-04-10 2020-04-10 Management method of multi-energy complementary energy hub equipment Pending CN113516269A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010278164.6A CN113516269A (en) 2020-04-10 2020-04-10 Management method of multi-energy complementary energy hub equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010278164.6A CN113516269A (en) 2020-04-10 2020-04-10 Management method of multi-energy complementary energy hub equipment

Publications (1)

Publication Number Publication Date
CN113516269A true CN113516269A (en) 2021-10-19

Family

ID=78060337

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010278164.6A Pending CN113516269A (en) 2020-04-10 2020-04-10 Management method of multi-energy complementary energy hub equipment

Country Status (1)

Country Link
CN (1) CN113516269A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114239292A (en) * 2021-12-21 2022-03-25 特斯联科技集团有限公司 Comprehensive evaluation method and system for low-carbon economic operation-oriented multifunctional demand park

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110244568A (en) * 2019-07-04 2019-09-17 华翔翔能电气股份有限公司 Energy hub model of industrial enterprise microgrid and multi-energy complementary optimization control method thereof

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110244568A (en) * 2019-07-04 2019-09-17 华翔翔能电气股份有限公司 Energy hub model of industrial enterprise microgrid and multi-energy complementary optimization control method thereof

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
刘方泽: "微能源网多能源耦合枢纽的模型搭建与优化", 《电力系统自动化》, vol. 42, no. 14, pages 91 - 98 *
吕天阳;: "一种改进的双能量源电动汽车功率分配策略", 农业装备与车辆工程, no. 03 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114239292A (en) * 2021-12-21 2022-03-25 特斯联科技集团有限公司 Comprehensive evaluation method and system for low-carbon economic operation-oriented multifunctional demand park
CN114239292B (en) * 2021-12-21 2022-05-31 特斯联科技集团有限公司 Comprehensive evaluation method and system for low-carbon economic operation-oriented multifunctional demand park

Similar Documents

Publication Publication Date Title
Sun A multi-objective optimization model for fast electric vehicle charging stations with wind, PV power and energy storage
Suganthi et al. Applications of fuzzy logic in renewable energy systems–a review
CN104065072B (en) A kind of micro-capacitance sensor running optimizatin method based on dynamic electricity price
CN107546781B (en) Micro-capacitance sensor multiple target running optimizatin method based on PSO innovatory algorithm
Elsied et al. An advanced energy management of microgrid system based on genetic algorithm
Qi et al. Low-carbon community adaptive energy management optimization toward smart services
CN112131733B (en) Distributed power supply planning method considering influence of charging load of electric automobile
CN111340274A (en) Virtual power plant participation-based comprehensive energy system optimization method and system
CN104392394B (en) A kind of detection method of micro-capacitance sensor energy storage nargin
Luo et al. Bi-level multi-objective optimization of design and subsidies for standalone hybrid renewable energy systems: A novel approach based on artificial neural network
Li et al. Tri-stage optimal scheduling for an islanded microgrid based on a quantum adaptive sparrow search algorithm
Beheshtikhoo et al. Design of type-2 fuzzy logic controller in a smart home energy management system with a combination of renewable energy and an electric vehicle
CN116014715A (en) Digital twinning-based comprehensive energy control method and system
Sari et al. New optimized configuration for a hybrid PV/diesel/battery system based on coyote optimization algorithm: A case study for Hotan county
Safari et al. Optimal load sharing strategy for a wind/diesel/battery hybrid power system based on imperialist competitive neural network algorithm
Lu et al. Online optimization of energy management strategy for FCV control parameters considering dual power source lifespan decay synergy
CN115241923A (en) Micro-grid multi-objective optimization configuration method based on snake optimization algorithm
Pu et al. A novel GRU-TCN network based Interactive Behavior Learning of multi-energy Microgrid under incomplete information
Xu et al. Optimal prime mover size determination of a CCHP system based on 4E analysis
Pan et al. Dual‐layer optimal dispatching strategy for microgrid energy management systems considering demand response
CN113516269A (en) Management method of multi-energy complementary energy hub equipment
Han et al. An optimal control method of microgrid system with household load considering battery service life
CN108009684A (en) A kind of micro-grid connection state energy management method comprising short-term load forecasting
Liang et al. Real-time optimization of large-scale hydrogen production systems using off-grid renewable energy: Scheduling strategy based on deep reinforcement learning
Yang et al. Evaluating impact of plug-in hybrid electric vehicle charging on power quality

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