CN116742678A - Comprehensive energy management system of master-slave control architecture and predictive control method - Google Patents

Comprehensive energy management system of master-slave control architecture and predictive control method Download PDF

Info

Publication number
CN116742678A
CN116742678A CN202310951159.0A CN202310951159A CN116742678A CN 116742678 A CN116742678 A CN 116742678A CN 202310951159 A CN202310951159 A CN 202310951159A CN 116742678 A CN116742678 A CN 116742678A
Authority
CN
China
Prior art keywords
power
energy storage
energy management
storage battery
day
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310951159.0A
Other languages
Chinese (zh)
Inventor
王亚雄
余庆港
范依莹
林良光
欧凯
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.)
Fuzhou University
Original Assignee
Fuzhou University
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 Fuzhou University filed Critical Fuzhou University
Priority to CN202310951159.0A priority Critical patent/CN116742678A/en
Publication of CN116742678A publication Critical patent/CN116742678A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/28—Arrangements for balancing of the load in a network by storage of energy
    • H02J3/32—Arrangements for balancing of the load in a network by storage of energy using batteries with converting means
    • 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
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0202—Market predictions or forecasting for commercial activities
    • 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
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
    • H02J13/00001—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by the display of information or by user interaction, e.g. supervisory control and data acquisition systems [SCADA] or graphical user interfaces [GUI]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
    • H02J13/00002—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by monitoring
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
    • H02J13/00006—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment
    • H02J13/00016—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment using a wired telecommunication network or a data transmission bus
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
    • H02J13/00006—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment
    • H02J13/00028—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment involving the use of Internet protocols
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/008—Circuit arrangements for ac mains or ac distribution networks involving trading of energy or energy transmission rights
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/20—The dispersed energy generation being of renewable origin
    • H02J2300/22—The renewable source being solar energy
    • H02J2300/24—The renewable source being solar energy of photovoltaic origin

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Finance (AREA)
  • Economics (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Game Theory and Decision Science (AREA)
  • Data Mining & Analysis (AREA)
  • Human Computer Interaction (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Primary Health Care (AREA)
  • Tourism & Hospitality (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention relates to a comprehensive energy management system of a master-slave control architecture and a predictive control method, wherein the system comprises an energy storage system, an energy storage converter, photovoltaic power generation equipment, a photovoltaic inverter, a load system, an energy management controller, an energy management center and the like; the energy management controller is a master node, the energy storage converter, the photovoltaic inverter and the load system are slave nodes, and the master node monitors the running state of the slave nodes and transmits signals to the slave nodes to adjust the working states of the energy storage system and the photovoltaic power generation equipment; the energy management center receives the information uploaded by the master node and stores the information in a database. According to the method, loads and photovoltaic power generation power in the day and the day are predicted on the basis of a database, a day-ahead energy management model which gives consideration to both the economical efficiency of the system and the durability of the energy storage battery and a day-ahead planned minimum-adjustment day-ahead energy management model are established, and the energy management optimization control is realized by transmitting the energy management model to a slave node through a master node. The system and the method are beneficial to improving the optimization effect of energy management.

Description

Comprehensive energy management system of master-slave control architecture and predictive control method
Technical Field
The invention relates to the technical field of energy management, in particular to a comprehensive energy management system of a master-slave control architecture and a predictive control method
Background
Renewable energy sources are widely applied due to the advantages of green, sustainable development, improvement of excessive exploitation of fossil energy sources, environmental pollution caused by unreasonable use and the like, but the problems of high cost and low energy utilization rate of an electric power system caused by uncertainty, uncontrollable property and the like of power generation are solved. The comprehensive energy system can be coupled with renewable energy sources, distributed power sources and the like, so that a large amount of the distributed power sources and the renewable energy sources can be promoted to be connected into a power grid, meanwhile, the complementary mutual-aid among multiple types of energy sources can improve the reliability of energy source supply, and the energy utilization rate is effectively improved while the system load requirement is met. The energy management optimization method of the comprehensive energy system can reasonably plan the power plan of the power supply according to the characteristics of the distributed power supply of the system so as to improve the economy, durability and the like of the system.
Because the comprehensive energy system comprises a plurality of types of energy sources and energy systems and the power supply characteristics, performance requirements and the like of different types of energy sources are different, for example, a battery energy storage system has higher requirements on service life and the like, so that the complexity of energy management is higher, and more factors need to be considered for the comprehensive energy system to optimally manage energy interaction between each power source and an external power grid. In addition, the conventional energy management is mainly optimized according to a day-ahead prediction result, the influence of various uncertain factors on renewable energy power generation and load demand is considered, the randomness is strong, certain deviation exists in the long precision of the day-ahead prediction time scale, the method cannot be directly applied to system scheduling, and the method is required to be adjusted according to the prediction result of a shorter time scale.
Disclosure of Invention
The invention aims to provide a comprehensive energy management system of a master-slave control architecture and a predictive control method, which are beneficial to improving the optimization effect of energy management.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows: a comprehensive energy management system of a master-slave control architecture comprises an energy storage system, an energy storage converter, photovoltaic power generation equipment, a photovoltaic inverter, a load system, an energy management controller, an energy management center and the like; the energy management controller is a master node, the energy storage converter, the photovoltaic inverter and the load system are slave nodes, and the master node monitors the running state of the slave nodes through an RS485 interface and transmits signals to the slave nodes to adjust the working states of the energy storage system and the photovoltaic power generation equipment; and the energy management center receives the information uploaded by the master node through the Ethernet and stores the information in the SQL database.
Further, the energy management controller establishes communication with the energy storage converter and the photovoltaic inverter through an RS485 industrial control bus and a MODBUS RTU communication protocol, and is used for issuing an optimization result of predictive control of the comprehensive energy system and comprehensively planning the output condition of the power supply; the energy management controller realizes information interaction with the energy management center through the Ethernet, and uploads loads, photovoltaic power generation power and meteorological data sent from the nodes to an SQL server database of the energy management center for a prediction function.
Further, the energy storage converter and the photovoltaic inverter both comprise a DC/AC converter and a controller; the DC/AC converter and the controller convert the generated direct current electric energy into alternating current with the same frequency and same phase as the power grid; the controller receives a control instruction sent by the main node through RS485 communication based on DSP control board development, packages information at intervals and feeds the information back to the main node; the controller of the energy storage converter receives a control signal issued by the main node, controls the charge and discharge mode and the power of the energy storage system in a P/Q mode, executes the energy management optimization result, and simultaneously communicates with the BMS (battery management system) through RS485 communication to acquire the state information of the battery pack so as to realize the protective charge and discharge of the battery; and an MPPT maximum power tracking algorithm is embedded in a controller of the photovoltaic inverter, so that the maximum power output by the photovoltaic power generation equipment is realized, and the controller is connected with the photovoltaic power generation equipment through RS485 communication to acquire information such as photovoltaic power generation power.
Further, the energy management center is connected with the main node through the Ethernet, and the main node packages and uploads the acquired information to the energy management center at intervals; the energy management center reads data uploaded by the main node according to IEC60870-104 standard communication and stores the data in an SQL server database; the energy management center configures an ODBC data source in the Windows operating system, so that prediction control software developed based on MATLAB can be connected with an SQL Server database to realize the functions of prediction and energy management optimization; and determining the device address to be controlled according to the optimized result by the MODBUS protocol point table, and transmitting a signal to the slave node through the master node.
The invention also provides a predictive control method of the comprehensive energy management system based on the master-slave control architecture, which predicts the load and photovoltaic power generation power before and in the day on the basis of a database, then establishes a day-ahead energy management model which gives consideration to the economical efficiency of the system and the durability of the energy storage battery and a day-ahead plan-adjustment minimum day-ahead energy management model, reduces the influence of predictive errors on the optimization effect, ensures the effectiveness of the day-ahead optimization plan, determines the issuing address according to a communication protocol point table and transmits the issuing address to the slave node through the master node, and realizes the energy management optimization control.
Further, the predictive control method includes the steps of:
step S1: based on a load and photovoltaic power generation power database, a machine learning method is adopted to establish a prediction model;
step S2: the method comprises the steps of predicting load and photovoltaic power generation power in the future, and establishing a future energy management model considering both the economical efficiency of the system and the durability of an energy storage battery according to time-of-use electricity price information of a region where the comprehensive energy system is located;
step S3: solving a day-ahead energy management model by adopting an improved gray wolf optimization algorithm, and planning an output plan of 24 hours in the future in one day by taking 1 hour as a time interval;
step S4: and (3) carrying out intra-day prediction based on the data acquired in real time, establishing an intra-day energy management model with minimum adjustment of a pre-day plan, rolling and correcting the pre-day plan at 15min as a time interval, and optimizing the real-time plan of 96 time periods of the future day.
Further, in step S1, detecting abnormal data in load and photovoltaic power generation power history data by adopting a statistic analysis method, processing the abnormal data by adopting a k-nearest neighbor method correction and similar daily data filling method, and normalizing the data processed by adopting a maximum and minimum normalization method; based on the time sequence convolution neural network, the characteristics of complex and nonlinear load and photovoltaic power generation power time sequence data are mined, and a prediction model is established.
Further, in step S2, in the day-ahead energy management model that gives consideration to both the system economy and the durability of the energy storage battery, the objective function of the durability of the energy storage battery is:
wherein N is life,i Represents the service life of the energy storage battery after the ith discharge depth, D od,i Represents the discharge depth of the ith time, which is the ratio of the discharge capacity of the battery to the rated capacity, alpha 1 -α 5 Representing fitting parameters; lambda% (i) represents the life loss rate at the ith depth of discharge, C uti_bat The operation loss cost of the energy storage battery in one scheduling period is represented, n represents the number of cyclic discharge times in one scheduling period, E bat Representing the rated capacity of the energy storage battery, wherein delta is the unit cost of the energy storage battery;
according to the load and photovoltaic power generation power prediction model, day-ahead prediction data of 1h interval and 24h duration are obtained, and the local time-of-use electricity price information is combined, so that an objective function considering both the system economy and the durability of the energy storage battery is established as follows:
C pv_cost (t)＝ξ pv ×P pv (t)
Wherein C is all Representing the total running cost of the system in one scheduling period, P grid (t) represents the interaction power of the system and the large power grid at the moment t, S buy (t) represents the electricity purchasing price of the large power grid at the moment t, S sell (t) represents electricity price selling to a large power grid at time t, C pv_cost (t) represents maintenance cost of photovoltaic power generation at time t, P pv (t) represents the photovoltaic power generation power, ζ, predicted at time t pv Representing a photovoltaic power generation maintenance coefficient;
the constraint conditions of the day-ahead energy management model are as follows:
(1) Power balance constraint
P load (t)＝P pv (t)+P bat (t)+P grid (t)
Wherein P is load (t) represents the load power predicted at t, P bat (t) represents the charge and discharge power of the energy storage battery at t;
(2) Energy storage battery charge-discharge power and power fluctuation constraint
Wherein P is bat_charge_max Represents the maximum charging power, P bat_discharge_max Represents the maximum discharge power, deltaP max Representing an upper limit of power fluctuation;
(3) Energy storage battery state of charge constraints
In SOC inital Indicating the initial state of charge, SOC of the energy storage battery min 、SOC max Respectively represent the minimum charge state and the maximum charge state of the energy storage battery, and SOC final Representing the state of charge of the energy storage battery after a scheduling period; in order to ensure the normal use of the energy storage battery in the next scheduling period, start-stop SOC difference constraint is required to be introduced, and epsilon represents the start-stop SOC difference;
(4) Interaction power constraint with power grid
-P grid_sell_max ＜P grid (t)＜P grid_buy_max
Wherein P is grid_sell_max Representing maximum power sold in interaction with a power grid, P grid_buy_max Representing the maximum purchase power interacting with the grid.
Further, in step S3, an improved gray wolf optimization algorithm is adopted to solve a day-ahead energy management model, and the charging and discharging power of the energy storage battery is optimized, and the specific method is as follows:
initializing parameters of an improved wolf optimization algorithm, and initializing a charging and discharging power sequence of the energy storage battery in a scheduling period according to model parameters and constraint conditions, namely improving the initial wolf group position of the wolf optimization algorithm; in the optimizing process, determining an fitness function of an optimizing algorithm, calculating fitness values according to the initial population and the objective function, and determining 3 energy storage battery power sequences with the lowest daily optimization objective function; updating the new position of the wolf group, namely the charging and discharging power sequence of the energy storage battery, according to the optimal positions of the 3 schemes and other feasible solutions;
in the updating process, a nonlinear convergence factor updating mechanism is adopted, so that global exploration and local searching are balanced better; in the process of updating the energy storage battery capacity planning by the wolf group hunting mechanism, adopting a hunting searching mechanism based on dimension learning, substituting an energy storage battery charging and discharging power sequence updated by the hunting searching mechanism based on dimension learning and a conventional result updated based on 3 optimal schemes into a fitness function, and selecting the minimum optimization objective function before the day under iteration;
The implementation method of the improved gray wolf optimization algorithm comprises the following steps:
the convergence factor adopts a nonlinear attenuation mechanism, and the formula is as follows:
in the method, in the process of the invention,represents the convergence factor, a employs a nonlinear update mechanism inial And a final Respectively represent the initial and final values, t, of the convergence factor max Representing a maximum number of iterations;
the hunting search mechanism based on dimension learning updates the wolf group position by the following method:
R i (t)＝||X i (t)-X i_GWO (t+1)||
wherein X is i (t) represents the position of the ith iteration of the ith wolf in the wolf cluster, namely the charging and discharging power sequence of the ith energy storage battery, X i_GWO Represents the position of the t+1st iteration of the ith wolf obtained by updating the conventional GWO, R i (t) represents X i (t) and X i_GWO Euclidean distance between, X i Neighborhood N of (t) i (t) satisfies the following formula:
N i (t)＝{X j (t)|D i (X i (t),X j (t))≤R i (t),X j (t)∈pop}
wherein N is i (t) represents that the compound satisfies R i (t) X in the radius range i Neighborhood of (t), D i X represents i (t) and X j Euclidean distance of (t), X j (t) is the position of a wolf in the wolf group; determination of X i After the neighborhood of (t), dimension learning is carried out, and the specific formula is as follows:
X i_DLH,d (t+1)＝X i,d (t)+rand*(X n,d (t)-X r,d (t))
wherein D represents one dimension of the problem dimension D, D is the length of the charge and discharge power sequence of the energy storage battery, X n,d (t) represents d-dimensional data of a wolf randomly selected from within the neighborhood, X r,d (t) d-dimensional data of a selected one of the wolves, X i_DLH,d (t+1) represents the result of multi-neighborhood learning;
the wolf group position obtained by hunting search mechanism based on dimension learning, namely the charging and discharging power sequence of the energy storage battery is X i_DLH (t+1), the conventional position updated from the optimal 3 solutions is X i_GWO (t+1) selecting an optimal value by comparing the fitness function sizes of the two;
where f (x) represents the fitness function to be optimized, i.e. the energy management objective function.
Further, the daily energy management model is as follows: in the daily optimization process, a minimum adjustment daily front plan is taken as an optimization target, and a daily energy management model is built according to the load and the photovoltaic power generation power which are predicted by a prediction model and have higher accuracy of 15min intervals and 3h duration, wherein the objective function is as follows:
wherein P is grid_s (k) Representing the power of the kth 15min in the t period of the large power grid, P bat_s (k) Representing the charge and discharge power of the kth 15min in the t period of the energy storage battery; mu represents penalty factor of large grid interaction power change, and (1-mu) represents penalty factor of energy storage battery power changeA seed; controlling the output of an energy storage battery or a large power grid to balance the power difference value generated by the prediction error in the daily optimization process by adjusting mu factor;
constraint conditions to be met by the daily energy management optimization model are as follows:
(1) Power balance constraint
P load_s (k)＝P pv_s (k)+P bat_s (k)+P grid_s (k)
Wherein P is load_s (k) Represents the k 15min base load power, P of ultra-short term prediction pv_s (k) Represents the kth photovoltaic power generation power of 15 min;
(2) The charging and discharging power, the power fluctuation and the SOC constraint of the energy storage battery are the same as before the same day;
(3) The interaction power of the system is the same as that of the power grid optimized before the same day;
the day-ahead plan is corrected by scrolling with 15min as a period, and the specific method comprises the following steps: firstly, solving an intra-day energy management model within a 15-min time interval by adopting an improved gray wolf optimization algorithm, and issuing a first action of an optimization result to a slave node through a master node to realize online optimization and control; then, the next 15min time interval is taken, the actual reserve of the energy storage battery of the system after the last interval optimization is taken as feedback, and then the main node acquires the load information, the photovoltaic power generation power, the meteorological parameters and the actual reserve information of the energy storage battery of the system in real time through RS485 communication and packages and sends the information to an energy management center; the prediction model in the energy management center predicts the load and the photovoltaic power generation power of the next 15min time interval and the duration of 3h according to the information acquired in real time, establishes an intra-day energy management model, and optimizes and issues control information again; the total output force of the power generation unit is gradually approximated to the actual load demand by rolling and optimizing 96 time periods, and the stable operation of the comprehensive energy system is ensured.
Compared with the prior art, the invention has the following beneficial effects:
1. the method comprises the steps of establishing a daily energy management objective function comprising energy storage battery operation loss cost based on discharge depth, power grid acquisition electricity cost and the like, optimizing the system economy, simultaneously considering the energy storage battery operation loss, and improving the service time of the energy storage battery.
2. According to the prediction result with higher daily precision, rolling correction is carried out on the daily schedule, the influence of uncertainty of the prediction information is reduced, and the accuracy and reliability of the comprehensive energy system energy management schedule are improved.
3. A gray wolf optimization algorithm integrating a nonlinear mechanism and a hunting search mechanism based on dimension learning is adopted to solve a day-before-day energy management model, the defect that a conventional algorithm is prone to being partially optimized is overcome, and the energy management optimization effect is improved.
Drawings
FIG. 1 is a schematic diagram of an energy management system according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a predictive control method according to an embodiment of the invention;
FIG. 3 is a flow chart of load and photovoltaic power generation power prediction in an embodiment of the present invention;
FIG. 4 is a diagram of the prediction result error in an embodiment of the present invention;
FIG. 5 is a graph of a relationship between pre-day optimization and intra-day optimization in an embodiment of the application;
FIG. 6 is a graph of the results of day-ahead energy management optimization with only economic considerations in an embodiment of the present application;
FIG. 7 is a graph of day-ahead energy management optimization results for both system economy and energy storage battery durability in an embodiment of the present application;
FIG. 8 is a graph showing the results of optimization of the daily energy management model in an embodiment of the present application.
Detailed Description
The application will be further described with reference to the accompanying drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present application. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
As shown in fig. 1, the embodiment provides an integrated energy management system of a master-slave architecture, which comprises an energy storage system, an energy storage converter, photovoltaic power generation equipment, a photovoltaic inverter, a load system, an energy management controller, an energy management center and the like. The energy management controller is a master node, the energy storage converter, the photovoltaic inverter and the load system are slave nodes, and the master node monitors the running state of the slave nodes through an RS485 interface and transmits signals to the slave nodes to adjust the working states of the energy storage system and the photovoltaic power generation equipment. The energy management center can receive information uploaded by the main node through the Ethernet, store the information in the SQL database and apply the information to a predictive control method to predict loads and photovoltaic power generation power in the day before and in the day, establish a day before energy management model which gives consideration to both system economy and energy storage battery durability and a day before energy management model which is planned to be adjusted least in the day before, reduce the influence of prediction errors and ensure the effectiveness of a day before optimization plan. And determining a transmitting address according to the optimized result by the communication protocol point table, and transmitting the transmitting address to the slave node through the master node.
In this embodiment, the energy management controller establishes communication with the energy storage converter and the photovoltaic inverter through an RS485 industrial control bus and a MODBUS RTU communication protocol, and is configured to issue an optimization result of predictive control of the integrated energy system, and overall plan the output status of the power supply; the energy management controller realizes information interaction with the energy management center through the Ethernet, and uploads loads, photovoltaic power generation power, meteorological data and the like sent by the nodes to an SQL server database of the energy management center for a prediction function.
In this embodiment, the energy storage converter and the photovoltaic inverter each include a DC/AC converter and a controller; the DC/AC converter and the controller convert the generated direct current electric energy into alternating current with the same frequency and same phase as the power grid; the controller receives a control instruction sent by the main node through RS485 communication based on DSP control board development, packages information at intervals and feeds the information back to the main node; the controller of the energy storage converter receives a control signal issued by the main node, controls the charge and discharge mode and the power of the energy storage system in a P/Q mode, executes the energy management optimization result, and simultaneously communicates with the BMS (battery management system) through RS485 communication to acquire the state information of the battery pack so as to realize the protective charge and discharge of the battery; and an MPPT maximum power tracking algorithm is embedded in a controller of the photovoltaic inverter, so that the maximum power output by the photovoltaic power generation equipment is realized, and the controller is connected with the photovoltaic power generation equipment through RS485 communication to acquire information such as photovoltaic power generation power.
In this embodiment, the energy management center is connected to a master node through an ethernet, and the master node packages and uploads the collected information to the energy management center at intervals; the energy management center reads data uploaded by the main node according to IEC60870-104 standard communication and stores the data in an SQL server database; the energy management center configures an ODBC data source in the Windows operating system, so that prediction control software developed based on MATLAB can be connected with an SQL Server database to realize the functions of prediction and energy management optimization; and determining the device address to be controlled according to the optimized result by the MODBUS protocol point table, and transmitting a signal to the slave node through the master node.
As shown in fig. 2, the embodiment also provides a predictive control method of the comprehensive energy management system based on the master-slave control architecture, which predicts the load and photovoltaic power generation power before and during the day on the basis of a database, then establishes a day-ahead energy management model which gives consideration to the economical efficiency of the system and the durability of the energy storage battery and a day-ahead plan adjustment minimum day-ahead energy management model, reduces the influence of the predictive error on the optimization effect, ensures the effectiveness of the day-ahead optimization plan, determines the issuing address according to the communication protocol point table and transmits the issuing address to the slave node through the master node, and realizes the energy management optimization control. The predictive control method specifically comprises the following steps:
step S1: based on a load and photovoltaic power generation power database, a machine learning method is adopted to establish a prediction model;
step S2: the method comprises the steps of predicting load and photovoltaic power generation power in the future, and establishing a future energy management model considering both the economical efficiency of the system and the durability of an energy storage battery according to time-of-use electricity price information of a region where the comprehensive energy system is located;
step S3: solving a day-ahead energy management model by adopting an improved gray wolf optimization algorithm, and planning an output plan of 24 hours in the future in one day by taking 1 hour as a time interval;
step S4: and (3) carrying out intra-day prediction based on the data acquired in real time, establishing an intra-day energy management model with minimum adjustment of a pre-day plan, rolling and correcting the pre-day plan at 15min as a time interval, and optimizing the real-time plan of 96 time periods of the future day.
As shown in fig. 3, a flow chart for predicting load and photovoltaic power generation. Firstly, load and photovoltaic power generation power historical data used for prediction are obtained, abnormal data are detected by adopting a simple statistic analysis method, and the abnormal data are processed by adopting a k neighbor method correction and similar daily data filling method. The repaired data is processed by adopting a maximum and minimum normalization method to improve training speed and prediction accuracy, and the data set is divided into a training set and a testing set.
Aiming at load prediction, taking historical load data at past 24 moments as input, taking load at the next moment as output, and building a prediction model based on a time sequence convolutional neural network through training of training set data; aiming at the photovoltaic power generation power prediction model, the photovoltaic power generation power prediction model based on the time sequence convolution neural network is established by taking weather parameters with high relevance as input and taking photovoltaic power generation power at the moment as output through training of training set data. In the process of training the neural network model, super parameters such as learning rate, dropout and iteration number have high influence on the model prediction accuracy, and parameters need to be adjusted to determine a high parameter combination so as to further improve the model prediction effect.
On the basis of the established load and photovoltaic power generation power prediction model, a day-ahead energy management model of the comprehensive energy system is established according to day-ahead predicted load and photovoltaic power generation power data, time-of-use electricity price information of a region where the comprehensive energy system is located and the like, wherein an objective function gives consideration to system economy and energy storage battery durability. According to the discharge depth of a certain cycle of the energy storage battery, a formula for describing the residual life of the energy storage battery can be established, and the cost loss caused by the discharge depth of each cycle can be quantified by quantifying the influence of the discharge depth of each cycle of the energy storage battery on the service life and combining the cost.
Wherein N is life,i Represents the service life of the energy storage battery after the ith discharge depth, D od,i Represents the discharge depth of the ith time, which is the ratio of the discharge capacity of the battery to the rated capacity, alpha 1 -α 5 Representing the fitting parameters. Lambda% (i) represents the life loss rate at the ith depth of discharge, C uti_bat The operation loss cost of the energy storage battery in one scheduling period is represented, n represents the number of cyclic discharge times in one scheduling period, E bat And the rated capacity of the energy storage battery is represented, and alpha is the unit cost of the energy storage battery.
Further, after the energy storage battery operation loss cost function is constructed, the operation maintenance cost of the photovoltaic power generation equipment and the electricity selling and purchasing cost of the power grid are combined to form the total operation cost of the system, and the formula is as follows:
C pv_cost (t)＝ξ pv ×P pv (t)
Wherein C is all The total operation cost of the system in a scheduling period is represented, the scheduling period is 24h, the control period is 1h, and P grid (t) represents the interaction power of the system and the large power grid at the moment t, S buy (t) represents electricity purchasing price from a large power grid at time t, S sell (t) represents electricity price selling to a large power grid at time t, C pv_cost (t) represents maintenance cost of photovoltaic power generation at time t, P pv (t) represents the photovoltaic power generation power at the moment t, and ζ pv Representing the maintenance coefficient of photovoltaic power generation.
Further, after the daily energy management objective function is established, power balance constraint, energy storage battery charging and discharging power and power fluctuation constraint, energy storage battery state of charge constraint and power constraint interacted with a power grid are introduced, so that a daily energy management model is perfected.
(1) Power balance constraint
P load_base (t)+P ev (t)＝P pv (t)+P bat (t)+P grid (t)
Wherein P is ev (t) represents the charging load of the electric vehicle at t time after the load level is optimized, P load_base And (t) represents the base load power predicted at t.
(2) Energy storage battery charge-discharge power and power fluctuation constraint
Wherein P is bat_charge_max Represents the maximum charging power, P bat_discharge_max Represents the maximum discharge power, deltaP max Representing the upper limit of the power ripple.
(3) Energy storage battery state of charge constraints
In SOC inital Indicating the initial state of charge, SOC of the energy storage battery min 、SOC max Respectively represent the minimum charge state and the maximum charge state of the energy storage battery, and SOC final Representing the state of charge of the energy storage battery after a scheduling period. In order to ensure the normal use of the energy storage battery in the next scheduling period, start-end SOC difference constraint is required to be introduced, epsilon represents the start-end SOC difference, and the value is 0.05.
(4) Interaction power constraint with power grid
-P grid_sell_max ＜P grid (t)＜P grid_buy_max
Wherein P is grid_sell_max Representing maximum power sold in interaction with a power grid, P grid_buy_max Representing the maximum purchase power interacting with the grid.
As shown in fig. 4, the energy storage battery power distribution is optimized for a flowchart of solving an energy management model for an improved gray wolf optimization algorithm. Initializing basic algorithm parameters such as the number of feasible solutions, the dimension of the position of the wolf, namely the number of optimization time periods, the number of iterations and the like for improving the population scale of the wolf optimization algorithm. And initializing a charging and discharging power sequence of the energy storage battery in a scheduling period according to the model parameters and the constraint conditions, wherein the charging and discharging power sequence is an initial population to be optimized as follows.
Wherein z represents the number of feasible solutions, P bat_start,z24 The charging and discharging power of the energy storage battery at the 24 th moment of the z-th population is represented. The fitness function formula for determining the optimization algorithm is:
in the optimizing process, the fitness value is calculated according to the combination of the initial population and the objective function, and the 3 energy storage battery output plans with the lowest daily optimization objective function are determined. The charge-discharge power sequence of the next generation energy storage battery is updated according to the optimal 3 schemes and the feasible solutions of other energy storage batteries.
Further, in the updating process, a nonlinear convergence factor updating mechanism is adopted, global exploration and local searching are balanced better, global optimization effect is improved, and robustness is enhanced. In the process of updating the energy storage battery output plan, a hunting search mechanism based on dimension learning is adopted, the energy storage battery output plan updated by the hunting search mechanism based on dimension learning and a conventional charge and discharge power sequence of the energy storage battery updated based on the optimal 3 energy storage battery output plans are brought into a fitness function, and an optimal solution under the iteration is selected.
And judging whether the maximum iteration times are reached, outputting the wolf-cluster position as an optimal solution if the maximum iteration times are satisfied, and taking the optimal solution as an energy storage battery output plan for 24 hours in the future, and continuing to perform iteration optimization until the iteration termination condition is satisfied if the maximum iteration times are not satisfied.
Then solving a day-ahead plan obtained by a day-ahead energy management model according to an improved gray wolf optimization algorithm, under the constraint conditions of meeting the running state of each device, guaranteeing the day-ahead plan, balancing and the like, combining the load and the photovoltaic power generation power which are predicted by a prediction model and have higher precision and have the period of 15min and the length of 3h, taking the minimum adjustment day-ahead plan as an optimization target in the day-ahead optimization process, and establishing a day-ahead energy management optimization model, wherein the objective function is as follows:
Wherein P is grid_s (k) Representing the power of the kth 15min in the t period of the large power grid, P bat_s (k) And the charge and discharge power of the kth 15min in the t period of the energy storage battery is shown. Alpha represents a penalty factor of large grid interaction power variation, and (1-alpha) represents a penalty factor of energy storage battery power variation. By adjusting the alpha factor, the power difference generated by the prediction error can be balanced by optimizing the process control energy storage battery or the large power grid output in the day.
Constraint conditions to be met by the daily energy management optimization model are as follows:
(1) Power balance constraint
P load_s (k)＝P pv_s (k)+P bat_s (k)+P grid_s (k)
Wherein P is load_s (k) Represents the k 15min base load power, P of ultra-short term prediction pv_s (k) Represents the k-th 15min photovoltaic power generation power.
(2) The charging and discharging power, power fluctuation and SOC constraint of the energy storage battery are the same as those of the prior art.
(3) The same as the grid interaction power optimized before the same day.
The process of solving the day-ahead energy management model by adopting the improved gray wolf optimization algorithm in fig. 4 is adopted, the day-ahead energy management model within the period of 15min is solved, and the first action of the optimization result is issued to the energy storage battery inverter through the energy management controller. And then in the next 15min period, taking the actual reserve of the energy storage battery of the system after the optimization of the previous period as feedback, and the energy management controller acquires the load information, the photovoltaic power generation power, the meteorological parameters and the actual reserve information of the energy storage battery of the system in real time through RS485 communication and packages and sends the information to an energy management center. And predicting the load and the photovoltaic power generation power with the period of the next 15min and the length of 3h by a prediction model in the energy management center according to the information acquired in real time, establishing an intra-day energy management model, and optimizing and issuing control information. Therefore, the total output force of the power generation unit gradually approaches to the actual load demand by rolling and optimizing 96 time periods, and the safe and stable operation of the comprehensive energy system is ensured.
FIG. 5 shows the relationship between the day-ahead energy management optimization and the day-in energy management optimization. The day-ahead energy management optimization stage is characterized in that the optimization period is 24h, and an output plan considering both the system economy and the durability of the energy storage battery is solved under the condition that the constraint conditions in the comprehensive energy system are met by combining local time-of-use electricity price information and considering the technical characteristics of each power supply unit according to the day-ahead load and the photovoltaic power generation power which are iteratively predicted by the established prediction model. The optimization process takes 1h as an interval, and the power of a power supply unit such as an energy storage battery and the like in each optimization interval is regarded as a variable which does not change. However, the prediction accuracy of the future is gradually reduced along with the increase of time, the future optimization result cannot meet the actual comprehensive energy system requirement, and the prediction results of the rolling correction load and the photovoltaic power generation power are required to be performed, so that the intra-day energy management optimization is performed. And the daily energy management optimization performs rolling optimization with 15min as a period when constraint conditions such as power balance, power supply unit output limit and the like are met according to ultra-short-term high-precision prediction information with a time scale of 3h, and rolling correction is performed on output arrangement of an output power supply at the next moment, so that the total output of a power generation unit is gradually approximated to the actual power generation requirement, and safe and stable operation of a micro-grid is ensured.
FIG. 4 is a graph of daily preload prediction errors for time series convolutional neural networks, long and short term memory neural networks, and counter propagating neural networks. The result in the graph shows that the prediction precision of the adopted time sequence convolution-based neural network prediction method is better than that of the other two neural networks, and the prediction result is closer to an actual curve.
FIG. 6 is a graph showing the results of optimization of a day-ahead energy management model that only considers economics. In the early morning low electricity price period, the large power grid meets the load demand, and the energy storage battery can purchase the power grid electric energy storage with low electricity price. In the first peak electricity price period, the load demand is met by the photovoltaic power generation power, and the rest photovoltaic power is stored by the energy storage battery and sold to the power grid to obtain high electricity price benefits. In the afternoon low electricity price period, the photovoltaic power generation power is low, the power grid supplements residual electric energy, and the energy storage battery can continuously purchase the electric energy of the power grid with low electricity price. During peak evening electricity price time, the energy storage battery provides a large amount of electric energy to reduce peak electricity price electric energy purchased from the power grid, and the system economy is improved.
FIG. 7 is a graph showing the results of optimization of a day-ahead energy management model that combines system economy and energy storage battery durability. In the figure, the energy storage battery can be charged in the low-electricity-price period and discharged in the high-electricity-price period, and compared with the optimization result without considering the durability of the energy storage battery, the power fluctuation degree of the energy storage battery is reduced, and the high-power discharging time is shortened. To further compare the advantages of the proposed day-ahead energy management model that compromise system economics and energy storage battery durability, the following table compares the optimized results:
The loss cost of the energy storage battery is reduced by 13507 yuan, the loss is reduced by 69.32%, and the durability is improved by 69.32%. The weight of the system selling electricity purchasing operation cost considering the energy storage operation loss is reduced by 6910 yuan compared with the cost without considering the energy storage, but the total operation cost of the system is 147361 yuan, which is reduced by 4.28 percent compared with 6595 yuan without considering the durability of the energy storage.
Fig. 8 shows a graph of the results of optimization of the daily energy management model, the top graph showing the day-ahead and daily schedule for the energy storage battery, and the bottom graph showing the day-ahead and daily schedule for interacting with the large grid. Compared with the day-ahead plan, the day-ahead plan is adjusted more frequently, but the whole is basically close to the day-ahead plan, and the total running cost of the system is 148829 yuan, which is only increased by 0.098% compared with the day-ahead global optimization, so that the day-ahead effectiveness is ensured.
It will be appreciated by those skilled in the art that 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 flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations 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.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the invention in any way, and any person skilled in the art may make modifications or alterations to the disclosed technical content to the equivalent embodiments. However, any simple modification, equivalent variation and variation of the above embodiments according to the technical substance of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims (10)

1. The comprehensive energy management system of the master-slave control architecture is characterized by comprising an energy storage system, an energy storage converter, photovoltaic power generation equipment, a photovoltaic inverter, a load system, an energy management controller, an energy management center and the like; the energy management controller is a master node, the energy storage converter, the photovoltaic inverter and the load system are slave nodes, and the master node monitors the running state of the slave nodes through an RS485 interface and transmits signals to the slave nodes to adjust the working states of the energy storage system and the photovoltaic power generation equipment; and the energy management center receives the information uploaded by the master node through the Ethernet and stores the information in the SQL database.
2. The master-slave architecture integrated energy management system of claim 1, wherein the energy management controller establishes communication with the energy storage converter and the photovoltaic inverter through an RS485 industrial control bus and a MODBUS RTU communication protocol, and is used for issuing an optimization result of predictive control of the integrated energy system and orchestrating the output condition of the power supply; the energy management controller realizes information interaction with the energy management center through the Ethernet, and uploads loads, photovoltaic power generation power and meteorological data sent from the nodes to an SQL server database of the energy management center for a prediction function.
3. The master-slave architecture integrated energy management system of claim 1, wherein the energy storage converter and the photovoltaic inverter each comprise a DC/AC converter and a controller; the DC/AC converter and the controller convert the generated direct current electric energy into alternating current with the same frequency and same phase as the power grid; the controller receives a control instruction sent by the main node through RS485 communication based on DSP control board development, packages information at intervals and feeds the information back to the main node; the controller of the energy storage converter receives a control signal issued by the main node, controls the charge and discharge mode and the power of the energy storage system in a P/Q mode, executes the energy management optimization result, and simultaneously communicates with the BMS (battery management system) through RS485 communication to acquire the state information of the battery pack so as to realize the protective charge and discharge of the battery; and an MPPT maximum power tracking algorithm is embedded in a controller of the photovoltaic inverter, so that the maximum power output by the photovoltaic power generation equipment is realized, and the controller is connected with the photovoltaic power generation equipment through RS485 communication to acquire information such as photovoltaic power generation power.
4. The integrated energy management system of a master-slave architecture of claim 1, wherein the energy management center is connected to a master node through an ethernet, and the master node packages and uploads the collected information to the energy management center at intervals; the energy management center reads data uploaded by the main node according to IEC60870-104 standard communication and stores the data in an SQL server database; the energy management center configures an ODBC data source in the Windows operating system, so that prediction control software developed based on MATLAB can be connected with an SQL Server database to realize the functions of prediction and energy management optimization; and determining the device address to be controlled according to the optimized result by the MODBUS protocol point table, and transmitting a signal to the slave node through the master node.
5. The predictive control method of the comprehensive energy management system based on the master-slave control architecture according to any one of claims 1 to 4 is characterized in that load and photovoltaic power generation power in the day before and in the day are predicted on the basis of a database, then a day before energy management model which gives consideration to both system economy and energy storage battery durability and a day before plan adjustment minimum day before energy management model are established, the influence of predictive errors on the optimization effect is reduced, the effectiveness of the day before optimization plan is ensured, a issuing address is determined according to a communication protocol point table and transmitted to a slave node through a master node, and energy management optimization control is realized.
6. The predictive control method as recited in claim 5, comprising the steps of:
step S1: based on a load and photovoltaic power generation power database, a machine learning method is adopted to establish a prediction model;
step S2: the method comprises the steps of predicting load and photovoltaic power generation power in the future, and establishing a future energy management model considering both the economical efficiency of the system and the durability of an energy storage battery according to time-of-use electricity price information of a region where the comprehensive energy system is located;
step S3: solving a day-ahead energy management model by adopting an improved gray wolf optimization algorithm, and planning an output plan of 24 hours in the future in one day by taking 1 hour as a time interval;
step S4: and (3) carrying out intra-day prediction based on the data acquired in real time, establishing an intra-day energy management model with minimum adjustment of a pre-day plan, rolling and correcting the pre-day plan at 15min as a time interval, and optimizing the real-time plan of 96 time periods of the future day.
7. The predictive control method according to claim 6, wherein in step S1, abnormal data in load and photovoltaic power history data is detected by a statistic analysis method, the abnormal data is processed by a k-nearest neighbor method correction and similar daily data filling method, and the data processed by the abnormal data is normalized by a maximum and minimum normalization method; based on the time sequence convolution neural network, the characteristics of complex and nonlinear load and photovoltaic power generation power time sequence data are mined, and a prediction model is established.
8. The predictive control method according to claim 6, wherein in step S2, in the day-ahead energy management model that combines both the system economy and the durability of the energy storage battery, the objective function of the durability of the energy storage battery is:
wherein N is life,i Represents the service life of the energy storage battery after the ith discharge depth, D od,i Represents the discharge depth of the ith time, which is the ratio of the discharge capacity of the battery to the rated capacity, alpha 1 -α 5 Representing fitting parameters; lambda% (i) represents the life loss rate at the ith depth of discharge, C uti_bat The operation loss cost of the energy storage battery in one scheduling period is represented, n represents the number of cyclic discharge times in one scheduling period, E bat Representing the rated capacity of the energy storage battery, wherein delta is the unit cost of the energy storage battery;
according to the load and photovoltaic power generation power prediction model, day-ahead prediction data of 1h interval and 24h duration are obtained, and the local time-of-use electricity price information is combined, so that an objective function considering both the system economy and the durability of the energy storage battery is established as follows:
C pv_cost (t)＝ξ pv ×P pv (t)
wherein C is all Representing the total running cost of the system in one scheduling period, P grid (t) represents the interaction power of the system and the large power grid at the moment t, S buy (t) represents the electricity purchasing price of the large power grid at the moment t, S sell (t) represents electricity price selling to a large power grid at time t, C pv_cost (t) represents maintenance cost of photovoltaic power generation at time t, P pv (t) represents the photovoltaic power generation power, ζ, predicted at time t pv Representing a photovoltaic power generation maintenance coefficient;
the constraint conditions of the day-ahead energy management model are as follows:
(1) Power balance constraint
P load (t)＝P pv (t)+P bat (t)+P grid (t)
Wherein P is load (t) represents the load power predicted at t, P bat (t) represents the charge and discharge power of the energy storage battery at t;
(2) Energy storage battery charge-discharge power and power fluctuation constraint
Wherein P is bat_charge_max Represents the maximum charging power, P bat_discharge_max Indicating maximum dischargePower, Δp max Representing an upper limit of power fluctuation;
(3) Energy storage battery state of charge constraints
In SOC inital Indicating the initial state of charge, SOC of the energy storage battery min 、SOC max Respectively represent the minimum charge state and the maximum charge state of the energy storage battery, and SOC final Representing the state of charge of the energy storage battery after a scheduling period; in order to ensure the normal use of the energy storage battery in the next scheduling period, start-stop SOC difference constraint is required to be introduced, and epsilon represents the start-stop SOC difference;
(4) Interaction power constraint with power grid
-P grid_sell_max ＜P grid (t)＜P grid_buy_max
Wherein P is grid_sell_max Representing maximum power sold in interaction with a power grid, P grid_buy_max Representing the maximum purchase power interacting with the grid.
9. The predictive control method according to claim 6, wherein in step S3, an improved gray wolf optimization algorithm is adopted to solve a day-ahead energy management model, and the specific method for optimizing the charge and discharge power of the energy storage battery is as follows:
Initializing parameters of an improved wolf optimization algorithm, and initializing a charging and discharging power sequence of the energy storage battery in a scheduling period according to model parameters and constraint conditions, namely improving the initial wolf group position of the wolf optimization algorithm; in the optimizing process, determining an fitness function of an optimizing algorithm, calculating fitness values according to the initial population and the objective function, and determining 3 energy storage battery power sequences with the lowest daily optimization objective function; updating the new position of the wolf group, namely the charging and discharging power sequence of the energy storage battery, according to the optimal positions of the 3 schemes and other feasible solutions;
in the updating process, a nonlinear convergence factor updating mechanism is adopted, so that global exploration and local searching are balanced better; in the process of updating the energy storage battery capacity planning by the wolf group hunting mechanism, adopting a hunting searching mechanism based on dimension learning, substituting an energy storage battery charging and discharging power sequence updated by the hunting searching mechanism based on dimension learning and a conventional result updated based on 3 optimal schemes into a fitness function, and selecting the minimum optimization objective function before the day under iteration;
the implementation method of the improved gray wolf optimization algorithm comprises the following steps:
the convergence factor adopts a nonlinear attenuation mechanism, and the formula is as follows:
In the method, in the process of the invention,represents the convergence factor, a employs a nonlinear update mechanism inial And a final Respectively represent the initial and final values, t, of the convergence factor max Representing a maximum number of iterations;
the hunting search mechanism based on dimension learning updates the wolf group position by the following method:
R i (t)＝||X i (t)-X i_GWO (t+1)||
wherein X is i (t) represents the position of the ith iteration of the ith wolf in the wolf cluster, namely the charging and discharging power sequence of the ith energy storage battery, X i_GWO Represents the position of the t+1st iteration of the ith wolf obtained by updating the conventional GWO, R i (t) represents X i (t) and X i_GWO Euclidean distance between, X i Neighborhood N of (t) i (t) satisfies the following formula:
N i (t)＝{X j (t)|D i (X i (t),X j (t))≤R i (t),X j (t)∈pop}
wherein N is i (t) represents that the compound satisfies R i (t) X in the radius range i Neighborhood of (t), D i X represents i (t) and X j Euclidean distance of (t)，X j (t) is the position of a wolf in the wolf group; determination of X i After the neighborhood of (t), dimension learning is carried out, and the specific formula is as follows:
X i_DLH,d (t+1)＝X i,d (t)+rand*(X n,d (t)-X r,d (t))
wherein D represents one dimension of the problem dimension D, D is the length of the charge and discharge power sequence of the energy storage battery, X n,d (t) represents d-dimensional data of a wolf randomly selected from within the neighborhood, X r,d (t) d-dimensional data of a selected one of the wolves, X i_DLH,d (t+1) represents the result of multi-neighborhood learning;
the wolf group position obtained by hunting search mechanism based on dimension learning, namely the charging and discharging power sequence of the energy storage battery is X i_DLH (t+1), the conventional position updated from the optimal 3 solutions is X i_GWO (t+1) selecting an optimal value by comparing the fitness function sizes of the two;
where f (x) represents the fitness function to be optimized, i.e. the energy management objective function.
10. The predictive control method as set forth in claim 6, wherein in step S4, the intra-day energy management model is: in the daily optimization process, a minimum adjustment daily front plan is taken as an optimization target, and a daily energy management model is built according to the load and the photovoltaic power generation power which are predicted by a prediction model and have higher accuracy of 15min intervals and 3h duration, wherein the objective function is as follows:
wherein P is grid_s (k) Representing the power of the kth 15min in the t period of the large power grid, P bat_s (k) Representing the charge and discharge power of the kth 15min in the t period of the energy storage battery; mu (mu)A penalty factor representing the change of the interaction power of the large power grid, wherein (1-mu) represents a penalty factor of the change of the power of the energy storage battery; controlling the output of an energy storage battery or a large power grid to balance the power difference value generated by the prediction error in the daily optimization process by adjusting mu factor;
constraint conditions to be met by the daily energy management optimization model are as follows:
(1) Power balance constraint
P load_s (k)＝P pv_s (k)+P bat_s (k)+P grid_s (k)
Wherein P is load_s (k) Represents the k 15min base load power, P of ultra-short term prediction pv_s (k) Represents the kth photovoltaic power generation power of 15 min;
(2) The charging and discharging power, the power fluctuation and the SOC constraint of the energy storage battery are the same as before the same day;
(3) The interaction power of the system is the same as that of the power grid optimized before the same day;
the day-ahead plan is corrected by scrolling with 15min as a period, and the specific method comprises the following steps: firstly, solving an intra-day energy management model within a 15-min time interval by adopting an improved gray wolf optimization algorithm, and issuing a first action of an optimization result to a slave node through a master node to realize online optimization and control; then, the next 15min time interval is taken, the actual reserve of the energy storage battery of the system after the last interval optimization is taken as feedback, and then the main node acquires the load information, the photovoltaic power generation power, the meteorological parameters and the actual reserve information of the energy storage battery of the system in real time through RS485 communication and packages and sends the information to an energy management center; the prediction model in the energy management center predicts the load and the photovoltaic power generation power of the next 15min time interval and the duration of 3h according to the information acquired in real time, establishes an intra-day energy management model, and optimizes and issues control information again; the total output force of the power generation unit is gradually approximated to the actual load demand by rolling and optimizing 96 time periods, and the stable operation of the comprehensive energy system is ensured.
CN202310951159.0A 2023-07-31 2023-07-31 Comprehensive energy management system of master-slave control architecture and predictive control method Pending CN116742678A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310951159.0A CN116742678A (en) 2023-07-31 2023-07-31 Comprehensive energy management system of master-slave control architecture and predictive control method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310951159.0A CN116742678A (en) 2023-07-31 2023-07-31 Comprehensive energy management system of master-slave control architecture and predictive control method

Publications (1)

Publication Number Publication Date
CN116742678A true CN116742678A (en) 2023-09-12

Family

ID=87918853

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310951159.0A Pending CN116742678A (en) 2023-07-31 2023-07-31 Comprehensive energy management system of master-slave control architecture and predictive control method

Country Status (1)

Country Link
CN (1) CN116742678A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117148875A (en) * 2023-10-30 2023-12-01 余姚市宏宇输变电工程有限公司 Photovoltaic panel corner control method and system based on energy storage cooperation and storage medium
CN118133145A (en) * 2024-05-07 2024-06-04 南京理工大学 Data center rack air outlet temperature prediction method based on support vector machine

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117148875A (en) * 2023-10-30 2023-12-01 余姚市宏宇输变电工程有限公司 Photovoltaic panel corner control method and system based on energy storage cooperation and storage medium
CN117148875B (en) * 2023-10-30 2024-01-16 余姚市宏宇输变电工程有限公司 Photovoltaic panel corner control method and system based on energy storage cooperation and storage medium
CN118133145A (en) * 2024-05-07 2024-06-04 南京理工大学 Data center rack air outlet temperature prediction method based on support vector machine

Similar Documents

Publication Publication Date Title
US11502534B2 (en) Electrical energy storage system with battery state-of-charge estimation
CN105846461B (en) Control method and system for large-scale energy storage power station self-adaptive dynamic planning
CN107248751B (en) A kind of energy storage station dispatch control method for realizing distribution network load power peak load shifting
CA2809011C (en) Adaptive energy management system
JP2023129546A (en) System and method for optimal control of energy storage system
CN110854932B (en) Multi-time scale optimization scheduling method and system for AC/DC power distribution network
CN116742678A (en) Comprehensive energy management system of master-slave control architecture and predictive control method
CN108288855B (en) Dynamic strategy simulation and optimization method based on micro-grid
CN103560532A (en) Monitoring system and monitoring method of megawatt battery energy storage power station
CN109066744B (en) Energy storage-containing power distribution network coordinated scheduling method and system
CN112636331B (en) Dynamic economic dispatching distributed optimization method and system for smart power grid
CN112583017A (en) Hybrid micro-grid energy distribution method and system considering energy storage operation constraint
CN105207259A (en) Energy-management-based micro-grid system dispatching method in grid connection state
Kapoor et al. Optimal charge/discharge scheduling of battery storage interconnected with residential PV system
CN105896575B (en) Hundred megawatt energy storage power control method and system based on self-adaptive dynamic programming
CN114897346A (en) Virtual power plant robustness optimization scheduling method considering uncertainty and demand response
CN116307087A (en) Micro-grid system energy storage optimal configuration method considering charging and discharging of electric automobile
CN110649638B (en) Optimization method of energy storage system for compensating wind power prediction error
US20160380444A1 (en) Method for managing the energy production of an energy system and associated management device
CN116131356A (en) Wind-solar storage grid-connected micro-grid optimal scheduling method
Clemmer et al. Design and evaluation of a next generation residential energy management system
Salazar et al. Optimum energy management of islanded nanogrids through nonlinear stochastic dynamic programming
CN114744632B (en) Method, device and storage medium for scheduling low-voltage direct-current interconnected power distribution network with quick charge load
CN117674113A (en) Comprehensive energy system energy management model prediction control method
CN115622065A (en) Multi-time-scale microgrid energy management method

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