A micro energy grid optimization scheduling method and system based on model predictive control

Abstract

The present invention relates to a method and system for optimizing the scheduling of a micro-energy network based on model predictive control. The method comprises: modeling various devices in an eight-in-one micro-energy network to generate mathematical models of each device; on a day-ahead time scale, using the mathematical models of various devices, while satisfying constraints, with the goal of minimizing the total cost within the optimization period, to obtain a day-ahead scheduling plan; on a real-time time scale, based on a model predictive control method, while satisfying constraints, with the goal of following the day-ahead scheduling plan and smooth scheduling, to obtain a real-time scheduling plan; and scheduling various devices in the eight-in-one micro-energy network in real time according to the real-time scheduling plan. In the present invention, by using a model predictive control method, the real-time rescheduling result tracks the day-ahead plan on the one hand, thereby ensuring the economy of real-time scheduling; on the other hand, it greatly reduces the impact of the uncertainty of cold, heat, and electricity loads and renewable energy output on the micro-energy network, and has strong robustness.

Description

Technical Field

[0001] The present invention relates to the technical field of micro energy network optimization scheduling, and in particular to a micro energy network optimization scheduling method and system based on model predictive control. Background Art

[0002] Micro energy network is a micro integrated energy system, which is a natural extension of micro grid in the context of "energy internet". Micro energy network includes the production, transmission, storage, conversion and utilization of different forms of energy such as cold, heat and electricity. It connects the power grid, natural gas pipeline network, heating pipeline network and other energy networks through energy hubs to meet the multi-load demand of cold, heat and electricity of end users. While taking measures according to local conditions and making full use of renewable energy, it realizes multi-energy complementarity and coordinated operation, and ultimately achieves the goal of environmental friendliness and sustainable development.

[0003] Effective optimization scheduling methods determine the energy management quality of micro-energy grids and the overall performance of the system. Common optimization time scales are divided into two scales: day-ahead and real-time. Day-ahead optimization is to predict the renewable energy output and load data of each period of the next day on the previous day, and arrange the power generation plan of each power source in each period of the next day according to certain economic criteria while meeting the load demand of the next day. The intermittent and volatile output of renewable energy has brought challenges to the optimization scheduling of micro-energy. How to deal with uncertainty has become a difficult problem in the day-ahead optimization scheduling of micro-energy grids. The existing processing methods are roughly divided into scenario-based stochastic optimization and robust optimization. The former requires the accurate probability distribution of uncertain factors to give a statistical optimal solution, while the latter requires optimizing the worst case that rarely occurs in practice in a predefined uncertainty set, making the optimization strategy too conservative. In addition, most of the current studies on the day-ahead optimization scheduling methods of micro-energy grids under a single time scale do not consider the prediction error caused by the uncertainty of renewable energy output. The results of decision-making based on day-ahead predictions are suboptimal or even infeasible in the actual operation of the system. Therefore, the energy management of micro-energy grids requires prediction and scheduling results of shorter time scales. The combination of multiple time scales can make the energy management of micro-energy grids more accurate and practical.

[0004] The two-stage optimization scheduling method greatly improves the utilization rate of renewable energy, but this open-loop control method has no feedback mechanism to correct the optimization control process. Model predictive control, as a modern control theory method, has been widely used in engineering practice. The idea of ​​rolling optimization and feedback correction can better solve the uncertainty problem and has strong robustness. However, most of the existing research based on model predictive control methods is aimed at microgrids or micro-energy networks with simple structures. The energy forms involved in micro-energy networks and the various links of energy generation, conversion, transmission, storage and utilization have not been fully considered.

[0005] The strong randomness of cooling, heating and electricity loads and renewable energy generation has put forward higher requirements for uncertainty analysis and optimization of micro-energy grid optimization scheduling. The day-ahead optimization method under a single time scale can no longer meet the requirements of microgrid optimization scheduling at this stage and in the future, mainly manifested in:

[0006] 1) The day-ahead optimal scheduling of micro-energy grids under a single time scale does not take into account the prediction error caused by the uncertainty of renewable energy output. The result of making decisions based on the day-ahead prediction is suboptimal or even infeasible in the actual operation of the system.

[0007] 2) Existing methods for dealing with uncertainty on the day-ahead time scale require accurate probability distribution of uncertainties to give a statistically optimal solution, or they need to optimize the worst-case scenario that rarely occurs in practice within a predefined uncertainty set, which makes the optimization strategy too conservative and has poor practical effects.

[0008] In order to overcome the errors in the scheduling results obtained at a single time scale, a combination of multiple time scales is currently used, the most important of which is the day-ahead-real-time two-stage scheduling method. The two-stage method can reduce the impact of uncertainty to a certain extent, but it still faces the following challenges in the optimization scheduling of micro-energy networks:

[0009] 1) The two-stage optimization scheduling method greatly improves the utilization rate of renewable energy, but this open-loop control method has no feedback mechanism to correct the optimization control process;

[0010] 2) The existing micro-energy grid systems have a single type of internal energy and few types of equipment, which makes it difficult to meet the actual needs of the increasingly complex micro-energy grid. Summary of the invention

[0011] In view of the defects and difficulties existing in the above-mentioned single time scale and two-stage scheduling methods, the present invention proposes a micro-energy network optimization scheduling method based on model predictive control. This method is based on the day-ahead-real-time two-stage optimization method, and continuously performs the process of prediction model-rolling optimization-feedback correction on the real-time time scale to finally obtain a real-time scheduling result, which can meet the actual requirements of the increasingly complex micro-energy network.

[0012] To achieve the above object, the present invention provides the following solutions:

[0013] A micro energy grid optimization scheduling method based on model predictive control, comprising:

[0014] Modeling various devices in the eight-in-one micro-energy network to generate mathematical models of various devices; the eight-in-one micro-energy network is a micro-energy network including typical devices of wind, light, rock, magnetism, fuel, heat, storage and load; the various devices include energy production equipment, energy storage equipment and energy conversion equipment; the energy production equipment includes distributed photovoltaic units, distributed wind turbines, solar collectors, solid oxide fuel cells; the energy storage equipment includes electrical energy storage and bedrock energy storage; the energy conversion equipment includes high-temperature water tanks, electric heat pumps, lithium bromide refrigerators and magnetic suspension refrigerators;

[0015] On a day-ahead time scale, the mathematical models of the various devices are used to obtain a day-ahead dispatch plan of the eight-in-one micro-energy grid with the goal of minimizing the total cost within the optimization period while satisfying the constraints;

[0016] In the real-time time scale, based on the model predictive control method, while satisfying the constraints, the real-time scheduling plan of the eight-in-one micro-energy network is obtained with the goal of following the day-ahead scheduling plan and smooth scheduling;

[0017] Various devices in the eight-in-one micro energy network are scheduled in real time according to the real-time scheduling plan.

[0018] Optionally, the modeling of various devices in the eight-in-one micro energy network to generate mathematical models of various devices specifically includes:

[0019] Modeling the distributed photovoltaic unit to generate a mathematical model of the distributed photovoltaic unit in is the output power of the photovoltaic array at time t; f PV is the derating factor; N is the number of photovoltaic panels; P RPV is the rated power of a single photovoltaic panel; I t is the actual light intensity at time t; I STC is the light intensity under standard test conditions; T STC is the standard test temperature; is the surface temperature of the photovoltaic panel at time t;

[0020] Modeling the distributed wind turbine generator set to generate a mathematical model of the distributed wind turbine generator set in is the fan output power at time t; v t is the wind speed at time t; v in and v out are the cut-in and cut-out wind speeds of the fan respectively; v r is the rated wind speed of the fan; P RWT is the rated output power of the fan;

[0021] Model the solar thermal collector and generate a mathematical model Q of the solar thermal collector. u =A p I(τα) e -A P U L (T p -T a );Q u A is the effective energy obtained by the solar collector per unit time; p and T p are the area and average temperature of the solar collector absorber, respectively; I is the solar irradiance; τ and α represent the effective transmittance and absorptivity, respectively, and e represents dimensionless; U L is the total heat loss coefficient; T a is the ambient temperature;

[0022] Modeling the solid oxide fuel cell to generate a mathematical model of the solid oxide fuel cell and Where P SOFC and H SOFC Respectively represent the electrical power and thermal power output of the solid oxide fuel cell; η h is the reversible thermodynamic efficiency of solid oxide fuel cell; η v and η g are voltage efficiency and fuel utilization, respectively; and F SOFC The lower calorific value of natural gas and the natural gas consumption per unit time respectively;

[0023] Modeling the electric energy storage to generate a mathematical model of the electric energy storage in Indicates the state of charge of the energy storage; σ e is the self-discharge rate of the electrical energy storage; and They are charging power and charging efficiency respectively; and are discharge power and discharge efficiency respectively; E EES is the electrical energy storage capacity;

[0024] Modeling the bedrock energy storage to generate a mathematical model of the bedrock energy storage in represents the charge state of bedrock energy storage; σ h is the self-heat release rate of bedrock energy storage; and are the heating power and heating efficiency respectively; and are heat release power and heat release efficiency respectively; Q BES is the bedrock energy storage capacity;

[0025] Modeling the energy conversion device to generate a mathematical model of the energy conversion device in and are the input power and output power of energy conversion device i at time t, η i is the conversion efficiency of energy conversion device i.

[0026] Optionally, satisfying the constraint condition specifically includes:

[0027] At the same time, system constraints, equipment output power constraints, electric energy storage element constraints, bedrock energy storage element constraints and electric vehicle constraints are met.

[0028] Optionally, the goal of minimizing the total cost within the optimization cycle specifically includes:

[0029] Total cost within the cycle minimum; where C represents the total cost within the period; T represents the period of the day-ahead optimal scheduling; N represents the number of energy conversion equipment; λ i and They represent the maintenance cost per unit output of energy conversion equipment i and the output power in time period t respectively; and denote the charging and discharging and charging and discharging heat costs of electrical energy storage and thermal energy storage in time period t, respectively; represents the power generation cost of the diesel generator in time period t; represents the cost of purchasing and selling electricity from the large power grid in time period t; represents the cost of purchasing natural gas; Represents the charging and discharging cost of electric vehicles.

[0030] Optionally, the goal of following the day-ahead scheduling plan and smooth scheduling specifically includes:

[0031] Satisfy the objective function of real-time optimization scheduling Where t is the current time; T s Optimize the cycle for real-time rolling; N s is the number of dispatchable equipment; Make decisions for real-time dispatchable equipment; is the reference value of the day-ahead scheduling plan; W err and W u is the coefficient matrix; It is the increment of dispatchable equipment output relative to the previous period.

[0032] A micro energy grid optimization dispatching system based on model predictive control, comprising:

[0033] A micro-energy network equipment modeling module is used to model various equipment in the eight-in-one micro-energy network and generate mathematical models of various equipment; the eight-in-one micro-energy network is a micro-energy network including typical equipment of wind, light, rock, magnetism, fuel, heat, storage and load; the various equipment include energy production equipment, energy storage equipment and energy conversion equipment; the energy production equipment includes distributed photovoltaic units, distributed wind turbines, solar collectors, and solid oxide fuel cells; the energy storage equipment includes electrical energy storage and bedrock energy storage; the energy conversion equipment includes high-temperature water tanks, electric heat pumps, lithium bromide refrigerators and magnetic suspension refrigerators;

[0034] A day-ahead optimization module is used to obtain a day-ahead dispatch plan of the eight-in-one micro-energy network by using mathematical models of the various devices under the day-ahead time scale, while satisfying constraints and taking the lowest total cost within the optimization period as the goal;

[0035] A real-time optimization module is used to obtain a real-time scheduling plan of the eight-in-one micro-energy network based on a model predictive control method under a real-time time scale, while satisfying constraints and aiming to follow the day-ahead scheduling plan and smooth scheduling;

[0036] The micro energy network equipment scheduling module is used to perform real-time scheduling of various equipment in the eight-in-one micro energy network according to the real-time scheduling plan.

[0037] Optionally, the micro energy network equipment modeling module specifically includes:

[0038] A distributed photovoltaic unit modeling unit is used to model the distributed photovoltaic unit and generate a mathematical model of the distributed photovoltaic unit. in is the output power of the photovoltaic array at time t; f PV is the derating factor; N is the number of photovoltaic panels; P RPV is the rated power of a single photovoltaic panel; I t is the actual light intensity at time t; I STC is the light intensity under standard test conditions; T STC is the standard test temperature; is the surface temperature of the photovoltaic panel at time t;

[0039] A distributed wind turbine modeling unit is used to model the distributed wind turbine and generate a mathematical model of the distributed wind turbine. in is the fan output power at time t; v t is the wind speed at time t; v inand v out are the cut-in and cut-out wind speeds of the fan respectively; v r is the rated wind speed of the fan; P RWT is the rated output power of the fan;

[0040] A solar thermal collector modeling unit is used to model the solar thermal collector and generate a mathematical model Q of the solar thermal collector. u =A p I(τα) e -A P U L (T p -T a );Q u A is the effective energy obtained by the solar collector per unit time; p and T p are the area and average temperature of the solar collector absorber, respectively; I is the solar irradiance; τ and α represent the effective transmittance and absorptivity, respectively, and e represents dimensionless; U L is the total heat loss coefficient; T a is the ambient temperature;

[0041] A solid oxide fuel cell modeling unit is used to model the solid oxide fuel cell and generate a mathematical model of the solid oxide fuel cell. and Where P SOFC and H SOFC Respectively represent the electrical power and thermal power output of the solid oxide fuel cell; η h is the reversible thermodynamic efficiency of solid oxide fuel cell; η v and η g are voltage efficiency and fuel utilization, respectively; and F SOFC The lower calorific value of natural gas and the natural gas consumption per unit time respectively;

[0042] An electric energy storage modeling unit, used to model the electric energy storage and generate a mathematical model of the electric energy storage in Indicates the state of charge of the energy storage; σ e is the self-discharge rate of the electrical energy storage; and They are charging power and charging efficiency respectively; and are discharge power and discharge efficiency respectively; E EES is the electrical energy storage capacity;

[0043] A bedrock energy storage modeling unit, used to model the bedrock energy storage and generate a mathematical model of the bedrock energy storage in represents the charge state of bedrock energy storage; σ h is the self-heat release rate of bedrock energy storage; and are the heating power and heating efficiency respectively; and are heat release power and heat release efficiency respectively; Q BES is the bedrock energy storage capacity;

[0044] An energy conversion device modeling unit is used to model the energy conversion device and generate a mathematical model of the energy conversion device. in and are the input power and output power of energy conversion device i at time t, η i is the conversion efficiency of energy conversion device i.

[0045] Optionally, the day-ahead optimization module and the real-time optimization module both include:

[0046] The constraint condition limiting unit is used to limit various devices in the eight-in-one micro energy network to simultaneously meet system constraints, device output power constraints, electric energy storage element constraints, bedrock energy storage element constraints and electric vehicle constraints.

[0047] Optionally, the day-ahead optimization module specifically includes:

[0048] Economically optimal unit, used to minimize the total cost within the cycle minimum; where C represents the total cost within the period; T represents the period of the day-ahead optimal scheduling; N represents the number of energy conversion equipment; λ i and They represent the maintenance cost per unit output of energy conversion equipment i and the output power in time period t respectively; and denote the charging and discharging and charging and discharging heat costs of electrical energy storage and thermal energy storage in time period t, respectively; represents the power generation cost of the diesel generator in time period t; represents the cost of purchasing and selling electricity from the large power grid in time period t; represents the cost of purchasing natural gas; Represents the charging and discharging cost of electric vehicles.

[0049] Optionally, the real-time optimization module specifically includes:

[0050] The minimum unit of the objective function is used to make various devices in the eight-in-one micro energy network meet the objective function of real-time optimization scheduling Where t is the current time; T s Optimize the cycle for real-time rolling; N sis the number of dispatchable equipment; Make decisions for real-time dispatchable equipment; is the reference value of the day-ahead scheduling plan; W err and W u is the coefficient matrix; It is the increment of dispatchable equipment output relative to the previous period.

[0051] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0052] The present invention provides a method and system for optimizing the scheduling of a micro-energy network based on model predictive control, the method comprising: modeling various devices in an eight-in-one micro-energy network to generate mathematical models of various devices; on a day-ahead time scale, using the mathematical models of the various devices, while satisfying constraints, with the goal of minimizing the total cost within the optimization period, obtaining the day-ahead scheduling plan of the eight-in-one micro-energy network; on a real-time time scale, based on a model predictive control method, while satisfying constraints, with the goal of following the day-ahead scheduling plan and smooth scheduling, obtaining the real-time scheduling plan of the eight-in-one micro-energy network; and scheduling various devices in the eight-in-one micro-energy network in real time according to the real-time scheduling plan. The present invention uses a model predictive control method, and the real-time rescheduling result tracks the day-ahead plan on the one hand, ensuring the economy of real-time scheduling; on the other hand, it greatly reduces the impact of the uncertainty of cold, hot, and electric loads and renewable energy output on the micro-energy network, and has strong robustness.

[0053] The present invention uses a model predictive control method to perform day-ahead and real-time two-stage optimization scheduling on a micro energy network including wind, light, rock, magnetism, fuel, heat, storage and load. The "eight-in-one" micro energy network system includes energy production equipment (including distributed photovoltaic / wind turbines / diesel generators / solar thermal collection systems / solid oxide fuel cells), energy storage equipment (including electric energy storage / bedrock energy storage), and energy conversion equipment (including high-temperature water tanks / electric heat pumps / lithium bromide refrigerators / magnetic levitation refrigerators). The day-ahead optimization takes the best economy as the goal to obtain a day-ahead scheduling plan. During the real-time optimization process, the cold, heat and electricity loads and renewable energy output within the next hour are re-predicted every 15 minutes, with the goal of minimizing the deviation from the day-ahead plan and optimizing the smoothness of the scheduling result. After three steps of prediction model, rolling optimization and feedback correction, the real-time scheduling result is finally obtained, which can meet the actual requirements of the micro energy network with increasing complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0055] Figure 1 A flowchart of a micro energy grid optimization scheduling method based on model predictive control according to the present invention;

[0056] Figure 2 A schematic diagram of the principle of a micro energy grid optimization scheduling method based on model predictive control according to the present invention;

[0057] Figure 3 A schematic diagram of an eight-in-one micro energy network provided by an embodiment of the present invention;

[0058] Figure 4 A schematic diagram of a real-time optimization process provided by an embodiment of the present invention;

[0059] Figure 5 This is a structural diagram of a micro energy grid optimization scheduling system based on model predictive control according to the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] The purpose of the present invention is to provide a micro-energy network optimization scheduling method based on model predictive control. Based on the two-stage optimization method of day-ahead-real-time, this method continuously performs the process of prediction model-rolling optimization-feedback correction on a real-time time scale, and finally obtains a 24-hour complete real-time scheduling result, which can meet the actual requirements of the increasingly complex micro-energy network.

[0062] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0063] Figure 1 A flowchart of a micro energy grid optimization scheduling method based on model predictive control according to the present invention; Figure 2 FIG. 1 is a schematic diagram of the principle of a micro-energy grid optimization scheduling method based on model predictive control of the present invention. Figure 1 and Figure 2As shown, the present invention provides a micro energy grid optimization scheduling method based on model predictive control, including:

[0064] Step 101: Model various devices in the eight-in-one micro energy network and generate mathematical models of various devices.

[0065] First, a mathematical model is established for each device within the micro energy network. The eight-in-one micro energy network is a micro energy network that includes typical devices of wind, light, rock, magnetism, fuel, heat, storage, and load. The various devices include energy production equipment, energy storage equipment, and energy conversion equipment. The energy production equipment includes distributed photovoltaic units, distributed wind turbines, solar collectors, and solid oxide fuel cells. The energy storage equipment includes electrical energy storage and bedrock energy storage. The energy conversion equipment includes high-temperature water tanks, electric heat pumps, lithium bromide refrigerators, and magnetic suspension refrigerators.

[0066] Figure 3 Schematic diagram of an eight-in-one micro energy network provided by an embodiment of the present invention. Figure 3 The "eight-in-one" micro-energy network of the present invention is a micro-energy network containing typical equipment of "wind, light, rock, magnetism, fuel, heat, storage, and load". The eight types of equipment correspond to distributed wind turbines, distributed photovoltaic units, bedrock energy storage, magnetic levitation refrigerators, solid oxide fuel cells, solar thermal collection systems, electric energy storage, and electric vehicles.

[0067] The energy production equipment involved in the micro-energy network includes distributed photovoltaics, wind turbines, diesel generators, solar thermal collection systems and solid oxide fuel cells; the energy storage equipment includes electrical energy storage and bedrock energy storage; the energy conversion equipment includes medium and high temperature water tanks, electric heat pumps, lithium bromide refrigerators and magnetic levitation refrigerators; the user's load demands include electricity, heat, cooling loads and electric vehicles.

[0068] The specific modeling method of each device in the micro energy network schematic diagram is as follows.

[0069] 1.1. Energy production equipment

[0070] 1.1.1) Distributed photovoltaic units

[0071] The output power of the photovoltaic power generation system is mainly affected by light intensity, ambient temperature and its own physical parameters, and can be expressed as:

[0072]

[0073] in, is the output power of the photovoltaic array at time t; f PV is the derating factor, which is used to describe the power drop of photovoltaic panels due to aging, dust, loss, etc., and is generally taken as 0.9; N is the number of photovoltaic panels; I tis the actual light intensity at time t; I STC P is the light intensity under standard test conditions; RPV is the rated power of a single photovoltaic panel; T STC The standard test temperature is 25°C; is the surface temperature of the photovoltaic panel at time t, which can be expressed as:

[0074]

[0075] in, is the ambient temperature at time t; v t is the wind speed at time t. In the present invention, the superscript or subscript t of a parameter represents the value of the parameter at time period t or time t, which will not be described in detail below.

[0076] 1.1.2) Distributed wind turbines

[0077] The output power of a grid-connected wind turbine can be expressed as a function of wind speed:

[0078]

[0079] in is the fan output power at time t; v t is the wind speed at time t, v in and v out are the cut-in and cut-out wind speeds of the fan respectively; v r is the rated wind speed of the fan, P RWT is the rated output power of the fan.

[0080] 1.1.3) Solar thermal collectors

[0081] A solar collector is a device that converts the sun's radiation into heat energy. It uses a coating to focus the scattered sunlight to achieve the purpose of heat collection. According to the law of conservation of energy, the effective energy Q obtained by the collector per unit time is u It is equal to the solar radiation energy S absorbed by the collector minus the energy Q dissipated by the collector to the surrounding environment. l :

[0082] Q u =SQ l (4)

[0083] S and Q l It is related to factors such as solar radiation, collector plate parameters and ambient temperature. The effective energy Q u It can be expressed as:

[0084] Q u =A p I(τα) e -A PU L (T p -T a ) (5)

[0085] A p and T p are the area and average temperature of the collector absorption plate respectively; I is the solar irradiance; U L is the total heat loss coefficient; T a is the ambient temperature; τ and α represent the effective transmittance and absorptivity respectively, and e represents dimensionless, i.e. (τα) e It represents the product of effective transmittance and absorptivity, and the quantity in brackets is dimensionless.

[0086] Taking the vacuum tube collector as an example, the total heat loss coefficient is the sum of the heat loss coefficient of the vacuum tube and the heat loss coefficient of the insulation box.

[0087] U L =U t +U b (6)

[0088]

[0089] U b =K b-a (8)

[0090] Where U t and U b are the heat loss coefficients of the vacuum tube and the insulation box respectively; K p-g and K g-a are the heat transfer coefficients of the heat absorbing plate and the glass tube, and the heat transfer coefficients of the glass tube and the surrounding environment; K b-a It represents the heat loss coefficient of the insulation box, which is determined by factors such as the thermal conductivity, thickness and surface area of ​​the insulation material.

[0091] 1.1.4) Solid Oxide Fuel Cells

[0092] Solid Oxide Fuel Cell (SOFC) is a fully solid-state energy conversion device that directly converts the chemical energy in fuel gas and oxidant gas into electrical energy. It has the structure of a general fuel cell. The input fuel is natural gas, and its output electrical power and thermal power are given by the following formula:

[0093]

[0094]

[0095] Where P SOFC and H SOFC Respectively represent the electrical power and thermal power output of SOFC; η his the reversible thermodynamic efficiency of the fuel cell, which is the ratio of the Gibbs free energy per unit fuel to the heat content of the fuel; and F SOFC The actual efficiency is obviously lower than the reversible thermodynamic efficiency. The main reason is that there is voltage loss and fuel utilization loss in the fuel cell reaction process. Therefore, the output electrical power and thermal power can be updated as follows:

[0096]

[0097]

[0098] Where η v and η g are voltage efficiency and fuel efficiency, respectively.

[0099] 1.2. Energy storage devices

[0100] 1.2.1) Electrical energy storage

[0101] The energy storage SOC (state of charge) can be expressed as:

[0102]

[0103] In the formula, and Respectively represent the state of charge of the energy storage at time t and time t-1; σ e is the self-discharge rate of the electrical energy storage; and are charging power and charging efficiency respectively, and are discharge power and discharge efficiency respectively; E EES is the energy storage battery capacity.

[0104] 1.2.2) Bedrock Energy Storage

[0105] Bedrock energy storage is to store heat in the bedrock by drilling holes in the underground bedrock, and then release the heat when needed. The energy conversion process has less loss and high conversion efficiency. Analogous to the SOC formula of electric energy storage, the SOC of bedrock energy storage can be expressed as:

[0106]

[0107] In the formula, and Respectively represent the bedrock energy storage charge state at time t and time t-1; σ h is the self-heat release rate of bedrock energy storage; and are the heating power and heating efficiency respectively, and are heat release power and heat release efficiency respectively; Q BES The bedrock energy storage capacity.

[0108] 1.3. Energy conversion equipment

[0109] The energy conversion equipment includes a high-temperature water tank, an electric heat pump, a lithium bromide refrigerator, and a magnetic levitation refrigerator. The conversion process is as follows: the heat energy of the fuel cell and the solar thermal collection system passes through the high-temperature water tank to make the hot water temperature reach the utilization requirements of the lithium bromide unit, and the refrigeration is supplied to cold users. Hot water that cannot be used for refrigeration enters the low-temperature water tank first, and the output reaches the specified temperature to meet the needs of hot water users; if there is still excess hot water, it enters the bedrock energy storage system for storage; if the hot water still cannot meet the user's needs, it is supplemented by the heat stored in the bedrock energy storage system. When the hot and cold load demand still cannot be met, it is supplied by electric heat pumps and magnetic levitation refrigerators.

[0110] Energy conversion devices can be represented by a unified model:

[0111]

[0112] In the formula, and are the input power and output power of energy conversion device i at time t, η i is the conversion efficiency of energy conversion device i.

[0113] 1.4. User load requirements

[0114] User loads include electricity, heat, cooling loads and electric vehicles, among which electricity, heat and cooling loads are obtained by prediction. Electric vehicles are an important part of the micro energy grid, playing an important role in system peak shaving, demand side response, reducing fossil energy consumption, and mitigating climate change. Considering the energy interaction between electric vehicles and micro energy grids, once connected, the micro energy grid regards electric vehicles as a special mobile energy storage device, and the charging and discharging power and charging and discharging time are controlled by the micro energy grid.

[0115] Considering the differences in user travel habits and electricity demand, the parameters of electric vehicles are different. For any electric vehicle, its relevant parameters can be expressed by the matrix Indicates that and They represent the time when the electric vehicle leaves and connects to the micro energy grid, and They represent the state of charge of the electric vehicle when it leaves and enters, respectively, and can be obtained by the Monte Carlo method:

[0116]

[0117]

[0118] Formula (16-17) indicates that the state of charge when leaving and connecting satisfies the normal distribution. and denote the average values ​​of the electric vehicle group when leaving and joining, σ and They represent the standard deviation of the electric vehicle group when leaving and joining. x represents the horizontal coordinate of the normal distribution function, and It represents the distribution law of the state of charge of all electric vehicles when leaving and connecting. Random sampling is performed in this distribution to obtain the specific value of the state of charge when leaving and connecting.

[0119] During the time from when an electric vehicle is connected to the grid to when it leaves the grid, the state of charge can be expressed by analogy with energy storage:

[0120]

[0121] In the formula, and Respectively represent the state of charge of the electric vehicle at time t and time t-1; σ ev is the electric vehicle self-discharge rate; and are charging power and charging efficiency respectively, and are discharge power and discharge efficiency respectively; E EV The battery capacity of electric vehicles.

[0122] Step 102: On a day-ahead time scale, the mathematical models of the various devices are used to obtain a day-ahead dispatch plan for the eight-in-one micro-energy network with the goal of minimizing the total cost within the optimization period while satisfying the constraints.

[0123] Day-ahead optimization scheduling is an important part of ensuring the safe and economic operation of the micro-energy network. The day-ahead optimization scheduling cycle is 24 hours and the time scale is 1 hour. Based on the 24-hour forecast sequence of renewable energy output and cooling, heating and power loads on the next day, the equipment model established in step 101 is used to meet the constraints of cooling, heating and power balance, equipment ramping, safety, etc., and with the goal of minimizing the total cost within the optimization cycle, to obtain the scheduling plan for the micro-energy network on the next day.

[0124] The total cost includes energy storage, heat storage, electric vehicle charging and discharging costs, energy conversion equipment maintenance costs, diesel generator power generation costs, interactive power purchase and sales costs with the large power grid, and natural gas purchase costs. The optimal dispatching scheme of the micro-energy network includes dispatching the output of the electric energy side, thermal energy side, and coupling equipment. The dispatching objects on the electric energy side are electric energy storage and electric vehicle charging and discharging power, diesel generator power generation power, and solid oxide fuel cell power generation power; the dispatching objects on the thermal energy side are bedrock energy storage charging and discharging power; the dispatching objects of the coupling equipment are electric heat pumps, magnetic levitation refrigerators, and lithium bromide refrigerators. The day-ahead optimization dispatching results (day-ahead dispatching plan) serve as a reference for real-time optimization dispatching and guide real-time optimization dispatching.

[0125] The objective function and constraints are introduced as follows.

[0126] 2.1. Objective function

[0127] The objective function of the day-ahead optimization dispatch is to optimize economic efficiency. Economic efficiency means that the total cost of the optimization period is minimized, including the energy storage cost. Thermal storage costs Electric vehicle charging and discharging cost C EV , energy conversion equipment maintenance costs, diesel generator power generation costs Cost of electricity purchase and sale with the large power grid and gas purchase costs The economic cost C can be expressed as:

[0128]

[0129] in,

[0130]

[0131] In the formula, C represents the total cost within the period; T represents the period of day-ahead optimization scheduling; N represents the number of energy conversion equipment; λ i and They represent the maintenance cost per unit output of energy conversion equipment i and the output power in time period t respectively; and denote the charging and discharging and charging and discharging heat costs of electrical energy storage and thermal energy storage in time period t, respectively; represents the power generation cost of the diesel generator in time period t; represents the cost of purchasing and selling electricity from the large power grid in time period t; represents the cost of purchasing natural gas during period t; K represents the charging and discharging cost of electric vehicles during period t. EES , K BES and K EVThey are the unit charging and discharging and unit charging and discharging heat costs of electric energy storage, thermal energy storage and EV (electric vehicle). a, b, c are the fuel cost coefficients of diesel generators; Represents the output of the diesel generator in time period t. represents the exchange power with the large power grid in time period t, where It means buying electricity from the big grid, and vice versa, it means selling electricity to the big grid; and are the electricity purchase and sale prices in time period t. t is the natural gas price in period t. m represents the number of electric vehicles, and They represent the charging power and discharging power of electric vehicle j in time period t, and They represent the charging efficiency and discharging efficiency of electric vehicle j respectively.

[0132] 2.2. Constraints

[0133] 2.2.1) System Constraints

[0134]

[0135]

[0136]

[0137]

[0138] Among them, formula (21) is the power constraint of the interconnection line between the micro energy grid and the external power grid, and formulas (22)-(24) are the power conservation of electric load, heat load and cooling load respectively. It is the maximum power allowed to be exchanged between the micro energy grid and the large power grid; outputting electrical power to the fuel cell; and are the input electrical powers of the magnetically suspended high-efficiency refrigerator and the electric heat pump, respectively; and They are photovoltaic and wind turbine output respectively. and They represent the output thermal power of the electric heat pump, the output thermal power of the solar power collection system and the input thermal power of the lithium bromide refrigeration unit respectively. and They are the output cooling power of the magnetic levitation high-efficiency refrigerator and the lithium bromide refrigerator respectively; and They are electrical load, heating load and cooling load respectively.

[0139] 2.2.2) Equipment output power constraints

[0140]

[0141]

[0142]

[0143]

[0144]

[0145] Among them, equations (25-29) are the power constraints of the diesel generator output electric power, the fuel cell output electric power, the electric heat pump output thermal power, the centrifugal chiller output cooling power, and the absorption chiller output cooling power. They are the maximum values ​​of the output power of the diesel generator, the output power of the fuel cell, the output power of the electric heat pump, the output power of the centrifugal chiller, and the output power of the absorption chiller. Since the output power of the fuel cell is related to the thermal power, when the output power meets the constraint, the output thermal power automatically meets the constraint.

[0146] 2.2.3) Constraints on electrical energy storage components

[0147]

[0148]

[0149]

[0150]

[0151] Among them, formula (30) is the constraint of the charging and discharging state of the energy storage battery, formulas (31) and (32) are the upper and lower limit constraints of the charging and discharging power of the energy storage battery, respectively, and formula (33) is the capacity constraint of the energy storage battery. and A binary variable representing the charge and discharge status of the energy storage battery, Indicates that the energy storage battery is charging at the current time t, otherwise Indicates that the energy storage battery is not charged at the current time t. Same reason. and Respectively represent the minimum and maximum charging power of the energy storage battery; and Respectively represent the minimum and maximum discharge power of the energy storage battery; and They are respectively the minimum and maximum charge states of the energy storage element (energy storage battery). Indicates the charge state of the energy storage battery at the current time t.

[0152] 2.2.4) Constraints of bedrock energy storage elements

[0153]

[0154]

[0155]

[0156]

[0157] Among them, formula (34) is the bedrock energy storage charging and discharging heat state constraint, formula (35) and (36) are the upper and lower limit constraints of the bedrock energy storage charging and discharging heat power, respectively, and formula (37) is the capacity constraint of the bedrock energy storage. The meaning of the parameters is the same as that of electric energy storage. In the present invention, the same superscript or subscript of the parameters represents the same meaning, and the meaning of the parameters can be derived from each other, which will not be repeated here.

[0158] For example, and A binary variable representing the charge and discharge status of the bedrock energy storage element, Indicates that the bedrock energy storage element is charged at the current time t, otherwise, Indicates that the bedrock energy storage element is not charged at the current time t; It means that the bedrock energy storage element is discharged at the current time t, otherwise, It indicates that the bedrock energy storage element is not discharged at the current time t. and They represent the minimum and maximum charging power of the bedrock energy storage element respectively; and They represent the minimum and maximum discharge power of the bedrock energy storage element respectively; and are the minimum and maximum charge states of the bedrock energy storage element, respectively; Indicates the charge state of the bedrock energy storage element at the current time t.

[0159] 2.2.5) Electric vehicle constraints

[0160]

[0161]

[0162]

[0163]

[0164] The meaning of formula (38-41) is the same as the energy storage constraint. Electric vehicles should also have the following constraints:

[0165]

[0166]

[0167]

[0168] In the formula, and They represent the state transition variables of the electric vehicle from charge to discharge and from discharge to charge, which are binary variables. Formula (42) indicates that the electric vehicle cannot be switched from charge to discharge or from discharge to charge in the same period. Formula (43) indicates that the charge-discharge conversion state value of a certain period is related to the discharge state of the adjacent period. c and t d are the time periods when electric vehicles are connected to and left the grid, respectively, and N EV It is a given integer used to constrain the number of charge and discharge times of EV after it is connected to the micro energy grid.

[0169] The step 102 uses the mathematical models of the various devices under the day-ahead time scale, and takes the lowest total cost within the optimization period as the goal while satisfying the constraints, to obtain the day-ahead scheduling plan of the eight-in-one micro-energy network. The day-ahead scheduling plan takes economic cost as the goal, and when each device satisfies the device constraints, the output of each device is obtained according to the device model, specifically, the day-ahead scheduling plan is based on the hourly scale, including the diesel generator power P DG , Solid oxide fuel cell power P SOFC 、Charging and discharging power of energy storage element P ch and P dis 、Electric vehicle charging and discharging power E ch and E dis , Interaction power with large power grid P grid , Power consumption of magnetic levitation refrigerator P MLBR , Electric heat pump power consumption P EHP ; Heat power of electric heat pump H EHP , solid oxide fuel cell thermal power H SOFC 、Bedrock energy storage charging and discharging power H ch and H dis 、Heat power consumption of lithium bromide refrigerator H LBR ; Magnetic suspension refrigerator cooling power F MLER , LiBr refrigerator cooling power F LBR .

[0170] Step 103: Under the real-time time scale, based on the model predictive control method, while satisfying the constraints, with the goal of following the day-ahead scheduling plan and smooth scheduling, obtain the real-time scheduling plan of the eight-in-one micro energy network.

[0171] The problem faced by real-time optimization scheduling is the uncertainty of renewable energy output and cold, heat and electricity load. Real-time prediction is more accurate than day-ahead prediction. Day-ahead scheduling plan has errors and cannot match the real-time system status. Real-time rescheduling is required based on day-ahead plan. Model predictive control can update scheduling decisions in real time according to the latest status of the micro-energy network. The single rolling cycle of real-time optimization scheduling is 1h, and the time scale is 15min. Every 15min, the cold, heat and electricity load and renewable energy output of 4 periods (15min each period) in the next hour are predicted. The scheduling results of the previous period are used as feedback information, the prediction sequence and the day-ahead scheduling results of the period are input into the real-time optimization model together. The objective function is solved according to the constraints to obtain the output of the four periods in the cycle. Then, the output of the first period of the four period output results is used as the scheduling result of the period, and the results of the remaining periods are ignored. The above operation is repeated when the next period arrives, and finally the real-time optimization scheduling result (real-time scheduling plan) of 24h (a total of 96 periods) is obtained. Compared with the day-ahead optimization, the constraint conditions of real-time optimization scheduling have two objective functions. The first part is to minimize the deviation of the dispatchable equipment output and energy storage output relative to the day-ahead planned reference value during the real-time optimization period; the second part is to minimize the sum of the dispatchable equipment output and energy storage output adjustments between adjacent time periods during the real-time optimization period. The principle of the model predictive control method is as follows.

[0172] Model predictive control is a type of model-based closed-loop optimal control method that takes into account future time steps and has a finite time domain. It consists of three parts: prediction model, rolling optimization, and feedback correction.

[0173] 3.1 Prediction Model

[0174] The function of the prediction model is to obtain the scheduling results within the cycle by minimizing the objective function based on the renewable energy output, cooling, heating and electricity loads in four time periods in a future cycle obtained through day-ahead planning and prediction, while meeting certain constraints.

[0175] Day-ahead optimal scheduling ensures the economic efficiency of micro-energy grid operation and guides intraday scheduling. To ensure that the scheduling results track the day-ahead plan, the objective function of real-time optimal scheduling is divided into two parts. The first part is to minimize the deviation of the dispatchable equipment output and energy storage output relative to the day-ahead plan reference value during the real-time optimization period; the second part is to minimize the sum of the dispatchable equipment output and energy storage output adjustments between adjacent time steps during the real-time optimization period.

[0176]

[0177] Where t is the current time; T s Optimize the cycle for real-time rolling; N s is the number of dispatchable devices; W err and W uis the coefficient matrix; It is a real-time dispatchable equipment decision-making, including the output of each unit and the energy storage charge status; is the reference value of the day-ahead scheduling plan; It is the increment of the dispatchable equipment output relative to the previous period. The power interaction with the large power grid maintains the day-ahead plan, and the constraints are the same as the day-ahead optimization.

[0178] 3.2 Scrolling Optimization

[0179] Rolling optimization and feedback correction distinguish the model predictive control method from traditional optimization methods. Each optimization process optimizes a total of four periods from the current period to the end of the cycle. When it reaches the next period, the optimization process is repeated, and rolling optimization is formed by continuously moving forward. This requires that a forecast of renewable energy output and cooling, heating and electricity loads within the cycle be made once in each period from the current period, and the forecast is also updated rollingly. Therefore, the optimization process of model predictive control is different from the offline one-time optimization method of traditional optimization methods, but is a repeated online rolling optimization. The rolling optimization of a limited period may not be able to obtain the global optimum, but it can continuously take into account the impact of uncertainty and make corrections in time. It is more adaptable to the actual process than relying only on the optimization of the model once, and has stronger robustness.

[0180] 3.3 Feedback Correction

[0181] By inputting the forecast information and historical scheduling information of the next period into the real-time optimization method, the scheduling result within a period can be obtained. In order to prevent the scheduling result from deviating from the ideal state due to uncertainty interference, only the scheduling result of the current period is used, and the scheduling result of the future period within the period is abandoned. In the next period, the updated real-time forecast information is used to correct the model-based forecast, and the optimization result of the previous period of 15 minutes is fed back to the input end to make the scheduling result more stable and smooth. Repeat the above steps for a new round of optimization.

[0182] Figure 4 A schematic diagram of a real-time optimization process provided by an embodiment of the present invention. Figure 4 As shown, It represents the day-ahead forecast set at time t, which consists of two parts: one is the electric heating and cooling load forecast, and the other is the renewable energy output forecast, which is expressed as

[0183] In step 103, based on the model predictive control method, the real-time scheduling plan of the eight-in-one micro-energy network is obtained based on the real-time time scale, while satisfying the constraints, with the goal of following the day-ahead scheduling plan and smooth scheduling. The real-time scheduling plan is a real-time scheduling scheme based on a 15-minute scale, including the diesel generator power P DG , Solid oxide fuel cell power P SOFC、Charging and discharging power of energy storage element P ch and P dis 、Electric vehicle charging and discharging power E ch and E dis , Power consumption of magnetic levitation refrigerator P MLBR , Electric heat pump power consumption P EHP 、Electric heat pump thermal power H EHP , solid oxide fuel cell thermal power H SOFC 、Bedrock energy storage charging and discharging power H ch and H dis 、Heat power consumption of lithium bromide refrigerator H LBR ; Magnetic suspension refrigerator cooling power F MLER , LiBr refrigerator cooling power F LBR The real-time dispatching scheme does not include the power interaction with the large grid, and the power interaction with the large grid directly follows the day-ahead plan.

[0184] The optimization in model predictive control is not only based on the model, but also utilizes feedback information, thus forming a closed-loop optimization. Through the model predictive control method, the real-time operation of the micro-energy grid follows the day-ahead plan as much as possible, while smoothly controlling the increase and decrease of output of each device caused by uncertainty.

[0185] Step 104: Perform real-time scheduling of various devices in the eight-in-one micro energy network according to the real-time scheduling plan.

[0186] The present invention proposes a detailed analysis and modeling of the power generation, conversion and storage equipment in the eight-in-one micro-energy grid system. In addition to the traditional equipment modeling, it adds the modeling of emerging micro-energy grid equipment such as solar thermal collection systems and bedrock energy storage; and proposes a two-stage optimization scheduling method based on model predictive control. The real-time scheduling plan obtained according to the method of the present invention performs real-time scheduling of various devices in the eight-in-one micro-energy grid, which can improve the optimization control accuracy of the micro-energy grid and has good robustness to the uncertainty of cold, hot and electric loads and renewable energy.

[0187] In summary, the present invention is a method for optimizing and dispatching a micro-energy network based on model predictive control. First, each energy device in the "eight-in-one" micro-energy network is modeled in detail to meet the actual requirements of the increasingly complex micro-energy network. Then, on the day-ahead time scale, the forecast data of renewable energy output and load in each period of the next day are used to obtain the day-ahead optimization result with the goal of economic optimization. Then, on the real-time time scale, based on the model prediction method, the cold, heat and electricity load and renewable energy output of the next hour are predicted every 15 minutes, with the goal of following the day-ahead plan and smooth scheduling, and the "prediction model-rolling optimization-feedback correction" process is repeated to finally obtain the real-time scheduling result. By using the model predictive control method, the real-time rescheduling result of the present invention tracks the day-ahead plan on the one hand, ensuring the economy of real-time scheduling; on the other hand, it greatly reduces the impact of the uncertainty of cold, heat and electricity load and renewable energy output on the micro-energy network, and has strong robustness.

[0188] Based on the micro energy network optimization scheduling method based on model predictive control provided by the present invention, the present invention also provides a micro energy network optimization scheduling system based on model predictive control. Figure 5 This is a structural diagram of a micro energy grid optimization scheduling system based on model predictive control of the present invention, such as Figure 5 As shown, the system comprises:

[0189] The micro-energy network equipment modeling module 501 is used to model various equipment in the eight-in-one micro-energy network and generate mathematical models of various equipment; the eight-in-one micro-energy network is a micro-energy network including typical equipment of wind, light, rock, magnetism, combustion, heat, storage and load; the various equipment include energy production equipment, energy storage equipment and energy conversion equipment; the energy production equipment includes distributed photovoltaic units, distributed wind turbines, solar collectors, and solid oxide fuel cells; the energy storage equipment includes electrical energy storage and bedrock energy storage; the energy conversion equipment includes high-temperature water tanks, electric heat pumps, lithium bromide refrigerators and magnetic suspension refrigerators;

[0190] The day-ahead optimization module 502 is used to obtain the day-ahead scheduling plan of the eight-in-one micro-energy network by using the mathematical models of the various devices under the day-ahead time scale, while satisfying the constraints and taking the minimum total cost within the optimization period as the goal;

[0191] A real-time optimization module 503 is used to obtain a real-time scheduling plan of the eight-in-one micro energy network based on a model predictive control method under a real-time time scale, while satisfying constraints and aiming at following the day-ahead scheduling plan and smooth scheduling;

[0192] The micro energy network device scheduling module 504 is used to perform real-time scheduling of various devices in the eight-in-one micro energy network according to the real-time scheduling plan.

[0193] The micro energy network equipment modeling module 501 specifically includes:

[0194] A distributed photovoltaic unit modeling unit is used to model the distributed photovoltaic unit and generate a mathematical model of the distributed photovoltaic unit. in is the output power of the photovoltaic array at time t; f PV is the derating factor; N is the number of photovoltaic panels; P RPV is the rated power of a single photovoltaic panel; I t is the actual light intensity at time t; I STC is the light intensity under standard test conditions; T STC is the standard test temperature; is the surface temperature of the photovoltaic panel at time t;

[0195] A distributed wind turbine modeling unit is used to model the distributed wind turbine and generate a mathematical model of the distributed wind turbine. in is the fan output power at time t; v t is the wind speed at time t; v in and v out are the cut-in and cut-out wind speeds of the fan respectively; v r is the rated wind speed of the fan; P RWT is the rated output power of the fan;

[0196] A solar thermal collector modeling unit is used to model the solar thermal collector and generate a mathematical model Q of the solar thermal collector. u =A p I(τα) e -A P U L (T p -T a );Q u A is the effective energy obtained by the solar collector per unit time; p and T p are the area and average temperature of the solar collector absorber, respectively; I is the solar irradiance; τ and α represent the effective transmittance and absorptivity, respectively, and e represents dimensionless; U L is the total heat loss coefficient; T a is the ambient temperature;

[0197] A solid oxide fuel cell modeling unit is used to model the solid oxide fuel cell and generate a mathematical model of the solid oxide fuel cell. and Where P SOFC and HSOFC Respectively represent the electrical power and thermal power output of the solid oxide fuel cell; η h is the reversible thermodynamic efficiency of solid oxide fuel cell; η v and η g are voltage efficiency and fuel utilization, respectively; and F SOFC The lower calorific value of natural gas and the natural gas consumption per unit time respectively;

[0198] An electric energy storage modeling unit, used to model the electric energy storage and generate a mathematical model of the electric energy storage in Indicates the state of charge of the energy storage; σ e is the self-discharge rate of the electrical energy storage; and They are charging power and charging efficiency respectively; and are discharge power and discharge efficiency respectively; E EES is the electrical energy storage capacity;

[0199] A bedrock energy storage modeling unit, used to model the bedrock energy storage and generate a mathematical model of the bedrock energy storage in represents the charge state of bedrock energy storage; σ h is the self-heat release rate of bedrock energy storage; and are the heating power and heating efficiency respectively; and are heat release power and heat release efficiency respectively; Q BES is the bedrock energy storage capacity;

[0200] An energy conversion device modeling unit is used to model the energy conversion device and generate a mathematical model of the energy conversion device. in and are the input power and output power of energy conversion device i at time t, η i is the conversion efficiency of energy conversion device i.

[0201] The day-ahead optimization module 502 and the real-time optimization module 503 both include:

[0202] The constraint condition limiting unit is used to limit various devices in the eight-in-one micro energy network to simultaneously meet system constraints, device output power constraints, electric energy storage element constraints, bedrock energy storage element constraints and electric vehicle constraints.

[0203] The day-ahead optimization module 502 specifically includes:

[0204] Economically optimal unit, used to minimize the total cost within the cycle minimum; where C represents the total cost within the period; T represents the period of the day-ahead optimal scheduling; N represents the number of energy conversion equipment; λ i and They represent the maintenance cost per unit output of energy conversion equipment i and the output power in time period t respectively; and denote the charging and discharging and charging and discharging heat costs of electrical energy storage and thermal energy storage in time period t, respectively; represents the power generation cost of the diesel generator in time period t; represents the cost of purchasing and selling electricity from the large power grid in time period t; represents the cost of purchasing natural gas; Represents the charging and discharging cost of electric vehicles.

[0205] The real-time optimization module 503 specifically includes:

[0206] The minimum unit of the objective function is used to make various devices in the eight-in-one micro energy network meet the objective function of real-time optimization scheduling Where t is the current time; T s Optimize the cycle for real-time rolling; N s is the number of dispatchable equipment; Make decisions for real-time dispatchable equipment; is the reference value of the day-ahead scheduling plan; W err and W u is the coefficient matrix; It is the increment of dispatchable equipment output relative to the previous period.

[0207] The optimization scheduling problem of micro-energy grid is mainly studied from two perspectives: equipment model establishment and optimization solution. In existing studies, the single time scale day-ahead optimization scheduling does not take into account the prediction error caused by the uncertainty of renewable energy output. It is difficult to solve the uncertainty of load and renewable energy in actual operation. The result of decision-making according to the day-ahead strategy lacks economy in the actual operation of the system, and is even not feasible. The existing day-ahead-real-time two-stage optimization scheduling methods are mostly open-loop control methods, and there is no feedback mechanism to correct the optimization control process. Even if rolling optimization measures are used in a few two-stage optimization scheduling methods, they are simple micro-energy grids, and the modeling of energy types and equipment is relatively simple, which is difficult to meet the needs of optimization scheduling of micro-energy grid systems with increasing complexity at this stage and in the future.

[0208] In view of the defects and difficulties of the above-mentioned single time scale and two-stage optimization method, the present invention proposes a micro-energy grid optimization scheduling method based on model predictive control. This method combines the optimization of two time scales: day-ahead and real-time. Day-ahead optimization takes economic optimization as the goal and obtains a day-ahead scheduling plan, which guides real-time scheduling. Real-time optimization goes through three steps: prediction model-rolling optimization-feedback correction. Following the day-ahead plan, the economic requirements of real-time scheduling are met. Feedback correction sends the current scheduling result to the input end, which greatly reduces the fluctuation of adjacent outputs caused by uncertainty, making the scheduling result smoother and more stable.

[0209] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0210] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A micro energy grid optimization scheduling method based on model predictive control, characterized in that: include: Modeling various devices in the eight-in-one micro-energy network to generate mathematical models of various devices; the eight-in-one micro-energy network is a micro-energy network including typical devices of wind, light, rock, magnetism, fuel, heat, storage and load; the various devices include energy production equipment, energy storage equipment and energy conversion equipment; the energy production equipment includes distributed photovoltaic units, distributed wind turbines, solar collectors, solid oxide fuel cells; the energy storage equipment includes electrical energy storage and bedrock energy storage; the energy conversion equipment includes high-temperature water tanks, electric heat pumps, lithium bromide refrigerators and magnetic suspension refrigerators; On a day-ahead time scale, the mathematical models of the various devices are used to obtain a day-ahead dispatch plan of the eight-in-one micro-energy grid with the goal of minimizing the total cost within the optimization period while satisfying the constraints; Under the real-time time scale, based on the model predictive control method, while satisfying the constraints, the real-time scheduling plan of the eight-in-one micro-energy network is obtained with the goal of following the day-ahead scheduling plan and smooth scheduling; The goal of following the day-ahead scheduling plan and smooth scheduling specifically includes: Satisfy the objective function of real-time optimization scheduling ;in for the current moment; Optimize cycles for real-time rolling; is the number of dispatchable devices; Make decisions for real-time dispatchable equipment; is the reference value of the day-ahead scheduling plan; and is the coefficient matrix; It is the increment of dispatchable equipment output relative to the previous period; Various devices in the eight-in-one micro energy network are scheduled in real time according to the real-time scheduling plan.

2. The method according to claim 1, characterized in that The modeling of various devices in the eight-in-one micro energy network to generate mathematical models of various devices specifically includes: Modeling the distributed photovoltaic unit to generate a mathematical model of the distributed photovoltaic unit ;in for The output power of the photovoltaic array at each moment; is the derating factor; is the number of photovoltaic panels; is the rated power of a single photovoltaic panel; for The actual light intensity at the moment; is the light intensity under standard test conditions; is the standard test temperature; for Surface temperature of photovoltaic panel at the moment; Modeling the distributed wind turbine generator set to generate a mathematical model of the distributed wind turbine generator set ;in for The fan output power at each moment; for Wind speed at the moment; and are the cut-in and cut-out wind speeds of the fan respectively; is the rated wind speed of the fan; is the rated output power of the fan; Modeling the solar thermal collector to generate a mathematical model of the solar thermal collector ;in is the effective energy obtained by the solar collector per unit time; and are the solar collector absorber plate area and average temperature respectively; is the solar irradiance; and represent the effective transmittance and absorption rate respectively, It means dimensionless; is the total heat loss coefficient; is the ambient temperature; Modeling the solid oxide fuel cell to generate a mathematical model of the solid oxide fuel cell and ;in and represent the electrical power and thermal power output of the solid oxide fuel cell, respectively; is the reversible thermodynamic efficiency of solid oxide fuel cells; and are voltage efficiency and fuel utilization, respectively; and The lower calorific value of natural gas and the natural gas consumption per unit time respectively; Modeling the electric energy storage to generate a mathematical model of the electric energy storage ;in Indicates the state of charge of the energy storage; is the self-discharge rate of the electrical energy storage; and They are charging power and charging efficiency respectively; and are discharge power and discharge efficiency respectively; is the electrical energy storage capacity; Modeling the bedrock energy storage to generate a mathematical model of the bedrock energy storage ;in Indicates the state of charge of bedrock energy storage; is the self-heat release rate of bedrock energy storage; and are the heating power and heating efficiency respectively; and are heat release power and heat release efficiency respectively; is the bedrock energy storage capacity; Modeling the energy conversion device to generate a mathematical model of the energy conversion device ;in and Energy conversion equipment exist The input power and output power at the moment, Energy conversion equipment conversion efficiency.

3. The method according to claim 1, characterized in that The constraint conditions are satisfied, specifically including: At the same time, system constraints, equipment output power constraints, electric energy storage element constraints, bedrock energy storage element constraints and electric vehicle constraints are met.

4. The method according to claim 1, characterized in that The goal of minimizing the total cost within the optimization cycle specifically includes: Total cost within the cycle Lowest; among represents the total cost within the period; Indicates the period of day-ahead optimization scheduling; Indicates the number of energy conversion devices; and Represents energy conversion equipment Maintenance cost per unit output and Output power during the time period; and Represents electrical energy storage and thermal energy storage in the time period The cost of charging, discharging and heat dissipation; Indicates the diesel generator in the time period The cost of electricity generation; Indicates during the period The cost of purchasing and selling electricity from the large power grid; represents the cost of purchasing natural gas; Represents the charging and discharging cost of electric vehicles.

5. A micro energy grid optimization scheduling system based on model predictive control, characterized in that: include: A micro-energy network equipment modeling module is used to model various equipment in the eight-in-one micro-energy network and generate mathematical models of various equipment; the eight-in-one micro-energy network is a micro-energy network including typical equipment of wind, light, rock, magnetism, fuel, heat, storage and load; the various equipment include energy production equipment, energy storage equipment and energy conversion equipment; the energy production equipment includes distributed photovoltaic units, distributed wind turbines, solar collectors, and solid oxide fuel cells; the energy storage equipment includes electrical energy storage and bedrock energy storage; the energy conversion equipment includes high-temperature water tanks, electric heat pumps, lithium bromide refrigerators and magnetic suspension refrigerators; A day-ahead optimization module is used to obtain a day-ahead dispatch plan of the eight-in-one micro-energy network by using mathematical models of the various devices under the day-ahead time scale, while satisfying constraints and taking the lowest total cost within the optimization period as the goal; A real-time optimization module is used to obtain a real-time scheduling plan of the eight-in-one micro-energy network based on a model predictive control method under a real-time time scale, while satisfying constraints and aiming to follow the day-ahead scheduling plan and smooth scheduling; The real-time optimization module specifically includes: The minimum unit of the objective function is used to make various devices in the eight-in-one micro energy network meet the objective function of real-time optimization scheduling ;in for the current moment; Optimize cycles for real-time rolling; is the number of dispatchable devices; Make decisions for real-time dispatchable equipment; is the reference value of the day-ahead scheduling plan; and is the coefficient matrix; It is the increment of dispatchable equipment output relative to the previous period; The micro energy network equipment scheduling module is used to perform real-time scheduling of various equipment in the eight-in-one micro energy network according to the real-time scheduling plan.

6. The system according to claim 5, characterized in that The micro energy network equipment modeling module specifically includes: A distributed photovoltaic unit modeling unit is used to model the distributed photovoltaic unit and generate a mathematical model of the distributed photovoltaic unit. ;in for The output power of the photovoltaic array at each moment; is the derating factor; is the number of photovoltaic panels; is the rated power of a single photovoltaic panel; for The actual light intensity at the moment; is the light intensity under standard test conditions; is the standard test temperature; for Surface temperature of photovoltaic panel at the moment; A distributed wind turbine modeling unit is used to model the distributed wind turbine and generate a mathematical model of the distributed wind turbine. ;in for The fan output power at each moment; for Wind speed at the moment; and are the cut-in and cut-out wind speeds of the fan respectively; is the rated wind speed of the fan; is the rated output power of the fan; A solar thermal collector modeling unit is used to model the solar thermal collector and generate a mathematical model of the solar thermal collector. ;in is the effective energy obtained by the solar collector per unit time; and are the solar collector absorber plate area and average temperature respectively; is the solar irradiance; and represent the effective transmittance and absorption rate respectively, It means dimensionless; is the total heat loss coefficient; is the ambient temperature; A solid oxide fuel cell modeling unit is used to model the solid oxide fuel cell and generate a mathematical model of the solid oxide fuel cell. and ;in and represent the electrical power and thermal power output of the solid oxide fuel cell, respectively; is the reversible thermodynamic efficiency of solid oxide fuel cells; and are voltage efficiency and fuel utilization, respectively; and The lower calorific value of natural gas and the natural gas consumption per unit time respectively; An electric energy storage modeling unit, used to model the electric energy storage and generate a mathematical model of the electric energy storage ;in Indicates the state of charge of the energy storage; is the self-discharge rate of the electrical energy storage; and They are charging power and charging efficiency respectively; and are discharge power and discharge efficiency respectively; is the electrical energy storage capacity; A bedrock energy storage modeling unit, used to model the bedrock energy storage and generate a mathematical model of the bedrock energy storage ;in Indicates the state of charge of bedrock energy storage; is the self-heat release rate of bedrock energy storage; and are the heating power and heating efficiency respectively; and are heat release power and heat release efficiency respectively; is the bedrock energy storage capacity; An energy conversion device modeling unit is used to model the energy conversion device and generate a mathematical model of the energy conversion device. ;in and Energy conversion equipment exist The input power and output power at the moment, Energy conversion equipment conversion efficiency.

7. The system according to claim 5, characterized in that The day-ahead optimization module and the real-time optimization module both include: The constraint condition limiting unit is used to limit various devices in the eight-in-one micro energy network to simultaneously meet system constraints, device output power constraints, electric energy storage element constraints, bedrock energy storage element constraints and electric vehicle constraints.

8. The system according to claim 5, characterized in that The day-ahead optimization module specifically includes: Economically optimal unit, used to minimize the total cost within the cycle Lowest; among represents the total cost within the period; Indicates the period of day-ahead optimization scheduling; Indicates the number of energy conversion devices; and Represents energy conversion equipment Maintenance cost per unit output and Output power during the time period; and Represents electrical energy storage and thermal energy storage in the time period The cost of charging, discharging and heat dissipation; Indicates the diesel generator in the time period The cost of electricity generation; Indicates during the period The cost of purchasing and selling electricity from the large power grid; represents the cost of purchasing natural gas; Represents the charging and discharging cost of electric vehicles.

Images