Electric heating integrated energy system optimization scheduling method based on IGDT theory

Abstract

The invention discloses an integrated electric heating energy system optimization scheduling method based on an IGDT theory, a scheduling method based on an information gap decision theory IGDT is introduced, and a risk pursuit opportunity IGDT scheduling model and a risk avoidance robust IGDT scheduling model are constructed. According to the model, factors such as output characteristics, starting and stopping states and climbing limitation of generator set systems such as wind power generation, photovoltaic power generation, a combined heat and power generation system and a fuel cell are fully considered, and coupling with charging and discharging states of an electric boiler, a heat pump, an electric energy storage device and a heat energy storage device is carried out. By using Cplex optimization software, the operation cost of the whole system is minimized while the optimal output power of each device in a given scheduling period is calculated. According to the method, the generalization and superiority of the IGDT in the aspect of coping with the information gap caused by uncertainty are verified through the comparison of the optimization scheduling results of two typical days and the comparison of the total cost of the robust strategy model and the opportunity strategy scheduling model under different deviation factors.

Description

Technical Field

[0001] The present invention relates to the technical field of energy scheduling, and in particular to an optimization scheduling method for an electric-thermal integrated energy system based on IGDT theory. Background Art

[0002] The integrated energy system, leveraging its multi-source synergy, has become a key platform for improving energy efficiency and promoting the uptake of renewable energy. As an innovative energy supply and management model, this system not only promotes the use of green energy but also reduces reliance on fossil fuels, marking a significant milestone in my country's energy supply transformation. However, due to the intermittent and random nature of wind power generation and the constraints of solar power generation on seasonal variations, meteorological conditions, and other factors, the actual output of renewable energy sources such as wind and solar energy is difficult to accurately predict.

[0003] Current research on strategies for handling uncertainty in renewable energy generation in power grids can be broadly categorized into two approaches. One approach proposes a robust optimization framework based on the extreme fluctuations of wind and solar power generation, then defines uncertainty using the concept of sets. Therefore, the definition of uncertainty sets is crucial for accurately characterizing potential variations in such robust models. Researchers have proposed robust reserve dispatch schemes and economic dispatch strategies that utilize multi-scenario analysis of extreme wind power scenarios. These robust models have been algorithmically implemented using Benders decomposition techniques, demonstrating their advantages. However, this approach primarily focuses on the dispatch participation of traditional generation units, making it difficult to find effective solutions in the new environment of high penetration of renewable energy generation equipment. Another approach employs stochastic optimization based on renewable energy generation forecasts, constructing dispatch plans in the form of probability density functions. The goal of stochastic optimization is to maximize or minimize the expected target value while satisfying the probability that the constraints hold true at a specific confidence level. Both strategies for handling uncertainty have their own advantages and disadvantages. Robust optimization ensures that all elements in an uncertain set satisfy the constraints, thus helping to improve system reliability. However, this comes with the added problem of reducing system economic efficiency. Stochastic optimization, on the other hand, aims to achieve "optimal" results from a probabilistic perspective. However, if the output or load of distributed power sources significantly exceeds predicted values during actual operation, this can result in decisions that fall outside the acceptable risk range. Furthermore, stochastic optimization must be performed given the distribution function and historical data of the uncertainties. If decision makers are faced with Knightian uncertainty, i.e., risks whose probability cannot be calculated or measured, stochastic optimization methods lose their effectiveness. Summary of the Invention

[0004] The purpose of the present invention is to provide an optimized scheduling method for an electric-thermal integrated energy system based on IGDT theory. By adopting the IGDT theory, namely the information gap decision theory, the uncertainty of wind and solar output and the influence of the decision maker's risk attitude are taken into account. According to the risk target preset by the dispatcher, the scheduling model is divided into a robust model and an opportunity model. The two models correspond to two different value orientations held by the dispatcher when dealing with system uncertainty. Combined with the scheduling results, the optimized scheduling of the electric-thermal integrated energy system is realized, aiming to solve the intermittent and random nature of wind power generation in the electric-thermal integrated energy system, and the efficiency and effectiveness of solar power generation are constrained by multiple factors such as seasonal changes, meteorological conditions and sunlight intensity, which makes it difficult to accurately predict the actual output of wind energy and light energy.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] An optimization scheduling method for an electric-thermal integrated energy system based on IGDT theory includes the following steps:

[0007] S100, collects operating parameters of various equipment and energy prices of the electric and thermal integrated energy system;

[0008] Various types of equipment in the electric and thermal integrated energy system include cogeneration systems, fans, thermal energy storage, photovoltaics, heat pumps, electric boilers, fuel cells, and electric energy storage equipment;

[0009] S200, establish an electric and thermal integrated energy system model;

[0010] S300: Establish an objective function for fuel cost, grid cost, equipment maintenance cost, environmental cost, and heat sales revenue:

[0011]

[0012] in

[0013]

[0014] C HE (t) = C he P he (t) (6)

[0015] Where: F is the total operating cost; N T is the total number of scheduling periods; C FU (t) is the fuel cost in period t; C EX (t) is the grid cost in period t; C ME (t) is the maintenance cost in period t; C CE (t) is the environmental cost in period t; C HE (t) is the heat sales revenue in period t; P ex(t) is the interaction power between the system and the large power grid during period t; C rs (t) is the electricity price during period t, C rb (t) is the electricity purchase price during period t; N M is the total number of units in the network; C Mi is the unit maintenance cost of unit i; P i (t) is the output of unit i during period t; P P is the conversion coefficient of electricity; h GT is the heating efficiency; C he P is the unit price of heat; he (t) is the heat load in the network during period t;

[0016] S400: Establishing constraints, including wind and solar output constraints, power balance constraints, energy storage balance constraints, decision variable boundary constraints, and some equipment ramping constraints;

[0017] S500: Establish an optimization model for the electric and thermal integrated energy system and call the Cplex solver to solve it. The solution algorithm includes nonlinear components and running or stopping state variables of devices such as gas turbines, electric boilers, fuel cells, fans, photovoltaics, heat pumps, and power grids in the objective function. The solution formula is:

[0018] min f(x,y)

[0019]

[0020] Where: The optimization variable x includes the power output of the gas turbine, electric boiler, fuel cell, heat pump, and other devices; the optimization variable y reflects the operating status of the fuel unit in operation or shutdown; the equality constraints are mainly based on the energy balance relationship within the system and the energy change law of the energy storage device; and the inequality constraints define the operating range of the equipment;

[0021] S600. Establish an electric-thermal integrated energy system scheduling model based on IGDT theory and select a fractional uncertainty model. The specific formula is:

[0022]

[0023] Where: α represents the fluctuation range of load demand, that is, uncertainty; represents the predicted value of load demand;

[0024] Actual value of load demand with uncertain factors The upper limit of the collection is The lower limit is

[0025] S700, builds opportunity model and robust model based on IGDT theory, and calls Cplex solver for solution;

[0026] S710. The IGDT scheduling model based on risk pursuit is a two-level planning model, and its formula is:

[0027]

[0028] Where: ε m is the chance deviation coefficient, DP OM is the expected profit threshold, DP0 refers to the best performance index that can be achieved by an exact electric and thermal system scheduling plan established based on the prediction of uncertain factors;

[0029] S720. A robust IGDT scheduling model based on risk aversion is formulated as follows:

[0030]

[0031] Where: m is the robust bias factor, DP RM is the robust return threshold.

[0032] Preferably, ε in the formula (9) m The value of is [0, 1), ε m There is a direct positive relationship between the specific value of and the decision maker's risk tolerance; compared with the method of operating solely based on electricity price fluctuations, the scheduling plan that takes load changes into consideration will lead to a decrease in total daily revenue; the actual electricity consumption is close to the lowest point in the expected range, which maximizes the daily profit calculated by the lower model; when the real-time load demand is the lower limit of the uncertainty set When , the daily profit of the lower model can reach the maximum value; the two-level programming model is transformed into the following single-level optimization model:

[0033]

[0034] In formula (10), δ m The value range of δ is [0, 1), m The specific value of is proportional to the decision maker's risk aversion; when the real-time load demand is When , which is the upper bound of the uncertainty set, the robust lower layer model can obtain the minimum value, and formula (25) is transformed into a single-layer optimization model:

[0035]

[0036] Preferably, the electric-thermal integrated energy system includes an electric power system and a thermal system; the electric power system includes a wind farm, a photovoltaic power station, a fuel cell, an electric boiler, a cogeneration system and a heat pump; the thermal system includes an electric boiler, a heat pump and a cogeneration system; the operating data of the electric-thermal integrated energy system includes the electric load, the thermal load power forecast, and the load forecast of the wind and solar output; the equipment parameters of the electric-thermal integrated energy system include the rated power and conversion efficiency, up and down climbing efficiency and maintenance cost of the output equipment, the energy release, energy storage power, total capacity, maintenance cost, efficiency and loss rate of the energy storage equipment.

[0037] Preferably, the output model of each device in the electric and thermal integrated energy system in S200 includes:

[0038] S210. The operation of electric boilers is regulated by the electricity price mechanism. Electric boilers work with the cogeneration system to increase electricity consumption during off-peak hours while meeting the heat load demand, thus coordinating the peak-valley differences between the electricity load and the heat load. The output model of the electric boiler is:

[0039] Q EB (t) = P EB (t)η ah (13)

[0040] Where: P EB (t) is the power consumption of the electric boiler in time period t; Q EB (t) is the heating power of the electric boiler in time period t; η ah The electric energy consumed by the electric boiler is dispatched as the system power load.

[0041] S220, cogeneration system including gas turbine and lithium bromide chiller;

[0042] S221. The mathematical model of the thermoelectric relationship of a gas turbine is:

[0043]

[0044] Where: Q GT (t) represents the waste heat generated by the gas turbine during the period t, P GT (t) represents the amount of gas generated by the gas turbine during the period t, η GT (t) represents the power generation efficiency of the gas turbine in the time period t; η L is the heat loss rate;

[0045] S222. The mathematical model of the thermoelectric relationship of lithium bromide refrigerator is:

[0046] Q GT-h (t) = Q GT (t)η h C OPh (15)

[0047] Where: Q GT-h (t) is the heating capacity of the lithium bromide refrigerator during time period t; C OPh is the thermal coefficient of the lithium bromide refrigerator, η h is the flue gas recovery rate of lithium bromide refrigerator;

[0048] S223. The fuel cost of the gas turbine in period t is:

[0049]

[0050] Where: Δt is the unit scheduling time; C GT (t) is the cost of fuel consumed by the gas turbine during period t; C CH4 is the price of natural gas; L LHV is the calorific value of natural gas;

[0051] S230, the fuel cost and electric power output characteristics of the fuel cell in time period t are:

[0052]

[0053] Where: C FC (t) is the fuel cost of the fuel cell in time period t; P FC (t) is the power generation of the fuel cell, η FC (t) is the conversion efficiency of the fuel cell;

[0054] S240, heating coefficient C of heat pump OPH The heat energy obtained by the user / the electricity or fuel energy consumed by the heat pump, C OPH =4.4;

[0055] S250, the energy storage system includes an electric energy storage system and a thermal energy storage system;

[0056] S251. The capacity of the power storage system and its charge and discharge capabilities must satisfy the following relationship:

[0057]

[0058] Where: E EES (t) is the capacity of the power energy storage system in time period t; τ is the self-discharge rate of the power energy storage system; P EES-ch (t), P EES-dis (t) are the charging and discharging power of the power storage system in time period t; η sch ,η sdis The efficiency of the electric energy storage system;

[0059] S252. The dynamic mathematical model of the thermal energy storage system is:

[0060]

[0061] Where: H HS (t) is the thermal energy storage capacity in time period t; μ is the thermal energy dissipation loss rate; P HS-ch (t), P HS-dis (t) are the heat absorption and release power of the thermal energy storage system in time period t; η hch ,η hdis are the efficiencies of the thermal energy storage system in time period t.

[0062] Preferably, the constraints include:

[0063] S410, wind and solar output constraints are:

[0064] Clean energy supplies include wind and solar power, and the amount of wind power that can be absorbed and utilized by the system is lower than expected:

[0065] 0.8P β,f ≤P β,t ≤P β,f (20)

[0066] Where: P β，f represents the predicted output of wind and solar power generation at time t, P β，t Indicates the actual output of wind and solar power generation at time t;

[0067] S420, the power balance constraint is:

[0068] S421. The electric power balance constraint formula is:

[0069]

[0070] Where: P I is the grid load during time period t, P grid is the grid power;

[0071] S422. The thermal power balance constraint formula is:

[0072] Q GT-h (t)+Q EB (t)+P HS-dis (t) = P HS-ch (t)+P he (t) (22)

[0073] S430, energy storage balance constraint is:

[0074] -γ ES,D C apES ≤P ES (t)≤γ ES,C C apES (twenty three)

[0075] λ min C apES ≤E ES (t)≤λ max C apES (twenty four)

[0076] E ES (0) = E ES (N T Δt) (25)

[0077] Where: P ES (t) is the power of energy storage in time period t; E ES (t) is the energy storage capacity in time period t; C apES is the total energy storage capacity; γ ES，D is the maximum charging rate of energy storage; γ ES，C is the maximum release rate of stored energy; max is the maximum state of charge of energy storage, λ min is the minimum state of charge for energy storage;

[0078] S440, the decision variable boundary constraints include the output constraints of the gas turbine, electric boiler, fuel cell, power grid, and heat pump. The constraint formula is:

[0079] P ε，min ≤P ε (t)≤P ε，max (26)

[0080] Where: P s (t) is the output of the gas turbine, electric boiler, fuel cell, power grid, and heat pump during period t; P ε，min is the upper limit of the output of gas turbine, electric boiler, fuel cell, power grid and heat pump, P ε，max It is the lower limit of the output of gas turbines, electric boilers, fuel cells, power grids, and heat pumps;

[0081] The ramp constraint formulas for S450, gas turbine, electric boiler, and fuel cell are:

[0082]

[0083] Where: is the ramp-up rate of gas turbine, electric boiler and fuel cell, is the ramp-down rate of gas turbines, electric boilers, and fuel cells.

[0084] Preferably, when the cogeneration system is in operation, high-quality heat energy is recovered by burning natural gas to drive a gas turbine to generate electricity, and the high-temperature waste heat flue gas emitted by the gas turbine is recovered by a bromide refrigeration machine and used for heating.

[0085] Preferably, the fuel cell is a proton exchange membrane fuel cell that uses natural gas as an initial energy source.

[0086] Compared with the prior art, the present invention has the following beneficial effects:

[0087] This paper constructs an optimized scheduling model for an integrated electric and thermal energy system. This model incorporates a scheduling method based on the information gap decision theory (IGDT), creating both a risk-seeking, opportunistic IGDT scheduling model and a risk-averse, robust IGDT scheduling model. This model fully considers factors such as the output characteristics, start / stop states, and ramp limits of generator sets such as wind power, photovoltaic power, combined heat and power, and fuel cells. It couples this model with the charge and discharge states of electric boilers, heat pumps, and electrical and thermal energy storage devices. Using Cplex optimization software, the model calculates the optimal output power of each device within a given scheduling cycle while minimizing the operating costs of the entire system.

[0088] On this basis, considering the output uncertainty of wind and photovoltaic power generation, the information gap decision theory is introduced into the electric-thermal integrated energy system. IGDT theory is an interval optimization method. However, unlike the confidence intervals relied on by robust optimization, the input of IGDT theory is the expected cost or benefit. This allows for better consideration of system economics while ensuring system robustness. Compared with stochastic optimization methods, IGDT theory does not rely on historical data or probability distributions. It can use the tolerable risk level as input to make decisions based on the decision maker's requirements, thus offering greater flexibility.

[0089] Taking into account the uncertainty of wind and solar power output, the present invention divides the scheduling model into a robust model and an opportunity model. The former tends to be conservative, while the latter tends to be risky. The two models correspond to two different value orientations held by the scheduler when dealing with system uncertainty. By comparing the scheduling results, the optimal scheduling of the electric and thermal integrated energy system is achieved. The robust model may lead to a significant increase in system operating costs due to its more cautious attitude. Therefore, on the premise of ensuring that the system has sufficient robustness, effective strategies should also be sought to reduce unnecessary economic burdens. At the same time, the opportunity model tends to accept a higher degree of risk, which makes it likely to suffer a greater impact when facing uncertain factors, so it is necessary to scientifically set the risk tolerance range to deal with risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 This is a flow chart of the IGDT-based electric and thermal integrated energy system modeling of the present invention;

[0091] Figure 2 This is a flow chart of the IGDT-based electrothermal integrated energy system optimization scheduling method of the present invention;

[0092] Figure 3This is a structural diagram of an electric and thermal integrated energy system according to an embodiment of the present invention;

[0093] Figure 4 This is a diagram of photovoltaic and electric heating load prediction output of a summer wind turbine in an embodiment of the present invention;

[0094] Figure 5 This is a diagram of photovoltaic and electric heating load prediction for a winter wind turbine in an embodiment of the present invention. DETAILED DESCRIPTION

[0095] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0096] In the embodiments of the present invention, the IGDT theory can overcome the shortcomings of both robust optimization and stochastic optimization. Like robust optimization, the IGDT theory is an interval optimization method. However, unlike the confidence intervals relied upon by robust optimization, the IGDT theory takes expected costs or benefits as input, thereby better balancing the economic efficiency of the system while ensuring system robustness. Compared with stochastic optimization methods, the IGDT theory does not rely on historical data or probability distributions. It can use the tolerable risk level as input to make decisions based on the decision maker's requirements, thus possessing greater flexibility and adaptability.

[0097] See also Figure 1 The present invention provides an IGDT-based electrothermal integrated energy system optimization scheduling method, comprising the following steps:

[0098] S100, collecting operating parameters of each device in the electric and thermal integrated energy system and energy prices;

[0099] S200, establish an electric and thermal integrated energy system model;

[0100] S300, establishing an objective function that takes into account fuel purchase costs, grid costs, equipment maintenance costs, environmental costs, and heat sales revenue;

[0101] S400: Setting wind and solar output constraints, power balance constraints, energy storage balance constraints, decision variable boundary constraints, and some equipment ramp constraints;

[0102] S500: Establish an optimization model for the electric and thermal integrated energy system and call the Cplex solver to solve it;

[0103] S600, introduce IGDT theory into the electric-thermal integrated energy system and establish an IGDT-based electric-thermal integrated energy system scheduling model;

[0104] S700 builds an opportunity model and a robust model based on IGDT, and calls the Cplex solver for solution.

[0105] The present invention will be described in detail below through specific embodiments in conjunction with the accompanying drawings.

[0106] For a certain place's electric and thermal integrated energy system, the integrated energy system includes wind turbines, solar photovoltaic cells, heat pumps, cogeneration systems, electric boilers, fuel cells, electric storage and heat storage, etc. The total number of time periods in the scheduling cycle is N T = 24, with each dispatch period being Δt = 1 hour. Within each dispatch period, the output and interaction power of each component remain constant, and the electricity trading price is set based on the time-of-use electricity price. Because batteries are used for electrical energy storage and thermal energy storage is performed using thermal storage tanks, and given the long duration of each dispatch period, all exhaust gas from the micro-turbine within a single dispatch period is fed into the lithium bromide absorption chiller. This model was developed in Matlab general-purpose modeling software and solved using the Cplex solver.

[0107] Step 1: Collect the operating parameters of each device in the electric and thermal integrated energy system and the energy price;

[0108] The collected operating parameters include the maximum and minimum output of the equipment, up and down climbing efficiency, and their respective maintenance costs. For details, please refer to Table 1; the energy release, storage power, efficiency and loss rate of the electric thermal energy storage are detailed in Table 2; the collected energy price is shown in Table 3.

[0109] Table 1

[0110]

[0111] Table 2

[0112]

[0113] Table 3

[0114]

[0115] Step 2: Establish an electric and thermal integrated energy system model;

[0116] The electric and thermal integrated energy system of the present invention mainly includes a cogeneration system, a fan, thermal energy storage, photovoltaic, heat pump, electric boiler, fuel cell, electric energy storage and other units.

[0117] a. Electric boiler

[0118] Electric boilers have been widely used in power systems due to their simple installation, flexible operation, and convenient maintenance and replacement. Electric boilers can be used to regulate their operation using an electricity price mechanism. They can be used in conjunction with cogeneration systems to meet thermal load requirements while increasing electricity consumption during off-peak periods. Therefore, electric boilers can be introduced to achieve electric-to-heat conversion and effectively coordinate the peak-to-valley differences between electrical and thermal loads. The output model is:

[0119] Q EB (t) = P EB (t)η ah (1)

[0120] Where: P EB (t) is the power consumption of the electric boiler in time period t; Q EB (t) is the heating power of the electric boiler in time period t; η ah The electric energy consumed by the electric boiler is dispatched as the system power load.

[0121] b. Combined heat and power system

[0122] The key equipment of the cogeneration system mainly consists of a gas turbine and a lithium bromide refrigerator. When the cogeneration system is in operation, it uses the high-quality heat energy recovered from the combustion of natural gas to drive the micro-turbine to generate electricity, while the high-temperature waste heat flue gas emitted by the micro-turbine is recovered by the lithium bromide refrigerator for heating. This paper selects the C65 model micro-turbine produced by Capestone Turbine Company in the United States as the object, and assumes that the impact of external environmental changes on power generation and fuel combustion efficiency during operation can be ignored. Its thermoelectric relationship mathematical model is:

[0123]

[0124] Where: Q GT (t) represents the waste heat generated by the gas turbine during the period t, P GT (t) represents the amount of gas generated by the gas turbine during the period t, η GT (t) represents the power generation efficiency of the gas turbine in the time period t; η L is the heat loss rate;

[0125] Q GT-h (t) = Q GT (t)η h C OPh (3)

[0126] Where: Q GT-h (t) is the heating capacity of the lithium bromide refrigerator during time period t; C OPh is the thermal coefficient of the lithium bromide refrigerator, η h is the flue gas recovery rate of lithium bromide refrigerator;

[0127] The fuel cost of the micro-turbine in period t is:

[0128]

[0129] Where: Δt is the unit scheduling time; C GT (t) is the cost of fuel consumed by the gas turbine during period t; C CH4The price of natural gas is set to 2.5 yuan / m 3 ;L LHV is the calorific value of natural gas, which is 9.7kW·h / m 3 .

[0130] c. Fuel Cell

[0131] Fuel cells can convert the chemical energy contained in hydrogen-containing fuels such as natural gas and methanol and oxygen into electrical energy in an efficient and environmentally friendly manner. This invention focuses on proton exchange membrane fuel cells, which use natural gas as the initial energy source and exhibit high energy conversion efficiency. Given that in the system architecture currently under discussion, fuel cells mainly play the role of a micro-power source for power allocation, their waste heat reuse is not included in the scope of consideration of this invention. The fuel cost and power output characteristics of the fuel cell in time period t are:

[0132]

[0133] Where: C FC (t) is the fuel cost of the fuel cell in time period t; P FC (t) is the power generation of the fuel cell, η FC (t) is the conversion efficiency of the fuel cell;

[0134] Microturbines, electric boilers, and fuel cells are typically classified as controllable units due to their compact size, ease of operation, fast startup, and flexible power regulation. The operating conditions of these units are limited by their power output level and ramp speed.

[0135] d. Heat pump

[0136] A heat pump is a device that uses a small amount of electricity or fuel as power to convert a large amount of low-quality heat at low temperatures into high-quality heat that can be used for heating. Its performance is measured by the heating coefficient C OPH It is generally defined as the heat energy obtained by the user / the electric energy or fuel energy consumed by the heat pump. In this invention, the heating coefficient C OPH Take 4.4.

[0137] e. Energy storage system

[0138] Energy storage systems can separate energy production and consumption in the time dimension, thereby achieving flexible allocation of energy between different time periods. In the power grid, in order to balance the difference between "source-load", in the electric-thermal joint scheduling model, the energy storage system includes two forms of power storage and thermal energy storage. Especially in the process of economic scheduling of the power grid, power storage can effectively smooth the load curve and reduce operating costs. Depending on the application requirements, power storage technology can be further divided into energy-type devices that focus on long-term energy storage, such as batteries, and power-type devices suitable for fast charging and discharging scenarios, such as supercapacitors. When the scheduling cycle is long, it is recommended to use an energy-type energy storage solution; when a fast response is required, it is more inclined to choose a power-type power storage method. Generally speaking, the capacity of the power storage system and its charging and discharging capabilities must satisfy the following relationship:

[0139]

[0140] Where: E EES (t) is the energy storage capacity in time period t; τ is the self-discharge rate of energy storage; P EES-ch (t), P EES-dis (t) are the charging and discharging power of the power storage system in time period t; η sch ,η sdis The efficiency of the electric energy storage system;

[0141] Due to the differences in scale and peak time between thermal load and electrical load, when encountering a situation where heat demand is low and electricity demand is high, some cogeneration systems may not be able to achieve optimal operation due to insufficient heat supply; conversely, when heat demand is high but electricity demand is relatively low, there may be an oversupply of electricity. This excess electricity is not only difficult to connect to the grid, but even if it can be connected to the grid, its cost is often too high, resulting in reduced overall scheduling efficiency and operating losses. By introducing thermal energy storage technology, the distribution of heat demand in different time periods can be effectively adjusted, thereby alleviating the scheduling difficulties caused by the imbalance of electricity and heat ratio, and realizing the effective integration and management of electricity and thermal energy resources. Thermal energy storage systems are usually composed of large heat storage tanks, heat storage tanks, and heat storage electric boilers. The characteristics of these components can be specifically described by multiple indicators such as capacity, input and output capabilities, and thermal conversion efficiency. Their working mechanism can be expressed by a dynamic mathematical model:

[0142]

[0143] Where: H HS (t) is the thermal energy storage capacity in time period t; μ is the thermal energy dissipation loss rate; P HS-ch (t), P HS-dis (t) are the heat absorption and release power of the thermal energy storage system in time period t; ηhch ,η hdis are the efficiencies of the thermal energy storage system in time period t.

[0144] Step 3: Establish an objective function that takes into account fuel purchase costs, grid costs, equipment maintenance costs, environmental costs, and heat sales revenue;

[0145] By integrating thermal energy storage technology to optimize the objective function, the combined heat and power system no longer needs to strictly match heat load fluctuations when providing heat, thereby enhancing the system's dispatch flexibility. From the perspective of power supply, to balance environmental protection and dispatch management needs, priority is given to renewable energy sources such as wind and solar power, and their utilization is maximized, while a maximum power point tracking strategy is implemented. This reduces the daily operating costs of the combined heat and power system, aiming to minimize the system's daily operating costs.

[0146] Establish the objective function of fuel cost, grid cost, equipment maintenance cost, environmental cost and heat sales revenue:

[0147]

[0148] in:

[0149]

[0150] C HE (t) = C he P he (t) (13)

[0151] Where: F is the total operating cost; N T is the total number of scheduling periods; C FU (t) is the fuel cost in period t; C EX (t) is the grid cost in period t; C ME (t) is the maintenance cost in period t; C CE (t) is the environmental cost in period t; C HE (t) is the heat sales revenue in period t; P ex (t) is the interaction power between the system and the large power grid during period t; C rs (t) is the electricity price during period t, C rb (t) is the electricity purchase price during period t; N M is the total number of units in the network; C Mi is the unit maintenance cost of unit i; P i (t) is the output of unit i during period t; P P is the conversion coefficient of electricity; h GT is the heating efficiency; C he P is the unit price of heat; he (t) is the heat load in the network during period t;

[0152] Step 4: Set wind and solar output constraints, power balance constraints, energy storage balance constraints, decision variable boundary constraints, and some equipment ramp constraints;

[0153] The constraints are designed with the goal of minimizing the daily dispatching cost of the electric and thermal integrated energy system. The constraints that minimize the total daily operating cost of the system include: wind and solar output constraints, power balance constraints, energy storage balance constraints, decision variable boundary constraints, and some equipment ramping constraints.

[0154] a. Wind and solar power output constraints

[0155] In terms of clean energy supply, the main focus is on wind and solar energy. Due to the instability of wind and photovoltaic power generation and the limitations of the power transmission network, the actual amount of wind power that can be absorbed and utilized by the system is often lower than expected:

[0156] 0.8P β,f ≤P β,t ≤P β,f (14)

[0157] Where: P β，f represents the predicted output of wind and solar power generation at time t, P β，t Indicates the actual output of wind and solar power generation at time t;

[0158] Constructing electric power balance constraints:

[0159]

[0160] Where: P l is the grid load during time period t, P grid is the grid power;

[0161] Construct thermal power balance constraints:

[0162] Q GT-h (t)+Q EB (t)+P HS-dis (t) = P HS-ch (t)+P he (t) (16)

[0163] b. Energy storage balance constraints

[0164] -γ ES,D C apES ≤P ES (t)≤γ ES,C C apES (17)

[0165] λ min C apES ≤E ES (t)≤λmax C apES (18)

[0166] E ES (0) = E ES (N T Δt) (19)

[0167] Where: P ES (t) is the power of energy storage in time period t; E ES (t) is the energy storage capacity in time period t; C apES is the total energy storage capacity; γ ES，D is the maximum charging rate of energy storage; γ ES，C is the maximum release rate of stored energy; max is the maximum state of charge of energy storage, λ min is the minimum state of charge for energy storage;

[0168] c. The decision variable boundary constraints include the output constraints of the gas turbine, electric boiler, fuel cell, power grid, and heat pump. The constraint formula is:

[0169] P ε，min ≤P ε (t)≤P ε，max (20)

[0170] Where: P ε (t) is the output of the gas turbine, electric boiler, fuel cell, power grid, and heat pump during period t; P ε，min is the upper limit of the output of gas turbine, electric boiler, fuel cell, power grid and heat pump, P ε，max It is the lower limit of the output of gas turbines, electric boilers, fuel cells, power grids, and heat pumps.

[0171] d. Some equipment climbing constraints include gas turbines, electric boilers, and fuel cells:

[0172]

[0173] Where: is the ramp-up rate of gas turbine, electric boiler and fuel cell, is the ramp-down rate of gas turbines, electric boilers, and fuel cells.

[0174] Step 5: Establish an optimization model for the electric and thermal integrated energy system and use the Cplex solver to solve it.

[0175] The solution algorithm includes nonlinear components and running or stopping state variables of some devices in the objective function. The standard solution form of this problem is:

[0176]

[0177] Where the optimization variable x represents the power output of a specific device; the optimization variable y reflects the operating state of the fuel cell, i.e., whether it is on or off. The equality constraints are primarily based on the system's internal energy balance and the energy variation patterns of the energy storage device; while the inequality constraints define the operating ranges of each component. To enhance the speed and efficiency of the model solution, the problem is formulated as a mixed-integer linear program and solved using the Cplex solver.

[0178] Step 6: Introduce the information gap decision theory into the electric-thermal integrated energy system and establish an electric-thermal integrated energy system scheduling model based on IGDT;

[0179] The scheduling model of the above-mentioned deterministic electric and thermal integrated energy system is based on the deterministic values of wind energy and solar energy for scheduling. During actual operation, due to the uncertainty of the output of wind energy and solar energy, the actual operation results and the predicted values deviate greatly, which will cause large economic losses. Therefore, the uncertainty of wind energy and solar energy is taken into account in the scheduling optimization process to provide decision-making guidance for the scheduling plan. The present invention uses the IGDT method to optimize and process the uncertainty of wind energy and solar energy. IGDT is a decision-making method for processing non-probabilistic uncertainty. This method is suitable for those uncertain situations that are difficult to express through probability or specific scenarios. Compared with traditional robust optimization technology, the information gap decision theory not only focuses on the stability of the system during the optimization process, but also takes into account its economic benefits, so that it can more effectively balance the safety and cost-effectiveness of the system operation. In this framework, the input of uncertainty is expressed as a fuzzy set. This uncertainty is described by the uncertainty set under the non-probabilistic model, such as the envelope model, fractional uncertainty model and ellipsoid model to describe uncertainty.

[0180] The fractional uncertainty model is selected as the research object, which is specifically expressed as follows:

[0181]

[0182] Among them: α represents the fluctuation range of load demand, that is, uncertainty; represents the predicted value of load demand;

[0183] That is: Uncertain factors (actual value of load demand ) has an upper limit of The lower limit is

[0184]

[0185] When the deterministic dispatch model is set to α = 0, it means that the load forecast is completely accurate. The optimal solution of the objective function obtained in this case is considered the reference standard. Different uncertainties and decision makers' different risk preferences will have a significant impact on the system's ultimate profitability and dispatch strategy.

[0186] The present invention divides the scheduling strategies of the system under uncertainty conditions into two categories: opportunity model (OM) and robust model (RM). The opportunity model tends to be a risky strategy, while the robust model tends to be a conservative strategy.

[0187] Step 7: Build opportunistic and robust models based on IGDT and call the Cplex solver for solution.

[0188] 1) Opportunistic IGDT scheduling model based on risk pursuit

[0189] The Opportunistic IGDT model reveals that the uncertainty of wind and solar generation can actually be converted into economic benefits during the dispatch process. By implementing proactive risk management strategies, the goal is to ensure that daily system revenue exceeds expectations while minimizing the adverse effects of uncertainty and achieving better economic benefits within a given load variation range.

[0190] The IGDT scheduling model based on risk pursuit is a two-level planning model, and its formula is:

[0191]

[0192] Where: ε m is the chance deviation coefficient, DP OM is the expected profit threshold, DP0 refers to the best performance index that can be achieved by an exact electric and thermal system scheduling plan established based on the prediction of uncertain factors;

[0193] This paper proposes a two-level programming model, in which the lower optimization objective is to maximize economic benefits by adjusting the risk preference strategy while ensuring that the real-time power demand remains within a preset uncertainty range; while the upper level is committed to reducing the uncertainty level in load forecasting. In this model framework, ε m Define the chance deviation coefficient, DP OM represents the expected profit threshold, and DP0 refers to the best performance indicator that can be achieved by an exact electric and thermal system scheduling solution established based on the prediction of uncertain factors, also known as a risk-neutral model. Given that the proposed solution has a certain degree of opportunistic characteristics, the target profit rate set by the lower model exceeds DP0, ε mThe value of is [0, 1). There is a direct positive relationship between the specific value of this parameter and the decision maker's risk tolerance. Compared with those methods that operate solely based on electricity price fluctuations, scheduling plans that take load changes into consideration may result in a reduction in total daily revenue. When taking a proactive approach to risk, if the actual electricity consumption is close to the lowest point in the expected range, the daily profit calculated by the lower model can be maximized. When the real-time load demand is the lower limit of the uncertainty set When , the daily profit of the lower model can reach the maximum value. Therefore, the two-level programming model can be transformed into the following single-level optimization model:

[0194]

[0195] 2) Robust IGDT scheduling model based on risk aversion

[0196] The robust IGDT model shows that load uncertainty will have a huge impact on the system scheduling results. The risk-avoiding scheduling decision is to ensure that the objective function of the electric and thermal integrated system obtains the optimal robustness while meeting the basic economic requirements, that is, to determine the limit value of the uncertain parameter fluctuation so that the daily profit of the electric and thermal integrated system can reach the expected value.

[0197] The robust IGDT scheduling model based on risk aversion is formulated as follows:

[0198]

[0199] Where: m is the robust bias factor, DP RM is the robust return threshold.

[0200] Similarly, δ m The value range of is [0, 1) and is proportional to the risk aversion of the decision. At the same time, for this formula, when the real-time load demand is When , which is the upper bound of the uncertainty set, the lower model can obtain the minimum value. Similarly, formula (10) can be transformed into:

[0201]

[0202] Based on the optimized scheduling model of the electric and thermal integrated energy system built on the IGDT framework, the system's operating data and equipment parameters are substituted into the above model and solved using Cplex. The obtained optimal scheduling results are analyzed.

[0203] This embodiment selects the following three scenarios for comparative analysis.

[0204] Scenario 1: Select the daily minimum cost of a typical day in summer and winter in an electric-thermal integrated energy system without the introduction of IGDT.

[0205] Scenario 2: Robust model cost in the electric-thermal integrated energy system introduced by IGDT on a typical day in summer and winter.

[0206] Scenario 3: Opportunity model cost of the electric-thermal integrated energy system introduced by IGDT on a typical day in summer and winter.

[0207] Table 4 shows the scheduling results. It shows that the typical daily cost in summer is X1 = 1150.01 yuan, and the typical daily cost in winter is X2 = 1746.84 yuan. After introducing the IGDT, the corresponding summer robust cost is 1311.01 yuan, and the opportunity cost is 989.01 yuan. In winter, the robust cost and opportunity cost are 1991.39 yuan and 1502.28 yuan, respectively.

[0208] Table 4

[0209]

[0210] From this, we can see that from the perspective of the opportunity model, as the deviation coefficient increases, the response uncertainty α increases, and the scheduling cost of the electric heating system decreases. This is because in the opportunity model, a larger α means that the decision maker is more optimistic about the potential benefits of response uncertainty, so the scheduling cost of the electric heating system decreases. From the perspective of the robust model, as the deviation coefficient increases, the uncertainty α value increases, and the total scheduling cost of the electric heating system increases accordingly. This is because in the robust model, a larger α means that the decision maker is more pessimistic about response uncertainty, which leads to an increase in the total cost of the electric heating system, a more conservative decision plan, and enhanced robustness.

[0211] In summary, an optimized scheduling model based on information gap decision theory (IGDT) was used to develop electricity usage strategies for both risk avoidance and opportunity pursuit scenarios. This approach not only effectively addresses the issue of unstable output power from renewable energy generation units, but also achieves a balance between risk management and capital investment based on the decision maker's attitude towards risk. This invention further demonstrates the application value of IGDT in practical energy management, providing dispatchers with a powerful tool to address the challenges posed by output fluctuations from renewable energy sources such as wind power generation, and to deal with the various uncertainties and associated risks encountered during system operation.

[0212] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. An optimization scheduling method for an electric-thermal integrated energy system based on IGDT theory, characterized by: The following steps are involved: S100, collects operating parameters of various equipment and energy prices of the electric and thermal integrated energy system; Various types of equipment in the electric and thermal integrated energy system include cogeneration systems, fans, thermal energy storage, photovoltaics, heat pumps, electric boilers, fuel cells, and electric energy storage equipment; S200, establish an electric and thermal integrated energy system model; S300: Establish an objective function for fuel cost, grid cost, equipment maintenance cost, environmental cost, and heat sales revenue: in C HE (t)=C he P he (t) (6) Where: F is the total operating cost; N T is the total number of scheduling periods; C FU (t) is the fuel cost in period t; C EX (t) is the grid cost in period t; C ME (t) is the maintenance cost in period t; C CE (t) is the environmental cost in period t; C HE (t) is the heat sales revenue in period t; P ex (t) is the interaction power between the system and the large power grid during period t; C rs (t) is the electricity price during period t, C rb (t) is the electricity purchase price during period t; N M is the total number of units in the network; C Mi is the unit maintenance cost of unit i; P i (t) is the output of unit i during period t; P P is the conversion coefficient of electricity; h GT is the heating efficiency; C he P is the unit price of heat; he (t) is the heat load in the network during period t; S400: Establishing constraints, including wind and solar output constraints, power balance constraints, energy storage balance constraints, decision variable boundary constraints, and some equipment ramping constraints; S500: Establish an optimization model for the electric and thermal integrated energy system and call the Cplex solver to solve it. The solution algorithm includes nonlinear components and running or stopping state variables of devices such as gas turbines, electric boilers, fuel cells, fans, photovoltaics, heat pumps, and power grids in the objective function. The solution formula is: min f(x,y) Where: The optimization variable x includes the power output of the gas turbine, electric boiler, fuel cell, heat pump, and other devices; the optimization variable y reflects the operating status of the fuel unit in operation or shutdown; the equality constraints are mainly based on the energy balance relationship within the system and the energy change law of the energy storage device; and the inequality constraints define the operating range of the equipment; S600. Establish an electric-thermal integrated energy system scheduling model based on IGDT theory and select a fractional uncertainty model. The specific formula is: Where: α represents the fluctuation range of load demand, that is, uncertainty; represents the predicted value of load demand; Actual value of load demand with uncertain factors The upper limit of the collection is The lower limit is S700, builds opportunity model and robust model based on IGDT theory, and calls Cplex solver for solution; S710. The IGDT scheduling model based on risk pursuit is a two-level planning model, and its formula is: object:mina Where: ε m is the chance deviation coefficient, DP OM is the expected profit threshold, DP0 refers to the best performance index that can be achieved by an exact electric and thermal system scheduling plan established based on the prediction of uncertain factors; S720. A robust IGDT scheduling model based on risk aversion is formulated as follows: object:minα Where: m is the robust bias factor, DP RM is the robust return threshold.

2. The method for optimizing and scheduling an electric-thermal integrated energy system based on IGDT theory according to claim 1, characterized in that: In the formula (9), ε m The value of is [0, 1), ε m There is a direct positive relationship between the specific value of and the decision maker's risk tolerance; compared with the method of operating solely based on electricity price fluctuations, the scheduling plan that takes load changes into consideration will lead to a decrease in total daily revenue; the actual electricity consumption is close to the lowest point in the expected range, which maximizes the daily profit calculated by the lower model; when the real-time load demand is the lower limit of the uncertainty set When , the daily profit of the lower model can reach the maximum value; the two-level programming model is transformed into the following single-level optimization model: object:mina In formula (10), δ m The value range of δ is [0, 1), m The specific value of is proportional to the decision maker's risk aversion; when the real-time load demand is When , which is the upper bound of the uncertainty set, the robust lower layer model can obtain the minimum value, and formula (25) is transformed into a single-layer optimization model: object:minα 3. The method for optimizing and scheduling an electric-thermal integrated energy system based on IGDT theory according to claim 1, characterized in that: The electric-thermal integrated energy system includes an electric power system and a thermal system; the electric power system includes a wind farm, a photovoltaic power station, a fuel cell, an electric boiler, a cogeneration system and a heat pump; the thermal system includes an electric boiler, a heat pump and a cogeneration system; the operating data of the electric-thermal integrated energy system includes the electric load, the thermal load power forecast, and the load forecast of wind and solar output; the equipment parameters of the electric-thermal integrated energy system include the rated power and conversion efficiency, up and down climbing efficiency and maintenance cost of the output equipment, the energy release, energy storage power, total capacity, maintenance cost, efficiency and loss rate of the energy storage equipment.

4. The method for optimizing and scheduling an electric-thermal integrated energy system based on IGDT theory according to claim 1, characterized in that: The output models of each device in the electric and thermal integrated energy system in S200 include: S210. The operation of electric boilers is regulated by the electricity price mechanism. Electric boilers work with the cogeneration system to increase electricity consumption during off-peak hours while meeting the heat load demand, thus coordinating the peak-valley differences between the electricity load and the heat load. The output model of the electric boiler is: Q EB (t)=P EB (t)η ah (13) Where: P EB (t) is the power consumption of the electric boiler in time period t; Q EB (t) is the heating power of the electric boiler in time period t; η ah The electric energy consumed by the electric boiler is dispatched as the system power load. S220, cogeneration system including gas turbine and lithium bromide chiller; S221. The mathematical model of the thermoelectric relationship of a gas turbine is: Where: Q GT (t) represents the waste heat generated by the gas turbine during the period t, P GT (t) represents the amount of gas generated by the gas turbine during the period t, η GT (t) represents the power generation efficiency of the gas turbine in the time period t; η L is the heat loss rate; S222. The mathematical model of the thermoelectric relationship of lithium bromide refrigerator is: Q GT-h (t)=Q GT (t)η h C OPh (15) Where: Q GT-h (t) is the heating capacity of the lithium bromide refrigerator during time period t; C OPh is the thermal coefficient of the lithium bromide refrigerator, η h is the flue gas recovery rate of lithium bromide refrigerator; S223. The fuel cost of the gas turbine in period t is: Where: Δt is the unit scheduling time; C GT (t) is the cost of fuel consumed by the gas turbine during period t; C CH4 is the price of natural gas; L LHV is the calorific value of natural gas; S230, the fuel cost and electric power output characteristics of the fuel cell in time period t are: Where: C FC (t) is the fuel cost of the fuel cell in time period t; P FC (t) is the power generation of the fuel cell, η FC (t) is the conversion efficiency of the fuel cell; S240, heating coefficient C of heat pump OPH The heat energy obtained by the user / the electricity or fuel energy consumed by the heat pump, C OPH =4.4; S250, the energy storage system includes an electric energy storage system and a thermal energy storage system; S251. The capacity of the power storage system and its charge and discharge capabilities must satisfy the following relationship: Where: E EES (t) is the capacity of the power energy storage system in time period t; τ is the self-discharge rate of the power energy storage system; P EES-ch (t), P EES-dis (t) are the charging and discharging power of the power storage system in time period t; η sch ,η sdis The efficiency of the electric energy storage system; S252. The dynamic mathematical model of the thermal energy storage system is: Where: H HS (t) is the thermal energy storage capacity in time period t; μ is the thermal energy dissipation loss rate; P HS-ch (t), P HS-dis (t) are the heat absorption and release power of the thermal energy storage system in time period t; η hch ,η hdi s are the efficiencies of the thermal energy storage system in time period t.

5. The method for optimizing and scheduling an electric-thermal integrated energy system based on IGDT theory according to claim 1, characterized in that: The constraints include: S410, wind and solar output constraints are: Clean energy supplies include wind and solar power, and the amount of wind power that can be absorbed and utilized by the system is lower than expected: 0.8P β,f ≤P β,t ≤P β,f (20) Where: P β，f represents the predicted output of wind and solar power generation at time t, P β，t Indicates the actual output of wind and solar power generation at time t; S420, the power balance constraint is: S421. The electric power balance constraint formula is: Where: P I is the grid load during time period t, P grid is the grid power; S422. The thermal power balance constraint formula is: Q GT-h (t)+Q EB (t)+P HS-dis (t)=P HS-ch (t)+P he (t) (22) S430, energy storage balance constraint is: -γ ES,D C apES ≤P ES (t)≤γ ES,C C apES (23) λ min C apES ≤E ES (t)≤λ max C apES (24) AND ES (0)=E ES (N T Δt) (25) Where: P ES (t) is the power of energy storage in time period t; E ES (t) is the energy storage capacity in time period t; C apES is the total energy storage capacity; γ ES，D is the maximum charging rate of energy storage; γ ES，C is the maximum release rate of stored energy; max is the maximum state of charge of energy storage, λ min is the minimum state of charge for energy storage; S440, the decision variable boundary constraints include the output constraints of the gas turbine, electric boiler, fuel cell, power grid, and heat pump. The constraint formula is: P ε，min ≤P ε (t)≤P ε，max (26) Where: P ε (t) is the output of the gas turbine, electric boiler, fuel cell, power grid, and heat pump during period t; P ε，min is the upper limit of the output of gas turbine, electric boiler, fuel cell, power grid and heat pump, P ε，ma x is the lower limit of the output of the gas turbine, electric boiler, fuel cell, power grid, and heat pump; The ramp constraint formulas for S450, gas turbine, electric boiler, and fuel cell are: Where: is the ramp-up rate of gas turbine, electric boiler and fuel cell, is the ramp-down rate of gas turbines, electric boilers, and fuel cells.

6. The method for optimizing and scheduling an electric-thermal integrated energy system based on IGDT theory according to claim 1, characterized in that: When the cogeneration system is in operation, high-quality heat energy is recovered by burning natural gas to drive a gas turbine to generate electricity, and the high-temperature waste heat flue gas emitted by the gas turbine is recovered by the bromide refrigeration machine for heating.

7. The method for optimizing and scheduling an electric-thermal integrated energy system based on IGDT theory according to claim 1, characterized in that: The fuel cell is a proton exchange membrane fuel cell that uses natural gas as an initial energy source.

Images