Island integrated energy system scheduling method considering uncertainty and multi-energy multi-type demand response

Abstract

The invention provides an island integrated energy system scheduling method considering uncertainty and multi-energy multi-type demand response, and belongs to the technical field of integrated energy system scheduling optimization. The problem of limitation of the existing research aiming at the island energy system in processing user demand response and unit output is solved; comprising the following steps: constructing an island integrated energy system and a unit output model thereof; the double-layer coordinated optimization scheduling model comprises an upper-layer optimization model and a lower-layer optimization model, and the upper-layer optimization model takes the maximum difference between the user load shedding penalty costs before and after optimization as the target; the lower-layer optimization model optimizes the output strategy of each time period by taking net income maximization as a target; solving the double-layer coordinated optimization scheduling model by adopting an improved self-adaptive genetic algorithm; the method is applied to optimal scheduling of the island integrated energy system.

Description

Technical Field

[0001] The present invention provides an isolated island integrated energy system scheduling method taking into account uncertainty and multi-energy and multi-type demand response, belonging to the technical field of integrated energy system scheduling optimization. Background Art

[0002] With the development of social economy, users' energy demand is growing, but remote areas are isolated due to resource and geographical limitations, facing the problem of insufficient energy supply. The isolated microgrid can not only operate independently to provide power to the load when the main grid fails or is under maintenance, improving the reliability of power supply, but also maximize the utilization of renewable energy, optimize energy configuration, and reduce operating costs, providing an effective technical approach to solve the problem of insufficient energy supply in remote areas. However, with the increase in the proportion of renewable energy, the isolated microgrid faces the problem of unstable power supply caused by the randomness and intermittent behavior of renewable energy, and the isolated integrated energy system provides a more comprehensive and flexible solution to these challenges by integrating multiple energy forms and energy storage technologies. Therefore, studying and optimizing the dispatching strategy of the isolated integrated energy system can not only improve energy utilization efficiency and ensure economy, but also enhance the stability of the system and meet the growing diversified energy needs of users.

[0003] At present, the research on isolated energy systems mainly focuses on optimization configuration, operation mode and multi-objective planning optimization. These studies provide theoretical basis and technical support for the efficient operation and management of isolated energy systems, such as improving the system's ability to cope with uncertainty by considering economy, flexibility and demand response, and realizing flexible energy scheduling and stable economic operation of microgrids through multi-agent systems. However, there are still some limitations in current research, especially in the lack of ability to handle user demand response and optimize unit output. Therefore, in order to further improve the operating efficiency and economy of isolated energy systems while ensuring the reliability and sustainability of the system, it is urgent to expand the isolated microgrid to the isolated integrated energy system. Summary of the invention

[0004] In order to solve the limitations of existing research on isolated energy systems in dealing with user demand response and unit output, the present invention proposes an isolated integrated energy system scheduling method that takes into account uncertainty and multi-energy and multi-type demand response.

[0005] The technical solution adopted by the present invention is: a method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response, comprising the following steps:

[0006] S1: Construct an isolated island integrated energy system and its unit output model, wherein the isolated island integrated energy system consists of a clean energy power supply unit, a solar thermal power station, a cogeneration unit and a ground source heat pump, wherein the clean energy power supply unit and the solar thermal power station are uncertain units, the cogeneration unit and the ground source heat pump are conventional units, and the unit output model includes an uncertain unit output model and a conventional unit output model;

[0007] S2: Construct a two-layer coordinated optimization scheduling model for an isolated integrated energy system: The two-layer coordinated optimization scheduling model includes an upper optimization model and a lower optimization model, wherein the upper optimization model aims to maximize the difference between the user load shedding penalty cost before and after optimization, and makes a load optimization strategy based on the prediction curve of user load and distributed generation power, combined with the user demand elasticity coefficient and the maximum load reduction controllable load information, and transmits the result to the lower optimization model. The lower optimization model optimizes the output strategy of each time period with the goal of maximizing net profit, and feeds back the running plan to the upper optimization model;

[0008] S3: An improved adaptive genetic algorithm is used to solve the two-layer coordinated optimization scheduling model.

[0009] Furthermore, the clean energy power supply unit includes a wind power unit and a photovoltaic power unit, and the uncertain unit output model includes a wind power unit output model and its constraints, a photovoltaic power unit output model and its constraints, and a solar thermal power station output model and its constraints;

[0010] The conventional unit output model includes the combined heat and power unit output model and its constraints, and the ground source heat pump output model and its constraints.

[0011] Furthermore, step S2 specifically includes:

[0012] S2.1: Scenario generation based on super Latin hypercube and scenario reduction method with synchronous back generation to deal with uncertainty in clean energy output;

[0013] S2.2: Constructing the upper optimization model: Taking multi-energy and multi-type demand response as the upper optimization decision variable, taking the difference of system load shedding penalty cost before and after optimization as the upper optimization operation target, constructing the upper optimization model;

[0014] S2.3: Construct the lower-level optimization model: Take the output of each unit in the isolated integrated energy system as the lower-level optimization decision variable, and take the maximization of the net profit on the internal energy supply side of the isolated integrated energy system as the lower-level optimization operation goal to construct the lower-level optimization model.

[0015] Furthermore, the implementation process of step S2.1 is as follows:

[0016] Step 2.1.1: Determine the number of sampling intervals;

[0017] Step 2.1.2: Randomly select sample values ​​in each sampling interval and calculate the probability of each sample value;

[0018] Step 2.1.3: Generate sampling matrix: Use hypercube sampling to generate a hypercube for each random variable, where each dimension represents a variable and each variable randomly selects sample values ​​in each interval;

[0019] Step 2.1.4: Check the number of sampling intervals: Check whether each sample in the sampling matrix meets the required number of sampling intervals. If so, stop sampling, count the scene data, and perform statistics and analysis on the existing scene data; if not, repeat step 2.1.2;

[0020] Step 2.1.5: Determine the number of scenes to be retained;

[0021] Step 2.1.6: Calculate scene distance and perform scene deletion and scene retention;

[0022] Step 2.1.7: Add the probability of the deleted scene to the retained scene;

[0023] Step 2.1.8: Check whether the number of scenes after deletion is equal to the number of scenes that need to be retained, and confirm whether the number of scenes currently retained has reached the preset target number. If so, stop sampling and count the retained data. If not, repeat step 2.1.6.

[0024] Furthermore, the objective function of the upper optimization model is expressed as follows:

[0025] ;

[0026] Where: is the upper optimization objective function, which represents the difference in load shedding cost before and after the upper optimization. and are the unit penalty costs of load shedding of the i-th type of load before and after the upper optimization, and They are the load shedding amounts of the i-th type of load in period t under w scenarios before and after the upper-layer optimization, I is the number of scenarios, and T is the total period.

[0027] Furthermore, the multi-energy multi-type demand response includes price-based and incentive-based multi-energy demand response models. The price-based multi-energy demand response model re-formulates the peak and valley time-of-use electricity and heat prices within the isolated integrated energy system based on the user demand elasticity matrix to guide users to change their energy usage methods. Its expression is as follows:

[0028] ;

[0029] Where: , are the load demands of the user class i before and after the implementation of the peak and valley time-of-use electricity and heat prices in period t, is the elasticity coefficient of the i-th type of load, j is the category of price change factors that affect the demand for this type of load, and are the original price and price change of the i-th type of load in period t respectively;

[0030] The incentive-based multi-energy demand response model is a load increase and decrease agreement signed between the operator of the isolated integrated energy system and the user, and its expression is as follows:

[0031] ;

[0032] Where: is the incentive demand response cost of the i-th load in period t, is the load reduction amount of the i-th type of load in period t, It is a trapezoidal price, which means the unit compensation price of the i-th type of load in the m-segment interval where the load can be reduced.

[0033] Furthermore, the constraints that the upper-level optimization model needs to meet include:

[0034] Supply and demand balance constraints: including electricity load supply and demand balance constraints, heat load supply and demand balance constraints, and gas load supply and demand balance constraints;

[0035] Demand response constraints: including load variation constraints, peak and valley power constraints, and user energy cost constraints.

[0036] Furthermore, the objective function of the lower optimization model is expressed as follows:

[0037] ;

[0038] Where: is the lower optimization objective function, which represents the maximization of the net benefit of the isolated island integrated energy system. The net benefit of the island integrated energy system is , are the electricity and heat benefits of the isolated island integrated energy system under scenario w, is the depreciation cost of the unit under scenario w, , , , They are the unit operation and maintenance cost, load shedding cost, incentive-based demand response cost and gas purchase cost under scenario w.

[0039] Furthermore, the constraints that the lower optimization model needs to satisfy are the constraints of each unit in the isolated integrated energy system, including:

[0040] Clean energy power supply unit constraints: including unit installed capacity constraints and output power constraints;

[0041] Constraints of CSP power station units: including constraints of the heat storage system in the heat storage link and constraints of the power conversion module in the power generation link. The constraints of the heat storage system include constraints of heat storage capacity, heat charging and discharging power, and charging and discharging power state. The constraints of the power conversion module include constraints of the power conversion module operation state, start and stop time constraints, electrical output power constraints, upward and downward regulation rate constraints, and input thermal energy constraints.

[0042] Cogeneration unit constraints: including ground source heat pump output constraints, heat exchanger output constraints, and waste heat boiler capacity constraints;

[0043] Ground source heat pump constraints: including ground source heat pump installed capacity constraints and energy supply power constraints.

[0044] Furthermore, the steps of using the improved adaptive genetic algorithm to solve the two-layer coordinated optimization scheduling model include:

[0045] Step 3.1: Input the raw data of the isolated integrated energy system;

[0046] Step 3.2: Initialize the iteration number of the two-layer coordinated optimization scheduling model, the upper optimization scenario, and the lower optimization scenario to 0;

[0047] Step 3.3: Solving the upper optimization model: After randomly generating the initial population, calculate the difference in the user load shedding penalty cost of the unit operation optimization plan corresponding to each individual, and calculate the individual fitness; then select the operation through the roulette method, adaptive mutation, crossover and elite retention, and generate a new population; repeat the above process until the maximum number of iterations of the upper optimization is reached, and output the optimal solution of the upper optimization model;

[0048] Step 3.4: Determine whether the two-layer coordinated optimization scheduling model has reached the maximum number of iterations. If not, pass the upper layer optimization result to the lower layer to solve the lower layer optimization model;

[0049] Step 3.5: Solving the lower-level optimization model: After randomly generating the initial population, solve the lower-level optimization, calculate the net benefit value of the internal energy supply side of the isolated integrated energy system, and calculate the individual fitness; then select the operation through the roulette method, adaptive mutation, crossover and elite retention, and generate a new population; repeat the above process until the maximum number of iterations of the lower-level optimization is reached, output the optimal solution of the lower-level optimization model, feed back the lower-level optimization results to the upper layer, and solve the upper-level optimization model;

[0050] Step 3.6: Until the two-layer coordinated optimization scheduling model reaches the maximum number of iterations, the optimal result is output.

[0051] The beneficial effects of the present invention compared to the prior art are as follows: the present invention enhances the coordination of the operation of the isolated island integrated energy system by establishing a linkage relationship of mutual transmission and mutual influence between demand response optimization of user load demand and unit output optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The present invention will be further described below in conjunction with the accompanying drawings:

[0053] Figure 1 It is a structural diagram of the isolated island integrated energy system of the present invention;

[0054] Figure 2 This is a schematic diagram of the energy flow inside a CSP plant;

[0055] Figure 3 A flowchart for dealing with uncertainty in clean energy;

[0056] Figure 4 This is a schematic diagram of the incentive-based multi-energy demand response ladder price;

[0057] Figure 5 Processing flow chart of the improved adaptive genetic algorithm. DETAILED DESCRIPTION

[0058] like Figures 1 to 5 As shown, the present invention provides an island integrated energy system scheduling method taking into account uncertainty and multi-energy multi-type demand response, constructs an islanded integrated energy system (IIES) of coupled solar thermal power stations taking into account the uncertainty of wind and light output, and constructs a two-layer coordinated optimization scheduling model of the islanded integrated energy system considering user demand response and optimized unit output on the basis of the islanded integrated energy system. The two-layer coordinated optimization scheduling model first takes multi-energy multi-type demand response as a decision variable, and constructs an upper-level optimization model with the goal of maximizing the difference in user load shedding penalty costs before and after optimization; secondly, by optimizing the unit output in each time period, a lower-level optimization model with the goal of maximizing the system net benefit is constructed. At the same time, the present invention proposes a scenario generation method based on super Latin hypercube and scenario reduction method of synchronous back generation for the uncertainty of clean energy, and adopts an improved adaptive genetic algorithm to solve the two-layer coordinated optimization scheduling model to ensure the effectiveness and practicality of the model solution.

[0059] The specific implementation steps of the method of the present invention are as follows:

[0060] S1: Constructing an isolated island integrated energy system and its unit output model: The combination of multiple energy sources has, to a certain extent, smoothed out the instability of a single energy supply due to resource reserves, seasonal changes, etc., enhanced the reliability of the system, adapted to the trend of future energy development, reduced dependence on traditional fossil energy, achieved green and low-carbon development, and also met the user's energy diversification needs. Therefore, the present invention expands the isolated island microgrid to a coupled solar thermal power station that takes into account the uncertainty of wind and light output to obtain an isolated island integrated energy system.

[0061] The isolated island integrated energy system constructed by the present invention is as follows Figure 1 As shown in the figure, it is composed of clean energy power supply units, solar thermal power stations, combined heat and power units, and geothermal heat pumps (GHP), which jointly supply the internal electricity and heat load requirements of the system. Among them, the clean energy power supply unit is composed of a wind power plant (WPP), a photovoltaic power plant (PV), and an electricity storage device (ESD); the solar thermal power station is composed of a collector field, a generator, and a heat storage device; the combined heat and power unit is composed of a gas turbine (GT), a waste heat recovery device (WHB), and a heat exchanger (CH).

[0062] The integrated energy system studied in this invention is in an island operation mode, disconnected from the external power grid and heat network, and the thermal power loads of users within the system are all supplied by internal units; the only energy input within the system network is natural gas, and the only gas load demand within the supply system is the GT operation demand.

[0063] The unit output models of uncertain units and conventional units included in the isolated integrated energy system include:

[0064] (1) Uncertain unit output model, including:

[0065] 1) Wind turbines:

[0066] Wind turbines are affected by the randomness of wind speed, and their output is uncertain. Wind speed can usually be described by Weibull distribution, which is expressed as follows:

[0067] (1);

[0068] Where: Wind speed The probability density function of is the shape parameter, is the scale parameter, the natural constant ≈2.71828.

[0069] Based on the analysis and calculation of the randomness of wind speed, the output model of wind turbine generator set is as follows:

[0070] (2);

[0071] Where: is the actual output power of WPP during period t, is the rated output power of WPP during period t, , , are the cut-in, rated and cut-out wind speeds, is the real-time wind speed during period t.

[0072] The constraints that need to be met to further construct the wind turbine are as follows:

[0073] (3);

[0074] Where: is the installed capacity of WPP, is the maximum installed capacity of WPP, , are the minimum and maximum output powers of WPP in period t, is the output power of WPP in period t.

[0075] 2) Photovoltaic generator set:

[0076] Photovoltaic generators are affected by the randomness of solar radiation, and their output is uncertain. Solar radiation can usually be described by Beta distribution, and its expression is as follows:

[0077] (4);

[0078] Where: is the probability density function of solar radiation intensity θ, is the gamma function, and are the solar irradiance and the maximum solar irradiance, and is the shape parameter of the Beta distribution.

[0079] Based on the analysis and calculation of the randomness of solar energy, the output model of the photovoltaic generator set is as follows:

[0080] (5);

[0081] Where: is the PV output power during period t, is the solar radiation area, The PV power conversion efficiency is is the solar radiation intensity during period t.

[0082] The constraints that need to be met to further construct the photovoltaic generator set are as follows:

[0083] (6);

[0084] Where: is the installed capacity of PV, is the maximum installed capacity of PV, , are the minimum and maximum output power of PV in period t, is the output power of PV during period t.

[0085] 3) Solar thermal power station:

[0086] The CSP power station is abstractly simplified into three parts: heat collection, heat storage and power generation. The input of the light field is the direct solar radiation in the normal direction. , the heat output to the heat transfer fluid after the conversion of the light field is The light field transfers part of the collected thermal energy to the power conversion module for direct power generation , the other part of the heat In addition, due to the limitations of the technology and absorption capacity of the solar thermal power station itself, the solar field will produce some abandoned light. ; The heat storage level of the heat storage system is , the heat dissipation transferred to the power conversion module is , used to maintain the power station's output power during periods of no sunlight or reduced sunlight caused by cloud cover; all heat transferred to the power conversion module is converted into electrical power, is the output power of the CSP station. The energy flow process inside the CSP station is as follows: Figure 2 shown.

[0087] In the heat collection process, the heat received by the CSP power station is expressed as follows:

[0088] (7);

[0089] Where: is the thermal energy collected by the light field during period t, is the conversion efficiency of the mirror field, is the direct solar radiation during period t, is the heat collection area of ​​the concentrator, is the solar incidence angle correction factor, is the time interval, , are the geometric efficiency and optical efficiency of the concentrator, respectively.

[0090] In addition to being directly transferred to the power conversion module to convert into electrical output power, the thermal energy obtained by the light field is also stored in the thermal storage system, which makes the solar thermal power station have good scheduling characteristics. This is also the biggest advantage compared to intermittent renewable energy generation technologies such as wind power and photovoltaics. Its expression is as follows:

[0091] (8);

[0092] Where: is the heat energy transmitted from the light field to the power conversion module for power generation during period t, is the thermal energy transmitted from the light field to the thermal storage system for storage during period t, is the amount of abandoned light in the light field during period t.

[0093] In the heat storage link, the heat storage system makes the CSP power station a dispatchable resource without increasing the uncertainty of the system, efficiently utilizes solar energy and efficiently converts it into heat energy for storage, and maintains the power station's electrical output power during periods of no sunlight or reduced sunlight caused by cloud cover. The heat storage capacity of the heat storage system is determined by the residual heat energy and the charging and discharging heat energy of the previous period, and its expression is as follows:

[0094] (9);

[0095] Where: is the heat storage capacity of the heat storage system in period t, It is the thermal energy that the thermal storage system transmits to the power conversion module during period t for power generation.

[0096] In the power generation stage, the electrical output power of the CSP plant is determined by the heat energy transmitted by the light field and the heat released by the heat storage system, as follows:

[0097] (10);

[0098] Where: is the electrical output power of the CSP plant in period t, is the cycle thermal efficiency, is the heat energy transmitted from the light field to the power conversion module for power generation during period t, It is the thermal energy that the thermal storage system transmits to the power conversion module during period t for power generation.

[0099] The constraints that need to be met to further build a solar thermal power station.

[0100] The following constraints must be met in the heat collection process:

[0101] (11);

[0102] Where: is the fluctuation of the collector field during period t, , are the minimum and maximum fluctuations of the collector field during period t, is the installed capacity of the collector field.

[0103] The following constraints must be met in the heat storage phase:

[0104] (12);

[0105] Where: is the heat storage capacity of the heat storage system in period t, , are the maximum and minimum heat storage capacities of the heat storage system during period t, , are the maximum charging power and maximum discharging power of the thermal storage system in adjacent scheduling time intervals, is the charging and discharging power state variable of the thermal storage system in period t, 1 represents heat release, and 0 represents heat charging.

[0106] The following constraints must be met in the power generation stage:

[0107] (13);

[0108] Where: is the operating state of the power conversion module in period t, 1 indicates operation, and 0 indicates shutdown; , They are the start and stop variables of the generator set in period t, 1 means start, 0 means stop; , The generator sets are The start and stop variables of the time period, 1 means start, 0 means stop; , are the minimum start and stop times of the power conversion module respectively; is the electrical output power of the power conversion module in period t, , are the maximum and minimum electrical output powers of the power conversion module, respectively; , They are the maximum upward rate and the maximum downward rate of the power conversion module respectively; is the maximum input thermal energy of the power conversion module.

[0109] (2) Conventional unit output model, including:

[0110] 1) Combined heat and power unit, its output model is as follows:

[0111] (14);

[0112] Where: is the power generation of CHP during period t, is the amount of natural gas consumed during period t, is the power generation efficiency of CHP, is the waste heat power of CHP power generation during period t, is the thermoelectric ratio, The heat supply for the heat exchange device during period t, is the waste heat entering the heat exchange device during period t, is the heating efficiency of the heat exchanger.

[0113] The constraints that need to be met to further build the cogeneration unit are as follows:

[0114] (15);

[0115] Where: , are the lower and upper limits of CHP output, , are the lower and upper limits of the heat exchanger output, respectively. is the installed capacity of the waste heat boiler.

[0116] 2) Ground source heat pump, its output model is as follows:

[0117] (16);

[0118] Where: is the heating power of the ground source heat pump during period t, is the amount of electricity entering the ground source heat pump during period t, is the electro-thermal efficiency of the ground source heat pump.

[0119] The constraints that need to be met for further construction of ground source heat pumps are as follows:

[0120] (17);

[0121] Where: is the installed capacity of the ground source heat pump, , They are the maximum and minimum power supply of the ground source heat pump respectively.

[0122] S2: Construct a two-layer coordinated optimization scheduling model for an isolated integrated energy system. For an isolated integrated energy system, when analyzing the overall optimization operation of its "source, load and storage", it is necessary to use the system's internal power generation and load demand as the basis for scheduling, and the load curve and power generation curve will be affected by the demand-side response. Therefore, the present invention constructs a two-layer coordinated optimization scheduling model for an isolated integrated energy system. The upper optimization model aims to maximize the difference in user load shedding penalty costs before and after optimization. According to the predicted curves of user load and distributed power generation, combined with the user demand elasticity coefficient and the maximum load reduction controllable load information, a load optimization strategy is made, and the result is passed to the lower optimization model. The supply side optimizes the output strategy for each time period based on the optimized load curve with the goal of maximizing net profit, and feeds back the operating plan to the upper optimization model to promote further optimization.

[0123] S2.1: Before constructing the upper optimization model, the uncertainty of clean energy output needs to be processed. Therefore, the present invention proposes a scenario reduction method based on super Latin hypercube scenario generation and synchronous back generation to process the uncertainty of clean energy output, such as Figure 3 As shown, the specific steps are as follows:

[0124] Step 2.1.1: Determine the number of sampling intervals. According to the probability distribution of wind and solar clean energy output, determine the number of sampling intervals that need to be divided.

[0125] Step 2.1.2: Randomly select sample values ​​in each sampling interval and calculate the probability of each sample value. In each sampling interval, randomly select sample values ​​according to the interval probability to form an initial scene set, and use the inverse function transformation to obtain the probability value of each sample point to ensure that the sample point conforms to the original probability distribution.

[0126] Step 2.1.3: Generate a sampling matrix. Using Latin hypercube sampling, generate a hypercube for each random variable, where each dimension represents a variable and each variable randomly selects sample values ​​in each interval.

[0127] Step 2.1.4: Check the number of sampling intervals. Check whether each sample in the sampling matrix meets the required number of sampling intervals. If it does, stop sampling, count the scene data, and perform statistics and analysis on the existing scene data; if it does not, repeat step 2.1.2.

[0128] Step 2.1.5: Determine the number of scenes to keep. The number of scenes to keep will affect the decision-making process for scene deletion.

[0129] Step 2.1.6: Calculate scene distances and perform scene deletion and scene retention. Calculate the distances between different scenes, and select and delete those scenes that are closest to other scenes based on the synchronous back-generation elimination algorithm.

[0130] Step 2.1.7: Add the probabilities of the deleted scenes to the retained scenes. The probability distribution of the deleted scenes is weighted to the retained scenes to ensure that the reduced set of scenes can represent the probability characteristics of the original set of scenes.

[0131] Step 2.1.8: Check whether the number of deleted scenes is equal to the number of scenes to be retained. Confirm whether the number of currently retained scenes has reached the preset target number. If so, stop sampling, count the retained data, and prepare for subsequent analysis or decision support; if not, repeat step 2.1.6.

[0132] S2.2: Constructing the upper optimization model: Taking the multi-energy and multi-type demand response strategy as the upper optimization decision variable, taking the difference in the system load shedding penalty cost before and after optimization as the upper optimization operation target, constructing the upper optimization model, the objective function of which is as follows:

[0133] (18);

[0134] Where: is the upper optimization objective function, which represents the difference in load shedding cost before and after the upper optimization. and are the unit penalty costs of load shedding of the i-th type of load before and after the upper optimization, and They are the load shedding amounts of the i-th type of load in period t under w scenarios before and after the upper-layer optimization, I is the number of scenarios, and T is the total period.

[0135] The multi-energy and multi-type demand response constructed by the present invention includes price type and incentive type. In addition, due to the coupling of multiple energy sources within the isolated integrated energy system, the supply and consumption of different energy sources affect each other, so the traditional power side is expanded to the field of heat and electricity multi-energy, and a multi-energy demand response model is constructed. The price type and incentive type multi-energy demand response models are as follows:

[0136] (1) Price-based multi-energy demand response model:

[0137] Based on the user demand elasticity matrix, the peak and valley time-of-use electricity and heat prices within the system are re-established to guide users to change their energy usage. The model is as follows:

[0138] (19);

[0139] Where: , are the load demands of the user class i before and after the implementation of the peak and valley time-of-use electricity and heat prices in period t, is the load elasticity coefficient of the i-th type, and j is the category of price change factors that affect the demand for this type of load. and are the original price and price change of the i-th type of load in period t respectively.

[0140] The load elastic coefficient model is as follows:

[0141] (20);

[0142] In the formula: when z=j, it is the self-elastic coefficient, when z≠j, it is the cross-elastic coefficient. is the original load of the i-th type of load in time period z, is the user load change of the i-th load in time period z. and are the original price and price change of the i-th type of load in period z respectively. and are the original price and price change of the i-th type of load in period j respectively.

[0143] (2) Incentive multi-energy demand response model:

[0144] The incentive-based multi-energy demand response model is a load increase and decrease agreement signed between the operator of the isolated integrated energy system and the user. The agreement includes the user's load reduction amount and the corresponding unit compensation price. The incentive-based multi-energy demand response cost (user participation in demand response benefits) model is as follows:

[0145] (twenty one);

[0146] Where: is the incentive demand response cost of the i-th load in period t (the user's benefit from participating in demand response), is the load reduction amount of the i-th type of load in period t, is a trapezoidal price, which indicates the unit compensation price of the i-th type of load in the m-segment interval where the load can be reduced. Figure 4 Place, and are the minimum and maximum load reduction of the i-th load respectively.

[0147] At the same time, the constraints that the upper-level optimization model needs to meet are as follows:

[0148] (1) Supply and demand balance constraints, including:

[0149] 1) Electric load supply and demand balance constraint, its expression is as follows:

[0150] (twenty two);

[0151] Where: The user power load satisfied by the isolated island integrated energy system after upper optimization, is the optimized upper-layer load shedding in period t under scenario w, is the power generation of WPP and PV in period t under scenario w, is the power generation of CHP in period t under scenario w, is the electric power output of the CSP plant at time t under scenario w, is the discharge amount of the power storage device in period t under scenario w, is the discharge efficiency of the energy storage device. is the change in WPP and PV power generation after the upper optimization, is the change in CHP power generation after upper optimization, is the change in power generation of the CSP plant after optimization at the upper level, The change in power generation of the power storage device after upper-level optimization.

[0152] 2) Heat load supply and demand balance constraint, its expression is as follows:

[0153] (twenty three);

[0154] Where: The user heat load satisfied by the isolated island integrated energy system after upper optimization, is the optimized hot load of the upper layer in time period t under scenario w, is the heat supply of the heat exchanger during period t under scenario w, is the heating power of the ground source heat pump in period t under scenario w, The heat storage device releases heat to meet the heat load ratio, is the heat release power of the heat storage device in period t under scenario w, is the heat release efficiency of the heat storage device. is the change in heating power of the heat exchange device after optimization of the upper layer, is the change in heating power of the ground source heat pump after optimization of the upper layer, The change in heating power of the heat storage device after optimization for the upper layer.

[0155] 3) Gas load supply and demand balance constraint, its expression is as follows:

[0156] (twenty four);

[0157] Where: is the gas load in period t under scenario w, is the amount of gas purchased from the external network during period t under scenario w.

[0158] (2) Demand response constraints, including:

[0159] 1) Load variation constraint:

[0160] The load reduction of the user does not exceed the maximum value that it can bear, and the constraints are as follows:

[0161] (25);

[0162] Where: is the original load demand of the i-th category of users in period t, To implement the mth type of demand response, the user's i-th type of load demand is: is the maximum change in load of the i-th type of user during period t.

[0163] 2) Peak-valley power constraint, expressed by the peak-valley power ratio threshold, is expressed as follows:

[0164] (26);

[0165] Where: and They are the minimum load during peak period and the maximum load during valley period respectively.

[0166] 3) User energy cost constraint, which is expressed as follows:

[0167] (27);

[0168] Where: and They are the energy consumption costs of the i-th type of load of the user in period t before and after the implementation of the m-th demand response.

[0169] S2.3: Construct the lower optimization model: Take the output of each unit in the system as the lower optimization decision variable, and take the maximization of the net profit of the internal energy supply side of the isolated island integrated energy system as the lower optimization operation goal to construct the lower optimization model. Its objective function is as follows:

[0170] (28);

[0171] Where: is the lower optimization objective function, which represents the maximization of the net benefit of the isolated island integrated energy system. The net benefit of the island integrated energy system is , are the electricity and heat benefits of the isolated island integrated energy system under scenario w, is the depreciation cost of the unit under scenario w, , , , They are the unit operation and maintenance cost, load shedding cost, incentive-based demand response cost and gas purchase cost under scenario w.

[0172] in,

[0173] (29);

[0174] Where: is the IIES electricity price in period t, and They are respectively the user electric load satisfied by the isolated island integrated energy system in period t under scenario w and the change in electric load satisfied after upper-layer optimization.

[0175] (30);

[0176] Where: is the heat selling price of the isolated island integrated energy system during period t, and They are respectively the user thermal charge satisfied by the isolated island integrated energy system in period t under scenario w and the thermal load variation satisfied after upper layer optimization.

[0177] (31);

[0178] Where: , , are the total investment cost, residual value and service life of unit 1 respectively.

[0179] (32);

[0180] Where: is the output of unit l in period t under scenario w, is the unit operation and maintenance cost of unit l.

[0181] (33);

[0182] Where: is the unit penalty cost of load shedding for the i-th type of load, is the total amount of load shedding of the i-th type of load in period t under scenario w.

[0183] (34);

[0184] Where: is the incentive demand response cost of the i-th type of load in period t.

[0185] (35);

[0186] Where: is the external gas purchase price during period t, is the gas purchase volume of the island integrated energy system before the upper optimization is implemented in time period t under scenario w, is the change in gas purchase of the isolated island integrated energy system after the upper-layer optimization is implemented in period t under scenario w.

[0187] At the same time, the constraints that need to be met by the lower-level optimization are the constraints that each unit in the system needs to meet.

[0188] S3: Solve the two-layer coordinated optimization scheduling model. Although traditional adaptive genetic algorithms have shown advantages in dealing with complex nonlinear problems, they still face challenges in practical applications such as difficulty in parameter adjustment, easy to fall into local optimal solutions and slow convergence speed. These problems limit their application efficiency and effectiveness in certain complex system optimization problems. In order to overcome these limitations, the present invention proposes an improved adaptive genetic algorithm for a two-layer coordinated optimization scheduling model. The algorithm improves the global search capability and convergence speed of the algorithm by dynamically adjusting the crossover and mutation probabilities and adopting an elite strategy to optimize the population, thereby effectively avoiding the risk of falling into local optimal solutions. The specific steps are as follows:

[0189] Step 3.1: Input the original data of the isolated integrated energy system, including typical predicted output of clean energy, unit output parameters, energy supply price, user load and other data.

[0190] Step 3.2: Initialize the two-layer coordinated optimization scheduling model, the number of iterations of the upper optimization scenario and the lower optimization scenario to 0.

[0191] Step 3.3: Solve the upper optimization model. After randomly generating the initial population, calculate the difference in the user load shedding penalty cost of the unit operation optimization plan corresponding to each individual, and calculate the individual fitness. Then, select the operation through the roulette method, adaptive mutation, crossover and elite retention, and generate a new population. The supply and demand balance and demand response constraints must be met during the solution of the upper optimization model. Repeat the above process until the maximum number of iterations of the upper optimization is reached, and output the optimal solution of the upper optimization model.

[0192] Step 3.4: Determine whether the two-layer coordinated optimization scheduling model has reached the maximum number of iterations. If not, pass the upper-layer optimization result to the lower layer to solve the lower-layer optimization model.

[0193] Step 3.5: Solve the lower-level optimization model. After randomly generating the initial population, solve the lower-level optimization, calculate the net benefit value of the energy supply side of the system, and calculate the individual fitness. Then, select the operation through the roulette method, adaptive mutation, crossover and elite retention, and generate a new population. The output constraints of each unit must be met during the solution of the lower-level optimization model. Repeat the above process until the maximum number of iterations of the lower-level optimization is reached, output the optimal solution of the lower-level optimization model, feed back the lower-level optimization results to the upper layer, and solve the upper-level optimization model.

[0194] Step 3.6: Until the two-layer coordinated optimization scheduling model reaches the maximum number of iterations, the optimal result is output.

[0195] Figure 5 This is an improved genetic algorithm processing step. The algorithm optimizes the population by dynamically adjusting the crossover and mutation probabilities and adopting an elite strategy. The crossover and mutation probability models are as follows:

[0196] (36);

[0197] (37);

[0198] Where: and are the crossover and mutation probabilities, respectively; and are the maximum and minimum crossover probability, respectively. , are the maximum and minimum values ​​of the mutation probability, respectively. and are the maximum fitness value and average fitness of the current population, respectively. is the larger fitness value among the crossover parents, is the fitness value of the mutant parent.

[0199] The present invention adopts an optimal preservation strategy to retain the best individual of each generation in the heredity. The method of the optimal preservation strategy is to compare the highest fitness value in the new generation group generated in each generation after the genetic operation with the highest fitness value in the previous generation group. If it is less than the highest fitness value of the previous generation, an individual in the new generation is randomly eliminated, and the individual with the highest fitness value in the previous generation is added to the new generation. The optimal preservation strategy can ensure that the current best individual will not be destroyed by genetic operations such as crossover and mutation, which is an important guarantee condition for the convergence of the improved adaptive genetic algorithm of the present invention.

[0200] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response, characterized by: The following steps are involved: S1: Construct an isolated island integrated energy system and its unit output model, wherein the isolated island integrated energy system consists of a clean energy power supply unit, a solar thermal power station, a cogeneration unit and a ground source heat pump, wherein the clean energy power supply unit and the solar thermal power station are uncertain units, the cogeneration unit and the ground source heat pump are conventional units, and the unit output model includes an uncertain unit output model and a conventional unit output model; S2: Construct a two-layer coordinated optimization scheduling model for an isolated integrated energy system: The two-layer coordinated optimization scheduling model includes an upper optimization model and a lower optimization model, wherein the upper optimization model aims to maximize the difference between the user load shedding penalty cost before and after optimization, and makes a load optimization strategy based on the prediction curve of user load and distributed generation power, combined with the user demand elasticity coefficient and the maximum load reduction controllable load information, and transmits the result to the lower optimization model. The lower optimization model optimizes the output strategy of each time period with the goal of maximizing net profit, and feeds back the running plan to the upper optimization model; S3: An improved adaptive genetic algorithm is used to solve the two-layer coordinated optimization scheduling model.

2. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 1, characterized in that: The clean energy power supply unit includes a wind power unit and a photovoltaic power unit, and the uncertain unit output model includes a wind power unit output model and its constraints, a photovoltaic power unit output model and its constraints, and a solar thermal power station output model and its constraints; The conventional unit output model includes the combined heat and power unit output model and its constraints, and the ground source heat pump output model and its constraints.

3. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 1, characterized in that: Step S2 specifically includes: S2.1: Scenario generation based on super Latin hypercube and scenario reduction method with synchronous back generation to deal with uncertainty in clean energy output; S2.2: Constructing the upper optimization model: Taking multi-energy and multi-type demand response as the upper optimization decision variable, taking the difference of system load shedding penalty cost before and after optimization as the upper optimization operation target, constructing the upper optimization model; S2.3: Construct the lower-level optimization model: Take the output of each unit in the isolated integrated energy system as the lower-level optimization decision variable, and take the maximization of the net profit on the internal energy supply side of the isolated integrated energy system as the lower-level optimization operation goal to construct the lower-level optimization model.

4. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 3, characterized in that: The implementation process of step S2.1 is as follows: Step 2.1.1: Determine the number of sampling intervals; Step 2.1.2: Randomly select sample values ​​in each sampling interval and calculate the probability of each sample value; Step 2.1.3: Generate sampling matrix: Use hypercube sampling to generate a hypercube for each random variable, where each dimension represents a variable and each variable randomly selects sample values ​​in each interval; Step 2.1.4: Check the number of sampling intervals: Check whether each sample in the sampling matrix meets the required number of sampling intervals. If so, stop sampling, count the scene data, and perform statistics and analysis on the existing scene data; if not, repeat step 2.1.2; Step 2.1.5: Determine the number of scenes to be retained; Step 2.1.6: Calculate scene distance and perform scene deletion and scene retention; Step 2.1.7: Add the probability of the deleted scene to the retained scene; Step 2.1.8: Check whether the number of scenes after deletion is equal to the number of scenes that need to be retained, and confirm whether the number of scenes currently retained has reached the preset target number. If so, stop sampling and count the retained data. If not, repeat step 2.1.

6.

5. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 3, characterized in that: The objective function of the upper optimization model is expressed as follows: ； Where: is the upper optimization objective function, which represents the difference in load shedding cost before and after the upper optimization. and are the unit penalty costs of load shedding of the i-th type of load before and after the upper optimization, and They are the load shedding amounts of the i-th type of load in period t under w scenarios before and after the upper-layer optimization, I is the number of scenarios, and T is the total period.

6. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 5, characterized in that: Multi-energy multi-type demand response includes price-based and incentive-based multi-energy demand response models. The price-based multi-energy demand response model re-formulates the peak and valley time-of-use electricity and heat prices within the isolated integrated energy system based on the user demand elasticity matrix to guide users to change their energy usage methods. Its expression is as follows: ； Where: , are the load demands of the user class i before and after the implementation of the peak and valley time-of-use electricity and heat prices in period t, is the elasticity coefficient of the i-th type of load, j is the category of price change factors that affect the demand for this type of load, and are the original price and price change of the i-th type of load in period t respectively; The incentive-based multi-energy demand response model is a load increase and decrease agreement signed between the operator of the isolated integrated energy system and the user, and its expression is as follows: ； Where: is the incentive demand response cost of the i-th load in period t, is the load reduction amount of the i-th type of load in period t, It is a trapezoidal price, which means the unit compensation price of the i-th type of load in the m-segment interval where the load can be reduced.

7. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 5, characterized in that: The constraints that the upper-level optimization model needs to meet include: Supply and demand balance constraints: including electricity load supply and demand balance constraints, heat load supply and demand balance constraints, and gas load supply and demand balance constraints; Demand response constraints: including load variation constraints, peak and valley power constraints, and user energy cost constraints.

8. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 3, characterized in that: The expression of the objective function of the lower optimization model is as follows: ； Where: is the lower optimization objective function, which represents the maximization of the net benefit of the isolated island integrated energy system. The net benefit of the island integrated energy system is , are the electricity and heat benefits of the isolated island integrated energy system under scenario w, is the depreciation cost of the unit under scenario w, , , , They are the unit operation and maintenance cost, load shedding cost, incentive-based demand response cost and gas purchase cost under scenario w.

9. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 8, characterized in that: The constraints that the lower optimization model needs to meet are the constraints of each unit in the isolated integrated energy system, including: Clean energy power supply unit constraints: including unit installed capacity constraints and output power constraints; Constraints of CSP power station units: including constraints of the heat storage system in the heat storage link and constraints of the power conversion module in the power generation link. The constraints of the heat storage system include constraints of heat storage capacity, heat charging and discharging power, and charging and discharging power state. The constraints of the power conversion module include constraints of the power conversion module operation state, start and stop time constraints, electrical output power constraints, upward and downward regulation rate constraints, and input thermal energy constraints. Cogeneration unit constraints: including ground source heat pump output constraints, heat exchanger output constraints, and waste heat boiler capacity constraints; Ground source heat pump constraints: including ground source heat pump installed capacity constraints and energy supply power constraints.

10. The method for dispatching an isolated island integrated energy system taking into account uncertainty and multi-energy and multi-type demand response according to claim 1, characterized in that: The steps of solving the two-layer coordinated optimization scheduling model using the improved adaptive genetic algorithm include: Step 3.1: Input the raw data of the isolated integrated energy system; Step 3.2: Initialize the iteration number of the two-layer coordinated optimization scheduling model, the upper optimization scenario, and the lower optimization scenario to 0; Step 3.3: Solving the upper optimization model: After randomly generating the initial population, calculate the difference in the user load shedding penalty cost of the unit operation optimization plan corresponding to each individual, and calculate the individual fitness; then select the operation through the roulette method, adaptive mutation, crossover and elite retention, and generate a new population; repeat the above process until the maximum number of iterations of the upper optimization is reached, and output the optimal solution of the upper optimization model; Step 3.4: Determine whether the two-layer coordinated optimization scheduling model has reached the maximum number of iterations. If not, pass the upper layer optimization result to the lower layer to solve the lower layer optimization model; Step 3.5: Solving the lower-level optimization model: After randomly generating the initial population, solve the lower-level optimization, calculate the net benefit value of the internal energy supply side of the isolated integrated energy system, and calculate the individual fitness; then select the operation through the roulette method, adaptive mutation, crossover and elite retention, and generate a new population; repeat the above process until the maximum number of iterations of the lower-level optimization is reached, output the optimal solution of the lower-level optimization model, feed back the lower-level optimization results to the upper layer, and solve the upper-level optimization model; Step 3.6: Until the two-layer coordinated optimization scheduling model reaches the maximum number of iterations, the optimal result is output.

Images