Virtual power plant regulation potential analysis method in spot mode

Abstract

The invention belongs to the technical field of virtual power plants, and particularly relates to a virtual power plant regulation potential analysis method in a spot mode, which comprises the following steps of: establishing translation, transferable and interruptible load regulation parameter classification models for virtual power plant users to participate in response; a multi-type adjustment resource classification clustering external characteristic typical database is constructed; carrying out aggregation calculation on typical classification marginal costs of different resource adjustment responses of the clustered virtual power plant users; based on an improved K-means clustering method, high-proportion new energy system spot market price and classification model peak-valley price difference interval clustering analysis is carried out, a virtual power plant user adjustable potential analysis model under a typical scene of a long-period spot mode is established, and the effect of improving participation of a virtual power plant in a spot market is achieved. It is of great significance to motivate user enterprises to adjust resources to participate in the market.

Description

Technical Field

[0001] The present invention belongs to the technical field of virtual power plants, and particularly relates to a method for analyzing the regulation potential of a virtual power plant in a spot market mode. Background Art

[0002] With the continuous advancement of the development of new energy, the scale of new energy is constantly expanding. On the one hand, the balance between power supply and demand and the difficulty of new energy consumption are prominent. Under the large fluctuations of new energy, the regulation capabilities and potentials of conventional energy sources have been deeply explored. Especially in the face of long-term scenarios of continuous windless and sunlightless new energy or long-term scenarios of continuous strong wind of new energy, virtual power plants aggregate various types of regulation resources such as controllable industrial loads, distributed photovoltaics and energy storage, electric vehicles, and commercial buildings. Deeply tapping the potential of virtual power plants has become an important means for ensuring power supply and consuming new energy. On the other hand, the response costs of various types of regulation resources involved in virtual power plants in the spot market mode are different, and the regulation response time, regulation capacity, duration, regulation rate, regulation accuracy, etc. are complex and variable. It is necessary to carry out clustering of typical classified regulation resources and typical spot market scenarios and conduct precise analysis in the spot market mode. Therefore, the analysis of the adjustable potential of virtual power plant users in a high-proportion new energy system under the clustering of typical market scenarios in the precise spot market mode is the basis for the operation of virtual power plants.

[0003] At present, scholars at home and abroad have carried out a large number of studies on the evaluation and analysis of the conventional regulation capabilities of virtual power plants. However, there is still a blank in the analysis of the regulation capabilities of virtual power plants under the complex spot market mode of high-proportion new energy systems. In terms of the evaluation and analysis of the conventional regulation capabilities of virtual power plants, most are based on model-data joint driving, clustering estimation, and multi-model dynamic combination. On the one hand, clustering division based on label characteristics is carried out to find similar operation curves. On the other hand, based on the combined analysis of multi-modal monitoring data, the perceiving and prediction of the adjustable potential of distributed resources are realized. For example, for the prediction of the adjustable capabilities of charging and swapping stations, multi-dimensional data such as the flow of people, electric vehicle heat maps, historical load data of charging and swapping stations, and working days and holiday periods need to be comprehensively analyzed. The comprehensive clustering method is used to study the user load characteristics and adjustable potential, aiming at the participation of adjustable loads in power grid dispatching, the integration of 5G technology, the use of an improved ant lion hybrid optimization system model, and the prediction of regulation capabilities. However, the types of flexible resources on the user side accessed by virtual power plants are diverse, the single-unit capacity is small, the locations are scattered, it is difficult to collect resource device information and real-time monitoring data, the resource characteristics are different, there are differences in time response characteristics, and it is difficult to build an aggregation model for resources; the description system of the regulation cost model of virtual power plants and the economic benefit model are incomplete. There are large differences in regulation resources under different typical spot market modes, and the problems of inaccurate analysis and response cannot be solved. The users of virtual power plants are complex and changeable: changes in external factors such as the market environment, economic situation, policies and regulations, and the production and operation conditions of users themselves may all affect the willingness and behavior of users, resulting in difficulties in the prediction, planning, and management and control of virtual power plants. The virtual power plant market is complex and changeable: the competition is fierce, the response modes are diverse, and it is difficult to calculate the regulation control cost, resulting in difficulties in price prediction and great difficulties in making trading quotation decisions for virtual power plants.

[0004] Existing research shows that by analyzing the output characteristics of high-proportion new energy, the accuracy of the analysis and evaluation of the regulation capabilities of virtual power plants under the spot market mode can be improved, and the virtual power plants can effectively participate in the declaration of the spot market and the adjustment of regulation responses. However, the existing research still has the following deficiencies: (1) There is relatively little research on how to classify the responses of regulation resources on the user side of virtual power plants, such as different typical industrial loads, distributed new energy and energy storage, and commercial buildings; (2) There is a lack of a method for analyzing the reliable response capabilities of regulation resources of virtual power plants under complex spot market modes; (3) There is no targeted research on quantitatively describing the analysis of typical classified regulation resources of virtual power plants by using the aggregation characteristics of new energy output, spot prices, and peak-valley price differences and typical scenarios. Summary of the Invention

[0005] Aiming at the above technical problem that the existing evaluation of the regulation capabilities of virtual power plant users lacks adaptability to multiple scenarios of spot market operation, the present invention provides a method for analyzing the regulation potential of virtual power plants under the spot market mode.

[0006] To solve the above technical problems, the technical solution adopted by the present invention is as follows: A method for analyzing the regulation potential of a virtual power plant in a spot market mode, comprising the following steps: S1. For two typical demands of consuming new energy and ensuring power supply, establish three load regulation parameter classification models of translatable, transferable, and interruptible for virtual power plant users to participate in response; S2. Considering the operation constraints of different types of regulation resources such as industrial loads, distributed energy storage, distributed photovoltaics, and industrial and commercial buildings connected to the virtual power plant, based on the load regulation parameter classification model formed in S1, establish a typical database of the external characteristics of different types of regulation resources classified and clustered; S3. Based on different types of regulation resources and the typical database formed in S2, calculate the typical classification marginal cost of the response of different regulation resources of virtual power plant users after clustering; S4. For four typical combined scenarios of wind, light, large, and small, based on the improved K-means clustering method, conduct clustering analysis on the spot market price and the price difference interval of the classification model of a high-proportion new energy system; S5. Based on the long-term prediction of spot market prices and the typical classification marginal cost, establish an analysis model for the adjustable potential of virtual power plant users in typical scenarios of the long-term spot market mode.

[0007] The method for establishing the translatable load regulation parameter classification model for virtual power plant users to participate in response in S1 is as follows: For virtual power plant users, the requirements for steel and glass as translatable loads are that the time is uninterrupted and the power magnitude remains unchanged at each moment before and after translation; for a certain translatable load , its power distribution vector is: (1) Wherein, is the starting time, is the continuous time period; use the 0-1 variable to represent the starting state of the time period, that is, represents starting to translate from the time period; The starting time period set is: ; represents that the load is not translated; and , indicating that the load is translated to start from , is the continuous distribution duration of the space where new energy is in large output or the spot price is low; the power distribution vector corresponding to is: (2) When , the corresponding power distribution vector is meaningless, and all of them are set to 0, that is: (3) Based on the power distribution vector, model the shiftable load, ; There are only two situations for the shiftable load after the virtual power plant participates in the response: a) not shifted; b) shifted to the period when new energy is in large generation or the spot price is low. The constraints are as follows: (4).

[0008] The method for establishing the classification model of adjustable parameters of transferable load for virtual power plant users to participate in the response in S1 is as follows: For the building users of the virtual power plant, their power can be transferred, but the time requirement is not an uninterrupted requirement. It can be adjusted within a certain time period and the total power remains unchanged. The transferable load power distribution vector is: (5) Use the 0-1 variable to represent the transfer status of the period , that is represents transferring in the period , represents not transferring; the power distribution vector of the load after transfer is: (6) The total power of the transferable load remains unchanged before and after the transfer. At time the transferred load power is between the maximum and minimum transfer powers; to avoid the load being transferred to multiple discontinuous periods, that is, the frequent start and stop of the switches and power sizes of air conditioners, refrigeration and heating electrical equipment, it is necessary to constrain the minimum continuous operation time of the transferred load, that is: (7) Among them, is the start time, is the duration, , are respectively the maximum and minimum allowable transfer powers of the load at time ;

[0009] The method for establishing a classification model of interruptible load regulation parameters for virtual power plant users to participate in response in S1 is as follows: For the load that can be reduced in non - continuous production enterprises such as steel and cement , as the participation response of virtual power plant users, it reduces the electricity consumption of users; use a 0 - 1 variable to represent the reduction state of the load that can be reduced or interrupted during the time period , that is means is reduced during the time period , means is not reduced, and the load model is: (8) Considering user satisfaction, it is necessary to constrain the minimum and maximum continuous reduction times and the reduction times, that is: (9) Among them, is the reduction coefficient of the load that can be reduced at the time node ; is the minimum continuous reduction time; is the maximum continuous reduction time; is the maximum reduction times.

[0010] The operation constraints of different types of regulation resources such as industrial loads, distributed energy storage, distributed photovoltaic, and industrial and commercial buildings connected to the virtual power plant in S2 are as follows: Operation constraints and external characteristics description of industrial loads: (10) Among them: represents the total adjustable power formed by aggregating J industrial loads of virtual power plant i at time t, , , respectively represent the adjustable load regulation power that can be translated equivalently, the transferable load , and the interruptible load equivalent coefficient of the th industrial load; represents the regulation capacity power constraint existing at time t during the operation of the th industrial load, , respectively represent the regulation equivalent coefficients corresponding to the main production load and the auxiliary production load during the operation of the th industrial load; Description of operating constraints and external characteristics of distributed energy storage: 1) During a battery charge / discharge cycle, the energy storage capacity value of the battery at each moment should be maintained within a certain range: (11) Where: is the state of charge of the battery energy storage at the moment; , respectively correspond to the upper and lower limits of the adjustable capacity of the battery energy storage; 2) Continuity constraint of the state of charge of energy storage: (12) Where: , are respectively the state of charge of the energy storage system at time t and time t-1; , are respectively the charging power and discharging power of the energy storage system at time t; and respectively correspond to the charging efficiency and discharging efficiency of the battery energy storage; is the time interval ( =1h); 3) Energy storage charge / discharge constraint: During operation, the battery energy storage should control the charge / discharge power each time not to exceed its rated value, and the total discharging power does not exceed the rated capacity of the energy storage: (13) Where: , respectively represent the maximum charging power and maximum discharging power of the energy storage system; Description of operating constraints and external characteristics of distributed photovoltaic: The operating constraint of distributed photovoltaic is the generable power that changes with time, and it participates in the aggregation response in the form of an interruptible load: (14) Where: , respectively represent the predicted values of the power generation of the distributed photovoltaic system and the maximum power generation at time t; is the equivalent interruptible load power value of the distributed photovoltaic; Description of operating constraints and external characteristics of industrial and commercial buildings: 1) Air conditioning system: Central air conditioners are installed in public buildings, with a load ratio of more than 40% and an adjustable potential ratio of 5% - 20%; assuming that the central air conditioning system in the building has been operating stably before the implementation of demand response, then considering the comfort of the indoor environment, the adjustable load constraint model of the central air conditioner is: (15) Wherein: is the maximum active power that can be reduced or increased by the centralized air - conditioning system; is the demand response duration; when in the heating mode, is the heating energy efficiency ratio of the centralized air - conditioning system, when in the cooling mode, is the cooling energy efficiency ratio of the centralized air - conditioning system; c is the specific heat capacity of air, and the specific heat capacity at constant pressure of air at a temperature of 300K is selected, with a value of 1.005 kJ / (kg·K); S is the heating area of the building; H is the average floor height of the building; is the density of air, and the dry air density at a temperature of 300K is selected, with a value of 1.177 kg / m 3 ; is the heating temperature set by the user; is the allowable indoor temperature; Among them, the shiftable load under different control modes of the air conditioner is equivalent to the following four scenarios: (16) Wherein, is the response capacity of the input current percentage of the control host. From receiving the response instruction to completing the power adjustment, the response preparation time is at the minute level, is the correction parameter; is the operating power of the air - conditioner host; is the host - limited current loading value; is the host operating output value; is the response capacity when remotely or locally controlling the water temperature, and the response preparation time is at the minute level, is the correction parameter; is the temperature adjustment value; is the response capacity for adjusting the compressor speed in the control frequency - conversion mode, is the frequency change value; is the rated frequency; is the response capacity when directly controlling to shut down some air conditioners; 2) Electric boiler: The heat - storage electric boiler is used for the domestic hot water use or heating of large public buildings such as office buildings, hotels, and shopping malls. Its load ratio can reach 30% - 40% in winter, and the adjustable potential reaches 15%; when the electric boiler has no heat - storage capacity, the calculation method of its constraint model is the same as formula (15); when the electric boiler has heat - storage capacity, the operation constraint is: (17) Wherein: is the heat change of the heat - storage medium after t moments of response, is the temperature of the heat - storage medium at the moment when the response is issued, is the specific heat capacity of the medium; is the density of the medium; V is the volume of the medium; is the temperature of the medium at time t; in the adjustable load model constraint conditions is the maximum active power that can be reduced or increased by the heat storage electric boiler; t′ is the demand response duration; is the efficiency of the electric boiler, is the heat transfer efficiency of the heat storage medium; is the mass flow rate of the boiler, is the water temperature at the secondary side outlet; is the water temperature at the secondary side inlet; 3) Lighting system: The load proportion of the lighting system is between 15% and 25%, and the adjustable potential proportion is between 3% and 10%. Its response preparation time is at the second level. The response capacity comes from two aspects. One is to ensure the normal work and production of users, and the method of sectional area control is adopted to turn off a part of the lighting equipment P1. The other is the lamps that could have been turned off but were not turned off during the non-working period P2; (18) Among them: P light is the power of the lamps originally turned on during work; λ1 is the reasonable turn-off coefficient obtained by optical measurement; P i is the lamps that can be turned off during the non-working period obtained by statistics; 4) Elevators and cold storage: The load proportion of the elevator and cold storage system is between 4% and 10%, and the adjustable potential proportion is between 1% and 5%. Its response preparation time is at the second level. The response capacity comes from turning off some idle elevators at night and shutting down the cold storage for a short time during the non-preparation time; (19) Among them, and respectively represent the equivalent interruptible elevator load and the cold storage refrigeration system load of the building load.

[0011] The method for calculating the typical classification marginal cost of the responses of different regulation resources of the virtual power plant users after clustering in S3 is: (20) Among them, represents the discount rate; represents the inflation rate; represents the service life of the regulation resource; represents the unit power retrofit cost of the regulation resource j; represents the unit power annual maintenance cost of the regulation resource j; represents the retrofit capacity of the regulation resource j; Represents the operating capacity of the regulation resource j; Represents the total regulation cost of aggregating a total of J regulation resources; c represents the typical regulation cost corresponding to the typical regulation power interval p.

[0012] The method for carrying out the clustering analysis of the price difference intervals of the spot market price and the classification model of the high-proportion new energy system in S4 is as follows: For the set of historical data scenarios p = {x1, x2, x3,..., x n}, where there are a total of n data samples (365 * 24), each sample is m-dimensional, assuming that the n data samples are divided into K classes, and the sample c i represents the center point of class K i , and N(K i ) represents the number of samples in class Ki. Represents the distance between samples, then: (21) Among them, r i is the class radius, which is the average value of the distances from all data within the class to the class center; λ is the ratio of the class radii; C ij is the midpoint between classes, which is the point that equally divides the line connecting different class centers according to the ratio of the class radii; is the class center density, which is the number of data within the class whose distance from the class center point is less than the class radius; is the class edge density, which represents the class dispersion, that is, among any two classes, the number of data whose distance to the midpoint between classes is less than the average value of the class radii; The index is the ratio of the data density within the class and the data density between classes; is the optimal number of clusters for K-means clustering determined based on the DBI index; the spot price and the peak-valley price difference formed after clustering are the corresponding typical libraries (X K1 , X K2 , X K3 ,...., X Kbest ; C K1 , C K2 , C K3 ,...., C Kbest ).

[0013] The method for establishing the virtual power plant user adjustable potential analysis model under the typical scenario of the long-term spot mode in S5 is as follows: (22) Among them, , , are the adjustable and transferable and interruptible regulation powers participated by the virtual power plant respectively. For the adjustable capabilities of the translatable and transferable types, when the spot price is higher than the iWhen the cost threshold of an adjustable power interval is reached, all the adjustment capabilities within this interval can be included in the response capacity interval; for interruptible adjustment capabilities, when the spot price is lower than its cost threshold, all its adjustment capabilities can be included in the response capacity interval.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: In response to the two typical response requirements of virtual power plants participating in new energy consumption and power supply guarantee, the present invention establishes a classification model for three types of load adjustment parameters, namely translatable, transferable, and interruptible, for virtual power plant users to participate in response, and gives parameter indicators for aggregating and classifying virtual power plant adjustment resources; taking the operation constraints of different types of adjustment resources such as industrial loads, distributed energy storage, distributed photovoltaics, and industrial and commercial buildings connected to the actual virtual power plant as the boundary, a typical database of the external characteristics of multi-type adjustment resource classification and clustering is constructed; and taking the transformation cost and operation and maintenance cost as the actual adjustment resource response cost, the typical classification marginal cost of different adjustment resource responses of virtual power plant users after aggregation calculation and clustering is obtained; in response to typical combined scenarios of new energy output with different scales of wind, light, large, and small, based on the improved K-means clustering method, the present invention conducts clustering analysis on the spot market price and the peak-valley price difference interval of the classification model of the high-proportion new energy system, benchmarks the spot price and the peak-valley price difference with the typical classification marginal cost, and establishes an analysis model for the adjustable potential of virtual power plant users in typical scenarios of the long-term spot mode, which is of great significance for improving the participation of virtual power plants in the spot market and motivating user enterprises to participate in the market with adjustment resources. Description of the Drawings

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings described below are only exemplary, and for those of ordinary skill in the art, without creative efforts, other implementation drawings can also be obtained based on the provided drawings.

[0016] The structures, proportions, sizes, etc. illustrated in this specification are only used to cooperate with the content disclosed in the specification for those who are familiar with this technology to understand and read, and are not used to limit the limiting conditions under which the present invention can be implemented. Therefore, they do not have technical substance. Any modification of the structure, change in the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed in the present invention.

[0017] Figure 1 It is a calculation flowchart of the present invention. Detailed Embodiments

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. These descriptions are only for further explaining the features and advantages of the present invention, rather than limiting the claims of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope protected by the present application.

[0019] The following will further describe in detail the specific implementation manners of the present invention in conjunction with the accompanying drawings and embodiments. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0020] A method for analyzing the regulation potential of a virtual power plant in a spot market mode, as Figure 1 shown, includes the following steps: Step 1: For two typical demands of absorbing new energy and ensuring power supply, establish a classification model of three load regulation parameters, namely translatable, transferable, and interruptible, for virtual power plant users to participate in response.

[0021] 1. Translatable load type: For some virtual power plant users, such as steel and glass, the requirement for translatable loads is that the time is continuous and the power magnitude remains unchanged at each moment before and after translation. For a certain translatable load , its power distribution vector is: (1) In the formula, is the start time, is the duration. Use the 0-1 variable to represent the start state of the period, that is, represents starting to translate from the period. The start period set is: represents that the load is not translated. And , represents that the load is translated to start from , is the continuous distribution duration of the space where new energy is abundant or the spot price is low. The power distribution vector corresponding to is: (2) When , , the corresponding power distribution vector is meaningless, and all of them are set to 0, that is: (3) Based on the power distribution vector, model the shiftable load. .

[0022] There are only two situations for the shiftable load after participating in the response of the virtual power plant: a) Not shifted. b) Shifted to the period when new energy is in large output or the spot price is low. The constraints are as follows: (4) 2. Shiftable load constraints: For some virtual power plant building users, their power can be transferred, but the time requirement is not for continuous operation. It can be adjusted within a certain period and the total power remains unchanged. The shiftable load The power distribution vector is: (5) Use the 0-1 variable to represent the transfer status of the period , that is represents transferring in the period , represents not transferring. The power distribution vector of the load after transfer is: (6) The total power of the shiftable load remains unchanged before and after transfer. At time , the transferred load power is between the maximum and minimum transfer powers. To avoid the load being transferred to multiple discontinuous periods, that is, the frequent start and stop of the switches and power of air conditioners, refrigeration, heating and other electrical equipment, it is necessary to constrain the minimum continuous operation time of the transferred load, that is: (7) In the formula, is the start time, is the duration, , are respectively the maximum and minimum allowable transfer powers of the load at time ;

[0023] 3. Interruptible load constraints: For the load that can be curtailed, such as non-continuous production enterprises like steel and cement , as the participation response of virtual power plant users, the electricity consumption of users is reduced. Use the 0-1 variable ​Indicates load that can be curtailed or interrupted Time period Curtailment status, i.e., Indicates During The time period is curtailed, Indicates Not curtailed, the load model is: (8) Considering user satisfaction, it is necessary to constrain the minimum and maximum continuous curtailment times and the number of curtailments, i.e.: (9) In the formula, Is Time node Curtailment coefficient of the load that can be curtailed Is the minimum continuous curtailment time Is the maximum continuous curtailment time Is the maximum number of curtailments

[0024] Step 2: Considering the operation constraints of different types of regulation resources such as industrial loads, distributed energy storage, distributed photovoltaics, and industrial and commercial buildings connected to the virtual power plant, establish a typical database of the external characteristics of multi-type regulation resource classification and clustering based on the load regulation parameter classification model formed in Step 1

[0025] 1. Description of the operation constraints and external characteristics of industrial loads: (10) In the formula: Represents the total adjustable power formed by aggregating J industrial loads of virtual power plant i at time t, , , Respectively represent the th industrial load's equivalent adjustable load regulation power , transferable load , interruptible load Equivalent coefficient; Represents the th industrial load's adjustable capacity power constraint existing at time t during operation, , Respectively represent the th industrial load's corresponding main production load and auxiliary production load Adjustment equivalent coefficient

[0026] 2. Description of the operation constraints and external characteristics of distributed energy storage: 1) During a battery charge / discharge cycle, the energy storage capacity values of the battery at each moment should be maintained within a certain range: (11) Where: is the state of charge of the battery's energy storage at time , correspond to the upper and lower limits of the adjustable capacity of the battery's energy storage respectively.

[0027] 2) State of charge continuity constraint: (12) Where: , are the state of charge of the energy storage system at time t and time t-1 respectively. , are the charging power and discharging power of the energy storage system at time t respectively; and correspond to the charging efficiency and discharging efficiency of the battery's energy storage respectively. is the time interval ( =1h).

[0028] 3) Energy storage charge / discharge constraint: During operation, the battery's energy storage should be controlled so that the charge / discharge power per time does not exceed its rated value, and the total discharging power does not exceed the rated capacity of the energy storage: (13) Where: 、 represent the maximum charging power and maximum discharging power of the energy storage system respectively.

[0029] 3. Distributed PV operation constraints and external characteristic description: The operation constraint of the distributed PV is the power generation power that changes with time, and it participates in the aggregation response in the form of an interruptible load: (14) Where: , represent the predicted values of the power generation power and the maximum power generation power of the distributed PV system at time t respectively; is the equivalent interruptible load power value of the distributed PV.

[0030] 4. Industrial and commercial building operation constraints and external characteristic description: 1) Air conditioning system: Central air conditioners are installed in public buildings, with a load ratio of more than 40% and an adjustable potential ratio of 5% - 20%. Assuming that the central air conditioning system in the building has been operating stably before the implementation of demand response, considering the comfort of the indoor environment, the adjustable load constraint model of the central air conditioner is as follows: (15) In the formula: is the maximum active power that can be reduced or increased by the central air conditioning system; is the duration of demand response; when in the heating mode, is the heating energy efficiency ratio of the central air conditioning system, and when in the cooling mode, is the cooling energy efficiency ratio of the central air conditioning system; c is the specific heat capacity of air, and the value of the specific heat capacity at constant pressure of air at a temperature of 300K is selected as 1.005 kJ / (kg·K); S is the heating area of the building; H is the average floor height of the building; is the density of air, and the density of dry air at a temperature of 300K is selected, with a value of 1.177 kg / m 3 ; is the heating temperature set by the user; is the allowable indoor temperature.

[0031] Among them, the shiftable load under different control modes of the air conditioner is equivalent to the following four scenarios: (16) In the formula, is the response capacity of the input current percentage of the control host. From receiving the response instruction to completing the power adjustment, the response preparation time is at the minute level, is the correction parameter; is the operating power of the air conditioner host; is the rated current loading value of the host; is the operating output value of the host; is the response capacity when remotely or locally controlling the water temperature. The response preparation time is at the minute level, is the correction parameter; is the temperature adjustment value; is the response capacity for adjusting the compressor speed in the variable frequency control mode, is the frequency change value; is the rated frequency; is the response capacity when directly controlling the shutdown of some air conditioners.

[0032] 2) Electric boiler: Regenerative electric boilers are generally used for domestic hot water supply or heating in office buildings, hotels, shopping malls and other large public buildings. Their load ratio can reach 30% - 40% in winter, and the adjustable potential is up to 15%. When the electric boiler has no heat storage capacity, the calculation method of its constraint model is the same as formula (15). When the electric boiler has heat storage capacity, the operation constraints are as follows: (17) In the formula: is the heat change of the heat storage medium after t moments of response, is the temperature of the heat storage medium at the moment when the response is issued, is the specific heat capacity of the medium; is the density of the medium; V is the volume of the medium; is the temperature of the medium at moment t; In the constraint conditions of the adjustable load model is the maximum active power that can be reduced or increased by the regenerative electric boiler; t′ is the duration of the demand response; is the efficiency of the electric boiler, is the heat exchange efficiency of the heat storage medium; is the mass flow rate of the boiler, is the water temperature at the secondary side outlet; is the water temperature at the secondary side inlet.

[0033] 3) Lighting system: The load ratio of the lighting system is 15% - 25%, and the adjustable potential ratio is 3% - 10%. Its response preparation time is at the second level. The response capacity comes from two aspects. One is to ensure the normal work and production of users, and the method of sectional area control is adopted to turn off a part of the lighting equipment P1. The other is the lamps that could have been turned off but were not turned off during the non-working period P2: (18) In the formula: P light is the power of the lamps that were originally turned on during work; λ1 is the reasonable turn-off coefficient obtained by optical measurement; P i is the lamps that can be turned off during the non-working period obtained by statistics.

[0034] 4) Elevators and cold storages: The load ratio of the elevator and cold storage system is 4% - 10%, and the adjustable potential ratio is 1% - 5%. Its response preparation time is at the second level. The response capacity comes from turning off some idle elevators at night and shutting down the cold storage for a short time during the non-preparation time: (19) In the formula, and respectively represent the equivalent interruptible elevator load and the cold storage refrigeration system load of the building load.

[0035] Step 3: Based on the typical database formed by different types of regulation resources and Step 2, calculate the typical classification marginal cost of the response of different regulation resources of virtual power plant users after clustering: (20) In the formula, represents the discount rate represents the inflation rate represents the service life of the regulation resource represents the unit power retrofit cost of regulation resource j; represents the annual maintenance cost per unit power of regulation resource j represents the retrofit capacity of regulation resource j represents the operating capacity of regulation resource j represents the total regulation cost aggregated by a total of J regulation resources; c represents the typical regulation cost corresponding to the typical regulation power interval p.

[0036] Step 4: For the four typical combined scenarios of wind, light, large, and small, based on the improved K-means clustering method, conduct clustering analysis on the spot market price and classification model price spread interval of the high-proportion new energy system.

[0037] For the historical data scenario set p = {x1, x2, x3,..., x n}, where there are a total of n data samples (365 * 24), each sample is m-dimensional, assuming that the n data samples are divided into K classes, and the sample c i represents the center point of class K i , N(K i ) represents the number of samples in class Ki, represents the distance between samples, then (21) In the formula, r i is the class radius, which is the average value of the distances from all data within the class to the class center; λ is the ratio of class radii; C ij is the midpoint between classes, which is the point that divides the line connecting different class centers according to the ratio of class radii; is the class center density, which is the number of data within the class whose distance from the class center point is less than the class radius; is the class edge density, representing the class dispersion, that is, the number of data whose distance from the midpoint between classes is less than the average class radius among any two classes; The index is the ratio of the data density within the class and the data density between classes; is the optimal number of clusters for K-means clustering determined based on the DBI index; the spot price and peak-valley price spread formed after clustering are the corresponding typical libraries (X K1 , X K2 , XK3 ,...., X Kbest ; C K1 , C K2 , C K3 ,...., C Kbest ).

[0038] Step Five: Based on the long - term prediction of spot market prices and the marginal costs of typical classifications, establish an analysis model for the adjustable potential of virtual power plant users in typical scenarios of the long - term spot model: (22) In the formula, , , are the adjustable power that can be shifted, transferred, and interrupted participated by the virtual power plant respectively. For the adjustable capabilities of the shiftable and transferable types, when the spot price is higher than the cost threshold of the i th adjustable power interval, all the adjustable capabilities of this interval can be included in the response capability interval; for the interruptible type of adjustable capability, when the spot price is lower than its cost threshold, all its adjustable capabilities can be included in the response capability interval.

[0039] The above only elaborates on the preferred embodiments of the present invention in detail. However, the present invention is not limited to the above - mentioned embodiments. Within the scope of knowledge possessed by those of ordinary skill in the art, various changes can be made without departing from the purpose of the present invention, and all such changes should be included within the protection scope of the present invention.

Claims

1. A method for analyzing the regulation potential of a virtual power plant in a spot mode, characterized in that: The following steps are involved: S1. Aiming at the two typical demands of absorbing new energy and ensuring supply, a load regulation parameter classification model of three types, namely, shiftable, transferable and interruptible, is established for users of virtual power plants to participate in the response; S2. Considering the operation constraints of different types of regulation resources such as industrial loads, distributed energy storage, distributed photovoltaics, and industrial and commercial buildings connected to the virtual power plant, a typical database of external characteristics of multi-type regulation resources classification and clustering is established based on the load regulation parameter classification model formed in S1; S3, based on different types of regulation resources and the typical database formed by S2, calculate the typical classified marginal costs of different regulation resource responses of clustered virtual power plant users; S4. Based on the improved K-means clustering method, the spot market price of high-proportion new energy systems and the price spread range of the classification model are clustered for four typical combination scenarios of wind, solar, large and small. S5. Based on the long-term forecast of spot market prices and typical classified marginal costs, a user adjustable potential analysis model for virtual power plants under typical scenarios of long-term spot models is established.

2. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The method for establishing the classification model of load adjustment parameters for virtual power plant user participation response in S1 is: For users of virtual power plants, the requirements for steel and glass as shiftable loads are that they are continuous in time and that the power level remains constant at every moment before and after the shift. , its power distribution vector is: （1） in, is the starting time, For duration period; use 0-1 variables express Time The initial state, that is express from The time period starts to shift; Starting period collection for: ; Indicates that the load is not translated; and , Indicates that the load is shifted from start, The duration of continuous distribution of new energy or low spot prices; The corresponding power distribution vector for: （2） when hour, , the corresponding power distribution vector is meaningless, and all of them are set to 0, that is: （3） Modeling of translatable loads based on power distribution vectors, ; There are only two situations for the load that can be shifted after the virtual power plant participates in the response: a) not shifted; b) shifted to the time interval of the new energy boom or the spot low price interval, with the following constraints: (4).

3. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The method for establishing the transferable load regulation parameter classification model for virtual power plant user participation response in S1 is: For users of virtual power plants and buildings, their power can be transferred, but the time requirement is not uninterrupted. It can be adjusted within a certain period of time and the total power remains unchanged. The load can be transferred The power distribution vector is: （5） Using 0-1 variables express Time The transition state, that is express exist Time period transfer, express No transfer; load after transfer The power distribution vector is: （6） The total power remains unchanged before and after the load transfer. The transfer load power is between the maximum and minimum transfer power values. To avoid the load being transferred to multiple discontinuous periods, i.e., frequent start and stop of the switches and power of air conditioners, refrigeration and heating electrical equipment, the minimum continuous operation time of the transfer load needs to be constrained, i.e.: （7） in, is the starting time, is the duration, , They are The maximum and minimum values ​​of power transfer allowed by the load at any moment; is the minimum continuous running time.

4. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The method for establishing the interruptible load regulation parameter classification model for virtual power plant user participation response in S1 is: For non-continuous production enterprises of steel and cement, the load can be reduced , as a response to the participation of users in the virtual power plant, reducing the user's electricity consumption; using 0-1 variables Indicates that load can be reduced or interrupted Time The reduction state, that is express exist The time slots were cut. express Without reduction, the load model is: （8） Considering user satisfaction, it is necessary to constrain the minimum and maximum continuous reduction time and the number of reductions, namely: （9） in, for Time Node The reduction factor of the load that can be reduced; is the minimum continuous reduction time; is the maximum continuous reduction time; is the maximum number of reductions.

5. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The operation constraints of different types of regulation resources connected to the virtual power plant in S2, such as industrial load, distributed energy storage, distributed photovoltaic, and industrial and commercial buildings, are as follows: Industrial load operation constraints and external characteristics description: （10） in: represents the total adjustable power formed by the aggregation of J industrial loads of virtual power plant i at time t, , , Respectively represent Industrial load equivalent load transfer power , transferable load , interruptible load Equivalence coefficient; Indicates The power constraint of the regulation capability of an industrial load at time t during operation is: , Respectively represent The industrial load corresponds to the main production load at time t in operation and auxiliary production load The adjustment equivalent coefficient of Distributed energy storage operation constraints and external characteristics description: 1) During a battery charge / discharge cycle, the battery energy storage capacity value at each moment should be maintained within a certain range: （11） in: for Battery energy storage charge status at all times; , They correspond to the upper and lower limits of the dispatchable capacity of battery energy storage respectively; 2) Energy storage charge state continuity constraints: (12) in: , are the charge states of the energy storage system at time t and time t-1 respectively; , are the charging power and discharging power of the energy storage system at time t respectively; and They correspond to the charging efficiency and discharging efficiency of battery energy storage respectively; is the time interval ( =1h); 3) Energy storage charging / discharging constraints: During the operation of battery energy storage, the charging / discharging power should be controlled not to exceed its rated value each time, and the total discharge power should not exceed the rated capacity of the energy storage: (13) in: , Respectively represent the maximum charging power and maximum discharging power of the energy storage system; Distributed photovoltaic operation constraints and external characteristics description: The operating constraints of distributed photovoltaics are the power that can be generated that varies with time, and they participate in the aggregate response as interruptible loads: (14) in: , They represent the power generation and maximum power generation prediction of the distributed photovoltaic system at time t respectively; is the equivalent interruptible load power value of distributed photovoltaics; Description of operation constraints and external characteristics of industrial and commercial buildings: 1) Air conditioning system: Centralized air conditioning is installed in public buildings, with a load ratio of more than 40% and an adjustable potential ratio of 5% to 20%. Assuming that the centralized air conditioning system in the building is already in a stable operating state before the implementation of demand response, considering the comfort of the indoor environment, the adjustable load constraint model of the centralized air conditioning is: （15） in: The maximum active power that can be reduced or increased for the centralized air conditioning system; is the duration of demand response; when it is heating condition, is the heating energy efficiency ratio of the centralized air conditioning system. When it is in cooling condition, is the cooling energy efficiency ratio of the centralized air conditioning system; c is the specific heat capacity of air, which is the constant pressure specific heat capacity of air at a temperature of 300K, and the value is 1.005 kJ / (kg·K); S is the heating area of ​​the building; H is the average floor height of the building; is the density of air, and the dry air density at a temperature of 300K is selected, which is 1.177 kg / m 3 ; Heating temperature set for the user; is the permissible indoor temperature; Among them, the shiftable load under different air conditioning control modes is equivalent to the following four scenarios: （16） in, The response capacity of the percentage of the input current of the control host is controlled. The response preparation time from receiving the response command to completing the power adjustment is in minutes. is the correction parameter; It is the operating power of the air conditioner host; Limit the current loading value for the host; Output value for host operation; The response capacity when controlling the water temperature remotely or locally, the response preparation time is in minutes. is the correction parameter; is the temperature adjustment value; To control the response capacity of adjusting the compressor speed in variable frequency mode, is the frequency change value; is the rated frequency; To directly control the response capacity when shutting down some air conditioners; 2) Electric boiler: Thermal storage electric boilers are used for domestic hot water or heating in large public buildings such as office buildings, hotels, and shopping malls. Their load share can reach 30% to 40% in winter, and their adjustable potential can reach 15%. When the electric boiler has no thermal storage capacity, the calculation method of its constraint model is the same as formula (15); when the electric boiler has thermal storage capacity, the operation constraint is: （17） in: In response to the heat change of the heat storage medium after t time, In response to the temperature of the heat storage medium at the time of issuing the command, is the specific heat capacity of the medium; is the density of the medium; V is the volume of the medium; is the temperature of the medium at time t; in the constraints of the adjustable load model is the maximum active power that can be reduced or increased by the thermal storage electric boiler; t′ is the duration of demand response; is the efficiency of the electric boiler, is the heat transfer efficiency of the heat storage medium; is the mass flow rate of the boiler, is the secondary side outlet water temperature; is the secondary side inlet water temperature; 3) Lighting system: The lighting system load accounts for 15% to 25%, and the adjustable potential accounts for 3% to 10%. Its response preparation time is in seconds. The response capacity comes from two aspects. One is to ensure the normal work and production of users by using the method of split-way area control to turn off part of the lighting equipment P1. The other is to turn off the lamps P2 that could have been turned off but were not turned off during non-working hours. （18） Where: P light is the power of the lamp originally turned on during operation; λ1 is the reasonable closing coefficient obtained by optical measurement; P i To obtain statistics of lamps that can be turned off during non-working hours; 4) Elevator and cold storage: The load of elevator and cold storage systems accounts for 4% to 10%, and the adjustable potential accounts for 1% to 5%. The response preparation time is in seconds. The response capacity comes from shutting down some idle elevators at night and shutting down cold storage for a short time during non-food preparation time. （19） in, and They represent the equivalent interruptible elevator load of building loads and the refrigeration system load of cold storage respectively.

6. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The method for calculating the typical classification marginal cost of different regulation resource responses of clustered virtual power plant users in S3 is: （20） in, represents the discount rate; represents the inflation rate; Indicates the useful life of the adjustment resource; represents the unit power transformation cost of regulating resource j; represents the annual maintenance cost per unit power of regulation resource j; represents the transformation capacity of regulating resource j; represents the operating capacity of the regulation resource j; represents the total adjustment cost of J adjustment resources; c represents the typical adjustment cost corresponding to different typical adjustment power intervals p.

7. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The method for carrying out cluster analysis of spot market prices and price spread intervals of classification models of high-proportion new energy systems in S4 is: Treat the cluster classification historical data scene set p={x1,x2,x3,...,x n }, there are n data samples (365*24), each sample is m-dimensional, assuming that n data samples are divided into K categories, with sample c i Represents class K i The center point, N(K i ) represents the number of samples in class Ki, represents the distance between samples, then: （21） Among them, r i is the class radius, which is the average distance from all data in the class to the class center; λ is the ratio of the class radius; C ij is the middle point between classes, which is the point where the line connecting the centers of different classes is equally divided according to the ratio of the class radius; is the class center density, which is the number of data within the class whose distance from the class center point is less than the class radius; is the class edge density, which indicates the inter-class dispersion, that is, the number of data between any two classes whose distance to the inter-class median point is less than the mean of the class radius; The indicator is the ratio of intra-class data density to inter-class data density; The optimal number of clusters for K-means clustering is determined based on the DBI index; the spot price and peak-valley price difference formed after clustering are the corresponding typical libraries (X K1 , X K2 , X K3 , ..., X Kbest ; C K1 , C K2 , C K3 , ..., C Kbest ).

8. The method for analyzing the regulation potential of a virtual power plant in a spot mode according to claim 1, characterized in that: The method for establishing the virtual power plant user adjustable potential analysis model in the typical scenario of the long-term spot mode in S5 is: （22） in, , , They are the shiftable, transferable and interruptible regulation power of the virtual power plant. For the shiftable and transferable regulation capabilities, when the spot price is higher than the i When the cost threshold of a power adjustment interval is reached, all adjustment capabilities in the interval can be included in the response capability interval; for interruptible type adjustment capabilities, when the spot price is lower than its cost threshold, all adjustment capabilities can be included in the response capability interval.

Images