CN114997715A

CN114997715A - Improved fuzzy C-means clustering-based combined cooling, heating and power system configuration method

Info

Publication number: CN114997715A
Application number: CN202210746304.7A
Authority: CN
Inventors: 李桐; 李青春; 张建; 唱友义; 刘淼; 李健; 梁晓赫; 张晔; 夏楠楠; 石泽文; 黄博南; 孙赫阳; 孙茜
Original assignee: State Grid Corp of China SGCC; State Grid Liaoning Electric Power Co Ltd; Electric Power Research Institute of State Grid Liaoning Electric Power Co Ltd
Current assignee: State Grid Corp of China SGCC; State Grid Liaoning Electric Power Co Ltd; Electric Power Research Institute of State Grid Liaoning Electric Power Co Ltd
Priority date: 2022-06-29
Filing date: 2022-06-29
Publication date: 2022-09-02

Abstract

The invention provides a cold-heat-electricity combined supply system configuration method based on improved fuzzy C-means clustering, and relates to the technical field of composite energy supply. The method improves the FCM algorithm, introduces a PFS index for evaluating the effectiveness of a data set geometric structure and a Vp index for evaluating the effectiveness of membership degree on the basis of the traditional FCM clustering algorithm, uses an entropy weight method to weight the PFS index and the Vp index for evaluating the effectiveness of membership degree so as to comprehensively evaluate the clustering effect, and then searches the optimal clustering number, the optimal fuzzy coefficient and the clustering result under the optimal parameters by a traversal method so as to achieve the purpose of optimal clustering. And (3) taking a meteorological and load data clustering result obtained based on an improved FCM algorithm as input, combining an NSGA-II algorithm and a PSO algorithm to complete solving, and realizing optimal CCHP equipment capacity configuration. The invention can obviously reduce the system operation cost and the carbon dioxide emission, and better exert the advantages of high energy utilization rate, energy saving and environmental protection of the CCHP system.

Description

Improved fuzzy C-means clustering-based combined cooling, heating and power system configuration method

Technical Field

The invention relates to the technical field of composite energy supply, in particular to a cold-heat-electricity combined supply system configuration method based on improved fuzzy C-means clustering.

Background

The development of human beings can not leave energy sources, and with the overuse of primary energy sources such as coal, petroleum and the like, the traditional energy source utilization structure brings rapid development to human society and simultaneously causes a series of development problems. First, coal, petroleum and other primary energy sources are about to be exhausted; secondly, the establishment of an energy utilization structure which overuses such energy sources brings great challenges to the ecological environment, such as air pollution, greenhouse effect and the like. Therefore, the problems of improving the energy utilization rate, reducing the emission of pollutants and carbon, and developing clean energy are important to research in various countries. A Combined Cooling, Heating and Power (CCHP) system is more and more favored by researchers due to the advantages of high energy utilization rate, energy saving, environmental protection, off-network operation and the like, and becomes a technical hotspot promoted by various countries.

The CCHP system is a composite energy supply system which is centered on a combined supply device, contains various units of energy supply, storage, conversion and the like, and is based on three energy flows of cold, heat and electricity. The CCHP system starts from the realization of multi-level utilization of energy, converts primary energy into electric energy through the generator, then further utilizes waste heat to generate electricity by using devices such as an absorption refrigerator and a heat exchange plate exchanger, and simultaneously provides multiple energy forms such as electric energy, heat energy and cold energy for users, so that the full utilization of the energy is promoted, and the purposes of reducing energy cost and reducing environmental pollution are achieved. In order to fully exert the advantages of good economy and high energy utilization rate of the CCHP system, the optimal matching of the source load and the load must be realized. The cold, heat and electric loads of the CCHP system are constantly changed, and if the capacity of each device of the system is unreasonable, reasonable utilization of energy cannot be realized, and energy waste can be caused. Therefore, on the basis of actual source load data, possible source load characteristic conditions are comprehensively considered, optimal configuration of the capacity of each device of the CCHP system is achieved, dynamic balance of the source load is achieved, and the method is an important prerequisite for efficient operation of the system.

The optimization of the CCHP configuration needs to consider a plurality of factors, and in the optimization model, long-term relevant data is generally needed to be evaluated, and the data is usually a large amount of meteorological data, load data and the like. However, the large amount of data can guarantee the accuracy of the optimization result as much as possible, but the calculation amount is greatly increased. In order to overcome the problem, a method for reducing scenes by using a clustering algorithm and replacing a large number of scene features with a small number of scenes is provided in the existing research. Among them, Fuzzy C-means clustering (FCM) is a commonly used clustering method. However, the FCM algorithm also has disadvantages, mainly: the clustering number of the algorithm needs to be manually set in advance; the quality of the clustering result is closely related to the size of the fuzzy coefficient, and the like. The ambiguity is a key parameter of the FCM algorithm, which reflects the ambiguity of the clustering result. Most of the current documents related to FCM set the ambiguity parameter as a default value, but the optimal ambiguity value is not necessarily a default value for different clustered objects. In addition, the different values of the cluster number also have influence on the subsequent calculation. Too many clusters are not favorable for solving calculation, and too few clusters are not representative. The clustering validity index is generally used for measuring the quality of a clustering result.

According to different targeted evaluation emphasis points, the clustering effectiveness index can be generally divided into an evaluation index considering the geometric structure information of the data set and an evaluation index considering the membership degree. The mentioned geometrical information of the data set specifically includes the degree of closeness and dispersion, the degree of dispersion and independence of the classification results, and so on. Currently, widely used evaluation indexes considering the geometry of a data set include: dunn index, CH index, G index, CI index, DB index, Sil index, I index, CS index, SF index, COP index, SV index, OS index, and the like. The indexes only evaluate the geometric distribution attribute of the clustering result, wherein most of the indexes are not exactly specified in application occasions and have good effect on a general data set, but the indexes cannot always obtain better clustering results on complex data sets with more outliers and larger data cross. In addition, the indexes only consider the geometrical structure of the data and do not relate to the evaluation of membership and fuzzy coefficient information, so the evaluation of the fuzzy clustering method is not very exact.

The fuzzy clustering effectiveness indexes considering the membership degree mainly comprise: the standard separation coefficient NPC and the standard separation drop NPE index, KYI index, the separation coefficient PC index and the separation drop PE index, VP index, OS index and the like, which only consider the membership degree information in the clustering division, have the following defects: first, these indicators generally tend to increase with increasing number of clusters; secondly, the blurring coefficient has a large influence on these indices; third, such indicators cannot evaluate the data geometry because they do not involve evaluation of the geometric information of the clustering results. Although most of the indexes considering the membership degree have the advantages of simplicity and easiness in calculation, the indexes are not ideal enough in measuring the effectiveness of clustering results.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a cold-heat-electricity combined supply system configuration method based on improved fuzzy C-means clustering, a PFS index and a Vp index are introduced on the basis of the traditional FCM clustering algorithm, and the PFS index and the Vp index are weighted by using an entropy weight method to comprehensively evaluate the clustering effect, so that the system operation cost and the carbon dioxide emission can be obviously reduced, and the advantages of high energy utilization rate, energy conservation and environmental protection of a CCHP system can be better exerted.

In order to solve the technical problems, the technical scheme adopted by the invention is as follows:

a cold-heat-electricity combined supply system configuration method based on improved fuzzy C-means clustering comprises the following steps:

step 1: the traditional fuzzy C-means clustering algorithm is improved, a PFS index considering the clustering effectiveness of the geometric structure information and a Vp index considering the clustering effectiveness of the membership degree are introduced on the basis of the original algorithm at the same time, and the PFS index and the Vp index are combined to form a comprehensive clustering evaluation index by using an entropy weight weighting form; and clustering the air temperature, the illumination intensity and the electricity/heat/cold load data through an improved FCM algorithm to obtain typical scenes and probability data of each scene in three seasons of a transition season, a refrigeration season and a heating season, and taking the obtained data result as the input of a CCHP optimization configuration model.

And 2, step: and constructing a CCHP system double-layer optimization configuration model based on an improved FCM algorithm. The outer layer is an optimization layer, and the aim is to minimize the annual average converted investment and operation cost of the system and minimize the annual carbon dioxide emission; the inner layer is an operation optimization layer and aims at minimizing daily operation cost; the inner layer model and the outer layer model are mutually nested.

And step 3: and (3) combining a fast non-dominated sorting genetic algorithm (NSGA-II) and a particle swarm optimization algorithm (PSO) to complete the solution of the CCHP system double-layer optimization configuration model based on the improved FCM algorithm, so as to obtain the optimal CCHP equipment capacity configuration.

The method for clustering the air temperature, the illumination intensity and the electricity/heat/cold load data in the step 1 is the same, and the clustering method is as follows:

step 1.1: acquiring annual hourly air temperature, illumination intensity or hotel electricity/heat/cold load data of the location of the system, recording the number of certain type of input data, recording the number of certain type of data as n, and setting a fuzzy coefficient m and a clustering number c;

step 1.2: clustering the input data set by using an improved FCM algorithm to obtain a clustering result S;

step 1.3: calculating PFS indexes and Vp indexes of the clustering result S, and respectively storing the PFS indexes and the Vp indexes into vectors y1 and y 2;

step 1.4: increasing the clustering number c by 1, and repeating the steps 1.2-1.3 until the clustering number is more than the clustering number

Then obtaining PFS index vector y1 and Vp index vector y2 with different clustering numbers under the same fuzzy coefficient, and obtaining yl and y2 by per unit of y1 and y2 according to the respective maximum values;

step 1.5: calculating weights wl and w2 of the two indexes by using y1 and y2 according to an entropy weight method;

step 1.6: calculating a comprehensive evaluation index vector PFS _ Vp ═ wl × y1 × w2 × y2 corresponding to the fuzzy coefficient m, and storing the comprehensive evaluation index vector PFS _ Vp in a matrix PFS _ P;

step 1.7: increasing the fuzzy coefficient m by 0.1, repeating the step 1.2 to the step 1.6 until the fuzzy coefficient is more than 5, and obtaining a comprehensive evaluation index matrix PFS _ P under different fuzzy coefficients;

step 1.8: m and c corresponding to the maximum comprehensive evaluation index in the matrix PFS _ P are the optimal fuzzy coefficient and the optimal clustering number.

The specific method of the step 2 comprises the following steps:

step 2.1: determining the types of various energy conversion equipment in the CCHP system, wherein the energy conversion equipment comprises a gas internal combustion engine, an energy storage device, an absorption refrigerating unit, a photovoltaic cell, an electric refrigerator, a gas boiler and a heat exchange device, and setting the connection mode among the various energy conversion equipment; the energy storage device comprises a storage battery and a heat storage tank; and then performing mathematical modeling on each energy conversion device respectively.

Step 2.2: and constructing an overall energy flow calculation mathematical model of the CCHP system according to the connection structure.

Step 2.3: obtaining a double-layer optimized configuration model of the CCHP system according to typical scenes of each data in three seasons, probability data of each scene and the constructed CCHP system mathematical model, wherein the typical scenes are obtained by improving an FCM algorithm;

the rated installation capacity of photovoltaic power generation, the rated installation capacity and the rated maximum charging and discharging power of the storage battery and the rated installation capacity of the heat storage tank are used as decision variables, and the annual reduced investment cost and the annual operation maintenance cost f are used as decision variables ₁ Annual carbon dioxide emission f ₂ Minimizing as a multi-objective function, and taking the running capacity of the CCHP system equipment as a constraint condition; establishing an outer layer configuration optimization model of the CCHP system by combining system data, equipment parameters and scene data;

on the basis of the capacity configuration of each device of the system given by the outer layer model, selecting the power generation power, the waste heat distribution ratio, the charge and discharge power of the storage battery and the heat storage and discharge power of the heat storage tank of the gas internal combustion engine at each time period of one day as decision variables, aiming at the minimum daily operation and maintenance cost of the system, and taking decision variable operation constraint and energy balance as constraint conditions; and establishing an inner layer operation optimization model of the CCHP system by combining system data, equipment parameters and scene data.

In step 3, the solving process of the double-layer model is as follows:

firstly, randomly generating a population by an outer layer model, and conveying the population to an inner layer model;

secondly, the inner layer model uses a PSO algorithm to solve an optimal operation scheme based on the equipment capacity information input by the outer layer according to meteorological and load data, and returns the optimal operation scheme to the outer layer model;

then, the outer layer model carries out non-dominated sorting and congestion degree calculation according to the operation scheme returned by the inner layer model, and generates a new population through selection, intersection and variation;

and finally, conveying the new population to the inner layer model, and performing cyclic substitution calculation. And when the number of selected generations reaches a specified value, stopping circulation and outputting the final generation of population.

The process for solving the outer layer multi-target capacity configuration problem based on NSGA-II is as follows:

step 3.1: the rated installation capacity of photovoltaic power generation, the rated installation capacity and the rated maximum charging and discharging power of the storage battery and the rated installation capacity of the heat storage tank are used as decision variables, and the annual reduced investment cost and the annual operation maintenance cost f are used as decision variables ₁ Annual carbon dioxide emission f ₂ The minimum mathematical model is used as an optimization objective function of the outer layer model, and the operation capacity of the CCHP system equipment is used as a constraint condition;

step 3.2: initializing the system by taking the equipment type, capacity, electricity/heat/cold load demand, wind speed/illumination intensity and energy price in the system as input parameters;

step 3.3: generating an initialized population P by using a multi-target function and input parameters in an outer layer model, and setting the population iteration number N to be 0 and the maximum iteration number to be D;

step 3.4: judging whether a first generation subgroup Q is generated or not, and if not, obtaining the first generation subgroup Q by utilizing three basic operations of selection, intersection and variation of a genetic algorithm after non-dominated sorting; if Q is generated, continuing the next step;

step 3.5: setting the evolution iteration number gen to be 2;

step 3.6: combining the parent population P with the child population Q, judging whether a new parent population Qt is generated, if not, simultaneously carrying out congestion degree calculation on individuals in each non-dominant layer through rapid non-dominant sorting, and finally selecting proper individuals according to the non-dominant relationship and the congestion degree of the individuals to form a new parent population Qt; if Qt is generated, continuing the next step;

step 3.7: carrying out three basic operations of genetic algorithm selection, crossover and mutation on the newly generated new father population Qt;

step 3.8: and judging a termination condition, terminating when the maximum iteration number D is reached, and outputting a Pareto optimal solution set. Selecting multiple groups of typical optimal solutions from the obtained Pareto optimal solution set to obtain capacity configuration schemes corresponding to the multiple groups of typical optimal solutions, and outputting an optimal equipment combination scheme, reduced investment cost and annual operation maintenance cost f ₁ Annual carbon dioxide emission f ₂ (ii) a Otherwise, the step 3.6 is returned.

The flow of solving the inner-layer optimal operation scheme problem based on the PSO is shown in fig. 5, and is specifically described as follows:

step 3.9: taking the lowest daily operation and maintenance cost in the inner-layer optimization operation model as an objective function, and taking decision variable operation constraint and energy balance constraint as constraint conditions;

step 3.10: taking the electricity/heat/cold load requirements, the wind speed/illumination intensity and the equipment capacity configured on the outer layer in the system as input parameters to carry out system initialization;

step 3.11: generating an initialization population P by using an objective function and input parameters in an inner layer model, setting the population iteration number R to be 0, setting the maximum iteration number to be G, setting the initial position W of each particle to be 0, and setting the initial speed V of each particle to be 0;

step 3.12: taking the operation and maintenance cost in the population S as a fitness function, then, taking the fitness function as a reference, evaluating the state of each particle, and selecting the optimal particles in the population and the individual optimal particles;

step 3.13: updating the historical optimal position of each particle;

step 3.14: updating the global optimal position of the group;

step 3.15: updating each particle velocity V ^G+1 And position W ^G+1 ；

Step 3.16: and judging a termination condition, terminating when the maximum iteration number G is reached, and outputting an optimal operation scheme.

Adopt the produced beneficial effect of above-mentioned technical scheme to lie in: the invention provides a cold-heat-electricity combined supply system configuration method based on improved Fuzzy C-Means clustering, which aims at solving the problem that Fuzzy coefficients and clustering numbers are difficult to determine in Fuzzy C-Means (FCM) clustering algorithm and improves the FCM algorithm. PFS indexes for evaluating the effectiveness of the geometric structure of the data set and Vp indexes for evaluating the effectiveness of membership degrees are introduced on the basis of a traditional FCM clustering algorithm, an entropy weight method is used for weighting the PFS indexes and the Vp indexes to comprehensively evaluate the clustering effect, and then the optimal clustering number, the optimal fuzzy coefficient and the clustering result under the optimal parameters are searched through a traversal method, so that the purpose of optimal clustering is achieved. And (3) taking a meteorological and load data clustering result obtained based on an improved FCM algorithm as input, and combining an NSGA-II algorithm and a PSO algorithm to complete solving, so that the optimal CCHP equipment capacity configuration is realized. Through the comparison of system operation before and after configuration, the reasonable photovoltaic power generation and electricity and heat energy storage configuration of the CCHP system based on the improved FCM algorithm is very necessary, the system operation cost and the carbon dioxide emission can be obviously reduced, and the advantages of high energy utilization rate, energy conservation and environmental protection of the CCHP system are better exerted.

Drawings

FIG. 1 is a flow chart of a calculation of an improved fuzzy C-means clustering algorithm according to an embodiment of the present invention;

FIG. 2 is a diagram of the architecture of a CCHP system with integrated electro-cooling-heating functions provided by an embodiment of the present invention;

fig. 3 is a flowchart of computing a double-layer optimized configuration model of a CCHP system according to an embodiment of the present invention;

FIG. 4 is a flowchart of a solution of an optimization model for the outer layer configuration of a CCHP system based on a genetic algorithm (NSGA-II) according to an embodiment of the present invention;

fig. 5 is a flowchart of solving an optimization model for inner layer operation of a CCHP system based on Particle Swarm Optimization (PSO) according to an embodiment of the present invention;

fig. 6 is a result of overall clustering of load and meteorological data based on the improved FCM according to the embodiment of the present invention; wherein, the graph (6a) is a cold load clustering result graph, the graph (6b) is a heat load clustering result graph, the graph (6c) is an electric load clustering result graph, the graph (6d) is an air temperature clustering result graph, the graph (6e) is an illumination intensity clustering result graph, and the graph (6f) is a probability statistical graph of each scene;

FIG. 7 is a FCM-based load and meteorological data overall clustering result according to an embodiment of the present invention; wherein, the graph (7a) is a cold load clustering result graph, the graph (7b) is a heat load clustering result graph, the graph (7c) is an electric load clustering result graph, the graph (7d) is an air temperature clustering result graph, the graph (7e) is an illumination intensity clustering result graph, and the graph (7f) is a probability statistical graph of each scene;

FIG. 8 is a graph showing the relationship between the fuzzy coefficient, the cluster number, and the comprehensive evaluation index according to the embodiment of the present invention;

fig. 9 shows Pareto frontiers and optimal solutions under the FCM and improved FCM clustering results provided in the embodiment of the present invention; wherein, the graph (9a) is a Pareto front comparison graph, and the graph (9b) is an optimal solution comparison graph;

FIG. 10 is a diagram illustrating multi-attribute decision-making indicator values for solutions based on the improved FCM algorithm according to an embodiment of the present invention;

FIG. 11 shows the daily operating cost and the total annual cost of a system in two configurations provided by an embodiment of the present invention; wherein, FIG. 11a is a daily operating cost comparison graph, and FIG. 11b is a yearly total cost comparison graph;

FIG. 12 shows the total carbon dioxide emissions per day and year around for the system in two configurations provided by an embodiment of the present invention; in this figure, 12a is a graph comparing the amount of carbon dioxide emitted per day, and 12b is a graph comparing the total amount of carbon dioxide emitted all year round.

Detailed Description

The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.

The invention provides a cold-heat-electricity combined supply system configuration method based on improved fuzzy C-means clustering. On one hand, the traditional FCM clustering algorithm is improved, and from the perspective of clustering effectiveness evaluation, clustering results are evaluated by introducing various evaluation indexes, so that the optimal clustering number and ambiguity parameters are searched, and the purpose of optimal clustering is achieved. And on the other hand, constructing a CCHP system double-layer optimization configuration model based on the improved FCM algorithm. The outer layer is an optimization layer, and the aim is to minimize the annual average converted investment and operation cost of the system and minimize the annual carbon dioxide emission; the inner layer is an operation optimization layer and aims at minimizing daily operation cost; the inner layer model and the outer layer model are mutually nested, and the NSGA-II and the PSO algorithm are combined to complete solution so as to realize the optimal CCHP equipment capacity configuration. The method of this example is specifically described below.

Step 1: the traditional fuzzy C-means clustering algorithm is improved, the improved FCM algorithm is used for clustering meteorological data such as air temperature and illumination intensity and electricity/heat/cold load data, typical scenes and probability data of each scene in three seasons of transition season, refrigerating season and heating season are obtained, the effectiveness of the improved algorithm is checked, and data input is provided for a CCHP optimization configuration model.

In order to comprehensively evaluate the advantages and disadvantages of the FCM clustering result, the embodiment introduces the PFS index considering the clustering effectiveness of the geometric structure information and the Vp index considering the clustering effectiveness of the membership degree on the basis of the original algorithm, and combines the PFS index and the Vp index by using the entropy weight weighting form to form the comprehensive clustering evaluation index.

PFS index: for samples of P (P >1) dimensional variables, a "pseudo-F statistical ratio" is defined as follows:

in the formula:

is a matrix

N is the number of samples, k is the number of classes,

are respectively asInter-class and intra-class scatter matrices of P-dimensional variable samples.

The expression of (a) is as follows:

in the formula: x is the number of _j Is the jth sample vector; v. of _i Is the ith class C _i The cluster center of (a); mu.s _ij The expression of (a) is as follows:

the relationship between the two is as follows:

in the formula:

is a mixed scatter matrix of P-dimensional variable samples.

The distance within the class is characterized,

the inter-class distance is characterized. In general, as the number of clusters k increases, the distance within a class will decrease and the distance between classes will increase. Furthermore, (n-k)/(k-1) increases with kAnd decreases. Therefore, as k increases, the PFS value increases and then decreases, i.e. the PFS may reach a maximum value at a certain k value, and this value is the best cluster number, i.e. finding the best classification number k is equivalent to finding the largest PFS value.

The VP index is defined as follows:

clustering validity index V _P Composed of two terms, the first reflecting compactness within a cluster, the b-th sample x _b The closer to the cluster center, the maximum membership max (μ) _ib ) The closer to the value 1. The second term reflects the degree of separation between clusters if x _b The closer to v _i Then, then

Close to 0. If it is not

Near 1/k, then x _b Belonging to all clusters with equal membership. Novel effectiveness index V _P Combining information of fuzzy closeness and separation. In the formula, mu _ib Representing membership functions of the ith sample in the class b cluster,

and representing membership functions of the ith sample in the class b cluster.

Entropy weight method: for the obtained matrix

u is the number of alternative schemes, g is the number of targets, and the entropy weight method comprises the following specific steps:

calculating a normalized matrix

Wherein:

calculating entropy of each target information

Wherein z is 1/ln (u) in the formula

Calculating respective target weights

As shown in fig. 1, the improved fuzzy C-means clustering method established in this embodiment includes constructing fuzzy validity indicators, obtaining a comprehensive evaluation clustering result by using an entropy weight method, and searching 3 parts by traversing the optimal clustering number/ambiguity parameter;

taking air temperature data as an example, the overall algorithm flow based on the improved fuzzy C-means clustering is as follows:

step 1.1: acquiring annual hourly air temperature data of a certain year at a certain hotel location, recording the number of input air temperature data as n, setting a fuzzy coefficient m to be 1.1 and a clustering number c to be 2;

Then obtaining PFS index vector y1 and Vp index vector y2 with different clustering numbers under the same fuzzy coefficient, and obtaining yl and y2 by per-unit of y1 and y2 according to the respective maximum values;

step 1.6: calculating a comprehensive evaluation index vector corresponding to the fuzzy coefficient m, calculating a comprehensive evaluation index vector PFS _ Vp ═ wl × y1 × + w2 × y2 corresponding to the fuzzy coefficient m, and storing the comprehensive evaluation index vector PFS _ Vp in a matrix PFS _ P;

The clustering of the illumination intensity and the hotel electricity/heat/cold load data is the same as the clustering process of the air temperature data, and the related parameters of the air temperature in the clustering process of the air temperature data are changed into the related parameters of the illumination intensity or the hotel electricity/heat/cold load data.

Step 2: and constructing a CCHP system double-layer optimization configuration model based on an improved FCM algorithm. The outer layer is an optimization layer, and the aim is to minimize the annual average converted investment and operation cost of the system and minimize the annual carbon dioxide emission; the inner layer is an operation optimization layer and aims at minimizing daily operation cost; the inner layer model and the outer layer model are nested with each other.

The CCHP system is first modeled mathematically. The mathematical modeling research of the CCHP system is a premise and a basis for constructing an optimized configuration model and an optimized operation model of the CCHP system. Specifically, the method comprises two aspects: firstly, modeling research is carried out on main equipment such as a gas internal combustion engine, an absorption refrigerating unit and the like in a system; and the other is to carry out modeling research on a system energy flow calculation model according to a system connection structure on the basis of establishing an equipment model.

Step 2.1: determining the types of various energy conversion equipment in the CCHP system, wherein the energy conversion equipment comprises a gas internal combustion engine, an energy storage device (a storage battery and a heat storage tank), an absorption refrigerating unit, a photovoltaic battery, an electric refrigerator, a gas boiler and a heat exchange device, and setting the connection mode among the various energy conversion equipment; the variable operating condition characteristics of the gas internal combustion engine and the time sequence correlation characteristics of the energy storage device are taken into account.

A CCHP system architecture related to an electricity-cold-heat integrated function is established, as shown in fig. 2, a gas internal combustion engine is used as a main energy supply device, primary energy natural gas is converted into electric energy to meet the requirement of an electric load, and meanwhile, the generated waste heat can be respectively transmitted to an absorption refrigerator and a heat exchange device according to a waste heat distribution proportion for refrigeration and heating. When the generated waste heat can not bear refrigeration and heating loads, the shortage is borne by the electric refrigerator and the gas boiler respectively. Furthermore, photovoltaic power plants are also capable of power supply, which can be supplemented by purchasing mains electricity from the grid if the electrical load is still in short supply. Energy storage devices such as a storage battery and a heat storage tank reasonably realize energy storage and release according to a formulated operation scheme, and further improve the supply and demand matching degree of the system.

Mathematical model of gas internal combustion engine: the gas internal combustion engine is the core of the CCHP system, which converts primary energy natural gas into electric energy, and the generated waste heat is used for refrigeration and heating. The gas internal combustion engine has different power generation efficiency and waste heat efficiency under different load rates, has variable working condition characteristics, and has the following mathematical model:

in the formula: v _mt Is the natural gas consumption in um ³ ；P _mt The unit of output electric power is kW;

and

respectively, a minimum value and a maximum value of the output electric power; q _mt The unit is kW;

and

respectively is the minimum value and the maximum value of the waste heat power; eta _mtP 、η _mtQ Respectively power generation efficiency and waste heat efficiency; l is _gas For the natural gas calorific value, the natural gas lower calorific value L in this example _gas ＝9.7(kW.h)/m ³ (ii) a Δ t is a scheduling time scale, which is taken in this embodiment for 1 hour; t is the number of the scheduling time segments.

The power generation efficiency is increased along with the increase of the load rate of the unit, and the waste heat efficiency is reduced along with the increase of the load rate of the unit; fitting to obtain the functional relation among the power generation efficiency, the waste heat efficiency and the load rate of the gas internal combustion engine as follows:

in the formula: p _mtN The rated power generation power of the gas internal combustion engine is kW.

Mathematical model of absorption refrigerating unit:

the absorption refrigerating unit can utilize the high-temperature flue gas produced by the gas combustion engine and the heat carried by the high-temperature cylinder sleeve water to cool a user, and the mathematical model of the absorption refrigerating unit is as follows:

in the formula: q _ac The refrigeration power of the bromine refrigerator; q _{ac_in} The unit is kW which is the input power of the bromine refrigerator; COP _ac The refrigeration coefficient of the bromine refrigerator is shown;

and

the minimum value and the maximum value of the refrigeration power of the bromine refrigerator are respectively.

The storage battery mathematical model is as follows:

the energy storage device can well smooth the fluctuation of photovoltaic output and load, and simultaneously reduces the strong coupling relation between energy and time sequence, thereby achieving the purpose of reasonably adjusting the energy according to the supply and demand relation. In order to realize more diversified control on the CCHP system, a storage battery is adopted as an electric energy storage device in the model of the embodiment. The storage battery needs to consider the characteristics of time sequence correlation, self-discharge, charge-discharge loss and the like in modeling, and the mathematical model is as follows:

in the formula: s _es (t) the remaining energy of the battery for a period of t;

and

respectively the minimum value and the maximum value of the residual energy of the storage battery; tau is _es The loss coefficient of the storage battery; Δ t is a scheduling time scale; eta _es，chr And η _es，dis The energy charge and discharge conversion efficiency of the storage battery; p _es，chr (t) and P _es，dis (t) charging and discharging power of the storage battery for a period of t, respectively;

and

respectively the charging power of the storage batteryA small value and a maximum value;

and

the minimum value and the maximum value of the discharge power of the storage battery are respectively.

Heat storage tank mathematical model:

the heat storage tank can store heat when the waste heat of the gas internal combustion engine is excessive, and release heat when the waste heat is insufficient, so that the utilization rate of the waste heat is effectively improved, the waste is avoided, and the operating cost is reduced. Similar to the storage battery, the heat storage tank also needs to consider the characteristics of time sequence association, self-discharge, charge and discharge loss and the like in modeling, and the mathematical model is as follows:

in the formula: s. the _hs (t) the residual energy of the heat storage tank in a period of t;

and

respectively is the minimum value and the maximum value of the residual energy of the heat storage tank; tau is _hs The loss coefficient of the heat storage tank; Δ t is a scheduling time scale; eta _hs，chr And η _hs，dis Respectively the energy heat release and heat storage conversion efficiency of the heat storage tank; p _hs，chr (t) and P _hs，dis (t) heat release and heat storage power of the heat storage tank at time t respectively;

and

the minimum value and the maximum value of the heat release power of the heat storage tank are respectively;

and

the minimum value and the maximum value of the heat storage power of the heat storage tank are respectively.

Mathematical models of photovoltaic power generation equipment:

for the convenience of modeling and calculation, the output of photovoltaic power generation is regarded as being only related to the illumination intensity and the air temperature, and the mathematical model is as follows:

P _pv ＝P _STC G(1+k _pv (T _c -25))/G _STC

in the formula: p _pv The output power of the photovoltaic cell is kW; g is the illumination intensity and the unit is W/m ² ；P _STC Is standard test condition (the incident intensity of sunlight is 1000W/m) ² The maximum test power at the ambient temperature of 25 ℃) is kW; g _STC The light intensity under standard test conditions is 1000w/m ² ；k _pv Is the power temperature coefficient in%/° c; t is _c Is the working temperature of the cell panel, in degrees C, the value of which passes through the ambient temperature T _r To estimate the temperature, T, of the photovoltaic module _c ＝T _r +30G/1000。

Mathematical models of other devices:

the relation between the input energy and the output energy of the gas-fired boiler and the heat exchanger can be embodied by the boiler efficiency and the heat exchange efficiency, and a mathematical model of the relation is as follows:

in the formula: q _ex _ _in The unit is kW which is the input power of the heat exchanger; q _ex The unit is kW which is the output heat of the heat exchanger;

and

respectively the minimum value and the maximum value of the output power of the heat exchanger; q _gb The unit is kW which is the heating power of the gas boiler;

and

respectively the minimum value and the maximum value of the heating power of the gas boiler; v _gb Is the gas consumption in m ³ ；η _gb 、η _ex The efficiency of the gas boiler and the heat exchanger, respectively.

The relationship between the refrigeration power of the electric refrigerator and the consumed power is as follows:

in the formula: q _ec The unit is kW which is the refrigerating power of the electric refrigerator; COP _ec Is the energy efficiency ratio of the electric refrigerator; p _ec The unit is kW, which is the power consumed by the electric refrigerator.

Step 2.2: according to the system structure and the mathematical models of the devices, the mathematical model for calculating the energy flow is constructed as follows:

in the formula: k is a radical of _mt Distributing the waste heat of the gas internal combustion engine to an absorption refrigerator for refrigeration; q _load.c 、Q _load.h The cold load and the heat load are respectively in kW; p _load The unit is kW which is the magnitude of electric load except the power consumed by the electric refrigerator; p _op The unit is kW for the self-use electric power of the system; k is a radical of _op The power utilization rate of the system is self-determined; p _grid The unit of the power purchasing power is kW for the power grid.

And obtaining an integral model of the CCHP system according to the input parameter scene obtained by the improved FCM algorithm, the corresponding probability and the established mathematical model of each device.

Step 2.3: and obtaining a double-layer optimized configuration model of the CCHP system according to typical scenes of all data in three seasons, probability data of all the scenes and the constructed CCHP system mathematical model obtained by improving the FCM algorithm.

on the basis of capacity allocation of each device of a given system of an outer layer allocation optimization model, selecting the power generation power, the waste heat allocation ratio, the charge and discharge power of a storage battery and the heat storage and discharge power of a heat storage tank of a gas internal combustion engine at each time period of one day as decision variables, aiming at the minimum daily operation and maintenance cost of the system, and taking decision variable operation constraint and energy balance as constraint conditions; and establishing an inner layer operation optimization model of the CCHP system by combining system data, equipment parameters and scene data.

Modeling the double-layer model of the CCHP system according to various parameter data obtained based on the improved FCM algorithm and the objective function of the double-layer model, wherein a flow chart of the double-layer optimization configuration model is shown in figure 3.

The establishing of the outer layer configuration optimization model of the CCHP system comprises the following steps:

build System object 1: mathematical model min f with minimum annual conversion investment cost and annual operation maintenance cost ₁ ：

In the formula, y _in The investment cost is converted for the system year; s is the total scene number; p _s Is the probability of scene s; t is a time period divided by one day; y is _op (t, s) is the running cost of the system in the t period under the s scene, and the specific expression is shown inAn inner layer optimization model;

annual conversion investment cost y _in The specific expression of (a) is as follows:

in the formula: p is a radical of _pvN 、E _hsN Rated installation capacity of the photovoltaic power generation and the heat storage tank is kW; p is a radical of _esN 、E _esN The unit is kW and KWh which are the maximum charge-discharge power and the rated installation capacity of the storage battery respectively; k is a radical of _pv 、k _hs The unit capacity cost of the photovoltaic power generation and the heat storage tank is unit/kW; k is a radical of formula _esp 、k _esE 、n _es The unit power cost, the unit capacity cost and the service life of the storage battery are respectively, and the unit is yuan/kW, yuan/kWh and year; n is _xt Planning the service life of the system, wherein the unit is year; l _tx For the discount rate, this example takes 0.05.

Build System object 2: mathematical model min f with minimum annual carbon dioxide emission ₂ ：

In the formula: f. of ₂ The carbon dioxide emission of the system is expressed in kg; p _s Is the probability of scene s;

the unit of the equivalent carbon dioxide emission of the carbon dioxide emission generated by the fuel gas and the electric quantity purchased by the power grid is kg, and the specific calculation formula is as follows:

in the formula:

the carbon dioxide conversion coefficients of natural gas and commercial power are respectively, and the units are kg/Nm and kg/kWh respectively; v _gas And P _grid And respectively purchasing gas quantity for natural gas and power for the power grid.

Constraint conditions are as follows:

in consideration of actual conditions such as capital, site and the like, the construction of the CCHP system has the following limitations:

in the formula: p is a radical of _pvN 、E _hsN Rated installation capacity of photovoltaic power generation and a heat storage tank is realized, and the unit is kW;

and

respectively the minimum value and the maximum value of the rated installation capacity of the photovoltaic power generation;

and

respectively the minimum value and the maximum value of the rated installation capacity of the heat storage tank; p _esN 、E _esN The maximum charge-discharge power and the rated installation capacity of the storage battery are respectively in kW and kWh units;

and

the maximum charge and discharge power of the storage battery is respectively the minimum value and the maximum value;

and

respectively, a minimum value and a maximum value of the rated installation capacity of the storage battery.

The inner-layer optimization operation model is specifically expressed as follows:

the inner layer optimization model takes the minimum daily operation and maintenance cost of the system as an objective function, and the mathematical model is as follows:

in the formula: y is _op The unit is element for the total operation cost of the system; p _s Probability of scene s; t is the cycle of the scheduling period; s is the total scene number; f _gas (t，s)、F _grid (t，s)、F _op (t, s) are respectively system fuel cost, power grid electricity purchasing cost and operation maintenance cost in a time period of t under the scene of s, units are elements, and specific expressions are as follows:

in the formula: c _gas 、C _grid The unit is Yuan/m and Yuan/kWh respectively; c _mt 、C _pv 、C _ac 、C _ex 、C _gb 、C _ec 、C _es 、C _hs The unit of the operation and maintenance cost is yuan/kW, and the unit of the operation and maintenance cost is respectively gas internal combustion engine, photovoltaic panel, absorption refrigerator, electric refrigerator, heat exchanger, storage battery and heat storage tank.

The decision variables are constrained as follows:

in the formula:

the maximum charging and discharging power of the storage battery is kW;

the ratio of the minimum load electric quantity and the maximum load electric quantity of the storage battery is obtained;

the minimum and maximum load heat proportion of the energy storage tank.

The energy balance constraint expression is as follows:

and step 3: and solving the double-layer model of the CCHP system according to the characteristic that the outer layer configuration optimization model and the inner layer operation optimization model have nonlinear optimization to form an optimal planning scheme set. The concrete expression is as follows:

the solving process of the double-layer model comprises the following steps: firstly, randomly generating a population by an outer layer model, and conveying the population to an inner layer model; secondly, the inner layer model uses a PSO algorithm to solve an optimal operation scheme based on the equipment capacity information input by the outer layer according to meteorological and load data, and returns the optimal operation scheme to the outer layer model; then, the outer layer model carries out non-dominated sorting and congestion degree calculation according to the operation scheme returned by the inner layer model, and generates a new population through selection, intersection and variation; and finally, conveying the new population to the inner layer model, and performing cyclic substitution calculation. And when the number of the selected generations reaches a specified value, stopping circulation and outputting the last generation of population.

The process for solving the outer-layer multi-target capacity allocation problem based on the NSGA-II is shown in FIG. 4, and specifically includes the following steps:

step 3.1: the rated installation capacity of photovoltaic power generation, the rated installation capacity and the rated maximum charging and discharging power of a storage battery and the rated installation capacity of a heat storage tank are used as decision variables, and the annual reduced investment cost and the annual operation and maintenance cost f are used ₁ Annual carbon dioxide emission f ₂ Smallest mathematical modelThe model is taken as an optimization objective function of an outer layer model, and the running capacity of CCHP system equipment is taken as a constraint condition;

step 3.3: generating an initialization population P by using a multi-target function and input parameters in an outer layer model, and setting the population iteration number N to be 0, wherein the maximum iteration number is D;

step 3.5: setting the evolution iteration number gen to be 2;

In fact, the solution of the multi-objective problem is not only an optimization problem but also a decision problem, and after the Pareto optimal solution set is obtained, the final compromise solution or optimal solution needs to be selected according to subjective and objective factors.

The solving result of the multi-target problem is a series of Pareto solutions, and the embodiment adopts a fuzzy multi-attribute decision method to select the optimal compromise solution, and the specific formula is as follows:

opt＝{o|μ ^o ＝max(μ ^τ )}

in the formula:

a gamma solution representing a tau-th objective function;

and

respectively representing a maximum value and a minimum value of a gamma solution of the tau-th objective function;

means for normalizing the γ th solution at the τ th objective function; mu.s ^τ Calculating the proportion of the sum of each attribute of each solution to the total; mu.s ^o Representing the maximum proportion of the sum of each attribute of each solution to the total; opt represents the selection of the solution with the largest overall proportion as the optimal solution.

A process for solving the problem of the inner-layer optimal operation scheme based on pso (particle Swarm optimization) is shown in fig. 5, and is described in detail as follows:

step 3.11: generating an initialization population P by using an objective function and input parameters in an inner layer model, setting the population iteration frequency R to be 0, setting the maximum iteration frequency to be G, setting the initial position W of each particle to be 0, and setting the initial speed V of each particle to be 0;

step 3.13: updating the historical optimal position of each particle;

step 3.14: updating the global optimal position of the group;

step 3.15: updating each particle velocity V ^G+1 And position W ^G+1 ；

In order to verify the effectiveness of the method of the embodiment, a CCHP energy station installed in a hotel is taken as a research object, a CCHP system architecture is shown in fig. 2, and the optimal configuration method of the embodiment is applied to perform optimal configuration of electricity, heat storage and photovoltaic devices, so as to achieve the maximum economy and environmental protection. The energy storage device can well alleviate the fluctuation of source charge and weaken the strong coupling of energy supply and demand on a time sequence, thereby promoting the energy utilization rate and reducing the operation cost. In addition, the clean environmental protection economy of photovoltaic power generation can reduce the power supply cost in hotel. In order to improve the comprehensive operation benefit of the hotel CCHP system, the storage battery, the heat storage tank and the photovoltaic panel with appropriate capacity are equipped for the hotel CCHP system by adopting the method on the basis of the existing equipment, so that the maximization of economic benefit and environmental protection benefit is realized.

To verify the effectiveness of the improved FCM in CCHP optimization configuration, the proposed optimization model based on the improved FCM algorithm is compared with the optimization model based on the conventional FCM algorithm. Firstly, using a traditional FCM algorithm to perform scene reduction on the meteorological data of the cooling, heating and power load, the air temperature and the illumination intensity of a hotel in 2020 as a whole by using the traditional FCM algorithm and an improved FCM algorithm, wherein a default value m of a fuzzy coefficient is 2, the number of clustering centers is set as a maximum value 19, and the clustering results of the traditional FCM algorithm and the improved FCM algorithm are respectively shown in fig. 6 and 7.

In the improved FCM algorithm reduction process, the relationship between the fuzzy coefficient, the clustering number and the comprehensive evaluation index is shown in fig. 8.

As can be seen from fig. 8, the overall evaluation index reaches a maximum value when m is 3.4 and c is 18, so that the optimal blurring coefficient is 3.4 and the optimal reduction number is 18, that is, the reduced scene set includes 18 typical scenes.

And (3) constructing a double-layer optimization configuration model according to the step 2, and respectively inputting the scenes obtained after the traditional FCM and the improved FCM are reduced into the model, wherein the obtained Pareto frontier is shown in FIG. 9. As can be seen from fig. 9, in the configuration based on the improved FCM algorithm, the annual average investment operation cost and the annual carbon dioxide emission are contradictory, and there is no ideal optimal solution. According to the multi-attribute decision method in the improved FCM, the multi-attribute decision index M of each solution is calculated as shown in fig. 10, and it can be seen that the μ value of solution No. 46 is the largest, so that solution No. 46 is selected as the best compromise solution. In addition, as can be seen from fig. 9, the Pareto frontier obtained by the configuration model based on the conventional FCM clustering algorithm is still inferior to that of the method provided in the present embodiment. Similar to the above, a compromise solution is selected from the solution, and it can be calculated, so that the annual average investment and operation cost of the optimized configuration scheme obtained by the method provided by the embodiment is reduced by 10.25%, and the annual carbon dioxide emission is reduced by 0.83%.

In order to further verify the superiority of the improved FCM clustering algorithm provided by the embodiment in solving the CCHP optimal configuration problem compared with the traditional FCM clustering algorithm, the actual load and meteorological data in 2020 and the obtained two optimal configuration schemes are brought into the inner layer optimal operation model, and the actual operation cost and carbon dioxide emission under the two schemes are calculated. As can be seen from fig. 11, in most cases, compared with the conventional FCM algorithm, the operation cost of the optimized configuration scheme obtained by applying the improved FCM algorithm is lower, the annual operation cost is reduced by 358962 yuan, and the reduction rate is 4.68%; the investment cost is reduced by 140090 yuan in the reduced year, and the reduction rate is 10.07%; the total cost is reduced by 499052 yuan, and the reduction rate is 5.51%.

Fig. 12 shows daily carbon dioxide emissions versus total annual carbon dioxide emissions. As can be seen from fig. 12, in most cases, compared with the conventional FCM algorithm, the optimized configuration scheme obtained by applying the improved FCM algorithm has lower carbon dioxide emission, the total emission is reduced by 365065kg, and the reduction rate is 4.19%.

The above results show that: on the basis of solving the CCHP optimization configuration problem, compared with the traditional FCM clustering algorithm, the improved FCM clustering algorithm provided by the embodiment has a better clustering effect, and can better reflect overall characteristics and rules of weather and load, so that the configuration scheme obtained by solving the optimization model can be more suitable for practice, and the economy and the environmental protection of a CCHP system are improved.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.

Claims

1. A cold-heat-electricity combined supply system configuration method based on improved fuzzy C-means clustering is characterized in that: the method comprises the following steps:

step 1: the traditional fuzzy C-means clustering algorithm is improved, a PFS index considering the clustering effectiveness of the data geometry structure information and a Vp index considering the clustering effectiveness of the membership degree are introduced on the basis of the original algorithm at the same time, and the PFS index and the Vp index are combined to form a comprehensive clustering evaluation index by using an entropy weight weighting form; clustering air temperature, illumination intensity and electricity/heat/cold load data through an improved FCM algorithm to obtain typical scenes and probability data of each scene in three seasons of a transition season, a refrigeration season and a heating season, and taking the obtained data result as the input of a CCHP optimization configuration model;

step 2: constructing a CCHP system double-layer optimization configuration model based on an improved FCM algorithm; the outer layer is an optimization layer, and the aim is to minimize the annual average converted investment and operation cost of the system and minimize the annual carbon dioxide emission; the inner layer is an operation optimization layer and aims at minimizing daily operation cost; the inner layer model and the outer layer model are mutually nested;

and step 3: and (3) solving the CCHP system double-layer optimization configuration model based on the improved FCM algorithm by combining a fast non-dominated sorting genetic algorithm NSGA-II and a particle swarm optimization algorithm PSO to obtain the optimal CCHP equipment capacity configuration.

2. The cooling, heating and power combined supply system configuration method based on the improved fuzzy C-means clustering is characterized in that: the method for clustering the air temperature, the illumination intensity and the electricity/heat/cold load data in the step 1 is the same, and the clustering method is as follows:

step 1.1: acquiring annual hourly air temperature, illumination intensity or hotel electricity/heat/cold load data of the location of the system, recording the number of certain types of input data, recording the number of certain types of data as n, and setting a fuzzy coefficient m and a clustering number c;

Then, obtaining PFS index vector y1 and Vp index vector y2 with different clustering numbers under the same fuzzy coefficient, and obtaining yl and y2 by per-unit of y1 and y2 according to the respective maximum values;

3. The method for configuring the combined cooling heating and power system based on the improved fuzzy C-means clustering according to claim 1 or 2, wherein: the specific method of the step 2 comprises the following steps:

step 2.1: determining the types of various energy conversion equipment in the CCHP system, wherein the energy conversion equipment comprises a gas internal combustion engine, an energy storage device, an absorption refrigerating unit, a photovoltaic cell, an electric refrigerator, a gas boiler and a heat exchange device, and setting the connection mode among the various energy conversion equipment; the energy storage device comprises a storage battery and a heat storage tank; then, mathematical modeling is carried out on each energy conversion device;

step 2.2: constructing an overall energy flow calculation mathematical model of the CCHP system according to the connection structure;

the rated installation capacity of photovoltaic power generation, the rated installation capacity and the rated maximum charging and discharging power of a storage battery and the rated installation capacity of a heat storage tank are used as decision variables, and the annual reduced investment cost and the annual operation and maintenance cost f are used ₁ Annual carbon dioxide emission f ₂ Minimizing as a multi-objective function, and taking the running capacity of the CCHP system equipment as a constraint condition; establishing an outer layer configuration optimization model of the CCHP system by combining system data, equipment parameters and scene data;

4. The method for configuring the combined cooling heating and power system based on the improved fuzzy C-means clustering of claim 3, wherein: in step 2.1, the mathematical model of each energy conversion device is specifically as follows:

the mathematical model of the gas internal combustion engine is as follows:

and

and

respectively is the minimum value and the maximum value of the waste heat power; eta _mtP 、η _mtQ Respectively power generation efficiency and waste heat efficiency; l is a radical of an alcohol _gas Is the heat value of natural gas; Δ t is a scheduling time scale; t is the number of scheduling time segments;

the power generation efficiency is increased along with the increase of the load rate of the unit, and the waste heat efficiency is reduced along with the increase of the load rate of the unit; the functional relation among the power generation efficiency, the waste heat efficiency and the load rate of the gas internal combustion engine is obtained by fitting:

in the formula: p is _mtN The rated power is the rated power generation power of the gas internal combustion engine, and the unit is kW;

the mathematical model of the absorption refrigerating unit is as follows:

in the formula: q _ac The refrigeration power of the bromine refrigerator; q _{ac_in} The unit is kW which is the input power of the bromine refrigerator; COP _ac The refrigeration coefficient of the bromine refrigerator is;

and

respectively the minimum value and the maximum value of the refrigeration power of the bromine refrigerator;

the storage battery mathematical model is as follows:

and

respectively the minimum value and the maximum value of the residual energy of the storage battery; tau is _es The loss coefficient of the storage battery; delta t is unit scheduling time; eta _es，chr And η _es，dis As energy of a storage batteryCharge and discharge conversion efficiency; p is _es，chr (t) and P _es，dis (t) charging and discharging power of the storage battery for a period of t, respectively;

and

respectively the minimum value and the maximum value of the charging power of the storage battery;

and

respectively the minimum value and the maximum value of the discharge power of the storage battery;

the heat storage tank mathematical model is as follows:

in the formula: s _hs (t) the residual energy of the heat storage tank in a period of t;

and

respectively is the minimum value and the maximum value of the residual energy of the heat storage tank; tau. _hs The loss coefficient of the heat storage tank; delta t is unit scheduling time; eta _hs，chr And η _hs，dis Respectively the energy heat release and heat storage conversion efficiency of the heat storage tank; p _hs，chr (t) and P _hs，dis (t) heat release and heat storage power of the heat storage tank at time t respectively;

and

and

the minimum value and the maximum value of the heat storage power of the heat storage tank are respectively;

the mathematical model of the photovoltaic power generation equipment is as follows:

P _pv ＝P _sTC G(1+k _pv (T _c -25))/G _STC

in the formula: p _pv The output power of the photovoltaic cell is kW; g is the illumination intensity and the unit is W/m ² ；P _STC The maximum test power under the standard test condition is kW; g _STC The light intensity under standard test conditions is 1000w/m ² ；k _pv Is the power temperature coefficient in%/° c; t is _c Is the working temperature of the cell panel, in degrees C, the value of which passes through the ambient temperature T _r To estimate the temperature, T, of the photovoltaic module _c ＝T _r +30G/1000；

Mathematical models of other devices:

the mathematical model of the gas boiler and the heat exchanger is as follows:

in the formula: q _{ex_in} The unit is kW which is the input power of the heat exchanger; q _ex The unit is kW which is the output heat of the heat exchanger;

and

respectively minimum output power of heat exchangerA value and a maximum value; q _gb The unit is kW which is the heating power of the gas boiler;

and

respectively the minimum value and the maximum value of the heating power of the gas boiler; v _gb Is the gas consumption in m ³ ；η _gb 、η _ex The efficiency of the gas boiler and the heat exchanger respectively;

the relationship between the refrigeration power of the electric refrigerator and the consumed electric power is as follows:

in the formula: q _ec The unit is kW which is the refrigerating power of the electric refrigerator; COP (coefficient of Performance) _ec Is the energy efficiency ratio of the electric refrigerator; p _ec The unit of the power consumed by the electric refrigerator is kW.

5. The cooling, heating and power combined supply system configuration method based on the improved fuzzy C-means clustering is characterized in that: in the step 2.2, according to the system structure and the mathematical models of the devices, the mathematical model for energy flow calculation is constructed as follows:

in the formula: k is a radical of _mt Distributing the waste heat of the gas internal combustion engine to an absorption refrigerator for refrigeration; q _load.c 、Q _load.h The cold load and the heat load are respectively the size, and the unit is kW; p is _load The unit is kW which is the magnitude of electric load except the power consumed by the electric refrigerator; p _op The unit is kW for the self-use electric power of the system; k is a radical of _op The power utilization rate of the system is self-determined; p _grid The unit of the power purchasing power is kW for the power grid.

6. The method for configuring the combined cooling heating and power system based on the improved fuzzy C-means clustering of claim 5, wherein: the step 2.3 of establishing an outer layer configuration optimization model of the CCHP system includes:

construction of System object 1: mathematical model min f with minimum annual conversion investment cost and annual operation maintenance cost ₁ ：

In the formula, y _in The investment cost is converted for the system year; s is the total scene number; p _s Is the probability of scene s; t is a time period divided by one day; y is _op (t, s) is the operation cost of the system at the time period t under the scene of s, and the specific expression of the operation cost is shown in an inner-layer optimization model;

in the formula: p is a radical of _pvN 、E _hsN Rated installation capacity of photovoltaic power generation and a heat storage tank is realized, and the unit is kW; p is a radical of _esN 、E _esN The maximum charge-discharge power and the rated installation capacity of the storage battery are respectively in kW and KWh; k is a radical of _pv 、k _hs The unit capacity cost of the photovoltaic power generation and the heat storage tank is unit/kW; k is a radical of _esp 、k _esN 、n _es The unit power cost, the unit capacity cost and the service life of the storage battery are respectively, and the unit is yuan/kW, yuan/kWh and year; n is _xt Planning the service life of the system, wherein the unit is year; l _tx The current rate is the current rate;

In the formula: f. of ₂ The unit is kg of carbon dioxide emission of the system; p is _s Is the probability of scene s;

in the formula:

the carbon dioxide conversion coefficients of natural gas and commercial power are respectively, and the units are kg/Nm and kg/kWh respectively; v _gas And P _grid Respectively the gas purchasing quantity of the natural gas and the power purchasing power of the power grid;

constraint conditions are as follows:

in the formula: p _pvN 、E _hsN Rated installation capacity of the photovoltaic power generation and the heat storage tank is kW;

and

and

and

the minimum value and the maximum value of the maximum charge-discharge power of the storage battery are respectively;

and

7. The method for configuring the combined cooling heating and power system based on the improved fuzzy C-means clustering of claim 6, wherein: the inner layer optimization operation model in the step 2.3 is specifically expressed as follows:

in the formula: y is _op The unit is element for the total operation cost of the system; p _s Is the probability of scene s; t is the cycle of the scheduling time interval; s is the total scene number; f _gas (t，s)、F _grid (t，s)、F _op (t, s) are respectively s sceneThe unit of the system fuel cost, the power grid electricity purchasing cost and the operation and maintenance cost in the period t is element, and the specific expression is as follows:

in the formula: c _gas 、C _grid The unit is Yuan/m and Yuan/kWh respectively; c _mt 、C _pv 、C _ac 、C _ex 、C _gb 、C _ec 、C _es 、C _hs The operating and maintaining costs of the gas internal combustion engine, the photovoltaic panel, the absorption refrigerator, the electric refrigerator, the heat exchanger, the storage battery and the heat storage tank are respectively unit/kW;

the decision variables are constrained as follows:

in the formula:

the maximum charging and discharging power of the storage battery is kW;

the minimum and maximum load heat proportion of the energy storage tank;

the energy balance constraint expression is as follows:

8. the method for configuring the combined cooling heating and power system based on the improved fuzzy C-means clustering of claim 7, wherein: in step 3, the solving process of the double-layer model is as follows:

finally, conveying the new population to the inner layer model, and performing cyclic substitution calculation; and when the number of selected generations reaches a specified value, stopping circulation and outputting the final generation of population.

9. The method for configuring the combined cooling, heating and power system based on the improved fuzzy C-means clustering of claim 8, wherein:

the process for solving the outer-layer multi-target capacity configuration problem based on NSGA-II is concretely as follows:

step 3.5: setting the evolution iteration number gen to be 2;

step 3.7: carrying out three basic operations of genetic algorithm selection, crossing and variation on the newly generated new father population Qt;

step 3.8: judging a termination condition, terminating when the maximum iteration number D is reached, and outputting a Pareto optimal solution set; selecting multiple groups of typical optimal solutions from the obtained Pareto optimal solution set to obtain capacity configuration schemes corresponding to the multiple groups of typical optimal solutions, and outputting an optimal equipment combination scheme, reduced investment cost and annual operation maintenance cost f ₁ Annual carbon dioxide emission f ₂ (ii) a Otherwise, the step 3.6 is returned.

10. The method for configuring the combined cooling, heating and power system based on the improved fuzzy C-means clustering of claim 9, wherein: the process for solving the inner layer optimal operation scheme problem based on the PSO is specifically described as follows:

step 3.13: updating the historical optimal position of each particle;

step 3.14: updating the global optimal position of the group;

step 3.15: updating each particle velocity V ^G+1 And position W ^G+1 ；