Multi-energy optimization method considering integration of multiple parks of carbon targets based on NSGA-RIME algorithm

Abstract

The invention relates to the technical field of multi-park energy units, and discloses a multi-energy optimization method considering integration of multiple parks of a carbon target based on an NSGA-RIME algorithm, and the method comprises the steps: S1, constructing a multi-park multi-energy-state double-layer digital-mechanism model; s2, mining and determining a carbon-energy double-layer target coupling relationship; s3, an MORIME algorithm is introduced; s4, multi-dimensional combination of an NSGA-II type algorithm and an MORIME algorithm is carried out; s5, carrying out competitive solution on the NSGA-II type algorithm and the MORIME algorithm; s6, providing a new scheduling method; s7, a to-be-solved double-layer multi-target model is determined; the invention provides an NSGA-RIME combined algorithm model aiming at double-layer multi-objective optimization for solving a double-layer multi-objective model so as to solve the complexity and the polymorphism of an energy system, the dual pressure of energy conservation and carbon emission, the solving requirement of a multi-objective optimization problem and the requirement of algorithm innovation and combined application. And optimization, upgrading and sustainable development of the park energy system can be promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-park energy units, and in particular to a multi-energy optimization method for integrating multiple parks taking into account carbon targets based on an NSGA-RIME algorithm. Background Art

[0002] With the continuous advancement of energy technology and the diversification of the park's energy structure, the energy units in each park (such as solar energy, wind energy, energy storage devices, traditional power generation equipment, etc.) have different characteristics and operating mechanisms. The complex interconnection between these energy units makes the operation and management of the entire energy system extremely challenging. Moreover, in the context of global climate change and environmental protection, energy conservation and reduction of carbon emissions have become the common goals of governments and enterprises around the world. For the park's energy system, how to achieve the dual goals of energy conservation and emission reduction while ensuring energy supply has become an urgent problem to be solved.

[0003] In addition, since energy systems often have multiple optimization objectives (such as energy saving, emission reduction, economy, etc.), these objectives often conflict and constrain each other, making it impossible to effectively solve the traditional single-objective optimization method.

[0004] To this end, a multi-energy optimization method based on the NSGA-RIME algorithm considering the integration of multiple parks with carbon targets is proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a multi-energy optimization method based on the NSGA-RIME algorithm that takes into account the integration of multiple parks with carbon targets, so as to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a multi-energy optimization method based on the NSGA-RIME algorithm taking into account the integration of multiple parks with carbon targets, comprising the following steps:

[0007] S1: Construct a two-layer digital-mechanism model of multi-park multi-energy states, and study and summarize the factors and data of concern such as the characteristics, interconnection relationships, and operation mechanisms of different multi-state energy units in each park;

[0008] S2: Mining and determining the coupling relationship between carbon and energy dual-layer targets. After collecting relevant information on the park's carbon targets and energy targets, characterize the functional relationship and data connection between the data-related models of the two aspects, so that it can achieve effective coupling of the two models in the carbon-energy dual-layer digital-mechanism model, and enable them to interact with each other in scheduling optimization;

[0009] S3: Introducing the MORIME algorithm. In the process of finding the optimal solution, the MORIME algorithm gradually covers the entire search space and retains the optimal decision;

[0010] S4: Multi-dimensional combination of NSGA-II algorithm and MORIME algorithm, which combines the algorithms in multiple dimensions to solve the two-level multi-objective optimization problem;

[0011] S5: Competitive solution of NSGA-II algorithm and MORIME algorithm. For a common two-level multi-objective optimization problem, it can be divided into two problems. Different combinations of algorithms form corresponding solution networks. The NSGA-II algorithm and MORIME algorithm are combined to form the NSGA-RIME algorithm.

[0012] S6: The new scheduling method is proposed to expand and supplement the existing optimization scheduling method to a certain extent, strengthen the scheduling capacity of the park, and enrich the existing scheduling methods;

[0013] S7: The two-level multi-objective model to be solved is established. When the above scheduling method is clear, the multi-objective optimization model input into the algorithm, as well as the constraints, objective functions and coupling relationships of the two-level multi-objectives are summarized;

[0014] S8: Use of NSGA-RIME algorithm and formation of dual Pareto front end. The dual-layer multi-objective model is input into the NSGA-RIME algorithm. The dual-layer multi-objective is input into different corresponding algorithm network layers respectively to obtain the corresponding Pareto front end decision set.

[0015] S9: Evaluation and decision integration. The dual Pareto front-end decision set includes different scheduling strategies. The two Pareto front-end decision sets are combined and an evaluation system is established based on them.

[0016] S10: Obtain the final single decision set, output the common parts of the two sets, screen the different parts through the relevant evaluation system, obtain the allowed relevant decisions, and incorporate them into the final result.

[0017] Preferably, in S1, when constructing a two-layer digital-mechanism model of multiple parks and multiple energy states, the collected data are summarized including: park energy unit characteristics and data collection in line with energy-saving targets and park energy unit characteristics and data collection in line with carbon targets.

[0018] Preferably, in S1, the park energy unit characteristics and data collection content that cooperate with the energy-saving target include: the operating mechanism and data of each park itself; the operating mechanism and data of the power network and other energy networks between parks; the operating mechanism and data summary of polymorphic energy in the park.

[0019] Preferably, in S1, the park energy unit characteristics and data collection content that match the carbon target include: inter-park carbon exchange characteristics and data collection; the park's own carbon indicators; and the park's energy units.

[0020] Preferably, in S2, mining and determining the carbon-energy dual-layer target coupling relationship includes: carbon-energy data model alignment; carbon-energy data model coupling relationship.

[0021] Preferably, in S1, a two-layer digital-mechanism model of multiple parks and multiple energy states is constructed, including an aggregated multi-dimensional mechanism model integrating the park network and the energy unit model; and an energy-carbon data model of energy units between parks and within the parks themselves.

[0022] Preferably, in S3, when the MORIME algorithm is introduced, the relevant parameters required by the algorithm include: SP is used to measure the distribution uniformity of the solution set generated by the algorithm; HV is an indicator for measuring the size of the hypervolume occupied by the solution set generated by the algorithm; GD is an indicator for measuring the distance between the solution set generated by the algorithm and the true frontier.

[0023] Preferably, in S4, the multi-dimensional combination of the NSGA-II algorithm and the MORIME algorithm includes: cross combination and front-to-back superposition.

[0024] Preferably, in S6, the new scheduling method includes: related scheduling technologies based on carbon monitoring and capture, application of carbon emission circulation network, and mobile energy storage vehicle scheduling method in conjunction with carbon scheduling.

[0025] Preferably, in S7, the dual-layer multi-objective objective function, constraint conditions and coupling relationships are:

[0026] Objective function:

[0027]

[0028]

[0029] σ1=f(x1,…,x n )#(3.3)

[0030] σ2=g(y1,…,y n )#(3.4)

[0031] Constraints:

[0032]

[0033] Coupling relationship:

[0034] H l (x1,…,x i ,…,x n )±Jk (y1,…,y j ,…,y m )=0 l=1,2...,n; k=1,2...,m#(3.7).

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention first studies and summarizes the factors and data of concern such as the characteristics, interconnection relationships and operation mechanisms of different polymorphic energy units between various parks, and on this basis combines the carbon emissions of energy units and the power supply carbon index of the power grid with the above factors to summarize the data information and operation principles of various energy units in various parks, so as to further establish a corresponding level of carbon energy double-layer data-mechanism model for various energy units in various parks while taking into account the dual goals of energy saving and carbon emissions. Since the above-mentioned operation data-mechanism model is established for the carbon-energy goal and has two different optimization goals, different The optimization objectives are optimized in layers. At this time, the conventional NSGA algorithm is not competent for such complex scenarios. Therefore, referring to its iterative version NSGA-Ⅱ algorithm and the MORIME algorithm which is also used to solve the multi-objective optimization results, they are combined and a NSGA-RIME combined algorithm model for double-layer multi-objective optimization is proposed to solve the double-layer multi-objective model, so as to solve the complexity and polymorphism of the energy system, the dual pressure of energy conservation and carbon emissions, the need to solve multi-objective optimization problems and the need for algorithm innovation and combined application. This method is helpful to promote the optimization and upgrading of the energy system in the park and its sustainable development.

[0037] The parts not involved in this method are the same as the existing technology or can be implemented by using the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the multi-energy optimization method of the present invention based on the NSGA-RIME algorithm taking into account the integration of multiple parks with carbon targets;

[0039] Figure 2 It is a schematic diagram of an aggregated multi-dimensional mechanism model integrating a campus network and an energy unit model of the present invention;

[0040] Figure 3 A flowchart of the MORIME algorithm of the present invention for solving the problem;

[0041] Figure 4 It is a flow chart for solving the problem when the NSGA-II algorithm of the present invention is cross-combined with the MORIME algorithm;

[0042] Figure 5 This is a schematic diagram of the competition between NSGA-II and MORIME algorithms of the present invention. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0044] See also Figure 1-Figure 5 The present invention provides a technical solution: a multi-energy optimization method based on the NSGA-RIME algorithm taking into account the integration of multiple parks with carbon targets, comprising the following steps:

[0045] S1: Construct a two-layer digital-mechanism model of multiple parks and multiple energy states, and study and summarize the factors and data of concern such as the characteristics, interconnection relationships, and operation mechanisms of different multi-state energy units between each park; on this basis, combine the carbon emissions of energy units and the power supply carbon indicators of the power grid with the above factors to summarize the data information and operation principles of each energy unit in each park, so as to further establish a corresponding level of carbon energy two-layer data-mechanism model for each energy unit in each park while taking into account the dual goals of energy conservation and carbon emissions.

[0046] In S1, when constructing a two-layer digital-mechanism model of multiple parks and multiple energy states, the collected data are summarized including: park energy unit characteristics and data collection in line with energy-saving targets and park energy unit characteristics and data collection in line with carbon targets.

[0047] Among them, in conjunction with the energy unit characteristics of the park to achieve energy-saving goals, the main goal of establishing optimized scheduling is energy saving, that is, to achieve maximum energy utilization efficiency (low loss) or minimum energy consumption on the park power grid side. Therefore, this should be used as a measurement indicator to summarize the data items that need to be included in the modeling and the related characteristics that need to be characterized in three dimensions: each park itself, the power network between each park and the power grid and other energy networks, and the polymorphic energy units of the park.

[0048] In S1, the park energy unit characteristics and data collection content that are in line with the energy-saving goals include: the operating mechanism and data of each park itself; the operating mechanism and data of the power network and other energy networks between parks; the operating mechanism and data summary of polymorphic energy within the park.

[0049] The operation mechanism and data of each park:

[0050] (1) First, from the perspective of operating mechanism, the types and composition of energy in the park are summarized, and on this basis, their interconnection relationships (energy input and output, network connection, and proportion of each item) are described. From the perspective of mainly caring about energy use, the corresponding logical relationships are selected and summarized, and it is required that they can ultimately reflect the interconnection relationships and specific functions of the various units within the park under normal operation.

[0051] (2) Secondly, from the perspective of operating data, the park is taken as a black box object, and its historical operating data is collected from the node level. A certain prediction mechanism is introduced to summarize the energy input and output data to form a corresponding time series model. At the same time, the input and output power of the node and the electric energy used in the minimum monitored time period are monitored in real time to obtain the corresponding data set.

[0052] Operation mechanism and data of inter-park power network and other energy networks:

[0053] (1) The operation mechanism of the inter-park power network and other energy networks is mainly reflected in the relationship between parks. On this basis, it is necessary to summarize the energy flow direction and type of energy flow between branches between nodes so that it can ultimately reflect the energy exchange relationship between the inter-park network.

[0054] (2) The operation data of the inter-park power network and other energy networks need to reflect the amount of energy flow between each node and its direction, that is, to specifically describe the specific situation of energy flow on the road, including related losses and transportation efficiency.

[0055] Operation mechanism and data summary of multi-state energy in the park:

[0056] (1) The main research object of the operation mechanism of multi-state energy in the park is a specific energy unit. Taking the mobile energy storage vehicle as an example, the mobile energy storage vehicle is a specific energy unit belonging to the park, and its operation mechanism needs to be studied: its main mobile compensation principle, the charging and discharging control principle of the energy storage vehicle, and other energy-related principles are summarized.

[0057] (2) The operation data of polymorphic energy in the park mainly need to collect multiple data in two aspects: 1* The historical operation data of the energy unit includes its energy input and output, energy conversion efficiency and other information. Such data can introduce corresponding prediction mechanisms to make it have a more detailed time scale; 2* The real-time operation data of the energy unit, which is arranged here considering the needs of subsequent scheduling, mainly reflecting the input and output power of the energy unit and the energy flow within the minimum monitoring time period.

[0058] Among them, the park energy unit characteristics and data collection that are in line with the carbon targets are another aspect of the unified level of the above summary, that is, the park energy unit characteristics and data collection that take into account the carbon targets should be summarized by imitating and analogizing the above collection methods for various carbon indicators, and are mainly reflected in three dimensions: carbon exchanges between parks, the carbon indicators of the park itself, and the carbon indicators of the park's energy units.

[0059] In S1, the park energy unit characteristics and data collection content that match the carbon target include: carbon exchange characteristics and data collection between parks; carbon indicators of the park itself; and park energy units.

[0060] Carbon exchange characteristics and data collection between parks:

[0061] Carbon exchange between parks mainly refers to carbon emissions that flow along with the flow of electricity on the grid side or between grid parks, that is, carbon emission flows in the usual sense, including various indicators such as branch carbon flow, branch carbon flow, branch carbon flow density, and node carbon potential. After collecting this information, the operation data-mechanism model can be built.

[0062] Carbon indicators of the park itself:

[0063] On the one hand, the carbon index of the park itself needs to be combined with part of the information given in the previous step (that is, the carbon emissions calculated by the park from the power grid and the use of electricity). On the other hand, it is necessary to take into account the additional carbon emissions within the park, that is, the specific data of carbon emissions generated in the form of changes such as energy form conversion and spatial position transfer. These data are integrated and matched with the above carbon indicators to form relatively unified data items, so that the carbon index of the park can be more comprehensively reflected.

[0064] Park Energy Unit:

[0065] The carbon targets of the park's energy units require research on each specific energy unit. Here are three situations: (1) The park's heating system and natural gas system will produce a certain degree of carbon emissions during the energy conversion process; (2) The park may be equipped with carbon capture equipment that can absorb carbon emissions in the park, or absorb carbon emissions while still generating or consuming energy; (3) The park's energy units absorb energy directly from the grid (not through the park) with the accompanying carbon emissions.

[0066] S2: Mining and determining the coupling relationship between carbon and energy dual-layer targets. After collecting relevant information on the park's carbon targets and energy targets, characterize the functional relationship and data connection between the data-related models of the two aspects, so that it can achieve effective coupling of the two models in the carbon-energy dual-layer digital-mechanism model, and enable them to interact with each other in scheduling optimization;

[0067] In S2, the mining and determination of the carbon-energy dual-layer target coupling relationship includes: carbon-energy data model alignment; carbon-energy data model coupling relationship.

[0068] Carbon-Energy Data Model Alignment:

[0069] In the data models of the above two aspects, there are data items with different but similar attributes. In order to optimize the subsequent scheduling processing and algorithm introduction of these data items, the data items that may have potential corresponding relationships are matched and aligned so that the corresponding data items can have the same data weight.

[0070] Carbon-energy data model coupling relationship:

[0071] There are certain connections between the data models of various aspects of carbon-energy. If this data relationship is not expressed, the final model will not be able to correctly reflect some detailed mechanisms within the system and lose key information. Here are a few limited relationships:

[0072] (1) Carbon-energy indicators change

[0073] The most common one is reflected in the following situations where the carbon-energy indicators change together: the park also generates carbon emissions when using energy on the grid side, and the carbon flow between parks changes with the energy flow between parks.

[0074] (2) Changes in carbon-energy indicators complementarity

[0075] After the introduction of carbon capture equipment, carbon emissions and energy between parks are complementary, showing a change state opposite to the above situation. Here, it is necessary to establish a correlation function between each data item to characterize the relationship.

[0076] Construct a two-layer digital-mechanism model of multiple parks and multiple energy states, including an aggregated multi-dimensional mechanism model that integrates the park network and energy unit model; an energy-carbon data model of energy units between parks and within the park itself.

[0077] Based on the data collected above, a two-layer digital carbon-energy mechanism model for each energy unit in the park is established according to the dual goals of energy conservation and carbon constraints. This model mainly includes two aspects: an aggregated multi-dimensional mechanism model that integrates the park network and the energy unit model; and an energy-carbon data model for energy units between parks and within the park itself.

[0078] Please also read Figure 2 , an aggregated multi-dimensional mechanism model integrating the campus network and energy unit model:

[0079] This model is mainly designed from the inter-park network model to form a corresponding carbon-energy communication network. On this basis, the content mechanism of the park nodes is filled to reflect the corresponding functions of the energy units.

[0080] Energy-carbon data model of energy units within and between parks:

[0081] This data model needs to cover two aspects of relevant data: historical data of data items in 96-point time series and real-time monitoring data. In terms of data items, it is necessary to collect the real-time power P of the node. i , the minimum time period for monitoring nodes is how much energy W i , the carbon emissions of the park are M i , the node carbon potential of the park i , the amount of power transmission between the nodes in the park within a specified time Wij , the carbon flow R between nodes ij , the proportion of energy units in the park p ik , and the assumed power and energy consumption of the campus nodes P ik , W ik , the energy conversion efficiency η of the park energy unit ik , the carbon emissions of the energy unit M ik , the property sig of the energy unit ik and more data items related to it.

[0082] S3: Introducing the MORIME algorithm. In the process of finding the optimal solution, the MORIME algorithm gradually covers the entire search space and retains the optimal decision;

[0083] Since the above operating data-mechanism model is established for the carbon-energy target and has two different optimization targets, different optimization targets need to be optimized in layers. At this time, the conventional NSGA algorithm is not competent for such a complex scenario. Therefore, referring to its iterative version NSGA-Ⅱ algorithm and the MORIME algorithm which is also used to solve the multi-objective optimization results, a NSGA-RIME combined algorithm model for two-layer multi-objective optimization is proposed to solve the two-layer multi-objective model.

[0084] The MORIME algorithm, also known as the multi-objective rime optimization algorithm, is a heuristic algorithm named after the rime phenomenon in nature. Rime is a type of ice crystal that forms on the surface of natural objects, like a layer of white fog wrapping around the object. This phenomenon is very common in nature. The design of the rime optimization algorithm is inspired by this phenomenon. In the process of finding the optimal solution, the algorithm will gradually cover the entire search space like rime and retain the optimal decision.

[0085] In S3, when the MORIME algorithm is introduced, the relevant parameters required by the algorithm include: SP is used to measure the uniformity of the distribution of the solution set generated by the algorithm; HV is an indicator to measure the size of the hypervolume occupied by the solution set generated by the algorithm; GD is an indicator to measure the distance between the solution set generated by the algorithm and the true frontier. SP is used to measure the uniformity of the distribution of the solution set generated by the algorithm. It evaluates the performance of the algorithm by calculating the distance between adjacent solutions in the solution set generated by the algorithm. The smaller the SP value, the more uniform the distribution of the solution set generated by the algorithm, and the better the diversity and coverage of the solution set.

[0086]

[0087] i,j=1,2,…,N po #(2.2)

[0088] (2) HV is an indicator that measures the size of the hypervolume occupied by the solution set generated by the algorithm. It evaluates the performance of the algorithm by calculating the hypervolume between the solution set generated by the algorithm and the true frontier. The larger the HV value, the larger the hypervolume occupied by the solution set generated by the algorithm, that is, the better the diversity and coverage of the solution set.

[0089]

[0090] (3) GD is an indicator that measures the distance between the solution set generated by the algorithm and the true frontier. GD evaluates the performance of the algorithm by calculating the distance between each solution in the solution set generated by the algorithm and the true frontier and averaging the distances of all solutions. The smaller the GD value, the closer the solution set generated by the algorithm is to the true frontier.

[0091]

[0092] Please also read Figure 3 In the MORIME algorithm, the relevant process of solving the problem is established as follows:

[0093] Start by initializing the frost population and rime parameters, use the improved soft frost algorithm to update the solution set position, use the hard frost algorithm to exchange information between the optimal individuals and ordinary individuals, obtain the optimal solution through the forward greedy mechanism, check whether the indicator reaches the range, and end if it reaches it. If not, return to use the improved soft frost algorithm to update the solution set position and recycle.

[0094] S4: Multi-dimensional combination of NSGA-II algorithm and MORIME algorithm, which combines the algorithms in multiple dimensions to solve the two-level multi-objective optimization problem;

[0095] Because the algorithm needs to solve a two-layer multi-objective optimization problem, it is decided to combine the algorithms in multiple dimensions. In S4, the NSGA-II algorithm and the MORIME algorithm are combined in multiple dimensions. The combination methods include: cross combination and front-to-back superposition.

[0096] Crossover combination:

[0097] Please also read Figure 4 In the NSGA-II algorithm, we extend some modules of the MORIME algorithm and add another data flow path on this basis, so that it can have the common advantages of both algorithms and can also handle dual-layer multi-objective problems. The process of solving the problem is as follows Figure 4 shown.

[0098] Front and back superposition:

[0099] Although both methods evolve the final solution by generating related methods of the population, Dans handles the population in different ways. Therefore, the front-to-back superposition method can make the result obtained by the algorithm absorb the respective advantages of the two different methods, and also achieve mutual communication between the two methods through population generation.

[0100] S5: Competitive solution of NSGA-II algorithm and MORIME algorithm. For a common two-level multi-objective optimization problem, it can be divided into two problems. Different combinations of algorithms form corresponding solution networks. The NSGA-II algorithm and MORIME algorithm are combined to form the NSGA-RIME algorithm.

[0101] Please also read Figure 5 , since we have obtained the multi-dimensional combination of algorithms, that is, the corresponding method to solve the problem, we now need to consider the algorithm solution so that the algorithm can have better performance and the interconnection of the two-layer objectives. Therefore, NSGA-II and MORIME algorithms are combined with each other to let the problem solution pass through this network to obtain the corresponding decision solution, such as Figure 5 shown.

[0102] For an ordinary two-layer multi-objective optimization problem, it can be divided into two problems (containing all the information of multi-objective optimization). Different combinations of algorithms form corresponding solution networks. The two problems are solved through the network to obtain two Pareto front ends. These two Pareto front ends are a set of solutions. They need to go through a process of scoring decisions to screen out the optimal decision.

[0103] S6: The new scheduling method is proposed to expand and supplement the existing optimization scheduling method to a certain extent, strengthen the scheduling capacity of the park, and enrich the existing scheduling methods.

[0104] After establishing the corresponding two-layer energy-carbon target operation mechanism-data model, a new multi-objective optimization NSGA-RIME algorithm was designed and improved. Therefore, combined with the actual situation of the park, the corresponding new scheduling method was organized and supplemented, and on this basis, the corresponding multi-objective optimization function to be input and its constraints were established. Finally, the NSGA-RIME algorithm was introduced to extract the decision, generate the optimal decision and discuss the relationship between the two-layer decisions.

[0105] The existing optimization scheduling method is expanded and supplemented to a certain extent, the scheduling capacity of the park is strengthened, and the existing scheduling means are enriched. Therefore, the following scheduling methods are proposed:

[0106] In S6, new dispatching methods include: relevant dispatching technologies based on carbon monitoring and capture, the application of carbon emission circulation networks, and dispatching methods for mobile energy storage vehicles in conjunction with carbon dispatching.

[0107] Related dispatching technologies based on carbon monitoring and capture

[0108] In addition to the thermal power plants on the grid side, there are also some multi-state energy units in the park that directly emit carbon or indirectly emit carbon during energy form conversion. Considering that the park may be equipped with carbon capture devices that absorb various forms of carbon emissions in the park and store them or generate electricity and other forms of energy, carbon monitoring equipment is added based on this device, and the means and monitoring data are fed back to the dispatching terminal in real time, so that the carbon emissions of the energy units can be coordinated among multiple energy units and multiple parks.

[0109] Application of carbon emission circulation network:

[0110] In some cases where the differences in carbon emissions are too large, a relevant carbon emission network can be established and superimposed on the power network. This method can be used to balance the carbon emission distribution in the park. This network is mainly transported by pipelines, and related carbon fixation technologies can also be used for mobile transportation when the spatial and temporal scales are small.

[0111] Mobile energy storage vehicle dispatching method in conjunction with carbon dispatching:

[0112] As mentioned earlier, power generation on the grid side is generally carried out by thermal power plants. This carbon emission is still calculated on the user side. The energy used by the user and the energy flow between the user network are simulated to propose a carbon emission flow. Therefore, a certain energy equivalent can be converted into an equivalent carbon emission equivalent according to actual conditions. At this time, the mobile energy storage vehicle can dispatch carbon indicators by moving.

[0113] The mathematical modeling of the optimization scheduling method is included in the existing model, that is, the generation of decisions: When the above scheduling method is clear, we need to summarize the multi-objective optimization model input into the algorithm, as well as the constraints, objective functions and coupling relationships of the two-layer multi-objective.

[0114] S7: The two-level multi-objective model to be solved is established. When the above scheduling method is clear, the multi-objective optimization model input into the algorithm, as well as the constraints, objective functions and coupling relationships of the two-level multi-objectives are summarized;

[0115] In S7, the dual-layer multi-objective objective function, constraint conditions and coupling relationships are:

[0116] Objective function:

[0117]

[0118] σ1=f(x1,…,x n )#(3.3)

[0119] σ2=g(y1,…,y n )#(3.4)

[0120] The objective function represents the energy-saving requirements and carbon emission requirements, wherein each objective function has multiple modules, each module is composed of the data items of the mathematical modeling described in the protection point 1 as independent variables, so that the above data can reflect the relationship between the optimal situation of the objective function effect.

[0121] Constraints:

[0122]

[0123] Each layer of the objective function requires corresponding constraints to cooperate, which are mainly divided into three parts: 1. The inequality optimization conditions of the combined function, which mainly reflect that the comprehensive data between certain indicators in the park needs to meet the relevant requirements of the power grid; 2. The constraints of a single independent variable, which mainly reflect that a certain carbon-energy indicator of a certain park or an energy unit in a certain park needs to exist within a certain range; 3. The correlation between a certain data and another data in the park, just find the corresponding correct function model.

[0124] Coupling relationship:

[0125] As mentioned above, there is a strong coupling relationship between the carbon-energy dual-layer multi-objectives. For example, in the power grid-park network, the carbon flow is introduced to simulate the flow of carbon emissions in the nodes, which is basically the same as the flow of electric energy in the network. Therefore, it is necessary to construct the corresponding relationship function between the data items, i.e., independent variables, between the carbon target and the energy-saving target to reflect the possible interaction relationship between them:

[0126] H l (x1,…,x i ,…,x n )±J k (y1,…,y j ,…,y m )=0 l=1,2...,n;k=1,2...,m#(3.7).

[0127] S8: Use of NSGA-RIME algorithm and formation of double Pareto front end. Input the double-layer multi-objective model into the NSGA-RIME algorithm, and input the double-layer multi-objective into different corresponding algorithm network layers respectively to obtain the corresponding Pareto front end decision set. First, input the related model constructed earlier into the NSGA-RIME algorithm proposed in protection point 2, and input the double-layer multi-objective into different corresponding algorithm network layers respectively to obtain the corresponding Pareto front end decision set. At this time, the set is for independent related decision information at two levels.

[0128] S9: Evaluation and decision integration. The dual Pareto front-end decision set includes different scheduling strategies. The two Pareto front-end decision sets are combined and an evaluation system is established based on them.

[0129] S10: Obtain the final single decision set, output the common parts of the two sets, screen the different parts through the relevant evaluation system, obtain the allowed relevant decisions, and incorporate them into the final result.

[0130] The double Pareto front-end decision set contains different scheduling strategies, but these scheduling strategies may be based on the same specific scheduling means, but there may be certain conflicts to some extent. Therefore, in this case, it is necessary to integrate the two decision sets and establish an evaluation system on this basis. The common parts of the two sets are output, and the different parts are screened through the relevant evaluation system to obtain the allowable relevant decisions and incorporate them into the final result.

[0131] The present invention first studies and summarizes the factors and data of concern such as the characteristics, interconnection relationships and operation mechanisms of different polymorphic energy units between various parks, and on this basis combines the carbon emissions of energy units and the power supply carbon index of the power grid with the above factors to summarize the data information and operation principles of various energy units in various parks, so as to further establish a corresponding level of carbon energy double-layer data-mechanism model for various energy units in various parks while taking into account the dual goals of energy saving and carbon emissions. Since the above-mentioned operation data-mechanism model is established for the carbon-energy goal and has two different optimization goals, different The optimization objectives are optimized in layers. At this time, the conventional NSGA algorithm is not competent for such complex scenarios. Therefore, referring to its iterative version NSGA-Ⅱ algorithm and the MORIME algorithm which is also used to solve the multi-objective optimization results, they are combined and a NSGA-RIME combined algorithm model for double-layer multi-objective optimization is proposed to solve the double-layer multi-objective model, so as to solve the complexity and polymorphism of the energy system, the dual pressure of energy conservation and carbon emissions, the need to solve multi-objective optimization problems and the need for algorithm innovation and combined application. This method is helpful to promote the optimization and upgrading of the energy system in the park and its sustainable development.

[0132] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

Claims

1. A multi-energy optimization method based on the NSGA-RIME algorithm that takes into account the integration of multiple parks with carbon targets, characterized by: The following steps are involved: S1: Construct a two-layer digital-mechanism model of multi-park multi-energy states, and study and summarize the factors and data of concern such as the characteristics, interconnection relationships, and operation mechanisms of different multi-state energy units in each park; S2: Mining and determining the coupling relationship between carbon and energy dual-layer targets. After collecting relevant information on the park's carbon targets and energy targets, characterize the functional relationship and data connection between the data-related models of the two aspects, so that it can achieve effective coupling of the two models in the carbon-energy dual-layer digital-mechanism model, and enable them to interact with each other in scheduling optimization; S3: Introducing the MORIME algorithm. In the process of finding the optimal solution, the MORIME algorithm gradually covers the entire search space and retains the optimal decision; S4: Multi-dimensional combination of NSGA-II algorithm and MORIME algorithm, which combines the algorithms in multiple dimensions to solve the two-level multi-objective optimization problem; S5: Competitive solution of NSGA-II algorithm and MORIME algorithm. For a common two-level multi-objective optimization problem, it can be divided into two problems. Different combinations of algorithms form corresponding solution networks. The NSGA-II algorithm and MORIME algorithm are combined to form the NSGA-RIME algorithm. S6: The new scheduling method is proposed to expand and supplement the existing optimization scheduling method to a certain extent, strengthen the scheduling capacity of the park, and enrich the existing scheduling methods; S7: The two-level multi-objective model to be solved is established. When the above scheduling method is clear, the multi-objective optimization model input into the algorithm, as well as the constraints, objective functions and coupling relationships of the two-level multi-objectives are summarized; S8: Use of NSGA-RIME algorithm and formation of dual Pareto front end. The dual-layer multi-objective model is input into the NSGA-RIME algorithm. The dual-layer multi-objective is input into different corresponding algorithm network layers respectively to obtain the corresponding Pareto front end decision set. S9: Evaluation and decision integration. The dual Pareto front-end decision set includes different scheduling strategies. The two Pareto front-end decision sets are combined and an evaluation system is established based on them. S10: Obtain the final single decision set, output the common parts of the two sets, screen the different parts through the relevant evaluation system, obtain the allowed relevant decisions, and incorporate them into the final result.

2. The multi-energy optimization method based on the NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1 is characterized in that: In S1, when constructing a two-layer digital-mechanism model of multiple parks and multiple energy states, the collected data are summarized including: park energy unit characteristics and data collection in line with energy-saving targets and park energy unit characteristics and data collection in line with carbon targets.

3. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 2 is characterized in that: In S1, the park energy unit characteristics and data collection content that are in line with the energy-saving goals include: the operating mechanism and data of each park itself; the operating mechanism and data of the power network and other energy networks between parks; the operating mechanism and data summary of polymorphic energy within the park.

4. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 3 is characterized by: In S1, the park energy unit characteristics and data collection content that match the carbon target include: carbon exchange characteristics and data collection between parks; carbon indicators of the park itself; and park energy units.

5. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1 is characterized in that: In S2, the mining and determination of the carbon-energy dual-layer target coupling relationship includes: carbon-energy data model alignment; carbon-energy data model coupling relationship.

6. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1 is characterized in that: In S1, a two-layer digital-mechanism model of multiple parks and multiple energy states is constructed, including an aggregated multi-dimensional mechanism model that integrates the park network and energy unit model; and an energy-carbon data model of energy units between parks and within the park itself.

7. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1 is characterized in that: In S3, when the MORIME algorithm is introduced, the relevant parameters required by the algorithm include: SP is used to measure the distribution uniformity of the solution set generated by the algorithm; HV is an indicator to measure the size of the hypervolume occupied by the solution set generated by the algorithm; GD is an indicator to measure the distance between the solution set generated by the algorithm and the true frontier.

8. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1 is characterized in that: In S4, the NSGA-II algorithm and the MORIME algorithm are combined in multiple dimensions, and the combination methods include: cross combination and front-back superposition.

9. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1, characterized in that: In S6, new dispatching methods include: relevant dispatching technologies based on carbon monitoring and capture, the application of carbon emission circulation networks, and dispatching methods for mobile energy storage vehicles in conjunction with carbon dispatching.

10. The multi-energy optimization method based on NSGA-RIME algorithm and integrated with multiple parks taking into account carbon targets according to claim 1, characterized in that: In S7, the dual-layer multi-objective objective function, constraint conditions and coupling relationships are: Objective function: σ1=f(x1,…,x n )#(3.3) σ2=g(y1,…,y n )#(3.4) Constraints: Coupling relationship: H l (x1,…,x i ,…,x n )±J k (y1,…,y j ,…,y m )=0 l=1,2…,n;k=1,2…,m#(3.7)。

Images