Multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation and control

Abstract

The application discloses a multi-energy regional autonomous optimization regulation method for virtual micro-grid regulation, and particularly relates to the technical field of power system optimization regulation, and the regulation method steps are as follows: corresponding data is collected, a virtual micro-grid is taken as a basic unit, a topology structure based on energy transmission is established, and meanwhile, a tie-in relationship is established between different virtual micro-grid units; a parallel energy supply scene is constructed, then based on equivalent power transmission power and actual power transmission power, and combined with the self state of each region, a corresponding evaluation function is established, a MOEDO algorithm is used to solve the problem of falling into a local optimal solution, and a corresponding scheduling strategy is obtained; the multi-energy regional autonomous optimization regulation method for virtual micro-grid regulation firstly gives a regional virtual micro-grid definition through multi-state energy transmission, normalizes the energy transmission means as a virtual tie-in definition, so as to solve the problem of other forms of energy transmission between different regions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system optimization regulation, and more particularly to a multi-energy regional autonomous optimization regulation method for virtual micro-grid regulation. BACKGROUND

[0002] With the promotion of new energy and environmental protection policies, more and more distributed clean energy and energy-saving technologies have been widely popularized. Most of the distributed clean energy and energy-saving loads show a relatively concentrated characteristic, such as being widely used in industrial, commercial, residential and other parks. In a single park, it is difficult to optimize and regulate multiple energies, and due to the differences in demand and supply between multiple parks, mutual cooperation and support are needed, so how to realize the optimization and regulation of multiple energies across parks, that is, the autonomous optimization and regulation of multi-energy regions, has important research significance and application value.

[0003] Therefore, in view of the above problems, a multi-energy regional autonomous optimization regulation method for virtual micro-grid regulation is proposed. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the present application provides a multi-energy regional autonomous optimization regulation method for virtual micro-grid regulation to solve the problems raised in the background art.

[0005] To achieve the above object, the present application provides the following technical scheme: a multi-energy regional autonomous optimization regulation method for virtual micro-grid regulation, the regulation method steps are:

[0006] Collecting corresponding data, constructing a target optimal function for measuring each node virtual micro-grid according to actual needs;

[0007] Taking virtual micro-grid as a basic unit, establishing a topology structure based on energy transmission, and establishing a tie-up relationship between different virtual micro-grid units, and then virtually transferring actual non-electric energy into actual electric energy transmission;

[0008] Constructing a parallel energy supply scenario: constructing a set of energy use power data of each virtual micro-grid without virtual tie-up, which is only powered by the grid side and exchanges energy with other virtual micro-grids at each time period;

[0009] Then, based on the equivalent power transmission power combined with the actual power transmission power, and combined with the state of each region, a corresponding evaluation function is established, each park generates a corresponding target function, and the MOEDO algorithm is used to solve to avoid falling into a local optimal solution, and the corresponding scheduling strategy is obtained.

[0010] Further, the virtual micro-grid is a regional complex that is relatively independent at the electric energy level and has multiple forms of energy.

[0011] Further, the parallel energy supply scenario is divided into an invariable part and a variable part, and each part is expanded and measured, wherein the invariable part is the relatively fixed plan of energy use of the virtual micro-grid area at each time scale within a day, and the variable part is the physical modeling of the energy channel, and for the transport pipeline or mobile energy storage vehicle, the energy path with strong randomness and no time period variation is realized.

[0012] Further, the invariable part of the parallel energy supply scenario first carries out data collection of energy supply: data acquisition of 100 data sites within 365 days in a year for each regional virtual micro-grid, to obtain the corresponding data set P:

[0013] P = [P ij ](i = 0, 1, 2, … 100; j = 0, 1, 2, … 365)

[0014] Wherein, P ij is the power usage of the node within any time period,

[0015]

[0016] Wherein P l is the time period power of the main energy unit or micro node in the node, W lall is introduced as the energy usage size of other non-electric energy units, t all is the total output time of the energy unit, P avg is the average power of new energy output in the region, P plant is the power of the power plant in the region, P G is the power supply on the grid side.

[0017] Then, the data is processed in a general time scale,

[0018] Through time scale fitting of power variation, for the region with large power variation and power peak value, the time scale description is:

[0019] For the power peak time period, the power description collection can collect all data sites; when the fluctuation size δ of the data site is greater than the specified value, the same processing is also done, and the specified value is:

[0020]

[0021] For data site homogenization or power trough, the final actual data set P' is obtained by reduction processing:

[0022] P' = [P ij ](i = 0, 1, 2, … n; j = 0, 1, 2, … 365, n < 100)

[0023] Then, based on the fitting process of PeepholeLSTM: further processing of the obtained data set, by using PeepholeLSTM to carry out corresponding power prediction fitting, the power data curve reflecting the characteristics of the region is obtained:

[0024] f t = σ (W f · [C t-1 , h t-1 , x t ] + b f )

[0025] i t = σ (W i · [C t-1 , h t-1 , x t ] + b i )

[0026] o t = σ (W o · [C t , h t-1 , x t ] + b o )

[0027]

[0028] h (t) = o (t) · tanh (c (t) )

[0029] P' is integrated into the form of x t input, the corresponding weight matrix W f , W i , W o , W (c) , b f , b i , b o is obtained by training the model, and until the model obtains the ideal data set h (t) .

[0030] Further, the variable part of the parallel energy supply scene first establishes a mobile energy storage energy supply channel,

[0031] Specifically, the corresponding energy transmission function is established,

[0032]

[0033] Where SOC i is the state of charge of the mobile energy storage vehicle, sum is the number of mobile energy storage vehicles that can be called, and l is the length of the regional path from all energy storage vehiclesi ;

[0034] Other non-periodic energy transfer variations, such as those occurring in energy channels with strong inertia and exhibiting daily planned changes or sudden, temporary variations, are collected to obtain a time-independent data sequence P. t Perform an exponential average, specifically:

[0035] v t =βP t +(1-β)v t-1

[0036] The corresponding sequence data v was obtained. t The sequence data is coupled with the preceding time dimension data according to the actual situation, so that the final output sequence data considers both the actual power consumption of the region and the actual energy transmission capacity of the region from the power perspective.

[0037] Furthermore, in the parallel power supply scenario, it is also necessary to accumulate and couple the constant and variable components. By accumulating and coupling the power of the constant and variable components, the final actual transmission power is obtained. The coupling weights are repeatedly calculated through regression, and finally, the daily power time series is output.

[0038] P ij =W con ⊙P'+W var ⊙P var

[0039] The actual electrical power exchanged between regions is obtained, and the power transmission of power lines with non-constant intraday multi-point curves is used to replace the actual energy transmission.

[0040] The technical effects and advantages of this application are as follows:

[0041] 1. Compared with existing technologies, this multi-energy regional autonomous optimization control method for virtual microgrid regulation first defines the regional virtual microgrid through multi-state energy transmission and standardizes the energy transmission means as virtual interconnection definition, thereby solving other forms of energy transmission problems between different regions.

[0042] 2. Compared with existing technologies, this multi-energy regional autonomous optimization control method for virtual microgrid regulation aims to accurately describe the power transmission of power tie lines. It quantifies the specific transmission situation by measuring the actual energy used, performs result-oriented power measurement, and finely classifies the power transmission source into variable and constant parts, coupling them accordingly to obtain the final power transmission of the simulated power tie line.

[0043] 3. Compared with the prior art, the multi-energy regional autonomous optimization control method for virtual micro-grid control combines power information and park energy consumption, involves an evaluation function to form an optimal park control function, forms a multi-objective optimization problem, and obtains the optimal control means by solving the problem. BRIEF DESCRIPTION OF DRAWINGS

[0044] Fig. 1 a flowchart of the control method of the present application;

[0045] Fig. 2 a schematic diagram of the fitting processing based on PeepholeLSTM in the present application. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present application will be described clearly and completely in the embodiments of the present application in combination with the drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0047] EMBODIMENT

[0048] The present application first determines three definitions, specifically the definition of virtual micro-grid, the definition of virtual connection and the definition of quasi-electric super-distance control. The definition of virtual micro-grid is to virtualize the concept of micro-grid by imagining virtual energy channels, and to regard a regional complex with relatively independent electric energy and multiple energy types as a nominal "virtual micro-grid". The specific definition conditions are as follows:

[0049] (1) The virtual micro-grid must be directly or indirectly supplied with electric energy by the power grid;

[0050] (2) There should be no corresponding electric power transmission path between the virtual micro-grids, or the electric power transmission scale between the virtual micro-grids can be almost negligible compared to the regional power consumption scale;

[0051] (3) The virtual micro-grid must have a relatively large scale, so that it has the necessary participation in control;

[0052] (4) The virtual micro-grid should have multiple energy units, i.e. the ability to generate or transmit non-electric energy;

[0053] (5) The virtual micro-grid should have the ability to exchange energy through non-electric energy transmission.

[0054] By introducing the above constraints, the specific concept of virtual micro-grid is actually defined and described. The control or scheduling performed below is based on the above basic definition unit.

[0055] The definition of virtual interconnection is based on virtual microgrids as basic units, establishing a topology structure based on energy transmission. This establishes interconnection relationships between different virtual microgrid units. Simultaneously, it aims to virtualize actual non-electrical energy transmission into actual electrical energy transmission, integrating it into overall control and providing a more accurate basis for control. The specific definition conditions are as follows:

[0056] (1) Virtual communication is based on actual non-electric energy transmission, and its specific methods include but are not limited to hydroelectric potential energy transmission, oil and gas energy transmission, mobile energy transmission, etc.

[0057] (2) Virtual interconnection is equivalent to the corresponding power transmission. The calculation of the amount of energy transmitted and the amount of actual energy replaced is equivalent to the transmission of actual power interconnection lines between different virtual microgrids.

[0058] (3) Virtual interconnection is not exactly the same as traditional power line transmission. The energy transmission carried out by virtual interconnection does not have the electrical characteristics of traditional power transmission and does not have power flow.

[0059] (4) Virtual communication is highly variable and changes in real time according to the actual transmission energy situation. It is time-specific and not constant.

[0060] Based on the above definitions, we point out that virtual interconnection is actually a pseudo-electrical channel between virtual microgrids that achieves the same effect as power interconnection lines through the transmission and exchange of polymorphic energy (other forms of energy). By defining the nodes in the virtual microgrid topology and the virtual interconnections between virtual microgrids, we establish the target structure for regulation. Therefore, the target structure to be regulated is clear and defined at the power transmission level.

[0061] The definition of pseudo-electrical distance control is that, after having a target to be regulated, based on the actual needs of different regions and the non-electrical energy transmission capacity, regulation is carried out from the relevant perspectives of power regulation within the established network structure, ultimately enabling the virtual microgrid region to achieve optimal performance in terms of energy consumption cost, carbon emission cost, energy margin, and other aspects. Therefore, pseudo-electrical distance control is defined as: generating optimized regulation equivalent to the power level by planning the multi-state energy transmission of different "virtual microgrids", even if there is no actual power transmission between the two regions being regulated.

[0062] As attached Figs. 1-2 The above describes a multi-energy regional autonomous optimization control method for virtual microgrid regulation. The control method steps are as follows:

[0063] Collect the corresponding data and construct the target optimal function to measure the virtual microgrid of each node according to actual needs;

[0064] Virtual micro-grid as a basic unit, establish based on the topology of energy transmission, while different virtual micro-grid units between the establishment of liaison relationship, then the actual non-energy energy transmission virtualization to the actual power transmission;

[0065] Construct parallel energy supply scenario: construct the use of energy power data set of each virtual micro-grid without virtual liaison in each time period only powered by the grid side and energy exchange with other virtual micro-grids;

[0066] Then based on the equivalent power transmission power combined with the actual power transmission power, combined with the state of each region, the corresponding evaluation function is established, each park produces a corresponding objective function, using MOEDO algorithm to avoid falling into local optimal solution, and the corresponding scheduling strategy is obtained.

[0067] Virtual micro-grid is a regional complex that is relatively independent in the field of electric energy and has multiple forms of energy.

[0068] Parallel energy supply scenario is divided into invariable part and variable part, and is measured respectively, wherein the invariable part is the relatively fixed energy plan of virtual micro-grid area in each time scale within a day; the variable part is the physical modeling of energy channel, for transportation pipeline or mobile energy storage vehicle, to realize the energy path with strong randomness and no time period variation.

[0069] The invariable part of the parallel energy supply scenario first carries out data collection of energy supply: data acquisition of 100 data points in 365 days in a year for each regional virtual micro-grid, to obtain the corresponding data set P:

[0070] P=[P ij ](i=0,1,2,…100;j=0,1,2,…365)

[0071] Among them, P ij is the power usage of the node in any time period,

[0072]

[0073] Among them, P l is the time period power of the main energy unit or micro node in the node, W lall is introduced as the energy size of other non-electric energy units, t all is the total output time of the energy unit, P avg is the average power of new energy output in the region, P plant is the power of the power plant in the region, P G is the power supply of the grid side;

[0074] Then the data is processed in a general time scale,

[0075] Through the time scale fitting power changes, power fluctuation, power peak area time scale description is:

[0076] For the use of electricity peak time period, power description collection can collect all data points; when the data point fluctuation size δ is greater than the specified value, the same processing is done, the specified value is:

[0077]

[0078] For data points homogenization or power trough, the final actual data set P' is obtained by reducing processing:

[0079] P' = [P ij ](i = 0, 1, 2, … n; j = 0, 1, 2, … 365, n < 100)

[0080] Then based on the fitting processing of PeepholeLSTM: the obtained data set is further processed, and the corresponding power prediction fitting is carried out by using PeepholeLSTM, and the power data curve reflecting the characteristics of the region is obtained:

[0081] f t = σ (W f · [C t-1 , h t-1 , x t ] + b f )

[0082] i t = σ (W i · [C t-1 , h t-1 , x t ] + b i )

[0083] o t = σ (W o · [C t , h t-1 , x t ] + b o )

[0084]

[0085] h (t) = o (t) · tanh (c (t) )

[0086] Integrate P' into x t form input, train the model to obtain the corresponding weight matrix W f , Wi , W o , W (c) , b f , b i , b o and until the model gets the ideal data set h (t) .

[0087] The variable part of the parallel energy supply scenario first establishes a mobile energy storage energy supply channel,

[0088] Specifically, the corresponding energy transmission function is established,

[0089]

[0090] Where the state of charge of the mobile energy storage vehicle is SOC i , the number of available mobile energy storage vehicles is sum, and the total energy storage vehicle distance from the regional path length l i ;

[0091] Other non-periodic energy transmission changes, for energy channels with strong inertia, have daily plan changes or temporary sudden changes, and become non-periodic, and corresponding collection is performed, obtaining a time-independent data sequence P t , exponential average processing, specifically:

[0092] v t = βP t + (1-β)v t-1

[0093] The corresponding sequence data v t is obtained, and the sequence data is coupled with the previous time dimension data according to the actual situation, so that the final output sequence data considers both the actual power consumption of the region and the actual energy transmission capacity of the region from the power level.

[0094] In the parallel energy supply scenario, the invariable part and the variable part also need to be accumulated and coupled, and the invariable part and the variable part are added and coupled to obtain the final actual transmission power size. The coupling weight is obtained by regression repeatedly, and the final output is the power time sequence within the day,

[0095] P ij = W con ⊙P'+ W var ⊙P var

[0096] The actual energy exchange power size between regions is obtained, and the power transmission of the power line with non-constant daily multi-point curve is used to replace the actual energy transmission situation.

[0097] For the actual situation of the region, the following several indicators are proposed as variables of the objective function:

[0098] (1) The actual utilization rate of new energy equipment, the actual situation of new energy equipment output accounting for actual use or power flow α:

[0099]

[0100] (2) The proportion of virtual contact energy transmission in the peak period of regional power consumption β:

[0101]

[0102] (3) The energy consumption of the park from the grid side P iG :

[0103] P iG =P i -P ij

[0104] Therefore, the target problem to be solved is:

[0105]

[0106] Through the solution of this problem, the corresponding energy distribution plan is obtained, which makes each virtual micro-grid obtain less power from the grid side and use as much new energy converted power as possible while ensuring that it can meet the load demand at peak time under the constraint of virtual contact.

Claims

1. A multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation and control, characterized in that: The regulation method step is: Collecting corresponding data, and constructing a target function for measuring each node virtual micro-grid according to actual needs; Taking the virtual micro-grid as a basic unit, establishing a topology structure based on energy transmission, and establishing a tie between different virtual micro-grid units, and then virtually transferring the actual non-electric energy into actual electric energy transmission; Constructing a parallel energy supply scenario: constructing a virtual micro-grid without virtual ties, and collecting energy power data sets of each time period of the virtual micro-grid only powered by the grid side and exchanging energy with other virtual micro-grids; Then, based on the equivalent electric power transmission power combined with the actual electric power transmission power, and combined with the state of each region, an evaluation function is established, each park produces a corresponding target function, and the MOEDO algorithm is used to solve to avoid falling into a local optimal solution, and the corresponding scheduling strategy is obtained. 2.The multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation and control according to claim 1, characterized in that: The virtual micro-grid is a regional complex that is relatively independent in the electric energy level and has multiple forms of energy. 3.The multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation and control according to claim 1, characterized in that: The parallel energy supply scenario is divided into an invariable part and a variable part, and each is expanded for measurement, wherein the invariable part is the relatively fixed plan of the virtual micro-grid region in each time scale within the day; the variable part is the physical modeling of the energy channel, and for the transportation pipeline or mobile energy storage vehicle, the energy path with strong randomness and no time period variation is realized.

4. The multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation according to claim 3, characterized in that: The invariable part of the parallel energy supply scenario first collects energy supply data: data is obtained from 100 data points in the standard 365 days of each regional virtual micro-grid within a year, and a corresponding data set P is obtained: P = [P ij ](i = 0, 1, 2,... 100; j = 0, 1, 2,... 365) where P ij is the node power usage for any time period, Where P l is the time period power of the main energy unit or micro node in the node, W lall is the energy consumption size of other non-electric energy units, t all is the total output time of the energy unit, P avg is the average power of new energy output in the region, P plant is the power generation of the power plant in the region, P G is the power supply on the grid side; Then the data is processed in a general time scale, Through time scale fitting of power changes, the time scale description of the power peak area is given for power variation and power peak area: For the power peak time period, all data points can be collected for power description collection; when the data point fluctuation size δ is greater than the specified value, the same processing is performed, and the specified value is: For data point homogenization or power trough, the final actual data set P' is obtained through reduction processing: P' = [P ij ](i = 0, 1, 2, … n; j = 0, 1, 2, … 365, n < 100) The fitting processing based on PeepholeLSTM: the data set obtained is further processed, and the power data curve reflecting the characteristics of the region is obtained through the use of PeepholeLSTM for corresponding power prediction fitting: f t = σ(W f · [C t-1 , h t-1 , x t ]+ b f ) i t = σ(W i · [C t-1 , h t-1 , x t ]+ b i ) o t = σ(W o · [C t , h t-1 , x t ]+ b o ) h (t) = o (t) · tanh(c (t) ) P' is integrated into x t Form input, corresponding training of the model to find the corresponding weight matrix W f , W i , W o , W (c) , b f , b i , b o And until the model gets the ideal data set h (t) .

5. The multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation according to claim 4, characterized in that: The variable part of the parallel energy supply scenario first establishes a mobile energy storage energy supply channel, Specifically, the corresponding energy transmission function is established, Wherein the state of charge of the mobile energy storage vehicle is SOC i , the number of the mobile energy storage vehicles that can be called is sum, and the total length of the region path is l i ; Other non-periodic energy transmission changes, for strong inertia energy channel has daily plan changes or temporary sudden changes, and is non-periodic, corresponding to the collection, get the time-independent data sequence P t , do exponential average processing, specifically: v t = βP t +(1-β)v t-1 The corresponding sequence data v is obtained t The sequence data is coupled with the previous time dimension data according to actual conditions, so that the final output sequence data considers both the actual power consumption of the region and the actual energy transmission capacity of the region from the power level. 6.The multi-energy regional autonomous optimization and regulation method for virtual micro-grid regulation and control according to claim 5, characterized in that: In the parallel energy supply scenario, the invariable part and the variable part also need to be accumulated and coupled, and the invariable part and the variable part are added and coupled to obtain the final actual transmission power size, and the coupling weight is obtained through regression repeatedly, and the final output is the power time sequence within the day, P ij = W con ⊙P'+W var ⊙P var The electric power size of the actual energy exchange between regions is obtained, and the power transmission of the power line with the non-constant multi-point curve within the day is used to replace the actual energy transmission.

Images