Distributed collaborative optimization method, system and device for electricity-gas coupled energy system

Abstract

The invention relates to the technical field of coupled energy collaborative optimization, in particular to a distributed collaborative optimization method, system and device for an electricity-gas coupled energy system, and the method comprises the steps: constructing a working model of the electricity-gas coupled energy system; constructing a load utility model based on the working model according to the conversion loss, the calculation coefficient and the conversion interval of the electricity-gas coupling energy system; and acquiring an electrical load and a gas load of a target area in real time, inputting the electrical load and the gas load into the working model and the load utility model, solving by adopting a completely distributed neural dynamic optimization algorithm, obtaining an optimal solution, and completing distributed cooperative control of the electricity-gas coupled energy system. According to the invention, the load energy consumption condition of the electricity-gas coupling energy system can be monitored and adjusted in real time, and distributed optimization is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of coupled energy collaborative optimization, and particularly to a distributed collaborative optimization method, system and device for an electric-gas coupled energy system. Background Art

[0002] Energy is an important foundation for supporting social life, economic development and technological innovation. With the high proportion integration of renewable energy, the rapid development of distributed energy production and energy storage technologies, and the diversified needs of end-users, the energy system is gradually developing from a single energy source to an integrated energy system. The complexity of the energy system has increased significantly, and the coupling of multiple energy supply systems such as electricity and gas has become a real demand. The integrated energy system can integrate and coordinately operate multiple energy sources (such as electricity, natural gas, etc.) and energy equipment (such as generators, boilers, etc.), and is an efficient operation scheme to promote the integration of source-network-load-storage, multi-energy complementarity and supply-demand coordination. In order to achieve the efficient collaboration and optimal scheduling of these energy systems, a theoretical system with distributed control as the core has emerged. Summary of the Invention

[0003] The object of the present invention is to provide a distributed collaborative optimization method, system and device for an electric-gas coupled energy system in view of the above technical problems.

[0004] To achieve the above object, on the one hand, the present invention provides a distributed collaborative optimization method for an electric-gas coupled energy system, including:

[0005] Constructing a working model of the electric-gas coupled energy system;

[0006] Based on the conversion loss, calculation coefficient and conversion interval of the electric-gas coupled energy system, constructing a load utility model based on the working model;

[0007] Real-time collecting the electricity load and gas load of the target area, inputting them into the working model and the load utility model, and using the fully distributed neural dynamic optimization algorithm to solve for the optimal solution, thereby completing the distributed collaborative control of the electric-gas coupled energy system.

[0008] Optionally, constructing the working model of the electric-gas coupled energy system includes:

[0009] Combining the balance relationship among electricity, gas and load of the equipment in the electric-gas coupled energy system, as well as the constraints of the equipment, the cost of fuel power generation equipment, and the cost of renewable energy equipment, to construct the working model.

[0010] Optionally, the balance relationship among electricity, gas and load of the equipment in the electric-gas coupled energy system is:

[0011]

[0012] Among them, p represents the input / output power of electricity and gas, and g represents the input / output power of gas; L is the power of the load demand; rg, fg, g2p, psd, gp, gsd, and p2g respectively represent distributed renewable power generation equipment, distributed fuel power generation equipment, distributed gas-to-electricity equipment, distributed power storage equipment, distributed gas production equipment, and distributed gas storage equipment and distributed electricity-to-gas equipment.

[0013] Optionally, the constraints of the equipment are:

[0014]

[0015] Among them, min and max represent the lower and upper bounds; φ represents electricity and gas; represents the slope limit of the equipment; represents the energy storage capacity of the energy storage equipment; represents the charge-discharge efficiency; represents the state of the energy storage equipment, "1" represents charging, and "-1" represents discharging; represents the charge-discharge capacity per unit time; represents the maximum charging capacity; represents the maximum discharge capacity; ψ p2g and ψ g2p respectively represent the ratios of electricity-to-gas and gas-to-electricity.

[0016] Optionally, the cost of the fuel power generation equipment is:

[0017]

[0018] Among them, is the power output of the i-th fuel power generation equipment; a i 、b i 、c i 、d i and are cost coefficients; and are respectively the lower and upper limits of the power output of the i-th fuel power generation equipment; is the ramp rate of the fuel power generation equipment.

[0019] Optionally, the cost of the renewable energy equipment is:

[0020]

[0021] Among them, is the power output of the i-th renewable energy equipment; a i 、b i 、c i, d i and are cost coefficients; and are the lower and upper limits of the power output of the i-th renewable energy device, respectively; is the ramp rate of the renewable energy device.

[0022] Optionally, the load utility model is:

[0023]

[0024] where, represents the electrical comprehensive load, represents the converted gas consumption, represents the converted electricity consumption, represents the gas and electricity utility coefficients at time t, α > 0, β > 0, and γ > 0 are compensation coefficients, and the first two terms of the equation represent the direct cost of energy consumption, and the third term represents the compensation for encouraging users to switch energy.

[0025] Optionally, the fully distributed neural dynamic optimization algorithm is:

[0026]

[0027] μ ∈ R np μ = [μ1; μ2;...; μ n ;

[0028] λ ∈ R r λ = [λ1; λ2;...; λ r ;

[0029] η ∈ R np η = [η1; η2;...; η n ;

[0030] L p ∈ R np×np

[0031]

[0032] where x = (x1, x2,..., x n ) T represents the state variables of all agents, where x i,t ∈ R qi ; λ represents the local constraint variables; μ and η are the shared variables related to the global constraints of the system; are the vector forms of the variables x, μ, λ, and η, respectively; ▽ is the gradient operator; f(x) ∈ R represents the objective function; g(x) represents the inequality constraint; The vector formed by g(x); where L is the Laplacian matrix, and I p is the p-order identity matrix; Ω represents the feasible operation region; is the matrix formed by the equality constraint A i ; is the constant vector of the equality constraint; R np , R r , R np×np , R nr respectively represent real number spaces of dimensions np, r, np×np, and nr.

[0033] On the other hand, the present invention provides a distributed collaborative optimization system for an electric-gas coupled energy system, including:

[0034] A model construction module, configured to construct a working model of the electric-gas coupled energy system, and based on the conversion loss, calculation coefficient, and conversion interval of the electric-gas coupled energy system, construct a load utility model based on the working model;

[0035] A data acquisition module, configured to collect in real time the energy supply data of the energy supply end and the load data of the load end in the target area;

[0036] A collaborative control module, configured to input the energy supply data and the load data into the working model and the load utility model, and use a fully distributed neural dynamic optimization algorithm to solve, obtain the optimal solution, and complete the distributed collaborative control.

[0037] On the other hand, the present invention provides a distributed collaborative optimization device for an electric-gas coupled energy system, including: a busbar trunking unit, an intelligent terminal, and a server. After collecting the energy supply data of the energy supply end and the load data of the load end in the target area through the busbar trunking unit, input them into the intelligent terminal to perform a fully distributed neural dynamic optimization solution in combination with the working model and the load utility model, obtain the optimal solution and generate a control signal, and then the control signal is issued through the busbar trunking unit. At the same time, the energy supply data, the load data, and the control process are input into the server for storage.

[0038] The beneficial effects of the present invention are:

[0039] The present invention constructs a working model for renewable energy power generation, gas production, and electric-gas conversion, which can refine the composition of multi-energy loads in the system; establishes a renewable energy cost function and a multi-energy load utility function model considering conversion loss, calculation coefficient, and conversion interval, improving the comprehensive energy utilization rate; proposes a neural dynamic optimization algorithm using a fully distributed method, which reduces the shared information sent by each energy node to its neighbor nodes, can monitor and adjust the load energy consumption of the electric-gas coupled energy system in real time, and achieve distributed optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0041] Figure 1 It is a topological diagram of a 41-node test system according to an embodiment of the present invention;

[0042] Figure 2 It is an energy output curve diagram according to an embodiment of the present invention. Among them, (a) is the optimal scheduling curve of electric energy output, and (b) is the optimal scheduling curve of natural gas output;

[0043] Figure 3 It is an energy supply and demand difference curve diagram according to an embodiment of the present invention. Among them, (a) is the deviation convergence situation between the total energy output and the total load demand, and (b) is the convergence situation of the auxiliary variables of all nodes;

[0044] Figure 4 It is a schematic diagram of energy flow and information flow in the distributed collaborative optimization process of an electric-gas coupled energy system according to an embodiment of the present invention;

[0045] Figure 5 It is a schematic diagram of the operation of an intelligent terminal according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0047] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0048] On the one hand, this embodiment provides a distributed collaborative optimization method for an electric-gas coupled energy system, including:

[0049] Construct a working model of the electric-gas coupled energy system;

[0050] Based on the conversion loss, calculation coefficient, and conversion interval of the electric-gas coupled energy system, construct a load utility model based on the working model;

[0051] Collect the electricity load and gas load of the target area in real time, input them into the working model and the load utility model, and use the fully distributed neural dynamic optimization algorithm to solve for the optimal solution, completing the distributed collaborative control of the electric-gas coupled energy system.

[0052] Specifically, in this embodiment, a working model of renewable energy power generation, gas production, and electric-gas conversion is constructed to refine the composition of multi-energy loads in the system; a renewable energy cost function and a multi-energy load utility function model considering conversion losses, calculation coefficients, and conversion intervals are established to improve the comprehensive energy utilization rate; a neural dynamic optimization algorithm using a fully distributed method is proposed. This algorithm enables each energy node to reduce the shared information sent to its neighbor nodes, and can monitor and adjust the load energy consumption of the electric-gas coupled energy system in real time and achieve distributed optimization.

[0053] Furthermore, constructing the working model of the electric-gas coupled energy system includes:

[0054] Combining the balance relationships among electricity, gas, and load of the equipment in the electric-gas coupled energy system, as well as the constraints of the equipment, the costs of fuel power generation equipment, and the costs of renewable energy equipment, to construct the working model.

[0055] Specifically, this embodiment proposes a working model of renewable energy power generation, gas production, and electric-gas conversion, which can refine the composition of multi-energy loads in the electric-gas coupled energy system, as follows:

[0056] For an integrated energy system containing multiple energy devices and loads, the balance relationships among electricity, gas, and load of the i-th device at time t are respectively:

[0057]

[0058] Among them, p represents the input / output power of electricity and gas, g represents the input / output power of gas; L is the power of the load demand; rg, fg, g2p, psd, gp, gsd, p2g respectively represent distributed renewable power generation equipment, distributed fuel power generation equipment, distributed gas-to-electricity equipment, distributed power storage equipment, distributed gas production equipment, distributed gas storage equipment, and distributed electricity-to-gas equipment.

[0059] In the electric-gas coupled system, all power generation equipment has upper and lower limits and rate limitations, as shown in formulas (3)-(4). The storage constraints are represented by formulas (5)-(7), including boundary constraints and charge / discharge slope limitations. Since P2G uses renewable generators to produce gas, its gas production is subject to certain constraints, as shown in formula (8); and gas-to-electricity is similar, as shown in formula (9). The mathematical expressions of these constraints are as follows:

[0060]

[0061] where min and max represent the lower bound and the upper bound respectively; φ represents electricity and gas; represents the slope limit of the device; represents the energy storage capacity of the energy storage device; represents the charge-discharge efficiency; represents the state of the energy storage device, "1" represents charging, and "-1" represents discharging; represents the charge-discharge capacity per unit time; represents the maximum charging capacity; represents the maximum discharging capacity; ψ p2g and ψ g2p represent the ratios of electricity-to-gas and gas-to-electricity respectively.

[0062] Due to the uncertainty and volatility of renewable energy, DFG is required to cooperate with it to jointly provide power supply. Therefore, when constructing the cost function of renewable energy, it is necessary to incorporate the direct operating costs and penalties that reflect the difference between predicted and actual energy production. This penalty term arises because when the actual output deviates from the prediction, DFG must adjust its own power generation strategy. To compensate for the impact of renewable energy on the power generation cost of DFG, a penalty component is introduced into the cost function, and this component decreases as the actual output gets closer to the prediction. From an optimization perspective, since the operating cost of renewable energy generating units is low, they should be operated at their capacity upper limit as much as possible. However, when the penalty term is exponentially related to the square of the power generation deficit, too much renewable energy will significantly increase the cost and cause the operating state to exceed the limit.

[0063] The cost function and local operation constraints of the fuel power generation equipment can be modeled as:

[0064]

[0065] where, is the power output of the i-th fuel power generation equipment; a i , b i , c i , d i and are cost coefficients; and are the lower and upper limits of the power output of the i-th fuel power generation equipment respectively; is the ramp rate of the fuel power generation equipment.

[0066] The relationship between the cost of the natural gas station and the natural gas production is shown in Equation (13). Since P2G uses renewable energy equipment to produce gas, its gas production will be limited, and its cost function is shown in Equation (14), while G2P is similar, and its cost function is shown in Equation (15):

[0067]

[0068] Among them, is the power output of the i-th renewable energy equipment; a i 、b i 、c i 、d i and are cost coefficients; and are the lower and upper limits of the power output of the i-th renewable energy equipment respectively; is the ramp rate of the renewable energy equipment.

[0069] It can be seen from the cost function that electricity and gas are coupled with each other, and their production costs are also coupled with each other.

[0070] Furthermore, a renewable energy cost function and a multi-energy load utility function model considering conversion losses, calculation coefficients, and conversion intervals are established, which can improve the comprehensive energy utilization rate, as follows:[[]]END]]

[0071] In an electricity-gas coupled energy system, if energy can be accurately generated according to load demand, the production cost can be minimized and energy consumption can be reduced in a distributed manner. In the past energy distribution plan, the load consisted of loads that must operate and controllable loads, and its calculation was usually in a fixed constant manner. However, with the development of the energy Internet, various newly emerging controllable loads can be simulated through mathematical modeling to consume various energies for the purpose. Based on conversion losses, calculation factors, and conversion intervals, the present invention refines the composition of controllable loads, conducts mathematical modeling on some of them, and enables controllable loads to participate in equality constraints and economic dispatch optimization in the form of variables. The system load is divided into the following parts:

[0072]

[0073] Among them, L t represents the overall system load at time t, representing the sum of all capacity equipment loads and consumer loads. The equivalent must-run power and gas load are respectively expressed as These are fixed. The coupled load on the user side is defined as the integrated energy load, expressed as The remaining non-transferable controllable load is expressed as

[0074] The electrical integrated load is the gas load that can be replaced by electricity and is expressed as The electrical integrated load function can be modeled as:

[0075]

[0076] where represents the converted gas consumption; represents the converted electricity consumption; μ e represents the relative power loss. Since the conversion loss between unit power and gas is fixed, μ e is usually constant. η e2g represents the conversion efficiency of converting electricity into gas. In addition, it is assumed that there are maximum and minimum constraints on electricity and gas consumption, for example:

[0077]

[0078]

[0079] Once the consumer demand is determined, it is fixed. and will be used as variables in the optimization operation to achieve the optimal energy ratio according to the real-time price and the energy production cost. Based on the above discussion, the utility function of the electricity-gas integrated load is modeled as:

[0080]

[0081] where represents the gas and electricity utility coefficients at time t, and α > 0, β > 0, and γ > 0 are compensation coefficients. The first two terms of the above equation represent the direct cost of energy consumption, and the third term represents the compensation to encourage users to switch energy.

[0082] Furthermore, a neural dynamic optimization algorithm using a fully distributed method is proposed, which enables each energy node to reduce the shared information sent to its neighbor nodes, specifically as follows:

[0083] In this step, the aforementioned economic dispatch can be converted into a distributed optimization problem, and the objective function can be described as:

[0084]

[0085] Among them, f(x) is the objective function; g(x) is the inequality constraint determined by local operation constraints; Ω is the feasible operation region determined only by the inequality constraint g(x). Note that f(x) and g(x) are convex functions on Ω and locally Lipschitz continuous. The background problem contains multiple capacity devices and loads, which can be regarded as multiple agents. Each agent i can receive information from its neighboring agents, and it has local constraints Ω i , the state x i,t ∈R qi ∈Ω i , the objective function f i (x i,t ), for all agents i ∈ {1, 2,..., n} and time t.

[0086] To establish a neural network model for solving the multi-objective optimization problem, multiple recurrent neural networks are used here to replace all agents. Based on the distributed continuous-time subgradient algorithm, the mathematical expression of the fully distributed neural dynamic optimization algorithm is designed as follows:

[0087]

[0088] Among them, x = (x1, x2,..., x n ) T represents the state variables of all agents, where x i,t ∈R qi ; λ represents the local constraint variable; μ and η are shared variables related to the global constraints of the system; f(x) ∈ R represents the objective function; and where L is the Laplacian matrix, I p is the p-order identity matrix; Ω represents the feasible operation region; is the matrix composed of the equality constraint A i .

[0089] Each agent involved in this embodiment includes two variables of electricity and gas, so q = q1 = q2; r and p represent the numbers of inequality constraints and equality constraints, respectively, where:

[0090]

[0091] Among them, is the constant vector of the equality constraint; R np , R r , R np×np , R nr represent real number spaces with dimensions of np, r, np×np, and nr, respectively.

[0092] Compared with the existing economic dispatch algorithms based on neural dynamics, this algorithm does not require the Lipschitz condition, does not involve parameter adjustment, has fewer common variables and a simple structure. Since the common variables are only composed of global constraints, the state variables are independently calculated by the agents under local constraints without participating in information transmission. Even if the data in the network is monitored, the state variables of the system cannot be accurately obtained, which has a good privacy protection effect.

[0093] Because both f(x) and g(x) are convex functions, if there exists a μ * ∈R np and a λ * ∈R r such that (x * , λ * , μ * ) is an L-saddle point, then x * is the optimal solution of formula (22). Therefore, the following conclusion can be drawn:

[0094] L(x * , λ, μ) ≤ L(x * , λ * , μ * ) ≤ L(x, λ * , μ * ) (24);

[0095] In addition, to verify the convergence of the neural dynamics algorithm, we propose the following three auxiliary variables:

[0096]

[0097] And define the Lyapunov function:

[0098]

[0099] Since f(x), g(x), and ||η|| 2 are convex and differentiable, it is not difficult to obtain Because the neural dynamics algorithm proposed in this embodiment includes the forward projection algorithm and has the properties of λ + g(x) = [λ + g(x)] + -[-λ - g(x)] + and . Taking the derivative of the Lyapunov function gives:

[0100]

[0101]

[0102] It can be obtained that According to the KKT conditions and Lyapunov stability theory, the above calculations prove that the fully distributed neural dynamic optimization algorithm proposed in this embodiment can asymptotically converge to a unique optimal solution within a large range.

[0103] On the other hand, this embodiment provides a distributed collaborative optimization system for an electric-gas coupled energy system, including: a model construction module, configured to construct a working model of the electric-gas coupled energy system, and based on the conversion loss, calculation coefficient, and conversion interval of the electric-gas coupled energy system, construct a load utility model based on the working model; a data acquisition module, configured to collect in real time the energy supply data of the energy supply end and the load data of the load end in the target area; a collaborative control module, configured to input the energy supply data and the load data into the working model and the load utility model, solve using the fully distributed neural dynamic optimization algorithm, obtain the optimal solution, and complete distributed collaborative control.

[0104] And a distributed collaborative optimization device for an electric-gas coupled energy system is developed, including: a busbar trunking unit, an intelligent terminal, and a server. After collecting the energy supply data of the energy supply end and the load data of the load end in the target area through the busbar trunking unit, input them into the intelligent terminal to perform fully distributed neural dynamic optimization solution in combination with the working model and the load utility model, obtain the optimal solution and generate a control signal, and then publish the control signal through the busbar trunking unit. At the same time, the energy supply data, the load data, and the control process are input into the server for storage.

[0105] Specifically, the distributed collaborative optimization device for the electric-gas coupled energy system developed in this embodiment integrates two types of energy, natural gas and electricity. The natural gas station, the urban power grid, and the photovoltaic power supply platform are used as the energy supply ends respectively, and common temperature control loads, commercial loads, learning tools, computing centers, lighting equipment, etc. are used as electric loads, and gas water heaters and gas stoves are used as gas loads. Four independently developed intelligent terminals are set in the platform, which are respectively used for the monitoring and control of the main power supply, photovoltaic power generation, energy storage equipment, and load side of the microgrid, and can complete functions such as operation data monitoring of the coupled energy, energy dispatching management and control, and load energy consumption optimization.

[0106] Among them, the intelligent terminal can satisfy the mutual communication of multiple intelligent terminals, can receive energy equipment and sensor data of multiple communication protocol types, supports interaction with the upper-level platform, can achieve distributed collaborative control and collaborative optimization, and achieve the expected task goals. Based on the algorithm theory support, the fully distributed neural dynamic optimization algorithm and the adaptive intelligent control algorithm are integrated and deployed to realize communication and collaborative operation between terminals.

[0107] The intelligent terminal integrates a low-level communication module with various serial communication functions such as RS485 / 232 interfaces, DB9 interfaces, and GPIO ports to collect data from low-level sensors. Device files are used to define relevant information about the device, the data format stored in the database, and operation commands for the device. Multi-terminal communication and interaction are realized based on TCP and UDP protocols. Based on the constructed electrical-gas coupling system, the intelligent terminal can achieve real-time data collection from the power grid, gas network, and low-level application devices, and realize interactive collaboration between multiple intelligent terminals under time-varying topologies, coordinated control between the intelligent terminal and the energy coupling network, and coordinated optimization between multiple information-energy coupling network nodes, thus realizing the real-time, stability, and economy of distributed collaborative optimization.

[0108] The overall control process of the intelligent terminal first initializes parameters after power-on and startup, and calls the distributed consensus power sharing collaborative control algorithm based on multi-intelligent terminal event-triggered communication to effectively improve the synchronization and stability and collaborative optimization ability of information-energy nodes, and overcome difficult problems such as high failure rate and poor effectiveness in multi-terminal control of the information-energy coupling network. Then, the terminal interaction system is run, and TCP and UDP are used to communicate and connect to terminal devices. Finally, by detecting online and offline nodes, device data is collected, stored in the local and edge servers, and periodically cleared. The real-time status information, energy node information, load node information are displayed through the interaction system, and it has a data query function and can adjust the status of the subnet in real time.

[0109] The distributed collaborative optimization mode of the entire electrical-gas coupling system under variable topology states is carried out according to the horizontal and vertical architectures. The intelligent terminal is connected to the cloud platform network on the upper layer and the low-level device ports on the lower layer, forming a cloud-edge-terminal collaborative control structure. To reduce the computing and storage burden of the terminal and accelerate the operation speed of the terminal algorithm, the present invention supports the interaction between the terminal and the cloud platform, thereby realizing distributed collaborative control and collaborative optimization.

[0110] The traditional cloud-edge collaboration model mainly uses the cloud as a big data analysis platform and the edge as the front-end link for data collection and preprocessing, and the functional division of each link is relatively clear. Therefore, the interaction of cloud-edge information flows mainly presents a vertical linear relationship, and is subject to access capacity and communication performance. Edge nodes can only be set at access points with good communication conditions. In comparison, the main architectural feature of the cloud-edge collaboration in the present invention is that its physical "cloud-edge-end" layer and the logical "perception-computing-decision-making" layer are no longer a linear relationship corresponding to a one-dimensional linear relationship, but a criss-crossing two-dimensional cross-collaborative relationship. On the one hand, using existing communication technology, the intelligent terminal connects the underlying device to the platform, and the terminal parses and processes the data protocol of the device, and then encapsulates it into an MQTT protocol frame, which is uploaded to the cloud platform through a remote wireless network. At the same time, in order to reduce the pressure of cloud data processing, the intelligent terminal does not upload all data to the cloud platform, but extracts valuable data after data cleaning for uploading. On the other hand, after the end-side data is processed, the processing results and related necessary data are uploaded to the cloud platform. The cloud platform summarizes, processes, analyzes and predicts the data in the region through the corresponding applications, and dispatches the global users in combination with the preset dispatch plan. The analysis results and dispatch arrangements are delegated to each terminal device, and the terminal device adjusts its own dispatch plan accordingly. The cloud platform summarizes and processes the power consumption and operating status data of the entire distribution network, analyzes it using intelligent algorithms, makes global load forecasts, and optimizes the decision-making plans of each region, thereby achieving safe, reliable, economical and efficient operation of the entire distribution network.

[0111] In order to evaluate the economic dispatch performance of the fully distributed neural dynamic optimization algorithm in the electricity-gas coupled energy system and prove its effectiveness, Figure 1 The experiment was conducted on the 41 nodes shown. Each load node can send its load information to the nearest distributed capacity node. The electric-gas coupled energy system consists of four power stations, a photovoltaic power station, three natural gas sources, a P2G and a G2P. Each capacity node exchanges information with its neighbors and performs local calculations through a distributed communication network.

[0112] Nodes 1-10 correspond to: power plant DG14, photovoltaic power generation PV1, natural gas sources GS1-3, P2G1 and G2P1. The energy units are unified here, that is, the power (gas) unit is 1p.u.=1MW, the gas unit is 1p.u.=79SCM / h, and the price unit is 1$ / MWh. The convergence and effectiveness of the proposed algorithm are proved. The total energy load of the 41-node electric-gas coupled energy network is initialized to 25 (pu) of electric load and 14 (pu) of gas load. The local target variable x is obtained by distributed computing. i. To reflect the randomization of the initial values of the neural dynamic algorithm proposed in this embodiment, the initial values of 10 capacity units are randomly set to (1.1, 2.3, 3.2, 4.7, 5.5, 0.6, 3.2, 4.1, 3.2, 1.7) (p.u.). By arbitrarily selecting the initial values, the trajectories of energy supply-demand imbalance, common variables, and outputs are obtained, as Figure 2 and Figure 3 shown.

[0113] The optimal scheduling curve of the electric energy output is as shown in Figure 2 (a). It can be seen that all distributed generation units reach a steady state under any initial value. At the same time, Figure 2 (b) shows the optimal scheduling curve of the natural gas output, which also finally reaches a steady state. At this time, when the convergence time reaches 60 seconds, it can be considered that the capacity cost of the electricity-gas coupling system reaches economic optimization.

[0114] From Figure 3 (a), it can be seen that the deviation between the total energy output and the total load demand gradually converges to 0. At this time, it can be considered that the electricity-gas coupling energy system has reached supply-demand balance. From Figure 3 (b), it can be seen that the auxiliary variables of all nodes finally converge to 0, which proves the consistency of the distributed neural dynamic optimization algorithm.

[0115] At the same time, in terms of device development, in order to enable the normal input and output of energy in the distributed collaborative optimization device of the electricity-gas coupling system, a busbar cabinet is introduced, which can converge and supervise the energy flows of the commercial power, three photovoltaic inverters, and the battery bank, sample and display the flow rate of the natural gas pipeline, and at the same time, the control signals are sorted and task dispatched by this busbar cabinet. The platform embeds the neural dynamic distributed regulation algorithm proposed in this chapter in 4 intelligent terminals, which can achieve distributed optimization with neighbors; at the same time, it is connected to the busbar cabinet and the server to achieve task dispatching and storage of the calculation results. The energy flow and information flow are as shown in Figure 4 . All energy first passes through the busbar cabinet and then flows to each load; the collected signals are also output from the busbar cabinet to the intelligent terminal and flow to the server for storage; the control signals are calculated from the intelligent terminal and sent to the actuators of each energy branch in the busbar cabinet respectively.

[0116] The neural dynamic distributed collaborative regulation algorithm is embedded in the intelligent terminal to monitor and adjust the energy consumption of the integrated energy microgrid experimental platform in real time to adjust the energy input situation. Figure 5Shown are the operating conditions of four intelligent terminals and their liquid crystal display interfaces. The electrical load includes common electrical appliances such as air conditioners, computers, and electric water heaters, and the gas load is a gas stove. In the initial stage, the stable operating state of the platform is an electrical load of 42.5 kW and a gas load of 2.7 kW. Since the electric water heater can be replaced by a gas water heater, this part of the load can be regarded as an electric-gas universal load. Therefore, after operating for 30 minutes, the gas water heater is used to replace the electric water heater. From Figure 5 As can be seen from the load sampling interface in the upper right corner, the electrical load drops to 36.7 kW and the gas load increases to 4.8 kW. At the same time, the four intelligent terminals enable a distributed control algorithm based on neural dynamics to optimize the scheduling of the energy input end, and achieve supply-demand balance and energy distribution after operating for 32 minutes.

[0117] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A distributed collaborative optimization method for an electric-gas coupled energy system, characterized in that: include: Construct a working model of an electricity-gas coupled energy system; Constructing a load utility model based on the working model according to the conversion loss, calculation coefficient and conversion interval of the electric-gas coupled energy system; The electricity load and gas load of the target area are collected in real time, input into the working model and load utility model, and solved using a fully distributed neural dynamic optimization algorithm to obtain the optimal solution, thereby completing the distributed collaborative control of the electricity-gas coupled energy system.

2. The distributed collaborative optimization method for the electric-gas coupled energy system according to claim 1 is characterized in that: The working model of the electricity-gas coupling energy system includes: The working model is constructed by combining the balance relationship between electricity, gas and load of the equipment in the electric-gas coupled energy system, as well as the constraints of the equipment, the cost of fuel power generation equipment, and the cost of renewable energy equipment.

3. The distributed collaborative optimization method for the electric-gas coupled energy system according to claim 2 is characterized in that: The balance relationship between electricity, gas and load of the equipment in the electric-gas coupled energy system is: Among them, p represents the input / output power of electricity and gas, g represents the input / output power of gas; L is the power required by the load; i represents the i-th device, t represents the time t, rg, fg, g2p, psd, gp, gsd, and p2g represent distributed renewable power generation equipment, distributed fuel power generation equipment, distributed gas-to-electricity equipment, distributed power storage equipment, distributed gas production equipment, distributed gas storage equipment, and distributed power-to-gas equipment, respectively.

4. The distributed collaborative optimization method for the electric-gas coupled energy system according to claim 3 is characterized in that: The constraints of the device are: Where min and max represent the lower and upper bounds, respectively; φ represents electricity and gas; Indicates the slope limit of the device; Indicates the energy storage capacity of the energy storage device; Indicates the charge and discharge efficiency; Indicates the state of the energy storage device, "1" indicates charging, and "-1" indicates discharging; Indicates the charge and discharge capacity per unit time; Indicates the maximum charging capacity; Indicates the maximum discharge capacity; ψ p2g and ψ g2p Represent the ratios of electricity to gas and gas to electricity, respectively.

5. The distributed collaborative optimization method for the electric-gas coupled energy system according to claim 4 is characterized in that: The cost of the fuel power generation equipment is: in, is the power output of the i-th fuel power generation equipment; a i 、b i 、c i ,d i and is the cost coefficient; and are the lower and upper limits of the power output of the i-th fuel power generation equipment; is the ramp rate of the fuel-fired power generation equipment.

6. The distributed collaborative optimization method for the electric-gas coupled energy system according to claim 5 is characterized in that: The cost of the renewable energy equipment is: in, is the power output of the i-th renewable energy device; a i 、b i 、c i ,d i and is the cost coefficient; and are the lower and upper limits of the power output of the i-th renewable energy device, respectively; is the ramp rate of renewable energy equipment.

7. The distributed collaborative optimization method for an electric-gas coupled energy system according to claim 1, characterized in that: The load utility model is: in, Indicates the comprehensive electrical load. Indicates the gas consumption after conversion, Indicates the amount of electric energy consumed after conversion, represents the gas and electricity utility coefficients at time t, α>0, β>0 and γ>0 are compensation coefficients, the first two terms of the equation represent the direct cost of energy consumption, and the third term represents the compensation to encourage users to switch energy.

8. The distributed collaborative optimization method for an electricity-gas coupled energy system according to claim 1, characterized in that: The fully distributed neural dynamic optimization algorithm is: Where x=(x1,x2,...,x n ) T Represents the state variables of all agents, where x i,t ∈R qi ;λ represents the local constraint variable;μ and η are shared variables related to the global constraints of the system; are the vector forms of variables x, μ, λ and η respectively; ▽ is the gradient operator; f(x)∈R represents the objective function; g(x) represents the inequality constraint; is the vector formed by g(x); Where L is the Laplace matrix, I p is the p-order identity matrix; Ω represents the feasible operation area; A is constrained by the equation i The matrix formed; is the constant vector of equality constraints; R np , R r , R np×np , R nr They represent real number spaces of dimensions np, r, np×np, and nr respectively.

9. A distributed collaborative optimization system for an electric-gas coupled energy system, implemented based on the distributed collaborative optimization method for an electric-gas coupled energy system according to any one of claims 1 to 8, characterized in that: include: A model building module, used to build a working model of the electric-gas coupled energy system, and to build a load utility model based on the working model according to the conversion loss, calculation coefficient and conversion interval of the electric-gas coupled energy system; The data acquisition module is used to collect the energy supply data of the energy supply end and the load data of the load end in the target area in real time; The collaborative control module is used to input the energy supply data and load data into the working model and the load utility model, and adopt a fully distributed neural dynamic optimization algorithm to solve, obtain the optimal solution, and complete distributed collaborative control.

10. A distributed collaborative optimization device for an electric-gas coupled energy system, implemented based on the distributed collaborative optimization method for an electric-gas coupled energy system according to any one of claims 1 to 8, characterized in that: include: A combiner box, an intelligent terminal and a server collect energy supply data of the energy supply end of the target area and load data of the load end through the combiner box, and then input the data into the intelligent terminal to perform fully distributed neural dynamic optimization solution in combination with the working model and the load utility model, obtain the optimal solution and generate a control signal, which is then issued through the combiner box. At the same time, the energy supply data, load data and the control process are input into the server for storage.

Images